UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Historical ecosystem modelling of the upper Gulf of California (Mexico) : following 50 years of change Lozano, Hector 2006

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_2006-200339.pdf [ 22.87MB ]
Metadata
JSON: 831-1.0074893.json
JSON-LD: 831-1.0074893-ld.json
RDF/XML (Pretty): 831-1.0074893-rdf.xml
RDF/JSON: 831-1.0074893-rdf.json
Turtle: 831-1.0074893-turtle.txt
N-Triples: 831-1.0074893-rdf-ntriples.txt
Original Record: 831-1.0074893-source.json
Full Text
831-1.0074893-fulltext.txt
Citation
831-1.0074893.ris

Full Text

H I S T O R I C A L E C O S Y S T E M M O D E L L I N G O F T H E U P P E R G U L F O F C A L I F O R N I A ( M E X I C O ) : F O L L O W I N G 50 Y E A R S O F C H A N G E . B y H E C T O R L O Z A N O B . S c , National University of Mexico, Mexico 1993 M . S., C I C E S E , Ensenada, Baja California, Mexico 1998 A THESIS S U B M I T T E D IN P A R T I A L F U L F I L M E N T O F T H E R E Q U I R E M E N T S F O R T H E D E G R E E O F D O C T O R O F P H I L O S O P H Y in T H E F A C U L T Y O F G R A D U A T E STUDIES (Zoology) T H E U N I V E R S I T Y O F BRITISH C O L U M B I A September 2006 © Hector Lozano, 2006 Abstract Ecological conditions in the upper G u l f of California (Northwest Mexico) have deteriorated significantly over the past seventy years through the removal of nutrients and freshwater from the Colorado River. Acting at the same time, uncontrolled exploitation has brought at least three fisheries to collapse and several endemic species, such as giant G u l f croaker (Totoaba macdonaldi) and the vaquita porpoise (Phocoena sinus), near to extinction. This motivated a study to quantify the ecological impacts attributable to diversion of the Colorado River on the dynamics and interactions o f the upper G u l f o f California (UGC) . In the absence of baseline studies for the pre-diversion period, this work has combined scientific information, newly-gathered local knowledge from fishers ( L F K ) , and information from industrial and artisanal fisheries to reconstruct for the first time models of the upper G u l f o f California for three key periods: 1950, 1980 and 2000. Reconstruction of these past ecosystem states are based on mass-balanced models (130 species considered in 50 functional groups) generated using the Ecopath and Ecosim approach and software. Dynamic simulations, with ecosystem model parameters fitted to all known biomass and fisheries data (historical time-series o f catch-at-age data and biomass surveys since 1970s reported by the Institute o f National Fisheries, Mexico), were used to track and emulate changes caused by water diversion and fishing within the limits set by fluctuations in climate such as the E l Nino events o f 1983, 1993 and 1997. The U G C is an ecosystem to be highly-dependent on detritus-benthic components. Biomasses appear to be largely controlled by lower trophic levels. A highly diverse artisanal fishery targeting a wide range of species, impacts all levels of the food web. These effects need to be taken into account to be considered in the management plans in the region. During the last five decades, change in trophic level indicated important impacts o f fishing, resulting in an average decline in the trophic level o f 0.02 trophic level/decade. Also, there was a notable loss in the abundance o f detritivores during the last 50 years, with a reduction o f 64% in the total estimated biomass of all the groups located in trophic level 2 and 2.5. This loss is affecting the incorporation of organic matter to higher trophic levels o f the food web, a critical function in any ecosystem. Overall, numerous changes in the energy flows and production rates were estimated during the research, indicating a large reduction in the energy and dynamics o f the upper Gulf. These changes affect the structure and function of this ecosystem, limiting its ability to recover from future natural or human forces. Simulations suggested that river flows o f only 1% of the undepleted level may produce increases of around 10% in the total biomass of the U G C , reflecting the enormous role o f the Colorado River in the productivity in the region. This is particularly important considering that this region provides 10% of the annual Mexican landings (INEGI, 2001), and provides a preliminary accounting o f the ecological and economic losses for Mexico during the last five decades since American river dams were built. Simulations o f the effect of reduced river flows and environmental (El Nino) influences on the U G C ecosystem revealed the superficiality o f our knowledge o f the quantitative process simulated in the models and reconstruction of past states o f natural aquatic ecosystems like the upper G u l f o f California. It is evident that any attempt to restore the U G C must be initiated with a u parallel study of the climate change in its influence in the region. This research suggested that human intervention can trigger rapid and significant changes in the structure and function o f small marine ecosystems such as the upper G u l f o f California. In the future, human interventions in river deltas, such as damming rivers and water abstraction, must be considered as one o f the major forces affecting the steady state evolution of these ecosystems, impacting their natural processes, productivity, biodiversity and the human communities that depend on them for food, work and culture. 111 TABLE OF CONTENTS Abstract i i Table of contents i v List of tables v i i i List o f figures ix Acknowledgements x i i CHAPTER I. Overview and Summary 1 1.1. Need and motivation for modelling the U G C 1 1.2. Research objectives 3 1.3. Biodiversity impacts of large dams 4 1.4. Modell ing ecosystems of the present and past: "Back to the Future" policy approach 5 1.5. Considering the effect o f climate change 8 1.6. Previous research 9 1.7. Thesis structure 10 CHAPTER II - A trophic model of the upper Gulf of California 12 2.1. Physical and biological characteristics 12 2.1.1. Historical variability of fluvial discharge 13 2.2. The fisheries o f the upper G u l f o f California 16 2.2.1. Fisheries Crises 18 2.2.2. First steps to address illegal and Unreported catches... 21 2.2.2.1. Methods 22 2.2.2.2. Results 24 2.2.2.3. Discussion 27 2.3. A n ecosystem model of trophic structure and fisheries interaction in the U G C for 1995-2000 35 2.3.1. Introduction 35 2.3.2. Methods 36 2.3.2.1. Constructing the trophic model 37 2.3.2.2. Aggregation of functional groups 38 2.3.2.3. Sources o f the basic input parameters 39 2.3.2.4. Sources of catch data 47 2.3.2.5. Balancing the model 48 2.3.2.6. Uncertainty and model validation 48 2.3.3. Results 49 2.3.3.1. Mortality: Fishing versus predation 50 2.3.3.2. Networkflow analysis 51 2.3.4. Discussion 53 2.3.4.1. Sensitivity analysis 55 2.3.4.2. Uncertainties in the data 55 2.3.4.3. Concluding remarks 56 iv 2.4. Tuning and simulating the upper G u l f o f California 60 2.4.1. Tuning the 2000 U G C model 61 2.4.2. Vulnerabilities 62 2.4.3. Ecosystem dynamics 63 2.4.4. The trophic impact o f sharks 64 2.4.5. The trophic impact of detritus 65 2.5. Summary 67 C H A P T E R III. Past ecosystem models for the U G C / C R D 68 3.1. Historical changes in the U G C / C R D : lessons from the past 68 3.1.1. The delta o f yesterday 68 3.1.2. The delta o f today 70 3.1.3. A changing fauna 73 3.2. Oral inputs to the U G C ecosystem reconstructions 77 3.2.1. Shirting baselines syndrome 79 3.2.3. Representativeness and Validity 84 3.2.4. Creating time-series of relative abundances 84 3.2.5. Results 85 3.2.5.1. Shifting o f environmental baselines U G C 87 3.2.5.2. Estimating past abundances based on Local Fishers Knowledge ( L F K ) 91 3.2.5.3. Agreement of L F K anecdotes with INR records.. 93 3.2.5.4. Opinion o f the fishers about the freshwater diversion 95 3.2.5.5. Opinions o f the fishers about the future o f their fisheries 96 3.2.3.6. Final remarks 97 3.3. Ecosystem models of the U G C in 1950 and 1980 . . . .101 3.3.1. Reconstructing the past 101 3.3.2. Modifying P /B and Q/B ratios 101 3.3.3. Diet composition 105 3.3.4. Catch Data 106 3.3.5. Biomass changes 110 3.4. Balancing the 1950 and 1980 models 123 3.4,1. Uncertainty in the 1980 and 1950 models 123 v CHAPTER I V . Connecting the past with the present: linking the 1950,1980 and 2000 models 129 4.1. Connection and fitting between the 1950, 1980 and 2000 models 129 4.1.1. Points to be considered in the reconstruction of the past 129 4.1.2. Tuning the 1950s and 1980s U G C Ecosystem Models 130 4.1.3. Improving the fitted model: climate influence 134 4.1.4. Vulnerabilities resulting from the tuning 135 4.2. Exploring ecosystem changes through time: Structure and function o f the U G C ecosystem in the last 50 years 137 4.2.1. Tracking food web changes: by trophic levels 138 4.2.2. Tracking food web changes: the loss of biomasses 141 4.2.3. Tracking ecosystem changes through time: Fishing mortalities 147 4.2.4. Tracking ecosystem changes through time: Network analysis 151 CHAPTER V . - Examining fishing and climate effects in the UGC 162 5.1. Exploring ecosystem effects of changes in fishing efforts under different policy objectives 162 5.2. Exploring changes in biodiversity under different fishing strategies 170 5.3. Evaluating tradeoffs 176 5.4. Exploring the climate influence on the food web 180 5.5. Exploring the effect of damming the Colorado River on the U G C ecosystem 186 5.6. Conclusions 190 CHAPTER V I . Summary and concluding comments 192 6.1. Recommendations 205 6.2. Conclusions 212 References 215 vi Appendix 1. Incentives to misreport Mexican fish landings from 1970-2000 236 Appendix 2. Diet composition for the 2000 model 242 Appendix 3. Diet composition for the 1980 model 245 Appendix 4. Diet composition for the 1950 model 248 Appendix 5. Questionnaire used in interviews with U G C fishers 251 Appendix 6. Market prices for the commercial species 257 Appendix 7. Fishing cost for the main fleets 258 Appendix 8. List o f the 70 scenarios for the search of the optimum fishing routine 259 Appendix 9. Summary o f sources o f information to built the 2000 model 261 Appendix 10. Summary o f sources o f information to built the 1980 model 263 Appendix 1 1 . Summary o f sources of information to built the 19500 model 265 vi i List of Tables. Table 1 .Summary o f qualitative estimations of misreporting catches 29 Table 2. Incentives for Mexican vessels in the G u l f of California to misreport catch 31 Table 3. Biomass estimation from 1992-1995 surveys 44 Table 4. Relative abundance o f cetaceans in the upper G u l f o f California 45 Table 5. Basic summary statistics from the network flow analysis o f the 2000 model 53 Table 6. Basic parameters of the U G C model 59 Table 7. Anecdotes and information from 49 fisher interviews 99 Table 8. Diets of sharks in the U G C estimated in the 1950, 1980 and 2000 models 106 Table 9. Biomass estimated by surveys in the U G C in 1967-1968 119 Table 10. Abundance of zooplankton in 1957 in the U G C 121 Table 11. Parameters of the 1980s model 126 Table 12. Parameters of the 1950s model 127 Table 13. F low indices of the U G C ecosystem in 1950, 1980 and 2000 161 v i i i List of figures Figure 1. G u l f o f California 15 Figure 2. Annual discharge of the Colorado River 16 Figure 3. Historical time series of the totoaba {Totoaba macdonaldi) landings in the G u l f of California. 19 Figure 4. Annual landing of shrimp in San Felipe (Baja California) 20 Figure 5. Temporal changes in the mean trophic levels in the G o C 21 Figure 6. Time series of total extraction of sharks and totoaba 25 Figure 7. Estimated total extraction of shrimp in the region of San Felipe, Northern Gul f o f California 26 Figure 8. Proportion of I U U fishing estimated in the U G C in 2000 26 Figure 9. Biomasses sampled of shrimps, "chanos" and crabs (right) in 1995 by the INP (Institute of National Fisheries) 43 Figure 10. Trophic aggregation of the 50 groups 50 Figure 11. Comparison o f fishing and predation mortalities 51 Figure. 12. M i x e d trophic impacts o f the 2000 model 57 Figure 13. M i x e d trophic Impacts of the eight fishery fleets 58 Figure 14. Tuning of the 2000 model 63 Figure 15. Simulated removal of sharks in the 2000 model 65 Figure 16. Simulated removal of detritus from the 2000 model : 66 Figure 17. Painting from 1845 of a steamer crossing the Colorado Delta 71 Figure 18. Colorado River discharge from 1910-2000 71 Figure 19. Sea surface salinity in the U G C 73 Figure 20. Relationship between shrimp landings and Colorado River discharge 75 Figure 21. Historical landings of the Gu l f corvine 75 Figure 22. Photo of Puerto Penasco, 1934 76 Figure 23. Totoabas of nearly 100 kg caught in San Felipe in 1940s 77 Figure 24. Shifting baselines in the U G C ; 80 Figure 25. Some o f the 49 fishers interviewed in the U G C in 2003 83 Figure 26. Age distribution of the 49 fishers interviewed 83 Figure 27. Depletion of fishing sites reported by L F K interviews 86 Figure 28. Map o f the fishing sites depleted in the U G C in the last 60 years 89 Figure 29. Change in the size of totoaba caught according to L F K analysis 90 Figure 30. Relative abundance estimated from L F K interviews 92 Figure 31. Concordance between relative abundances estimated by the L E K and biomasses estimated by surveys. 94 Figure 32. Summary of the Spearman correlation analysis 95 ix Figure 33. Opinions of 49 fishers o f the upper G u l f of California. 96 Figure 34. Future of the fisheries in the upper G u l f of California according to the perspective of 49 fishers 97 Figure 35. Change in P /B from 1950 and 1980 compared to 2000 103 Figure 36. Changes in Q/B from 1950-2000 104 Figure 37. Change in landings from 1980 compared to 2000 109 Figure 38. V P A based on length-frequency of totoaba 115 Figure 39. Changes in proportion o f the biomass from 1950, 1980, 2000 116 Figure 40. Mean trophic level o f the benthic fish in 1952 in the U G C 118 Figure 41. Absolute biomass of zooplankton and phytoplankton estimated by surveys in 1950 122 Figure 42. Pedigree indices for the 1950, 1980 and 2000 models 125 Figure 43. Time-series of annual fishing mortality from 1950 to 1978 131 Figure 44. Estimation o f annual fishing mortality of blue shrimp 132 Figure 45. Fitting to time series plot for the shrimps and totoaba 133 Figure 46. Improving the fitted 2000 model with the Colorado River. discharge from 1950-2000 as a forcing function 135 Figure 47. Vulnerabilities obtained during the tuning process from the 1950s to the 2000 137 Figure 48. Trophic levels of the 1950s, 1980s and 2000 models 140 Figure 49. Reduction of the mean trophic level o f the catch froml950-1980 140 Figure 50. Total biomass by trophic level o f the past and present food web models of the U G C 141 Figure 51. Changes of the biomass of invertebrates from the 1950, 1980 and 2000 models 143 Figure 52. Changes in the biomass of fish groups from the 1950, 1980 and 2000 models 143 Figure 53. Marine mammals biomass change from 1950 to 1980 to 2000 145 Figure 54. Decline of the ratio between benthic and pelagic biomass from the 1950,1980 and 2000 models 146 Figure 55. Changes in the fishing mortality from the 1950, 1980 and 2000 models 148 Figure 56. Reduction of predation o f sea lions by sharks 1950-2000 150 Figure 57. Reduction in the predation mortality by sharks 151 Figure 58. Results from the M i x e d Trophic Impact routine 153 Figure 59. Tropic impacts of sharks 1950-2000 154 Figure 60. Trophic impacts of totoaba 1950-2000 155 Figure 61. Trophic impacts o f vaquita 1950-2000 157 x Figure 62. Decline o f the overhead 1950-2000 159 Figure 63. Mean change o f the total catch after 50 years 166 Figure 64. Mean change o f the profit forecast after 50 years 168 Figure 65. Relationship between the restoration goals for vaquita , with the total biomass of the upper G u l f o f California 169 Figure 66. Profits predicted on the recovery of vaquita 170 Figure 67. Measure of biodiversity with Q-90 index 171 Figure 68. Relationship between the biodiversity measured by Q-90 index and the final value o f the catch 174 Figure 69. Cluster dendrogram for the monetary value 175 Figure 70. Depletion of the stocks 1950-2000 178 Figure 71. Tradeoffs between economical and diversity goals 1950-2000 179 Figure 72. Monthly anomalies of SST 1992-1997 181 Figure 73. Comparison o f sampled and calculated phytoplankton biomasses 183 Figure 74. Changes in relative biomasses 1974-2002 associated with the Colorado River floods 185 Figure 75. Biomass profiles estimated under undepleted flows of the river 188 Figure 76. Changes in the biomass of the 1950 model after 50 years o f simulation with different levels of the Colorado River discharges 190 x i Acknowledgements I am very grateful and extend my sincere thanks to my supervisor Dr Tony Pitcher for his mentorship, co-operation and wealth of knowledge that was supplied during our weekly meetings. Also , Dr Pitcher gave me several opportunities to present my research on the international scene. M y appreciation to the members o f my committee: Dr Daniel Pauly for encouragement to excellence and guidance during my path; to Dr Col in Brauner for his enthusiasm and unfaltering support. I also thank to Niggel Haggan for his valuable comments throughout my research, field trip and writing process. The author acknowledges the invaluable input and help of the following individuals and organizations during this project: Dr Tony Pitcher, D r Migue l Cisneros (Institute of National Fisheries, Mexico) and Dr Francisco Arreguin ( C I C I M A R , Mexico) for their support during the Workshop organized in L a Paz, Baja California, Mexico (2002); Armando Rosas and Julian Gallardo (CRLP-Ensenada) for shrimp information; Miguel Lavin (CICESE) for oceanographic data; Bertha Lavaniegos (CICESE) for zooplankton data; Jose Campoy and Martha Roman (Biosphere Reserve, upper G u l f of California, Mexico) for totoaba data; Peggy Turk and Richard Cudney ( C E D O ) for social, economic and fisheries data; Oscar Pedrin for biomass surveys; Ivonne Ortiz (University o f Washington) for vaquita information; Staff o f U N I S O N (University o f Sonora, Mexico) for fisheries data and equipment; Scripps Institute o f Oceanography (La Jolla, C A ) for 1957 surveys data and finally, I would like to express my gratitude to the fishers o f the upper G u l f o f California for their hospitality, friendliness and sharing their memories. This project would not have been possible without the financial support provided by the scholarships granted by C O N A C Y T (Mexico) and the University off British Columbia Graduate Fellowship. Also , I thank to the University o f British Columbia for the "Cec i l and Kathleen Morrow" scholarship provided in 2003 to realize the field study in Baja California, Mexico. Finally, I would like to thank Patty Rojo, my wonderful wife for supporting and encouraging me to pursue this degree. To my parents, brother (Eugenio) and sister (Pily) and their deep belief in me. Gracias de todo Corazon. x i i Chapter I. Overview and Summary. 1.1. Need and motivation for modelling the upper Gulf of California. Despite the fact that the G u l f of California is the world's youngest sea, created when the San Andres faults split the Northwest o f Mex ico and allowed the Pacific to flow in, it is recognized as one o f the most productive marine ecosystems in the world. Furthermore, its richness produces 40% of Mexico ' s total landings and it provides jobs to nearly 30,000 fishermen (ESfEGI, 2001). A critical and unique area o f this large marine ecosystem is its northernmost region, the upper G u l f of California ( U G C ) , including one of the world's great desert deltas, the Colorado River Delta (CRD). This delta provides 10% of the national landings and has supported important fisheries since 1920s. The historical richness of this area has been explained mainly by the nutrients and sediments provided by the Colorado River and by extreme tidal and upwelling processes (Carraquiry and Sanchez, 1999). Unfortunately, the C R D / U G C exemplifies human efforts to divert and control the major rivers in the world. A s in many other river deltas in the world, such as the Amazon (Barthem et al., 1995) or the Mississippi (Arthington and Welcomme, 1995), the C R D has suffered a combination of river impoundment and diversions since 1905, mainly because o f a series of U .S . dams and irrigation projects (1935: the completion of the huge Hoover dam; 1963: completion of Glen Canyon dam) to divert approximately 18 bill ion m 3/year o f freshwater and up to 450 mil l ion o f metric tonnes/year of sediment that once were delivered into the upper gulf (Van Andel , 1964; Carbajal et al., 1997; Carraquiry and Sanchez, 1999). This anthropogenic intervention has had a devastating effect not only on the delta wetlands, but also on the estuarine and marine waters of this ecosystem (Briggs and Cornelius, 1998; Brusca et al 2001; Glenn et al., 2001) with an evident alteration in the hydrography of the C R D / U G C system. In turn, this has produced a permanent increase of salinity from 22-35 before the dams were built and also during 1 controlled releases (Townsend, 1901; Lavin et al, 1998; Lavin, 1999) to current salinities that are typically in the range o f 35-45 (Alvarez-Borrego, 1975; Lavin, 1999). The increase in salinity in the delta and upper G u l f has fundamentally changed spawning and nursery ground conditions that are critical to hundred of species, including invertebrates, fish, reptiles, marine mammals and birds that use the delta and its marine zone (Sykes, 1937). Several examples illustrate the importance of the freshwater, sediments and nutrients delivered by the Colorado River. The endemic population o f bivalves (mainly Mulina colorandensis) has shown a reduction o f more than 90% from pre-dam densities (Kowaleski et al., 2000; Rodriguez et al., 2001; Flessa et al., 2001). Blue shrimp catches (Litopenaeus stylirostris) and their post-larva abundances are positively correlated with the previous year's discharge (Galindo-Bect and Glenn, 2000; Aragon-Noriega and Calderon-Aguilera, 2000; Calderon-Aguilera et al., 2002). Finally, there is the remarkable case of the gulf corvina (Cynoscion othonopterus), endemic to the G u l f o f California, but not been seen in the upper gulf for 40 years, but which returned in large numbers after the floods of E l Nino 1992-93 (Zengel et al, 1995; Cudney-Bueno and Turk, 1998). Also , important reductions of refuge habitats for more than 100,000 migratory birds visiting the area each winter have been documented (Boyer, 1996; Brusca et al, 2001). These examples illustrate that the Colorado River once played a major role in not only supplying nutrients, but also providing low-salinity nursery areas for larval shrimps and fish fry where euryhaline predators could not penetrate. Uncontrolled fisheries have brought several endemic species near to extinction. The giant G u l f Croaker (Totoaba macdonaldi), has the unenviable distinction of being the first marine fish listed under CITES and E S A . The vaquita porpoise (Phocoena sinus) is also endangered (Cisneros-Mata et al, 1995; Roman-Rodriguez and Hamman. 1997; Jaramillo-Legorreta, 1999; D'Agrosa et al, 2000), but the role of increased salinity in their key habitats is unknown (Glenn et al, 2001). Because ecological conditions in the C R D / U G C have deteriorated significantly over the past 70 years, resulting in an 2 economic and ecological crisis during the late 1980s, this region was declared a Biosphere Reserve in 1993 (Diario oficial, 1993; McGuire and Greenberg, 1993). 1.2. Research objectives. This study aims to use an ecological theory to evaluate trophic changes in the upper G u l f over the past 50 years as a result o f the elimination of nutrients by the series of dams built along the Colorado River and by the intense fishing pressure that has been imposed over the past several decades in this ecosystem. The two main objectives are: (1) to describe the ecosystem of the U G C / C R D using trophodynamic ecosystem simulation models representative of 1950, 1980 and 2000 with the help o f Ecopath and Ecosim software (EwE; Christensen et al., 2000); and (2) to investigate possible policy goals based on restoration of delta ecosystems that are under similar conditions as the upper Gu l f or that could be used in the near future, evaluating the cost and benefits of fisheries among economic, ecological and social criteria. It is expected that this knowledge may improve the understanding of these kinds o f human interventions and enable the incorporation o f the results from this study into policies and monitoring programs. The following are planned tasks associated with each objective: Objective I: 1. Review of historic landings, as reported by the Institute o f National Fisheries, Mexico (INP). 2. Examination of fishing effort, discards, bycatch and fisheries value associated with these harvests. 3. Construction of mass-balanced E w E models of the U G C system representative of the 1950, 1980 and 2000 periods, covering major fluxes: phytoplankton and primary production, zooplankton and secondary production, major fish species and their fisheries, marine mammals and their food consumption. 4. To include qualitative information and knowledge from local fishermen of the 3 U G C (San Felipe, Golfo de Santa Clara and Puerto Penasco) as an additional tool to estimate some parameters required by E w E for the 1950 and 1980 models. Objective II: 1. Examination o f changes in the ecosystem form and process by comparing the models constructed according to objective I. This evaluation is performed using dynamic simulations in Ecosim. 2. Generate forecasts o f ecological, economic and employment responses to possible management alternatives using the E w E approach. 3. Employ dynamic Ecosim simulations using the Colorado River discharge as a forcing factor to track and emulate the effect of inter-annual climate variations such as E l Nino /La Nina in the structure and function o f the U G C . 1.3. Biodiversity impacts of large dams. Major dams meet one or more of the following criteria: are at least 150 m high; have a 3 ' 3 volume of at least 15 mil l ion m . They have a reservoir capacity of at least 25 km or a generation capacity of at least one Gigawatt (McCul ly , 1996). In 1998, there were 306 major dams in the world with 57 planned for the near future (Revenga et al., 1998). The Colorado River has the third greatest number of major dams in the world, i.e. a total of 12 (just behind the Parana and Colombia rivers with 14 and 13 major dams, respectively). In general, dams and their reservoirs impact freshwater and marine biodiversities by: 1. Blocking movement of migratory species up or down rivers, causing extirpation or extinction of genetically distinct stocks or species. 2. Changing turbidity and sediment levels, trapping silt in reservoirs deprives downstream deltas and estuaries of maintenance materials and nutrients that help make them productive. This could change the normal seasonal estuarine discharges, which can reduce the supply of entrained nutrients, impacting the food chains that sustain 4 fisheries in estuarine deltas. 3. Changing conditions in rivers flooded by reservoirs: running water becomes still , silt is deposited and deepwater zones, temperature and oxygen conditions are changed, making the environment unsuitable for river and delta species. In the case o f estuarine and marine impacts, dams are associated with changes in seasonal flows, turbidity and productivity as in the North Caspian, where dams caused an increase in salinity from 8 to 19 (Rozengurt and Hedgepeth, 1989). The estuarine mixing zone was compressed and moved up to the delta, reducing the nutrient load (especially phosphorus) and the sediment load was reduced by as much as 5 times. In the Volga, biomass of phytoplankton, zooplankton and benthic organisms decreased by as much as 3.5 times and a substantial part o f the Volga flood plains, that served as nursery grounds for many valuable fishes, was transformed into drying swamps or deserts (Rozengurt and Hedgepeth, 1989). The Volga diversion resulted in a progressive deterioration and significant decline in natural recruitment of commercial species, producing a substantial reduction from three to five times for three sturgeon species, and up to 10 times for bream (Abramis brama), Caspian roach (Rutilus rutilus) and nearly 100 times for the commercial fishery of Caspian herrings (Alosa kessleri volgensis). In a few extreme cases, such as the stellate sturgeon 'sevruga' (Acipenser stellatus), it was necessary to release fry reared in hatcheries throughout a period o f 20 years in order to save this species from extirpation in this region (Rozengurt and Hedgepeth, 1989). 1.4. Present and past ecosystem models: The 'Back to the Future' policy approach. Failure to predict fish stock collapses and their economic and ecological cascade of effects in aquatic ecosystems have resulted in severe critiques single of species fisheries science (Hutchings et al, 1997; Pitcher et al, 1998). It is believed that some depletions and collapses around the world are even worse than we had thought (e.g., large fish, Myers and Worm 2003; fish biomass, Christensen et al. 2003; whales, Roman and Palumbi 2003; sharks, Baum et al. 2002, Schindler et al. 2003: turtles, Hays et al. 2003). 5 Increasing attention has focused on the importance interdisciplinary information about past abundances in determing the extent o f the collapse and the potential for restoration (Pitcher et al, 2005; Saenz-Arroyo et al, 2005). In the U G C , changes to the marine fauna (e.g., the decline of sharks, shrimps, benthic communities and a near extinction of totoaba and vaquita) are parallel to drastic changes in the terrestrial ecology of this region. The first Spanish expeditions to the G u l f of California (e.g. Francisco de Ul loa , 1523) described the upper G u l f as a lush region bordered by vast areas o f riparian corridors and tidal wetlands that supported beavers, otters and thousand o f birds, with deer, Cimarron rams, wolves and mountain lions. These pristine conditions seemed to persist until the twentieth century (according to the maps, flora and fauna described by Sykes, 1937 and Osorio-Tafall, 1947). It has been suggested that the lack of communication between scientists, policy makers and the public has also contributed to failures of fisheries science to predict major collapses around the world during the last century. It is for this reason that the combination o f different disciplines to examine traditional fisheries science questions could result in a new tool that may help fisheries scientists better explain the nature o f stocks and marine ecosystems. In order to answer these critical questions, a new fishery policy approach was proposed in 1998 (Pitcher et al, 1998a) based on a 'rebuilding' plan derived from the architecture of past ecosystems. Basically, this new methodology, called 'Back to the Future' (BTF), attempts to understand the history of ecosystems using the entire spectrum of tools and knowledge available, i.e., incorporating the help and knowledge of maritime historians, archaeologists, ecological economists, fisheries ecologists and also, decades and sometimes hundreds of years of experience of local and indigenous people. B T F provides an integrative approach to the strategic management of marine ecosystems, including new methods for describing past ecosystems that could help in development strategies for the new fisheries that could meet criteria for sustainability and responsibility (Pitcher et al, 2005). Also , B T F is particularly useful when the effect of a specific perturbation (natural or anthropogenic) needs to be evaluated by comparing 6 the structure o f this system before and after such perturbations. This was one o f the main reasons why this new methodology was employed in the quantification of ecological impacts in the upper G u l f that may be attributable to the Colorado River diversion. B T F allowed the reconstruction of this ecosystem (1950s model) for a specific period before the completion and filling of the Glenn Canyon Dam in 1960 in order to compare its trophic structure and function with the results o f 30 and 50 years later (1980s and 2000s models). Changes which resulted in the system due to this anthropogenic intervention were then evaluated and, according to the B T F method, restored past conditions used as a guide to evaluate the costs and befits o f future restoration policies. The B T F approach applied in the upper G u l f o f California project began with the construction o f a present-day ecosystem simulation model (referred to as '2000s' model because it averages fisheries and abundances from 1995-2003). The structure and parameters o f the balanced 2000 model were used as a framework for assembling the past models for the 1950s and 1980s models. During the construction of the two past models, data gained from archives, historical documents, archaeological information (clam fossils were used to estimate past abundances) and interviews vis-a-vis traditional environmental knowledge plus opinions from local fishers o f the upper G u l f were included. Once the present and past models were mass-balanced, real time series o f fishing efforts and mortalities were used to validate the dynamic simulations of these models. Also , by using a time series o f the biomass for two main species exploited in the region in the last 60 years (shrimps, Peneaus stylirostris, P. californiensis and totoaba, Totoaba macdonaldi), it was possible to tune the three models (see details in chapter HI) and to evaluate the changes in the trophic structure and function o f the upper Gu l f o f California from 1950 to 2000. A n exploratory search for sustainable fisheries was then conducted to examine tradeoffs between conservation and fishing that might help in the understanding of this ecosystem and contribute to future decisions among all the stakeholders. Finally, it was possible to challenge the present and past models with forcing functions such as the Colorado River discharge in order to emulate the effects of climate change in the upper Gulf, a critical element to consider in the modelling of past ecosystems. 7 1.5. Considering the effect of climate change in the modelling the past and present of the upper Gulf of California. The U G C is a tidal shallow sea, with approximately 70 % o f its area less than 200 m, and therefore it responds rather quickly to meteorological variability (Reyes and Lavin, 1997). This area is surrounded by deserts and is characterized by atmospheric and marine isolation. Mountain ridges on both sides o f the gulf create an upper gulf climate that is more Mediterranean than marine (Leal and Lavin, 1999). For these reasons, E l Nino/Southern Oscillation (ENSO) events have marked effects on both the physical and biological components of the upper Gulf. In the first case, Alvarez-Borrego and Schwartzlose (1979) documented a stronger invasion o f Tropical Pacific water masses into the G u l f during the 1957 E l Nino. Baumgartner and Christensen (1985) concluded that the principal source of inter-annual variability in the sea level climate o f the G u l f o f California is the changing intensity of the equatorial circulation associated with E N S O phenomena. The biological impacts of climate change in the G u l f can be evaluated at different levels; in the case o f inter-annual E N S O events, some results from Valdez-Holguin and Lara-Lara (1987) suggested an increase in phytoplankton biomass during the E N S O 1982. Baumgartner et al. (1985) have shown that siliceous phytoplankton assemblages preserved in the laminated sediments of the anoxic Guaymas basin (Central G u l f o f California) are correlated to inter-annual sea-level anomalies. They have also reported that the E N S O periods are generally marked by increases in preserved abundances of total siliceous assemblages and, in particular, by a greater number of individuals from species whose distributions are limited to tropical and subtropical waters. Also , the E N S O 1982-1983 was detected in the G u l f of California by a change in the habitual distribution patterns of silicoflagellates (small size phytoplankters with a siliceous framework), where the incursion of 'hot' waters (and tropical species) caused the decline of the silicoflagellate population in the southern part o f the Gulf, and inhibited the influence o f ' c o l d ' waters of the G u l f (Perez-Cruz and Molina-Cruz, 1988). 8 In general, environmental factors have a stronger impact on short-lived species, such as crustaceans and small pelagic fishes. Lopez-Martinez et al. (2003) evaluated the inter-annual variation in the growth of brown shrimp (Farfantepenaeus californiensis) in the G u l f o f California. They established a relationship between temperature and growth o f the brown shrimp (optimum temperature for growth was 25°C), and concluded that growth was favorable only during moderate the E N S O events (1986 and 1992), whereas during more intense events (1982-1983), had adverse effects on growth. These examples of the effect o f climate change on the physical and biotic elements o f the G u l f o f California reflect the need to consider the role of climate change component in the Ecosim scenarios. It is worth noticing that the E w E approach can account for population parameters such as natural mortality, growth and recruitment, all o f which are affected by inter-annual climate changes. 1.6. Previous Research. A n Ecopath model of the north region of the G u l f o f California (the northern G u l f of California and the upper G u l f of California cannot be considered synonymous since their physical, chemical oceanographic conditions are different) was constructed by Morales-Zarate (2001). This Ecopath model represents the main biomass flows in this system with special emphasis on the dynamics of blue shrimp, Liptopeneaus stylirostris, (80% of the total shrimp catch). The model structure consisted o f 29 functional groups distributed in an area of 7,172 k m 2 , from the Colorado Delta southward to the large islands of Tiburon and Angel de la Guarda. The average depth o f the area used in this model was 200 m. Intriguingly, the most affected groups were impacted more by predation and competition than by fishing pressure. The model constructed by Morales-Zarate (2001) and Morales-Zarate et al. (2004) is the only relevant point of comparison for future results from my own upper G u l f model. 9 1.7. T h e s i s s t r u c t u r e . Chapter II presents an exploratory analysis o f illegal and unreported catches in the region to estimate the total extractions to be included in the present-day model. The sources o f information required for building the 2000 trophic model in the upper G u l f of California and Colorado River Delta are provided. Finally, the function and structure of this ecosystem for 2000 is presented through the results from network flow analysis. A quantitative analysis o f the sensitivity and uncertainties o f the data employed is discussed. A qualitative comparison between the former pristine conditions and the changing fauna of the upper G u l f before and after the water diversion is provided in Chapter IH. The results from the oral inputs and opinions from interviews o f the local fishers are presented. The information gathered from this analysis was partially used to build past time series o f relative abundances of the fauna in the region, a critical issue for assembling the 1950s and 1980s trophic models. Chapter IV presents the connection of the past and present models, providing the basis for quantifying changes in the structure and function for the upper G u l f from 1950 to 2000 (considering changes in biomasses, fishing mortalities and energetic flows). This chapter also incorporates the climate change influence into the tuning process for the three models. Chapter IV illustrates how the ecological role of a single functional group can be explored through changes in the form and structure of the past states of the upper G u l f of California ecosystem and how these changes can be used as a first approach to quantify ecological impacts attributable to diversion o f the Colorado River. 76 scenarios for optimal fishing strategies under economic, social, ecological and rebuilding criteria were analyzed to find optimum fishery strategies in the upper G u l f are documented. Changes in biodiversity in the region under different fishing regimens are also discussed. 10 Chapter V I presents the summary and concluding comments. Benefits and limitations o f the 50-year modelling o f the upper G u l f are discussed. 11 Chapter II. A trophic model of the upper Gulf of California and Colorado River Delta. 2.1. Physical and biological characteristics. The study area was selected because of the historical influence of the Colorado River that once flowed 70 km south from the present day mouth (Lavin and Sanchez, 1999; Flessa et al, 2001). This flow affected an area of approximately 4,550 km 2 (Fig. 1). The development and evolution of the U G C and CRD have been controlled by interaction of two geological processes: (1) the Colorado River with an average sediment supply calculated as 160 x 106 tonnes/year (van Andel, 1964); and (2) an extensive tidal regime in the delta region considered to be amongst the largest in the world (up to 10 m), and one which has very strong tidal currents (Alvarez-Borrego, 1983; Carraquiry and Sanchez, 1999). The northern half of the U G C is very shallow, with a minimum depth of 2 m and a maximum of 30 m, while the southern half descends to 120 m (Lavin et al, 1998). The climate is continental arid, with scant rainfall and low humidity (Paden, 1991). According to the non-linear shelf model (using a constant discharge of freshwater of 2000 m3/s) published by Carbajal et al. (1997), the pre-dam salinity pattern (12-34) is radically different from the present salinity of ~37, with profound impact on the U G C circulation (Lavin et al, 1999). The estuary circulation is now driven by the high evaporation rate in the area of 900 cm/year with only 6.8 cm of yearly mean precipitation (Carbajal et al. 1997; Lavin et al, 1999). These changes have transformed the UGC into an inverse or negative estuary, with higher salinity toward the head and lower salinity toward the ocean (Lavin et al, 1998). The saltier water formation in the upper gulf that sinks and flows southeastwards during late winter is one of the mechanisms that maintains a salt balance in the region (Lopez, 1997). Dam projects have trapped much of the Colorado's sediment load. Historically, a charge of 160 millions t/year of sediment was delivered into the delta (van Andel, 1964), while the sediment load today is almost zero (Carraquiry and Sanchez, 1999). Strong tidal waves (up to 10 m in the mouth) have removed historically deposited 12 sediments, and it is likely that the dams have engendered a sediment starvation phase in the delta. Consequently, the delta is passing through an erosional phase and exports sediments to the Northern Gulf at rates similar to those of unaltered river flows prior to 1935 (van Andel, 1964; Cupul, 1994). Despite the ecological impact caused by the construction of dams, the U G C is one of the richest marine ecosystems in the world (Cudney-Bueno and Turk, 1998). The load capacity of its food chain has been evaluated by estimating the high primary productivity reported over recent decades (Zeitzchel, 1969; Alvarez-Borrego et al., 1975; Alvarez-Borrego et al., 1983; Brinton et al., 1986; Valdez-Holguin and Lara-Lara. 1987; Millan-Nunez, E. 1992; Millan-Nufiez et al. 1999; Garcia-Pamanes and Lara-Lara, 2001). The high fertility is explained by the tidal currents, wind-driven upwelling and the plateau topography (Lavin et al., 1999), resulting in numerous areas of upwelling mainly along the eastern coast of Baja California, (Argote et al., 1998) with the productive potential to maintain large food chains with no freshwater input (Millan-Nunez et al. 1999). 2 . 1 . 1 . Historical variability of fluvial discharge. During the nineteenth century, tidal bores of up to ten meters were registered in the delta, allowing a regular commercial navigation of steamer boats to cross the delta and go up the river (Sykes, 1937). Today, the Colorado is no longer used for navigation and its discharge has been diverted by more than 20 dams and dozens of irrigation projects (Carbajal et al., 1997). The completion of the two US major dams, the Hoover Dam (1935) and the Glen Canyon Dam (1960), had the most drastic impact on the amount and timing of fresh water that reached the Gulf (Lavin and Sanchez, 1999). The first documented intervention of the Colorado was in the early 1900s when a discharge of 18 x 109 m3/year was recorded (Sykes, 1937). In 1934, when the Hoover dam started to store water in Lake Mead, the discharge decreased by 80% (2.8 x 109 m3/year) and showed a slight recovery between 1940 and 1952. The completion of the Glen Canyon dam in the late 1960 was responsible for a 96% discharge reduction compared with the beginning of 13 the century (Fig. 2). Some strong precipitation events such as the E l Nino of 1983 and 1993 released more water to the G u l f (Lavin and Sanchez, 1999; Brito-Castillo et al, 2002). Indeed, the rare Colorado fresh water discharge during March and A p r i l o f 1993 provided an opportunity to evaluate its importance. A n evident reduction of salinity was detected (contrary to the now normal inverse estuarine situation) and a dilution pattern was observed of up to 70 km from the river mouth, verifying that the hydrography of the shallow (<30m) upper G u l f is affected by the water delivered by the Colorado River (Lavin and Sanchez, 1999). 14 Figure 1. Gu l f of California (or 'Sea of Cortez') N W Mexico . The bold line represents the a area of 4,550 k m 2 produced by the historical influence o f the Colorado River (70 km from the river's mouth). Significant places referred to in the text are shown. The dashed line represents the Biosphere Reserve declared in 1993 by the Mexican government to protect the delta and marine fauna of the northern gulf, including the endemic species o f vaquita porpoise (Phocoena sinus) and giant gulf croaker (Totoaba macdonaldi). Core zone is a no fishing area. 15 3000 -, 1920 1930 1940 1950 1960 1970 1980 1990 Figure 2. Annual discharge of the Colorado River below the Hoover Dam over 80 years. The horizontal bars represent the years when the Hoover Dam (1934) and Glenn Canyon Dam (1960) were completed. Note that an increase of freshwater was delivered into the upper Gulf during strong precipitation events like E l Nino 1982-83 and 1993 (Data from Lavin and Badan, 1997; Lavin and Sanchez, 1999). 2. 2. The fisheries of the upper Gulf of California. Despite the adverse conditions resulting from the diversion of fresh water, nutrients, and sediments from the Colorado River since the 1920s, the productivity of the upper Gulf of California has been challenged by its years of exploitation. Although today's activities such as tourism and farming are important, fishing is still the main economic activity, with more than 70 species of fish, mollusk, crustacean and echinoderm exploited. The upper Gulf provides 15% of Mexico's total landings, a significant amount considering that Mexico produces 1.3 million metric tonnes of fish per year and is among the world's 20 leading fish producers (Hernandez and Kempton, 2003). 16 Major fisheries in these waters started at the beginning of the 20 Century. The last 80 years have seem significant changes in fishing methods, efficiency and intensity. In the 1920s, canoes were used to capture totoaba with silk and cotton nets under an importation agreement with the U S . These were replaced slowly in the late 1950s by large nylon gillnets, while outboard motors allowed larger vessels and longer trips that continued until the fishery collapsed in 1977. The fishery has remained closed since then, though illegal totoaba fishing is known to have taken place during at least the last decade (D'Agrosa 1995; Cisneros-Mata et al, 1997). Soon after the totoaba fishery was initiated, shrimp trawlers from the central G u l f (Guaymas, Sonora) started to fish in the upper Gulf, and the first company to buy and commercialize shrimp began in 1920 under Japanese owners. Today, the shrimp trawl fishery represents the principal source o f income in the upper Gulf, followed by a highly diverse, mixed and opportunistic small-scale fleet that now exceeds 800 boats, employing gillnets, longlines, traps and hookah diving. These boats fish at all trophic levels, causing fishing mortalities from top predators to benthic filter feeders. There are three small fishing ports in the upper G u l f of California, San Felipe, Puerto Penasco and E l Golfo de Santa Clara (Fig. 1). Also , the indigenous Cocopa community has been fishing in the delta. Today, the 225 Cocopa fishers are the only people allowed by Mexican law to fish in the core zone of the Biosphere Reserve (Cudney-Bueno and Turk, 1998; Ballinas, 2002). A detailed description of the Cocopa community is presented in Chapter III. Some small pelagic species in the central and northern G u l f have been negatively impacted by climate fluctuations (Cisneros-Mata and Ffammann, 1995; Lopez-Martinez et al, 2003). A notable example was documented in the central G u l f with the Pacific sardines (Sardinops Sagax caeruleus) the most important species in the G u l f wet fish fishery. This fishery declined dramatically just after the 1983 E l Nino, as can be seen from the national fishery statistics (Hernandez and Kempton, 2003). Chapter V discusses the role of climate and large scale changes in the productivity of the upper Gulf, including specific documented cases and simulations obtained from the models. 17 2.2.1. Fisheries crises in the upper Gulf of California. In 80 years of exploitation, three major crises occurred in uncontrolled fisheries in the upper Gulf. The first disaster: the collapse of the totoaba fishery, occurred in the early 1970s (Fig. 3) when this species was nearly driven to extinction. A second major crisis occurred during the mid 1980s when enormous schools of Pacific sharp-nose shark vanished (Cudney-Bueno and Turk, 1998). The third fishery collapse came just a couple of years later, when a sudden decline in the production (from 560 tonnes in 1988 to 235 tonnes in 1990) of shrimp was detected (Fig 4). A catastrophic mix of low stock levels, too many fishers and no incentives to conserve the resources resulted in the collapse of the shrimp fishery, which brought about the bankruptcy of BANPESCA, a Mexican federal bank that administrated the fisheries (Hernandez and Kempton, 2003). In an attempt to restore the ecology and economy of the upper Gulf of California, the Mexican government declared the upper Gulf a protected area, thus establishing the Colorado River Delta and upper Gulf of California Biosphere Reserve (Diario Oficial, 1993; McGuire and Greenberg, 1993). During the last 50 years, fisheries catches in the southern Gulf have gradually shifted from long-lived, high trophic level species to short-lived, low trophic species (Fig. 5; Sala et al, 2004); a global pattern that has been described as a worldwide tendency called 'fishing down marine food webs' (Pauly et al. 1998a, 1998b; 2000a; Myers and Worm, 2003). The scenario found in the upper Gulf is no different; the results obtained indicated an overall reduction of 0.2 trophic levels in the catches recorded in the last 50 years, with a clear decline of top predators (sharks and totoaba) during the 1970-80s and an increase in the catch of low trophic species such as corvinas, chano and crabs (these results are detailed in full in section 4 of this chapter). Sala et al. (2004) pointed out that the coastal fisheries of the southern Gulf of California are unsustainable, and their management needs to be re-evaluated to prevent further degradation of coastal food webs. These recommendations are supported by the results obtained in this document, and they could be easily extended to the UGC. 18 1930 1940 1950 1960 1970 F i g u r e 3. Historical time series of landings of the endemic giant gulf croaker or totoaba {Totoaba macdonaldi) in the Gul f of California. This species was exploited from the 1920s until its virtual extinction in the late 1970s. Data from Flanagan and Hendrickson (1976). 19 200 A 0 - I — i — i — i — i — i — i — i — i — i — i — i — i — i — i — i — i — i — i — i — i — i — i — i — i — i — i — i — i 1975 1980 1985 1990 1995 2000 Figure 4. Annual landing of shrimps (Peneaus stylirostris and P. californensis) in one of the most important ports in the northern Gulf of California, San Felipe (Baja California). During the late 1980s a dramatic crisis in the production of shrimp resulted in the Mexican government declaring the upper Gulf of California (including the Colorado River Delta) a Biosphere Reserve in 1993. Data provided by CRIP-Ensenada. 20 > i-J o o H 4.3 i 4.2 -4.1 -4 -3.9 3.8 3.7 3.6 H 3.5 160 r- 120 h 80 h 40 a ttf) cl JO "c3 -4-> O H 1970 1980 1990 2000 Mean Trophic Level • Maximum fish size Figure 5. Temporal changes in mean trophic level and maximum fish size (total length) of coastal fisheries landings in the Southern Gulf of California. Data retrieved from Sala et al., 2004. 2.2.2. First steps to address illegal and unreported catches in the Gulf of California. Illegal and unreported fishing occurs in practically all marine ecosystems and, unfortunately, the Gulf of California is no exception. The Mexican authorities declared in 2002 that, for some resources, illegal fishing can reach up to 30 percent of the total production (Weiner, 2002). This increases the threat to already over-exploited organisms, and makes it critically important to include estimates of illegal and unreported catches in the trophic simulation models. Since marine fisheries began in the Gulf of California, foreign fleets have been a major factor in the sharp decline of large predators such sharks and totoaba. Before and after World War II, American ships openly exploited every school of tuna and every swarm of sardines that they could find, along with sea lions taken for pet food and the extensive use 21 of shark's livers for vitamins and iron. Japanese fleets also fished sardines and sharks intensely in the G u l f (Steinbeck, 1940; Magallon-Barajas, 1987). Those activities clearly marked the rise o f unregulated fisheries in the Gulf. Law enforcement was lax, and economic development strategies that changed with successive governments, resulted in decades o f practically unregulated fishing. In the 1990s, the Mexican government took the first steps to create effective systems of licensing and permits. Using the Mexican Navy to police the outlaws, the government created biosphere reserves along the Gulf, incorporated accurate fishing seasons, established mandatory turtle excluder devices (TEDs) in commercial shrimp trawlers, and created international committees for the recovery o f endangered species. One of the most important regulations was to create the National Fisheries Chart (CNP) in 2000 in attempt to reverse the increasing fishing pressure. This section presents qualitative estimations o f misreported and unreported catches of the principal species exploited in the G u l f (totoaba, shrimps, sharks, sardines and squids), based on the history of their exploitation. While the incentives to misreport catches and estimates presented here are open to discussion, they do represent a critical first step in the analysis of extent o f illegal fishing and design o f effective measures to combat I U U fishing in the G u l f of California. 2.2.1.1. Methods. Intensive historical research was conducted to summarize the history of fisheries in the G u l f of California from 1900 to 2004. The review included a wide range of information, from scientific and historical documents to published interviews. The compiled history includes pelagic fisheries (sardine, Opistonema libertate, and anchovies, Anchoa hellery, A. eschiand), bottom trawling fisheries (blue and brown shrimps) and the historical fishery of totoaba divided by decades from 1900 to 1970, and by five year periods from 1970 to 2004. Relevant aspects in social development, changes in technology and 22 management regimens in the G u l f were also synthesized. This information is presented in Appendix 1. The 80 years of integrated history and important historical changes in Mexican regulatory regimens since the 1920s were employed to estimate incentives to misreport catches only from Mexican vessels in the G u l f (this analysis did not consider the American and Japanese fleets that were fishing from the 1920s to the 1960s). Table 1 presents three possible categories to misreport catches: (1) Discards; (2) Illegal; and (3) Unmandated. These categories were defined according to the criteria proposed for estimating illegal and unreported catches from marine ecosystems by Pitcher et al. (2002). Incentives to misreport can occur due to a broad spectrum of motives, from changes in technology (i.e. processing facilities) or regulations (i.e. change in quotas), to changes in the market (i.e. discarding low value fish) and social factors (some species are for local consumption). Table 1 presents qualitative estimations o f misreporting catches. U p arrows indicate an incentive to misreport, down arrows indicate incentive not to misreport the catches. Table 1 includes a brief description o f the circumstances that could produce a possible incentive to misreport the landings. Table 2 gives estimates of misreporting of the six principal exploited species - totoaba, shrimp, sharks, sardines, squid and demersal resources - in the G u l f o f California from 1920 to 2003: these estimates are presented in a qualitative fashion, ranked in the following gross categories: None (0-1%), Low (1-3%), Low/Medium (2-6%), Medium (3-12%); Medium/High (4-24%) and High (5-25%). These estimates were utilized for each of the three categories mentioned (discard, illegal and unreported). To convert the qualitative estimates presented in Table 2 to a meaningful shape, a series of anchor points is needed at least for the last periods. Unfortunately, at the current stage of this project, no quantitative anchor estimates for the magnitude of illegal landings were obtained. Future work and collaboration with Mexican authorities w i l l be needed to quantify total extractions. The qualitative estimates were used to get an exploratory approximation of 23 the misreported catches. Historical catch data series o f totoaba (1930s to 1970s), shrimp (1950s to 2003) and the landings from 1974 to 2003 reported by the Institute of National Fisheries, Mexico (INP) were multiplied by the percentage of misreporting the catches presented in Table 2. The total extractions estimated for the species considered were incorporated into the three trophic ecosystem models (1950, 1980, 2000) developed in this project. 2.2.2.2. Results The exploratory results obtained suggest that the catches of sharks reported in the G u l f of California may have been underestimated within a range of 1-25%; a value that includes both discards and illegal fishing. The upper limit o f this range was mainly concentrated from 1975-85, and an important reduction o f R J U activities was apparent at the beginning of the 1990s (Fig. 6). The landings of sharks reported by the INP for the three main ports in the upper G u l f in 2000 were approximately 0.17 t/km2/year. This value was multiplied by the 25% of missing catch estimated and a total extraction of 0.23 t/km2/year was applied to the Ecopath model for present day conditions in the upper Gulf. Also , Figure 6 presents the I U U catches calculated for totoaba since the 1920s until its collapse in 1977, and those values were employed in the 1950 and 1980 trophic models constructed in the region. In the case of shrimp catches, this fishery is under tight regulation, including law enforcement by the Mexican Navy in the upper Gulf. I U U catches for this resource were estimated in the range of 1-14% of the total production for the upper Gulf. The history of its exploitation indicated that higher incentives to misreport the catches occurred during 1977-1987 and just before the shrimp collapse at the end of the 1990s. Apparently, the regulations established in the early 1990s have reduced I U U fishing o f this resource at least in the northern Gulf. Figure 7 presents the total estimated extraction of shrimp in the upper G u l f (San Felipe). The official catch of shrimp (blue shrimp) reported by the M P in the upper G u l f (three ports) was approximately 0.70 t/km /year, and so when the I U U estimation was applied, a catch of 0.82 t/km 2/year was estimated. This value was included 24 in the 2000 Ecopath model (Section 3 o f this chapter). Historical background, data consulted and incentives to misreport suggest that I U U catches for the sardine, squid and demersal resources (crabs, octopus, clams and sea cucumber) fisheries are low, probably less than 5-10%. Figure 8 shows the percentage o f I U U fishing in the upper G u l f for the most exploited species considered in this exploratory analysis. 1970 1975 1980 1985 1990 1995 2000 Figure 6. Estimated total extraction (landings reported + I U U fishing) of shark (upper) and totoaba (bottom) in the northern G u l f o f California (thin line) in comparison to reported catch (thick line). The total estimated extractions of totoaba and sharks were utilized in the models of the upper G u l f for the 1950s and 1980s. 25 800-1 600 H r 400 H 200 H 0 -I—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i 1975 1980 1985 1990 1995 2000 Figure 7. Estimated total extraction (landings reported + I U U fishing) o f shrimp in the region of San Felipe, Northern G u l f of California (thin line) in comparison to reported catch (thick line). I U U was higher during the 1970s. Apparently, regulations established since the early 1990s have reduced I U U fishing of shrimp, at least in the Northern G u l f of California. 0 J Blie shrirrp Sharks Demersal Squids Sardines Resources • Catch m IUU Figure 8. Proportion of I U U fishing estimated for five resources in the upper G u l f of California in 2000. The total harvest estimated for these- resources (landings reported + IUU) were the input values (landings) employed in the Ecopath model for present day conditions (Chapter II, section 3). 26 2.2.2.3. Discussion The results presented in this section must be taken as preliminary; nevertheless they exhibit some trends in the incentives to misreport catches during the history of the G u l f o f California fisheries. Illegal, discards, and unreported catches in the G u l f are a current problem that Mexican authorities are attempting to tackle. For example, there were about 1,200 permits (in mid 1990s) for boats in the G u l f o f California and Pacific Ocean, and it is estimated that 20 to 30 percent o f their catch was being taken illegally (Weiner, 2002). During the last decade, the Mexican government has taken initiatives to combat the ecological and economic crises that occurred during the late 1980s in the Northern Gulf. New regulations include establishing a Biosphere Reserve in the upper section o f the Gulf, and creating the N O M (Mexican Official Standards) as an instrument to manage the most important fisheries and to avoid an increasing trend in fishing effort. Furthermore, it has enforced a mandatory use of turtle excluder devices, founded international committees for the recovery o f the vaquita ( C I R V A ) , created variable closed seasons for shrimp by area and deploys the Mexican Navy to patrol areas of ecological and economic interest. From 1994 to 2000, the National Consultative Committee for Sustainable Fisheries (Mexico) promoted the use o f N O M s for 14 fisheries in Mexico, including regulations such as permits, gear specifications, season and area closures, size and quota limits. It has also established sanctions including administrative sanctions or judicial penalties for illegal fishers; however, the reason that only 14 fisheries are regulated by N O M s is because of the length o f time required by federal law to consult all the stakeholders, before implementing the N O M (Hernandez and Kempton, 2004). These regulations marked the birth of a new administration that not only tried to include scientific participation by management but also tried to stop the tendency to change the programs and fisheries policies every six years after each presidential election. A l l o f these structural changes implemented since 1994 had the goal of reducing fishing effort for most o f the fisheries in Mexico, but in a parallel way, they have the potential to reduce 27 R J U (illegal, unregulated and unreported) fishing. Perhaps, some obstacles are still present: inadequate scientific resources to address many artisanal fisheries, lack o f local government participation in management, and refusal by government scientists to share detailed data (Hernandez and Kempton, 2004). The current fisheries policies implemented in Mexico seem to respond to long-term problems, and at least in theory, stop the growth o f illegal fishing documented during the 1970s and 1980s. Incentives and estimates for unreported catches presented here require more accurate and detailed information from the dozens o f fishing camps along the Gulf, especially as to how their catches are registered by local offices and how their reports are sent to the National Fisheries Institute. Direct consultation with skippers and crew on G u l f fishing vessels w i l l definitely play an important role in abating or eliminating I U U fishing. However, these estimations are open to discussion and collaboration from Mexican authorities. A clear need o f anchor points has been expressed to convert the qualitative estimations to misreport the catches to values that can provide confidence intervals throughout a Monte Carlo simulation based on likely error ranges, as suggested by Pitcher et al., 2002 and Ainsworth and Pitcher 2005. In the next stage, this I U U project must try to establish anchor points in order to quantify the impact o f I U U fishing and obtain a better approximation of the total extraction in the G u l f of California. A quantitative estimation of I U U fishing activities is just one o f a series of effective measures to combat this problem in the G u l f of California and worldwide. After all, the G u l f of California is recognized not only by its cultural relevance to Mexico , i.e., providing fishing for hundreds of years to several indigenous communities, but also by its richness in producing 40% of the total marine production in Mexico. Consequently, there is an obligation to estimate the true magnitude of illegal and unreported fishing. 28 Table 1. Summary o f qualitative estimates of unreported catches. Up arrows indicate an incentive to misreport, down arrows indicate an incentive to report correctly. A brief description of the circumstances that could produce the incentive to misreport landings is included. Numbers correspond to the reference numbers presented in Appendix 1. Year Regulation or Management regime Other 1920-1929 Agreement between U.S. and Mexico for importation of totoaba (f) (52). Totoaba sport fishery began (|)(11). Totoaba fishery in the central Gulf started to decline, so that 6 German fishermen followed totoaba migration to the upper Gulf (f) (11). Totoaba juveniles taken accidentally during shrimp fishing (f )(13). High price of totoaba meat in US and Asian markets. Change to outboard motors (t) (39). 1930-1939 Closed season for penaeid shrimp first introduced (j) (10). Japanese trawlers fishing in the Mexican Pacific coast and located the main trawling areas in the same decade (t)(14). 17 sardine boats modified for shrimp fishing (f)(14). 1940-1949 Closed season for totoaba from Apri l to May to protect spawners (J.X52). "Before and After World War II, American and Japanese ships took every school of tuna and every swarm of sardines they could, along with sea lions for pet food and sharks to use the livers to remedy iron-poor or tired-blood" (f) (8 in 9). 1950-1959 1955 a new Biosphere Reserve's nuclear zone was established, (J,) (10). 1950 shrimp fishery (offshore and small scale sectors) grew exponentially (t )(14). Double-ring trawls introduced into the shrimp fishery (|) (14). 1960-1969 Shrimp fishery expanded to the mouth of the GoC (f) (14). Late 1960s Shrimp trawls fishermen gradually reduce mesh size (implies increase of bycatch) (f) (16). 1970-1974 No regulatory measures which fix catch quotas for totoaba units(t)(38). Mexican immigration to the Gulf in a kind of 'Gold Rush' for shrimp which was known as 'Pink Gold Rush' (t)(l)- Refuge zone declared in 1955 now (1974) declared a reserve zone banning all fishing in the core zone (J.)(32). 1975-1979 Mexican Law protect totoaba. Fishing banned ( i ) (18). Mesh size regulation introduced in 1977 as a management measure in the shrimp trawl fishery ( |X 16). Drastic declines in totoaba, commercial and sport fishing was outlawed, but totoaba gillnetting continued until 1990s and it was a primary cause of incidental kills of vaquita (f) (49).Unknown de number of small boats (pangas) fishing in the GoC, but there is an estimation of 5,000 - 7,000 pangas fishing (t)(19). Institute of National fisheries, Mexico (Institute Nacional de la Pesca, INP) began monitoring programs to evaluate shrimp spawning and recruitment each year until 1987(J.)(14). 1985-1989 E E Z (Economic Exclusive Zone) established Q)(60). Approximately, 24,000 sharks killed in 1987 in Central GoC (t)(24). 161 tonnes of totoaba poached in 1985 (T)(64). 29 Table 1. Continuation. Year Regulation or Management regime Other 1990 1994 Trawler forced to use turtle-and finfish excluder devices Q)(15). P R O F E P A -government environmental enforcement agency more vigilant on activities in the Reserve (|)(6). Inshore fleet of the upper GoC began to hammer chano ( M megalops) to supply the Asian market for surimi (t)(15, 27). Increase of sport fishing in the upper Gulf (]) (1). Shark camps in the central Gulf grew exponentially, with > 200,000 sharks killed between 1985-93 (f)(24). Near collapse of shrimp fishery provoked diversification of small scale fishing activity and target species switch (T)(6, 56,). Establishment of the Biosphere Reserve, U G C ( i ) (12). Use of radar to fish sardine (69). 1995 1999 Turtle excluder devices mandatory in commercial shrimp fishery Q) (7, 44). No fishing in the core area of the Biosphere Reserve, also, no trawling in depths lower than 9m (|) (26). Shrimp trawlers and sardine purse seines fishing squids with appropriate gear (f)(58). Estimated 200,000 tonnes of bycatch for the GoC catch (t)(28). Mexico created C I R V A : International Committee for the recovery of the Vaquita (j) (26). Use of GPS technology to track banks and fishing sites (69). 2000 CNP (National Fisheries Chart) established no more increments in fishing effort in the upper Gulf with 700 tonnes of shrimp per year U)(26). It is difficult to know the real number of boats fishing the GoC because they can move easily from state to state bordering the Gulf, and cross it from side to side following target species (1) (62). 2001 For the first time the closed season for shrimps was variable by area (|) (46, 49). More than 1,600 km of gillnet were sold in Sonora (Central Gulf) (t)(9). As many as 150 turtles per day were harvested and killed during the peak of the season (just in San Carlos/Guaymas area-Central GoC) (T)(9). 2002 Trawling activity banned in the Biosphere Reserve (J.) (26, 62). "Fishermen, businessmen, scientists and even some federal officials say at least 12,000 unregulated fishing boats, probably more, now at large in the GoC "(T)(62). Jeronimo Ramos, the National Fisheries commissioner said about 1,200 permits existed for boats in the GoC and Pacific Ocean, estimating 20-30% of the catch was being taken illegally (f)(9)-2003 PROFEPA confiscated 11,000 kg of turtle in the upper Gulf during 8 days)(t)(30). CONAPESCA gave 2,400 permits for shrimp fishing for the season 2002-03,but there are between 6,000 to 7,000 boats fishing shrimps in the GoC, all year around (including closed season) (f) (30). 30 Table 2. Continuation. Footnotes: 1. In 1923 shrimp fishery starts with totoaba juveniles taken accidentally (Majallon-Barajas, 1997). 2. Bycatch of Vaquita began (Vidal, 1995). 3. 17 sardine boats modified for shrimp fishing (Majallon-Barajas, 1997). 4. Japanese trawlers fishing the Mexican Pacific coast locate the main trawling areas during this decade (Majallon-Barajas, 1997). 5. The most important changes in boats during the development of fishery in the GoC was the use of outboard motors for small boats were formerly propelled by oars at the end of the 1920s (Chute, 1928; Chute, 1930). 6. Closed season for totoaba from March to May to protect spawners (Cisneros-Mata et al., 1995). 7. American ships took every school of tuna, sardine and shark that they could (Steinbeck, 1940 in Shwartz, 2003). 8. The Japanese (besides Americans) came too during 1940s (Majallon-Barajas, 1997). 9. Completion of Highway 8 in 1942 and the railroad in 1947 increased sales into U.S. markets (Boyer, 1996). 10. 1950 shrimp fishery (offshore and small scale sectors) grew exponentially. Poorly recorded (Majallon-Barajas, 1997). 11. A new Biosphere Reserve nuclear zone was established (1955) in the upper GoC, fishing in the mouth of the Colorado River is prohibited (Diario Oficial de la Federacion, 1955). 12. Closed season for totoaba changed according to its abundance (Cisneros-Mata et al., 1995). 13. Totoaba landings start to decline (Cisneros-Mata et al., 1995). 14. Double-ring trawls introduced in the shrimp fishery (Majallon-Barajas, 1997). 15. In 1955, the shrimp C P U E (Catch/Boat) had its maximum in Guaymas (central GoC), the number of shrimp boats is unknown, but estimated around 800 (Majallon-Barajas, 1997). 16. Shrimp fishery expanded to the mouth of the GoC 17. Sardine fishmeal plants and canneries installed in Guaymas, central GoC (Cisneros-Mata et al., 1995). 18. Although prohibited by law, dynamite is also used to kill totoaba directly (Ramirez, 1962). 19. High shrimp prices; there was a Mexican immigration to the GoC in a kind of 'Gold Rush' for shrimp which was known as 'Pink Gold rush' (Boyer, 1996). 20. Approximately 30 large boats (shrimp trawlers) dedicated partially or exclusively to totoaba fishing in the U G C (Arvizu and Chavez, 1970). 21. No regulatory measures to set catch quotas for totoaba or to limit the number of fishing units (Arvizu and Chavez, 1970). 22. Pacific sardine fleet moved to Guaymas (central GoC) for fishing during the winter season (Cisneros-Mata et al., 1995). 23. Mexican Law protected totoaba - all fishing banned (Diario Oficial de la Federacion, 1975). 24. Totoaba gillnetting continued until 1990's and it was a primary cause of incidental kills of vaquita (Cisneros-Mata et al, 1995). 25. Increase of activity of the Pacific sardine fleet in the GoC (Cisneros-Mata et al., 1995). 26. Mesh size regulation in the trawl fishery introduced in 1977 as a management measure in the shrimp trawl fishery (Lluch-Belda, 1977). 27. Approximately 28 Peruvian boats entered to the GoC after the collapse of the Peruvian anchovy fishery (Cisneros-Mata et al., 1995). 28. Monofilament gill nets became the tool of the panga fishermen; 5,000 - 7,000 pangas fishing in the GoC (Sea Watch, 2003b). 29. The Institute of National Fisheries, Mexico (INP) began monitoring programs to evaluate shrimp spawning and recruitment each year (Majallon-Barajas, 1997). 30. 344 000 tonnes of bycatch in the shrimp fishery from 1977-80 (Grande-Vidal, 1980). 31. 1978 started the commercial exploitation of demersal resources in the GoC with 12 trawlers (Grande-Vidal, 1980). 32. Still fishing totoaba (Campoy-Fabela, 2002). 33. 35 vaquitas killed per year (1985-1992) in shark gillnets (Vidal, 1995). 32 34. Expansion of the Pacific sardine fishery in the GoC (Cisneros-Mata et al., 1995). 35. Increase of the Pacific shrimp catches reaching levels similar to the peak ones (Morales-Bojorquez, 2001). 36. More than 80% of the total catch of shrimp in Mexico is caught in the GoC (Majallon-Barajas, 1997). 37. It has begun the commercial fishery of demersal resources with a potential biomass exploitable of order of 1,374,000 tonnes per year (Grande-Vidal, 1980). 38. Between 1987 and 1989, 10 pangas caught 8,000-10,000 sharks each season in San Francisquito, central GoC (Sea Watch, 2003c). 39. Between 1985 and 1993 over 200,000 sharks were killed in GoC (Sea Watch, 2003c). 40. Drastic decline of Pacific sardine fishery in the GoC (Cisneros-Mata et al., 1995). 41. Increase of demersal fishing, where chano represents 16% of the finfish fishery in the U G C (Instituto Nacional de Pesca, 2000). 42. Tonnes of chano and other species left to rot on the sandy streets of the Golfo de Santa Clara, U G C (McGuire and Valdez-Gardea, 1997). 43. Not enough ice and equipment to process and freeze the landings in the U G C (personal communication from fishers interviewed, 2003). 44. 12 vaquitas were killed in 1993 during 8 months of chano fishing in the U G C (McGuire and Valdez-Gardea, 1997). 45. Trawler activities forced to use turtle- and finfish excluder devices (Diario Oficial de la Federacion, 1992). 46. In 1992, the Mexican authorities banned the use of gillnets with a mesh size > 25 cm (Diario Oficial de la Federacion, 1992). 47. Gulf corvina {Cynoscion othonopterus), return in large numbers to the upper Gulf after 40 years (Vazquez-Leon and McGuire, 1993). 48. Shrimp bycatch ratio estimated at 1:10 (Brusca et al., 2001). 49. Establishment of the upper Gulf of California/ Colorado River delta Biosphere Reserve. In its core zone: all commercial fisheries banned. Increased regulations for most fisheries (Reserva de la Biosfera y alto Golfo de California, 2003). 50. In 1994, a gillnet fleet backed by unknown financiers appeared in Sonora, fishermen and scientists say it slaughters thousands of sharks solely for their fins ($300 a pound in Asian markets; Weiner, 2002). 51: In 1995, almost no shark at all in central GoC, the big shark processors moved to their next target (Sea Watch, 2003c, d). 52. Trawlers forced to carry turtle-excluder devices and no fishing in the core area of the Biosphere Reserve, also, no trawling in esteros or depths lower 5 fathoms (~9 m) (Campoy-Fabela, 2002). 53. Turtle excluder devices mandatory in commercial shrimp fishery (Vidal, 1995). 54. At least 70 species of fish, mollusks, crustaceans and echinoderms are regularly caught by the artisanal fleet in the U G C , approximately 40% of these are designated for international market (Cudney-Bueno, 1998). 55. Estimated 200,000 tonnes of bycatch in the shrimp fishery for the GoC (Anuario de Pesca, 1996). 56. Drastic drop in the Monterrey sardine landings (Sanchez-Velazco et al., 2002). 57. It is difficult to know the real number of boats fishing in the GoC (Lopez-Martinez et al., 1999). 58. As many as 150 turtles per day being harvested and killed during the peak of the season in central GoC (Weiner, 2002). 59. More than 1,600 km of gillnet were sold in Sonora (central GoC; Weiner, 2002). 60. Jeronimo Ramos, the National Fisheries Commissioner said about 1,200 permits existed for boats in the GoC and Pacific Ocean, estimating 20-30% of the catch was being taken illegally (Weiner, 2002). 61. Fishers, businessmen, scientists and even some federal officials say at least 12,000 unregulated fishing boats in the GoC (Weiner, 2002). 62. Ratio between shrimp and discarded fauna is 1:10, using turtles and finfish excluder devises could be 1:6 (Campoy-Fabela, 2002). 63. 140 shrimp boats in the UGC, but an undetermined number of boats from other parts (Guaymas, Topolobampo, Mazatlan, Yavaros) enter to the Reserve (Campoy-Fabela, 2002). 33 64. PROFEPA confiscated 11,000 kg of turtle (corvina) in the U G C during 8 days (NOTIMEX, 2003). 65. C O N A P E S C A gave 2,400 permits for shrimp fishing for the season 2002-03, but there are between 6,000 to 7,000 boats fishing shrimps in the GoC, all year around (including closed season; Velazco-Rodriguez, 2003). 66. There is no control for foreign boats in the GoC, it is necessary to make a record of these boats (Velazco-Rodriguez, 2003). 67. Jumbo squid fishery in central GoC began, fishing from April to June (Morales-Bojorquez et al., 2001). 68. Japanese vessels appeared in 1979 fishing squid (Ramirez, 1985). 69. Jumbo squid fishery collapsed, associated with over fishing, recruitment and El Nino, 1982-83. (Morales-Bojorquez et al., 2001). 70. Shrimp trawlers and sardine purse seines fishing squids with appropriate gear (Markaida and Sosa-Nishizaki, 2001). 71. Squid fishery almost disappeared from the central GoC at the end of 1990s (Morales-Bojorquez et al, 2001). 72. During 1980s Federal Legislation gave exclusive rights of shrimp exploitation to cooperatives (avoiding private investment); this policy produced an evident increase of shrimp catches that in most of the cases was a product for international market (Hernandez and Kempton, 2003). 73. Most of the fisheries status for Mexico was uncertain (Hernandez and Kempton, 2003). 74. In early 1990s, the federal agency in charge of fisheries demonstrated signals of inefficiency, two elements determined such inefficiency, the fishery policy from 1970 did not change significantly and the growth in artisanal fisheries created problems of monitoring and enforcing law (Hernandez and Kempton, 2003). 75. 500 shrimp trawlers estimated to fish inside the Biosphere Reserve (Anonymous, 2002). 76. S E M A R N A P created the National Consultative Committee for Sustainable Fisheries to institutionalize the participation of fishers, scientists and other stakeholders, such as local authorities and NOG's (SEMARNAP. 1998). 77. N O M ' s were not enough to correct the over-exploitation trend that more than 80% of the fisheries in Mexico was undergoing, where only 14 fisheries in the country were regulated under NOM's (Hernandez and Kempton, 2003). 78. The N F C (National Fisheries Chart) shows that 82% of Mexican fisheries are fully or over-exploited, justifying policies to reduce the effort in all those fisheries (Hernandez and Kempton, 2003). 34 2.3. An ecosystem model of trophic structure and fisheries interactions in the upper Gulf of California and Colorado River Delta for 1995-2000. 2.3.1. Introduction The declines and losses o f major fish stocks worldwide attributable to intensive industrial fishing in the past century have been extensively documented. This point to a dismal failure o f management based on traditional single species stock assessment has been unsuccessful, pointing out the failure o f fishery science during the past decades with dozens o f collapses and crises. Unfortunately, the fisheries in Mexico are no different. 82% o f the stocks are reportedly overexploited with an increasing growth in the small-scale fishing effort (Hernandez and Kempton, 2005). The single species approach, applied in the G u l f o f California and in Mexico in general, is inappropriate for tropical small-scale fisheries as it was developed for temperate regions where single species fisheries are prevalent and where fish are harvested by a few gear types at most (Munro, 1979). A different scenario is presented in tropical regions where typically the fisheries have a multi-species, multi-gear nature and harvesting occurs at different stages of the life histories. The single species approach is also weak as it fails to consider the following critical aspects: (1) biological and environmental interactions occurring at varying temporal and spatial scales; (2) the changes produced by humans (e.g. habitat degradation, pollution and increasing fishing); and (3), the influence of short-or long-term climatic changes, resulting in limited fishery policies with narrow views and unrealistic expectations of levels o f sustainable exploitation. Disappointingly, fishery science has failed to avoid fish stock collapses, generating frustration among scientists during the last decade and the desire to use new approaches with the potential to answer more complex, ecosystem-based questions and ultimately to manage and rebuild ecosystems (Pitcher 1998a, 1998b, Pitcher and Pauly, 1998). 35 For these reasons, this study uses ecosystem modelling to explore the structure, function and energetic dynamics of the upper G u l f o f California. The modelling is focused on trophic interactions among 50 functional groups of the upper Gulf, providing assessments of energy fluxes and partitioning of resources among predator groups, including extractions by fisheries. In addition, a network analysis quantifies flows for a comparison with models of past ecosystems in the upper Gulf, examining changes in form and function that are important for management. The mass-balance ecosystem model for the G u l f o f California in 1995-2000 developed in this chapter covers more than 120 species and is used as the basis for a dynamic model that can track and emulate changes caused by water diversion, fishing and fluctuations in climate over the past 60 years. In an ecosystem context, the health and integrity (associated with both structure and function) o f the upper G u l f of California and Colorado River Delta are discussed through a present-day food web model that describes the interplay of predators, prey and fisheries. Chapter ILT presents models o f the system as it was in 1950 and 1980 and presents the 'Back to the Future' approach to restoration ecology. 2.3.2. Methods. The first step to building a mass-balance trophic model with Ecopath is to define the objective and the area. Here, the objective was to create an ecosystem model to evaluate the impacts of the water diversion from the Colorado River. The area was defined according to the historical Colorado River influence in the upper G u l f that used to flow 70 km from the river's mouth (Carbajal et al, 1997; Lavin, 1999; Flessa, 2001). This pattern produces a triangular area o f nearly 4,550 k m 2 (Fig. 1). Approximately 80% of the model area was declared a Biosphere Reserve in 1993 (Diario Oficial, 1993). A l l commercial fisheries are prohibited in the core zone of this reserve except for traditional practices by the aboriginal Cocopa people living in the Delta, and clam harvesting by local residents, ( S E M A R N A T , 2003). The model area includes the three coastal fishing communities of San Felipe, E l Golfo de Santa Clara and Puerto Pefiasco (Fig. 1), which 36 have supported important fisheries since 1920 and which now provide 15% of the national fish production. 2.3.2.1. Constructing the trophic model for the CRD/UGC ecosystem. Ecopath mass-balance models account for trophic interactions averaged over a pre-defined area and time period at multiple trophic levels among organisms within the ecosystem area defined (Polovina, 1984; Christensen and Pauly, 1992; Christensen et al., 2000). Dynamic routines (named Ecosim and Ecospace) use the mass-balanced model generated by Ecopath to simulate changes that may include effects of human activities, including fisheries, other disturbances and stressors on the biological components in the system (Walters et al., 1997; 2000), providing an effective tool for evaluating ecosystem impacts. More comprehensive descriptions of the basis, scope and pitfalls o f Ecopath with Ecosim (EwE) can be found in Christensen and Walters (2004) and the freely distributed software is available at www.ecopath.org. According to the methodology proposed by Christensen and Walters (2004), the general approach o f E w E is summarized below: Ecopath uses a series of simultaneous linear equations, one for each functional group to quantify the energetic flows among trophic groups according to the law of conservation of mass or energy (Equation 1). The net production of a functional group equals the total mass removed by its predators and fisheries plus its net migration and its energy or mass that flows to detritus. The master equation is described as: Production - Predation - Other mortality - Exports = 0 or Production=Fishing mortality + Predation mortality + Other mortality + Biomass Accumulation + Net Migration B, • PBj • E E , = Y i + SBj • (Q/B)j • DC,j + E X j Where: 37 B, and Bj are biomasses o f prey (i) and predators (j) respectively; PBj is the production/biomass ratio, equivalent to total mortality (Z) in most circumstances (Allen, 1971); EEj is the ecotrophic efficiency or the proportion o f production utilized in the system Yj is the fisheries catch per unit area and time. Q/Bj is the consumption/biomass ratio of the predator DCj, is the fraction of prey / in the diet o f predator j E X j is the export o f i. The input data for any particular functional group are: P / B , Q / B , B , E E and D C ; however, Ecopath requires D C as an input while any one o f the other parameters can be estimated by mass balance i f the other three are known. Normally E E is estimated, but in cases when biomass is unknown, it is possible to obtain an estimate by making an assumption about E E (0.95, for example). However, in the case of particular groups in this U G C model (totoaba, vaquita porpoise and blue shrimp), ontogenetic changes were represented by splitting these groups into recruits or pre-adults and adults. 2.3.2.2. Aggregation of functional groups. The 120 species considered in the model were organized into groups based on the functional roles of the species, and which took their food habits and natural history into consideration. Groups with important conservation values were kept as individual groups, producing single groups for vaquita porpoises (Phocoena sinus), giant gulf croakers or totoaba (T. macdonaldi) and the endemic clam of the Colorado delta, Mulina coloradensis. In the case of species of economic interest, such as a gulf croaker called 'chano' (Micropogonias megalops), and the blue (P. stylirostris), brown (P. californiensis) and rock (Sycyonia spp) species o f shrimp were also modeled as unique groups. In order to evaluate the importance of the estuarine waters of the upper G u l f as a nursery area (Cisneros-Mata et al, 1995; Glenn; Calderon-Aguilera et al., 2000; Aragon-38 Noriega et al., 2001, Brusca et al., Glenn et al., 2001) juveniles or pre-adults of vaquita porpoise, totoaba and shrimps were also included in the model. Our classification resulted in 50 ecological groups (Table 6). 2.3.2.3. Sources of the basic input parameters. Biomass: Biomass (B), the ratio of production to biomass (P/B), and the ratio of consumption to biomass (Q/B) represent the 'basic input parameters' o f Ecopath models. Numerous sources o f information were used to estimate these basic parameters. In general, biomasses were estimated mainly from the databases generated from 1991 to 1995 by the INP (National fisheries Institute, Mexico). This institution, through its regional center of fishery research located in Ensenada, Baja California (CRIP-Ensenada), has conducted several surveys and trawl samplings in the shallow waters of the upper G u l f to estimate biomasses of benthic organisms: INP generated maps of abundance per unit area for more than 20 species from 1991-1995 (CRIP-Ensenada, 1995). The following sections summarize how the biomass was estimated for the principal model groups (phytoplankton, zooplankton, benthos, seabirds, fish marine mammals and detritus). A detailed list o f the references employed for. constructing the 50 functional groups in our G u l f o f California model is presented in Appendix 9. Primary production: The biomass of primary producers was the sum of the four producers considered in the model (phytoplankton, macro-algae, seagrasses and seaweeds). Primary production in the U G C can be two or three times greater than that in the open Pacific or Atlantic Oceans at similar latitudes (Zeitzchel, 1969). Biomass estimates were based on local surveys by Millan-Nufiez, et al. (1999), where, according to chlorophyll concentrations and depths registered during their surveys, average primary production associated with phytoplankton was estimated as 28.9 t/km 2 . The intertidal macrophyte communities of the upper G u l f comprise more than 39 19 major taxa (mainly Ulva, Porphyra, Sargassum, and Colpomenia) covering more than 60% of the rocky habitats (Littler and Littler, 1981). The mean macrophyte net production per square metre reported in 1980 for the upper G u l f was 47.9 mg C day"1 (equivalent to a standing biomass o f 1.145 t/km 2), a value that was employed for the 1980 and 2000 models (see Chapter III for details). N o production estimates for macrophytes, seagrasses and seaweeds were found for the time period modeled (1995-2000), and so their biomasses were estimated by Ecopath (Table 5). Nevertheless, these groups are wel l distributed throughout the G u l f o f California (Steward and Norris, 1981; Little and Little, 1981; Meling-Lopez and Ibarra-Obando, 1999) and more research on their biomass and distribution in the upper G u l f is needed. For this reason, the biomass estimated for the early 1980s was used for this group. During the tuning and fitting o f the past and present models (Chapter Til), a compatible estimation of the biomass predicted for the 1980s by Ecopath is compared with that obtained from surveys and the abundance results for the 2000 model. Zooplankton: abundances of zooplankton and microzooplankton were taken from local surveys by Farfan and Alvarez-Borrego (1992) and Garcia-Pamanes and Lara-Lara (2001), respectively. In general, the upper Gu l f is a very rich region with zooplankton biomass that can reach up to 154 mg/m 3 (Montague Island, Figure 1). A n average biomass for the upper G u l f from was estimated at 26.3 t /km 2 (Farfan and Alvarez-Borrego, 1992). This functional group includes copepods (Calanus pacificus), cladocerans (Penila spp) and chaetognaths (Saggita enflata). P /B and Q/B ratios were taken from Chavez et al, (1993) and Garcia-Pamanes (2001), respectively. Benthos: Information available in the literature is scarce for the benthos community in the U G C . Some biomasses of invertebrates were estimated from a rigorous study done by the INP from 1992-1995 (CRLP-Ensenada, 1995), that aimed at estimating the bycatch in shrimp trawls; estimates are provided for Cnidaria, Mollusca, Arthropoda and Echinodermata. In the absence of data between 1995-2000, information reported by Felix-Pico (1975), Romero (1980) and from Salazar-Vallejo (1990) was used 40 (Table 1). The Q/B ratios were generally estimated by using the empirical formula reported for these groups by Pauly et al, 1993. Since there was no information on mortality rates for most of these groups, P /B ratios were taken from other Ecopath models for the central and northern G u l f o f California (Arreguin et al, 2002; Morales-Zarate, 2004). Seabirds: abundances o f seabirds on the coast and islands o f the U G C were based on visual census (Peresbarbosa and Mel l ink , 1994; Mel l ink et al, 1996 and 1997). During the winter o f 1991, more than 100,000 birds of 36 species (most of them shorebirds, Calidris mauri, black skimmer, Rhynchops niger and American avocets, Recurvirostra americana) were observed in the upper G u l f (Morrison et al, 1992). Due to its importance for shorebirds, the Colorado River Delta has been incorporated into the Western Hemisphere Shorebird Reserve Network. In the summer, this habitat is less important, but resident populations of brown pelican (Pelecanus occidentalis), double-crested cormorants (Phalacrocorax auritus), gulls (Larus delawarensis), white terns (Sterna caspia) are abundant (Peresbarbosa and Mell ink, 1994). For the model, seabirds were split into two gross groups. The first group is represented by shorebirds (planktivorous birds, such as Calidris mauri, willets, avocets, etc) l iving in the deltas located in the mouth of the rivers (less than 2% of the area modeled. The numerical abundances o f this group can easily reach more than 80,000 birds during the winter (Morrison et al, 1992). A conservative abundance of 40,000 resident birds with a mean body weight o f 50g was used to estimate the biomass of 0.001 t /km 2 used in the Ecopath model. The second group comprises seabirds such as pelicans, gulls, cormorants. A gross estimation of 18,000 birds in the region was used (based on number of pairs of birds recorded in Montague Island by Peresbarbosa and Mel l ink, 1994) and using a mean body weight of 600g (mainly gulls, terns and pelicans), it results in a biomass estimated at 0.003 t/km 2 . Fish. Fish stock biomasses were estimated from different sources. In the case of demersal species, surveys done by INP-CRIP-Ensenada (1992-1995) were used. 41 Using the method of Cushing (1971) based on swept area, the maps o f abundances (individual/hour) were transformed to an average biomass (t/km ). In most cases, the 1995 survey was used as an average year and to set the biomass values for the Ecopath model. These local surveys reported the abundance o f more than 20 species (juveniles and adults of totoaba, T. macdonaldi, corvinas Cynoscion spp, snappers, Lutjanus spp., shrimps, Peneaus spp., crabs, Callinectes, rays, Urobatis spp, guitarfish, Rhinobatos sharks, Carcharhinus limbarus, Heterodontus spp) with ecological and economic value and other fish with minor economic value (aggregated in the model as 'other fish'). Table 3 presents a comparison between the biomasses estimated by the INP surveys and the biomasses employed to construct the trophic models reported by Morales-Zarate et al. (2004) and this thesis. Figure 9 shows three examples o f the species sampled (shrimp, crabs and chano) during the 1995 surveys. In some cases, and for single species (e.g. totoaba), Virtual Population Analysis ( V P A ) was applied to the historical time-series o f catch-at-age data to estimate fishing mortalities and biomass back to 1950 (more details are presented in Chapter HI). 42 Figure 9. Contour map of abundances of chano (left) and crabs (right) in 1995. The abundance of more than 20 benthic species (crustaceans and fishes) was estimated from four years of surveys performed by the INP (Institute of National Fisheries) from 1991-1995 (CRIP-Ensenada, 1995), and these biomasses were employed in the 2000 model (maps were provided by the Institute of National Fisheries, Mexico). Maxima Abundance Scales: chano, 3600 individuals/hr; crabs, 7500 individual/hr. 43 2 Table 3. Biomasses estimated (t/km ) from the maps of abundance generated in the Northern G u l f o f California by the National Fisheries Institute and CRJP-Ensenada. Biomasses used in another Ecopath model for this region (Morales-Zarate et al, 2004) are compared with those utilized in this thesis. See references for complete citations. 1991 Surveys 1992 1994 1995 Survey average Morales-Zarate model. This thesis Sharks 0.02 0.28 0.0 0.05 0.09 0.47 0.28 Serranids 0.06 0.02 0.16 0.11 0.09 0.29 0.16 Sciaenids 0.77 0.11 3.99 0.33 1.29 0.35 3.99 Rhinobatids 0.45 0.35 0.0 0.22 0.26 0.33 0.35 Rays 1.31 0.89 14.6 0.56 4.35 0.56 4.35 Flatfishes 0.02 0.02 0.33 0.05 0.10 1.51 0.33 Marine mammals. The high abundance of marine mammals is one of the reasons why the upper G u l f is recognized nationally and internationally as a 'hot spot' o f diversity (Jaramillo-Legorreta et al., 1999). Sea lions, Zalophus californiensis, are abundant from a colony off San Felipe (Baja California), and at least seven species of cetaceans were recorded after a period (1986-1987) o f freshwater release from the Hoover and Gleen Canyon dams (Silber et al, 1994). Biomass estimates used the numerical abundances reported by Zavala-Gonzalez and Mell ink, 2000 (for sea lions) and by Silber et al. 1994 (for dolphins and cetaceans) multiplied by the average weight for each species (Table 4). A n estimate of 567 vaquitas (Jaramillo-Legorrreta, 1999) was used to estimate biomass. Tagging studies by Urban (2002) suggest that there is a permanent population of fin whale (Balaenoptera physalus) in the central Gu l f of California that makes feeding journeys into the upper Gulf. Table 4 shows the estimated numerical abundances of marine mammals in the region. 44 Table 4. Numerical abundance of cetaceans in the upper G u l f of California, according to; (1) Silber et al. 1994; (2) Jaramillo-Legorreta et al, 1999 and by (3) Zavala and Mel l ink , 2000. See references for complete citations. Species Name Numerical abundance Zalophus californiensis California sea lion 20,000 3 (Central GoC). Delphinus delphis Common dolphin 14,239' Tursiops truncates Bottlenose dolphin 1,416" Phocoena sinus Vaquita 5672 Balaenoptera physalus Fin whale 215' Balaenoptera physalus Fin whale 215' Orcinus orca Killer whale 17' Balaenoptera edeni Bryde's whale 7' Escgrictius robustus Gray whale 3' Detritus: there was no available information for detritus accumulation in the area, but an empirical relationship established by Pauly et al, 1993 was used to obtain a rough estimate o f this parameter: logD = 0.954 logPP + 0.863 logE-2.41 Where D is detrital biomass in g C m 2 , PP is primary production in g C m 2 year _ I and E is the euphotic layer depth (m). For the model, a PP =115 g C m 2 (according to Mi l l an et al, 1999) and average E = 20 m were used. At this point, no information of irradiance has been found for the area, so the euphotic depth was estimated at 20 m. The fit o f the regression equation to the data is not very good, but as suggested by Pauly et al, (1993), it might be considered sufficient in cases where no other information is available. A n estimation of 28.7 t /km 2 was used employed for the 2000 model, a value that seems to be reasonable, considering the more than 450 millions of metric tonnes of sediment/year that once reached the upper G u l f (Van Andel, 1964; Carbajal et al, 1997; Carraquiry and Sanchez, 1999). 45 Production per unit of biomass (P/B) In this kind o f mass-balance model, the ratio of production to biomass, P /B is assumed to equal total mortality, Z (Allen, 1971). Therefore, this production parameter was calculated for commercial exploited stocks as the total of fishing (F) and natural mortalities (M). The P /B value for non-exploited species was obtained directly from natural mortality, as estimated from the empirical equation proposed by Pauly (1980): M = K° 6 5 * L j n f " ° 2 7 9 * T ° ' 4 6 3 where K is the von Bertalanffy growth constant; L , n f is the asymptotic length expressed in centimetres and T is the average water temperature in °C. The average temperature in the upper G u l f used was 23.5°C, according to the mean SST distribution for the period 1984-2000 (Lavin et al., 2003). Appendix 9 presents the sources o f information employed for estimations of P /B for the non-fish groups included in the model. Consumption per unit of biomass (Q/B). The Q/B ratio represents the food intake by a group over a specified time period (consumption) divided by its biomass. Q/B was calculated using the holistic method proposed by Pauly et al. (1990) was employed according to the following equation: Q/B = 10.67 * 0.0313™ * W i n f " ° 1 6 8 * 1.38 p / * 1 . 8 9 H r f where TK is an expression for mean annual habitat temperature (TK=1000/T°C + 273.1); P / i s 1.0 for top predators and zooplankton feeders; and a value of zero for other feeders. W , n f is the maximum weight of the fish, estimated from the asymptotic length given by FishBase (Frose and Pauly, 2000; 2001). Kd is the food type (0 for carnivores and 1 for herbivores and detritivores). 46 Diet composition. The diet composition matrix was assembled as percentage weight or volume of the annual fraction that each prey contributes to the overall diet o f the predator (Christensen et al., 2004). Several local reports were used, but when upper G u l f data were not specifically available, values were taken from the same species from the central or northern G u l f o f California models documented previously (see appendix 9 for a complete list o f references). The diet composition matrix is presented in Appendix 2. 2.3.2.4. Sources of Catch Data. Most o f the commercial fisheries catch data used in the model were provided by the regional fisheries offices of the Institute o f National Fisheries located in San Felipe (Baja California), Puerto Peiiasco and the G u l f o f Santa Clara (Sonora). Databases for shrimps and historical landings o f totoaba were obtained from the CRTP-Ensenada (Regional Fisheries Centre, Ensenada, Baja California). For some particular invertebrates (clams, Chione spp, Tivela styultorum, snails, Strombus spp, sea cucumber, Brandtothuria impatiens, octopus, Octopus digueti, crabs, squids, Lolliguncula spp) exploited by hookah divers from Puerto Peiiasco, the information utilized was provided by Richard Cudney-Bueno and Peggy Turk from C E D O (Study Center of Deserts and Oasis, Sonora, Mexico) and from Cudney-Bueno and Turk-Boyer (1998). Fishing fleets were divided into eight sectors: offshore shrimp trawlers, offshore finfish fishery, artisanal shrimp (gillnet 5 cm), artisanal gillnet from 6-10 cm, artisanal gillnet higher than 10 cm artisanal long-liners, traps and hookah divers. A s outlined above, illegal and unreported catches were estimated for the region according to the methodology proposed by Pitcher et al. (2000 & 2005). The total extractions estimated were included in the 2000 model. Recreational catches o f San Felipe (and more recently in Puerto Peiiasco) are probably significant in this ecosystem. A guesstimate o f 5% of the official landings recorded in San Felipe was added for sharks, jacks (Caranx spp, Seriola spp), wrasses (Halichoeres spp), snappers, serranids (Paralabrax spp, Ephinephelus spp), corvinas, chanos and flounders (Paralichthys spp) in addition to the previous I U U adjusted estimates. 47 2.3.2.5. Balancing the model. Model imbalance is usually signified by thermodynamically unrealistic values of output parameters. Usually, the first iteration produced values o f ecotrophic efficiency (EE) greater than 1.0 (the ecotrophic efficiency represents the proportion of production that is consumed by predators, and cannot exceed 1.0). During the first run, twenty-two out o f the 50 groups were thermodynamically unbalanced with an average E E o f 4.66 ± 5.02 SD (range from 1.12 to 19.08). Most o f the unbalanced groups were fish (13 groups) with an average EE= 2.52 ± 1.12. Invertebrates represented the most unbalanced groups (7 groups) with an average E E = 9.67± 6.64; the other two groups out o f balance were sea lions and seabirds. The first step taken to achieve balance was to minimize cannibalism within groups (i.e., toothed cetaceans, including orcas and dolphins; sharks, snappers and others) liberating this energy to other groups (as suggested by Christensen et al. 2000). The second step was to reduce the consumption rates for those groups without local information reported; hence, the diets of their predators were adjusted in order to decrease the consumption values (shifting the consumption pressure to other preys). Since only six groups out of the 50 were unbalanced and E E ranged from 2.16 + 1.17 SE , Q/B values were reduced between 5-15 % for the following groups: shrimps, crabs, small pelagics, corvinas, totoaba and sharks. After these changes, the model was found to be in mass-balance. Final parameters employed in the 2000 model are presented in Table 6. 2.3.2.6. Uncertainty and model validation. After tuning the model, I proceeded to evaluate its consistency through the respiration to biomass ratio (R/B), which, as expected, increased with trophic levels, as actively swimming organisms have a higher R / B ratio (note that R / B does not account for specific differences in metabolic rates between marine mammals and fish groups). After I verified that the model was both balanced and consistent, the Ecoranger routine, built into the E w E software by Christensen and Walters (2004), was applied to evaluate the uncertainty 48 associated with all input parameters. Using a Bayesian approach and performing several iterations using randomly selected values and specified frequency distribution (normal distribution for this model), it was possible to minimize residuals (according to the least squares criterion). The best model fitted was selected for use with the dynamic simulations in Ecosim. The 'pedigree' routine in E w E , serves as a sensitivity analysis for documenting the effect of inputs on estimated parameters. The pedigree index (P) measures the amount o f local data used (i.e., minor uncertainty in the inputs) among the five basic categories of models: Biomass (B), Production to biomass (P/B), the ratio of consumption to biomass (Q/B), and diets and catches for each of the functional groups. The range of P is from 0 for data not rooted locally to 1.0 for data that are fully rooted in local data (Christensen et al., 2004). For the model presented in this paper, 227 categories from the 50 functional groups were analyzed and a value o f 0.64 for this index was obtained. 2.3.3. Results. The hundreds of trophic connections among the 50 model groups make it difficult and somewhat incomprehensible to represent them through a food web diagram; instead, Table 6 shows the basic parameters of the upper G u l f of California ecosystem. This synthesis represents the first step in evaluating the structure of the system, a potential attribute in resource management. The ecosystem model spans more than four trophic levels. The top predators were represented by sharks larger than 120 cm (Carcharhinus limbatus, Sphyrnia lewini, Heterodontus mexicanus) at a trophic level of 4.2 Most of the eco-groups occur at a trophic level between 2.5 and 3.5. O f these, croakers, corvinas, rays, guitarfish, flounders, snails, murex, octopus, crabs, and shrimps are mainly dependent on detritus food chains. Given the relatively large number o f groups occurring at these similar trophic levels, competitive interactions among these groups no doubt occur in addition to the predator-49 prey relationships. Trophic Level Figure 10. Trophic aggregation of the 50 groups (> 120 species) in the upper Gulf model, show that the system is largely controlled by the lower trophic levels. 2.3.3.1. Mortality: Fishing versus Predation. In terms of overall predation, the mortality imposed by fishing is the second top predator. Sharks took more biomass of vertebrates than fishing from the system (0.204 t/km2). Figure 10 illustrates the distribution of fishing mortality across the principal groups and compares this to the predation imposed by mortality for each group. Fishing pressure is spread out in all trophic levels, imposing high mortalities to top predators, but also at the bottom of the food web. The fishery is the only predator of large sharks in the upper Gulf and it removed more biomass of sharks, corvinas and chanos than natural predators. These results indicated that the fishery has a large impact on several species of the upper Gulf, including key commercial species such as shrimp, Sciaenids, scombrids, and sharks. Clearly, fishing, sharks and sea lions are major predators in the upper Gul f of California system. 50 2n • l in ing • Predators Figure 11. Comparison o f fishing mortality and predation mortality rates on individual prey groups. 2.3.3.2. Network flow analysis. Interactions among ecosystem components were evaluated using the 'mixed trophic impact' routine built into E w E (Christensen et al., 2000) that displays positive or negative changes in biomass for a group when other biomasses change. In general, the groups that cause large increases are detritus, phytoplankton, seagrasses, zooplankton and chano. A summary of the mixed trophic impact results is given in Figure 12. It is important to note the impact of detritus on commercial species such as shrimps, snails, clams, crabs, octopus, chano, corvinas and more. The top predators in the system (excluding fisheries), such as sharks, totoaba, sea lions and vaquitas, had negative impacts on intermediate groups (jacks, croakers, corvinas hakes, etc), thus releasing lower groups from predation pressure. A n increase in fishing by artisanal long-liners had a negative impact on large and medium sharks, but a positive effect on their prey, which can be explained by the reduction in predation and competition as noted above. 51 To evaluate the role of fisheries in this ecosystem, a first approach was to use this trophic impact routine. Results indicated that the fishery has a big impact on the CRD/TJGC ecosystem. The large scale sector, represented by the offshore shrimp trawlers, had a negative effect on many groups, not only affecting benthic resources, but also producing a negative cascade of impacts at intermediate and high trophic levels (Fig. 13). The large range of species caught by this non-selective fishery is an important factor in determining the health and management of this area. Figure 13 dramatically shows that the diverse small-scale fishery sector (using gillnets from 2 to 8 inches, long-lines, traps and hookah divers) imposed pressure on a wide range o f species (from sharks to crabs or snails). There is, however, a fundamental difference, in that the artisanal gear not only targets specific species or groups, but also has a very low discard rate, compared with approximately 7 kg (range from 2-10 kg) o f discarded organisms (mainly fish) per 1 kg o f shrimp by offshore trawlers. Discards represented 26% of the total catch in the system, 89% o f which originated from shrimp trawling. Moreover, the highly diverse small-scale sector, at least in this region, has the ability to fish at all trophic levels, niches and eco-groups. This imposes greater fishing mortalities on most of the groups, something that needs to be considered in evaluating the total impact o f the fisheries and management plans on the region. Summary statistics for the system are presented in Table 5, and a full description of these results (flows, consumptions and indices) w i l l be found in Chapter ILT as part of a comparison with models representing 1950 and 1980. 52 Table 5. Basic summary statistics from the network flow analysis o f the 2000 model. Parameter Value Units Sum of all consumptions 2,139.2 t /km 2 /year Sum of all exports 1,058.7 t /km 2 /year Sum of all respiratory flows 784.1 t /km 2 /year Sum of flows into detritus 1,197.7 t /km 2 /year Total system throughput 5,178.3 t /km 2 /year Net primary production 1,169 t /km 2 /year Total biomass (excluding detritus) 1,06.5 t /km 2/year Total catches 12.9 t /km 2 /year Mean trophic level of the catch 2.89 -Gross efficiency (catch/net primary production) 0.011 -2.3.4. Discussion. One of the benefits of ecosystem modelling approach is the potential to identify which areas or groups in the system need more research and understanding. In the case of the C R D / U G C ecosystem, the outcomes from the unbalanced model suggest that benthic communities (polychaetes, sea cucumbers, snails, scallops, oysters, clams, murex, sea stars, octopus, crabs and others) are the weakest link. The model revealed that the benthic groups require more information about their biology, production and consumption in order to understand their ecosystem role, including their response to human activities. The trophic imbalances of these eco-groups must be resolved by a better understanding of their roles in the food web. The mixed trophic impact analysis indicated that the fishery, as one of the three top predators in the system, had a big influence on the C R D - U G C ecosystem. The large-scale industrial fleet (mainly shrimp trawlers) not only has an impact at the lower trophic levels of benthic resources but also produces a cascade of effects that reach higher in the food 53 web, mainly through the high rate of discards which eliminates organisms that could otherwise be consumed by higher trophic levels. The diverse small-scale sector, on the other hand, fishes at practically all trophic levels, but the artisanal gear is more selective than offshore trawlers and there are almost no discards. The shrimp trawler effort has been reduced in the area, and trawlers have been banned in the core zone of the biosphere reserve since the mid 1990s. However, these regulations do not seem to be sufficient, as because the small-scale sector also has impacts. A l l sectors o f the fishery must be assessed and managed, and it is not completely accurate to label shrimp trawlers as a 'bad fishery' and the small-scale sector a 'good fishery'. The trophic impact analysis also shows the important role of detritus in the upper G u l f for all consumer organisms and for the fishery. This response could be explained by the Colorado River delivering hundreds of millions of tonnes of nutrients and organic matter every year (180 x 10 6 t/year; van Andel, 1964). This is a key factors for promoting productivity o f lower trophic levels (particularly to those groups which are linked to the detritus flows), and thus leading to higher catches. The negative impact of predation by large fish (sharks, totoaba, scombrids, and serranids) suggests 'top-down' control (Carpenter et al, 1985) in the system, but at the same time, the positive impact of benthic groups (including detritus) shows a potential for 'bottom-up' control (Carpenter et al, 1985), indicating a more realistic 'mixed ' control of the food web in the upper G u l f o f California. Comparison o f energy flows with other marine ecosystems around Mexico supports the suggestion by Morales-Zarate et al. (2004) that the Northern section of the G u l f o f California is a dynamic system in a mature stage of development. The total system throughput (sum of all flows) suggests that the upper Gu l f is a relatively small, but highly productive system. This is corroborated by observations made over decades that the upper G u l f and delta are a feeding ground and area of protection for larval and juvenile stages of 54 many fish and vertebrates (Sykes, 1937; Alvarez-Borrego, 1975; Perez-Mellado, 1980; Pedrin-Osuna et al., 2001; Campoy-Fabela, 2002). 2.3.4.1. Sensitivity analysis. The basic model parameters (B, P / B , Q/B and EE) presented in Table 6 were varied between - 5 0 % and +50% to evaluate model sensitivities. The model was insensitive (producing a response of less than 15% in the estimated parameters of other groups) to parameter changes in several groups, such as small pelagics, myctophids, wrasses, turtles, seabirds and sea cucumbers. In contrast, the model was most sensitive to groups at the top and bottom o f the food web. For example, a 50% decrease in the biomass of the sharks resulted in an increase of 32% in vaquita and 42% in sea lions. Corvinas, croakers, and crabs showed similar sensitivity to a decreased biomass o f totoaba. The greatest sensitivities are associated with zooplankton and accumulation of detritus. A 50% reduction in the P /B ratio of zooplankton produced an increase of 95% in the E E o f phytoplankton. Unfortunately, this analysis shows that the model was most sensitive to the groups with high uncertainty concerning their biology in the upper Gulf. These groups included sharks, marine mammals, benthic communities o f sessile and semi-sessile epibenthos, meiobenthos, and detritus. Also , lower trophic levels were very sensitive to changes in the simulated accumulation of detritus. 2.3.4.2. Uncertainties in the data. Weaknesses in the biomass and diet estimations have been pointed out in previous sections o f this thesis. In addition, we are unsure of the predator and prey relationships of juvenile totoaba and vaquita, perhaps the most important species of conservation concern in the U G C . Again, the model provides estimates of the role of various predators in the system (sharks, sea lions, dolphins, vaquita, serranids, scombrids and totoaba), but the diet information for some of these predators is unsatisfactory. Also, cannibalism is estimated to be less than 5% in groups such as sharks, serranids, scombrids, or Sciaenids 55 (including totoaba), but this needs to be calculated more carefully in the future with better biological data. The U G C has been recognized as an important nursery and reproductive habitat for several species (from totoaba to shrimps) and the predation pressure imposed on their eggs, larvae and juveniles has not been fully included in this work (only three groups of juveniles were considered in the model). 2.3.4.3. Concluding remarks. M y model of the upper G u l f of California and Colorado River Delta in 1995-2000 represents a synthesis o f our current knowledge o f the biota and fisheries. The results confirm that the upper G u l f is a key habitat for diversity and richness in the G u l f o f California and in Mexico. Sensitivity analysis indicated that the model was sensitive to the biomass of top predators and detritus; however, i f the biomasses of top predators are increased by more than 15%, the model does not balance. Hence, obtaining local biomass values from field surveys is a critical aspect for the future. Trophic impact analysis suggests that accurate estimations o f productivity at lower trophic levels are equally relevant. The model also highlights many o f the uncertainties concerning the biological knowledge of the area. These issues not only include all trophic levels and many of the heavily exploited species but also suggest that they have to be resolved by a better understanding of the biology. In general, the study demonstrates the benefits of exploring the role of top predators and the energy flows required to sustain their interactions in the system. It is not possible to understand the ecosystem effects of the Colorado water diversion or fishing until the interactions among the species in the food web are evaluated and understood. 56 Sharks TXT ' Totoaba Sea lions • 1 1 — > — ' — — 1 1 * — 1 1 1 Vaquita U M „ Corvinas Blue sbrirrp 0.3 -0.3 I 1^ I ^ SF$ arp *F$r EPS Figure 12. Mixed trophic impacts of the principal groups of the upper Gulf model, representing the direct and indirect impacts that a small increase in the biomass of the groups on the vertical axis would have on those on those on the horizontal axis. The black bars above the lines represent positive impacts, whereas the shaded bars are negative impacts. The impacts are relative, but are comparable between groups. See text for details. Fisheries are shown to right of vertical line. 57 Figure 13. Mixed trophic Impacts of the eight fishing fleets included in the C R D - U G C model, representing the direct and indirect impacts that a small increase in the fleets along the vertical axis would have on the species groups on the horizontal axis. The black bars above the lines represent positive impacts, whereas the shaded bars are negative impacts. Fisheries are shown to right of vertical line. 58 Table 6. Basic parameters for the upper G u l f o f California model. Bo ld numbers were parameters calculated by E w E . Group name Trophic B P/B Q/B EE level (t/km2) (year"1) (year"1) Seagrasses 1.0 0.745 15 - 0.68 Seaweeds 1.0 1.281 15.34 - 0.71 Phytoplankton 1.0 28.972 60 - 0.85 Macroalgae 1.0 1.145 60 - 0.39 Detritus 1.0 28.751 - - 0.21 Benthic meoifauna 2.0 6.796 6.7 25.00 0.95 Sea cucumber 2.0 0.069 4.1 5.47 0.46 Mulina coloradensis 2.0 0.005 1.14 23.00 0.31 Bivalves 2.0 0.0558 1.14 25.00 0.95 Zooplankton 2.1 26.327 18 60.00 0.59 Juv. Blue shrimp 2.1 0.137 12 60.00 0.95 Crabs 2.2 1.162 3.76 10.74 0.93 Small demersal fish 2.2 2.640 1.821 6.81 0.95 Gerreidae 2.2 0.747 2.13 6.97 0.96 Croakers 2.2 1.621 0.719 3.90 0.98 Sessile Epibenthos 2.3 2.354 2.7 15.00 0.94 Blue shrimp 2.3 0.583 4 28.94 0.92 Brown shrimp 2.5 0.326 4 28.94 0.42 Planktivorous birds 2.4 0.00002 9 45.00 0.00 Octopus 2.4 0.080 3.45 9.51 0.72 Rock shrimp 2.5 0.773 4 28.94 0.87 Other fishes 2.5 6.999 1.85 6.40 0.99 Corvinas 2.5 3.988 0.76 7.80 0.99 Semi-sessile 2.6 1.360 2.2 8.20 0.95 epibenthos Myctophids 2.7 1.891 2.276 5.78 0.76 Guitarfish 2.7 0.351 1.78 6.30 0.87 Small pelagics 2.7 0.001 3.76 11.20 0.99 Baleen whales 2.8 0.021 0.05 10.90 0.32 Flounders 2.8 0.331 0.881 1.74 0.94 Rays 2.8 4.345 2.192 7.20 0.65 Sea turtles 3.0 0.003 0.2 3.50 0.95 Stomatopods 3.1 0.882 4.38 15.60 0.90 Jellies 3.1 0.376 25 40.00 0.80 Chano 3.2 1.620 1.937 9.62 0.98 Squids 3.2 0.170 4.879 22.50 0.92 Grunts 3.3 0.815 1.873 6.78 0.88 Serranids 3.4 1.253 0.164 3.23 0.86 Snappers 3.4 0.196 0.76 5.40 0.56 Wrasses 3.4 0.434 1.1 7.80 0.51 Juv. Totoaba 3.5 0.002 2.8 14.00 0.95 Jacks 3.5 0.388 0.619 3.18 0.73 Pre-adult Vaquita 3.6 0.00001 6.0 40.00 0.95 Seabirds 3.6 0.001 5.4 40.00 0.93 Sharks < 120 cm 3.6 2.990 0.35 3.20 0.99 59 Table 6. Continuation. Group name Trophic level B P/B Q/B EE Vaquita 3.6 0.006 0.6 27.00 0.01 Toothed cetaceans 3.7 0.236 0.2 27.00 0.79 Sea lions 3.8 0.291 0.4 26.75 0.70 Scombrids 3.9 1.704 1.107 3.20 0.98 Totoaba 4.0 0.016 0.466 4.80 0.90 Sharks 4.2 0.159 0.278 3.00 1.00 2.4. Tuning and simulating the upper Gulf of California and Colorado River Delta ecosystem using Ecosim. It is possible to use the mass-balance solutions from the Ecopath model to produce dynamic simulations (Ecosim) that explore the direct and indirect ecological effects o f fisheries, perturbations and even scenarios o f climate effects (Walters et al, 1997; Christensen et al, 2000; Pauly et al, 2000; Walters et al, 2000; Christensen and Walters, 2004), providing potential information for management and policy (Pitcher et al, 2005). In Ecosim, the predator-prey interactions are mainly moderated by the vulnerability parameters (v) that reflect the relative time that prey are vulnerable to predation, representing the relative strength of bottom-up forcing (donor control, or density-dependent ratio, v = 1) and top-down control (predator control, or Lotka-Volterra, v=8); the v 's can also represent an intermediate state (Walters et al, 2000; Christensen and Walters, 2004). For this reason, the time o f foraging is one of the parameters to consider in the tuning of the simulation models. Ecosim uses a system of differential equations that expresses flux rates among biomass pools as a function of time varying biomass and fishing rates (for equations see Walters et al, 1997, 2000; Christensen et al, 2004). These equations are solved using an Adams-Bashford integration routine (by default) or by Runge-Kutta 4 t h order routine ( if selected; Christensen et al, 2004a). A statistical measure of goodness of fit to the time series data is measured as the sum of squared deviations (SS): this goodness of fit is generated each time Ecosim is run. 60 2.4.1. Tuning the 2000 U G C model. Biomasses predicted by the Ecosim simulations can be fitted using time series data of relative abundance (e.g., catch per unit effort C P U E ) or absolute abundance estimates (e.g., survey biomasses). This process known as 'tuning' provides adjusted models (based on stock assessment data) that can track changes in biomass that are known to have occurred in the past. The time series fitting uses either fishing effort or fishing mortality data as a driving factor for the Ecosim model runs. In the case o f the U G C / C R D model, it was necessary to estimate the fishing mortality from 1983 to 2000 for each exploited group in the upper G u l f (read as a C S V file). These estimations were obtained as a result of dividing the catches published (adding the illegal, unreported and unregulated fishing or 1TJU as a percentage estimated in Chapter n, section 2) by the National o f Institute o f Fisheries over the biomasses reported in the region from 1978 to 2003 (Perez-Mellado, 1980; Magallon-Barajas, 1988; Nava-Romo, 1994; CRIP-Ensenada, 1996; Lopez et al., 1997; Aragon-Noriega et al, 1999). In particular cases, such as totoaba, biomass was estimated by applying a V P A to the catch at length reported by Roman-Rodriguez and Hammann (1997) and its posterior conversion to age through the von Bertalanffy growth function documented by Cisneros-Mata et al. (1995). This stock assessment allows the estimation of the biomass of totoaba from 1950 to 2000, used in the present day model as well as in the 1950 and 1980 models for the upper G u l f presented in Chapter HI. Fishing mortalities drive the changes in simulated catches and biomasses in the model. Biomasses estimated by surveys and stock assessments were compared to those calculated by Ecopath during the fitting routine. A s the tuning process requires a continuous time-series of fishing mortality (in this case for 1978-2003), it was necessary to keep fishing mortality constant at the 1978 value (or to the earliest previous value) for those groups with no fishing mortality rate (no biomass estimation). 61 2.4.2. Vulnerabilities. The second component to consider in the tuning process is the vulnerability (V) in predators/prey interactions. In general, the higher the vulnerability, the more that biomass is affected by predation (or fishing). In contrast, i f prey is unable to find refuge from their predators, v w i l l be lower. In Ecosim, the vulnerabilities are set to 1.0 for bottom-up control, while v=2.0 is considered as mixed control, and values higher than 2.0 represent top-down control (Christensen et al., 2004). Based on the trophic level of each group, the vulnerabilities in the model were set initially from 1.2 to 4, and adjustments to specific groups were made after they were compared with the time-series o f biomass during the tuning process. The tuning process in the upper G u l f was also weighted by the biomasses and catches o f locally-exploited species. The major weight (50%) was put on the catches of shrimps, followed by their biomass (weighted 25% higher than the rest of the groups). Shrimps receive a high emphasis because their biology and population ecology have been the subject o f intense study in the upper G u l f for decades (Felix-Pico, 1975; Perez-Mellado, 1982; Perez-Mellado and Findley, 1985; Magallon-Barajas, 1988, CRIP-Ensenada/MP, 1996; Aragon-Noriega et al, 1999; Aragon-Noriega, 2000; Aragon-Noriega and Calderon-Aguilera, 2000, 2001; Calderon-Aguilera et al., 2002; Aragon and Garcia-Juarez, 2002; S A G A R P A 2003). The shrimp fishery is also the most important source o f income for the upper Gulf, representing probably the best-known species in the upper G u l f and probably in the entire Sea of Cortez which accounts for 50% of Mexico 's shrimp production (Garcia-Caudillo et al, 2000). Figure 14 shows the fitting for the 2000 model, where the predicted biomass of shrimps is compared with the time series of biomass observed (from surveys) in the upper G u l f [Note that fitting plots for other species are presented in Chapter IV- tuning of the 2000, 1980 and 1950 models]. 62 Figure 14. A representative run (screen-shot) from Ecosim for the upper G u l f o f California/Colorado River Delta model displaying the predicted biomasses of shrimp (line) and biomasses obtained from surveys (dots) in the upper G u l f from 1978-2003 (see text for references). 2.4.3. Ecosystem dynamics. There are many ways to model and explore interactions within a food web; in this section, two examples illustrate how the ecosystem role of a single U G C species may be explored using Ecosim (Christensen et al., 2000). The 20-year time period for simulations was chosen simply because it was considered to be of adequate duration to explore emergent changes in the system. Both examples, a simulated removal of top predators and detritus from the mass-balanced model, were used to track the changes produced in the biomass trajectories in the system, an essential step in understanding the biotic interactions and the responses of the components. In many cases, functional groups appeared to reach equilibrium after 20 years, while in other cases, the dynamics of some functional groups appeared to maintain some transient dynamics. The dynamic simulations presented in this section do not include any response to physical or environment factors. Chapter IV 63 presents an analysis of more complex disturbances in the Colorado River discharge, changes in fishing efforts and climate variability. 2.4.3.1. The impact of sharks. Figure 15 presents a simulated removal o f large sharks (black-tip, Charcharhinus limbatus, horn shark, Heterodontus mexicanus, and smooth-hound shark, Rhizoprionodon longurio) from the 2000 model, resulting in important changes at the top o f the food chain in the upper G u l f system. With a reduction of 90% in the biomass o f sharks over 20 years, the biomass of totoaba increased by 110%>. Similarly, the reduction o f mortality imposed by sharks resulted in a growth of 90% in the sea lion population. In the case o f vaquita, its biomass trajectory showed a quick recovery of 40% during the first 10 years, reaching 130% of its initial population at the end o f the simulation. Conversely, depletion of squids, jack, crabs, sardines, scombrids, (all by less than 40%), is mainly explained by increased mortality from their predators, such as sea lions, dolphins and vaquita. 64 Biomass (End/Start) 0 1 2 3 Corvinas Squids Serranids Small pelagics Scombrids Chano Toothed cetaceans Vaquita Totoaba Sea lions Figure 15. The top ten predicted biomass changes resulting from the gradual reduction of sharks by 90% during a 20-year simulation in the present-day U G C model. The figure presents increases in biomasses of various species such as sea lions, dolphins, vaquita, corvinas and totoaba (right bars) in response to the reduction of predation imposed by sharks. The decline in the biomasses of scombrids, jacks, serranids, and squids is explained as a response to their predators. Overall prey vulnerability was set at 2.5, which represents a mixture of bottom-up and top-down forces structuring the upper Gulf ecosystem. 2.4.3.2. The impact of detritus. Figure 16 shows a series of potential indirect trophic cascade effects produced by a simulated removal of detritus over 20 years. These effects were perceived in all trophic levels of the system. The biomass of lower trophic level groups such as corvinas, rays, flounders, clams and other benthic invertebrates decreased by 30-60% of their current status. Sea cucumbers declined dramatically to 10% of present population in just 20 years. The -40% depletion in the biomass of totoaba and vaquita after 20 years is mainly explained by the reduction in the abundance of their prey. This simulated removal of detritus confirmed that the historical detritus and nutrients accumulation from the 65 Colorado River is a critical attribute in the upper Gulf of California ecosystem, where 43 out of the 50 groups included in the model showed a 20% change in their original biomass. Biomass (End/Start) 0 0.2 0.4 0.6 0.8 1 1.2 1.4 Jellies Squids Sea turtles Sharks Totoaba Vaquita Blue shrimp Flounders Rays Corvinas Bivalves Juv. Shrimp Sea cucumber Figure 16. Predicted biomass changes attributable to simulated removal of detritus from the upper Gulf of California model over a period of 20 years. A l l the groups changed 20 percent or more (43 out of the 50 groups included in the model). Detritus reduction depleted lower trophic level groups, such as sea cucumbers, juvenile shrimps, clams, shrimps, and crabs. Note the dramatic reduction of sea cucumbers after the 20-year simulation. In turn, the decline in the population of sharks, totoaba and vaquita is a result of the decline of their prey. 66 2.5. Summary. The first step in answering critical questions about the environmental consequences of the huge dams along the Colorado River was the construction of a dynamic E w E model of present-day conditions. The model describes the interplay of predators, prey, and human fisheries using fifty functional groups (representing 130 species) l iving in this marine ecosystem. The results presented in this section show that the upper G u l f responds to both top-down and bottom-up changes. They underscore the role of the Colorado River as the principal source of sediment and detritus to the upper G u l f o f California. Understanding the process and interactions within this complex ecosystem, including the role of both low and high trophic level groups and the impact of fishing mortalities, can promote and support plans for conservation and management. Chapter III presents a E w E model that attempts to reconstruct past stages of the upper G u l f ecosystem and capture changes in the structure and form essential to quantify ecological impacts attributable to the diversion of the Colorado River. 67 Chapter III. Historical ecosystem models for the upper Gulf of California and Colorado River Delta. 3.1. Historical changes in the upper Gulf of California/Colorado River Delta: lessons from the past. 3.1.1. The delta of yesterday. The Gulf of California (then named the Vermillion Sea and later, Sea of Cortez), was first described during the sixteen century by Spanish expeditions (e.g. Francisco de Ulloa, 1523). Early explorers reported jaguars, mountain lions, wolves, Cimarron rams, deer, otters and beavers, a legendary abundance of birds, sea turtles, dolphins, whales and hundreds of fish and invertebrate species living in the estuarine and marine habitats which were formed where the Colorado River met the Gulf of California, resulting in the delta of the Colorado River (Knifing 1931, 1932; Sykes, 1937; Osorio-Tafall, 1943; Leopold, 1948). According to the maps, flora and fauna described by Sykes (1937) and Osorio-Tafall (1943), these pristine conditions continued until the twentieth century. As much as 70% of the Colorado River's silt load was carried to the upper Gulf (Thompson, 1968; Luecke et al., 1999). The annual overflows of the Colorado fertilized the waters and soils of the delta. In the 1870s, extensive steam navigation was developed for the commerce of cotton and sugar (Shipek, 1965). The delta's richness is further increased by the tidal action of 5m on average (up to 10 m, Lavin 1999), an unusually high ebb and flow that used to extend the tidal estuary up to 50 km upriver (Luecke et al, 1999). The delta of the Colorado, referred to as a "jaguar-infested jungle" by Aldo Leopold in the 1920s (Leopold, 1949), once covered more than 7,700 k m 2 with water with inputs of 7,000 m V and a transport of sediments at a rate that resulted in up to 50 m thick deposits in some areas (Varady et al., 2001). The sediment load from the Colorado has been estimated in the order of 160 x 106 t/year (van Andel, 1964), producing during the 68 Quaternary a cone of sediments of more than 7,700 k m 2 at the mouth of the river (Carraquiry and Sanchez, 1999). These high rates of sediment discharge interacted with the strong tidal regimen to create tidal bores (>3m) that more than once have sunk large ships operating in the upper Gulf, with the loss of several human lives (Sykes, 1937). Since the Spanish explorers, navigating the mouth of the Colorado and its channels, has been described as 'dangerous' because of the strength and magnitude of the spring tides and the tidal bores (Sykes, 1937; Osorio-Tafall, 1943; Leopold, 1949). Today, tidal bores are still a feature of the region, but observers report maximum heights of about 1 m (Carbajal et al., 1997). The interruption of the sediment discharge for almost a century appears to have suspended the natural construction and development of the Colorado delta (Meckel, 1975). Sediments deposited before the diversion of the Colorado River are now re-used by the tides, suggesting that this deltaic system is going through an erosional stage (Carraquiry and Sanchez, 1999). Humans first attempted to control the natural flow of the Colorado River during the late 1800s, shortly after, in 1922, the first Colorado River Compact was signed. Twelve years later, the colossal Hoover Dam was completed and started to store water, resulting in the formation of Lake Mead. The conclusion of the Glenn Canyon Dam in 1960 was responsible for a 96% reduction of the original river flow (Fig. 2). Since then, the fresh water delivered to the upper G u l f has been about 4% of that at the beginning of the century (Lavin, 1999). Except for years with strong precipitation events, such as ' E l N ino ' , not a single drop reaches the upper G u l f of California (Vandivere and Vorster, 1984). Figure 18 shows the discharges of the Colorado at the Southerly International Boundary (SIB) and an undepleted discharge estimate by Cohen et al (2001), revealing the magnitude of the water impounded and lost by the upper G u l f ecosystem by the series of dams along the Colorado. Although the Colorado delta can be influenced by a natural process (climatic oceanographic, geotectonic), the water diverted by the dams is now the major factor affecting the delta (Carraquiry and Sanchez, 1999). 69 Unfortunately, as the water of the Colorado River was diverted, the future of native people living in the delta and along the Colorado was ignored and they were left out of discussions of the management plans. There are several Indian communities that have lived, hunted and fished in the Colorado waters for centuries (Chemehuevis, Havasupais, Hualapais, Southern Paiutes, Southern Ute, Yumas, Mohave, Navajos and Cocopa). The Cocopas subsisted by fishing and farming in the margins and delta of the Colorado River before colonization. For Cocapas, the core zone of the present Biosphere Reserve represents the territory where their culture and fishing tradition (Bowen, 2004) developed over thousands of years. Today, they are the only people allowed by the Mexican government to fish in the core area of the Biosphere Reserve and approximately 225 out of the 700 Cocapas in the delta fish in the upper G u l f (Ballinas, 2002): 80% o f their fishing effort is concentrated on the gulf corvina (Cynoscion othonopterus), resulting in earnings of $2,000 to $3,000 per fisher/season (Moreno-Mena and Suarez-Sanchez, 2002). In 2002, the C N D H (The National Commission o f Human Rights) recommend to S A G A R P A (Agricultural and Fisheries Secretariat, Mexico) a more active participation of Cocapas in social development and policies in the upper gulf (Ballinas, 2002). 3.1.2. The delta of today. With more than 20 dams, the Colorado River Basin is now called 'the world's largest plumbing system' (Fradkin, 1981) and it is considered to be one of the most highly regulated rivers in the world (Andrews, 1991). Like many other large river deltas, such as the Amazon (Barfhem et al, 1995) or the Mississippi (Arthington and Welcomme, 1995), river impoundment of the Colorado for a century has had devastating effects not only in the wetlands, but also in the estuarine and marine habitats of the upper G u l f (Briggs and Cornelius, 1998; Brusca et al 2001; Glenn et al, 2001). 70 Figure 17. Painting from 1845 of a steamer crossing the Colorado Delta; at the bottom two people from the indigenous Cocapas community are represented as the first habitants of the region (reproduced from Sykes, 1937). From 1850 to the 1900s the Delta supported an intense commerce of cotton and sugar transported by steamers up river as far as Yuma. 1910 1920 1930 1940 1950 1960 1970 1980 1990 MM Discharge —•— Estimated undepleted discharge Figure 18. Colorado River discharge at the southerly International Boundary from 1910-1998. The black bars represent the recorded river discharge and the line reflects the undepleted discharge of the Colorado estimated by Cohen et al. (2001). 71 During the twentieth century, Colorado water reaching the G u l f was reduced by more than 95%, from an annual average o f 29 x 10 9 m 3 from 1896 to 1921 (Fradkin, 1981) to an annual average of 2.6 x 10 9 m 3 from 1984 to 1999 (Glenn et al, 1999). Some of the changes produced in the upper G u l f and its delta by the reduction of Colorado water are discussed next: 1. A n erosion process has begun in the region, and like other river deltas at risk, such as the Ni l e (Stanley and Warne, 1993), the Colorado delta has actually begun to decrease in size (Carraquiry, 1999). 2. The combination of river impoundment and diversions has reduced delta wetlands to about 5% o f their original extent (Brigghs and Cornelius, 1998), which have been reduced to about 5% of their original extent. Native forests of cottonwood and wi l low have yielded to sand and mudflats dominated by the non-native salt cedar, decreasing the habitat value of the riparian forest (Briggs and Cornelius, 1997; Luecke et al, 1999). 3. Much of the upper delta has been converted to irrigate farmland, levees and channels that have changed the area significantly (Luecke et al, 1999; Brusca et al, 2001). 4. A n alteration in the hydrography and currents of estuarine and marine habitats in the upper G u l f has occurred (Carbajal et al, 1997; Lavin and Sanchez, 1999). 5. A n increase of salinity in the lower delta has inhibited agriculture activity. The diversion of freshwater combined with the high evaporation typical of the upper G u l f has produced a permanent increase of salinity in the region, with average values changing from 23-35 before dams to current salinities that are typically in the range of 35-45 (Lavin et al, 1998; Lavin, 1999). Today, the upper G u l f is a negative estuary, where salinity increases toward the mouth of the river; this occurs throughout the year, despite the seasonally reversing temperature gradient (Lavin et al, 1998). Figure 19 72 presents a comparison between the current salinity in the upper Gulf and that estimated by Carbajal et al (1997) using a constant discharge of freshwater of 2000 m3/s. Figure 19 . Sea surface salinity values in the upper Gulf of California. Left panel shows the surface salinity (0-10 m) under present conditions. Right shows salinity distribution simulated using a constant discharge of freshwater of 2000 m3/s (Carbajal et al. 1997). Since the construction of the Hoover Dam in 1935, the hydrography of this region has been profoundly modified, and today the upper Gulf is an inverse estuarine system (with higher salinities in the mouth of the river). 3.1.3. A changing fauna. The increase in salinity has degraded both the delta conditions that are critical for spawning and the nursery grounds of hundred of species, including invertebrates, fish, reptiles, marine mammals and birds that use the delta and its marine zone. Several examples illustrate the importance of the freshwater, sediments and nutrients delivered by the Colorado River: 1. The bivalve mollusk Mulina colorandensis was once the most abundant species of clam inhabiting the delta, but today, only a small relict population survives near the river's mouth. Kowaleski et al. (1996) estimated that approximately 500 billion shells 73 make up the Delta's beaches, where 85-95% of the shells are from this species (Kowaleski et al, 2000; Rodriguez et al, 2001; Flessa et al, 2001). The population of these bivalves has been reduced from 50 clams/m2 (before dam construction) to approximately 3 clams/m2 in 2000 (Rodriguez et al. 2001). This dramatic change is mainly explained by the change in salinity discussed above (Flessa et al, 2001). 1. Blue shrimp catches (Litopenaeus stylirostris) and their post-larvae abundances are positively correlated (P<0.05) with the rate of freshwater delivered by the Colorado, (Galindo-Bect and Glenn, 2000; Aragon-Noriega and Calderon-Aguilera, 2000; Calderon-Aguilera et al, 2002). Lower salinity would increase the nursery area of this species, increasing the survival of its early life stages (Galindo-Bect and Glenn, 2000). A parallel circumstance has been reported in the southwestern Gulf of Mexico by Garcia (1991), where the recruitment of white shrimp (Peneaus setaceous) was positively correlated with river discharges, resulting in an expansion in estuarine nursery habitat. Figure 20 shows the important role of the Colorado River discharge in the production of blue shrimps in the upper Gulf. 2. It is remarkable that the Gulf corvina (Cynoscion othonopterus), an endemic Gulf of California fish that had not been seen in the upper gulf for 20 years (1970-90), returned in large numbers to the delta after the floods of E l Nino 1992-93 (Zengel et al, 1995; Cudney-Bueno and Turk, 1998). Since then, an important fishery was established in the region (Fig 21). Figure 22 shows a photograph taken in 1934 in Puerto Penasco showing the abundance of Gulf corvina and totoaba in the waters of the upper Gulf of California (Hendrickson, 1979). 74 Figure 20. Blue shrimp (Litopenaeus stylirostris) landings were significantly correlated (P<0.05) with Colorado River discharge in the same year. This relationship is explained by an increase of the nursery area and survival of larvae as result of the freshwater input, revealing the importance of the Colorado River in the ecology and economy of the upper Gulf of California. Data from Galindo-Bect and Glenn (2000). o ed u 5000 -i 4000 A 3000 A S 2000 A 1000 ~ i r 1985 1990 1995 2000 Figure 21. Historical landings of the Gulf corvina (Cynoscion othonopterus) in the upper Gulf of California. This species vanished in the upper Gulf for more than 40 years until it returned in early 1990s, activating a new fishery in the region. 75 Figure 22. Photo of Puerto Penasco, 1934 including Gulf corvinas (C. othonopterus) and totoaba (Totoaba macdonald!) taken by L . M . Huey (1934) and published by Hendrickson(1973). These cases illustrate that the Colorado River once played a major role not only supplying sediments, nutrients and organic matter, but also providing low-salinity nursery areas for larval shrimps and fish fry where euryhaline predators could not penetrate. Moreover, uncontrolled fisheries have brought several endemic species such as the giant Gulf Croaker or totoaba and the vaquita porpoise near to extinction (Cisneros-Mata et al., 1995; Roman-Rodriguez and Hammann. 1997; Jaramillo-Legorreta, 1999; D'Agrosa et al., 2000); however, the role of increased salinity in their key habitats is unknown (Glenn et al., 2001). Figure 23 displays some photos taken in San Felipe from 1926 to 1957 showing that, in the past, the upper Gulf was the home of 100 kg totoaba; today these colossal totoaba are gone. 76 Figure 23. Totoabas of nearly 100 kg caught in San Felipe the 1940s (left) and 1954 (right), confirming that huge totoabas were formerly common in the upper Gulf; today these colossal totoabas have vanished. Photos courtesy of Desert Fishes Council (www.desertfishes.org). 3.2. O r a l inputs to U G C ecosystem reconstructions: Results from interviews fishers of the U G C dur ing May-June 2003. As explained in Chapter I, the modelling work in this thesis uses the 'Back to the Future' approach that integrates knowledge of ecological, economic and social fields to reconstruct past stages of marine ecosystems. This interdisciplinary methodology may help to answer new questions about what an ecosystem was like before large commercial fisheries or any other human stress factors (i.e., dam construction). In order to understand changes occurring throughout time, the B T F approach incorporates T E K and L E K (Traditional and Local Environmental Knowledge) that provides information about the use of natural resources by the local people of the ecosystem modeled (Pitcher et al, 2005). This thesis addresses the participation of local fishers' knowledge (hereafter termed L F K ) from the communities of the upper Gulf as a method to promote participation and knowledge in the enhancement of ecosystem models and the resulting potential plans for conservation and restoration. 77 The local ecological knowledge concept can be defined as the knowledge held by a specific group of people about their local ecosystem (Nygren, 1999). Although it is challenging to incorporate T E K into scientific analysis (Johannes et al. 2001), there are several studies of terrestrial ecosystems that have combined scientific data with local and traditional knowledge in order to make better decisions in sustainable resource management (Nazarea-Sandoval, 1995; Berkes and Folke, 1998; Berkes et al., 2000, Huntington, 2000; Berkes et al, 2003). Only a few studies have addressed the potential existence and use of such knowledge in marine ecosystems. The B T F approach has proved the value of L E K in the reconstruction of marine ecosystems around the world (Western and Eastern coasts of Canada and Hong-Kong; Pitcher et al. 2005), providing critical insights of the past states of these systems. It is important to keep in mind that L E K is not a result of a systematic scientific study; its strength is in a lengthy series of local observations (Folke et al, 2003). The environmental knowledge of fishers could be very useful in countries such as Mexico, where historical records of landings and abundances before the 1970s are few or absent. The potential use of fishers' perceptions and historical anecdotes has helped to quantify important declines of sharks, turtles and big groupers in the central G u l f of California over the past 60 years (Saenz-Arroyo et al, 2005). In this thesis, significant effort has been made to incorporate the perception, historical anecdotes and environmental knowledge from fishers into the building of trophic models of former states of this rich marine ecosystem during the past 60 years. The L F K survey (Appendix 5) had two purposes: (1) to gain information about the fishers' perception of past states of the U G C , including losses of abundances, diversity and trophic structure by fishing and effects of the Colorado River Diversion; and (2) to obtain estimates of relative past abundances of non-commercial species, a critical aspect for the building of past trophic models of the U G C . 78 3.2.1. Shifting baselines syndrome. A n important element to consider in the L F K analysis is how the perception of the health of the upper G u l f o f California has changed throughout generations of fishing or by the presence of the water diverted from the Colorado River. These changes in the perception of the state of any environment from generation to generation have been described as 'Shifting Environmental Baselines' by Pauly (1995). Also , this phenomenon has been reported not just for fisheries, but also in social sectors (www.shiftinglines.org), explaining why people could be so tolerant of depletion o f resources or increasing violence, respectively. Important evidence o f shifting environmental baselines has been described in the central G u l f o f California by Saenz-Arroyo et al. (2005), where old fishers mentioned that 85% of the species had been depleted compared to just 10% by young fishers. In addition, perceived maximum size o f the largest G u l f grouper caught (Mycteroperca Jordan!) was significantly different among the three generations of fishers, ranging from 84 kg (mentioned by old fishers) to 63 kg reported by young fishers (Saenz-Arroyo et al, 2005). Comparisons of the catch records of the tuna boats in the G u l f o f Mexico with recent surveys also found that the current population of the oceanic white tip shark (Carcharhinus longimanus) is less than 1% of the density recorded by tuna fishers in the 1950s (Baum and Myers, 2004). A similar trend of depletion has been well documented for totoaba in the upper G u l f of California, where more than five decades of exploitation resulted in the collapse o f its fishery in the late 1970s and current risk of extinction (Cisneros-Mata et al, 1995; Roman-Rodriguez and Hammann. 1997). As a result of the intense fishing pressure imposed since the 1930s on totoaba, the average size of this species was apparently reduced as could be seen in the photo archives kept in San Felipe, Baja California where totoaba of 170-180 cm in length and more than 100 kg were landed in the Port from the 1920s to the 1950s. This range of size contrasts greatly with the largest totoaba of only a 79 60-80 cm length caught in the 1980s as the best catch possible in the area (announced in tourist promotional of San Felipe, www.mexfish.com). This drastic decline in the average size of totoaba documented through historical archives is shown in figure 24. The following paragraphs present anecdotal evidence from fisher interviews that supports the change in the perception of the environmental degradation of the U G C over three generations. It is particularly important to identify baseline shift in countries such as Mexico, where approximately 65% of the population is under 30 years of age (INEGI 2002), and the perception of previous health, richness and diversity of their environment is not fully visualized or understood, resulting in a society that tolerates the loss of biodiversity and which does not appreciate the conservation efforts that most be taken in order to protect or restore their ecosystems. Figure 24. Photo archives of totoaba landed in San Felipe in 1926 (left), 1956 (centre) and 1990 (right), where a reduction of abundances and mean size of this species has been observed after 60 years of intense exploitation. This dramatic change in size was used as preliminary evidence to explore the phenomenon of shifting baselines acting in the upper Gulf of California (see text for details). The left and centre photographs are courtesy of www.sanfelipe.com.mx, and the right photo was taken by Malt Quilter (1990; courtesy of www.mexfish.com). The study area for the L F K survey was the three fishing ports of the upper Gulf of California: San Felipe (Baja California), Puerto Penasco and Golfo de Santa Clara (Sonora). Data were collected applying semi-structured interviews (Huntington, 2000) with direct verbal interaction with the fishers during the field trip from April 17' to May SO 20 of 2003. The questionnaire was designed to gain information about the resources, abundances, gear and fishing sites in the region, based on the memory of fishers of different age groups. Because the fishers' use of common names for fish was different among the villages, the interview included flash cards with 55 pictures and common names of the principal species of sea mammals, birds, commercial and non commercial fish, sharks, shrimps and crustaceans reported to live in the area. The L F K questions covered the following aspects (Appendix 5 presents the interview in its English translation): • Personal information (name, age, date, location). • Fishing experience. • Fishing areas and seasons for main target species. • Estimation of largest animal caught (just for totoaba, sharks and corvinas). • Fishing gear. • Percentages of discards. • Estimations of illegal and unreported fishing. • Past abundances by decades from 1950 to 1990 for the major groups considered in the model. • Possibility of local extinctions. • Perspective of their future as fishers. • Effect of the water diversion from the Colorado River. • Impact of the Biosphere Reserve on the upper Gulf of California. Interviewees were selected by snowball sampling (Berg, 2001), a method that relies on referrals from initial subjects to generate additional subjects. This method, where individuals are selected from the population in a nonrandom manner, was appropriate because shifting baselines have recently been reported in the Central G o C (Saenz-Arroyo et al. 2005). M y survey focused on interviewing people who began to fish 40 to 60 years ago in order to obtain information about the last 50 years. The technique of snowball sampling reduces search costs but also reduces the likelihood that the sample will 81 represent a good cross section from the population (Huntington, 2000). The degree to which the sample differs from the population is also unknown (Wright, 2002). The L F K database was built from 49 interviews of fishers covering the industrial and small-scale fleets. A l l the interviews were performed individually and mainly during the afternoon, just after the fishers completed their work and met at the beach to socialize. The location of the interviews was variable; in some cases on the beach, on pangas, on trawlers or in the fishers' homes (Fig. 25). A t the beginning of each interview, the B T F concept was explained, emphasizing the need to obtain perceptions of the upper G u l f during the 1950s (just before the completion of the Glenn Canyon Dam in 1960). This resulted in younger fishers suggesting older retired colleagues to interview. They, in turn, introduced me to older fishers in their houses. Figure 26 shows the distribution of age of the fishers interviewed in each of the ports visited. The information provided by questionnaires along with any additional comments was processed to ensure anonymity, according to the requirements of the Behavioral Research Ethics Board of the University of British Columbia. The interviews lasted between 60 and 90 minutes using paper questionnaires and leaving space for the most relevant phrases, anecdotes and parts. The interviews were conducted in Spanish, and some of the relevant comments were translated into English. This fieldwork and survey were supported by the 'Cec i l and Kathleen Morrow Scholarship 2002' awarded to the author in 2003. Also , this field trip followed technical and ethical recommendations proposed by Bunce et al. (2000), which suggest that interviews be conducted in a respectful manner and minimize the disruption to people's routines. Following these recommendations, the interviews were conducted in private after the fishers accomplished their activities. Unfortunately, the native Cocopa community was not interviewed during the fieldwork. They live in a remote area of the San Luis Rio Colorado (Sonora), on a small ranch named ' E l Mayor ' which is 45 km from the main road to San Felipe, Baja California. No public transport is available and the author was not able to visit the site. 82 Figure 25. Some of the 49 fishers interviewed in the upper Gulf of California from April 17 to May 20, 2003. The semi-structured interviews took place in the fishers' houses (first picture from the left), pangas (center), trawlers (right), or on the beach (not shown). The questionnaire aimed to record the perception of the resources, abundances, gear and sites of fishing in the region based on the memory of fishers of different ages. This fieldwork was supported by the 'Cecil and Kathleen Morrow Scholarship 2002'. 25-1 20 H Puerto Penasco Gotib de Sarta Clara San Felipe • Old • Middle-aged • Yoirg Figure 26. Age distribution of the 49 fishers interviewed in the three main ports of the upper Gulf of California. The three generations of fishers from the three fishing ports were grouped: young (15-30 years old, n = 7), middle-aged (31-55, n = 24) and old (>55,n=18). 83 3.2.3. Representativeness and Validity The interviews aimed at gathering detailed information about more than 30 groups of organisms over the past 50 years, but the number and type of interviews can bias results. Each interview is unique and it is possible that, in another situation, the same person could have given slightly different answers (Flick, 1998). In addition, fishers were aware that their opinions would be published in this thesis and possibly in a scientific journal, and therefore, it is possible that they gave 'socially desirable' answers about the Colorado River diversion or increased their estimates of past abundances. Moreover, results from a small number of interviews can be biased (Yli-Pelkonen and Kohl , 2005), and it is noted that the L F K results from the upper Gulf of California are based on a relatively small number of interviews (n = 49); this is less than 5% of the fishers' population. No recreational, occasional or aboriginal fishers were considered in the L F K interviews, and the results, trends and perspectives expressed by the fishers interviewed may be different from those in other fishing camps. But, although the results may be biased (Berg, 2001) and these issues should be acknowledged for future comparison with results of other interviews in the upper Gulf, most of the trends obtained through the L F K process were shared by fishers from the different fishing camps in the region. Moreover, given the wide experience (more than 1,000 years of combined experience) and deep perspectives of the interviewees on marine resources, the results of the L F K analysis presented in this thesis can at least serve as a preliminary estimate of historical changes in the upper Gulf of California. 3.2.4. Creating time-series of relative abundances For each of the functional groups of organisms included in the interviews, an index of relative abundance compared to current status was assigned as: increasing (+1), decreasing (-1), or stable (0). These perceptions were ascribed to five decades from 1950 84 to 2000. The average relative abundance of the main l iving groups (totoaba, shrimps, corvinas, sharks, vaquita and whales) per decade was calculated from 1950 to 1990 according to a decade's perception of the L F K . A l l the fishers interviewed were considered 'experts'; therefore, no weighting by experience was applied and the abundances estimated by 'old or expert' or 'young or novice' fishers and their perceptions of abundance were equally averaged. The relative abundance time series from the interviews was converted into absolute abundance, assuming the same average and amplitude of change as the stock assessment data so that it could be incorporated into the 1950 and 1980 trophic models of the upper gulf. This process was employed only for those species without published references of their past abundances (i.e., sea mammals and non-commercial species). In the case of totoaba, a stock assessment ( V P A ) was utilized to estimate biomass in the 1950s. In the results presented in this chapter, a comparison is shown between the abundances estimated by L F K and those obtained by stock assessments ( V P A ) and historical surveys. 3.2.5. Results. 3.2.5.1. Shifting of environmental baselines in the upper Gulf of California. The majority of fishers interviewed were of the opinion that physical and biological conditions in the upper G u l f have deteriorated. The reasons for these changes were mainly explained by a catastrophic combination of decades of intense fishing and by the elimination of practically all the nutrients delivered by the Colorado River. The L F K analysis also provided a guideline for specific areas that were productive in the past, and that were used for daily fishing, but which are only visited sporadically today. One of the specific questions asked of the fishers in the L F K analysis focused on their preferred sites for fishing with special emphasis on reduction and changes in catch. According to the opinions of the 49 fishers, a mean number of 4.2 out of the 11 main fishing sites in the U G C have been depleted (Figure 27). The perception of this loss is 85 different with each generation. For example, old fishers expressed that an average of 5.6 sites (s.d. ± 2.8) has been depleted in the last 50 years; meanwhile, in contrast, middle-age fishers declared an average of 4.1 sites (s.d. ± 2.2) being depleted, and young fishers reported that, on average, only 2.7 U G C fishing sites have been reduced. These results taken from the anecdotes and opinions of three generations of fishers confirms that perspectives on the richness of the U G C have changed. The number of sites depleted that were reported by each of the 49 fishers was plotted in Figure 27. A Kruskal-Wallis test was also applied to evaluate the significance of the differences in the number of fishing areas depleted (Fig. 27). This test proved that there is a significant difference (X2 = 12.75, d.f.=2, P< 0.005) in the number of depleted sites as reported by three generations of fishers. -o Q. -o 8 n 60 C 4H tc o 4> X> E 3 C c 2 ^ 0 J Old Middle-Aged Young Figure 27. Opinions about the number of fishing areas depleted in the upper Gulf of California from 49 fishers interviews. The difference in the perception of the sites depleted among the three generation of fishers was significant (X2 = 12.75, d.f.=2, PO.005), indicating a shifting of ecological baselines in the region. The reduction of the fishing sites was explained by the fishers as a result of decades of overfishing and elimination of nutrients from the Colorado River. 86 Besides the temporal changes in the number of fishing sites described by the T F K analysis, fishers reported an important spatial change in the distribution of these fishing areas. Most fishers reported a depletion o f the fishing sites located in the East coast of the U G C (coast of Sonora), with major declines in the catches from sites like: ' E l Borrascoso', ' E l Tornillar' , ' L a Choya' , ' L a Salina' and ' E l Desemboque' (Figure 28). According to the testimony of old and middle-age fishers of Puerto Pefiasco and Santa Clara (Sonora), these sites were once rich in totoaba, sharks, sierras and chanos. In contrast, young fishers and middle-age fishers declared that these species are no longer on the East coast of the upper Gulf. In order to catch sharks, chanos and sierras, fishers need to go as far as 30-40 km west, sometimes reaching the other half of the G u l f (Fig. 28). In this case, it is worth mentioning the fishing site named ' E l Canalon' (deep Channel), located in the central and deeper waters of the U G C which, according to old and middle-age fishers, was once a very productive site for sharks, specifically from the 1960s to the 1980s (Sphyrna spp, Mustelus spp, Carcharhinus limbatus) with biomasses estimated in the range 2-4 t/km in the entire upper Gulf. Numbers appear to have declined significantly, and the site is now used only sporadically for shark fishing. In contrast, a minor depletion of the fishing sites located in the West coast of the upper G u l f was found by the L F K analysis. According to the perspective of the fishers interviewed, they prefer to fish in sites such as 'Consang Rocks ' , ' E l Coloradito', ' E l Moreno' and 'Puertecitos' which are located in the west half of the G u l f (Fig. 28). Fishers of San Felipe (Baja California, West coast of the Gulf), they very often expressed concern about the number of fishers from the East Coast (Puerto Pefiasco and Santa Clara) who are fishing for shrimps, sierras and corvinas in "their" waters. Old and some middle-age fishers from San Felipe mentioned that this conflict with the East Coast fishers is relatively 'new', beginning at the end of the 1980s. According to old fishers, they caught totoaba, turtles, sierras, and shrimps within 5-10km of their fishing camps and there was no need to travel or "invade" the fishing sites of other fishing communities, much less travel to the other half of the G u l f (4-6 hrs) in order to fish. 87 The opinions of the three generations of fishers confirmed that the waters which surround the mouth of the Colorado River (including Montague and Core Islands) are the most productive region of the upper Gulf. They identified this region as an important nursery habitat for the species exploited (i.e. shrimps). Important fishing regions reported by the fishers in the mouth of the Colorado River are: ' E l Chinero', 'Punta Sargento', ' E l Machorro' and 'Pelicano Island'. However, according to the L F K , even in this small area, productivity has declined in the last 50 years, perhaps due to intense efforts by the Mexican Government which has declared the mouth of the River as a protected area since 1955. The three generations o f fishers underscore the importance o f nutrients and fresh water delivered by the Colorado River to the survival o f larvae and to good catches of blue shrimp close to the mouth of this river. Another piece of evidence vis-a-vis the shifting of environmental baselines in the upper G u l f revealed by the L F K analysis was the decline in large totoaba caught in the region by the three generation of fishers. Old fishers reported that the biggest totoaba caught during the 1950s ranged from 130-170 cm (average of 156cm ± 19cm); middle-age fishers declared a range from 120-160cm with an average of 141 cm ± 13cm. In contrast, young fishers estimated the biggest totoaba caught was within 110 to 152 cm with a mean of 131 cm ± 19cm. Figure 29 shows the decline in the mean size of totoaba caught by the three generations of fishers during the last 50 years in the U G C . These differences in the mean size of totoabas reported by the three age-groups of fishers were significant (one-way A N O V A , p<0.05). The declining trend in the reduction of the mean size of this species, as reported by the L F K , supports the observations made by Cisneros-Mata et al. (1995) and Roman-Rodriguez and Hammann (1997) i.e. that this reduction in size could be interpreted as a response of the species to several decades of intense exploitation. 88 32 31"4 30 Montague Island Pta. Sargento AdaiiBay Sonora La Pinta estuary ^ Las Almejas LaSalina El desemboque Baja California -115 Figure 28. Comparison of the depletion of fishing sites (shaded areas) of the upper Gulf of California from 1950-1980 (upper map) and 2000 (lower map) as described by 49 local fishers. According to their perspective, the east coast of the Gulf has been depleted in the last 50 years, resulting in the best sites for fishing being located on the west coast. 89 170 160 150 140 130 120 110 100 Old Middle-Aged Young Figure 29. Mean size of the biggest totoaba caught by three generations of fishers as recalled of the upper Gulf of California (n=49). The differences in the mean size of this species reported by the three groups were significant (one-way A N O V A , P<0.05). The results of the L F K analysis suggest the existence of a shifting of ecological baselines in the upper Gulf. It was perceived that young fishers are very tolerant of the loss and collapse of their fisheries. For example, three major crises were documented in the region when totoaba (late 1970s), sharks (mid 1980s) and shrimps (early 1990s) were depleted to levels that resulted in the collapses of their fisheries. In a general sense, young fishers have the perception that old fishers always exaggerate the richness and diversity of the region and that it is not true that the upper Gulf once supported colossal populations of totoaba, sharks, Pacific sierras and other predators (suggesting equal richness in the abundances of their prey). Such a rapid shift over just a few decades in the perspective of the degradation of the environment is a very important red flag to be considered by both the Mexican government and management authorities if they wish to promote conservation of the region (through courses, seminars or other educational means) and to raise the ecological and economical value of the upper Gulf of California. 90 3.2.5.2. Estimating past abundances in the upper Gulf of California based on LFK. Figure 30 shows the estimated relative abundance for eight groups from 1950 to 1990 from information gathered from the 49 fishers interviewed. Most of the groups show a decreasing trend (some groups such as sea lions have been increasing), and according to the L F K results, over the past 60 years, fishery resources have been depleted by 86% in the upper Gulf. The results include several examples o f shifting baselines. Older fishers suggested that sea turtles (mainly the green turtle, Chelonia mydas agassisii) had populations 6-8 times higher than today's, while younger fishers thought that this species was never really abundant in the upper Gulf, and suggested an average abundance close to 1.5 times higher during 1980 than in 2000. A similar response was apparent when young and middle-age fishers were questioned about the heyday o f the totoaba (1940-60); they thought that while totoaba was 2-5 times more abundant and important for the economy of the region, it is was not as abundant as suggested by old fishers, who thought it was up to 20 times higher. Some of the young fishers declared that the old fishers always exaggerate. Sharks provided a parallel observation, where the baseline had clearly shifted across the generations. A remarkable example was attributed to the remembered history of the corvinas; while young fishers had the impression that corvinas are more abundant today than during the 1980s, this point of view contrasts with those who fished during the 1950s and the 1960s. The latter fishers reported that the high abundances of this species created a problem during the gillnet fishery of the totoaba. In contrast, in the case of the vaquita, all the fishers agreed that this species has always been rare. This observation is congruent with results presented by Ortiz (2003), who suggests that the vaquita has always been rarely seen in the upper G u l f and that (1950s) its maximum number was no greater than 2,000-5,000. 91 3H 2 -1 -0-1 • Corvinas 1 1 1 1 1 1950 1960 1970 1980 1990 Sea lions 0 i 1 1 1 1 1 1950 1960 1970 1980 1990 Whales Vaqiita o 1950 1960 1970 1980 1990 1950 1960 1970 1980 1990 Figure 30. Past abundances (relative to 2000) for eight taxonomic groups living in the upper Gulf of California (black dots). These abundances were estimated from 49 interviews with local fishers of the region. Sample size for 1950 period, 18; 1960, 33; 1970, 36; 1980, 44; 1990, 49 interviews. It seems that the rich upper Gulf of former times, when it likely supported great abundances of totoaba, corvinas, sharks, turtles and seabirds, lives in the memories of old fishers, but these memories have not traveled across the three generations to today's 92 young fishers, who do not appreciate the magnitude of the depletion in the region (Saenz-Arroyo et al, 2005). It is critical for restoration and management that that young fishers and people in Mexico visualize and understand the previous states of their ecosystems. 3.2.5.3. Agreement of L F K with I N P records. Unfortunately, there are no biomass surveys of the region conducted over long periods of time that can be used to compare the past L F K abundances from 1950-2000. In a few cases, such as the totoaba, the extent of the knowledge of this species allows us to estimate its past biomass with a V P A stock assessment. The perceived L F K abundance was converted to an absolute index by scaling the series so that the average and the amplitude of change could be compared with the V P A . Figure 31 shows the L F K trajectory for totoaba and its concordance with V P A biomasses from 1950 to 1975. The agreement between both series was measured using the Spearman Rho nonparametric coefficient of correlation; shows a significant (a=0.05) 82% concordance between the abundances estimated by interviews and the stock assessment. Figure 31 displays a visual agreement between the trends for shark, shrimp and corvina abundances suggested by the fishers and the estimated biomass and landings recorded by INP. The case of corvinas is particularly interesting, where past abundances from L F K suggest an increase since the 1990s. The corvina had been absent in the upper Gulf since the 1970s until its return in large numbers after the E l Nino of 1993 (Zengel et al, 1995; Cudney-Bueno and Turk 1998). Figure 31 presents the L F K trend of the abundance of the gulf corvina and its correspondence with the catches recorded since the beginning of the 1990s when its commercial exploitation began. Figure 32 summarizes a Spearman correlation analysis among the trajectories for the L F K and the biomasses and C P U E (kg/boat) reported by INP; the positive correlation for totoaba, corvinas and flounders is statistically significant (a=0.05) while a significant negative trend for flounders indicated that the average fisher's perception contradicts the trend of C P U E from INP data. Overall, there is a satisfactory agreement among the L F K abundances and the reported trajectories of biomasses, but more information is required in both sectors to validate these results. 93 1975 1980 1985 1990 1995 2000 1950 1960 1970 1980 1950 1960 1970 1980 1990 2000 1950 1960 1970 1980 1990 2000 Figure 31. Concordance between relative abundances estimated by the L F K and biomasses estimated by INP surveys (shrimp) or stock assessment (totoaba) from 1950 to 2000. In cases with no absolute biomasses, a relative time-series of biomasses ( C P U E , kg/boat) was used for visual agreement. Note the recovery of corvinas described by the fishers and its agreement with the reported return of this species to the upper Gulf after 30 years (Zengel et al., 1995; Cudney-Bueno and Turk 1998). 94 - 1 -0. Spearman Rho -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0. Totoaba Sharks Shrimps Corvinas Serranids Rays Flounders Small pelagics Chano Crabs Bivalves -i—i i i i i i i i i_ • • • • Figure 32. Summary of the outcome of a Spearman correlation analysis among the trajectories of perceived abundance from LFK. and the biomasses (shrimps), stock assessment (totoaba) and C P U E (sharks, corvina, serranids, rays, flounders, chano, crabs and bivalves). Only the gray bars are statistically significant (a=0.05). 3.2.5.4. Opinion of the fishers about the freshwater diversion from the Colorado River . As detailed in Chapter II, the water from the Colorado River has been diverted by consecutive dams since 1934. Analysis of the L F K data shows that the ecological impact of this water reduction is well perceived by the fishers of the upper gulf, with more than 80% considering these dams to have a negative impact on the Mexican ecosystem (Fig. 33). None of the fishers who participated in the L F K survey considered the huge dams along the Colorado River to have had positive effects on the diversity and fisheries of the upper Gulf. Less than 10% interviewed said that the dams have had no effect on productivity. 94% were younger than 55, suggesting again a generational shift of the environmental baselines in the region. 95 Positive Negative No effect Figure 33. Opinion of 49 fishers from the upper Gulf of California on the impact of the historical water diversion of the Colorado River by U.S. dams. 3.2.5.5. Opinion of the fishers about the future of their fisheries. The 49 fishers interviewed are negative and pessimistic: more than 70% of them thought that catches will be worse in the future (Fig. 34). Fishers were adamant that their sons not become fishers. Many factors influence this firm and determined answer: corruption among authorities and decision makers, a lot of illegal fishing in the area as there is no real control of the numbers of operating pangas, and low and changing market prices. These views likely reflect considerable problems for fisheries management in the region. 96 IOO-I IT. •J~J c u bt) B C w p — 60 40 20 n=49fehers Better Worse Same Does not know Figure 34. The dismal future of fisheries in the upper Gulf of California, according to 49 fishers living in the region. This pessimism is a reflection of more complex social issues (illegal fishing, corruption and lack of control of fishing effort) as well as economic issues (low and changing market prices). Another question in my interviews related to possible benefits of the establishment of the Biosphere Reserve in their community. The general perception of this protected area is positive (only 14% of the interviewees declared a negative effect) and fishers from all three communities believed that this reserve has helped the marine resources of this ecosystem. 3.2.5.6. Final remarks. A recent study of the central section of the Gul f of California reported a shifting of environmental baselines, where for example, old fishers remember an abundance four times greater of the Gulf grouper (Mycteroperca Jordan) than younger fishers, while only a few young fishers appreciate that large species of sharks, groupers and turtles have ever been abundant in the central G o C (Saenz-Arroyo et al. 2005). The L F K conducted during this thesis confirms the existence of a shifting of ecological baselines phenomenon in the upper section of the Gulf of California. 97 In the absence of baseline ecological studies of the pre-diversion period, more than a thousand collective years of experience of the 49 fishers interviewed represents a valuable source of information that may be incorporated into quantitative modelling to evaluate the possible impacts of the Colorado dams. The L F K analysis represents the only way (with a few exceptions from the fossil records) to estimate past abundances for many non-commercial species (seabirds, marine mammals, fishes and invertebrates) in the area. Table 7 presents some of the relevant anecdotes captured by the L F K interviews, giving an impression of the past richness of the region and its gradual change over 50 years to a polluted and endangered ecosystem. Overall, this L F K material raises a red flag pointing out that besides the water diversion, there are critical factors such as pollution, corruption and drug traffic. (For example, 367 dolphins, 51 sea lions, 8 whales and more than 200 seabirds died close to San Felipe after consuming N K 1 9 , a synthetic cyanide compound used for tracking drugs at night, Azuela de la Cueva et al, 1995). From the fishers' perspectives, the future of the upper G u l f of California w i l l be based on tourism. Although just two time-series of past abundances described by the fishers significantly correlated to the biomass reported, the overall agreement for the rest of the 19 species analyzed was satisfactory. Hence, the Local Ecological Knowledge from fishers is employed alongside scientific biomass surveys and fishery information to construct, tune and validate the past trophic models for this ecosystem. The following section describes the methodology and results from these models. 98 Table 7. Some relevant anecdotes from 49 fishers from the upper Gulf of California interviewed from April 17 t h to May 20 t h 2003. Source Date Quote Interview with a 83 year-old fisher from Puerto Pefiasco Interview with a 78 year-old fisher from Puerto Pefiasco Interview with a 76 year-old fisher from E l Golfo of Santa Clara Interview with a 74 year-old fisher from Puerto Pefiasco Interview with a 73 year-old fisher from San Felipe Interview with a 71 year-old fisher from E l Golfo of Santa Clara Interview with a 64 year-old fisher from El Golfo de Santa Clara Interview with a 57 year-old fisher from Puerto Peiiasco 1950 1950 1950-1960 1950-1960 1955-1960 1950-1960 1960-1970 1970 "We used to fish totoabas of 120 kg (without head and tail) and today, just 15-18 kg"; " 9 totoabas — 1 ton "; "Sometimes, we used dynamite to fish totoaba... and the rest of the fish (killed) were left floating on the beach "; "20 tonnes of totoaba with just one trough (gillnet) in the Choya region (10 minutes from Puerto Pefiasco) "; "None of the fishers work from May to October (no totoaba) and we left to rest the product". "We used our kitchens to stores totoabas, there was not enough ice in the town "; "Mountains of 10-20 tonnes (up to 6 m in height) of totoaba' heads on the beach...flies were a nightmare, but seabirds (sea gulls and pelicans) ate all that garbage"; "Before (1950s), while we were sleeping, we could hear (nights without moon), the sound of totoabas on the beach... it was beautiful". "With the running of totoaba, there were orcas entering to the upper Gulf"; "Most of the sharks (around 80%) were caught on the coast, today, we have to travel 20-30 miles "; "No more Canada geese and Monarch butterflies (in the upper Gulf) ". "Before the 1950s; there were Japanese boats hiring people (from Puerto Pefiasco) to cut heads and tails of totoabas "; "Also, they (Japanese) used the sea lion meat as bait". "Turtles were sold on the market, in the same way that chicken or pigs "; " Up to 2 tonnes of sharks per trip, 20-30 minutes from the beach (San Felipe) ". "During low tides, we could find large puddles of freshwater with turtles"; "I used to fish easy 15-20 turtles with harpoon at 10mfrom shoreline"; "Eveiy year during May was possible to fish Marlin for approximately 1.5 months". Note: no reports were found to support this anecdote. "When the water of the river (Colorado) changed, we started to fish other things (species) such as: rays, chano, corvinas"; "ln the 1960s, chano and corvinas tagging in the gillnets were problem for the totoaba fishing, covering the entire net'. "The water from the Colorado (river) affects the spawning of the shrimp, the water is the food of shrimps"; "In 1970s I caught 1 ton ofbaqueta (serranid) with 1 tank of gasoline; today I need 3 tanks for 200 kg ". 99 Table 7. Continuation. Source Date Quote Interview with a 45 year-old "Sea lions are a problem for the sierra fishing, they fisher from San Felipe 1990 eat it, so we have to use a small chinchorro (gillnet) to distract the sea lions ". Interview with a 42 year-old "We want to obey the law, but anyway, there is fisher from Puerto Pefiasco 1990 corruption. You can buy a tourism permit for $ 500 dollars ". "There is no reason to apply for a legal license for fishing, you always can bribe anybody (authorities) "; "Drugs are a horrible problem among the young fishers, I have nothing to envy to this junky generation ". Interview with a 38 year-old "All the species use the water (from the Colorado fisher from Puerto Pefiasco 1990 River) for breeding and more closed seasons are needed, at least 6-7 months for each product (species) ". Interview with a 36 year-old 1995 "Tourism is the future, it is 100% secure". fisher from Puerto Pefiasco. Interview with a 29 year-old "We fish for necessity, not for pleasure. We don't have fisher from Puerto Pefiasco 1995 life or social insurances, retirement plans or infonhabit (a low rate government loan for buying houses) ". Interview with a 27 year-old "The floating garbage is a real problem, it scares all fisher from E l Golfo de Santa 1995-2000 the fishes ". Clara Interview with a 26 year-old "Pollution is affecting and killing everything, there is fisher from Puerto Pefiasco 1995-2000 diesel everywhere "; "Tourism will help us to survive ". 100 3.3. Ecosystem models of the upper Gulf of California in 1950 and 1980. 3.3.1. Reconstructing the past. The first step in assessing the impact of Colorado River diversion on the structure and function of the upper G u l f o f California was to construct a model of its present state (Chapter II). The original intention was to build a model of the 1900s, prior to Hoover Dam construction in the early 1930's. This proved impractical, as few data for any species or groups, including L F K materials; were available before 1950, so it was decided to build two past models, one from 1950 (based on oceanographic expeditions, fossil records and L F K ) , and another from 1980 (with more accurate data on biomass and fishing mortalities). The main goal of building the 1950s and 1980s models was to run them forward to the current model (2000) to quantify possible changes in biomasses and dynamics of the species l iving in the upper G u l f after 50 years of water diversion. The present day mass-balance model described in Chapter II served as a skeleton for building these two past models. A s likely past conditions had already been considered with the need to build past models in mind, no functional groups were added and the basic input parameters (B, P / B , Q/B and diets) were modified when documentation was available. The following paragraphs describe the changes implemented. 3.3.2. Modifying P/B and Q/B ratios. The total mortality (P/B) for several of the model groups (corvinas, cabrillas, chano, hakes, guitarfish, wrasses, crabs, sea cucumbers) was lower in 1980 than in 2000 because they were fished less heavily in 1980 and 1950. However, for a few heavily depleted groups such as totoabas and sharks, the P /B values are lower in 2000 than in 1950 and 1980. Figure 30 indicates the changes made to P /B among the major groups, where positive values (percentages) represent higher fishing pressures in the past and negatives represent an increase in fishing mortality since the 1950s. Ful l explanations and sources of information employed to calculate P /B values for the 2000 model are in Chapter II. 101 Not surprisingly, the most significant changes in P /B between 1950 and 2000 related to highly exploited large fish such as sharks, totoaba and serranids (Fig. 35), and species with historical economic commercial values (shrimps). In some cases, such as marine turtles, the L F K revealed that these organisms were heavily fished and often sold in public markets in the region during the 1950s. Mortality imposed by former turtle fishing resulted in a decrease of 35% in the total mortality of sea turtles from 0.272 (1950 model) to 0.20 (2000 model). Conversely, 'Fishing down marine food web' in the G u l f of California (Sala et al., 2004) and changes in the market demand have increased mortality in species that were not exploited before the 1990s (crabs, chanos, corvinas, rays, guitarfish), so that a reduction of their P /B values for the 1950s and 1980s models was required (Fig. 35). The rate of consumption per unit of biomass per year (Q/B) may have been lower in the past due to the larger individuals present in the populations of the 1950s, and less so in the 1980's. A few small changes in Q/B were also needed in the 1950s model during the balancing process. Figure 36 shows the adjustments of Q/B for the two past models. Overall, the Q/B of only 12 groups out of 50 was modified less than 20%. 102 % change ofP/B corrpared to present P/B •100 -50 0 50 100 150 200 250 Sharks Totoaba Blueshrirrp Sharks 120 cm Sea turtles Sarranids Pre-adult Vaquita Rockshnrnp Squids Toothed cetaceans Sea lions Vaquita Seabirds Stomatopods Juv. Totoaba Myctophids Baleen vvhales Jellies Brown shrirnp S-sessile epibenth Sea cucumber Benthic msoifauna Sessile Epibenthos Mulina Bivalves Juv. of Shrimp Seag s^ses Seaweeds Macroalgae Corvinas Octopus Snappers Jacks Flounders Scontrids Crabs Croakers Small pelages Grunts Small denxrsal Other fishes Wrasses Chano Guitarfish Rays _i L J i I i I i I • 1950 • 1980 Figure 35. Change in Production to biomass ratio (P/B) of major groups in the upper Gulf of California from 1950 (white bars) and 1980 (grey bars) compared to 2000. Note P/B is equivalent to total mortality (Z). 103 % change of CYB compared to current O B -20 -15 -10 -5 10 Sea cucirrber Totoaba Toothed cetaceans Sea lions Vaquita Croakers Serranids Seabirds Flounders Sea turtles Re-adult Vaquita Grunts Guitarfish Small demersal fish Storratopods Juv. Totoaba ivyctcphids Baleen whales Crabs Squids Jellies Ftenkt. birds Ftckshrirrp Bue shrirrp Brown shrirrp S-sessil Bpibenth Benthic rreorfauna Sessil Epiberithos Mjlina Bvalves Juv. of Shrirrp Zooplankton Wasses Jacks Snappers F^ys Cther fishes Ivbrine rrcjarres Sharks <120cm Corvinas Scombrids Cctopus Chano Sharks Small pelagics • 1950 • 1980 Figure 36. Change in Q/B of major groups in the upper Gulf of California from 1950 and (white bars) and 1980 (gray bars) compared to 2000. 104 3.3.3. Diet composition The diet matrix from the 2000 U G C model described in Chapter II (Appendix 2) was used as a base for the trophic links and diets needed to built the 1950s and 1980s models (Appendix 3 and 4, respectively). This decision was taken based on the quality and quantity o f diets reported in the upper G u l f after the mid 1980s (appendix 9 presents a summary of the sources of information employed), where the 2000 diets represented the best approach to rebuild the past trophic interactions in the region. It was assumed that predator preferences have changed very little in the past 50 years and that the changes in these interactions reflect the changes in abundances of the prey (i.e., increasing their vulnerabilities to predation during high abundances and vice versa). Table 8 shows changes in the diets estimated for sharks among the three models where the reduction of totoaba ingested is explained by the documented decline of the population of this species, or the increasing consumption of sea lions is a response to their increasing abundances (according to the L F K analysis). The 'Fishing down in the food web' trend observed in the upper G u l f is also assumed to have affected top predators such as sharks. This explains the increase in the preference for small fish such as corvinas and chanos in lower trophic levels instead of the large serranids and scombrids that were common in the region during 1950s ( L F K analysis: in the opinion of local fishers). Appendices 3, 4 and 5 present the diet matrix of the 2000, 1980 and 1950 models, respectively. 105 Table 8. Estimated percentages of the prey ingested by large sharks in 2000, 1980s and 1950s in the upper Gulf of California. Last column on the right presents the percentage of changes in the preys consumed between 2000 and 1950. Prey 2000 1980 1950 Change(%) Totoaba 0.002 0.002 0.071 - 97 % Small sharks 0.09 0.22 0.17 - 47 % Serranids 0.027 0.03 0.039 -30 % Scombrids 0.25 0.27 0.33 -24 % Corvinas 0.08 0.07 0.082 - 2 % Vaquita lxlO" 5 l x l O 5 l . l x lO" 5 <1% Toothed cetaceans 0.014 0.015 0.012 + 8% Sea lions 0.09 0.08 0.06 + 50% Other fish 0.23 0.16 0.15 + 53 % Croakers 0.09 0.08 0.04 + 120% Chanos 0.12 0.07 0.04 + 200 % 3.3.4. Catch data. 1980s model. Several sources of information were used to assemble the 1980s catch. Most of the commercial landings were obtained from yearly INP reports. The final catch input in the 1980s model represents the average catch from 1978 to 1983. Shrimp, octopus, clam, sea cucumber and crab, landings were estimated from the official landings reported in San Felipe from 1977 to 1983, and provided by CRIP-Ensenada (Regional Fisheries Centre, Ensenada, Baja California), an office extension of the INP. The fishing fleet structure for this model was practically the same as for the 2000 model, but the artisanal traps and hookah diving fleets were not included because they were undeveloped in the early 1980s. Illegal and unreported catches were estimated for the major species exploited according to the analysis presented in Chapter II. Each input catch value used in the model represents the total extraction calculated for the group in the average year of 1980. 106 Discards from the industrial shrimp trawlers were considered to be of the same magnitude (10:1) and fish composition as those reported for the 2000 model. Vaquita was a special case since they were caught during the 1980s in all types of gillnets, including totoaba nets greater than 25 cm in mesh size (now prohibited by federal law), as well as legal nets for shrimps, chano, sierra, sharks and rays (ranging from 5 to 15 cm inches in mesh size; D'Agrosa et al, 1995). Based on mortality estimates at E l de Santa Clara, 13 vaquitas were killed annually in shrimp nets, 17 by the chano fishery, 7 in nets set for sharks and rays and 2 in nets for sierra and mackerel, for a total minimum annual mortality of 39 vaquitas during the early 1980s (D'Agrosa et al, 1995). Vidal (1990) calculated that at least 110 vaquitas have been caught in totoaba and shark nets since the 1970s and he estimated that 23 vaquitas/year were killed during fishing activities. 1950s model. Catch information for the 1950s model came from historical records of landings from CRIP-Ensenada, including records of shrimp and totoaba landings in the Port of San Felipe, Baja California. Information and time-series of landings (totoaba, sharks and shrimps) reported by Arvizu and Chavez (1970), Flanagan and Hendrickson (1976) and Magallon-Barajas (1987) were also considered and compared with those reported by CRIP-Ensenada. For a few species without official records of landings (sea turtles, corvinas, chanos and clams), the L F K (Local Fishers Knowledge) testimonies revealed that these species were exploited during the 1950s at a low intensity, and estimations of the catches were calculated using information from L F K interviews. According to the official records and L F K , the major species caught during the 1950s in the upper Gulf were totoaba and shrimps, followed by a few sharks and clams. According to L F K informants, local markets used to sell corvinas, flounders and sea turtles ( L F K testimonies). For the 1950 model, only four types of gear were considered: offshore shrimp trawlers, artisanal shrimp gillnets, artisanal gillnets larger than 15 cm, and fishing rods. Unreported totoaba and shrimp catches estimated in Chapter II were added to the landings, and the final catch value was incorporated into the Ecopath model. The fish 107 composition of discards from trawls was considered the same as in 2000, but the ratio was lower. Fishers reported catching approximately 3-5 kg of 'trash' (low value fauna) with each kg of shrimp. So, for the 1950s model, the ratio of discards was set as 4:1 for the offshore fleet. This assumption was based on the opinions o f local fishers that, during the last decade, it has been harder and harder to catch shrimp and each day more trawls are needed to reach the quota. They stated that "today, just garbage [low price fish] is found". Accordingly, the trash fish discards were considered lower in the 1950s than those estimated in the 2000 model. No discards were included for the other three fleets. According to the L F K analysis, no vaquitas were caught accidentally in gillnets; however, a conservative value of 0.00019 t/km (equivalent to 25 vaquitas caught per year or 0.5% o f the 5,000 population estimated) was considered more realistic and was included in the analysis. Figure 37 summarizes the changes in the landings estimated from 1950 and 1980 in comparison to those reported in 2000. 108 % change of 1980 landings compared to landings in 2000 Crabs Chano Oai ts Corvinas Scorrixids Other fishes Rock shrirrp Small pelagics Croakers Blue shrimp Flounders Brown shrimp Octopus Jacks Rays Sharks Sharks <120cm Serranids Squids Snappers Figure 37. Change in landings of major groups in the upper Gulf of California from 1980 to 2000. Data provided by the Institute of National Fisheries (INP). 109 3.3.5. Biomass change. 1980s Model . For the 1980s model, 70% of the 50 input biomass values were obtained from local surveys and published reports from direct biomass samplings in the upper Gulf using trawling surveys; 24% were calculated by Ecopath and just 6% of the biomasses were estimated using L F K modifications to the 2000 model values (sea lions, seabirds and sea turtles). Perez-Mellado (1980) and Perez-Mellado and Findley (1985) estimated the densities of sharks and rays based on a swept area method. In the case of fish groups, biomasses were calculated from surveys from 1978 to 1980 (Grande-Vidal, 1981), which reported abundances of jacks, corvinas, chano, serranids, scombrids, snappers, flounders, wrasses, grunts, and other fish. For invertebrate groups, a trawl sampling by Felix-Pico (1975) was employed to calculate the biomass of crabs, squids, jellies, blue and brown shrimps and rock shrimps (Trachypenaeus similes pacificus). Samples taken by Felix-Pico (1975) not only cover abundances of shrimp post-larvae (mainly blue shrimps), but also record standard lengths (5-40 mm). He reported an average abundance of 0.7995 post-larvae/m (originally 7,995 PL/ha). This abundance was transformed into tonnes/km using: W = 0.008 T L 3 Where, W = weight (g) T L = total length (cm) Using a T L = 2.5 cm, the larvae have an average weight of 0.125 g and so an abundance of 0.7995 larvae/m2 produces a biomass of 0.099 t/km 2. In general, the biomasses estimated from the surveys by Felix-Pico (1975); Perez-Mellado (1980) and Perez-Mellado and Findley (1985) were based on the swept area method and their conversion to t/km 2 was done according to the equation proposed 110 byGulland(1971): Biomass = [K 3 • C] / [T • S A ' V] Where: K 3 = conversion from minutes to hours C = catch (kg) T = trawling time (hours) SA = swept area (0.18 km 2/hr was estimated using a speed boat of 4 km/h and two nets of 12 m width). V = vulnerability factor (default value = 0.5). Another source of information employed to obtain past abundances was the number of mollusk shells (mainly bivalves) preserved in the southernmost portion of the Colorado River Delta in Baja California (Kowaleswski et al, 2001; Flessa et al, 2001; Rodriguez et al, 2001). These deposits of bivalves or 'cheniers' (primarily the endemic species of Mulina coloradensis) are found as far as 80 km south of the mouth of the Colorado River, and they have supported an abundance of 3 to 5 clams per square meter since the 1980's (Kowaleswski et al. 2001 and Flessa et al, 2001). The biomass value for these two groups was calculated assuming a density of 3 clams/m 2 with an average weight of 5g (without shells) distributed in a conservative area of 12 k m 2 (west side of the delta), resulting in a biomass of 0.04 t/km 2 for clams and 0.004 t/km 2 for M. coloradensis (approximately 10% of the 1980-2000 mollusk composition of these cheniers was represented by this species; Kowaleswski et al, 2001). In the case of the infaunal communities (polychaetes group), their biomass was estimated from core samples (with abundances reported in mg/m 2) taken during 1984, and published by Fernandez-Alamo (1992). Some information was available for marine mammals. Surveys by Vil la (1986) reported 18 baleen whales in the upper Gulf (mainly Golfo de Santa Clara) during 1984. Using an average weight of 30 tonnes per whale over the 4,500 k m 2 of area modeled resulted in a 111 biomass of 0.12 t/km 2 for this group. Two sporadic orcas (Orcinus orcd) were observed in the Golfo de Santa Clara in 1988 (Delgado-Estrella et al., 1994), and their biomass was incorporated into the toothed cetaceans. Dolphins biomass was assumed to be the same as that found in the 2000 model (based on the 1993 aerial surveys). Vaquita, its biomass was estimated using the 1986 survey that reported 506 vaquitas living in the upper Gulf (Vidal et al., 1999). Using an average weight of 35 kg/vaquita, results in a biomass of 0.004 t/km 2. The biomasses of sea lions were estimated using L F K based on the abundances reported for the mid 1990s and were used for the 2000 model (Chapter II). Stock assessment was used to establish the biomass of totoaba during the last five decades. A n age-structured Virtual Population Analysis (VPA) used catch-at-length data from 1986 to 1991 (Figure 38, 1,280 totoaba analyzed) reported by Roman-Rodriguez and Hammann (1997). The catch-at-length distribution was converted to catch-at-age (assuming a stable age distribution) according to the growth parameters of this species reported by Cisneros-Mata et al. (1995). This age distribution was then used to obtain past population sizes of totoaba retroactively from 1990 until 1950 through the V P A . The first step of this stock assessment was to calculate the annual natural mortality (M) using the model of Pauly (1984): Log M = -0.0066 - 0.279*Log L i n f + 0.6543* Log K + 0.4634 Log T Where the growth parameters required for this model (K=0.152 and Lj n f = 169.9 cm) were reported by Cisneros-Mata et al. (1995), and the average surface temperature of 20°C during spawning of totoaba (16-28 °C; Arvizu and Chavez 1970) resulted in a natural mortality of 0.27. The annual fishing mortality of Z = 0.414 for 1990 was reported by Cisneros-Mata et al. (1995) based on commercial catch-at-age, producing a fishing mortality of 0.141 that was used as terminal F for the V P A . Also, this figure presents the biomass estimated by the V P A , where a biomass of totoaba of 0.093 t/km 2 was calculated for 1980. 112 Zooplankton biomass was determined from 1974 oceanographic cruises (Farfan and Alvarez-Borrego 1992) that reported biomasses as high as 154 mg/m 3 (during August near the Colorado River mouth). A n average of 12 mg/m 3 (dry weight) was used for the average year. This value was converted to t/km 2 using the ratio of 15:1 between dry weight and wet weight (according to Lavaniegos-Espejo and Lara-Lara, 1990). Assuming that the reported concentration is proportional to surface area, we get 0.012 t/km2/day or 4.38 t/km2/year times the conversion factor to wet weight, producing a biomass of 65.7 XJ km for the average year. Phytoplankton biomass was incorporated into the model according to abundances in the range of 0.5-16 mg/m 3 for 1982 published by Valdez-Holguin and Lara-Lara (1987). Similar abundances of phytoplankton in the upper Gulf were found in 1986 by Millan-Nunez et al. (1999). Macrophyte biomass was calculated from a mean macrophyte net production of 47.9 mg C hr"1 m"2 in Puerto Pefiasco (Littler and Littler, 1981). In the absence of local data for sea lions, seabirds and turtles, their abundances were estimated using L F K from the 49 interviews conducted in the upper Gulf during 2002 (Chapter II presents a full description of this analysis). The abundance estimated for the 1980s by the fishers was based on the biomasses reported for these groups during the mid 1990s (Tables 5 and 6). Finally, the detritus pool was recalculated from the formula presented by Pauly et al. (1993): Log D = -2.41 + 0.954 Log PP + 0.863 Log E Where, D= detritus standing stock (gC/m ) PP = primary productivity (gC/m /year) E = Euphotic depth (m) Using the primary productivity of 2.76 ± 0.53 gC/m 2/day reported by Valdez-Holguin et 113 al. (1985) and an average depth of 20m for the euphotic zone, the estimated detritus pool amounts to 65.9 t/km 2 . Figure 39 indicates proportional changes in the biomass estimates for the 50 model groups in the 1980s and 2000s models. It is worth noting that groups represented by small and highly dynamic organisms such as sardines, anchovies and squids exhibit the largest changes in biomass during the last 20 years; these changes are as high as 15-fold. These species are characterized by large population fluctuations (fossil records have shown that the populations of these small pelagics may change because of natural causes or fishing (Baumgartner et al. 1992; Lozano-Montes, 1997). It seems that these fluctuations in the biomass estimated in both models represent a red flag for the 1980 model, indicating that more accurate information is needed for these groups. 114 gure 38. Upper panel presents the length-frequency distribution for 1,280 totoaba (25 mm intervals) collected from 1986 to 1991 in the upper Gulf of California by Roman-Rodriguez and Hammann (1997). These data were used to estimate past biomasses using (age-structured) virtual population analysis (lower graph, white circles). For comparison, the reported catch is presented in dark triangles. 115 75 C - I o (>j Octopus Small pelagics Snappers H O 3^ Squids y p p Rounders ft — 3 Totoaba S* o 5 <n> SrtxxW deirersal fish © 3 w Jacks g?- 3. Benthic meoilauna O w 3 Scombrids 3 S^> "3 wrasses -J Sj O o Qiano -j " 3 "3 Snappers — • O Sessile epibenthos -r | vO M . Grabs " , , O Vlanne mojarres-< O 3 Other fishes -EL -p O Crabs 3 *g 1 - 1 1 Seabirds -| "j "3 ~r Seaweed II 2 rt> Brown shrimp-| i—• „ Sharks O p o' Myctophids "—' 3 3 Vaquita rjj 2. 5 S-sessile epibenthos "3 "—' <s> Jellies-j £ P x Serranids* $ Q. o> S [ l a r k s 120 cm-' 3 C/5 S 5-1 w r *s Baleen whales SO & Toothed cetaceans OO Croakers 4 3" O ^ Grunts -p o Blue shrimp-2 O rm Stomatopods -i i Planktivorous birds -) 9 Mulina 3 3 *§ Bivalves H S* CJQ **• 1 - 1 2 • § 3 Sea turtles tn IX'tritus Juv. Totoaba -I Sea lions o Phytoplankton -j re ft *3 Juv. fShrirnp"3 Macroalgae -f Rock shrimp h Rock shrimp-I Pre-adultS Vaquita Corvinas H p T"' Seagrasses >< O as proportion of2000 biomass Biomass as proportion to 2000 biomass c Sea turtles Octopus Sharks<120cm Snappers StoiTHtopods Small pelagicsH Serranids Small demersal fish Jacks Wrasses Guitarfish Squids Marine mojarres Croakers Bivalves Totoaba Juv. Totoaba Flounders Vaquita 4 Chano -j Jellies Mulina Other fishes Scombrids Blue shrimp-| Brown shrimp Curvinas H Planktivorous birds Zooplankton Benthic meoifauna Croakers ^ Grunts Pink. Birds Seaweeds - | Crabs Detritus HI Baleen whales Rock shrimp Rays Seabirds Sharks Seagrasses Semi-sessil epibenthos Macroalgae Phytoplankton Toothed cetaceans Detritus Sessile epibenthos -| Sea lions 1950s Model . Building a 1950s model and running it forward to the 2000 model represents a unique opportunity to monitor how biomasses have changed throughout time in the upper G u l f as a result of natural causes, e.g., climate factors, human presence (dams and fishing) or some combination. This work represents a first attempt to go back in time to address critical question of what the upper G u l f of California was like before the 1960s when the completion of Glen Canyon Dam drastically depleted freshwater flows. However, since the earliest data for many model groups is from the 1980s, constructing a 1950s model represents a significant challenge. Despite the uncertainty and weakness of the 1950s model, it presents a framework to address questions of major ecological impact. Contributions from fisheries scientists, local fishers, historians and conservationists are welcome in order to improve this model. The 1950s model exemplifies the potential role of interdisciplinary information in fisheries science and how qualitative data from other disciplines can be used to answer biological questions, explore past states of the ecosystem and try to quantify impacts previously thought to be inaccessible. Information from different sources such as local fisher knowledge ( L F K ) , fossil records of clams (Mulina coloradensis), historic archives, oceanographic expeditions and published reports were important elements in modelling the upper G u l f as it was 50 years ago. A n intensive search during my 2003 field trip to the Northwest of Mexico, southern California and Arizona was conducted in order to obtain past abundances of the fauna in the upper Gulf. Several lists of fish records and field notes since 1947 were obtained. Major sources of information include the Fish Collection at the University of California at Los Angeles and the Marine Vertebrates Collection, Scripps Institution of Oceanography, University of California at San Diego, L a Jolla, California. These records from the late 1940s and 50s, represent the first scientific expeditions to collect benthic organisms from the upper G u l f of California. They contain a complete list, description 117 and identification of the fish collected with bottom trawling nets; however, it was not possible to estimate absolute biomasses (t/km 2) for the list o f species recorded for these two 1950 expeditions using swept area estimates because no specific information about the trawling time and speed of the vessels was given. The information contained in these collections was augmented by two other fish collections assembled during the 1970s and 1980s by the University of Arizona and Centro Ecologico de Sonora (Ecology Centre of Sonora). Trophic levels for each species in the sample list were taken from FishBase (Frose and Pauly, 2000). The analysis shows a reduction in the mean trophic level of fish since the 1950s (Fig. 40), and the trend obtained is in agreement with the results found by Sala et al. (2004). Observed changes may have resulted from intense fishing by shrimp trawlers that were in the region for more than 60 years, supporting the 'fishing down in the food web' concept described by Pauly et al. (1998a). Figure 40. Mean trophic level of the benthic fish collected during nineteen trawl surveys conducted in the upper G u l f of California since 1952. The trend indicates a reduction of the mean trophic level of benthic fauna, possibly associated with intense shrimp trawling for more than 60 years. 118 From 1967 to 1969, the Fishery Agency (now named Institute of National Fisheries, INP) conducted a series of bottom trawling surveys in the upper Gulf of California to establish the composition and abundance shrimp fishery bycatch. These trawl samples represent the earliest biomass estimates of local benthos. The biomasses of biomasses of totoaba, corvinas, groupers, cabrillas {Ephinephelus spp), sierras (Scomberomorus spp), chanos, flounders, 'other fishes' (Batistes spp, Pomadasys spp, Mugil spp, Pomacanthus spp and more), crabs, squids, octopus reported by Chavez and Arvizu in 1970 (Table 9) were to cross-validate the 1950 relative biomass estimation obtained from L F K interviews. The relative biomass estimated by L F K was converted to absolute values (t/km2) and incorporated into the 1950s model. The absolute biomass estimates of these groups after the balancing process are presented in Table 12. Table 9. 1968 biomasses from upper Gulf of California surveys (Chavez and Arvizu, 1970). These absolute abundances (t/km2) were employed to estimate relative biomass for specific groups based on the factor of conversion that represents the average change of abundance perceived from 1950 to 1970 by the 49 fishers of the upper Gulf of California interviewed during 2003. Group Absolute biomass from 1968-1969 surveys (Chavez and Arvizu, 1970). (t/km2) Factor of conversion from absolute biomass to relative biomass 1970 and 1950. (times) Relative biomass from L F K for 1950 model based on 1968-1969 surveys. (t/km2) Blue Shrimps 0.23 8x 1.86 Corvinas 0.80 15x 12.03 Flounders 0.25 9x 2.26 Chanos 0.61 15x 9.27 Crabs 0.29 9x 2.68 Squids 0.14 • l l x 1.59 Zooplankton biomass was estimated from samples collected in 1956 and 1957 by the CalCOFI expeditions run by Scripps Institution of Oceanography (La, Jolla, California) in the Gulf of California, including seven stations in the northernmost region of the Gulf. Zooplankton abundance -reported by these expeditions was based on a displacement 119 volume method (ml/1000m3). The conversion of volume displaced to dry weight (mg/m3) used was suggested by Lavaniegos and Lara-Lara (1983) (Lavaniegos-Espejo and Lara-Lara, 1990; Lavaniegos-Espejo personal communication). Table 10 shows the abundances of zooplankton collected in 1957, their conversion to dry weight and finally to biomass. A n average biomass of 78.1 Xl km 2 was estimated and incorporated into the 1950s model. Figure 36 presents a comparison of the zooplankton abundances estimated in the 1950's, 1980s and 2000s models, and shows an apparent trend of reduction in the zooplankton abundances. At the moment, there is no conclusive evidence that this decline of zooplankton in the upper Gulf is a real trend in the G o C or that it represents an overestimation of the biomass calculated in the 1950s (with a considerable SD = 34 t/km 2). However, between 1950 to 1970 and the early 1990s zooplankton levels dropped by 80% in the southern portion of the California current (off southern California) as sea surface temperatures increased, resulting in a reduction of upwelling and food supply for zooplankton (Roemmich, and McGowan. 1995). More research in the upper G o C is needed to confirm the possible decline of zooplankton and the role of climate and river inflow during the last 60 years. 120 Table 10. Relative abundance of zooplankton (cm 3/1000m 3) collected by Scripps in the CalCOFI expeditions to the Gulf of California in 1956-57. Conversion to absolute values (t/km2) according to Lavaniegos and Lara-Lara (1983). Expedition Dry weight Wet Weight Biomass mg/m3 mg/nf t/km2 April 1956 60.6 184.9 67.5 April 1957 121.2 369.6 134.9 June 1957 60.6 184.9 67.5 Aug 1957 37.9 115.6 42.2 Average 70.1 ±34.2 213.8 ± 94.3 78.1 ± 34.4 The waters of the Gulf of California were described as "fabulously rich" by expeditions conducted in 1939 and 1940 on the research vessel ' E . W . Scripps' by Scripps Institution of Oceanography, SIO (SIO, 1969). These two expeditions represented the first systematic sampling of the entire Gulf, and for the first time, phytoplankton was collected from the surface down to 60 m at intervals of 10 m (Gilbert and Allen, 1943). A n average superficial value of primary productivity 12 mgC/m 3/day (within the range 0.5 to 24.4 mgC/m 3/day reported in 1940) was used to estimate the biomass of phytoplankton for the 1950s model. This average production was converted to an integrated primary production (gC/m7day) using the factor of 57.68 reported for the shallow waters of the upper Gulf by 2 2 Zeitzchel (1969), resulting in 0.201 C/m /day that is equivalent to 73.13 t/km for the average year, value employed for the Ecopath model. A similar primary productivity of 0.27 gC/m 2/day (biomass = 98.55 t/km2) was obtained in 1960 in the Central G o C region (Angel de la Guarda Island) by SIO (Scripps Institution of Oceanography, cruise TO-60-1). Primary productivity reported by Millan et al. (1999) and abundances of zooplankton obtained by Farfan and Alvarez-Borrego (1992) were incorporated into the 2000 model (Fig. 36). 121 The biomasses of non-exploited groups such as baleen whales, dolphins, sea lions, seabirds and turtles were estimated using the relative abundances from the L F K analysis (Chapter II). These were used to modify the 1990's absolute biomasses estimated by surveys and other sources of information (Appendix 10 and 11). Vaquita biomass for the 1950s period was calculated using the average carrying capacity for this species estimated by Ortiz (2002) i.e., 5,000 vaquitas (Fig. 41), resulting in a biomass of 0.038 t/km 2. s 80 60 H 40 H ffi 20 H Scripps (1957) Zeitzschel(1968) Brinton and Townsend (1980) Valdez-Holguin et al. (1985) Farfan and Arvarez-Borrego (1992) Milan etal. (1999) 1950 1960 1970 • Zboplarkton 1980 1990 2000 M Phytoplankton Figure 41. Absolute biomass of zooplankton and phytoplankton estimated for the 1950s (based on Scripps cruises in 1957 and Zeitzchel, 1968), 1980s (Brinton and Townsend, 1980; Valdez-Holguin et al., 1985) and 2000 models from Farfan and Alvarez-Borrego (1992), and Millan et al. (1999). There is no confirmation of the apparent decline of zooplankton in the upper Gulf of California, but a similar trend has been reported in the southern California Current in the last 60 years as a response to increasing sea surface temperatures (Roemmich and McGowan, 1995). 122 3.4. Balancing the 1980 and 1950 Ecopath models. In the draft 1980 Ecopath model, seventeen out of 50 groups were unbalanced with an average E E of 6.7 ± 7.8, where 11 were fish groups, 5 invertebrates and 2 marine mammals. The higher values of E E were associated with invertebrates and benthic fish. For the 1950 model, the number of unbalanced groups was higher than that in the 1980 model. A total o f 27 groups were out of balance with E E as high as 41.8 (average E E = 14.3 ± 12.5), reflecting the higher uncertainty in this model. Both models were balanced manually with adjustments to diet matrices, reducing both cannibalism (liberating this energy to other groups) and predatory consumption (shifting the consumption pressure to other prey with lower EE) . For example, the biomasses of the groups consumed by sharks in the 1950 model (sierras, hakes, totoaba, sea lions, corvinas) were increased (by less than 20%) in order to meet the high predation mortality imposed by the abundant population o f sharks (biomass = 4.1 t/km 2). Final adjustments in the balancing process reduced consumption rates, Q/B, between 5% and 15% for some of the benthic groups (i.e., crabs, octopus, sea cucumber, semi-sessile and sessile epibenthos). The final balanced parameter values for the 1980 and 1950 models are shown in Tables 13 and 14. 3.4.1. Uncertainty in the 1980s and 1950s models. Unbalanced groups in the 1950 and 1980 models drew attention to the areas and groups in the upper G u l f o f California that are not very well understood, raising a red flag for more intense research in the benthic communities (fish and invertebrates). A more robust biology of these groups could help reduce uncertainty about the role of benthic fauna in the upper Gulf, including their groups' interactions with humans. The trophic imbalances of these groups should be resolved in future models by improving understanding of their biology rather than by solving the linear equations of the model. Once both models were balanced, their consistencies were evaluated through the respiration to biomass ratio (R/B). The general trend in these models was for R / B was to increase at higher trophic levels, as expected, where more active species are found. 123 Another element included in the Ecopath software that assists is the sensitivity analysis in the pedigree routine. The pedigree index (P) categorizes the origin o f the sources of information used to build the model, measuring the uncertainty associated according to the type of data for each of the five basic parameters of the model (Biomass, P /B , Q/B, diet composition and catches). The P index ranges from 0 for data with no local input to 1 for local data. This P index can be used as a measure of fit for Ecopath models. Analysis of 211 categories across the 50 groups in the 1980s model gave a value of 0.59. In the 1950s model, the pedigree routine reflects the higher uncertainty of the model with a P = 0.47, mainly explained by the indirect method of L F K employed to estimate the biomass of more than 20 groups. Figure 42 shows the P index calculated among the 1950, 1980 and 2000 models. The predicted biomasses from the past models are compared with those estimated from stock assessment or surveys in Chapter IV in order to evaluate discrepancies and improve their accuracy. Chapter IV also links the two past models with the 2000 model in a first attempt to quantify changes in biomasses and dynamics of the species living in the upper G u l f after 50 years of water diversion. 124 u 2 tu — P U 0.8 0.6 0.4 0.2 A 0 • • 1950 1 1980 2000 Figure 42. Pedigree indices for the 1950, 1980 and 2000 models built in the upper Gulf of California. This index categorizes the origin of the sources data for the models. It ranges from 0 for data with no local input to 1 for local data. 125 Table 11. Basic parameters for the 1980s upper Gulf of California model. Values in bold have been calculated using Ecopath. The sources of information used to build this model are presented in appendix 10. Group name T L Habitat B P/B Q/B E E % (t/km2) (year 1) (year"') Sharks 4.32 0.75 2.215 0.86 2.5 0.99 Sharks < 120 cm 3.61 1 3.405 0.78 3.0 0.99 Totoaba 3.63 0.9 0.073 0.37 4.8 0.64 Toothed cetaceans 3.87 0.95 0.250 0.20 27.0 0.78 Sea lions 3.74 1 0.27 0.40 16 0.75 Vaquita 3.84 0.8 0.004 0.60 18.0 0.00 Grunts 2.03 1 1.789 0.67 3.9 0.95 Carangids 3.48 0.95 1.045 0.65 3.2 1.00 Corvinas 2.29 1 1.931 0.58 7.2 0.95 Lutjanids 3.01 1 1.46 0.32 3.2 0.96 Scombrids 4.16 1 3.871 0.66 3.4 0.93 Groupers 3.39 1 0.391 2.87 5.3 0.73 Seabirds 4.16 0.3 0.002 5.40 40.0 0.52 Rays 2.4 1 2.12 0.64 7.2 0.54 Flounders 2.19 1 1.681 1.35 4.8 0.92 Wrasses 3.41 0.7 0.969 0.48 7.6 0.95 Chanos 2.91 1 3.596 0.72 8.6 0.95 Sea turtles 2.65 0.2 0.003 0.21 4.0 0.22 Pre-adult Vaquita 3.92 0.9 4.68E-07 8.00 40.0 0.95 Sciaenids 2.78 0.9 0.877 1.37 16.5 0.96 Gerreidae 2.15 0.85 3.128 1.20 6.6 0.87 Guitarfish 2.55 0.7 2.820 0.47 3.3 0.95 Small demersal fish 2.15 1 8.004 0.85 7.5 0.95 Other fishes 2.39 1 12.234 0.99 6.2 1.00 Octopus 2.82 0.9 1.241 3.94 8.8 0.95 Stomatopods 3.03 1 0.882 4.40 15.5 0.96 Juv. Totoaba 3.06 1 0.002 2.80 14.0 0.17 Myctophids 3.18 0.7 2.599 2.28 7.2 0.95 Baleen whales 3.24 0.2 0.12 0.05 11.0 0.79 Crabs 2.18 1 1.83 2.87 10.7 0.97 Squid 3.49 1 1.14 5.64 22.5 0.96 Small pelagics 3.24 0.9 2.673 2.59 10.3 0.99 Jellies 3.25 0.8 0.448 25.00 40.0 0.52 Plankt. birds 2.4 0.02 0.00002 9.00 45.0 0.00 Rock shrimp 2.57 0.7 0.205 4.95 28.7 0.78 Blue shrimp 2.53 1 0.63 5.98 28.7 0.62 Brown shrimp 2.58 0.9 0.153 4.00 28.7 0.72 Semi-sessile epibenthos 2.38 0.7 1.639 2.20 8.5 0.95 Sea cucumber 2 0.95 0.0358 4.10 5.6 0.95 Benthic meoifauna 2.1 1 16.646 6.70 25.0 0.95 Sessile epibenthos 2.34 0.85 4.533 2.70 15.0 0.95 126 Table 11. Continuation Group name T L Habitat B P/B Q/B E E % (t/km2) (year "') (year _ 1) M. coloradensis 2.16 0.095 0.005 1.14 23.0 0.99 Bivalves 2.15 0.055 0.048 1.15 25.0 0.92 Juv. of Shrimp 2.41 0.5 0.099 12.00 60.0 0.61 Zooplankton 2.25 1 65.71 18.00 60.0 0.85 Seagrasses 1 0.1 0.17 15.00 - 0.95 Seaweeds 1 0.2 0.38 15.24 - 0.95 Phytoplankton 1 1 44.79 60.00 - 0.92 Macroalgae 1 0.7 1.232 60.00 - 0.65 Detritus 1 1 28.751 - - 0.28 Table 12. Basic parameters for the 1950s upper G u l f of California model. Values in bold have been calculated using Ecopath. The sources of information used to build this model are presented in appendix 11. Group name T L Habitat B P/B Q/B EE % (t km 2 ) (year') (year1) Sharks 4.12 0.6 4.134 0.48 2.5 0.76 Toothed cetaceans 3.95 0.9 0.279 0.2 27 0.85 Totoaba 3.89 1 2.146 0.56 4.8 0.93 Scombrids 3.72 1 5.67 0.58 3.1 0.94 Sea lions 3.69 0.95 0.185 0.4 16 0.97 Seabirds 3.57 0.35 0.002 5.4 40 0.93 Vaquita 3.53 1 0.038 0.6 18 0.17 Sharks <120 cm 3.46 0.95 4.227 0.48 3 0.91 Sciaenids 3.32 1 1.594 5.64 22.5 0.96 Pre-adult Vaquita 3.3 1 0.000107 8 40 0.95 Groupers 3.18 1 4.102 0.87 5.2 0.95 Carangids 3.17 0.9 5.038 0.55 3.2 0.95 Baleen whales 3.17 0.3 0.441 0.05 11 0 Myctophids 2.97 0.55 4.049 2.28 7.2 0.95 Small pelagics 2.96 1 3.587 2.29 8.3 0.88 Juv. Totoaba 2.84 1 0.014 2.8 14 0.46 Lutjanids 2.8 1 13.323 0.48 3.2 0.95 Wrasses 2.77 0.7 4.668 0.36 7.6 0.95 Croakers 2.65 1 12.764 0.51 3.9 0.95 Guitarfish 2.54 0.7 3.324 0.41 3.3 0.95 Stomatopods 2.54 1 16.174 4.4 15.5 0.95 Octopus 2.51 0.9 2.254 3.5 8.8 0.95 Flaounders 2.47 1 2.261 0.78 4.8 0.7 Rays 2.35 1 4.31 0.41 7 0.86 Sea turtles 2.35 0.8 0.097 0.27 4 0.4 Grunts 2.34 0.9 6.019 0.98 16.5 0.95 127 Table 12. Continuation. Group name T L Habitat B P/B Q/B EE % (t km 2 ) (year1) (year"1) Chanos 2.33 1 9.2017 0.61 8.6 0.98 | Other fishes 2.28 1 25.8175 0.91 6.2 0.95 Corvinas 2.26 1 12 0.72 7.2 0.98 Brown shrimp 2.19 1 0.247 4.4 28.7 0.12 Rock shrimp 2.17 1 0.3202 4.95 28.7 0.31 Gerreidae 2.17 0.95 13.964 1.13 6.6 0.95 Small demersal fish 2.16 1 34.578 0.94 7.5 0.95 Plankt. Birds 2.15 0.04 0.0000605 9 45 0 Blue shrimp 2.15 1 1.86 5.98 28.7 0.26 Zooplankton 2.14 1 78.114 18 60 0.61 M. coloradensis 2.15 0.15 0.0186 1.14 23 0.01 Bivalves 2.15 0.15 0.4298 1.15 25 0.25 Sessile Epibenthos 2.14 0.85 2.6065 2.7 15 0.95 Semi-sessile epibenthos 2.05 0.9 1.7156 2.2 8.2 0.95 Crabs 2.02 1 2.6802 2.87 10.7 0.84 Juv. of Shrimp 2.02 0.6 0.053 12 60 0.95 Sea cucumber 2 0.95 0.0659 4.1 5.6 0.43 Benthic meoifauna 2 1 18.6195 6.7 25 0.95 Phytoplankton 1 1 73.167 60 - 0.94 Macroalgae 1 0.7 1.3626 60 - 0.95 Seagrasses 1 0.15 0.3333 15 - 0.95 Seaweeds 1 0.25 0.5128 15.24 - 0.95 Detritus 1 1 65.89 - - 0.46 128 Chapter IV. Connecting the past with the present: tuning the 1950,1980 and 2000 models. 4.1. Connection and fitting between the 1950,1980 and 2000 models. A n Ecopath model represents a description of an ecosystem at one particular time and it gives no evaluation of possible changes in biomass associated with differences in fishing mortalities through a period of time. Responses of the biomass pools to fishing mortality and other factors in the system may be tested using the dynamic simulator Ecosim. These dynamic simulations may be run alongside biomasses estimated from direct surveys or stock assessment data to adjust the ecosystem model to reflect changes that have been observed and documented. This process is known as 'tuning' (Christensen et al., 2004). The basic procedure and details for the tuning of the Ecopath ecosystem models were explained in Section 4 of Chapter II. 4.1.1. Points to be considered in the reconstruction of the past. One problem had to be overcome before connecting these models and simulating dynamics of the upper Gulf from the 1950s to the present. There is no reliable diet information or bycatch estimates from the 1950s. Both Ecopath and Ecosim simulations are highly sensitive to the initial diet matrices, since these determine the base predation mortality rates and the rates of effective search for prey by predators. This issue was addressed by using the 2000 Ecopath model as an initial state including its diet matrix (with the best scientific information available) and bycatch proportions as first guesses for the 1950s model. The main benefit of this approach is that (unknown) diet composition and biomasses for the 1950s model remain consistent with those implied by the 2000 model. The criteria of how well the above scheme worked was to compare the time-series of biomass from 1950 to 2000 predicted by Ecosim with those obtained from direct surveys 129 in the upper Gulf. A second approach, which involved the connection of the past and present models, was to adjust the 'feeding time' parameter in Ecosim that determines how a given group adjusts its time spent on feeding (implying vulnerability to predation) in response to changes in prey availability. L o w parameter values (0.2) did not change the overall biomasses predicted by Ecosim for the three species used in the tuning. 4.1.2. Tuning the 1950s and 1980s upper Gulf of California Ecosystem Models. To tune the 1950s and 1980s models, it was necessary to generate annual fishing mortalities for the two main species, totoaba and shrimps, that have been exploited commercially since the 1950s. This process involved calculating the fishing mortality rate imposed by trawling and gillnet fleets on these three groups in the Ecopath base year as Fjto = Yjio/Bio, where Yji0 is mean catch (1950-2000) o f group i by fleet j, and Bi0 is the mean biomass during the year (estimated by direct surveys in the area or by stock assessment). The mean yearly catch was then added to the I U U (illegal, unreported and unregulated fishing) estimations for these three species presented in chapter II. Sharks were not considered in the tuning of past models due to the lack of reliable biological and fishing information for this group. The landings recorded in the upper Gu l f (mainly San Felipe Port) for the species involved in the tuning process were provided by CRIP-Ensenada, an office extension of the INP. Information and time-series of landings (totoaba and shrimps) reported by Arv izu and Chavez (1970), Flanagan and Hendrickson (1976) and Magallon-Barajas (1987) were also considered and compared with those reported by CRIP-Ensenada. In the absence of biomass estimation, fishing mortality was held at the previous year's level (or to the earliest previous value). For totoaba, a continuous time-series of annual fishing mortality from 1950 to 2000 (read as a C S V file) was estimated from biomasses calculated from V P A (details o f this stock 130 assessment are explained in Chapter II, section 4) and from the historical records of landings reported by CRIP-Ensenada from 1940 until 1978, when this fishery was closed. After 1978, fishing mortality of this species was assumed to remain constant at the 1977 value (F=0.11). Figure 43 presents the annual fishing mortality, catch and biomass of totoaba used in the tuning process. 1950 1955 I960 1965 1970 1975 1980. Catch —&— V P A - » - F Figure 43. Annual fishing mortality from 1950 to 1978 (black circles) of the Gulf giant croaker or totoaba (Totoaba macdonaldi) in the U G C using stock assessment to estimate past biomasses from catches (solid line) reported by the CRIP-Ensenada ( V P A biomass; white triangles) until 1978 when this fishery collapsed and was banned. The time-series of fishing mortality of totoaba was one of those employed to tune and connect the past and present U G C models. In the case of blue shrimps from 1950-1965 it was necessary to use the L F K to estimate past abundance because there are no surveys or other sources of local abundance information before 1965. These L F K estimates were combined with formal scientific biomass surveys in the area from the end of the 1960s until 1995 (Felix-Pico, 1975; Perez-Mellado, 1980; Perez-Mellado and Findley, 1985; CRIP-Ensenada, 1995). Figure 131 44 represents an example of how L F K estimates (black circles) were included in the estimation of past biomasses (1950-1960). L F K was a key element in producing the time-series of annual fishing mortalities from 1950-2000 required to tune the three models of the U G C . 24 1.8 o S o -C o U 1.2 H 0.6 0.0 Blue shrimp r 1.6 1.2 0.8 0.4 i- 1- 0 1950 1960 1970 1980 1990 2000 A Biomass from LFK A Biomass from surveys —-Catch - • — F S3 Figure 44. Estimation of annual fishing mortality (circles) of blue shrimp (L. stylirostris) in the upper Gulf of California from 1954 to 2000 using catch data reported by the INP (line), and local biomass surveys (white triangles) conducted by Felix-Pico, 1975; Perez-Mellado, 1980; Perez-Mellado and Findley, 1985; and CRIP-Ensenada, 1995. Perception of the average abundances from 1950 to 1960 (black triangles) estimated from interviews with local fishers (see Chapter III). The annual fishing mortality was used to drive the fitting and connection of the 1950s, 1980s and 2000 U G C models. During this tuning process (through the fitting of the trophic models using a time series of biomass, catch and fishing mortality), each Ecosim run generated a statistical measure of the goodness of fit of the time series data employed was generated each time Ecosim was run. This goodness of fit is represented by a weighted sum of squared deviations (SS) of log biomasses from log predicted biomasses. Using this criterion, it was possible to 132 search for vulnerabilities that gave the best 'fits' of Ecosim to the time series of biomass and catch, and then selecting the best overall model with the lowest SS. Figure 45 shows the best model fitted (SS=27.42) comparing the biomass and catches observed from 1950 until 2000 of totoaba and blue shrimps with those estimated by the model fitted through the 50-year time-series of fishing mortalities estimated for both species. Shrimp B Shrimp C [•' *'" _J^' ^ Totoaba B Totoaba C h 'ft IK 1 1950 1962 1974 1986 1950 1962 1974 1986 Figure 45. Fits to time series (screen capture) for the biomass and catch of shrimps and totoaba, the two principal species exploited in the upper Gulf of California since the 1950s. Dots indicate time series data since 1950 from assessments. Lines show Ecosim simulation results for the fitting of biomasses (B) and catches (C). Residual sum of squares (SS) has a value of 27.42. This value was compared with that obtained from the tuning after environmental factors were included. Vertical axis is expressed in t/km 2. 133 4.1.3. Improving the fitted model: Climate influence. The interaction between the ocean and the atmosphere is a key element to be considered in the modelling of marine ecosystems. Changing environments can change ecosystems and these changes can be manifested by a shift in productivity, abundance and distribution o f many species (Polovina et al, 1994; Francis and Hare 1994, Hayward, 1997; Pitcher et al, 2005). In order to try to capture the complex interactions between biology and physics in the upper Gulf, an environmental process was included into the fitted model. The climate influence incorporated in the tuning was represented by a time-series from 1950-2000 of the Colorado River discharge below the Hoover Dam (Fig. 2). This environmental forcing function includes the changes in the river flow associated with strong precipitation years during E l Nino events such as 1957, 1983, 1993 and 1997 (Lavin, 1999 and Lavin et al, 2003): Changes in water and sediments delivered by the Colorado River from 1950-2000 affected only primary production. Including the environmental factor resulted in an improvement in fit, the overall SS=22.61 was 18% lower than without considering climate factors (SS= 27.42; Figure 46). This improved model including the climate series was used for the dynamic simulations and quantification of the water diversion impact in the upper G u l f of California. 134 Shrtmp B i f,i\ Shrimp C TolooD* B * 1 i * 1 (\ \ \ l v \\ * K J l •|f •I h i ' V f J * i i \ r g % HB» W« t*H t»S MC (952 WW |W» Figure 46. Improved fit of the upper Gulf of California Ecosim model (screen capture) including the Colorado River discharge from 1950-2000 as a forcing function affecting primary productivity. Taking environmental influence into account improved the model, resulting in an overall SS=22.61 compared with a SS=27.42 obtained without the climate factor (Figure 45). 4.1.4. Vulnerabilities resulting from the tuning. In Ecosim, vulnerabilities (V) are assigned to individual predator/prey relationships, indicating whether the biomass of different groups is controlled primarily by predator or by prey. In Ecosim, vulnerabilities range from 1 to co; when V takes high values ('top down'), a high proportion of the biomass is vulnerable to predation. If V is closer to 1.0 ('bottom up'), then prey have the opportunity to find refuge from predators. In the three U G C models, Vs were based both on tuning to biomass data and on the trophic level of the prey. The principle behind this is that higher trophic level organisms have been heavily depleted in ecosystems (Cheung et ai, 2002). Initially, the Vs during the fitting 135 process were allocated from 2.0 to 8.0. After each run, the V s of each group were checked against the time-series biomass data during the tuning. This methods of estimating V s produced realistic responses to changes in fishing. Figure 47 shows the average predator vulnerability ( A V P ) results o f the fitted model (including environmental influence), displaying a significant (P<0.05) linear relationship with the trophic level. This pattern is in agreement with one of the Back to Future assumptions, which states that large predators (located in higher trophic levels) have a broader spectrum of diet. These top predators (sharks, sierras, totoaba, serranids) are more generalist, and therefore require a longer time to search for food, increasing their vulnerability to predation, including the mortality imposed by fishing (T. Pitcher, 2006 pers. comm.). A second point to be considered in the pattern of A V P s found could be the decline in biomass of some key potential prey (small pelagics, cephalopods, Sciaenids and rhinobathids) since the 1970s according to surveys by Felix-Pico (1975); Perez-Mellado (1980); Perez-Mellado and Findley (1985); and CRIP-Ensenada (1995). Reduced prey abundance could have resulted in an increase in the amount of time for search, increasing their vulnerability to other predators or by fishing gear. This trend of high V s associated with large predators has been found in other ecosystem models, e.g., in British Columbia (C. Ainsworth, 2006 pers. comm.), and in less intensity in Hong Kong (W. Cheung, 2006 pers. comm.), indicating the need to explore and to quantify this assumption in the future using different trophic models of other marine ecosystems. 136 10 n APV = 2.364 TL -2.231 R 2 = 0.71 o o o > OH < 8 o < 0 2 3 4 5 Trophic Level (TL) Figure 47. Vulnerabilities resulting from the 1950s model fitted with local biomass surveys, L F K and catches recorded in the upper Gulf that generate a 50-year annual time-series of fishing mortality to fit the 1950s, 1980s and 2000 models. The linear trend observed (P<0.05) between trophic level and the average predator vulnerability is in agreement with the assumption that large predators are more generalist and they need longer time searching for food, increasing their vulnerability to predation. 4.2. Exploring ecosystem changes through time: Structure and function of the upper Gulf of California ecosystem in the last 50 years. One of the main objectives of this project is to quantify the ecological and economic impact of water diversion by Colorado River diversion on the upper Gulf of California. In order to achieve this goal, it is essential to quantify ecosystem states prior to dam construction (Gleen Canyon Dam, 1961), and then to compare them with the current state of the system. The following sections present trophic structures of the U G C ecosystem for the 1950s, 1980s and 2000 periods (fitted models) which focus on biomass flows among the 137 components and species of ecological and commercial relevance, with the purpose of identifying parameters that allow for evaluating the impact of the water diversion and fisheries in this marine ecosystem. 4 .2 .1 . T r a c k i n g f o o d w e b c h a n g e s : I n f o r m a t i o n r e v e a l e d b y t r o p h i c l e v e l s . During the last decade, trophic level structures has been used to evaluate fishing and other human effects on marine ecosystems. For example, Pauly et al. (1998) demonstrated a steady reduction of the mean trophic level of fisheries landings from 1950 to the present, suggesting that fisheries increasingly concentrate on the more abundant, fast growing fishes and invertebrates near the bottom of the aquatic food web. These findings represent examples of how trophic levels generated by Ecopath models could be used to quantify human impact on marine ecosystems. Trophic levels have been used beyond these generalizations. Thus, Pauly and Christensen (1995), who assigned trophic levels to all fish and invertebrates caught and reported in F A O global fisheries statistics 9Pauly and Christensen (1995) showed that the primary production required (also using transfer efficiencies estimated by Ecopath) to sustain the present world fisheries was much higher than previously estimated: 8% for the global ocean and between 25-35% for coastal shelves, from which 90% of the world catches originate. In the case of the upper Gulf of California, the estimation of trophic levels and the standard error of these trophic levels (the square root of the omnivory index) estimated from each of the three trophic models constructed were employed to evaluate the changes in trophic structure from 1950 to 2000. The results from the trophic levels aggregation (through the 50 years modeled) indicated a reduction in the number of groups between T L 3.5 and 4.5 (Fig. 48). Simultaneously, there were two more groups within T L = 2-2.4 in the 2000 model than in the 1950s model (Fig. 49). A reduction in the mean T L of the catch from 1950 to 2000 was also observed (Fig. 44), where a decline rate of 0.02 T L - per decade was quantified from 1950-1980 and a 0.1 T L decline per decade from 1980 to 2000. The trend in the reduction of the mean T L of the catch is corresponds to the results reported by Sala et al. (2004) that intense fishing in the entire Gulf of California has 138 reduced the mean trophic level of the landings in the past three decades; this response has also been described for other marine ecosystems associated with intense fishing, and it is known as 'Fishing down marine food webs' (Pauly et al., 1998). A second result obtained from the analysis of trophic levels aggregation revealed that almost 70% of the l iving groups considered in the 50-years of simulation could be categorized within T L of 3.0 or lower, indicating that the upper G u l f is a marine ecosystem that has been largely controlled by lower trophic levels, thus supporting the observations reported of Sykes (1937), Felix-Pico (1975) and Perez-Mellado (1980). 139 20 -, Trophic Level • 1950 • 1980 B2000 Figure 48. Trophic level aggregation of the 50 groups considered in the 1950s, 1980s and 2000 trophic models of the upper Gulf of California. The results indicate a reduction in the number of large predators (TL of 4 and 4.5) during the last 50 years. 3.2-i 1950 1960 1970 1980 1990 2000 Figure 49. Reduction of the mean trophic level of the catch estimated from the three trophic models of the upper Gulf of California revealing a decline rate of 0.02 T L per-decade was quantified from 1950-1980 and a 0.1 T L -decline per decade from 1980 to 2000. These results confirmed the Fishing Down in the Marine Food Web trend reported by Sala et al. (2004). 140 4.2.2. Tracking food web changes: The loss of biomasses. The trophic states of the upper Gulf ecosystem constructed for the 1950s, 1980s and 2000 cover important changes in biomasses over the past 50 years. For example, there was a reduction of 64% in the total estimated biomass in trophic levels 2 and 2.5 (Fig. 50), where most of the groups that occur in this range (shrimps, crabs, chanos, corvinas, croakers) are dependent on detritus food chains, emphasizing the relevance of the Colorado River as a main source of sediments and nutrients in the upper Gulf (Sykes 1937; Mackerel, 1975; Carbajal et al, 1997, Carraquiry and Sanchez, 1999). 200 -i 1 2 2.5 3 3.5 4 Trophic Level 1950 - B - 1 9 8 0 ^ ^ 2 0 0 0 Figure 50. Total biomass by trophic level of the past and present food web models of the upper Gulf of California, displaying an important reduction (64%) of the biomass of organisms located between trophic levels 2 and 2.5 from 1950-2000. Most organisms in this trophic level range dependent on detritus food chains. 141 The loss of biomass between 1950 and 2000 was experienced not just by detritivores, but also by apex predators of both invertebrates and fish groups. Figure 51 shows the changes in the biomass of the main invertebrate groups. The important role played by commercial fishing extractions in the composition and abundance of the marine biota in the upper Gulf was confirmed by major changes in the biomass of large fish predators such as totoaba and sharks during the last 50 years. Figure 52 shows the changes in abundance from 1950-2000 of the main fish groups from detritivores to top predators. Significant changes were observed. Sharks, declined b y almost 2 t/km2 between 1950 to the present day, based on the biomasses estimated from L F K interviews. This dramatic decline in sharks in the upper Gulf (Weiner, 2002) correlates with the almost complete disappearance of large sharks and mantas in the Midriff Island region located in central Gulf of California as a response to excessive fishing mortality imposed by industrial and artisanal gillnets and long-liners. More than 200,000 sharks were killed between 1985-93 in the central Gulf of California, and during the mid 1990s, some species of shark (i.e. Pacific sharp nose shark) practically disappeared from the central Gulf (Knudson, 2001; Sea Watch, 2003a,c,d). The loss of biomass of totoaba is equally dramatic (Fig. 52), a decline, described by the past and present models, which coincided with the collapse of this endemic large predator of the GoC due to a tragic combination of the degradation of their spawning area located in the estuarine waters of the upper GoC because of the diversion of the Colorado River (Arvizu and Chavez, 1970; Sanchez, 1992; Cisneros-Mata et al, 1995; Roman-Rodriguez and Hammann, 1995) with intense fishing mortality since the 1930s. The near extinction of totoaba in turn led to bankruptcy of the fishery in 1977 (Flanagan and Ffendrickson, 1987; Cisneros-Mata et al, 1997). 142 O pq Biomass of Invertebrates Octopus Blue slrrinp Squids Crabs • 1950 • 1980 Bivalves Sea cucumber 12000 Figure 51. Changes in the total biomass of invertebrates estimated from the 1950, 1980 and 2000 models of the upper Gulf of California. CO 0 Fishg-oups 1 n Chano Corvinas Scombrids Sharks Small Guitarfish Snappers Flounders Totoaba pelagics • 1950 • 1980 12000 Figure 52. Changes in the total biomass of the main fish groups (including apex predators) estimated from the food web models constructed for the 1950, 1980 and 2000 periods of the upper Gulf of California. 143 The decline and loss of biomass detected in the invertebrate and fish groups also occurred with the marine mammal groups. Figure 53 presents the changes in the biomass of the four groups of marine mammals showing a general trend of reduction over the past 50 years. The only group that shows a recovery in the abundance estimated is sea lions, no doubt due to the reduction of the predation mortality imposed by a declining population of sharks as a result of intense fishing. The direct impact of fishing on sea mammals has also been documented in the area, where the well known case of the high accidental mortality of vaquita caused by fishing nets has been defined as one of the main reasons for the decline of this endemic species (Turk, 1989; D 'Agrosa et al., 1995; Jaramillo et al., 1999; D 'Agrosa et al., 2000). Another factor is the deterioration o f the vaquita's habitat in the upper G u l f (Silber, 1990; Ortiz, 1999; Campoy, 2001; Urban, 2002). Reasons for this deterioration, include a reduction in the abundance of prey (small fish and invertebrates) combined with the extreme human impact such as pollution and the uses of killer luminescent substances during drug trafficking. In January of 1993, a total of 367 dolphins (most of them Delphinus capensis); 51 sea lions (Zalophus californicus), 8 whales (Balaenoptera acuturostrata, B. edeni and B. physalus) and hundreds of seabirds were kil led (according to tissue samples) by a synthetic cyanide substance classified as N K 1 9 , which is used as a luminous marker during drug activities at night in the upper G u l f ( P R O F E P A , 1995). The increase of sea lion biomass revealed by the trophic models is supported by the local fishers'perception ( L F K : Chapter HI). According to the survey conducted in 2003, more than 80% of the artisanal long-liners and gillnet fishers interviewed (n=49) mentioned that encounters with sea lions had caused either injuries or severe losses to their catches. These fishers have adopted a new technique of using a small net to distract and feed the sea lions while the long-line or commercial gillnet is deployed. Unfortunately, the expanding population of sea lions in the upper G u l f during the last 10-20 years has increased, thus stepping up attacks by fishers. It appears that marine mammals are steadily losing the fight. For example, from 1990 to 1993, approximately 13% of the 79 stranded mammals (most of them were sea lions) presented signs of injury from either harpoons or bullets (Delgado-Estrella et al., 1994). 144 ca i o 5 0.5 i 0.4 0.3 H S 0.2 J o . H o J Marine rremrals Baleen whales Toothed cetaceans Sea Boris Vaquita • 1950 1980 12000 F i g u r e 53. Changes in the total biomass of the marine mammals in the 1950, 1980 and 2000 food web models of the upper Gulf of California. The temporal analysis of biomasses also show the change in the ratio of benthic over pelagic biomass of the 50 groups modeled from 1950 to 2000 in the U G C . Figure 54 presents a clear trend of reduction in this ratio from 2.8 in the 1950s model to 1.5 in the 2000 model. This indicates an important depletion in the abundance of benthic groups (most of them associated with detritus food chains) since the diversion of the Colorado River at the end of the 1940s. The decline of the non-commercial benthic species in the upper Gulf is a direct result of the damage caused by dams built along the Colorado River and dramatically-reduced flows into the Gulf (Fig 54). Scientific monitoring of this ecosystem before 1975 was scarce (Brusca et al, 2001; Knudson, 2003), and the few documents published on past abundances in the region focused on commercial species such as totoaba and shrimps (Arvizu and Chavez, 1970; Magallon-Barajas, 1987; Brusca et al, 2001). However, the trophic models of the upper Gulf show important changes in 145 the abundance of those groups which are characterized by either scarce or no scientific facts because they have not been exploited in the area. They do, however, have the potential to indicate changes in the trophic structure of this ecosystem. 1950 1960 1970 1980 1990 2000 Figure 54. Decline of the ratio between benthic and pelagic biomass of the 50 groups in the trophic models built of the upper Gulf for the periods of 1950, 1980 and 2000. Loss of total biomass in the low T L s combined with the reduction in the number of large, high-TL predator groups during the last 50 years indicates the effect of two forces acting negatively in the upper Gulf: the practically null input of sediments and nutrients from the Colorado River caused by the dams (affecting detritus food chain organisms), and intense fishing of large fish (sharks, totoaba, serranids), resulting in a 'fishing down in the marine food web' response in the upper Gulf. The changes in abundance and composition of the T L aggregations during the last 50 years in the upper Gulf cannot be explained just by the water and sediment diversion imposed since the 1940s. Other factors such as fishing and probably pollution (including cyanide substances used during drug activities) that may have contributed to these changes, but the high values of E E (>0.4) of detritus found among the past and present models indicates that there has been a high demand of detritus in the system throughout the 50-years modeled. 146 4.2.3. Tracking ecosystem changes through time: Predation and Ashing mortalities. The previous section presented sharp and continuous declines in the simulated biomasses of invertebrates, fish and marine mammals in the upper G u l f from 1950-2000. These declines may have been the result o f two negative forces acting in the area: the degradation of the physical and chemical properties caused by the diversion of the Colorado River, plus high fishing mortality and by-catch pressure imposed by increased fishing in the region. This section examines how fisheries could affect the food web structure (including indirect recruitment aspects) based on the historical reconstruction of the mortality imposed by fishing activities from 1950-2000. The time-series of fishing mortalities were estimated using continuous time-series o f biomasses (results from V P A and a series of local surveys conducted by INP) and the official time-series data for catches reported by the government. The details of this 'tuning' process are fully explained in Section 4 of Chapter II (2000 model) and Section 1 o f Chapter IV (1980 and 1950 models). The impact of fishing on the structure of the upper G u l f was evaluated with the multi-species model built for the 2000 period (Chapter II). Results revealed that fishing could be identified as the second top predator (just below 0.204 t /km 2 consumed by large sharks). Also , fishing mortality, spread throughout all trophic levels, indicated that fishing has a large impact on the structure of the U G C in present day conditions. The fishing impact is evaluated in this section in a historic perspective including 50 years of biomasses predicted by Ecosim as a response to the interaction between predation and fishing mortalities. Fishing mortality rates estimated from 1950 to 2000 revealed abrupt changes in the average annual rates, ranging from 0.05 to 1.5 for chanos or from 0.18 to 1.07 for sharks (Fig. 55). The overall trend of the 50-year fishing mortality indicated an intense exploitation of large fish such as totoaba during the 1950's. This exploitation then shifted to sharks in the 1980s and then to small fish such as chanos and corvinas at the end of the 1990s (Fig. 55). This trend caused the two major fishery collapses 147 experienced in the U G C (totoabas in the 1970s and sharks at the end of the 1980's) as well as the F D M F W trend reported by Sala et al, 2004. Chano Scorrbrids Small Snappers Totoaba pelagics • 1950 11980 • 2000 Figure 55. Changes in the fishing mortality in the upper Gulf of California obtained from the past and present trophic models; revealing a shifting of the species fished from large fish such as totoaba and sharks before the 1980s to small benthic groups such as chanos and corvinas. Another emerging property of the trophic models constructed with Ecopath is the ability to quantify important interactions between predators and their prey through predation mortality. It was calculated by dividing the sum of the consumption of the group iby the other groups over the biomass of the group i (Christensen et al, 2004). Predation mortality was used to evaluate the predation imposed on top predators in the trophic structures of this ecosystem. Sharks, heavily exploited since mid 1960s, had very high rates of fishing mortality that reached their maximum values during the mid 1980s. Unfortunately, these fishing mortalities produced not only the collapse of the shark fishery in the region (Cudney-Bueno and Turk, 1998), but also the virtual disappearance 148 of Pacific sharp-nose sharks in the area (approximately 200,000 sharks were killed in central Gulf of California from 1985-91; Sea-Watch, 2003b). The decline of sharks in the upper Gulf of California by intense fishing exerted since the 1970s has resulted in the reduction of the mortality of their main prey, sea lions, and a subsequent increase in the biomass of these mammals in the region (Fig. 56). The increased biomass of sea lions predicted by the 2000 model accords with the perspective of 80% of the 49 fishers interviewed in the upper Gulf, who reported an existing conflict with sea lions, confirmed by either the consumption or damage to fish caught. This interaction has reached such a level that guns are used to scare or ki l l sea lions; from 1990 to 1993, 13% of the 79 stranded mammals (most of which were sea lions) presented signs of harpoon marks or bullet wounds (Delgado-Estrella et al, 1994). In the case of shark prey (i.e. corvinas, chanos, totoaba, etc,) that also are commercially exploited in the upper Gulf or caught accidentally during fishing activities (vaquita, sea turtles, dolphins), a reduction of their predation mortality rates by the high fishing mortality imposed on sharks (Fig. 57) did not produce the expected increases in their biomasses in the 2000 model. According to the ecology theory, it is possible to experience an increased recruitment of any species when a decline of its apex predator is experienced (Walters et al, 2000). The results from the predator mortality analysis suggest that sharks are currently at low abundance, and they impose a lower predation pressure (compared to that exerted in the 1950s) on species such as totoaba and vaquita. However, the high rate of incidental killing of these species during fishing activities must be reduced in the future in order to recover the size of the populations of these two endemic species to the levels that existed in the upper Gulf 50 years ago. One of the most revealing results of the analysis of the predation mortality was indicated by the potential indirect impact of fishing activities on the trophic interactions among commercial and non-commercial species living in the upper Gulf of California. More details of fishing effects on trophic structures of the U G C ecosystem are displayed by the trophic impact routine included in the network analysis (Section 2.4 of Chapter I V ) . 149 u >> 1 CO I an e IS en tc c o 0.8 0.6 0.4 -\ 0.2 H 0.3 0.2 0.1 ' m 1950 1960 1970 1980 1990 2000 • Predation on sea Bons •F sharks • Bbmass sea lions Figure 56. Reduction of the predation mortality of sea lions (circles, left axis) imposed by large sharks from 1950 to 2000. The increased sea lion biomass (triangles, right axis) is attributable to intense fishing imposed on sharks (squares, left axis) since the 1970s. 150 Figure 57. Reduction of the predation mortality rates of the main shark prey (1950-2000) predicted in the upper Gulf of California by the three trophic models. These changes are hypothesized as results of the intense fishing mortality (solid line) imposed on sharks in the area during 1980s, resulting in a sharp reduction of shark stocks and collapse of this fishery at the beginning of the 1990s. Due to the low values of mortality estimated for vaquita, it was necessary to plot their values on the right axis. 4.2.4. Track ing ecosystem changes through time: Network analysis. One interesting aspect of food webs obtained through Ecopath applications is that they can be used to quantify human impacts through the analysis of the trophic interactions and flows between most groups in the ecosystem, both at the lower part of the ecosystem, where the bulk of the flows occurs, and also to describe what happens at higher trophic levels, especially those that are commercially exploited. There are several ways to evaluate the functioning of an ecosystem, and when Ecopath models are used to describe the structure and function of marine ecosystems, they include the theoretical ecology proposed by Ulanowicz (1986) that quantify attributes of the system related to its development, i.e., detritus recycling and the impact of predation on the system. While this 151 theory may be in need of revision (Christensen and Walters, 2004); it does represent a way to quantify Ecopath food webs and helps to consolidate knowledge vis-a-vis ecosystem function. The Ecopath network analysis routine comprises several analyses that quantify the interactions among the groups, inter alia, impact o f predation, energy flows and recycling of detritus. The first network approach used to evaluate changes in the U G C ecosystem in the last 50 years was the analysis o f the M i x e d Trophic Impact (Christensen et al., 2000) which displays the direct and indirect trophic impacts of any group on others in the system. The magnitudes o f these impacts are relative, but they are comparable between groups. This analysis revealed that the greatest impacts are caused by the lower trophic levels; for instance, the positive impact of detritus on the upper G u l f ecosystem is noticeable not just in lower trophic groups, but also in the simulated increase of detritus produced indirectly by the increment in the biomass of high trophic levels such as that of totoaba and vaquita as a result of the rise in their prey which is associated with detritus food chains (Fig. 58). Figure 58 gives a quick overview of the importance of detritus in the productivity of the fisheries developed in the upper G u l f in the last 50 years, where the biomasses of species targeted by industrial and artisanal fisheries in the area (ranging in trophic level from 2.0 to 4.2) have been positively influenced by sediments, organic matter and detritus transported by the Colorado River into the upper Gulf. The M i x e d Trophic Impact routine (1986; Ulanowicz and Puccia, 1990), was also made it possible to evaluate the impact of industrial and artisanal fisheries. Which can be considered as a 'predator' occupying a mean trophic level of 4.1, slightly lower than that of the top predators, i.e. large sharks. For instance, the last 50 years of fishing and extensive extraction of sharks in the upper G u l f have resulted in a reduction of their impact on sharks from -0.5 to -0.23 which could be interpreted as a decline in competition and cannibalism rates among sharks (Fig. 59). 152 Figure 58. Direct and indirect impacts of simulated increases of detritus on the biomass of the main living groups considered in the 1950, 1980 and 2000 trophic models of the upper Gulf of California. These results reveal the positive impact of detritus on the system, including the theoretical increments in the biomasses of the target species fished by industrial and artisanal gears (last 6 groups on the right). The shaded bars above the x-axis represent positive impacts, whereas the black bars below are negative impacts on the biomasses. These impacts are relative, but comparable between groups. Fisheries are shown to right of vertical line. 153 Figure 59 reproduces the historical exploitation of sharks documented in the upper Gulf, where the peak of the fishery reached in the mid 1980s, is represented by the maximum value of the Mixed Trophic Impact analysis of -0.44 (long-lines) with a subsequent decrease to -0.31 in 2000 (long-lines). In general, sharks were positively impacted by groups such as sea lions, totoaba, corvinas and serranids, which have been reported as the main prey of shark species in the region (Perez-Mellado, 1980; Compagno et al, 1995; Compagno, 1999; Morales-Zarate et al, 2004). In the case of the endemic totoaba, the results reveal an important reduction in the predation by sharks (>80%) since the 1980s (Fig. 60). 1950 1980 2000 Long-lines • Gillnet (>10cm) 1 ailnet(lOcm) ailnet(5cm) Finfish Trawlers 1 Detritus 1 Squids 1 Scorrbrids D Serranids Corvinas D Sea lions • Totoaba Sharks 1 -0.5 -0.25 O 0.25 Relative irrpact -0.5 -0.25 D • • I • I 0.25 -0.5 -0.25 0 0.25 Figure 59. Results from the Mixed Trophic Impact routine displaying the impact of the main groups on sharks, including fishing activities in the last 50 years in the U G C ecosystem. Black bars represent negative impacts (decrease in the biomass of sharks), whereas shaded bars are positive impacts (increasing in the biomass of sharks) as a result of the increment in the biomasses of the groups or fishing mortality (fishing gears) in the y-axis. In general, the negative impact of the long-line fishing on sharks during the last 50 years reproduces the trend of the catches observed for the shark fishery in the area. Fisheries are shown above the horizontal line. 154 Figure 60 indicates the negative impacts of artisanal gillnets on the biomass of totoaba, considered to be the second highest negative impact on the abundances of totoaba in the 1950s (just behind the predation by sharks). This figure also indicates the trend of declining totoaba culminating in collapse of this fishery at the end of the 1970s (Cisneros-Mata et al., 1995; Pedrin-Osuna et al., 2001). The positive impacts of long-line fishing on the biomass of totoaba observed in the 1980 and 2000 models are explained by the high fishing mortality imposed since the 1970s on its main predators, i.e., sharks. 1950 Long-liners • Qllnet (>10cm) OOlnet(lOcrri) Cillnet(5cm) 1 Trawlers 1 Detritus | | Meoifauna D Small pelagics 0 Squids • Chints 1 1 Scorrbrids • Serranids D Sharks | i -0.4 -i i i 0.2 0 0.2 1980 • • I 2000 Z ] I • • • Relative irrpact -0.4 -0.2 0 0.2 0.4 -0.4 -0.2 0 0.2 0.4 Figure 60. Trophic impacts on totoaba by the fisheries and main groups in the U G C ecosystem in the last 50 years, revealing the negative effect (black bars) caused by intense gillnet fishing. Also it is worth noting the reduction in predation by sharks due to their high mortality imposed by long-line fishing during the 1970' and 1980s. See text for details. Fisheries are shown above the horizontal line. 155 In the case of the endemic populations of vaquita, the analysis shows a reduction in the predation mortality imposed by sharks in the 1980 and 2000 models, l ikely due to overexploitation of shark stocks by longliner fisheries. It is worth noting that the negative impact of gillnet fishing on the biomass of these mammals was higher than predation by sharks in the 1950s model, suggesting that by-catch of vaquita has occurred since the beginning of the fishery. The reduction of the impact of gillnets on vaquita in the 1980 and 2000 models could be explained by their legal protection in the upper Gulf, and the use of devices to avoid their incidental ki l l ing, plus the fact that the upper G u l f o f California has been declared as a Biosphere Reserve (Diario Oficial, 1993; Jaramillo-Legorreta, 1999; Ortiz, 1999; Brusca et at., 2001). Expected positive impacts on the biomass of vaquita (shaded bars in Figure 61) by the groups that they feed on species such as crabs, corvinas, croakers and grunts were observed through the 50 years o f trophic modelling in the upper Gulf. It is significant that the main preys of vaquita are species dependent on detritus food chains, indicating the relevance of the sediment and detritus delivered by the Colorado River. 156 1950 1980 2000 Long-liners Gillnet (>10cm) Gillnet (10cm) Gillnets (5cm) Trawlers Grabs* Stomatopods Octopus Grunts Serranids Qoviras G-oakers Sharks Squids • • D U • • • • • • D I D • -0.8 -0.6 -0.4 -0.2 0 02 -0.8 -0.6 -0.4 -0.2 0 0.2 -0.8 -0.6 -0.4 -0.2 0 02 Rehtive impact Figure 61. Results from the Mixed Trophic Impact routine indicating both the impact of the main groups and fishing on vaquita during the last 50 years in the U G C ecosystem and the negative effect of gillnet gear on the biomass of these endemic marine mammals. Results also indicate the high predation that vaquita suffered from sharks during the 1950s. Intense fishing exerted on sharks during the 1980s no doubt caused further reduction of this predation. The shaded bars represent the positive impacts of groups such as octopus, crabs, croakers, and grunts that are common prey of vaquita. Fisheries are shown above the horizontal line. While the network analysis does not produce dynamic or quantitative results, and cannot predict how biomass will change with time or with fishing mortality, it does provide valuable information about the structure of the ecosystem, displaying, in a snapshot, which parts of the ecosystem play a major role. In this section, some network attributes (Ulanowicz, 1986; Ulanowicz and Kay, 1991) are presented to describe holistic properties of the upper Gulf of California ecosystem in the last 50 years. One of the emerging properties of the energy flow indices incorporated into the network analysis is the overhead measure that represents the maximum energy in reserve. Overhead is a measure of the system's ability to withstand unexpected perturbation, an important consideration 157 in the study of the structure and function of any ecosystem (Ulanowicz, 1986). In the upper G u l f ecosystem, the overhead obtained from the past and present models shows a clear declining trend from 1950 to 2000 (Fig. 62). This trend could be interpreted as a loss in the energy storage contained in the historical hundreds o f millions of tonnes of nutrients and organic matter (180 millions tonnes/year; Van Andel , 1964) once deposited every year by the Colorado River, but which ceased at the end of the 1960s due to the construction o f the Hoover and Glen Canyon Dams. This result supports the hypothesis made by van Andel (1964) and Cupul (1994) that the upper gulf is suffering an erosional and starvation phase because the sediments and nutrients that were deposited during the pre-dam period have been exported and consumed at similar rates to those before the 1940s. The (-50%) reduction of Total System Throughput (TST; total flux through the system) in the system from 1950 to 2000 is particularly significant. This is one o f the most important attributes of the network analysis because it represents the relative size of the ecosystem, incorporating all the internal consumptions and exports o f the system including commercial fishing. The range of the TST obtained from the past and present models in the upper G u l f is 35% higher than an Ecopath model of the Northern G u l f of California that included the upper G u l f and its Delta region (Morales-Zarate et al, 2004). Also , they are relatively smaller (-20%) compared to other ecosystems models in the Central G u l f o f California and G u l f of Mexico (Arreguin et al. 2002; Arreguin et al, 1996). 158 £3 o 25000 20000 15000 10000 5000 I 1950 1960 1970 1980 1990 2000 Figure 62. Decline of the overhead measure obtained from past and present Ecopath trophic models of the upper Gulf of California. The overhead measures the energy in reserve in the system and it can represent how the system may meet unexpected perturbation (Ulanowicz, 1986). This declining trend is interpreted as a loss of the energy accumulated in the sediments deposited in the region attributable to huge dams along the Colorado River. Another key attribute that declined (-50%) over the past five decades in the region was the Total System Biomass (TSB) expressed as t/km /year (Table 13), indicating that in 50 years the loss of the average 180 millions of tonnes of nutrients and sediment delivered per year from the Colorado River combined with overfishing practices (i.e. totoaba, sharks and shrimps fisheries collapsed) eradicated approximately 7.7 x 105 tonnes for the 4,500 km 2 of the upper Gulf of California and its delta. This observation is supported by other attributes obtained, i.e., the Proportion of Total Flows associated with detritus that also presented a decline of 30%, suggesting a loss in the number of interactions in the lower trophic levels where the detritus food chains dominate. This hypothesis is supported by the decline in the proportion of the biomass of benthic and pelagic groups over the last 50 years, indicating that groups directly affected by detritus and nutrients delivered by the Colorado River have declined in abundance. This trend reproduces the history of many benthic species, e.g., endemic populations of clams were reduced to 5% 159 of pre-dam abundances (Kowaleski et al, 2000; Rodriguez et al, 2001). A second perspective provided by the network analysis was a general framework of the structure and changes of fisheries among the three trophic models. These results are summarized in Table 13. A key result of this analysis is the reduced trophic level of the catch recorded in the upper G u l f during the last 50 years, during which a loss of 0.25 T L was estimated by the models (Table 13). This trend accords with the average reduction in the trophic levels of catches recorded for the entire G u l f of California since the 1970s. This reduction is the result of an intense fishing mortality that gradually shifts the catches from a long-life high trophic level species to that of a short lifespan species (Sala et al, 2004). This result confirms that the industrial and artisanal fisheries in the upper G u l f are negatively impacting this ecosystem, including the near extinction of totoaba from over-fishing (Cisneros-Mata et al, 1995; Roman-Rodriguez and Hammann, 1997) and the high incidental mortality of vaquita due to gillnets (Jaramillo-Legorreta, 1999; D'Agrosa, 2001). Sala et al. (2004) concluded that the coastal fisheries of the southern G u l f are unsustainable and that management needs to be re-evaluated in order to prevent further degradation of coastal food webs. These suggestions are supported by the results obtained in this section and they can easily be extended to the upper Gu l f of California. The intense mortality imposed by the multiple fisheries developed in the upper G u l f in the last 50 years was also indicated by the increasing trend of the total catches obtained from the network analysis of the past and present models. A n increase of almost 300% in the total catch was estimated in the period modeled, representing an increase in the total extraction of marine resources of almost 35,000 tonnes per year for the entire upper G u l f during the 2000s over catch levels in the 1950s (Table 13). 160 Table 13. Flow indices of the upper Gulf of California ecosystem based on the results of the Network Analysis applied in the three trophic models constructed for the 1950, 1980 and 2000 periods. Ecosystem attributes 1950 model 1980 model 2000 model Change Remarks Total System Throughput (t/km2/year). 12,106 8,132 5,922 - 48 % Relative size of the system (better than the sum of the biomass). Overhead (t/km2/year). 13,024 9,417 5,998 - 46 % Maximum energy in reserve and how the system can meet unexpected perturbation. Total System Biomass excluding detritus (t/km2/year). 347 233 176 - 49 % Could the nutrients delivered by the Colorado change the carrying capacity? Proportion of total flows originated from detritus. 0.32 0.26 0.23 - 32 % Relevance of Colorado River as main source of detritus accumulation. Ratio of biomass of Benthic/pelagic groups. 2.73 1.57 1.43 - 52 % Reduction of groups directly affected by detritus & nutrients from CR (i.e. clam population was reduced to 5% of abundance). Total catches, (t/krrf/year). 3.5 13.4 12.1 + 300 % 10-15% of Mexico total landings originated in the U G C Mean Trophic Level of the Catch. 3.14 3.09 2.89 - 0.25 T L Intense fishing pressure, suggesting F D M F W . 161 Examining Ashing and climate effects on the upper Gulf of California. Chapter IV illustrated how past and present models can be used to explore how single functional groups affect form and structure of the upper Gulf of California ecosystem. These changes can then be used as a first approach to quantify ecological impacts attributable to diversion of the Colorado River. Although the Back to the Future approach adopted in this research entails reconstruction of past ecosystem states of the U G C , it also emphasizes the potential of the dynamic simulations to explore changes in the U G C as a result of either natural factors (influence of climate) or human influences (habitat destruction or effects of fishing). Using this new methodology, this Chapter sets out to explore the influence of climate, ocean change and fishing on the marine ecosystem of the U G C . Understanding the effects of these three factors in marine ecosystems (not just in the Gulf of California) may not only allow us to forecast ecological and economic impacts more rapidly, but also simultaneously develop more robust policies against climate and ocean change. To this end, the second part of this Chapter includes various types of climate forcing functions to explore the role of climate change in the upper Gulf of California. 5.1. Exploring ecosystem effects of changes in Ashing effort under different policy objectives. Two main aims in building present and past upper Gulf of California models were to explore the effect of changes in mortalities imposed by fishing fleets operating in the area, and to investigate possible optimum fishing strategies under specified economic, social and ecological objectives. Dynamic simulations (changes of fishing mortalities over time) based on trophic models can guide decision makers as to which directions they can take in future management plans, but are of course subject to all of the uncertainties associated with such models. 162 The fishing optimization presented in this Chapter is an exploratory approach to maximize four of the critical 'objective functions' considered for management of marine ecosystems defined in Ecosim (Christensen et al, 2004). (1) Maximize fisheries rent, i.e., profits from sale of catch (each species has a market price) after deducting the costs of fishing (both fixed and variable). The objective is to focus on the fishing efforts of the most lucrative species. Appendices 7 and 8 present the market price used for each commercial species and the fixed and variable costs for the fleets used in the optimization. (2) Maximize social benefits, defined here as direct employment in the fisheries. For each gear type or fishery sector, the number of jobs per catch value is specified in the model, and the optimization favors the most labour intensive gear. The benefits of this objective are calculated as numbers of jobs relative to the catch. (3) Maximize rebuilding of species: this objective reflects the external pressure on policy makers and stakeholders to preserve or rebuild the population of charismatic or indeed any given species. Fishing mortalities across gear types are adjusted to maximize the biomass of groups that receive a high weighting value from the user, in this case, it focused on the rebuilding of the two endemic species that are endangered in the U G C : totoaba and vaquita. The biomasses used during the optimization corresponded to those incorporated into the 2000 E w E model. (4) Maximize ecosystem structure or 'Ecological'. This objective is based on the network analysis presented in Chapter i n and it refers to the ecosystem 'maturity' concept described by Odum (1971 and 1988) in which mature systems are dominated by large, long-lived organisms. Optimization for ecology often implies a reduction in the fishing effort for all gear types in order to maximize the biomass of the groups that receive a user-set weighting value (Christensen et al, 2004). 163 The fishing optimization search also allows the modeler to specify the weights for one or more of these objective functions, based on the management priorities established for each scenario. Basically, by changing relative fishing mortalities, the multi-dimensional Davidson-Fletcher-Powell search algorithm included in the Ecosim 'policy search' routine will seek an optimum solution based on the weighting assigned to the objectives (Walters et al, 2002; Christensen et al., 2004). The search iteratively changes the fishing mortality of all the gears employed (a total of six for the 2000 model) to maximize the objective specified (or a mix of the four objectives) over a simulation of the 50 years. The catches (including estimates of illegal fishing), discards, fishing efforts, relative operating costs and market prices used in the mass-balanced model of the U G C for present day conditions were incorporated into the optimum fishing routine built into Ecosim (Christensen et al., 2004) and run for 50 years (also, this model was calibrated with a time-series of fishing efforts; see details in Chapter II). The optimum search maximizes the chosen objectives and provides a forecast of economic values, numbers of jobs, catches and biomasses at the end of the 50-year runs for all the groups exploited in the ecosystem. Some scenarios resulted in drastic depletions of vaquita and/or totoaba, and in extreme cases, their extinction (vaquita) or extirpation from the upper Gulf (totoaba). Assumptions, uncertainty of the results and the practicality of implementing any reduction of the fishing effort in the U G C are discussed at the end of this section. Using the 2000 landings, discards and fishing mortalities of the upper Gulf as a starting point, important changes in the total mean catch (all commercial groups) were found after the 76 scenarios were run under a 50-year simulation. Figure 56 shows the theoretical tradeoffs between extreme scenarios where no fishing was allowed (high weighting on rebuilding or ecological objectives) or those cases where excessive fishing efforts were employed without considering the depletion of charismatic species (scenarios with high values on economy and number of jobs). This was as expected, because economic and social, as opposed to ecological weightings, were assigned to maximize catches. 164 However, a few scenarios which focused on ecological attributes, for example scenario number 53 (Appendix 8), a achieved a hypothetical harvest of up to 0.6 t/km2/year greater than that of the present landings (Fig. 63). Although this scenario emphasized the ecological objective, it was run with the same fishing efforts for each gear type employed in the 2000 mass-balance model. For this reason, scenario 53 was considered a 'realistic' option as opposed to two other scenarios with the potential of higher harvests than those of 2000 were considered as unrealistic because they were designed with no fishing (zero value as the economic objective). A similar pattern was found in the hypothetical catch obtained under the rebuilding goal, where only scenario 51 predicted a catch 0.3 t/km /year higher than that estimated in 2000 under the same fishing efforts (score 1 in economic objective), but which emphasized the recovery of totoaba and vaquita). It is noteworthy that only 2 out of the 44 ecological and rebuilding scenarios tested produced harvests slightly higher than the catches estimated in 2000 with the same fishing efforts as those of the 2000 model. The original goal of this exploratory analysis was not to incorporate the results from the scenarios described above into future policies for the management of the fisheries in the region, instead, these 'win-win' scenarios (i.e., scenario 53) simply represent a preliminary stage and they must be taken as an exploration exercise which indicates that both conservation and sustainable fisheries could reach their goals. Future detailed analyses of the win-win solutions must be conducted in order to get more solid and realistic options that could set the basis for long-term policies which would include both restoration of endangered species and viability of the fishing sectors. 165 ca o ca 4= I -o c o o ca Qi. Ecororrical Social rxofogcal Rebuilding Figure 63. Mean change in total catch between the 2000 model and those forecast 50 years later in the upper Gulf of California (expressed in t/km /year). The dotted line represents the 2000 total catches estimated in the region. The simulation was run under four major criteria employing 76 different scenarios in order to explore optimum fishing strategies (see text for details). Error bars represent standard deviation. Using the 2000 market prices ($US/kg; Appendix 6) and operating and fishing costs as detailed in Appendix 7, it was possible to calculate the mean change in profit after the 50-year ecosystem simulation under the 76 scenarios considered (Fig. 64). The observed trend met with the expectations, i.e., economic and social fields must be the most profitable objectives whereas restoration and rebuilding objectives were predicted to be less profitable (in comparison with the value obtained for the commercial fishing entered in the 2000 model). Even though the ecological scenarios were less profitable than the economic and social goals, a few of them (i.e. scenario 47) resulted in theoretical profits higher than those estimated in 2000; however, they were considered unrealistic because of the proposed drastic reduction in fishing. Also, it is worth noticing that the mean profit of the rebuilding objective (including specific scenarios for vaquita, totoaba and the 166 entire ecosystem) was lower than the current profits and that only scenario 76 produced profits above the 2000 model (upper limit of standard deviation of the mean profit of the rebuilding objective, Figure 64); it is, however, considered unrealistic due to its properties (zero values in economic and social objectives). Figure 65 displays the relationship between the rebuilding of vaquita and the total biomass of the upper Gulf (excluding detritus), suggesting that recovery efforts and plans for this endemic marine mammal could also result in an increase of the overall biomass of the region. This recovery is explained as a direct result of a reduction in fishing pressure, but there were scenarios that could match the economic conditions used at the start of the 50-year simulation (Fig. 65). It is important to mention that several restrictions and simplifications were adopted during the simulations and the win-win scenarios obtained are far from the real conditions. Further analysis and local scientific and fisher's input into scenario interpretation is needed before incorporating into management policies. Fisher's participation is critical to their compliance with any future management regime. On the other hand, Figure 66 exhibits a negative (but non significant) trend between profit and the recovery of vaquita. This negative trend is explained by all the restrictions imposed on the fleets operating in the upper Gulf in addition to the high mortality imposed by incidental killings (gillnet fleets) and the depletion of potential prey caught by this marine mammal. Considering just the economic value and profits, the recovery of the vaquita population to double its current level as a result of closing the fisheries could lead to a loss in profits of up to $0.61/kg/km (54%). This theoretical exercise illustrates the complexity in choosing ecosystem restoration goals; the social component and sustainable fisheries are elements that could not be excluded from any realistic restoration goal. 167 2 2 a 1.5 + o 0.5 Eccroriical Social Ecological Rebuilding Figure 64. Mean change in the profit obtained from commercial fishing ($US/kg/km ) between the catches of 2000 in the upper Gulf of California and those forecast after 50-years of ecosystem simulations. The dotted line represents the value of the catch at the start of the simulation (see text for details). Error bars represent standard deviation. 168 Figure 65. Relationship between the restoration goals for vaquita with the total biomass of the upper Gulf of California ecosystem (Y-axis) and the biomass of vaquita after a 50-year simulation. This simulation considers only restoration and rebuilding objectives (34 scenarios), but shows that conserving efforts to preserve vaquita could achieve an overall rise of the biomass in the region (This significant relationship <0.05 is displayed by the line). 169 2 <a >> o in CD C3 13 o 1 . 5 0 . 5 H 0 . 0 0 4 0 . 0 0 8 0 . 0 1 2 o o o S3) 0 . 0 1 6 Biomass of Vaquita (t/km ) Figure 66. Relationship between the profits ($US) predicted after 50 years of simulated fishing in the upper Gulf of California (industrial and small scale fleets) and the recovery of the population of the endemic vaquita. The negative trend displayed (non-significant) that the profits decreased as the biomass of vaquita increased, and it agrees with that which was expected due to a reduction in fishing efforts. 5.2. Exploring changes in biodiversity under different fishing strategies. Quantifying changes in biodiversity across scenarios is another critical component to include in the policies and strategies of managing marine ecosystems. This exploratory analysis used the Q-90 biodiversity index for comparison among the 76 scenarios designed. Q-90 is a modified Kempton index (Ainsworth and Pitcher, 2006), designed to be used specifically in trophic models generated by Ecopath. Q-90 measures the effects of the different fishing strategies on whole ecosystem biodiversity, through the changes in the slope of the cumulative curve between the 10 and 90 percentiles (Fig. 67). The main 170 modification of Q-90 is that each functional group is considered a single species where biomass is analogous to the number of individuals, a vital element to estimate richness and biodiversity (i.e., Shannon- Weaver/Simpson index; Margalef, 1968). This methodology avoids extinctions by maintaining the same number of functional groups; but sets a low non-zero for scenarios with critical depletion (high weights on economic objectives). To increase the sensitivity of the index to group depletions, a filter is passed over the biomass profile during each year of the simulation (Ainsworth and Pitcher, 2004). For this exploratory analysis, an 80% filter was adopted, meaning that i f a biomass of a given functional group falls below 80%, that group is omitted from the Q-90 calculation and it will result in a reduction of the measured biodiversity of the system. l o g R , log R 2 l o g - a b u n d a n c e Figure 67. Measure of biodiversity with Q-90 index modified from Kempton and Taylor (1976) for trophic modelling of marine ecosystems (Ainsworth and Pitcher, 2004). S represents the number of functional groups (analogous to number of species) in the reference model; R i and R2 are lower and upper 10 percentiles of the species abundance distribution. Figure courtesy of Ainsworth and Pitcher (2004). 171 Using the changes of this biodiversity index, it was also possible to explore the relative ecological impact of different fishing strategies. This methodology was designed to support the use of marine ecosystem modeling to find optimal fishing strategies. The incorporation of biodiversity into the search for optimal fishing will increase the likelihood of obtaining more realistic restoration targets that satisfy multiple (social and economic, ecological) management objectives. Using this methodology, Q-90 index was calculated for each of the 76 scenarios that included economic, social, ecological and rebuilding objectives. As explained in Section 5.1 (an economic objective maximizes profits, a social objective maximizes employment, an ecological objective increases the biomass of long-lived species and a rebuilding objective includes risk aversion for specific species; Walters et al, 2002), these objectives span the spectrum of human use versus conservation, and include both total fishing with no consideration of the ecology and no fishing in the upper Gulf. Figure 68 shows the relationship between the biodiversity measured by the Q-90 index and the value of the catch (expressed in $US/km 2/year) after a 50-year period of simulations for the 76 scenarios categorized by social, economic, ecological and rebuilding criteria. After 50 years of simulation, none of the 16 economic scenarios tested produced a higher biodiversity than 2000 (Q-90 = 5.1); however, they were the most profitable ecosystems (in terms of the value of the commercial species). As expected, rebuilding and ecological objectives produced the highest biodiversities in the long-term, reaching up to 10% more biomass for the total ecosystem in comparison to that calculated in 2000 in the upper Gulf. It is remarkable that, at the end of the 50-year period of simulation, only seven out of 76 scenarios produced both biodiversities higher than 2000 and theoretical harvests and commercial value above the 2000. The majority of these scenarios were considered unrealistic because they implied total or drastic reductions to the fisheries in the region. However, scenario 53 (Appendix 8) was the only one that simultaneously considered the same fishing efforts as 2000, and at the end of the run, produced a 'win-win' ecosystem because its biodiversity and catch value were higher than those at the start of the dynamic simulation. 172 The biodiversity resulting from the 50-year simulations of the 76 scenarios produced consistent results according to each of the four objectives implemented. For example, under the monetary value ($US), the biodiversity produced by the rebuilding scenarios was more comparable and closer to that obtained under the economic goals (Fig 68). For this reason, a hierarchical clustering analysis was employed to categorize the 76 scenarios based on Ward's minimum variance method ( A N O V A sum of squares between Q-90 and $ value at the end of the simulation) to estimate the minimum distances between single linkages. Figure 69 shows the cluster dendogram in which economic scenarios were clearly separate from ecological and rebuilding scenarios. The results show that the four objectives employed in the search for optimum fishing undoubtedly produced statistically different outputs that were in agreement with the different weights used under the economic, social, ecological and rebuilding goals. The overall results of this exploratory analysis confirm that fishing and conservation could be compatible, but the low number of scenarios with this 'win-win' characteristic suggests a high risk for endangered vaquita and totoaba. The Mexican government's current management policies for the upper Gulf chiefly focus on conservation and controlling the size of the fishing effort. These policies are in general agreement with the win-win scenario criteria for long-term, sustainable fishing in the region. In this way, fishing effort in the Gulf of California has been regulated efficiently at least since early 1990s when the first steps by the Mexican government were taken to create effective systems of licensing and permits, creating Biosphere reserves, incorporate accurate fishing seasons, establishing mandatory turtle excluder devices in commercial shrimp trawlers, createe committees for the recovery of endangered species (i.e., vaquita and totoaba), use the Mexican Navy to police the outlaws and one of the most important regulation was to createe in 2000 the National Fisheries Chart (CNP) to avoid the increasing trend of fishing efforts. 173 2.5 -, 3.5 5.5 4 4.5 5 Q-90 Index • Economic A Social O Ecological O Rebuilding Figure 68. Relationship between the biodiversity measured by the Q-90 index and the final value of the catch (expressed in $US/km 2/year) after 50 years of simulation under economic (square), social (triangle), ecological (circle) and rebuilding (diamond) criteria. Horizontal and vertical dotted lines represent the 2000 model values at the start of the simulation. 174 69 34 14 16 ro 11 12 15 .67 61 62 63 24 60 29 73 51 71 66 38 34 26 59 43 52 47 10 33 45 9 64 7 2 8 1 67 68 66 74 5 17 19 75 3 4 76 Or >-Economic L Economic-Ecological b>- Economic-Social Ecological L Rebuilding Figure 69. Cluster dendogram for the monetary value (based on market prices) and biodiversity index (Q-90) of the 75 scenarios of fishing optimization in the U G C . The cluster was obtained using Ward's minimum variance method ( A N O V A sum of squares between both variables) to estimate the minimum distance between single linkages. 175 5.3. Examining Tradeoffs A fundamental way to explore the benefits of the restoration goal for the upper Gulf is to compare the economic value of the fisheries operating in the region. The benefits estimated for each of the 76 scenarios are based on the Net Present Value. The management policy for marine ecosystems clearly states that economic considerations are not sufficient and that future policies and restoration goals must include all sectors of the ecosystem, such as social and biodiversity components (Sumalia and Walters, 2003, 2004; Pitcher et al., 2005). In this exploratory analysis, a biodiversity index Q-90 (Ainsworth and Pitcher 2004) and profits based on the Net Present Value (based on the 2000 model) were included in the simulations to compare the restoration goals for each scenario. Tradeoffs between economic values and biodiversity (measured by the Q90 index) for the 76 scenarios were considered for each of the three models of the upper Gulf of California (1950, 1980 and 2000) where the optimized fishing simulations were grouped according to their objectives (economic, social, ecological and rebuilding) on the x-axis. The profits resulting from the fishing simulations are presented on the left of the y-axis (no discount rates were included). The right side of this Figure illustrates the biodiversity of each scenario as measured by Q-90 statistics (Ainsworth and Pitcher 2004). Before examining the tradeoff of the restoration goals employed in the region, it is worth mentioning the depletion of exploited stocks in the upper Gulf, including a loss of 6% in the biodiversity between 1950 and 1980 and approximately 8% from 1950 to 2000 (Fig. 70). In general, it was found that 'economic' and 'social' scenarios produced ecosystems with higher profits, but they tended to sacrifice diversity. Vice versa, the 50-years of simulation indicated that all the scenarios categorized as 'ecological' and 'rebuilding' resulted in ecosystems with the highest biodiversities, but with profits lower than those estimated in the 2000 model (Fig. 71). The results obtained from the 'mixed' scenarios do not display a definite pattern, but it could be said that they represent a balance between 176 the four criteria considered. The 'success' of future restoration goals and fishing plans must include a full evaluation in each of the social, economic and ecological fields with the participation of experts in order to increase the likelihood of developing restoration scenarios and targets with more solid and realistic backgrounds and to meet with the management regimes needed in the upper Gulf. The exploratory analysis presented in this chapter illustrates the need to perform a more complete and detailed analysis of the cost-benefits of fishing strategies before considering any policy goal in the upper Gulf and its biosphere reserve. The social factor is considered to be one of the key elements in future plans for the management of this region because, as illustrated in section 3.2 (Chapter III), the complexity of local fishers of both industrial and small scale fleets could result in sudden protests and impediments to establishing reductions in quotas key to achieving more sustainable management for some of these sectors. For this reason, it is recommended that educational programs be introduced among the several fishing camps along the upper Gulf with the goal of emphasizing two critical aspects. The first aspect would focus on the enormous relevance of the interaction between non-commercial species and those that have been historically exploited. This effort could result in a better appreciation of charismatic species such as vaquita, other marine mammals or turtles in the future by fishers who may expect to find a positive relationship between the biodiversity of the upper Gulf and their catches. The second factor that could be considered in these educational programs would be towards the vision and value of long-term sustainable exploitation - the food and livelihoods security concept - explaining the relevance and reasons for attempting to restore biomasses of the long lived target species such as totoaba, sharks or corvinas for the future richness and fishing of upcoming fishing generations. 177 50-i 40 30 H 20 A io H h 5 X u — •—-1950 1960 1970 1980 1990 2000 • Bbrrass A Diversity (Q90) Figure 70. Depletion of stocks (fish and invertebrates) estimated in the upper Gulf of California from 1950 to 2000: taken from the three Ecopath models presented in this thesis. Broken line represents a reduction in the biodiversity of this ecosystem (estimated by Q-90 Index), with a loss of approximately 8% in the last five decades. 178 1980 Figure 71. Tradeoffs between economic values ($US) and biodiversities estimated in the upper Gulf of California from 1950 to 2000. See text for details. 179 5.4. Exploring influence of climate on the food web structure and productivity of the upper Gulf of California. One of the lessons illustrated by the last three decades of fishing is that not all changes in marine fish stocks are due to fishing pressure. Examples in practically all productive marine ecosystems have shown that for many fish stocks, climate change and long-term natural fluctuations also play an important role (i.e. Polovina et al, 1994; Schwing et al, 2002; Lavin et al, 2003). Centuries of fossil records from periods prior to industrial fishing demonstrate long-term patterns from 400 A D to 1980 A D of low and high abundances of sardines and anchovies in the Pacific current which concur with regimens of climate change (Baumgartner and Christensen, 1985; Baumgartner et al, 1993; Lozano-Montes, 1997). The Gulf of California is strongly affected by climate fluctuations such as E l Nino, which is the strongest identified cause of interannual anomalies in the region (Robles and Marinone, 1987; Marinone, 1988; Ripa and Marinone, 1989; Lavin et al, 1997). Baumgartner and Christensen (1985) found that the interannual changes in the ocean climate of the Gulf of California (indicated by sea level and temperature anomalies) were explained by the E N S O cycle. Direct evidence of high surface temperatures, reduction in salinity (0.1 to 0.3) and changes in the thermocline in the Gulf of California were reported during the 1992 E l Nino (Fernandez-Barajas et al, 1994). Soto et al. (1999) also used satellite infrared images of the G o C to evaluate interannual anomalies of sea surface temperature from 1984-1995; these images showed that the most important interannual sea surface temperature (SST) anomalies relate to E l Nino events. These observations were echoed by the results from 1984 to 2000 reported by Lavin et al. (2003), who found a statistically significant positive trend between SST and interdecadal variation of the Pacific Ocean, where the 1988-1999 L a Nina produced the largest negative anomaly in SST during the period 1984-2000 (Fig. 72). 180 Figure 72. Time series of the monthly SST anomalies (after removing the mean, 25.47 °C) in the Gulf of California reported by Lavin et al. (2003). The most notable positive anomalies were explained by E l Nino events in 1992 and 1997 and the largest negative anomaly was due to the 1988-1989 L a Nina. In addition to E l Nino and L a Nina events, there are important interdecadal changes in the rainfall reflected in the Pacific Decadal Oscillation (PDO), resulting in direct changes in the water, nutrients and sediments delivered by the Colorado River into the upper Gulf. For example, the entire Gulf of California was below normal rainfall during the 1990s decade, with a drought period that began in 1992 in the Northern region (Brito-Castillo et al., 2002). Changes in the Pacific climate observed from 1998-2001 suggest that the Pacific is entering its cold phase. The current knowledge of the P D O phase is insufficient to predict which kind of period, either below or higher than normal humidity will happen in the summers of the Gulf of California (Hare and Mantua, 2000). Due to the practically zero correlation (r = 0.03) between long-term summer fluctuations of rainfall and P D O index from 1921 to 1999 in the Gulf of California (Brito et al, 2002), it was decided not 181 to include the P D O index as a forcing function of the trophic models of the upper Gulf. Instead, the strongest anomalies in temperature and changes in precipitation due to E l Nino and L a Nina events were employed during this exploratory analysis as the main environmental forces interacting with the trophic models presented. In order to explore possible effects of environmental factors on the components of the food web in the U G C ecosystem, one of the built-in routines of Ecosim (Christensen et al., 2004) was used to evaluate not only the impact of physical and environmental parameters on the trophic interactions driven by primary production but also the cascade of effects on the system. This routine, which is known as 'forcing function', is generally applied directly to primary producers (Christensen et al., 2004). Figure 73 presents the forcing function I used in the simulations, in which three E l Nino events (1983, 1993 and 1997) and one La Nina 1988-1989 event were considered. The forcing function was derived the scaling of the biomasses of phytoplankton obtained by local surveys and those calculated by the model through Monte Carlo runs. This scaling was incorporated into Ecosim in order to explore the effect of dams along the Colorado River on the trophic interactions driven by primary production. When applied as a forcing function on primary producers, this time series not only produced a cascade of trophic impacts on primary consumers, as was expected, but also affected top predators, such as sharks and marine mammals (including vaquita). Primary producers (mainly phytoplankton) clearly flourished as a result of the strong E l Nino recorded in 1983, possibly due to the addition of nutrients brought with the floods of the Colorado River. The boost of primary production was associated with increases in the biomass of primary consumers and lower trophic levels that simultaneously provided energy to higher trophic levels. Figure 74 illustrates this cascade of energy through the whole food web as a result of the enrichment of nutrients delivered during E l Nino 1983 and, on a smaller scale, in E l Nino 1993 and 1997. 182 70 60 -£ 50 - A ^ T 0 J_ c 40 -f • ^ j= 30 - • cu 1 -£ -I 2 0 - I CQ . 1 10 -o -I—• 1—. 1 1978 1982 1986 1990 1994 1998 2002 Figure 73. Time series of the biomass of phytoplankton obtained by surveys (triangles) in the upper Gulf of California from 1974 to 2002 compared with that estimated by the model (black circles). Bars represent standard deviation of the mean. This time series was used as a forcing function in Ecosim to explore the effects of dams on the trophic interactions driven by primary production (see text for details). In some extremes cases, e.g., blue shrimps, crabs and sardines, it was found that the biomasses of these groups increased up to five times in E l Nino events. This behavior echoes the elevated catches of shrimp in the upper Gulf during the Colorado River floods (Galindo-Bect and Glenn, 2000). The mechanism through which the river discharge might have affected the shrimp fishery is unknown (Galindo-Bect and Glenn, 2000); however, lower salinity may have improved the survival of early life stages by providing an expansion of the nursery areas which could have resembled past conditions of the upper Gulf as described by Sykes (1937), Osorio-Tafall (1943) and Leopold (1949). Aragon-Noriega and Calderon-Aguilera (2000) demonstrated that blue shrimp post-larvae were significantly more abundant in those years when the freshwater flow reached the upper Gulf (mainly by E l Nino 1993 and 1997). It is also worth mentioning that, during the L F K analysis (Chapter ITI), the local fishers reported a migration of sub-adult blue A A T A A t 183 shrimp from estuaries to the marine waters during the sporadic discharges of the Colorado. They also stated that their catches are strongly influenced by the freshwater delivered by the Colorado River (more details are explained in section 3.2 of Chapter III). Besides a significant relationship between catches and freshwater as presented by Galindo-Bect and Glenn, 2000), there is also a significant relationship between the interannual variations in the growth of the brown shrimp and the temperature in the Gulf of California, where growth of this species was favorable only when during E l Nino of moderate or weak intensity, as in 1993. However, during more intense events, such as those in 1983, the effect on growth was adverse (Lopez-Martinez et al., 2003). The results presented in Figure 74 must be taken as preliminary, i.e., cannot yet be used as a confident prediction of future changes in the abundance of stocks or population of species in the upper Gulf during E l Nino events. There are several points to be considered in the modelling of the interaction between marine ecosystems and climate change. For example, it is important to realize that fish populations are not in a steady state; rather, they are affected by spatial and temporal changes in the environment and they respond to the environment in different ways. It is widely known that the biological responses to climate variability are both complex and not well understood. Indeed, life cycles and reproduction are key elements in future evaluations of global warming and climate variability. 184 6 n 44 O IB 1 < 2 Top predators T—I—I—I—I 1—I—I—I—I—I—I 1—I I—1—I—I—I—I—I—1—I—I 1—I 1 1—I—I—I—1—I 1970 1975 1980 1985 1990 1995 2000 4H 2 H < 2 • Sharks Toot lied cetaceans —o— Sea liens Middle trophic levels 0 H—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i 1970 1975 1980 1985 1990 1995 2000 Flounders —a— Turtles —°— Demersal fish «e 4 2 2 Lower trophic levels i—i—i—i—i—i—i—i—i—i—i—i—i—I—I—I—I—I—I—I—I—I—I—I—I—I—I—r—i—I—I—I—I—I 1970 1975 1980 1985 1990 1995 2000 • Zooplankton —o— Blue shrimps Crabs Figure 74. Simulated changes in relative biomass of nine groups of organisms from 1974 to 2002 as a result of the changes in the water and nutrients delivered during the floods of 1983, 1993 and 1997 associated with the E l Nino events. The model suggests that biomasses increased across the entire food web of the upper Gulf of California. 185 5.5. Exploring the effect of damming of the Colorado River on the biomasses of the upper Gulf of California. Recent research has confirmed the importance of freshwater input water on the ecology and future restoration of the upper Gulf and its delta (Glenn et al. 1992, 1996, 1999, 2001; Zengel et al, 1995; Morrison et al, 1996; Lucke et al, 1999; Pitt et al, 2000; Cohen et al, 2001). Before upstream impoundments and diversions, the Colorado River contributed with 30,000 1 0 6 m3/year, supporting tremendous biological productivity and diversity (Lucke et al, 1999). More recently, flood releases have been significantly correlated to a rise in shrimp catches in the upper Gulf (Galindo-Bect et al, 2000; Aragon-Noriega and Calderon-Aguilera, 2000; Aragon-Noriega and Garcia-Juarez, 2002), indicating a relationship between the potential renewal of estuarine habitats and the amount of water delivered. It is worth mentioning that hydrographic studies have shown that the effects of the water from the Colorado River reach as far as 70 km from the river mouth (Lavin and Sanchez, 1999). The results of section 5.3 provided evidence that regional rainfall in the upper Gulf is strongly influenced by E l Nino events and P D O and that this rainfall results in increments in freshwater, sediments and nutrients delivered to the region. These inputs simulate increased primary production. Climate, rainfall and primary productivity in the upper Gulf are tightly coupled. The history of the Colorado River underscores the need to explore the phenomena associated with interruptions to freshwater inputs to coupled atmosphere and marine ecosystems. Reconstructed discharge volumes of the Colorado River at the Southerly International Boundary based on the US Bureau of Reclamation (1952) as reported by Cohen et al. (2001), were used to explore the potential effects of the dams built along the Colorado River from 1950-2000. Figure 18 presents the undepleted discharge of the Colorado River that was used as forcing function affecting the primary production of the 1950 model was used as a primary production forcing function in the 1950 model (this mass-186 balanced model was fitted with data originating from surveys and stock assessments as explained in Chapter IV). The estimated undepleted discharge was incorporated into Ecosim as a forcing function throughout 50 years of simulation. No changes in fishing pressure, consumptions or diets were made during this analysis. At the end of the simulation, the final simulated biomass for each of the 50 groups was compared with that estimated from the surveys, stocks assessments and L F K analysis used to build the 1950 model (details of the biomasses and data sources are presented in Chapter III). Figure 75 shows the differences in the biomasses estimated by the model. The major differences were found in the primary producers and consumers, such as phytoplankton and zooplankton. Significant changes were also noted in the lower trophic level groups, such as crabs and benthic fish, which are dependent on detritus food chains, Overall, the simulation results confirmed the historic role of the Colorado River as the main mechanism that provides nutrients and support richness in the upper Gulf (Sykes, 1937; Osorio-Tafall, 1943, Leopold, 1949, Alvarez-Borrego, 1999, Glenn et al, 2001); an increase in the primary production of up to three times the normal amount was obtained after a 50 year scenario in which the huge dams (i.e. Glenn Canyon Dam) were removed from the 1950 model. Biomasses of higher trophic groups, such as totoaba, vaquita and sharks also increased when the system was exposed to undepleted flows of nutrients and water from the Colorado River. These results echoed the observations reported by Glenn (1998), i.e., that vaquita and totoaba may be more numerous in the estuarine region of the upper Gulf after flooding years. Vis-a-vis conservation considerations, it is important to remember that the upper Gulf (mainly Montague, Gore and Pelicano Islands) is used by at least 74 species of bird, including approximately 200,000 wintering shorebirds living in the mud lands (Mellink et al., 1996, 1997). The western population of North America's white pelican population, which has been on the decline for many decades, relies on the Delta as a migratory stopover (Brusca et al., 2001). 187 Different levels of the reconstructed undepleted flow from the Colorado River by Cohen et al. (2001) were used to explore a possible minimum discharge that could produce an increase in the overall productivity of the upper Gulf. Several 50-year simulations using 100%, 50%, 20% and 1% of the undepleted flow were carried out to evaluate the effects of the nutrients delivered by the Colorado River. Figure 75. Biomass profiles of the upper Gulf of California ecosystem in 1950 with 50-years simulations using the undepleted Colorado River flows estimated by Cohen et al., 2001 (line). This biomass, presented in the dark shaded area, is compared to the biomass of the 1950 mass-balanced model after tuning with observed data and climate factors (see details in Chapter IV). Both profiles have an equal fishing series and equal fishing pressure. 188 Figure 76 indicates the results from these simulations, in which the highest total biomass was found when the runoff of the Colorado River was not diverted by dams (100% undepleted), resulting in a 220% rise in the total biomass of the system. Even undepleted flows of only 1% produced increases of approximately 10% in the total biomass of the upper Gulf after 50 years, revealing the enormous role of the Colorado River in the regions' productivity. This is particularly important considering that this region provides 15%o of the Mexican landings (INEGI, 2001), confirming the ecological and economical losses for Mexico during the last five decades as a result of US diversion of Colorado River. Loss of Colorado River water has modified the hydrographic circulation in the upper Gulf (Lavin and Sanchez, 1999); furthermore, the loss in sediments has halted delta construction and exposed the entire deltaic structure to the destructive hydrodynamic forces in the sedimentary basin, promoting resuspension and erosion of sediments in the estuarine basin (Carraquiry and Sanchez, 1999). These negative forces have produced an undisputable deterioration of the upper Gulf of California with severe impact on the marine fauna including its endemic species, such as the well documented case of the totoaba and vaquita. The increase in the total biomass of the system resulting from only a 1%> surplus of water from the river was an unexpected result; however, these results confirmed the observations made by Cohen et al. (2001), who reported that little water might be sufficient to conserve the existing riparian and wetland habitats in the upper Gulf. This observation is supported by a water balance study which suggested that even when there is no floodwater release, vegetation, including native trees and marsh plants, is supported by the scarce amount of water from the agricultural return flows (Cohen et al., 2001). Even modest flood releases are sufficient to induce over bank flooding and to germinate new cohorts of native trees (Zamora-Arroyo et al, 2001). A n analysis of the vegetation response to past flow events estimated that 0.5% of the mean annual flow from the Colorado River could be sufficient to maintain the vegetation and wetlands of the northern region of the Colorado Delta (Cohen et al, 2001). 189 Log % of undepleted discharge from the Colorado River Figure 76. Changes in the total biomass of the 1950 upper Gulf of California model after 50 years of trophic simulation under different percentages of undepleted Colorado River discharge. 5.6. Conclusions. The results presented in this section are preliminary. It is not suggested that these results provide a realistic goal for future management plans or policies in the upper Gulf of California. There are some gaps in the biological and physical fields, such as upwelling processes, seasonal modulation of the discharges, tidal mixing, water-mass formation due to salinity gradients, vertical and horizontal distribution of isohalines, sediment transport, and accumulation. A l l these need to be improved in the future in order to understand the relationship between water discharge from the Colorado and productivity of this system. These results illustrate not only the relevance of considering the environmental influence on marine ecosystems, but also the superficiality of our knowledge of the quantitative 190 process of simulation and reconstruction of past states of natural aquatic ecosystems such as the upper G u l f of California. A n y attempt to restore the upper G u l f must begin with a parallel study o f the climate changes and their influences in the region. It w i l l be crucial in the next project to establish the differences between past climate regimens with those observed during the present day because the extent of these differences w i l l determine the potential degree of restoration. In addition to the climate influence, future restoration research wi l l require ecological, social and economical criteria in order to choose the best restoration goal. The author of this thesis believes that results presented in this chapter could serve to guide future research with the aim of recapturing the former diversity and health states of the upper G u l f o f California. 191 Chapter V I Summary and Concluding Comments. The overall objective of this thesis was to use a ecological theory to evaluate trophic changes in the upper Gulf over the past 50 years as a result of the elimination of nutrients by the series of dams built along the Colorado River and by the intense fishing pressure that has been imposed over the past several decades in this ecosystem. The 50-year trophic modelling of the upper Gulf of California faced several challenges which included having to cope with poorly-known aspects of the basic biology of key species (commercial and non-commercial species) as well as the relative paucity of research and data in the fisheries field. Better information in these fundamental areas is critical to evaluate possible changes in the system as a result of direct or indirect human presence, such as fishing or elimination of nutrients by dam construction, respectively and to develop restoration strategies that are robust to natural variability and climate and realistic and acceptable to fishers. The following specific aims directed this thesis: to develop present and former states of the U G C ecosystems using trophic ecosystem models representative of 1950, 1980, and 2000 covering their major fluxes: phytoplankton and primary production, zooplankton and secondary production, major fish species and their fisheries, marine mammals and their food consumption. Qualitative information and knowledge from local fishers of this region (San Felipe, Baja California; Puerto Pefiasco and Golfo de Santa Clara, Sonora) was also gathered as an additional tool in order to estimate relative abundances of non-commercial species during the 1950s and 1980s. Changes in the U G C ecosystem form and function were explored by comparing the models using time-changes based on dynamic simulations in Ecosim and by studying the effects of inter-annual climate variations such as E l Nino/La Nina in the structure and function of the U G C . This research required a complex integration of several methodologies in the ecological, social and physical fields in order to evaluate the history of a complex marine ecosystem. 192 In Chapter II, I evaluated the incentives to misreport catches during the history of the fisheries of the G u l f of California in order to incorporate more realistic catches into the trophic models of the region. Discards, illegal and unreported catches in the G u l f o f California are a current problem which Mexican authorities are attempting to tackle. For example, in 2002 approximately 1,200 permits were granted to boats in the G u l f of California and Pacific Ocean, and it is estimated that 20 to 30 percent of their catch was illegal (Weiner, 2002). The incentives and estimates for unreported catches reported in Chapter II require more accurate and detailed information from the dozens of fishing camps along the Gulf, especially as to how their catches are registered by local offices and how their reports are sent to the National Fisheries Institute. A direct consultation with owners and workers of vessels operating in the gulf w i l l definitely play an important role in abating or eliminating I U U fishing. However, these estimations are open to discussion and collaboration with Mexican authorities in order to establish realistic anchor points to convert the qualitative estimations to misreport catches into values that can provide confidence intervals throughout a Monte Carlo simulation based on likely error ranges, as it was obtained to evaluate the impact of I U U fishing in other marine ecosystems (Pitcher et al., 2002; Ainsworth and Pitcher 2004). The evaluation of illegal fishing in the upper G u l f o f California is a fundamental aspect in the future management of this ecosystem; that is of significant cultural relevance to Mexico, provided fishing for hundreds of years to several indigenous communities and even today, produces 10% of the total marine production. Consequently, there is an obligation to estimate the true magnitude of illegal and unreported fishing in this area. It is worth mentioning that, during the last decade, the Mexican government took the initiative to combat the ecological and economic crises of the late 1980s in the Northern G u l f by implementing new regulations which included: (1) the establishment of a Biosphere Reserve in the upper section of the Gulf; (2) the creation of N O M (Mexican Official Standards) as an instrument to manage the most important fisheries and to avoid an increasing trend in fishing effort; (3) the mandatory use of turtle excluder devices; founding international committees for the recovery of the vaquita ( C I R V A ) ; and (4) a variable closed season for 193 shrimp by area, using the Mexican Navy to patrol areas of ecological and economic interest. While some obstacles remain, i.e., inadequate scientific resources to address many artisanal fisheries, lack of local government participation in management and the refusal by government scientists to share detailed data (Hernandez and Kempton, 2004). Mexico's current fisheries policies, seem to respond to long-term problems, prevent the growth of illegal fishing documented during the 1970s and 1980s. Chapter II incorporates total extractions (reported + illegal fishing) from the fisheries in the upper Gulf into the 2000 trophic model. This enables the model to produce more realistic results of the interaction and impact of fisheries as one of the main predators in the upper Gulf. Overall, the results of Chapter LI show that the mortality imposed by fishing could be considered as the second top predator in the region, just behind the pressure imposed by large sharks. The 2000 model also confirms that fishing pressure in the U G C is dispersed throughout all trophic levels, imposing high mortalities not only on top predators but also on the bottom of the food web. For example, the large-scale industrial fleet (mainly shrimp trawlers) not only has impacts at the lower trophic levels of benthic resources, but also produces a cascade of negative effects that reach higher in the food web, mainly through the high rate of discards, thus eliminating organisms that have the potential to be consumed by higher trophic levels. Chapter II also displays the important role of detritus (associated with nutrient loads) transported by the Colorado River, where addition of these elements to the system result in increases of biomass. This response could be explained by the 180 millions of tones per year of nutrients and organic matter that used to be delivered each year by the Colorado River, promoting the productivity of lower trophic levels (especially to those groups which are linked to the detritus flows). These, in turn provide energy for higher trophic levels, leading to higher catches. A similar response has been shown in real conditions in other marine ecosystems, where an increase of detrital biomass provoked a rapid functional response in detritivores with a subsequent positive influence on higher trophic levels (Graf, 1992). Due to the substantial relevance of detritus and nutrients in 194 the trophic interactions of the upper Gulf, it could be said that this system is driven not only by these elements (suggesting a bottom-up system; Carpenter et al, 1985), but also the significant impact of predation by large fish (sharks, totoaba, scombrids, and serranids) also indicates the influence of 'top-down' forces (Carpenter et al, 1985), indicating a more realistic 'mixed' control of the food web in the upper Gulf of California. Overall, the results obtained from the model suggest that Sykes's original description (Sykes, 1937) of the upper Gulf and its delta as an ecosystem driven by the detritus delivered by the Colorado River still holds. The last section of Chapter II compares of the energy flows estimated for 2000 in the upper Gulf with those reported in other marine ecosystems in the Gulf of California and Gulf of Mexico. The suggestion by Morales-Zarate et al. (2004) that the northern section of the Gulf of California is a dynamic system at a mature stage of development is supported by the results found in Chapter II. The total system throughput (sum of all flows) suggests that the upper Gulf is a relatively small, but highly productive system. This is corroborated by observations made over decades that the upper Gulf and delta are a feeding ground and area of protection for larval and juvenile stages of many fish and vertebrates (Sykes, 1937; Alvarez-Borrego, 1975; Perez-Mellado, 1980; Pedrin-Osuna et al, 2001; Campoy-Fabela, 2002). Several useful insights emerged from the balancing process of the 2000 model presented in Chapter II. It was possible to identify which areas or groups in the system need more research and understanding, The unbalanced 2000 model suggested that benthic communities (polychaetes, sea cucumbers, snails, scallops, oysters, clams, murex, sea stars, octopus, crabs and others) are the weakest link; more information about their biology, production and consumption is required in order to understand their interactions in the system, including their response to human activities. In future, the trophic imbalances of these eco-groups must be resolved through biological research and a better understanding of their roles in the food web rather than by solving the linear equations of Ecopath. 195 The second half of Chapter II presents a critical aspect for the building and modelling of marine ecosystems, the process known as tuning. It provides adjusted models (based on stock assessment data) that can track changes in biomass that are known to have occurred in the past. The fitting of the 2000 model was based on the fishing mortality obtained from time series data (1983-2000) of relative abundance (e.g., catch per unit effort C P U E ) or absolute abundance estimates (e.g., survey biomasses) for each of the main groups exploited in the region. These estimates were based on several databases obtained with considerable effort by the Mexican Government; including the biomasses surveyed by the National Institute of Fisheries, and biomasses reported in the region from 1978 to 2003 (Perez-Mellado, 1980; Magallon-Barajas, 1988; Nava-Romo, 1994; Crip-Ensenada, 1996; Lopez et al, 1997; Aragon-Noriega et al, 1999). The tuning process was also weighted by the biomasses and catches of shrimp recorded in the region. The incentive to put more emphasis on this group is because its biology and population ecology have been the subject of intense study in the upper Gul f for decades (Felix-Pico, 1975; Perez-Mellado, 1982; Perez-Mellado and Findley, 1985; Magallon-Barajas, 1988, CRLP-Ensenada/INP, 1996; Aragon-Noriega et al, 1999; Aragon-Noriega, 2000; Aragon-Noriega and Calderon-Aguilera, 2000, 2001; Calderon-Aguilera et al, 2002; Aragon and Garcia-Juarez, 2002; S A G A R P A 2003). The shrimp fishery is the most important source of income in the upper Gulf, probably representing the best-known species there, if not in the entire Sea of Cortez (50% of the Mexican shrimp production originates in the Gulf of California; Garcia-Caudillo et al, 2000). As Figure 14 indicates, a correlation between the predicted biomass of shrimps with that estimated by local surveys was obtained, resulting in a mass-balanced model that is also fitted with observed data, thus giving it some degree of validation. In Chapter III, the elements and points to be considered for building the trophic models of the upper Gulf of California representative of the periods of 1950s and 1980s were developed. These periods were carefully selected because they include two of the main events in the history of the upper Gulf: the completion and filling of the Glenn Canyon 196 Dam (1961), and the intense E l Nino recorded in the G u l f in 1982-83. One critical element incorporated in the modelling of the past states of this ecosystem was the L F K analysis (Local Fishers Knowledge) comprising 49 semi-structured interviews conducted with fishers of the upper G u l f (San Felipe, Golfo de Santa Clara and Puerto Pefiasco). The L F K survey had two purposes: (1) to gain information about the fishers' perspectives about possible changes in the perception of past states of the U G C , including losses of abundances and diversity as a result of fishing and the effects of the Colorado River Diversion; and (2) to obtain estimations of relative past abundances of non-commercial species, a critical aspect for the building of past trophic models of the U G C . Through the L F K analysis, which was the only available source of information, it was possible to obtain estimations of past abundances of non-commercial species (i.e. vaquita, turtles, sea lions, dolphins, whales,) before the 1980s. A t the same time, L F K provides evidence that supports the observations made by Saenz-Arroyo et al., 2005, who pointed out the potential use of fishers' perceptions and historical anecdotes to quantify important declines of megafauna in the G u l f o f California because, according to their research, sharks, turtles and big groupers have declined in the central G u l f o f California over the past 60 years. This phenomenon known as shifting environmental baselines, represents all the inter-generational changes in the perception of the state of the environment (Pauly, 1995); however, this phenomenon has been reported not only for fisheries or environmental perception, but also in social sectors (www.shiftinglines.org), explaining why people can be so tolerant of the depletion of resources or increasing violence, respectively. Some of the results from the L F K analysis presented in Chapter m not only revealed significant differences among the three generations of fishers with respect to the number of fishing sites depleted in the last 50 years, but also provided evidence to support a shifting of environmental baselines phenomenon in the region. The reduction of the fishing sites was explained by the fishers as the result of decades of overfishing and the elimination of nutrients from the Colorado River. In addition, significant differences 197 were found in the mean size of the biggest totoaba caught among the three generations of fishers, providing even more evidence of the existence of shifting ecological baselines. It was perceived through the L F K analysis that young fishers of the upper G u l f are very tolerant of the loss and collapse of their fisheries. Three major fishery crises have been documented in the region: the totoaba fishery collapse in the late 1970s, followed by the collapse of the shark fishery in the mid 1980s, and more recently, the collapse of the shrimp fishery in the early 1990s. Overall, it was perceived that young fishers have the impression that old fishers always exaggerate the richness and diversity of the upper G u l f in past decades and that the old fishers are not correct when they state that the upper G u l f once supported colossal populations of totoaba, sharks, Pacific sierras and other predators (suggesting equal richness in the abundances of their preys). Such a rapid shift in the perspective of the degradation of the environment (in just a few decades) is a very important red flag to be considered by both the Mexican government and management authorities i f they wish to promote conservation of the region through courses, seminars or any other educational tools in order to raise the ecological and economic value of the upper G u l f o f California. Another red flag raised by the L F K (besides the elimination of nutrients delivered by the Colorado River) was the pollution of this area by agricultural runoff and the recent drug traffic. A recent report confirmed the mortalities of 367 dolphins, 51 sea lions, 8 whales and more than 200 seabirds in San Felipe (Baja California) were caused by consuming N K 1 9 , a synthetic cyanide compound used for tracking illegal drugs at night (Azuela de la Cueva et al., 1995). From the fishers' perspective, the future of the upper Gu l f o f California w i l l rely heavily on tourism. The second half of Chapter III presents all the interdisciplinary sources of information used to build the 1980s and 1950s models, including scientific expeditions and fish collections by Mexican and U S organizations, fossils records, L F K , and dozens of scientific reports. This extensive source material was used to calculate the hundreds of parameters considered in the trophic modelling. Both past models were balanced and their 198 uncertainties were presented and discussed in section 3.4.1. It is worth mentioning that, before balancing the 1950 and 1980 models, there were areas and groups (benthic fish and invertebrates) in the models that were immediately out of balance, indicating that the biology of these groups in the upper Gulf is not very well understood, raising a flag for more intense research. The trophic imbalances of these groups should be resolved in future models by improving the understanding of their biology rather than by solving linear equations of the model. Chapter III also presents another element, i.e., the pedigree index (?) routine in Ecopath that assists in the sensitivity analysis. This index, which ranges from 0 where data are drawn from outside the system to 1 for data from local surveys and stock assessment, categorizes the origin of the sources of information used to build the model, measuring the uncertainty associated with the type of data for each of the five basic parameters of the model (Biomass, P/B, Q/B, diet composition and catches). The overall P index of the 211 categories measured in the 1980s model 0.59 and 0.47 for the 1950s, indicates a higher uncertainty associated with the biomass estimations from the L F K analysis incorporated into the model. Finally, this chapter discussed critical points to be considered in the reconstruction of past models, such as catchability which was considered constant over the 50-years of modelling, and because of this, no changes to fishing mortality were included. Predation mortality was also considered to be constant in the linkage among the Ecopath models. The objective of Chapter IV was to connect the 1950s and 1980s models with the 2000 present-day condition model. This connection was made through dynamic Ecosim simulations using biomasses estimated from direct surveys or stock assessment data in order to adjust the ecosystem models to reflect changes that have been observed and documented. Two additional problems needed to be overcome before connecting the three models. First, there was no reliable diet information or bycatch estimates from the 1950s. Both Ecopath and Ecosim simulations are highly sensitive to the initial diet matrix, as it determines the base predation mortality rates and the rates of effective search 199 for prey by predators. A second relevant aspect to be considered is bycatch (including discards and 1TJU) because this parameter determines the initial fishing mortality rates, which are then adjusted annually in Ecosim according to the tuning method described above. The best approach to dealing with these issues was to use the 2000 Ecopath model's initial state including its diet matrix (with the best scientific information available) as a first guess for the 1950s diets and reconstruct the (implied) 1950s trophic state. The main benefit of this approach is that (unknown) diet composition and biomasses for the 1950s model remain consistent with those implied by the 2000s model. The tuning process presented in Chapter IV involved a wide range of sources of information, from the L F K and stock assessments (including illegal and unreported catches estimated in Chapter II) to local surveys conducted by the Mexican government since the early 1970s. During this tuning process, a statistical measure of'goodness of fit of the time series data employed was generated each time Ecosim was run. Goodness of fit is represented by a weighted sum of squared deviations (SS) of log biomasses from log predicted biomasses. Using this criterion, it was possible to search for vulnerability (Vs) estimations that gave the best 'fits' of Ecosim to the time series of biomass and catch incorporated into the fitting; and then select the best overall model fitted (lower SS generated). In order to improve the fitted model (1950-2000), the interaction between the ocean and the atmosphere, a relevant aspect in the modelling of marine ecosystems, was incorporated. Natural variability (ENSO) was also considered in the trophic models as changes in the environment can be manifested by a shift in productivity, abundance and distribution of many species (Polovina et al, 1994; Francis and Hare 1994, Hayward, 1997; Pitcher et al, 2005). The complex interactions between biology and climate in the upper Gulf were simulated by running the fitted model was run under the climate influence of the 1950-2000 time-series of the Colorado River discharge below the Hoover Dam, which included several overflows associated with El Nino events: 1957, 1982, 1993, 1997 (Lavin, 1999 and Lavin et al, 2003). The effect of the changes in water, sediments and nutrients delivered 200 by the Colorado River from 1950-2000 was incorporated into the fitted model affecting only primary production. Overall, the incorporation of the environmental factors resulted in an 18% improvement of the fitted model. The improved fitted model (that includes the climate series) was used for the dynamic simulations and quantification of'the water diversion impact on the upper Gulf of California presented in Chapter V . It is worth mentioning that the vulnerabilities resulting from the tuning and climate incorporation of the 1950-2000 model displayed a realistic response to changes in fishing, in which the A V P (average predator vulnerability) displayed a significant (P<0.05) linear relationship with the trophic level estimated by Ecosim. This pattern is in agreement with the assumption that large predators have a broader spectrum of diet, and thus require a longer time to search for food, increasing their vulnerability to predation and mortality imposed by fishing (T. Pitcher, 2006 pers. comm.). This pattern also suggested that a decline in the biomass of potential prey resulted in the need for predators to spend more time searching for food and thus create a greater vulnerability to be caught by other predators or by fishing gear. The increased vulnerability of predators resulting from a decline in their prey is in agreement with the surveys conducted by Felix-Pico (1975); Perez-Mellado (1980); Perez-Mellado and Findley (1985); CRLP-Ensenada (1995), indicating a realistic decline of the benthic fauna in the upper Gulf. This trend of high V s associated with large predators has been found in other ecosystem models in British Columbia (C. Ainsworth, 2006 pers. comm.), but with less intensity in Hong Kong models (W. Cheung, 2006 pers. comm.), thus indicating the need to explore and quantify this assumption in the future, using different trophic models of other marine ecosystems. Generating a scale from 'healthy' to 'unhealthy' attributes for the entire U G C by quantifying the possible ecological impact of diverted waters resulting from a series of dams along the Colorado River on the upper Gulf of California was one of the main objectives and motivations of this thesis project. The effect of the water diversion from the Colorado River was based on the quantification of energy flows of the past models of 201 the upper Gulf, and it was compared to the flows estimated for the current state of the system (2000 model). The quantification of the energy flows provide a form and structure of the system, two key characteristics for understanding the complexity, and simultaneously, a way to quantify changes through time in the upper Gulf of California. By analyzing the flows over the 50-years of modelling, it was possible to track changes in the food web through changes in trophic levels, depletions of biomasses and changes in predation and fishing mortalities. Network analyses and emergent properties obtained from the past and present models represent just one of the many approaches that are used worldwide to define the structure and function of ecosystems. The combination of these methods represents an important step in the further comprehension of marine ecosystems like the upper Gulf of California through time and space. The connection between network analyses and changes in the biomasses estimated by Ecopath suggests that the 'Back to the Future' strategy (Pitcher and Pauly, 1998; Pitcher, 2001, Pitcher et al 2005) could be useful for a comprehensive understanding of the ecosystem dynamics in the evaluation of both human activities such as overfishing and activities that threaten diversity. The ecosystem approach used in the modelling of the U G C for the past 50 years determined the positive role of detritus in the system for all consumer organisms and for fisheries. This response could be explained by the Colorado River delivering at least 180 millions of tonnes of nutrients and organic matter every year, key factors for promoting the productivity of lower trophic levels (especially to those groups which are linked to the detritus flows), which in turn leads to higher catches. A similar response has been reported in other marine ecosystems, in which an increase of detrital biomass provoked a rapid functional response in detritivores with a subsequent positive influence on higher trophic levels (Graf, 1992). However, the loss in energy storage contained in the hundreds of millions of tonnes of nutrients and organic matter historically deposited every year by the Colorado River, but which ceased with the Hoover and Glen Canyon Dams at the end of the 1960s, supports the hypothesis made by van Andel (1964) and Cupul (1994) that the upper gulf is suffering an erosional and starvation phase because the sediments and 202 nutrients that were deposited during the pre-dam period have been exported and consumed at similar rates to those before the 1940s. Also, the loss of overhead in an ecosystem has been interpreted as a measure of the potential energy in reserve (Arreguin-Sanchez and Valerio, 1996; Morales-Zarate et al., 2004), suggesting that the current state of the upper Gulf of California has less potential for growth and is more subject to 'perturbation' than in the 1980s and 1950s. Overall, the results obtained from the models suggest that Sykes's original description (Sykes, 1937) of the upper Gulf and its delta as an ecosystem driven by the detritus delivery by the Colorado River still holds. The results from the network analysis from 1950-2000 revealed that the detritus group (which also includes nutrients and sediments) produced a positive cascade of effects from the bottom of the trophic structure. These results agree with the high productivities reported during decades in the upper Gulf (Zeitzchel, 1969; Alvarez-Borrego et al., 1975; Alvarez-Borrego et al, 1983; Brinton et al., 1986; Valdez-Holguin and Lara-Lara. 1987; Millan-Nunez, E . 1992; Millan-Nunez et al. 1999; Garcia-Pamanes and Lara-Lara, 2001). The high fertility of these waters has been explained by the interaction between high nutrient levels (historically delivered by the Colorado River). Tidal currents, wind, and the plateaus' topography (Lavin et al., 1999), result in numerous areas of upwelling mainly along the eastern coast of Baja California, (Argote et al., 1998) which seems to have the potential to maintain large food chains with no freshwater input (Millan-Nunez et al. 1999), this is likely because of residual sediment from pre-dam periods, but is now a finite resource. Perhaps the reason why the sediment load from the Colorado River is almost zero today (Carraquiry and Sanchez, 1999) is because strong tidal waves (up to 12m from the mouth) removed the previously deposited sediments (before the completion of the huge dams), suggesting an erosional phase in the upper Gulf that is increased by an exportation of sediments to the Northern Gulf at rates similar to those of the unaltered river flow prior to 1935 (van Andel, 1964; Cupul, 1994), and thus creating a starvation phase of nutrients and sediments in the delta as a result of the construction of the dams. 203 Chapter V explored the influence of climate, ocean change and fishing on the marine ecosystem of the U G C . Understanding the effects of these three factors on the upper G u l f may allow for greater accuracy in predicting their ecological and economic impacts and for simultaneously developing stronger policies that are more robust vis-a-vis climate and ocean change interactions. The study of several ecosystems and the effects caused by changes in fishing efforts and diversity (measured by the Q-90 index) under different policy objectives (76 scenarios under consideration) was undertaken in the research presented in this chapter. It is remarkable that, at the end of the 50-year period of simulation, only seven out of the 76 scenarios produced biodiversities as well as theoretical harvests and commercial values higher than those of the 2000 model. The majority of these scenarios were considered unrealistic because they implied total or drastic reductions to current fishing efforts in the upper Gulf. Only a few scenarios resulted in diversities higher than those of 2000. Although the overall results of this exploratory analysis suggest that fishing and conservation could be compatible, the low number of scenarios with this 'win-win ' characteristic suggests a substantially high risk for endangered vaquita and totoaba. The current management policies of the Mexican government in the upper G u l f focus highly on conservation and controlling the size of the fishing effort. These efforts are in conjunction with the win-win scenarios criteria for a long-term sustainable fishing in the region. The second part of Chapter V takes environmental influences into account (Colorado River discharge as a forcing factor) in the search for optimal fishing and a detailed analysis of environmental uncertainties in the 2000 model of the upper G u l f of California is also discussed. The second part of Chapter V presents an exploratory analysis of possible effects of environmental factors on the components of the food web in the U G C ecosystem. This approach included a time-series of the water delivered by the Colorado River, where the interannual changes in the flows are explained by the E l Nino events o f 1983, 1993 and 1997. The results obtained showed a direct increase in the primary production in the region as a result of nutrients being delivered during the floods. The richness in primary production also produced increases of up to five times the original biomasses in most o f 204 the species considered in the model, from primary consumers to top predators such as sharks and marine mammals. However, these results were considered preliminary and there was no suggestion that they could be used in the future prediction of environmental effects or global warming due to the substantial uncertainty of the data and the complex and poor understanding of the interactions between ocean, climate and marine populations. During the application and integration of all the methods used in the reconstruction of the history of the upper Gulf, several assumptions and limitations of these methods were discussed in the relevant chapters, emphasizing the need to examine the impact of these assumptions in the future. This represents one of the rationales of this research, i.e., to address the deficiencies and weaknesses of the required parameters which could be examined in greater depth and whose unbalanced values could be resolved once a better knowledge of their biology is available. 6.1. Recommendations. When quantifying human impact on marine ecosystems through analyses of trophic models, it is vital to estimate realistic extractions realized by fishing rather than simply using official statistics provided by governments. It has been demonstrated that important differences could occur between catches reported and catches obtained (Pitcher et al, 2002; Ainsworth and Pitcher 2005). In the specific case of the upper G u l f o f California, the incentives and estimates for unreported catches presented in this thesis require more accurate and detailed information from the dozens of fishing camps along the Gulf, especially as to how their catches are registered by local offices and how their reports are sent to the National Fisheries Institute. A direct consultation with owners and workers of vessels operating in the gulf w i l l definitely play an important role in abating or eliminating R J U fishing. However, these estimations are open to discussion and to collaboration with Mexican authorities. Anchor points are needed in order to convert the qualitative estimations to misreport the catches into values that can provide confidence 205 intervals throughout a Monte Carlo simulation based on likely error ranges. These anchor points have been used as key elements to evaluate the impact of IUU fishing on other marine ecosystems (Pitcher et al., 2002; Ainsworth and Pitcher 2005). In the next stage, this IUU project must include this information in order to quantify the impact of IUU fishing and obtain a better approximation of the total extractions in the Gulf of California. A quantitative estimation of IUU fishing activities represents just one of a series of effective measures to combat this problem both in the Gulf of California and worldwide. After all, the Gulf of California is recognized not only because of its cultural relevance to Mexico, providing fishing to several indigenous communities for hundreds of years, but also because its richness produces 40% of the total marine production; consequently, there is an obligation to estimate the true magnitude of illegal and unreported fishing. As mentioned, for many groups empirical data are either sparse or non-existent, resulting in the use of diets reported in other models for the same region, but numerous shortcomings were noted. For this reason, an extensive and intensive stomach sampling program in the upper Gulf would provide more accurate estimates of the trophic interaction and quantities consumed by the key species, reducing the uncertainty associated with the role that diets played in the outputs of the Ecopath models. The second area of the model that could be improved with more accurate data is the biomass estimation; the biomasses used for many groups are poorly estimated, especially at lower trophic levels. The parameters employed in the model should be constantly revised or replaced with new estimates. Finally, seasonal variations in energy flows and system attributes were not addressed in the network analysis presented in this thesis. The upper Gulf of California is subjected to a seasonal variation in both biotic and biotic factors (Alvarez-Borrego, 1975; Alvarez-Borrego and Schwartzlose, 1979; Alvarez-Borrego, 1999; Lavin, 1999; Brusca et al, 2001; Lavin et al, 2003), and such variations could influence the energy flows and attributes of the upper Gulf, making this another reason for further investigation in the area. 206 The balancing process o f the 2000 model confirmed that empirical data for many groups are sparse or non-existent, thus resulting in the use o f diets reported in other models for the same region; however, numerous shortcomings were noted from those models. Weaknesses in the diet matrixes used to build the present and past models have been noted in many sections of this research. Additional comments - input from workshop with scientists, fishers and Cocapas- are warranted regarding small juvenile fish (i.e., totoaba) in the diets of predators. The three trophic models presented in this thesis provide estimates of the contribution by various predators or groups of predators; however, it must be clear that the diet information for many of these predators is not completely accurate and more research on consumption is recommended. The importance of cannibalism in large predators such as sharks needs to be calculated more carefully. The second area of the trophic models presented in this research that could be improved with more accurate data is the biomass estimation; the biomasses used for many groups are poorly estimated, especially at lower trophic levels. This is a key element in the construction of trophic models because biomass estimates serve to scale these models. Biomass estimates are poor for many groups. This is especially true for lower trophic groups, but they also occur at higher level groups such as sharks, whales, sea lions and dolphins. The past and present models would be more reliable i f more confidence could be placed in the biomass estimates. The parameters used in the model should be constantly revised or replaced with new estimates. It is important to mention that the process required to build these trophic models is essentially open-ended. The data available for inputs must be constantly added or revised. The models presented here should be considered a first step, and there is an open invitation to critique the structure, the input data and the assumptions in order to improve the models during subsequent exercises. The data presented in this model are available from the author upon request. It is worth pointing out that the 50 years' modelling of the upper G u l f presented in this research does not include the recreational fishing sector whose impact on the region must be studied carefully in the future. This further research w i l l provide information to 207 support the general consensus (according to the LFK analysis) that sports fishers produce large amounts of discards and do not record the number of fish that they catch. It will be necessary to get information from the Tourism Secretariat and from the several sport fish companies in Puerto Pefiasco (Sonora) and San Felipe (Baja California) in order to estimate the number of trips and fish caught by sport fishers. In addition, seasonal variations in energy flows and system attributes were not addressed in the network analysis presented in this thesis. The upper Gulf of California is subject to seasonal variation in both biotic and biotic factors (Alvarez-Borrego, 1975; Alvarez-Borrego and Schwartzlose, 1979; Alvarez-Borrego, 1999; Lavin, 1999; Brusca et al, 2001; Lavin et al., 2003), and such variations could influence the energy flows and attributes of the upper Gulf. The next phase of this research must be aimed at refining the temporal and spatial resolutions of the past and present models. Temporally, the emphasis will be placed on understanding the relative role that fisheries and environments have played on the ecosystem of the upper Gulf in the last decade. Spatially, the emphasis will be on developing a version of the ecosystem model which considers the diversity of habitats in the region, such as the Colorado Delta, mud areas, and the wetlands of the upper Gulf. Another important recommendation exposed by this research refers to the role and impact of fishing imposed by both large and small scale fleets. It was evident that large-scale industrial fleets (mainly shrimp trawlers) not only have an impact on the lower trophic levels of benthic resources, but that they also produce a cascade of negative effects that can reach the higher levels of the food web; this was explained by the high rate of discards that eliminates organisms that can potentially be consumed by higher trophic levels. In contrast, the diverse small-scale sector was more selective than offshore trawlers, with practically no discards. However, the small scale sector is fishing at practically all trophic levels. It is important to consider this impact because the Mexican government recently reduced the shrimp trawler effort in the upper Gulf, and trawling has been banned in the core zone of the biosphere reserve since the mid 1990s. However, these regulations do not seem to be sufficient because the small-scale sector also creates an impact. All sectors of the fishery must be assessed and managed, and it is not 208 completely accurate to label shrimp trawlers as a 'bad fishery' and the small-scale sector as a 'good fishery'. Important recommendations could be drawn from the information received during the Local Fishers Knowledge interviews. The opinion of the fishers of the upper Gulf interviewed suggested the existence of a shifting of ecological baselines in the upper Gulf. Furthermore, young fishers indicated that they are very tolerant of the loss and collapse of their fisheries, referring to the three major crises in the region when totoaba (late 1970s), sharks (mid 1980s) and shrimps (early 1990s) were depleted to levels that resulted in the collapse of their fisheries. In a general sense, young fishers have the perception that old fishers always exaggerate the richness and diversity of the region, and that it is not true that the upper Gulf once supported the colossal populations of totoaba, sharks, Pacific sierras and other predators (suggesting equal richness in the abundances of their prey). Such a rapid shift in the perspective of the degradation of the environment (just a few decades) is a very important red flag to be considered by the Mexican government and management authorities i f they wish to promote conservation of the region through courses, seminars or any other educational tools in order to raise the ecological and economic value of the upper Gulf of California. For example, it is necessary to promote the conservation of vaquita emphasizing its colossal ecological value as an endemic species. This is because many fishers have the impression that their fisheries (shrimp trawlers and shrimp gillnets) have been closed or limited due to vaquita, and they believe that this marine mammal is just a myth in the upper gulf and that the biosphere reserve is just a government hoax in order to make money for its own benefit. Important suggestions could be drawn from the L F K analysis. It confirmed that the artisanal fisheries in the upper Gulf are complex and highly diverse. Each community has its own seasons, regions and fishing methods. Changes in tides, changes in the length of daylight and moon cycles play a role in the fishing activities of this region. Also, non-natural factors such as fuel prices, equipment costs, and market prices affect their fishing. A l l these variables that occur simultaneously must be considered for future management 209 plans in the upper G u l f of California. Also , the L F K revealed an intense desire on the part o f the fishers to participate in and to be informed of all the decisions and management plans for the upper Gulf, including the Biosphere Reserve. The biomass and trophic interactions presented in this research were also used as an initial analysis to evaluate both the diversity and economic richness of the upper Gulf, and their possible worth today i f the dams had not been constructed. This ecosystem-based evaluation (considering economic, social ecological aspects) aimed at providing information that could be used for future management plans. However, the analysis showed that most of the scenarios considered were not feasible since they suggested unreasonable changes in the activities of some fleets, and such outcomes were considered unviable. The restoration of species such as vaquita and totoaba were just a few of the dozens of scenarios run and which produced reasonable reductions in fishing. Overall, it could be said that fishing and conservation could be compatible in the upper Gulf, but the low number of scenarios that achieved the conservation goals indicated a substantial risk not just for endangered species, such as vaquita and totoaba, but also for heavily exploited groups such as shrimps and sharks. After an exhaustive search, the overall results suggest that it is necessary to examine the tradeoffs presented among the win-win scenarios in more detail in order to obtain more realistic solutions. In the case of the exploratory analysis of the climate influence on the food web of the upper G u l f of California, the results obtained must be taken as preliminary, and they are far from being used as a confident prediction of future changes in the abundance of the stocks or in the population of species in the upper G u l f during following E l Nino events. There are some considerations that were raised during this analysis: it is important to consider that fish populations are not in a steady state because they are affected by spatial and temporal changes in the environment, and furthermore, fish populations respond to the environment in different ways. It is recommended that future biological research focus on life cycles and that reproduction be carried out in the upper G u l f in order to more accurately evaluate global warming and future E l Nino and L a Nina events. This 210 knowledge could be used as a first attempt to forecast similar responses of commercial fish populations, as well as their food and predators. Climate and fish management is a complex interaction, and in the future, it will be necessary to devise quotas and spatial management schemes in order to respond more rapidly to early warnings and more effectively to worse risks. Finally, it could be useful to ensure that all stakeholders and coastal communities in the Gulf of California are aware of risks and agree with contingency plans in order to avoid future fishery collapses and economic crashes not just in the upper Gulf, but also for the entire Gulf of California. A cooperative international agreement assuring sufficient quantities (10-15%) of water, delivered at the right time, will be needed in order to conserve the delta and improve the current productivity of the marine ecosystem of the upper Gulf of California ecosystems for future generations. Some of the results presented in Chapter V (section 5.3) provided evidence that regional rainfall in the upper Gulf is strongly influenced by E l Nino events and P D O and that this rainfall results in increments in freshwater, sediments and nutrients delivered to the region. These waters are subsequently enriched by a growth in primary production. The relationship between climate, rainfall and primary productivity in the upper Gulf is very close and the history of the Colorado River has confirmed the effects of climate change on the delivery of nutrients and fresh water, emphasizing the need to explore the climatic phenomenon coupled with the atmosphere and marine ecosystems. Also, results from the simulations presented in Chapter V , in which the highest total biomass was found when the runoff of the Colorado River was not diverted by dams (100% undepleted), resulting in a 220% rise in the total biomass of the system. Even undepleted flows of only 1% produced increases of approximately 10% in the total biomass of the upper Gulf after 50 years, revealing the enormous role of the Colorado River in the regions' productivity. This is particularly important considering that this region provides 15% of the Mexican landings (INEGI, 2001), confirming the ecological and economical losses for Mexico during the last five decades as a result of the damming of the Colorado River by US dams. Vis-a-vis conservation considerations, it is important 211 to remember that the upper G u l f (mainly Montague, Gore and Pelicano Islands) is used by at least 74 species of birds, including approximately 200,000 wintering shorebirds that depend on the mudflats and wetlands (Mellink et al, 1996, 1997). The western population o f North America's white pelican population, which has been on the decline for many decades, relies on the Delta as a migratory stopover (Brusca et al., 2001). 6.2. Conclusions. The aim of the present research was to identify and analyze trophic changes in the upper G u l f of California resulting from water diverted by the series of dams along the Colorado River and by fishing over the last 50 years; however, we cannot begin to understand the ecosystem effects of fishing or of any other human impact until we understand the nature and history of interactions among the species or groups at all levels of the food web and the dynamics of the region. The three Ecopath models provide a summary of our current knowledge of the biomass, consumption, production, food web and tropic flows in the upper G u l f of California ecosystem occurring over the past 50 years. A n examination of overall evidence resulting from the construction of the three trophic models of the upper G u l f of California relating to the past and present confirmed some o f our knowledge, such as the important role played by sediments and nutrients delivered by the Colorado River in the structure and function of this ecosystem. The analysis undertaken showed a clearer picture of the great complexity of this ecosystem: The relevance of the lower trophic groups that were directly affected by the depletion of nutrients associated with the sediments of the Colorado River; which in turn resulted in a cascade of negative effects on the trophic web, affecting both overall productivity and fishing in the region. For example, 43 out o f the 50 groups included in the models showed at least a 20% of reduction in their biomass after a simulated 212 reduction of detritus over a period of 20-years. This is particularly important because understanding the process and interactions within the system can aid in both the future formation of long-term management policies and the restoration of the region. The chapters presented here are examples of approaches to characterizing the responses of a whole community which has changed as a result of human efforts, and they represent preliminary steps towards developing sustainable strategies for the restoration of the region. One of the goals achieved during this research was to produce models of the past state of the upper Gulf which were generated from the present, taking into consideration the former series of fisheries and climate influence both before and after the construction of the Glenn Canyon Dam in the early 1960s. During this process, a wide variety of multidisciplinary sources of information and methodologies was amalgamated with the trophic models in order to reconstruct for the first time the history of this rich marine ecosystem. Perhaps the past models were subject to some uncertainty due to climate influence, which played a critical role in the reconstruction of the upper Gulf. The past models were adjusted using a time series of climate data, incorporating the effect of several E l Nino events (1957, 183, 1993 and 1997) in the primary production estimated in the region. Some of the results presented in this research could be used as Ecosystem Reference Points (ERP), which could help in future management purposes in the same way as Biological Reference Points (BRP) proposed by Hilborn and Walters (1992) for fish stock assessments. For example, the evidence of loss in biomass and number of high trophic levels during the 50-years of modelling which suggests a 'Fishing down in the food web' phenomenon could be viewed as a symptom of ecosystem deterioration by overfishing. During the last five decades, the information provided by the trophic levels indicated important impacts by fishing, resulting in an average decline rate in the trophic levels of 0.02 TL-per decade. Also, a notable loss in the abundance of detritivores during the last 50 years was observed; it was estimated that a 64% reduction in the total biomass of all the groups located in trophic level 2 and 2.5 took place. This loss is affecting the incorporation of organic matter to higher trophic levels of the food web which is a critical function of any ecosystem. Overall, the numerous changes in the energy flows and 213 production rates which were estimated during the research indicated an imminent loss in the energy and dynamics of the upper Gulf. These changes affect the structure and function of this ecosystem, limiting its ability in the future to recover to its natural state. This research proved that human intervention can trigger rapid and significant changes in the structure and function of small marine ecosystems such as the upper Gulf of California. Based on the results illustrated by this research, it is possible to visualize the future of the upper Gulf of California. Basically, there are two possible paths that the fisheries in the upper Gulf may take in the next decades. One is to continue with the same trend observed during the last five decades, in which intense harvesting and fishing pressure on all levels of the food chain will remain high, and industrial and small scale fleets will continue to catch the maximum allowed by the system in a short term. This would result in a sequential depletion of the resources, and the fisheries would then move lower down in the food web of the upper Gulf. It is predicted that fish and invertebrates species that are currently discarded or have no commercial value would thus dominate the composition of the catch and future markets of the upper Gulf. Alternatively, the research conducted in this thesis confirms the need to integrate ecological elements such as the construction of dams or reductions in fishing pressure into future changes in the region. In the future, the maximization of yield in the upper Gulf must be established under an ecosystem-based management with all the uncertainties imposed by climate changes. This approach must be a priority on the stakeholders' agenda because, i f it is not a priority, there is a risk that the fisheries in the upper Gulf will attain neither stability nor profitability in the future. 214 References. Abitia-Cardenas, L . A . , J. Rodriguez-Romero and F. Galvan-Magana. 1990. Observaciones troficas de tres especies de peces de importancia comercial de Bahia Conception, Baja California sur, Mexico. Investigaciones Marinas. CICFMAR, 5(1): 57-61. Ainsworth, C. and T. J. Pitcher. 2004. Modifying Kempton's Biodiversity index for use with dynamic ecosystem simulation models, p. 91-93. In: Pitcher, T. (ed) Back to the Future: advances in methodology for modelling and evaluating past ecosystems as future policy goals. Fisheries Center Research Reports, 12(1): 34-39. Allen, R. 1971. Relation between production and biomass. Journal of the Fisheries Research Board of Canada, 28: 1573-1581. Alvarez-Borrego, S. 1983. Gulf of California, p. 77-92. In: B . H . Ketchem (Ed). Estuaries and Enclosed Seas. Elsevier Scientific Pub. Co. New York. U S A . Alvarez-Borrego, S. 1999. The Colorado River Estuary and upper Gulf of California, Baja, Mexico. Ecological Studies, 144: 331-340. Alvarez-Borrego, S., B . P. Flores-Baez, and L . A . Galindo-Bech. 1975. Hidrologia del alto Golfo de California LT. Condiciones durante invierno, primavera y verano. Ciencias marinas, 2(1): 21-36. Alvarez-Borrego, S. and R. Schwartzlose. 1979. Masas de agua del Golfo de California. Ciencias Marinas, 6: 43-63. Alvarino, A . 1969. Zoogeografia del Mar de Cortes: Quetognatos, Sifonoforos y Medusas. Anales del Instituto de Biologia. Universidad Autonoma Nacional de Mexico. Series. Ciencias del Mar y Limnologia, 1: 11-54. Andrews, E . D . 1991. Sediment transport in the Colorado River Basin, p. 221-234. In: Colorado River Ecology and dam management. Proceedings of a symposium, May 24-25, 1990, Santa Fe, N M . Nazt. Acad. Press, Washington, D C . U S A . Anonymous. 2002. Shrimpers and Mexican government compromise on fishing in reserve. M P A News, 4 (5): http://depts.washington.edu Anonymous. 2003. Mexico faces tough battle in saving marine reserve. Planet-Save Network. Http://www.planetsave.com. Oct-28, 2002. Anonymous. 2004. Current Animations of SST and O R L : SST animations of recent E l Nino (warm) events. Sonora Desert Coastal Conservation, www.cdc.noaa.gov. Anuario de Pesca. 1996. Anuario de Estadistica Pesquera. Secretaria del Medio Ambiente, Recursos Naturales y Pesca, Mexico. 141p. Aragon-Noriega, E . A . 2000. Ecologia del reclutamiento del camaron azul Litopenaeus stylirostris (Stimpson, 1871) en el alto Golfo de California. Ph.D. Thesis. Centra de Investigation Cientifica y Education Superior de Ensenada. Ensenada, B . C . Mexico. 117 p. Aragon-Noriega, E . A . , C. Cervantes-Valle, A . R . Garcia-Juarez and L . E . Calderon-Aguilera. 1999. Distribution y abundancia de la poblacion desovante de camarones del norte del Golfo de California durante el verano de 1996. Ciencia y Mar, 3(9): 37-48. Aragon-Noriega, E . A . and L . E . Calderon-Aguilera. 2000. Does damming of the Colorado River affect the nursery area of blue shrimp Litopenaeus stylirostris (Decapoda: 215 Penaeidae) in the upper Gulf of California? Revista de Biologia Tropical, 48(4): 867-871. Aragon-Noriega, E . A . and L . E . Calderon-Aguilera. 2001. Age and growth of shrimp (Decapoda: Penaeidae) postlarvae in the upper Gulf of California. Journal of Ichthyology and Aquatic Biology, 4(3): 99-104. Aragon-Noriega, E . A . and Garcia-Juarez. 2002. Postlarvae recruitment of blue shrimp Litopenaeus stylostris (Stimpson, 1871) in anti-estuary conditions due to anthropogenic activities. Hidrobiologica, 12(1): 37-46. Arellano-Martinez, R., M . J. De L a Cruz-Aguero and J. Gomez. 1997. New records and range extensions of fishes on the Pacific coast of Mexico. Revista de Biologia Tropical, 45(2): 936-947. Argote, M . L . , A . Amador, and M . F . Lavin. 1998. Barotropic eulerian residual circulation in the Gulf of California due to M2 tide and wind stress. Atmosfera, 11: 173-197. Arreguin-Sanchez, F. 1999. Octopus-red grouper interaction in the exploited ecosystem of the northern continental shelf of Yucatan, Mexico. Ecological Modelling, 129: 119-129. Arreguin-Sanchez, F . , E . Valero-Pacheco and E . A . Chavez. 1993a. A trophic box of the coastal fish communities of the southwestern Gulf of Mexico, p. 197-205. In: V . Christensen and D . Pauly (Eds), Trophic models of Aquatic Ecosystems. 1993. I C L A R M . Manila, Philippines. Arreguin-Sanchez, F. , J .C. Seijo and E . Valero-Pacheco. 1993b. A n application of Ecopath II to the North continental Shelf Ecosystem of Yucatan, Mexico, p. 269-278. In: V . Christensen and D. Pauly (Eds), Trophic models of Aquatic Ecosystems. 1993. I C L A R M . Manila, Philippines. Arreguin-Sanchez, F. , M . Contreras, V . Moreno, R. Burgos, and D . V aides. 1996. Population dynamics and stock assessment of the red grouper (Eppinephelus morio) fishery on the Campeche Bank, p. 449-467. In: Arreguin-Sanchez, F. , J .L. Munro, M . Balgos, D . Pauly (Eds). Biology of tropical groupers and snappers. 48. International Center for Living Aquatic Resources Management, Manila, Philippines. Arreguin-Sanchez, F . , E . Arcos and E . A . Chavez. 2002. Flows of biomass and structure in an exploited benthic ecosystem in the Gulf of California, Mexico. Ecological Modelling, 156: 167-183. Arthington, A . H . 1995. The condition of large river systems of the world, p 44-75. In: N.B. Armantrouth (Ed). Condition of the World's aquatic habitats. Proceedings of the World Fisheries Congress, Theme I. Science Publishers Inc. Lebanon, U S A . Arvizu-Martinez, J. and H . Chavez. 1970. Synopsis of the biology of totoaba (Cynoscion macdonaldi, Gilbert, 1890). F A O Fisheries Synopsis, 108. 74 p. Arvizu-Martinez, J. 1987. Fisheries activities in the Gulf of California, Mexico. CalCOFI Report, 28: 32-36. Aurioles, D. , B.J. Le Boeuf and L . T . Findley. 1993. Registros de pinipedos poco comunes para el Golfo de California. Revista de Investigation Cientifica. Universidad Autonoma de Baja California Sur, 1(1): 13-19. Azuela de la Cueva, A . , V . Ramirez Navarro, J .L. Calderon, R. Gluyas, A . Fuentes, A . O. Sandoval, and M . Celis. 1995. Mortandad de Mamiferos marinos y aves marinas en el alto Golfo de California. Informe Final. Procuraduria Federal de Protection al 216 Ambiente. Secretaria del Medio Ambiente, Recursos Naturales y Pesca. 39p. Baddyr, M . 1991. Mortality of the squid, Loligo vulgaris in relation to the artisanal fishing site in the south-east of South Africa. S.Afr. J. Sci., 9: 11-26. Ballinas, V . 2002. They request to Semarnat and Sagarpa to avoid violations to guarantees of fishermen. L a Jornada News (Mexico). April 13:11-12. Barbosa-Ledesma, I.F., J .A. Gamboa-Contreras and J. A . Ake-Castillo. 2000. Stomapods (Crustacea: Hoplocarida) from the Gulf of Tehuantepec, Mexico. Ciencias Marinas, 26: 169-181. Barrera-Guevara, J .C. 1990. The conservation of Totoaba macdonaldi (Gilbert), (Pisces: Sciaenidae) in the Gulf of California, Mexico. Journal of Biology, 37: 201-202. Barthem, R.B . , M . C . Ribeiro and M . Petrere, Jr. 1991. Life strategies of some long-distance migratory cat-fish in relation to hydroelectric dams in the Amazon Basin. Biological Conservation, 55: 339-345. Baum, J .K. , R . A . Myers, D . G . Kehler, B . Worm, S.J. Harley, and P.A. Doherty. 2002. Collapse and conservation of shark populations in the northwest Atlantic. Science, 299: 389-392. Baum, J.K. and R. A . Myers. 2004. Shifting baselines and the decline of pelagic sharks in the Gul f of Mexico. Ecol. Lett. 7: 135-145. Baumgartner, T. and N . Christensen. 1985. Coupling the Gulf of California to large-scale interannual climatic variability. J. Mar. Res. 43: 825-848. Baumgartner, T. , V . Ferreira-Bartrina, H . Scharader and A . Soutar. 1985. A 20 year varve record of siliceous phytoplankton variability in the Central Gulf of California. Marine Geology, 64: 113-129. Beland, R. 1953. The effect of channelization on the fishery of the lower Colorado River. California Fish and Game, 30 (3): 137-139. Berg, B . L (Ed.). 2001. Qualitative research methods for the social sciences. Fourth Edition. Needham Heights, C A . U S A : 126p. Berkes, F . and C . Folke, (Eds.). 1998. Linking social and ecological systems: Management practices and social mechanisms for building resilience. Cambridge: Cambridge University Press. U S A . 334 p. Berkes, F . , J. Colding, and C . Folke. 2000. Rediscovery of traditional ecological knowledge as adaptive management. Ecological Applications, 10: 1251-1262. Berkes, F . , J. Colding, and C. Folke. (Eds.). 2003. Navigating social-ecological systems: Building resilience for complexity and change. Cambridge: Cambridge University Press. 13 lp. Boyer, P. 1996. History of Puerto Pefiasco. C E D O (Centro de Estudiosde Desiertos y Oceanos, Sonora, Mexico) News, 9(2): 15-18. Bowen, T. S. 2004. Archaeology, biology and conservation on islands in the Gulf of California. Environmental Conservation, 31(3): 199-206. Brey, T. 1999. A collection of empirical relations for use in ecological modelling. Naga, The I C L A R M Quarterly, 22(3): 24-28. Briggs, M . K . and S. Cornelius. 1998. Opportunity for ecological improvement along the lower Colorado River and its delta. Wetlands, 18(4): 513-529. Brinton, E . , A . Fleminger, and D . Siegel-Causey. 1986. The temperate and tropical 217 planktonic biotas of the Gulf of California. CalCOFI Report, 27: 228-269. Brito-Castillo, L . , A . Leyva-Contreras, A . V . Douglas and D . Lluch-Belda. 2002. Pacific Decadal Oscillation and the filled capacity of dams on the rivers of the Gulf of California continental watershed. Atmosfera, 15: 121-138. Brownell, R. Jr. 1986. Distribution of the Vaquita, Phocoena sinus, in Mexican waters. Marine Mammal Science, 2(4): 299-305. Brucker, A . W . , K . A . Johnson and J.D. Field. 2003. Conservation strategies for sea cucumbers: can a CITES Appendix II listing promote sustainable international trade?. SPC Benche-de-Mer information Bulletin, 18: 24-33. Brusca, R . C . , J. Campoy-Fabela, G . Castillo-Sanchez, R. Cudney-Bueno, L . T . Findley, M . A . Cisneros-Mata, M . Nevarez-Martinez, and G. Hammann. 1995. The rise and fall of the Pacific sardine, Sardinops sagax caeruleus Girad, in the Gulf of California, Mexico. CalCOFI Report, 36: 136-142. Brusca, R . C . , J. Campoy-Fabela, G . Castillo-Sanchez, R. Cudney-Bueno, L . T . Findley, J. Garcia-Hernandez, E . Glenn, I. Granillo, M . E . Hendricks, J. Murrieta, C . Nagel, M . Roman and P. Boyer. 2001. A case of two Mexican biospheres reserves. The upper Gul f of California/Colorado River Delta and Pinacate/Gran Desierto de Altar Biosphere Reserves. Report for the Unidad C I A D Guaymas, Sonora, Mexico. 96 p. Brush, S.B. and D. Stabinsky. 1996. Valuing local knowledge. Island Press. Covelo, California. U S A . 305 p. Bunce, L . , P. Townsley, R. Pomeroy and R. Pollnac. 2000. Socioeconomic manual for coral reef management. I U C N . The World Conservation Union. Australia Institute of Marine Science. Townsville, Australia. 62 p. Calderon-Aguilera, L . E . , E . A . Aragon-Noriega, H . A . Licon, G . Castillo-Moreno and A . Maciel-Gomez. 2002. Abundance and composition of penaeid postlarvae in the upper Gul f of California. Contribuciones al estudio de los crustraceos del Pacifico Este, 1:176-194. Campoy-Fabela, J. 2001. L a vaquita de mar (Phocoena sinus) en la reserva de la biosfera alto Golfo de California y delta del Rio Colorado. Informe Inedito. Mexico. 13 p. Campoy-Fabela, J. 2002. Situation de la problematica relacionada con la flota mayor de puerto Pefiasco y la proxima temporada de pesca de camaron 2002-2003. S E M A R N A T . Mexico. Nota informativa 28: 2-3. Campoy-Fabela, J. 2003. Evaluando el impacto de la red de arrastre en la Reserva de la Biosfera alto Golfo de California y delta del Rio Colorado: consideracones biofisicas y socioeconomicas. S E M A R N A T . Mexico 76p. Carbajal, N . , A . Souza, and R. Durazo. 1997. A numerical study of the ex-ROFI of the Colorado River. Journal of Marine Systems, 12: 17-33. Carpenter. S. R., J. F. Kitchell and J.R. Hodgson. 1985. Cascading trophic interactions and lake productivity. Bioscience, 35: 634-639. Carraquiry, J. D. and A . Sanchez. 1999. Sedimentation in the Colorado River delta and upper Gulf of California after a nearly a century of discharge loss. Marine Geology, 158: 125-145. C E D O (Centra de Estudios de Desiertos y Oceanos, Mexico). 1994. L a vaquita del Golfo de California. C E D O , Intercultural A . C . Serie Educational. Puerto Pefiasco, Sonora Mexico. 29 p. 218 CephBase. 2001. CephBase team. Census of marine Life, http://www.cephbase.dal.ca Cetacea. 2001. Supported by: Whale & Dolphin Conservation Society, Environmental Investigation Agency, Nature net & Discovery Channel online. Cheshire, United Kingdom, http://www.cetacea.org/. Cisneros-Mata, M . A . , G . Montemayor-Lopez and M . J . Roman-Rodriguez. 1995. Life history and conservation of Totoaba macdonaldi. Conservation biology, 9(4): 806-814. Cisneros-Mata, M . A . , M . Nevarez-Martinez and G . Hammann. 1996. The raise and fall of the Pacific sardine, Sardinops sagax caeruleus Girard, in the Gulf of California, Mexico. CalCOFI Report, 36: 136-142. Cisneros-Mata, M . A . , L . W . Botsford and J. F. Quinn. 1997. Projecting viability of Totoaba Macdonaldi, a population with unknown age-dependent variability. Ecological Applications, 7(3): 968-980. Chao, L . N . 1995. Sciaenidade. Corvinas, barbiches, bombaches, corvinatas, corvinetas, corvinillas, lambes, pescadillas, roncachos, verrugatos, p. 34-57. In: Fischer, W. , F. Krupp, W. Schneider, C. Sommer, K . E . Carpenter & V . Niem (Eds.) Guia F A O para identification de especies para los fines de la pesca. Chavez, H . and J. Arvizu. 1970. Estudio de recursos demersales del Golfo de California, 1968-1969 III. Fauna de Acompanamiento del camaron (peces finos y 'basura'). Recursos pesqueros III. Reporte Final. Secretaria de Pesca. Mexico. 113p. Chavez, E . A . , M . Garduno and F. Arreguin-Sanchez. 1993. Trophic Dynamic structure of Celestum lagoon, southern Gulf of Mexico, p. 45-53. In: Christensen, V . and D . Pauly (Eds). Trophic models of Aquatic Ecosystems. 1993. I C L A R M . Manila, Philippines. 389 p. Christensen, V . 1995. Ecosystem maturity - towards quantification. Ecological Modelling 77: 3-32. Christensen, V . and D . Pauly. 1992. Ecopath II. A software for balancing steady state ecosystem models and calculating network characteristics. Ecology Modelling, 61: 169-185. Christensen, V . and D . Pauly. 1993. Flow characteristics of aquatic ecosystems, p.213-231. In: Christensen, V . and D. Pauly (Eds). 1993. Tropical models of aquatic Ecosystems. International Centre for Living Aquatic Resources Management ( I C L A R M ) , Manila, Philippines. 389 p. Christensen, V . , C . J. Walters and D. Pauly. 2000. Ecopath with Ecosim Version 4. Help system. University, of British Columbia, Fisheries Centre, Vancouver, Canada and I C L A R M , Penang , Manila. 105 p. Christensen, V . , S. Guenette, J.J. Heymans, C.J. Walters, R. Watson, D . Zeller, and D . Pauly. 2003. Hundred-year decline of North Atlantic predatory fishes. Fish and Fisheries, 4: 1-24. Christensen, V . and C . J. Walters. 2004. Ecopath with Ecosim: methods, capabilities and limitations. Ecological modelling, 172: 109-139. Christensen, V . , C . J. Walters and D. Pauly. 2004. Ecopath with Ecosim: a user's guide. Fisheries Centre, University of British Columbia, Vancouver, Canada. 154 p. Chute, G.R. 1928. The totoaba fishery of the Gulf of California. Cai. Fish and Game, 14(4): 275-281. 219 Chute, G.R. 1930. A regal-soup stock. Cai. Fish and Game, 16(1): 23-35. Clemente-Rojo, M . 1967. Apuntes historicos de la frontera de la Baja California. Museo de Historia de Ensneada, Baja California, Mexico. 25 p Cohen, M . J. , C . Henges-Jeck, G . Castillo-Moreno. 2001. A preliminary water balance for the Colorado River delta, 1992-1998. Journal of Arid Environments, 49: 35-48. Collette, B . B . 1995. Scombridae. Atunes, bonitos, caballas, estorninos, melva etc, p 1521-1543. In: Fisher, W. F. Krupp, W. Schneider, C. Sommer, K . E . Carperter, V . Niem (Eds). Guia F A O para la identification de especies para los fines de la pesca, V o l 3. Pacifico Centro-Oriental. F A O , Rome. Collette, B . B . 2003. Family Scombridae Rafinesque 1815 - mackerels, tunas, and bonitos. Cai. Acad. Sci. annoted checklist of fishes, 19: 28-42. Collette, B . B . and J.L. Russo, 1984. Morphology, systematic and biology of the Spanish mackerels (Scomberomorus, Scambridae). Fish Bull. 82(4): 545-692. Compagno, L . J . V . 1999. Check list of living elasmobranches, p. 471-498. In: W. C. Hamlett (Ed). Sharks, skates, and rays: the biology of elasmobranch fishes. John Hopkins University Press, Maryland. U S A . Compagno, L . J . V . , F. Krupp and W. Schneider. 1995. Tiburones, p. 647-744. In: Fisher, W. F. Krupp, W. Schneider, C . Sommer, K . E . Carperter, V . Niem (Eds). Guia F A O para la identification de especies para los fines de la pesca, 3. Pacifico Centro-Oriental. F A O , Rome, Italy. Craig, J. A.1926. A new Fishery in Mexico. California Fish and Game, 12(4): 166-169. Craig, W. l . 1960. Food and feeding, p. 35-46. In: J. L . Baxter (Ed). A study of the yellowtail tuna. Fishbull. 110. Dept. O f Fish and Game. Marine Resources operations. CRIP-Ensenada. 1995. Atlas pesquero de la fauna de acompanamiento del camaron del alto Golfo de California. Ensenada, Baja California, Mexico. http ://inp. semarnap. gob. mx Crustacean. 2001. Crustacean Society website. Virginia Institute of Marine Science. http://www.vims.edu/tcs/ Cudney-Bueno, R. and P. Turk. 1998. Pescando entre mareas del alto Golfo de California. C E D O (Centro de Estudios de Desiertos y Oceanos, Mexico). 154 p. Cudney-Bueno, R. 2000. C E D O and fisheries: a historical perspective. C E D O (Centro de Estudios de Desiertos y Oceanos, Mexico) News, 2(9): 13-19. Cupp, E . E . , and W. E . Allen. 1944. Plankton diatoms of the Gulf of California obtained by Allan Hancock Pacific expedition of 1937. Contributions of the Scripps Institutions of Oceanography. New Series, 14: 61-73. Cupul, A . 1994. Flujos de sedimentos en suspension y nutrientes en la cuenca estuarina del Rio Colorado. MSc . Thesis, Universidad Autonoma de Baja California, Mexico. 148 p. D'Agrosa, C , O. Vidal and W. C . Graham. 1995. Mortality of the Vaquita (Phocoena sinus) in gillnet fisheries during 1993-94. Rep. Int. Whal. Commn. (Special issue 16): 281-289. D'Agrosa, S., C . Lennert-Cody and O. Vidal. 2000. Vaquita bycatch in Mexico's artisanal gillnets fisheries: driving a small population to extinction. Conservation Biology, 14:110-119. Dale, A . R., E . DeMartin and K . Plummer. 1984. The feeding habits of juvenile -small 220 adult barred sand bass (Paralabrax nebulifer) in near shore waters off Northern San Diego County. C a L C O F I Reports, 25: 105-111. Diario Oficial de la Federacion. 1955. Decreto por el que se establece la zona de la desembocadura del Rio Colorado y delta del Rio Colorado como zona de refugio. Diario Oficial de la Federacion, Mexico. 19 de Septiembre, 1955: 2-3. Diario Oficial de la Federacion. 1975. Decreto por el que se establece veda indefinida de la pesqueria de totoaba en el Golfo de California. Diario Oficial de la Federacion. Mexico, 1 de agosto de 1975:11-13. Diario Oficial de la Federacion. 1992. Secretaria de Pesca. Diario Oficial de la Federacion. Mexico D .F . 13 de Febrero: 32-34. Diario Oficial de la Federacion. 1993. Decreto por el que se declara area natural protegida con el caracter de Reserva de la Biosfera, la region conocida como alto Golfo de California y Delta del Rio Colorado, ubicada en las aguas del Golfo de California y los municipios de Mexicalli, B . C . , Pefiasco y San Luis Rio Colorado, Sonora. Diario Oficial de la Federacion. Mexico, Jueves 10 de Junio de 1993: 24-28. Delgado-Estrella, A . , J. G . Ortega-Ortiz and A . Sanchez-Rios. 1994. Varamiento de mamiferos marinos durante la primavera y otofio y su relation con la actividad humana en el norte del Golfo de California. Anales del Instituto de Biologia. Universidad National Autonoma de Mexico. Serie Zoologia, 65(2): 287-295. Ellen, R., P. Parkes and A . Bicker (Eds). 2000. Indigenous Environmental Knowledge and its transformations. Critical Anthropological perspectives. University of Kent at Canterbury. U K . 329 p. Eschmeyer, W. N . , E . S. Herald and H . Hammann. 1983. A field guide to Pacific coast fishes of North America. Houghton Mifflin Company, Boston, U .S .A . 336 p. F A O . 1995. Catalogo para las especies susceptibles a la pesca de la region Pacifico Centro-Oriental. I. Invertebrados. Roma, Italia. 365 p. Farfan, C . and S. Alvarez-Borrego. 1992. Zooplankton biomass of the northernmost Gulf of California. Ciencias Marinas, 18(3): 17-36. Felix-Pico, E . 1973. Agrupaciones de fauna (Macro invertebrados). Section IV. Biologia, p. 254-261. In: Universidad Autonoma de Baja California. Unidad de Ciencias Marinas. Reporte final a la Direction de Acuacultura de la Secretaria de Recursos Hidraulicos de la segunda etapa del estudio quimico sobre la contamination por insecticidas en la desembocadura del rio Colorado. Tomo II. Ensenada, B . C . Mexico. 394 p. Felix-Pico, E . F. 1975. Aspectos de la Ecologia de macroinvertebrados con referenda especial al camaron en la desembocadura del Rio Colorado en el alto Golfo de California. Thesis. Universidad Autonoma de Baja California, Mexico. 178 p. Fernandez-Alamo, M . A . 1992. Distribution y abundancia de los poliquetos (Annelidae-Polychaeta) en el Golfo de California, Mexico, durante los meses de Marzo y Abril de 1984. Investigaciones Marinas, CICLMAR, 7(2): 75-89. Fernadez-Barajas, M . E . , M . A . Monreal-Gomez and A . Molina-Cruz. 1994. Thermohaline structure and Geotrophic flow in the Gulf of California during 1992. Ciencias Marinas, 20: 267-286. Finn, J.T. 1980. Flow analysis of models of the Hubbard Brook ecosystem. Ecology, 6: 562-571. 221 FishBase. 2006. FishBase: A global System Information on Fishes, www.fishbase.org Flanagan, C A . and J.R. Hendrickson. 1976. Observations on the commercial fishery and reproductive biology of totoaba (Cynoscion macdonaldi) in the northern Gulf of California. Fishery Bulletin, 74: 531-544. Flessa, K . , Schoene, B . and Kowalewski, M . 2001. New uses for the dead: Reconstructing baseline conditions on the Colorado delta. Annual Meeting, Society for Conservation Biology. July 2001. 11 p. Folke, C , J. Colding, and F. Berkes. 2003. Synthesis: building resilience and adaptive capacity in social-ecological systems, p. 352-387. In: F. Berkes, J. Colding and C . Folke (Eds.). Navigating social-ecological systems: Building resilience for complexity and change. Cambridge: Cambridge University Press. U .S .A . 469 p. Fradkin, P. L . 1981. A river no more: the Colorado River and the West. Alfred Knop. New York. U S A . 360 p. Francis, R. C . and S. Hare. 1994. Decadal scale regime shifts in the large marine ecosystems of the north-east Pacific: a case for historical sciences. Fisheries Oceanography, 3: 279-291. Frose, R. and D . Pauly. 2000. Fishbase 2000: concepts, design and data sources. I C L A R M . Philipinnes. 344 p. Frose, R. and D . Pauly (Eds). 2001. Fishbase. World Wide Web electronic publication. www.fishbase.org Galindo-Bect, M . , E.P. Glenn, H . Page, L . A . Galindo-Bect, J. Ayon-Hernandez, R . L . Petty, J. Garcia-Hernandez and D. Moore. 2000. Penaeid shrimp landings in the upper Gulf of California in relation to Colorado River freshwater discharge. Fishery Bulletin, 98: 222-225. Galvan-Magana, F. , and H . J. Nienhuis. 1989. Seasonal abundance and feeding habits of sharks of the lower Gulf of California, Mexico. California Fish and Game, 75(2): 74-84. Gallardo-Cabello, M . , M . Jacob-Cervantes and X . Chiappa-Carrera. 1991. Analisis de los habitos alimentarios y del aporte de carbono de las presas a la dieta de la sardina crinuda, Opisthonema libertate, gunther 1866 (Pisces: Clupeidae) en el Golfo de California. Anales del Instituto de Ciencias del Mar y Limnologia. Universidad Nacional Autonoma de Mexico, 8(1): 37-48. Garcia, A . 1991. Spawning stock recruitment of white shrimp in the southwerstern Gulf of Mexico. Trans-America. Fisheries Society, 120(4): 519-527. Garcia-Borbon J .A. , J. Madrid-Vera, S. Ramos-Cruz, A . R. Garcia Juarez, A . R . Chavez-Herrera, D . Castro-Castro, E . Zarate-Becerra, L . Belendez-Moreno, E . Morales-Bojorquez, P. Loreto-Campos, S. Hernandez, A . Ramos-Montiel, M . Anguiano-Carrazco and C . Alvarado-Sarabia. 2004. Initio de la temporado de veda 2004 para la pesqueria de camaron del Oceano Pacifico Mexicano. Instituto Nacional de la Pesca. S A G A R P A . Informe de Investigation. Mazatlan, Sinaloa, Mexico. 26 p. Garcia-Caudillo, J . M . , M . A . Cisneros-Mata and A . Balmori-Ramirez. 2000. Performance of a bycatch reduction device in the shrimp fishery of the Gulf of California, Mexico. Biological Conservation, 92: 199-205. Garcia-Gomez, M . 1976. Breve analisis de cuatro temporadas de pesca camaronera en Puerto Pefiasco, Sonora (Mexico). Thesis. Escuela Nacional de Ciencias Biologicas. 222 Instituto Politecnico Nacional. Mexico. 79 p. Garcia-Pamanes, J. and J. R. Lara-Lara. 2001. Microzooplankton grazing in the Gulf of California. Ciencias Marinas, 27(1): 73-90. Garcia-Rodriguez, F.J. 1999. Cambios espaciales y estacionales en la estructura trofica y consumo del lobo marino de California, Zalophus californicus, en la region de las grandes islas, Golfo de California. Thesis. Centro Interdisciplinario de Ciencias Marinas-INP. L a Paz, B C S , Mexico. 89 p. Gilbert, J. Y . And E . Allen. 1943. The phytoplankton of the Gulf of California obtained by the ' E . W. Scripps' in 1939 and 1940. Journal of Marine Research, 5(2): 89-111. Glenn, E . P., R. S. Felger, A . Burquez and D. S. Turner. 1992. Cienega de Santa Clara: endangered wetland in the Colorado River delta, Sonora, Mexico. Natural resources journal, 32:817-824. Glenn, E . P., C. Lee., R. Felger, S. Zengel. 1996. Effects of the water management on the wetlands of the Colorado River delta, Mexico. Conservation biology, 10: 1175-1186. Glenn, E . P., J. Garcia-Hernandez, C. Congdon and D. Lucke. 1999. Status of the wetlands supported by agricultural draining water in the Colorado River delta, Mexico. Horticultural Science, 34: 16-21. Glenn, E.P. , F. Zamora-Arroyo, P. L . Nagler, M . Briggs, W. Shaw, and K . Flessa. 2001. Ecology and conservation biology of the Colorado River delta, Mexico. Journal of Arid Environments, 49: 5-15. Gomez-Pompa, A . and R. Dirzo. 1995. Reservas de la biosfera y otras areas naturales protegidas de Mexico. C O N A B I O . 67 p. Graf, G . 1992. Benthic-pelagic coupling: a benthic view. Oceanography and Marine Biology, 30: 149-190. Grande-Vidal, J . M . 1980. Evaluation biotecnologica de los recursos demersales vulnerables a redes de arrastre en el fondo del Golfo de California 1978-1980. Secretaria de Pesca. Informe Tecnico 3. INP. Mexico. 27 p. Gul, K . J. 1961. Lower California. Guidebook. Clark Company. California, U S A . 213 p. Gulland, J.A. 1971. Manual de metodos para la evaluation de poblaciones de pesca. Edited by Arancibia and F A O . 85 p. Guevara-Escamilla, S., M . A . Huerta-Diaz, E . Felix-Pico, C . Farfan and C . Mathews. 1973. Estudio quimico sobre la desembocadura del Rio Colorado. Tomo JJ. Reporte final a la direction de Acuacultura de la Secretaria de Recursos Hidraulicos, Universidad Autonoma de Baja California. Erisenada. Baja California, Mexico. 173 p. Hare, S. R. and N . J. Mantua. 2000. Empirical evidence for north Pacific regime shifts in 1977 and 1989. Progress in oceanography, 47: 103-145. Hayward, T . L . 1997. Pacific Ocean climate change: atmospheric forcing, ocean circulation and ecosystem response. Trends in Ecology and Evolution, 12: 150-154. Hendrickx, M . E . 1985. Diversidad de los macroinvertebrados bentonicos acompanantes del camaron en el area del Golfo de California y su importancia como recurso potential. Chapter 4: 95-148 p. In: Yanez-Arancibia, A . (Ed). Recursos pesqueros potenciales de Mexico: La pesca acompanante del Camaron. Uiversidad Nacional Autonoma de Mexico. Hendrix, M . E . 1990. The stomatopod and decapod crustacean collected during 223 G U A Y T E C JJ cruise in the Central Gulf of California, Mexico with the description of a new species of Plesionika bate (Caridae: Pandalidae). Revista de Biologia Tropical, 38:35-53. Hendrickson, J. R. 1973. Study of the marine environment of the northern Gulf of California. Thecnical Report to Goddard Spaceflight Center, Greenbelt, Maryland. U S A . 106 p. Hensley, D . A . 1995. Paralichtydae. Lenguados, p. 459-471. In: Fisher,W. F. Krupp, W . Schneider, C. Sommer, K . E . Carperter, V . Niem (Eds). Guia F A O para la identification de especies para los fines de la pesca, V o l 3. Pacifico Centro-Oriental. F A O , Rome. Herendeen, R. 1989. Energy intensity, residence time, energy and ascendancy in dynamic ecosystems. Ecology Modelling, 48: 19-44. Hernandez, G . , E . Glenn, I. Granillo, M . E . Hendricks, J. Murrieta, C . Nagel, M . Roman and P. Turk, J. Campoy-Fabela. 2002. Evaluando el impacto de la red de arrastre en la Reserva de la Biosfera alto Golfo de California y delta del Rio Colorado: consideraciones biofisicas y socioeconomicas. S E M A R N A T . Mexico. 67p. Hernandez, A . and W. Kempton. 2003. Changes in fisheries management in Mexico: Effects of increasing scientific input and public participation. Ocean and Coastal Management, 46: 507-526. Hernandez-Ayon, M . , S. Galindo-Bect, S. Alvarez-Borrego. 1993. Nutrient concentrations are high in the turbid waters of the Colorado River delta. Estuarine, Coastal and Shelf Science, 37: 593-602. Hernandez-Trujillo, S. and A . Esquivel-Herrera. 1989. Los copepodos y quetognatos del Golfo de California en la primavera de 1985. Investigaciones Marinas. CICEVIAR, 4(2): 151-164. Hilborn, R. And C. J. Walters. 1992. Quantitative Fisheries Stock Assessment. Choice, Dynamics and Uncertainty. Chapman and Hall, Routledge, 569 p. Hobson, E.S. 1976. The rockwrasse, Halichoeres semicinctus, as a cleaner-fish. California Fish and Game, 62(1): 73-78. Hohn, A . A . , A . J. Read, S. Fernandez, O. Vidal and L . T. Findley. 1996. Life history of the vaquita, Phocoena sinus (Phocoenidae, Cetacea). Journal of Zoology, 239: 235-251. Hutchings, J .A. , C . Walters and L . R. Haedrich. 1997. Is scientific inquiry incompatible with government information control?. Can. J. Fish. Aquat. Sci. 54: 1198-1210. Huntington, H.P. 2000. Using traditional ecological knowledge in science: Methods and applications. Ecological Applications, 10: 1270-1274. Hays, G . C , A . C . Broderick, B.J. Godley, P. Luschi, and W.J. Nichols. 2003. Satellite telemetry suggests high levels of fishing-induced mortality in marine turtles. Mar. Ecol. Prog. Ser., 262: 305-309. EVIMA Inc. org. 2001. International Marine Mammal Association. Online Resources for Marine Mammal Conservation, http://www.imma.org/index.html. INEGI. 2001. Direction General de Estadistica. Encuesta Nacional de Poblacion y Vivienda 2001. Instituto Nacional de Estadistica, Geografia e Informatica, Mexico. 141p. INEGI, 2002. XII Censo General de Poblacion y vivienda 2000. Instituto Nacional de 224 Geografia Estadistica e mformatica. Mexico. 171p. INP (Instituto Nacional de la Pesca). 2000. Sustentabilidad y pesca responsable en Mexico. Evaluation y manejo 1999-2000 S E M A R N A P . 104p. Jaramillo-Legorreta, A . , L . Rojas-Bracho and T. Gerrodette. 1999. A new abundance estimate for vaquitas: first step for recovery. Marine Mammal Science 15:957-973. Kempton, R . A and L . R. Taylor. 1976. Models and statistics for species diversity. Nature, 262: 816-820. Kniffin, F .B . 1931. The primitive cultural landscape of the Colorado delta. University of California, Berkley, California. U S A . 94p. Kniffin, F .B . 1932. The natural landscape of the Colorado Delta. University of California, Berkley, California. U S A . 67p. Knudson, T. 2001. A sea teeming with greed. A dying Sea. Special Report. www.sacbee.com/static/archive/news/proiects/dyingsea/2001.htlm Kowaleswski, M . , G . Avila-Serrano, K . Flessa and G . Goodfriend. 2001. Dead's delta former productivity: two trillion shells at the mouth of the Colorado River. Annual meeting Society for Conservation Biology, 32: 46-59. Lavaniegos-Espejo, B. and R. Lara-Lara. 1990. Zooplankton of the Gulf of California after 1982-1983 E l Nino event: biomass distribution and abundance. Pacific Science, 44(3): 297-310. Lavin, M . F . and S. Sanchez. 1999. On how the Colorado River affected the hydrography of the upper Gulf of California. Continental Shelf Research, 19: 1545-1560. Lavin, M . F , E . Beier and A . Badan. 1997. Estructura hidrografica y circulation del Golfo de California: escalas estacional e interanual, p. 46-53. In: Lavin, M . F . (Editor), Contribuciones ala Oceanografia Fisica en Mexico. Mono gratia, 3. Union Geofisica Mexicana. Mexico. Lavin, M . F . , V . M . Godinez and L . G . Alvarez. 1998. Inverse-estuarine features of the upper Gulf of California. Estuarine, Coastal and Shelf Science, 47: 769-795. Lavin M . F . , E . Palacios-Hernandez and C . Cabrera. 2003. Sea surface temperature anomalies in the Gulf of California. Geofisica International, 42(3): 1-13. Leal, J .C. and M . F. Lavin. 1999. Comparacion del viento costero y marino en la region norte del Golfo de California durante el invierno de 1994. G E O S , 22(1):12-17. Leon-Portilla, M . 1972. Descubrimiento en 1540 y primeras noticas de Isla Cedros. Calafia. Universidad de Baja California, Mexico. Gaceta 2: 8-10. Leopold, A . 1948. A sand county almanac. Ballantine Books. 57 p. Luecke, D. , Ch. Congdon, E . Glenn, C . Valdes-Casillas and M . Briggs. 1999. A delta once more: restoring riparian and wetland habitat in the Colorado River delta. Environmental Defense Fund. E D F publications. Washington. D C . U S A . 65p. Littler, M . M . and D . S. Littler. 1981. Intertidal macrophyte communities from Baja California and the upper Gulf of California: relatively constant vs environmentally fluctuating systems. Marine Ecology Progress series, 4: 145-158. Lindeman, R . L . 1942. The trophic-dynamic aspect of ecology. Ecology, 23: 399-418. Lluch-Belda, D . 1970. Crecimiento y mortalidad del lobo marino de California. Anuales de la Escuela Nacional de Ciencias Biologicas, 18: 25-36. 225 Lluch-Belda. D . 1977. Diagnostico, modelo y regimen O p t i m o de la pesqueria de camaron en altamar en el noreste de Mexico. PhD. Thesis. Escuela Nacional de Ciencias Biologicas. Instituto Politecnico Nacional. Mexico. 430 p. Lopez, M . 1997. A numerical simulation o f water mass formation in the northern Gulf of California during winter. Continental Shelf Research, 17(13): 1581-1607. Lopez-Martinez J. , C. Salinas-Zavala, A . Muhlia-Melo, R. Morales-Azpeitia, E . Alcantara-Razo. 1997. Evaluation del potential pesquero del camaron de roca, cacahuate o japones Sicyonia spp en el litoral Sonorense. Final report. Secretaria del Medio Ambiente Recursos Naturales y Pesca. Centro de Investigaciones Biologicas del Noroeste. Guaymas, Sonora, Mexico. 48 p. Lopez-Martinez, J. , E . Chavez, S. Hernandez-Vazquez and E . Alcantara-Razo. 1999. Potential yield of a rock, Sicyonia penicillata, in the northern Gulf of California. Crustacean, 72(6): 581-590 Lopez-Martinez J. , F. Arreguin-Sanchez, S. Hernandez-Vazquez, A . R. Garcia-Juarez, W. Valenzuela-Quinonez. 2003. Interannual variation of growth of the brown shrimp Farfantepenaeus californiensis and its relation to temperature. Fisheries Research, 61: '95-105. Luecke, D .F . , Ch. Congdon, E . Glenn, C. Valdes-Casillas and M . Briggs. 1999. A delta once more: Restoring riparian and wetland habitat in the Colorado River delta. Environmental Defense Fundation Report. 37 p. McCully, P. 1996. Silenced rivers. The ecology and politics of large dams. Zed books, London and New Jersey. 350 pp. McEachran, J.D. 1995a. Dasyatidae. Rayas-latigo, p. 752-755. In. Fisher,W. F. Krupp, W. Schneider, C . Sommer, K . E . Carperter, V . Niem (Eds). Guia F A O para la identification de especies para los fines de la pesca, V o l 3. Pacifico Centro-Oriental. F A O , Rome. McEachran, J.D. 1995b. Urolophidae. Rayas redondas, p. 786-792. In: Fisher,W. F. Krupp, W. Schneider, C. Sommer, K . E . Carperter, V . Niem (Eds). Guia F A O para la identification de especies para los fines de la pesca, Vol . 3. Pacifico Centro-Oriental. F A O , Rome. McGuire, R. T. and J.B. Greenberd (Eds). 1993. Maritime Community and Biosphere Reserve: Crisis and Response in the upper Gulf of California. Occ. Pap. 2, Bur. Applied Res. Anthro. University of Arizona, Tucson. U S A . 169 p. McGuire, R . T . and G . C. Valdez-Gardea. 1997. Endangered species and precarious lives in the upper Gulf of California. Culture and Agriculture, 19(3): 101-107. McKay, R. J. and M . Schneider. 1995. Haemulidae. Burros, corocos, chulas, gallinazos, roncos, p. 1136-1173. In: Fisher,W. F. Krupp, W. Schneider, C . Sommer, K . E . Carperter, V . Niem (Eds). Guia F A O para la identification de especies para los fines de la pesca. Pacifico Centro-Oriental. F A O , Rome. Magallon-Barajas, F . J. 1987. The Pacific shrimp fishery of Mexico. CalCOFI Report, 28: 43-52. Mangatalle, D. 1996. Coastal state requirements for foreign fishing. F A O technical paper: Legislative study. Rome, Italy. 57p. Margalef , R. 1968. Perspectives in ecology theory. University Press, Chicago. U S A . l l l p . 226 Markaida, U . and Sosa-Nishizaki, O. 2001. Reproductive biology of jumbo squid, Dosidiscus gigas, in the Gul f of California, 1995-1997. Fisheries Research, 54: 63-82. Meckel, L . D . 1975. Holocene sand bodies in the Colorado delta area, northern Gulf of California, p. 239-365. In: Broussard, M . C . (Ed.), Deltas and models for exploration. Houston Geological Society, T X . U S A , Meling-Lopez, A . E . and S.E. Ibarra-Obando. 1999. Annual life cycles of two Zostera marina populations in the Gulf of California: contrasts in seasonality and reproductive effort. Aquatic Botany, 65: 59-69. Mellink, E . , E . Palacios and S. Gonzalez. 1996. Notes of nesting birds of the Cienega de Santa Clara, northwestern Sonora, Mexico. Western Birds, 27: 202-204. Mellink, E . , E . Palacios and S. Gonzalez. 1997. Non-breeding water-birds of the delta of the Colorado River, Mexico. Journal of Field Ornothology, 68: 113-123. Meza, R. 2001. Opinion tecnica de la pesqueria riberefia de camaron en el Golfo de Santa clara (reserve de la Biosfera alto Golfo de California y delta del Rio Colorado) temporadas 1995-2001. S A G A R P A . E l Golfo de Santa Clara, Sonora, Mexico. 19 p. Millan-Nunez, E . 1992. Spatial distribution of the phytoplankton in the Northern Gulf of California. Ciencias Marinas, 18(2): 101-117. Millan-Nunez, R., E . Santamaria-del-Angel, R. Cajal-Medrano and O A . Alcides. 1999. The Colorado River delta: a high primary productivity ecosystem. Ciencias Marinas, 25(4): 509-524. Molina, R. E . and F. A . Manriquez. 1997. Stomach contents of two planktivorous fishes of the Gulf of California during summer 1991. Ciencias Marinas, 23(2): 163-174. Molina-Valdez, D and M . A . Cisneros-Mata. 1988. Prospeccion y evaluation de la totoaba (Totoaba macdonaldi) en el Golfo de California. Informe tecnico, CRIP, Guaymas. Mexico. 16p. Morales-Bojorquez, E . , M . A . Cisneros-Mata, M . Nevarez-Martinez and A . Hernandez-Herrera. 2001. Review of stock assessment and fishery biology of Dosidiscus gigas in the Gulf of California, Mexico. Fisheries Research, 54: 83-94. Morales-Zarate, M . V . 2001. Modelo ecologico de flujos de biomasa en la region norte del Golfo de California. Master Thesis. Centra de Investiaciones Biologicas del Noreste (Centra de Investigaciones Biologicas del Noroeste, Mexico). 76 p. Morales-Zarate, M . V . , F. Arreguin-Sanchez, J. Lopez-Martinez and S.E. Lluch-Cota. 2004. Ecosystem trophic structure and energy flux in the Northern Gulf of California, Mexico. Ecological Modelling, 174: 331-345. Moreno-Mena, J. A . and L . A . Suarez-Sanchez. 2002. Estudio soc ioeconomic© de la comunidad indigena Cucapa. Instituto National Indigenista. Instituto de Investigaciones Sociales-UABC. Ensenada, Baja California. Mexico. 38p. Morrison, R., R . K . Ross, and M . Torres. 1992. Aerial surveys of shorebirds wintering, p. 31-42. In:, Canadian Ministry of the Environment (Ed.). Mexico: Some preliminary results. Prog. Note, Canadian Wildlife Service. Ottawa. Canada. Morrison, J. I., S. L . Postel and G . Gleick. 1996. The sustainable use of water in the lower Colorado River basin. Pacific Institute. Oakland, California. U S A . 77 p. Moser, H . G . and E . H . Ahlstrom. 1996. Myctophidae: lanternfishes, p. 387-475. In: H . G . Moser (Ed). The early stages of fishes in the California Current region. CalCOFI Atlas, 33: 1505-1511. 227 Musick, J. and A . Harbin. 2000. Marine estuarine and diafromous fish stocks at risk of extinction in north America (exclusive of Pacific salmonids). Fisheries, 25(11): 378-386. Myers, R. A . A n d B. Worm. 2003. Rapid worldwide depletion of predatory fish communities. Nature, 423: 280-283. Nava-Romo, J. M . 1994. Impactos a corto plazo y largo plazo en la diversidad y otras caracteristicas ecologicas de la comunidad bentico-demersal capturada por la pesqueria de camaron en el norte del alto Golfo de California, Mexico. Thesis. Instituto Tecnologico y Estudios Superiores de Monterrey (Campus Guaymas, Mexico). 89 p. Nazarea-Sandoval, V . D . 1995. Local knowledge and agricultural decision making in the Philippines. Cornall University Press. U S A . 221 p. Nevarez-Martinez, M . and E . Morales-Bojorquez. 1997. E l escape proporcional y el uso del punto de referenda biologico F%BR, para la explotacion del calamar gigante, Doscidicus gigas, del Golfo de California, Mexico. Oceanides, 12(2): 83-94. Nygren, A . 1999. Local knowledge in the environment-development discourse: From dichotomies to situated knowledge. Critique of Anthropology, 19(3): 267-288. NOTJJVIEX. 2003. Combate P R O F E P A captura ilegal de especies protegidas. Notimex Mexico. April 1: 3-4. Odum, E.P. 1969. The strategy of ecosystem development. Science, 164: 264-270. Odum, E.P. 1971. Fundamentals of Ecology. W . B . Saunders, Philadelphia. P A . U S A . 574 P-Odum, E.P. 1988. Self organization, transformity and information. Science, 104: 262-270. Optiz, S. 1993. A quantitative model of the trophic interactions in a Caribbean coral reef ecosystem, p. 259-267. In: Trophic models of Aquatic Ecosystems. 1993. Edited by V . Christensen and D . Pauly. I C L A R M . Manila, Philippines. Ortiz, I. 1999. Impacts of fishing and habitat alteration on the population dynamics of the vaquita, (Phocoena sinus). Thesis. School of Aquatic and Fishery Sciences. University of Washington. U S A . 72 p. Osorio-Tafall, B .F . 1943. Algunos datos sobre la zoologia economica del mar de Cortez. Zoologia Economica, 4(3): 59-71. Paden, C . 1991. Tidal and atmospheric forcing of the upper ocean in the Gulf of California. 1. Sea surface temperature variability. Journal of Geophysical Research, 96(10): 18,337-18359. Pauly, D. 1980. On the interrelationship between natural mortality, growth parameters and mean environmental temperature in 175 fish stocks. Journal du Conseil International pour l'Exploration de la Mer, 39: 175-192. Pauly, D . 1986. A simple method for estimating the food consumption offish population from growth data of food conversion experiments. U.S. Fish. Bull. 84(4): 827-840. Pauly, D . 1995. Anecdotes and the shifting baseline syndrome of fisheries. Trend in Ecology and Evolution, 10: 430-436. Pauly, D. 1997. Small-scale fisheries in the tropics: marginality, marginalization and some implications for fisheries management, p. 40-49. In: E . L . Pikitch, D .D. Huppertand M . P. sissenwine (Eds). Global trends: fisheries management. American 228 Fisheries Society Symposium 20, Bethesda. Pauly, D . , V . Sambilay, Jr. and S. Optiz. 1993. Estimates of relative food consumption by fish and invertebrates populations, required for modelling the Boliao reef ecosystem, Phillipines, p. 236-251. In: Trophic models of Aquatic Ecosystems. 1993. Edited by V . Christensen and D. Pauly. I C L A R M . Manila, Philippines. 389 p. Pauly, D . and V . Christensen. 1995. Primary production required to sustain global fisheries. Nature, 374: 255-257. Pauly, D . , V . Christensen, J. Dalsgaard, R. Frose, and F. Torres Jr. 1998a. Fishing down marine food webs. Science, 279: 860-869. Pauly, D . , R. Frose and V . Christensen. 1998b. How pervasive is 'Fishing down marine food web?'. Science, 282: 1383-1402. Pauly, D. , R. Frose and V . Christensen and M . L . Palomares. 2000. Fishing down aquatic food webs. American Scientist. 88: 46-51. Pauly, D . , V . Christensen and C. Walters. 2000. Ecopath, Ecosim and Ecospace as tools for evaluating ecosystem impacts of fisheries. ICES Journal of Marine Science, 57: 697-706. Pedrin-Osuna, O .A . , J. H . Cordova-Murrieta and M . Delgado-Marchena. 2001. Crecimiento y mortalidad de la totoaba, Totoaba macdonaldi, del alto Golfo de California. INP. S A G A R P A . Mexico. Ciencia Pesquera, 15: 131-140. Peresbarbosa, E . and E . Mellink. 1994. More records of breeding birds from Montague Island, northern Gulf of California. Western Birds, 25: 201-203. Perez-Cruz, L . L . , and Molina-Cruz, A . 1988. E l Nino 1983: Effect on the distribution of the silicoflagellates in the Gul f of California. Ciencias Marinas, 14(3): 9-38. Perez-Espana and Arreguin-Sanchez. 2001. A n inverse relationship between stability and maturity in models of aquatic ecosystems. Ecological Modelling, 145: 189-196. Perez-Mellado, J. 1980. Analisis de la fauna de acompanamiento del camaron capturado en las costas de Sonora y Sinaloa, Mexico. Master thesis. Instituto Tecnologico de Estudios Superiores de Monterrey, Sonora, Mexico. 86 p. Perez-Mellado, J. And L . T . Findley. 1985. Evaluacion de la ictiofauna acompanante del camaron capturado en las costas de Sonora y norte de Sinaloa, Mexico, p. 201-254. In: Yanez-Arancibia, A (Ed.), Recursos pesqueros potenciales de Mexico: la pesca acompanante del camaron. Programa universitario deAlimentos/Instituto Ciencias del Mar y Limnologia, UNAM/Instituto Nacional de la Pesca. Mexico. Pitcher, J.T. 1998a. 'Back to the Future': A novel methodology and policy goal in fisheries. Fisheries Centre Research Report. University of British Columbia, Canada, 6(5): 4-7. Pitcher, J.T. 1998b. Ecosystem simulation models and the new 'Back to the Future' approach to fisheries management, p 21-23. In: D. Pauly (ed). The use of Ecopath with Ecosim to evaluate strategies for sustainable exploitation of multi-species resources. Fisheries Centre Research reports, 6(2). Vancouver, British Columbia, Canada. Pitcher, J. T. and D . Pauly. 1998. Rebuilding ecosystems, not sustainability, as the proper goal of fishery management, p. 311-329. In T. J. Pitcher, P.J.B. Hart and D. Pauly (Eds). Reinventing Fisheries Management. Chapman and Hall, London. Pitcher, J.T. 1999. Rebuilding as a new goal for fisheries management: Reconstructing 229 the past to salvage the future. Ecological Applications, 10(2): 126-135. Pitcher, J.T. 2000. Fisheries management that aims to rebuild resources can help resolve disputes reinvigorate fisheries science and encourage public support. Fish and Fisheries, 1(1): 99-103. Pitcher, J.T. 2001. Fisheries managed to build ecosystems. Reconstructing the past to salvage the future. Ecological applications, 11(2): 601-617. Pitcher, T. J. 2004a. Introduction to the methodological challenges in 'Back to the Future' research, p. 4-10. In: Pitcher T. J. (Ed) Back to the Future: advances methodology for modelling and evaluating past ecosystems as future policy goals. Fisheries Centre Research Reports, 12(1). Vancouver, British Columbia, Canada. Pitcher, T. J. 2004b. What was the structure of past ecosystems that had many top predators?, p. 18-20. In: Pitcher T. J. (ed) Back to the Future: advances methodology for modelling and evaluating past ecosystems as future policy goals. Fisheries Centre Research Reports, 12(1). Vancouver, British Columbia, Canada. Pitcher, T. J. 2004c. The problem of extinctions, p. 21-28. In. Pitcher T. J. (ed) Back to the Future: advances methodology for modelling and evaluating past ecosystems as future policy goals. Fisheries Centre Research Reports 12(1). Vancouver, British Columbia, Canada. Pitcher, T.J . , R. Watson, R. Forrest, H . Por Valtysson, and S. Guenette. 2002. Estimating illegal and unreported catches from marine ecosystems: A basis for change. Fish and Fisheries, 3: 317-339. Pitcher, T.J. and R. Forrest. 2004. Challenging ecosystem simulation models with climate change: the 'perfect storm', p. 29-39. In: Pitcher T. J. (Ed) Back to the Future: advances methodology for modelling and evaluating past ecosystems as future policy goals. Fisheries Centre Research Reports 12(1). Vancouver, British Columbia, Canada. Pitcher, T.J . , C H . Ainsworth, E . A . Buchary, W . L . Cheung, R. Forrest, N . Haggan, H . Lozano-Montes, T. Morato and L . Morissete. 2005. Strategic management of marine ecosystem using whole-ecosystem simulation modelling: The 'Back to the future' policy approach. In: Strategic Management of Marine Ecosystems (Levner, E . , I. Linkov and J . M . Proth Editors) Strategic Management of Marine Ecosystems. N A T O Sceince Series. TV Earth and Environmental Science. 50: 199-258. Pitt, J . , D .F . Luecke, M . J. Cohen, E . P. Glenn and C . Valdes-Castillas. 2000. Two Countries, one river: managing for nature in the Colorado River delta. Natural Resources Journal, 40: 819-864. Polovina, J.J. 1984. Model of a coral reef ecosystem I. The Ecopath model and its application to trench frigate shoals. Coral Reefs, 3(1): 1-11. Polovina, J.J., G . C . Mitchum, E . Graham, P. Craig, E . E . Demartin and E . Flint. 1994. Physical and biological consequences of a climate event in the central North Pacific. Fisheries Oceanogaphy, 3: 15-21. P R O F E P A (Procuraduria Federal de Protection al Ambiente). 1995. Mortandad de mamiferos marinos y aves marinas en el alto Golfo de California. Informe final. Procuraduria Federal de Protection al Ambiente. Subprocudaruria de Recursos Naturales. Secretaria de Medio Ambiente, Recursos Naturales y Pesca. Mexico. 63 p. 230 Ramirez, G . E . 1962. Proyecto de fomento pesquero para ayudar a la rehabilitation economica de la tribu Sen. Secretaria Ind. Com. Dir. Gral. Pesca. Ind. Con. Trab. Divulg. 3(26): 1-17. Ramirez, R. M . 1985. Composition de la captura del calamar gigante en el Golfo de California durante 1981.CEBASIO, 10: 123-137. Reserva de la Biosfera y alto Golfo de Caifornia y delta del Rio Colorado. 2003. Programa de Conservation y Manejo. S E M A R N A T . Mexico. 141p. Reyes, C . A . and F. Lavin. 1997. Effects of the autumn-winter meteorology upon the surface heat loss in the northern Gulf of California. Atmosfera, 10: 101-123. Revenga, C , S. Murray, J. Abramovitz and A . Hammond. 1998. Water-shields of the World: Ecological value and vulnerability. World Watch Institute, Washington, D C . U S A . 49 p. Richey, J. E . , R . C . Wissman and A . H . Devol. 1978. Carbon flow in four lake ecosystems: A structural approach. Science, 202: 1183-1186. Ripa, P. and S.G. Marinone. 1989. Seasonal variability of temperature, salinity, vorticity and sea level in the central Gulf of California inferred from historical data. Quart. J. Roy. Met. Soc , 115: 887-913. Robles, J. M . and S. G . Marinone. 1987. Seasonal and interannual thermoaline variability in the Guaymas Basin of the Gulf of California. Cont. Shelf. Res., 7(7): 715-733. Rodriguez, C , K . W . Flessa and D. Dettman. 2001. Effects of upstream diversion of Colorado River water on the estuarine bivalve mollusk, Mulina coloradensis. Conservation Biology, 15: 249-258. Roemmich, D. and J. McGowan. 1995. Climatic warming and the decline of zooplankton in the California Current. Science, 267: 1324-1326. Rojas-Bracho, L . 1998. Evaluation genetica y riesgo de endogamia en una especie endemica criticamente amenazada del Golfo de California: la vaquita (Phocoena sinus Norris y McFarland, 1958). PhD. Thesis. Facultad de Ciencias Marinas, Universidad Autonoma de Baja California, Ensenada, B C . Mexico. 113 p. Rojas-Bracho, L . And B. Taylor. 1999. Risk factors affecting the vaquita Phocena sinus. Marine Mammal Science, 15: 974-989. Rojas-Herrera, A . and X . Chiapa-Carrera. 2002. Feedings habits of the spotted rose snapper Lutjanus gutatus in the coast of Guerrero, Mexico. Ciencias Marinas, 28(2): 1-15. Rojo-Vazques, J .A. , F. Arreguin-Sanchez, E . Godinez-Dominguez and M . Ramirez-Rodriguez. 1999. Gillnet selectivity for the spotted rose snapper (Lutjanus guttatus) and the Amarillo snapper (Lutjanus argentiventris) in Navidad Bay, Jalisco, Mexico. Ciencias Marinas, 25(1): 145-152. Roman-Rodriguez, M . and G . Hamman. 1997. Age and Growth of totoaba, Totoaba macdonaldi, (Sciaenidae) in the upper Gulf of California. Fishery Bulletin, 95: 620-628. Romero, C . G . 1980. Sistematica, biologia y ecologia de los anomuros (Crustacea decapoda) de la Laguna de Percebu, alto Golfo de California, B C . Mexico. MSc . Thesis. Universidad Autonoma de Baja California, Mexico. 161p. Rozengurt, M . and I. M . Hedgpeth. 1989. The impact of alternate river flow on the ecosystem of the Caspian Sea. Reviews in Aquatic Science, 1: 337-362. 231 Ruesink, J.L. 1999. Ecosystem models as caricatures: The case of Prince William Sound, p 87-89. In: Okey, T. A . and D . Pauly. 1999. Atrophic mass-balance model of Alaska's Prince William Sound Ecosystem, for the post-spill period 1994-1996. Fisheries Centre research Reports. 7 (4). Vancouver, British Columbia, Canada. Saenz-Arroyo, A . , R. M . Callum, J. Torre, M . Carino-Olvera and R. Enriquez-Andrade. 2005. Rapidly shifting environmental baselines among fishers of the Sea of Cortez. Proceedings of the Royal Society, 1726: 1-6. Sala E . , O. Abusto-Oropeza, G . Paredes and G . Thompson. 2003. Spawning aggregation and reproductive behavior of reef fish in the Gulf of California. Bulletin of Marine Science, 72: 103-121. Sala E . , O. Abusto-Oropeza, M . Reza, G . Paredes and L . Lopez-Lemus. 2004. Fishing down coastal food webs in the Gulf of California. Fisheries management, 29(3): 19-25. Salazar-Vallejo, S. I. 1990. Soft-bottom polychaetes (Annelidae: Polychaeta) of IslaRasa, Gulf of California. Ciencias Marinas, 16(4): 75-85. Sanchez, R. 1992. Algunas consideraciones socioeconomicas para la conservation de la vaquita y la totoaba. Sixth bi-annual dialogue. Critical issues in United States-Mexico. U C Mexus, San Diego, C A . U S A . July: 24-25. Sanchez-Velazco, L . , J. Valdez-Holguin, B . Shirasgo, M . A . Cisneros-Mata, A . Zarate. 2002. Changes in the spawning environment of Sardinops sagax caeruleus in the Gulf of California during E l Nino 1997-1998. Estuarine, Coastal and Shelf Science, 54: 207-217. Schindler, D . E . , T . E . Essington, J.F. Kitchell, C . Boggs, R. Hilborn. 2002. Sharks and tunas: fisheries impacts on predators with contrasting life histories. Ecological Applications, 12(3): 735-748. Scripps Institution of Oceanography, University of California. 1969. Cruise TO-60-1. Data Report, 69(4): 57-65. Sea Watch. 2003a. Nets- killers of the Sea. http://www. Seawatch.org/2003/seaofcaortes. Sea Watch. 2003b. Mexico's marlin.. .doomed? http://www.Seawatch.org/2003/seaofcaortes. Sea Watch. 2003c. The Midriff Island Shark disaster. http://www. Seawatch.org.2003/seaofcaortes. 2003/seaofcaortes. SeaWatch. 2003d. Where have all mantas gone?, http://www.seawatch.org. S E M A R N A T (Secretaria de Medio Ambiente y Recursos Naturales). 2003. Programa de conservation y manejo. Reserva de la Biosfera alto Golfo de California y delta del Rio Colorado. Informe Inedito. April , 2003. 142 p. S E M A R N A P (Secretaria de Medio Ambiente, Recursos Naturales y Pesca, Mexico). 1998. Informe de labores 1997-1998. Capitulo 2. 41p. S E M A R N A P (Secretaria de Medio Ambiente, Recursos Naturales y Pesca, Mexico). 1999. Mortandad de lobos marinos registrada en las costas de Sonora. http://www.semarnap.gob.mx/quincena/qui-34/mira.htm Press release from P R O F E P A , emitted March 4, 1999. 79 p. Silber, K . G . 1990. Distributional relations of cetaceans in the Northern Gulf of California, with special reference to the vaquita, Phocoena sinus. PhD Thesis. University of California, Santa Cruz, C A . U S A . 116 p. 232 Silber, K . G . , M . Newcomer, P. Perez-Cortez and G . Ellis. 1994. Cetaceans of the northern Gulf of California: distribution, occurrence and relative abundance. Marine Mammal Science, 10(3): 283-298. Shipek, F. 1965. Lower California Frontier: articles from the San Diego Union (1870). Dawson's Book shop, Los Angeles. California, U S A . July22: 37-40. Smith-Vaniz, W. 1995. Carangidae. Jureles, pampanos, zapateros, cocineros, macarelas, p. 940-986. In: W. Fisher, F. Kropp, W. Schneider, C . Carpenter (Eds). Guia F A O para la identification de especies afines de la pesca del Pacifico Centro-Oriental. 16. F A O Rome, Italy. Squires, H.J . and J .H. Barragan. 1991. Lolliguncula panamensis (Cephalopoda: Loliginidae) from the Pacific Coast of Colombia. U N D P / F A O proyecto para el desarrollo de la pesca maritima en Colombia, Bogota. 67 p. Stanford, J. A . and J. V . Ward. 1986. The Colorado River system, p. 21-44. In: B.R. Davies and K . F. Walker (Eds). The Ecology of river systems; Dr. W. Junk, Dordrecht, The Netherlands. Stanley, D. J. and G . Warne. 1993. Nile Delta: recent geological evolution and human impact. Science, 260: 628-634. Steinbeck, J. 1940. Sea of Cortez: A Leisurely Journal of Travel and Research, p. 17-23. In: Mark Shwartz. 2003. Voyage will retrace Steinbeck-Ricketts trip to Sea of Cortez. Stanford Report, 3. Electronic file retrieved on December 21, 2003. http://news-service.stanford.edu/news/2003/december3/cortez-123.html Stewart, J. G . and J. N . Norris. 1981. Gelidiaceae (Rhodophyta) from the northern Gulf of California, Mexico. Phycologia, 20(3): 273-287. Sumalia, U . R. and C.J. Walters. 2003. Intergenerational discounting, p. 19-25. In: Sumalia, R. (Ed).Three essays on the Economics of fishing. Fisheries Centre Reports 11(3). Vancouver, British Coluymbia, Canada. Sumalia, U . R. and C.J . Walters. 2004. Intergenerational discounting: a new intuitive approach. Ecological Econimics, 16(3). Vancouver, British Coluymbia, Canada. Schwing, F. B . , T. Murphie and P. Green. 2002. A climate index for the North Pacific. Geophys Res, 6: 1635-1638. Sykes, G . C . 1937. Delta, estuary and lower proportion of the channel of the Colorado River 1933-1935. Carnegie Institution of Washington and the American Geographical Society of New York. Washington, D . C . I 13 p. Talent, L . G . 1982. Foods habits of the gray smoothound, Mustelus californicus, the brown smoothound, Mustelus henlei, the shovelnose guitarfish, Rhinobatos productus and the bat ray, Myliobatis calirnica. California Fish and Game, 68: 224-234. Tershy, B.R., A . Acevedo-G., D . Breese and C S . Strong. 1993. Revista de Investigaciones Cientificas. Univesidad Autonoma de Baja California Sur, 1 (Numero Especial S O M M E M A 1): 31-35. Thompson, R . W . 1968. Tidal sedimentation of the Colorado River delta, northwestern Gulf of California. Geological Society of America. 87p. Townsed, C H . 1901. Dredging and other records of the U.S. Fish Commission Steamer 'Albatros' with biography relative to the work of the vessel. U S Fish Commission Report f o r i 900: 387-563. Turk, P.J. 1989. Assessment of the potential impact of fishing activities in the northern 233 Gulf of California on the Phocoena sinus population. A grant report submitted to the Center for Marine Conservation. C E D O , Mexico. Ulanowicz, R. E . 1986. Grow and Development: Ecosystem phenomenology. Springer-Verlag, New York. 203 p. Ulanowicz, R. E and C . J. Puccia. 1990. Mixed trophic impacts in ecosystems. Coenoses, 5: 7-16. Ulanowicz, R. E and J J . Kay. 1992. A computer package for the analysis of ecosystem flow networks. Environ. Software, 6: 131-142. Urban, J.R. 2002. Marine mammals in the Gulf of California: conservation and management. Back to the Future in the Sea of Cortez: A Workshop on the Restoration of Past Ecosystems as Policy Goals for Fisheries. Novemeber 2-5, 2002. C I C I M A R , L a Paz, Baja California, Mexico. Valdez-Holguin, J. E- and J. R. Lara-Lara. 1987. Primary productivity in the Gulf of California, effects of E l Nino 1982-1983 event. Ciencias Marinas, 13: 34-50. Van Andel, T . H . 1964. Recent marine sediments of the Gulf of California, p. 121-136. In: Van Andel, T .J . , Shor, G . G . (Eds), Marine Geology of the Gulf of California: A symposium. A m . Assoc. Pet. Geol. 3. Los Angeles, C A . U S A . Vandivere, W. And B. Vorster. 1984. Hydrology analysis of the Colorado River floods of 1983. Geo-Journal, 9: 343-350. Varady, R. G . , K . B . Hankins, A . Kaus, E . Young and R. Merideth. 2001. The sea of Cortes: nature, water, culture and livelihood in the lower Colorado River basin and delta - an overview of issues, policies and approaches to environmental restoration. Journal of Arid environments, 49: 195-209. Vasquez-Leon, M . , and T.R. Mcguire. 1993. The Mexican Shrimp industry: politics of efficiency. M A S T , 6(2): 59-73. Velarde, E . , M . Tordesillas and L . Rocio-Esquivel. 1994. Seabirds as indicators of important fish population in the Gulf of California. CalCOFI Report, 35: 137-143. Velasco-Rodriguez, V . 2003. No hay control de pesca en mares Mexicanos: advierte el P V E M . Boletin del Senado, P V E N . April 21, 2003. Mexico. Vidal, 0.1995. Population biology and incidental mortality of the vaquita, Phocoena sinus. Report of the International Whaling Commission. Special Issue, 16: 247-272. Vidal, O. and J.P. Gallardo-Reynoso. 1996. Die-offs of marine mammals and seabirds in the Gulf of California, Mexico. Marine Mammal Science, 12(4): 627 :635. Vidal, O., R. L . Brownell Jr. And L . T. Floyd. 1999. Vaquita, Phocoena sinus, p.357-373. In: Ridgway, S. and Richard Harrison (Eds). Handbook of marine mammals. The second book of dolphins and the porpoises. San Diego, C A . U S A . Vil la, R .B. 1986. Megaptera novaeangliae, la ballena jorobada en el Golfo de Santa Clara, Mar de Cortes, Mexico. Anales del Instituto de Biologia. Serie Zoolica. Universidad Nacional Autonoma de Mexico. Mexico, 57(2): 481-484. Villareal-Chavez, G . 1992. Algunos aspectos de la biologia de Callinectes arcuatus (Crustacea: ecapoda: Portunidae) en el delta del Rio Colorado, Mexico. Proceedings of the San Diego society of Natural History, 10: 1-4. Villareal-Chavez, G . and C M . Carmona-Cedillo. 1999. Estructura de la comunidad bentonica (infaunal) en el delta del Rio Colorado, Mexico. Revista de Biologia Tropical, 47(4): 809-813. 234 Walters, C , D . Pauly and V . Christensen. 1997. Structuring dynamic models of exploited ecosystems from trophic mass-balance assessment. Reviews in Fish Biology and Fisheries, 7: 139-172. Walters, C , V . Christensen., D . Pauly and J. F. Kitchell. 2000. Representing density dependence consequences of life history strategies in aquatic ecosystems: Ecosim II. Ecosystems, 3(1): 70-81. Walters, C , V . Christensen., D. Pauly and J. F . Kitchell. 2002. Searching for optimum fishing strategies for fishery development, recovery and sustainability, p. 11-15. In: Pitcher T.J. and Cochrane, k. (Eds). The use of ecosystem models to investigate multi-species management strategies for capture fisheries. Fisheries Centre Research Report, 10(2). Vancouver, British Columbia, Canda. Walther, A . 1993. Origen de Mexicalli. Universidad de Baja California. Mexico.63 p. Watson, W. 1996. Haemulida: grunts, p. 1002-1011. In: H . G . Moser (Ed). The early stages of fishes in the California Current region. California Cooperative Oceanic Fisheries Investigations (CalCOFI), Atlas 33. Weiner, T. 2002. In Mexico, Greed kills fish by the seaful. The New York Times. New York, U S A . April 10, 2002: 3-4. Whitehead, P.J.P., G.J. Nelson and T. Wongratana. 1988. F A O Species Catalogue. Clupeoid fishes of the world (Suborder Clupeoidei). A n illustrated catalogue of the herrings, sardines, pilchards, sprats, shads, anchovies and wolf-herrings. Part 1 -Chirocentridae, clupeidae and pristigaste. F A O Fish. Synopsis, 125(7): 1-303. Wright, L . M . 2002. Building local knowledge for developing health policy through key informat interviews. Journal of Extension, 40(1): 34-53. Yli-pelkonen, V . , and J. Kohl. 2005. The role of local ecological knowledge in sustainable urban planning: perspectives from Finland. Sustainability: science, practice and policy. E-journal (1). http://eiournal.nbii.ors/archives/vollissl/0410-007.yli-pelkonen.html. Zamora-Arroyo, F. , P. Nagler, M . Briggs, M . Radtke, H . Rodriguez, J. Garcia, C . Valdes, H . Huete and E . Glenn. 2001. Regeneration of native tress in response to flood releases from the United States into the delta of the Colorado River, Mexico. Journal of Arid Environments, 49: 163-171. Zavala-Gonzalez, A . and E . Mellink. 2000. Historical exploitation of the California Sea lion, Zalophus calirnianus, in Mexico. Marine fisheries Review, 62(1): 35-39. Zeitzchel, B. 1969. Primary productivity in the Gulf of California. Marine Biology, 3: 201-207. Zengel, S.A., V . J . Meretsky, E.P. Glenn, R.S. Felger, D. Ortiz. 1995. Cienega de Santa Clara, a remanent wetland in the Colorado River delta (Mexico): vegetation distribution and the effect of water flow reduction. Ecological Engineering, 4: 19-36. Zetina-Rejon, M . J. 1999. Influencia de la pesca de camaron en la estructura del ecosistema lagunar Huizache-Caimanero, Sinaloa, Mexico. Msc. Thesis. Centra Interdisciplinario de Ciencias Marinas. L a Paz, B .C.S . Mexico. 93 p. 235 Appendix 1. History of the main fisheries in the Gulf of California, Mexico. Event Only aboriginal Cucapa communities were fishing in the upper Gulf before 1500s reaching up to 20,000 people (2,4). Delta area of 780,000 ha (to decrease to 70,000 ha in 1990s; 3, 37). Seasonal fishing camps in the upper Gulf grew into families and communities: Puerto Pefiasco, Golfo de Santa Clara and San Felipe became permanents (1). First Colorado River diversion started (3). Totoaba adults fished for local consumption in Central Gulf (14). Sport fishing began in Puerto Pefiasco with the first hotel for American tourists (2). Totoaba sport fishery began (11). Totoaba fishery in the central Gulf started to decline, so that 6 German fishermen followed totoaba migration to the upper Gulf (11). The Water diversion of the Colorado River is signed (6). Offshore shrimp fishery began with two US boats (14). Commercial totoaba fishery in the upper Gulf (5). In 1923 began shrimp fishery, with totoaba juveniles taken accidentally (14). Agreement between US and Mexico for totoaba (52). The most important changes in boats during the development of fishery in the GoC were the use of outboard motors for small boats which were formerly propelled by oars (39, 40). Hoover Dam became the first of 20 Major Dams on the Colorado River (6). 17 sardine boats modified for shrimp fishing (14). Bycatch of vaquita began (7). Physical and chemical conditions of the upper Gulf altered and affecting endemic species (totoaba and vaquita) and commercial species such as shrimp (52). Closed season for totoaba from 20/03/40 to 01/05/40 to protect spawners (52). Totoaba sport fishery declined (66). "Before and After World War II, American and Japanese ships took every school of tuna and every swarm of sardines they could, along with sea lions for pet food and sharks to use the livers to remedy iron-poor or tired-blood" (8 in 9). In 1946, the highest exportation of totoaba to the US market with 1,299 t of fillets (52). Completion of Highway 8 in 1942 and the railroad in 1947 linked the Mexican production of the U G C to US markets (1). 1950 shrimp fishery growth exponentially (14). 1955 a new Biosphere Reserve nuclear zone established, fishing in the mouth of the Colorado river is prohibited (10). Closed season for totoaba changed from 01/03 to 01/04 or 15/05 according to their abundances (52). Totoaba landings started to decline (52). Double-ring trawls introduced into the shrimp fishery (14). 1955 shrimp C P U E (Catch/Boat) has its maximum with 45 t (heads off) in Guaymas (central Gulf; 14). The freshwater input to the upper Gulf stops due to completion of the Glenn Canyon Dam and filling of Lake Powell (12). Shrimp fishery expanded to the U G C (10). Sardine fish meal plants and canneries come to central Gulf (Guaymas) (53). Sardine landings started to decrease (53). Although, it has been prohibited by law, dynamite is also used to kill totoaba directly (41 in 39). Late of 60's Shrimp trawls fishermen gradually reduce mesh size (implies increase of bycatch)(10). 236 Appendix 1. Continuation. Event Pacific sardine fishery collapsed and moved to central Gulf (53). Sardine fishing in the GoC was initiated on a massive scale in 1971 (67). Jumbo squid fishery in central Gulf began (54). High shrimp prices; government invest in boats, equipment and a harbor en Puerto Pefiasco (1). Mexican immigration to the Gulf in a kind of 'Gold Rush' for shrimp, which was known as 'Pink Gold' (1). E l Nino event (1972) (22). 73 species were collected from the bottom of the U G C (13 species are might be endemic), juveniles of most fish species occur in areas of high turbidity in the mouth of the Colorado River (36). Refuge zone declared in 1955 declared a reserve zone in 1974 banning all fishing in the U G C (32). 1970, there are around 12 large boats (shrimp trawlers) dedicated partially or exclusively to totoaba fishing in Puerto Pefiasco and 18 small boats in Golfo of Santa Clara (38). There are no regulatory measures which fix catch quotas for totoaba or which limit the number of fishing units (38). Fishing totoaba is prohibited during the priod from the April 1 to May 15 of every year (39). Mexican Law protects totoaba; fishery banned (18). Totoaba listed as threatened under CITES (13). Drastic declines in totoaba, but commercial and sport fishing outlawed, but totoaba gillnetting continued until 1990s and it was a primary cause of incidental kills of vaquita (53). Mesh size regulation introduced in 1977 as a management measure in the shrimp trawl fishery (16 in 14). 1977, 28 Peruvian boats (120 t) entered the Gulf after the collapse of the Peruvian anchovy fishery (53). 1979 high squid abundance triggered increase in fleet size, Japanese vessels appeared (17). Lake Powell is now filled and some water can reach the upper Gulf during strong precipitation events (6). Monofilament gill nets became the tool of the panga fishermen; 5,000 -7,000 pangas fishing in the Gulf (19). Institute of National fisheries (INP) began monitoring programs to evaluate shrimp spawning and recruitment each year until 1987 (14). 1978 saw onset of commercial exploitation of demersal resources in the GoC with 12 trawlers (43). 344, 000 tonnes of bycatch in the shrimp fishery from 1977-80 (43). A n average of 6 marlins a day were catching (in Cabo San Lucas, BCS, according to one fisherman; 20). Still fishing totoaba (38). 35 vaquitas killed per year (1985-1992) in gillnets (7). Shrimp landings were higher than normal - suggests that spawning in promoted in post E l Nino conditions (54). More than 80% of the total catch of shrimp in Mexico (40,000 to 55,000 t heads on) is caught in the Gulf of California (14). The shrimp fleet doubled without any increase in the total catch and the catch per boat diminished from 39 t (heads off) per year in 1971 to 15 tonnes in 1980 (14). Expansion of the Pacific sardine fishery in the Gulf (53). Newly installed processing plants increased demand for Pacific sardine (53). Strong E l Nino event (1982-83) and associated with an increase of water discharge from the Colorado River in the upper Gulf (22). Jumbo squid fishery collapsed (54). Seven species of cetacean have been recorded after a period (1986-87) of freshwater release in the upper Gulf (34). 237 Appendix 1 . Continuation. Event From 1985 to 1993 over 200,000 sharks were killed (24). Late of 1980s near vanish of Pacific sharp nose shark (55). 1984, shrimp landings higher than normal - associated with El Nino conditions-(56). EEZ established (57). Anchovy first appeared in commercial landings (53). Collapse of shrimp industry at the end of 1980s (21). Drastic decline of Pacific sardine fishery in the Gulf (53). Catches of chano represents 16% of the finfish fishery in Puerto Pefiasco and it growth to 86% in 1992 (51). 1989, a total of 816 'pangas' and 266 shrimp boats in the upper Gulf (64). 161 tonnes of totoaba poached in 1985 (f)(64). High market price (US markets) for the totoaba meat, increases illegal fishing of this species (64). Small boat fishermen in the upper Gulf began to hammer chano to supply the Asian market for surimi, the processed with 700 t processed in the Golfo of Santa Clara alone (15). From Jan to Oct 1993, 12 vaquitas were killed during 16,000 hours of chano fishing (el Golfo de Santa Clara small scale fleet around 200 pangas)(15). Total vaquita population is 400-500 individuals (44). Trawles forced to use turtle- and finfish excluder devices (15). Near collapse of shrimp fishery provoked diversification of small scale fishing activity and target species switch (6). Upper Gulf has a fleet around 600 pangas (21). 1993 for the firt time at least 35 vaquitas died accidentally in the fisheries each year (7). In 1992, the Mexican authorities banned the use of gillnets with a mesh size greater than 25 cm (45). Fishing improved significantly following Colorado River floods in the early 1980s, blue shrimp landings increased, and the Gulf corvina (Cynoscion othonopterus), a fish that had not been seen in the upper Gulf for 40 years, returned in large numbers.(21). Santa Clara, San Felipe and Puerto Penasco's fleet were largely privatized during the 1993-94 season (21). Establishment of the upper Gulf of California/ Colorado River Delta Biosphere Reserve: Core Zone: all commercial fisheries banned except for traditional fishing by the Cucapa people in the delta and clam harvesting by local residents; buffer zone: increased regulations for most fisheries (12). There was an increase of sport fishing in the upper Gulf, first tourist mega-project (hotel Plaza Las Glorias), today at least 13 mega-projects are planned for the U G C (1). 1994, a gillnet fleet backed by unknown financiers appeared in Sonora, fishermen and scientists say it slaughters thousands of shark solely for their fins ($US 300 a pound in Asian markets)(9). 1990, population of giant mantas in Southern Gulf had dwindled to nothing (victims of harpooning and gillnets)(25). The shark fishermen bait their nets with marine mammal meat (24). 1993 E l Nino event linked to increased water discharge from the Colorado River, water release from Lake Powell (22). PROFEPA enforces activities in the Reserve (6). Shrimp bycatch ratio estimated at 1:10(6). 1995, almost no shark at all, the big shark processors moved to their next target (24). Nevertheless, the management plan, finalized in 1996, allows offshore trawling in the buffer zone of the Biosphere reserve; trawlers will be required to carry turtle-excluder devices (21). No fishing in the core area of the Biosphere Reserve, also, no trawling in estuaries or depths lower than 9m (26). Increase of shrimp landings after E l Nino event and water delivered by the Colorado River (26). Shrimp trawlers started to use turtle and finfish excluder devises in the upper Gulf (26). Turtle excluder devices mandatory in commercial shrimp fishery (7). Only 20 new squid licenses recommended (68). 238 Appendix 1. Continuation Event At least 70 species of fish, mollusks, crustaceans and echinoderms are regularly caught by the artisanal fleet, approximately 40% of these are designated for international market (27). Mexico created CIRVA: International Committee for the recovery of the Vaquita (26). 212,000 t of Monterrey sardine in central Gulf recorded pre-El Nino (59). 1997: 1,144 shrimp boats in the Gulf using two nets of 24 m (60).. 1997, strong E l Nino event (22). 1998, only 59,000 tonnes of Monterrey sardine recorded from the Central Gulf - drop associated with warm temperatures of E l Nino 1997 (59). Water released from the Lake Powell into the U G C (6). 1995, most fishermen of the upper Gulf thought that the richness had declined between 70 to 80% (19). Increase of sardine landings associated with post E l Nino cooler SST (59). The landings shrimp catches in U G C correlated significantly with Colorado River discharge (35). The spawning and nursery grounds of totoaba have been altered by control of the Colorado River flow and agricultural waste (negative impact on the survival of very early life stages of totoaba)(42). 1997 last census puts vaquita population at 567 individuals (47). 1998, CEDO published extensive research of inshore fisheries in the U G C (27). A n atlas of abundance for the 17 most abundant fish species in the U G C was developed with a list of 98 fish species caught by trawlers (50). There were 200 (1993) pangas in the Gulf of Santa Clara that increase to 600 pangas in 2000 (26). CNP (Carta Nacional Pesquera) established to increase the fishing effort in the upper Gulf with 700 tons of shrimp per year, not including Puerto Pefiasco (26). " It is difficult to know the number of boats fishing in the gulf (61). Mexican law makes it illegal to destroy seabed structure and use fishing methods with high bycatch rates (6). NAFS official list of marine fishes at risk of extinction include six species from the GoC (33). Delta maintains about 70,000 ha of wetland (6). The CNP (Carta Nacional Pesquera) disbanded bycatch of vaquitas must be lower than 0.2 /year (48). The GoC is the most productive region for fisheries in Mexico (51). 46,000 people living inside of the Biosphere Reserve, where fishing and tourism are the principal activities (51). More than 1,600 km of gillnet were sold in Sonora (central Gulf). 40,000 turtles are killed per year on the west Coast of Mexico (9). The high price of turtle meat and shark fins, founded on male folklore long predating (31). October: PROFEPA (Secretary of Environmental Protection) said that it was possible to remove the shrimp trawlers during 15 days from the Biosphere Reserve (16). WWF/CEDO meeting with gillnet fishers concluded that 75% of them in the upper Gulf were prepared to change fishing practice/ jobs (6). Closed season for shrimps was variable by area for the first time (61). A total of 1,500 small boats in the upper Gulf (46). The productivity of mollusk, Mulina Coloradensis of the delta system has fallen at least 95% since dams built (49). In the GoC the offshore shrimp fishery caught 87% of this resource, inshore 13% (51). An official government survey found 238 pangas (49% of the official number registered) in El Golfo de Santa Clara fishing without permit (65). Mr. Jeronimo Ramos, the National Fisheries commissioner said about 1,200 permits existed for boats in the GoC and Pacific Ocean, estimating 20-30% of the catch was being taken illegally (9). "Fishermen, businessmen, scientists and even some federal officials say at least 12,000 unregulated fishing boats, probably more, now at large in the GoC (9). The long-liners land as much as 20 t a day of dorado in the Port of Guaymas alone, along with unrecorded illegal catches like sea turtles (SUS 200 dlls a piece in Mexican black market)(9). Shrimp fleet haul up 10 pounds of life for every pound of shrimp (9). 2,500 t of shrimp inside of the Biosphere Reserve are "harvested" (probably 1,800 tonnes by trawl fleet and the rest by the inshore fleet)(26). Official shrimp season from Sept-Oct to Feb-March (26). 140 shrimp boats in the upper Gulf, but an undetermined number of boats from other parts (Guaymas, Topolobampo, Mazatlan, Yavaros) enter to the Reserve (26)."It is necessary to know the real number of pangas fishing without permission in the Gulf of Santa Clara (16). September: emergency law introduced prohibiting trawling and restricting the types of gillnets allowed introduced. Triggered a dispute with shrimp trawlers. Resolved in November, with an agreement allowing 130 trawlers in the Reserve (63). 239 Appendix 1. Continuation Year Event Ref. 2003 PROFEPA confiscated 11,000 kg of turtle in the upper Gulf during 8 days (30). CONAPESCA gave 2,400 permits for shrimp fishing for the season 2002-03, but there are between 6,000 to 7,000 boats fishing shrimps in the GoC, all year round (including closed season)(31).; Fish brokers bought 150 to 200 tonnes of fish a week, today they buy 10 to 15 tonnes (19). Trawling activity banned in the Biosphere Reserve (16). 30, 19, 31, 16, 1 Boyer. P. 1996. 2 Walther, A . 1993. 3 Sykes, G. 1937. 4 Leon-Portilla, M . 1972. 5 Alvarez-Borrego, S. 1999. 6 Brusca, R.C., J. Campoy-Fabela, G. Castillo-Sanchez, R. Cudney-Bueno, L.T. Findley, J. Garcia-Hernandez, E. Glenn, I. Granillo, M.E . Hendricks, J. Murrieta, C. Nagel, M . Roman and P. Turk-Boyer. 2001. 7 Vidal, 0.1995 8 Jonh Steinbeck. 1940. 9 Weiner, T. 2002. 10 Diario Oficial de la Federacion. 1955. 11 Craig, J. A . 1926. 12 Reserva de la Biosfera y alto Golfo de California y delta del Rio Colorado. 2003. 13 Barrera-Guevara, J. C. 1990. 14 Magallon-Barajas, F. J. 1987. 15 McGuire, R T . and G. C. Valdez-Gardea. 1997. 16 Lluch-Belda. D. 1977. 17 Ramirez, R. M . 1985. 18 Diario Oficial de la Federacion. 1975. 19 Sea Watch. 2003a. 20 Sea Watch. 2003b. 21 Vasquez-Leon, M . , T.R. Mcguire. 1993. 22 Lavin, M.F. and S. Sanchez. 1999. 23 Molina-Valdez, D and M . A . Cisneros-Mata. 1988. 24 Sea Watch. 2003c. 25 SeaWatch. 2003d. 26 Campoy-Fabela, J. 2002. 27 Cudney-Bueno, R. and P. Turk. 1998. 28 Anuario de Pesca. 1996. 29 Anonymous. 2004. 30 N O T I M E X . 2003. 31 Velasco-Rodriguez, V . 2003. 32 Campoy-FabelaJ. 2003. 33 Musick, J. A. , and R. Harbin. 2000. 34 Silber, K . G . , M . Newcomer, P. Perez-Cortez and G. Ellis. 1994. 35 Mellink, E., E. Palacios and s. Gonzalez. 1997 36 Guevara-Escamilla, S., M . A . Huerta-Diaz, E. Felix-Pico, C. Farfan and C. Mathews. 1973. 37 Luecke. D.F., Ch. Congdon, E. Glenn, C. Valdes-Casillas and M . Briggs. 1999. 38 Arvizu, Land H . Chavez. 1970. 39 Chute, G.R. 1928. 40 Chute, G.R. 1930. 41 Ramirez, G.E. 1962. 42 Cisneros-Mata, M . A . , L.W. Botsford and J. F. Quinn. 1997. 43 Grande-Vidal, J .M. 1980. 44 D'Agrosa, C , O. Vidal and W. Graham.1995. 240 45 Dirario Oficial de la Federacion. 1992. 46 Camopoy, J.R. 2001. 47 Jaramiilo, L .A . , L. Rojas and T. Gerrodette. 1999 48 C E D O (Centro de Estudios del Desierto y Oceano). 2001. 49 Glenn, E.P., F. Zamora-Arroyo, P. L. Nagler, M . Briggs, W. Shaw, and K . Flessa. 2001. 50 CRIP-Ensenada. 1995. 51 INP (Instituto Nacional de la Pesca). 2000. 52 Cisneros-Mata, M . A. , G. Montemayor-Lopez, M.J . Roman-Rodriguez. 1995. 53 Cisneros-Mata, M . A . M . Nevarez-Martinez, G. Hammann. 1996. 54 Morales-Bojorquez, E., M . A . Cisneros-Mata, M . Nevarez-Martinez and A . Hernandez-Herrera. 2001. 55 Cudney-Bueno, R. 2000. 56 Galindo-Bect, M . , E.P. Glenn, H . Page, L . A . Galindo-Bect, J. Ayon-Hernandez, R.L. Petty, J. Garcia-Hernandez and D. Moore. 2000. 57 Mangatalle, D. 1996. 58 Markaida, U . and O. Sosa-Nishizaki. 2001. 59 Sanchez-Velazco, L. , J. Valdez-Holguin, B. Shirasgo, M . A . Cisneros-Mata, A . Zarate. 2002. 60 Garcia, M . A . , M . A . Cisneros-Mata and A . Balmori-Ramirez. 2000. 61 Lopez-Martinez, J., E. Chavez, S. Hernandez-Vazquez and E. Alcantara-Razo. 1999. 62 Anonymous. 2003. 63 Anonymous. 2002. 64 Turk, P.J. 1989. 65 Meza,R. 2001. 66 Flanagan, C A . and J.R. Hendricson. 1976. 67 Arvizu-Mariinez, J. 1987. 68 Nevarez-Martinez, M and E. Morales-Bojorquez. 1997. 69 Personal communication from fishers of the U G C in 2003. 241 Appendix 2 . Diet matrix built for the 2000 trophic model of the U G C . Prey \ Predator 1 2 3 4 1 Sharks > 120 cm 2 Sharks 120 cm 0 . 0 9 0 3 Totoaba 0 . 0 0 2 4 Toothed cetaceans 0 . 0 1 4 5 Sea lions 0 . 0 9 0 6 Vaquita 0 . 0 0 0 7 Croakers 0 . 0 9 0 0 . 0 2 6 0 . 0 1 2 0 . 0 4 1 8 Jacks 0 . 0 0 3 0 , 0 2 5 9 Corvinas 0 . 0 8 0 0 . 0 0 1 0 0 0 2 0 . 0 0 2 1 0 Serranids 0 . 0 2 7 0 . 0 2 4 0 . 0 0 7 0 . 0 0 2 1 1 Scombri ds 0 . 2 5 0 0 . 0 1 7 0 . 0 0 2 0 . 0 4 1 1 2 Snappers 0 . 0 1 0 1 3 Seabirds 0 . 0 0 1 1 4 Rays 1 5 Flounders 0 . 1 2 0 1 6 Wrasses 0 . 0 0 4 1 7 Chano 0 . 1 2 0 0 . 0 3 7 0 . 1 3 0 0 . 0 1 3 1 8 Sea turtles 0 . 0 0 0 1 9 Pre-adult Vaquita 0 . 0 0 0 2 0 Grunts 0 . 2 5 1 0 . 1 9 5 2 1 Gerreidae 0 . 0 9 4 0 . 0 0 2 0 . 0 7 6 2 2 Guitarfish 0 . 0 3 1 0 . 0 7 4 0 . 0 0 6 2 3 Small demersal fish 0 . 0 3 2 0 . 0 5 8 0 . 0 1 0 2 4 Other fishes 0 . 2 3 0 0 . 3 0 1 0 . 0 3 3 0 . 2 0 1 2 5 Octopus 0 . 0 5 2 2 6 Stomatopods 0 . 0 3 2 0 . 0 8 0 0 . 0 6 2 2 7 Juv. Totoaba 2 8 Myctophids 0 . 0 1 7 2 9 Baleen whales 0 . 0 0 0 3 0 Crabs 0 . 2 4 7 3 1 Squids 0 . 1 4 1 0 . 0 7 7 3 2 Small pelagics 0 . 0 1 2 0 . 2 2 6 0 . 1 7 1 3 3 Jellies 3 4 Planktivorous birds 3 5 Rock shrimp 3 6 Blue shrimp 3 7 Brown shrimp 3 8 Semi-sessile epibenthos 3 9 Sea cucumber 4 0 Benthic meoifauna 4 1 Sessile epibenthos 4 2 Mulina coloradetisis 4 3 Bivalves 4 4 Juv. of Blue Shrimp 4 5 Zooplankton 4 6 Seagrasses 4 7 Seaweeds 4 8 Phytoplankton 4 9 Macroalgae 5 0 Detritus 0 . 0 1 1 0 . 1 0 7 0 . 1 2 9 0 . 0 6 4 0 . 0 0 7 0 . 0 0 9 0 . 2 3 2 0 . 0 1 8 0 . 2 1 7 0 . 0 2 8 0 . 3 0 6 0 . 3 4 5 0 . 1 7 5 0 . 1 1 8 0 . 1 9 4 0 . 1 6 4 0 . 0 8 6 0 . 0 6 8 0 . 0 5 5 0 . 0 1 0 0 . 0 1 1 0 . 1 7 7 0 . 3 4 7 0 . 0 2 5 0 . 0 0 1 0 . 0 4 0 0 , 0 2 4 0 . 0 1 0 0 . 2 4 8 0 , 0 9 1 0 , 0 1 1 0 . 0 7 1 0 0 1 0 0 . 0 5 0 0 . 0 3 5 0 . 0 4 7 0 . 2 4 0 0 . 1 4 1 0 . 0 2 0 0 . 0 1 1 0 . 0 7 6 0 . 8 9 6 0 . 1 6 9 0 . 0 2 7 0 . 0 7 2 0 . 0 1 6 0 . 0 0 4 0 . 0 0 1 0 . 0 1 7 0 . 0 3 4 0 . 5 8 1 0 . 0 2 5 0 . 0 1 0 242 Appendix 2. Continuation. 9 1 0 I I 1 2 1 3 1 4 1 5 1 6 1 7 1 8 1 9 2 0 2 1 2 2 2 3 2 4 2 5 2 6 2 7 2 8 2 9 3 0 3 1 3 2 3 3 3 4 3 5 3 6 3 7 3 8 3 9 4 0 4 1 4 2 4 3 4 4 4 5 4 6 4 7 4 8 4 9 5 0 0 . 0 3 2 0 . 0 3 1 0 . 1 8 8 0 . 1 5 6 0 . 0 1 7 0 . 0 1 3 0 . 0 0 1 0 . 1 0 8 • 0 . 0 0 6 0 . 0 5 4 0 . 6 0 9 0 . 0 6 0 0 . 0 8 8 0 . 0 7 3 0 . 0 3 9 0 . 0 3 1 0 . 0 6 8 0 . 0 9 9 0 . 0 0 2 0 . 0 0 2 0 . 0 0 2 0 . 1 7 2 0 . 2 2 9 0 . 0 0 8 0 . 0 2 1 0 . 0 5 6 0 . 0 1 6 0 . 0 0 1 0 . 0 9 8 0 . 6 0 1 0 . 1 3 0 0 . 0 0 0 0 . 0 6 5 0 . 0 0 8 0 . 0 1 2 0 . 0 0 1 0 . 1 7 1 0 . 0 9 3 0 . 0 2 3 0 . 6 8 4 0 . 0 5 2 0 . 1 7 8 0 . 0 3 0 0 . 6 3 1 0 . 2 2 1 0 . 0 4 6 0 . 1 4 1 0 . 0 2 7 0 . 4 8 9 0 . 0 3 9 0 . 0 0 1 0 . 0 0 1 0 . 0 0 2 0 . 4 3 6 0 . 0 4 8 0 . 2 5 2 0 . 0 2 4 0 . 0 0 2 0 . 0 3 6 0 . 0 6 0 0 . 0 0 7 0 . 0 0 7 0 . 1 2 8 0 . 1 2 2 0 . 2 1 9 0 . 1 1 6 0 . 1 0 9 0 . 0 0 3 0 . 0 0 1 0 . 0 0 1 0 . 0 0 2 0 . 0 1 0 0 . 0 1 5 0 . 0 1 5 0 . 0 1 5 0 . 0 2 7 0 . 0 4 0 0 . 0 0 6 0 . 0 2 1 0 . 7 4 7 243 Appendix 2. Continuation. 1 0 1 1 1 2 1 3 1 4 1 5 1 6 1 7 1 8 1 9 2 0 2 1 2 2 2 3 0 . 0 0 0 2 4 2 5 2 6 0 . 0 1 0 0 . 0 5 6 2 7 2 8 0 . 0 7 2 0 . 0 2 1 2 9 3 0 0 . 0 0 3 0 . 0 2 8 3 1 3 2 0 . 1 1 8 0 . 0 0 1 3 3 0 . 1 7 0 3 4 3 5 0 . 0 2 9 3 6 0 . 0 0 5 0 . 0 0 3 3 7 0 . 0 0 1 3 8 0 . 0 1 3 0 . 0 2 3 3 9 4 0 0 . 1 3 3 0 . 0 7 0 0 . 1 2 3 0 . 1 0 4 0 . 0 8 9 4 1 0 . 0 2 9 0 . 0 2 6 4 2 4 3 4 4 0 . 0 6 0 4 5 0 . 7 5 2 1 . 0 0 0 0 . 9 7 8 0 . 2 1 7 0 . 3 5 7 0 . 3 4 4 0 . 3 2 7 0 . 1 2 1 0 . 1 6 6 0 . 1 0 0 0 . 1 0 0 4 6 0 . 1 2 1 0 . 0 0 1 0 . 0 0 5 4 7 0 . 0 2 5 0 . 1 2 4 0 . 0 5 9 0 . 0 8 3 0 . 1 0 5 4 8 0 . 2 1 4 0 . 3 7 2 0 . 1 6 3 0 . 1 0 6 0 . 4 0 0 0 . 6 0 0 0 . 6 0 0 0 . 9 0 0 4 9 0 . 0 4 6 0 . 0 3 5 0 . 0 1 3 0 . 0 1 6 0 . 0 2 8 0 . 1 1 8 0 . 1 1 3 0 . 0 9 5 5 0 0 . 0 4 8 0 . 5 7 9 0 . 2 7 3 0 . 2 4 8 0 . 3 7 1 0 . 3 0 2 0 . 8 2 4 0 . 5 3 1 0 . 5 3 4 0 . 6 0 0 0 . 4 0 0 0 . 3 0 0 244 / Appendix 3. Diet matrix built for the 1980 trophic model of the U G C Prey \ Predator l 2 3 4 5 6 1 Sharks > 120 cm 0 . 0 0 1 2 Sharks 120 cm 0 . 2 2 0 3 Totoaba 0 . 0 0 2 4 Toothed cetaceans 0 . 0 1 5 5 Sea lions 0 . 0 8 0 6 Vaquita 0 . 0 0 0 7 Croakers 0 . 0 8 0 0 . 0 2 1 0 . 0 3 9 0 . 0 7 4 8 Jacks 0 . 0 0 9 0 . 0 9 3 9 Corvinas 0 . 0 7 0 0 . 0 0 3 0 . 0 0 2 0 . 0 0 2 0 . 0 1 9 1 0 Serranids 0 . 0 3 0 0 . 0 2 1 0 . 0 2 4 0 . 0 4 9 0 . 0 4 4 0 . 0 0 8 1 1 Scombrids 0 . 2 7 0 0 . 0 0 1 0 . 0 0 1 1 2 Snappers 0 . 0 0 2 0 . 0 1 0 1 3 Seabirds 0 . 0 0 0 1 4 Rays 0 . 1 5 8 1 5 Flounders 0 . 0 1 2 0 , 0 1 3 1 6 Wrasses 0 . 0 0 7 1 7 Chano 0 . 0 1 1 0 . 1 3 3 0 . 1 6 3 0 . 0 2 0 0 . 0 0 2 1 8 Sea turtles 0 . 0 0 0 1 9 Pie-adult Vaquita 0 . 0 0 0 2 0 Grunts 0 . 3 0 2 0 . 1 5 0 0 . 1 2 7 2 1 Gerreidae 0 . 0 6 6 0 0 0 1 0 . 0 6 7 0 . 0 3 2 0 . 0 1 1 2 2 Guitarfish 0 . 0 4 2 0 . 2 3 3 0 . 0 0 8 2 3 Small demersal fish 0 . 0 0 1 0 . 0 0 5 2 4 Other fishes 0 . 1 6 0 0 . 2 2 7 0 . 0 7 4 0 . 1 4 5 0 . 3 8 2 0 . 1 4 8 2 5 Octopus 0 . 0 3 7 0 . 1 0 2 0 . 1 9 1 2 6 Stomatopods 0 . 0 1 2 0 . 0 2 5 0 . 2 6 9 0 . 2 1 3 2 7 Juv. Totoaba 0 . 0 0 0 2 8 Myctophids 2 9 Baleen whales 0 . 0 0 0 0 . 0 0 0 3 0 Crabs 0 . 3 5 2 0 , 0 1 7 0 . 1 0 2 3 1 Squids 0 . 2 5 0 0 . 1 8 6 0 , 1 0 2 0 1 2 3 3 2 Small pelagics 0 . 0 6 0 33 Jellies 34 Planktivorous birds 35 Rock shrimp 3 6 Blue shrimp 37 Brown shrimp 3 8 Semi-sessile epibenthos 39 Sea cucumber 4 0 Benthic meoifauna 4 1 Sessile epibenthos 4 2 Mulina coloradensis 43 Bivalves 4 4 Juv. of Blue Shrimp 45 Zooplankton 4 6 Seagrasses 47 Seaweeds 4 8 Phytoplankton 49 Macroalgae 50 Detritus 0 . 3 3 4 0 . 0 0 6 0 . 0 1 0 0 . 0 0 5 0 . 3 1 0 0 . 0 5 9 0 . 0 3 7 0 . 0 1 3 0 . 0 1 0 0 . 3 3 4 0 . 2 6 5 0 . 0 2 7 0 . 0 1 5 0 . 0 0 8 0 . 0 6 0 0 . 0 1 0 0 . 0 0 1 0 . 0 2 6 0 . 6 4 7 0 . 0 3 1 0 . 0 0 2 0 . 0 3 5 0 . 0 0 0 245 Appendix 3. Continuation. 1 3 1 4 1 5 1 6 2 3 4 5 6 7 0 . 0 0 1 9 1 0 1 2 1 3 1 4 0 . 0 0 1 0 . 0 3 1 1 5 0 . 0 1 6 1 6 0 . 0 0 7 1 7 0 . 0 0 1 1 8 1 9 2 0 0 . 0 0 1 2 1 0 . 0 0 8 2 2 ' 2 3 0 . 1 0 6 0 . 0 2 3 2 4 0 . 0 1 6 0 , 0 1 3 0 . 0 0 9 0 . 0 1 7 2 5 0 . 0 3 5 0 . 0 1 4 0 . 1 6 4 2 6 0 . 0 7 7 0 . 0 1 3 2 7 2 8 0 . 8 4 1 2 9 3 0 0 . 0 1 3 0 . 0 1 2 0 . 0 1 2 3 1 3 2 0 . 0 6 7 3 3 3 4 3 5 3 6 0 . 0 0 1 0 . 0 0 1 0 . 0 0 8 3 7 0 , 0 0 0 0 . 0 0 0 3 8 0 . 0 3 5 0 . 0 3 4 0 . 1 6 1 3 9 0 . 0 0 1 4 0 0 . 0 5 3 0 . 0 3 4 0 . 0 8 0 4 1 0 . 0 0 2 0 . 5 2 0 4 2 0 . 0 0 0 4 3 0 . 0 0 0 4 4 0 . 0 1 4 4 5 4 6 4 7 4 8 4 9 5 0 0 . 6 9 6 0 . 8 6 1 1 7 1 8 \ 9 2 0 2 1 2 2 2 3 2 4 2 5 2 6 2 7 0 . 0 1 0 0 . 0 1 1 0 . 0 2 5 0 0 0 9 0 . 1 1 0 0 . 0 0 9 0 . 0 0 7 0 . 0 6 2 0 . 0 1 1 0 . 0 6 3 0 . 0 3 9 0 . 0 1 4 0 . 1 3 0 0 . 0 1 3 0 . 0 0 3 0 . 0 1 2 0 . 0 1 9 0 . 0 0 5 0 . 0 3 3 0 . 0 0 2 0 . 0 0 0 0 . 2 5 9 0 . 0 1 0 0 . 0 0 2 0 . 0 0 2 0 . 0 0 7 0 , 0 3 9 0 . 0 0 1 0 . 0 4 8 0 . 0 1 1 0 . 2 4 3 0 . 0 0 1 0 . 1 8 8 0 . 0 0 9 0 . 0 1 2 0 . 7 7 1 0 . 0 2 2 0 . 0 0 4 0 . 2 3 3 0 . 0 0 4 0 . 3 1 1 0 . 0 9 1 0 , 0 0 1 0 . 4 6 7 0 . 0 6 5 0 . 0 0 0 0 . 0 0 1 0 . 0 0 0 0 . 0 0 9 0 . 0 2 8 0 . 1 3 9 0 . 2 7 7 0 . 0 0 5 0 . 0 0 1 0 . 0 1 0 0 . 0 1 9 0 . 0 0 1 0 . 0 6 7 0 . 1 9 3 0 . 0 3 1 0 . 0 5 1 0 . 1 7 5 0 . 0 3 2 0 . 1 9 3 0 . 1 9 3 0 . 0 0 7 0 . 0 0 2 0 . 0 0 1 0 . 0 8 5 0 . 2 8 9 0 . 1 5 8 0 . 0 1 2 0 . 0 0 1 0 . 0 0 1 0 . 0 0 2 0 . 6 6 8 0 . 0 5 3 0 . 0 0 1 0 . 0 0 0 0 . 0 9 0 0 . 0 2 7 0 , 0 0 2 0 . 0 3 3 0 . 0 1 5 0 . 2 7 2 0 . 6 9 0 0 . 0 3 2 0 . 6 2 1 0 . 0 3 0 0 . 2 8 9 0 . 0 2 9 0 . 0 6 7 0 . 3 2 3 246 Appendix 3. Continuation. 3 0 3 1 2 3 4 5 6 7 9 1 0 1 2 1 3 1 4 1 5 1 6 1 7 1 8 1 9 2 0 2 1 2 2 2 3 2 4 2 5 0 . 0 0 1 2 6 0 . 0 0 9 2 7 2 8 0 . 2 2 3 2 9 3 0 0 . 0 0 1 3 1 0 , 0 0 2 3 2 0 . 0 1 5 3 3 0 , 0 1 7 3 4 3 5 0 . 0 0 6 3 6 0 . 0 1 1 3 7 0 . 0 0 4 3 8 0 . 0 1 9 3 9 4 0 0 . 0 7 2 4 1 0 . 0 2 1 4 2 4 3 4 4 0 . 0 0 0 4 5 0 . 7 4 3 4 6 0 . 0 3 6 4 7 0 . 0 4 1 4 8 4 9 0 . 0 2 2 5 0 0 . 7 6 0 0 . 0 0 7 0 . 0 0 0 0 . 0 0 0 0 , 0 1 7 0 . 0 2 8 0 . 0 2 7 0 . 1 3 9 0 . 0 8 9 0 . 1 2 4 0 . 1 6 6 0 . 2 3 9 0 . 0 3 7 0 . 0 2 5 0 . 3 0 8 0 . 2 8 7 0 . 2 8 9 0 . 0 2 1 0 . 1 1 6 0 . 1 1 9 0 . 0 6 1 0 . 0 8 3 0 . 0 0 1 0 0 9 3 0 . 1 0 6 0 . 0 1 3 0 . 0 1 3 0 . 1 3 0 0 . 1 1 8 0 . 3 2 9 0 . 0 0 6 0 . 0 0 8 0 . 0 0 8 0 . 0 8 8 0 . 0 2 9 0 . 0 1 1 0 . 0 1 2 0 . 0 1 9 0 . 0 8 3 0 . 0 9 9 0 . 5 0 8 0 . 5 4 2 0 . 4 9 8 0 . 4 3 1 0 . 8 5 6 0 . 7 2 4 0 . 6 7 8 0 . 8 7 0 0 . 8 8 2 0 . 5 8 4 247 Appendix 4. Diet matrix built for the 1950 trophic model of the U G C . Prey \ Predator I 2 3 4 5 6 7 8 9 1 0 1 1 1 Sharks > 120 cm 0 . 0 0 1 2 Sharks 120 cm 0 . 1 7 8 3 Totoaba 0 . 0 7 1 4 Toothed cetaceans 0 . 0 1 1 5 Sea lions 0 . 0 6 0 6 Vaquita 0 . 0 0 0 7 8 Croakers Jacks 0 . 0 3 5 0 . 0 3 5 0 . 2 7 2 0 . 3 4 0 0 . 2 8 2 0 . 0 9 2 9 Corvinas 0 . 0 8 2 0 . 1 2 7 0 . 0 4 6 0 . 1 0 5 0 . 0 4 7 0 . 0 9 2 1 0 Serranids 0 . 0 3 9 0 . 1 7 7 0 . 2 3 2 0 . 1 5 4 0 . 0 7 6 0 . 0 4 2 1 1 Scombrids 0 . 3 2 9 0 . 2 4 6 0 . 0 4 2 0 . 0 4 7 1 2 Snappers 0 . 1 1 8 0 . 0 8 8 1 3 Seabirds 0 . 0 0 0 0 . 0 0 0 1 4 Rays 1 5 Flounders 1 6 Wrasses 0 . 0 2 1 0 . 0 2 0 1 7 1 8 Chano Sea turtles 0 . 0 3 7 0 . 0 0 1 0 . 0 3 7 0 . 0 4 1 0 . 0 2 5 0 . 0 3 1 0 . 0 9 9 1 9 Pre-adult Vaquita 0 . 0 0 0 0 . 0 0 0 2 0 Grunts 0 . 3 9 8 0 . 1 6 3 2 1 Gerreidae 0 . 0 2 7 0 . 0 1 0 0 . 6 3 1 2 2 Guitarfish 2 3 Small demersal fish 0 . 0 8 3 2 4 Other fishes 0 . 1 5 1 0 . 4 1 2 0 . 2 2 6 0 . 2 7 9 0 . 3 4 6 2 5 Octopus 0 . 2 3 6 0 . 2 2 4 0 . 0 3 6 0 . 0 2 2 0 . 0 0 4 2 6 Stomatopods 0 . 0 4 4 0 . 2 3 0 0 . 1 1 8 0 . 1 6 5 0 . 0 1 7 0 . 1 9 7 0 . 0 2 3 2 7 Juv. Totoaba 0 . 0 0 1 2 8 Myctophids 0 . 2 4 7 2 9 Baleen whales 0 . 0 0 0 3 0 Crabs 0 . 0 5 1 0 . 0 1 8 0 . 0 0 0 3 1 Squids 0 . 1 4 9 0 . 0 4 6 0 . 1 3 9 0 . 0 8 3 0 . 0 2 9 0 . 0 4 9 3 2 Small pelagics 0 . 0 6 2 0 . 2 3 6 0 0 6 1 0 . 8 1 3 33 Jellies 34 Planktivorous birds 3 5 Rock shrimp 3 6 Blue shrimp 37 Brown shrimp 3 8 Semi-sessile epibenthos 39 Sea cucumber 4 0 Benthic meoifauna 4 1 Sessile epibenthos 4 2 Mulina coloradensis 43 Bivalves 4 4 Juv. of Blue Shrimp 45 Zooplankton 4 6 Seagrasses 47. Seaweeds 4 8 Phytoplankton 49 Macroalgae 50 Detritus 0 . 0 0 7 0 . 5 7 9 248 Appendix 4. Continuation. 1 3 1 4 1 5 1 6 1 7 1 8 1 9 2 0 2 1 2 2 ' 2 3 2 4 2 5 2 6 2 7 2 8 1 2 3 4 5 6 7 8 9 0 . 1 0 4 1 0 I I 1 2 0 . 0 2 0 0 . 0 1 2 0 . 0 1 1 0 . 0 4 5 0 . 0 2 4 1 3 0 . 0 2 2 1 4 1 5 0 . 0 0 1 0 . 0 0 7 0 . 0 3 9 1 6 0 . 0 4