UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Uniting historic perspectives, human behaviour, and habitat use to assess the future for overfished seahorses O'Donnell, Kerrie Patricia 2011

You don't seem to have a PDF reader installed, try download the pdf

Item Metadata

Download

Media
[if-you-see-this-DO-NOT-CLICK]
ubc_2012_spring_o'donnell_kerrie.pdf [ 5.58MB ]
Metadata
JSON: 1.0072414.json
JSON-LD: 1.0072414+ld.json
RDF/XML (Pretty): 1.0072414.xml
RDF/JSON: 1.0072414+rdf.json
Turtle: 1.0072414+rdf-turtle.txt
N-Triples: 1.0072414+rdf-ntriples.txt
Original Record: 1.0072414 +original-record.json
Full Text
1.0072414.txt
Citation
1.0072414.ris

Full Text

UNITING HISTORIC PERSPECTIVES, HUMAN BEHAVIOUR, AND HABITAT USE TO ASSESS THE FUTURE FOR OVERFISHED SEAHORSES  by Kerrie Patricia O’Donnell  B.A., Brown University, 2002  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE STUDIES (Zoology)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)  November 2011  © Kerrie Patricia O’Donnell, 2011  Abstract  This dissertation casts new light on reconciling fisheries and conservation. This reconciliation is particularly challenging for small-scale fisheries, because they are data-poor, but it is necessary to sustain biodiversity and food security. My research focuses on a typical small-scale fishery in the Central Philippines that catches seahorses. I begin by providing an historic reconstruction of seahorse catch rates, the first of its kind despite a multi-decade, global trade for seahorses. I quantified fisher perceptions of the decline in trade to set historic baselines. Results showed that relying solely on recent decades of data could underestimate IUCN Red List extinction risk for historically exploited fish such as my study species, Hippocampus comes. My research contributed robust approaches to incorporating fisher knowledge into quantitative assessments. I developed a novel approach to correct for inaccuracies in fishers’ memories of past events and showed, for the first time, that conservation assessments are sensitive to assumptions made when analyzing local knowledge. Comparisons I made between catch rate trends estimated by resource users and those obtained using standard approaches (catch landings or underwater surveys) suggest that fisher interviews or logbooks are a reasonable proxy for more costly ‘scientific’ methods. In contrast, interviews do not allow inferences about the absolute value of catch rates at scales relevant to management/conservation. My findings raise questions about the assumed sustainability of small-scale fisheries and identified new strategies to protect them. I conducted a novel analysis of spatial fishing  ii  behaviour that illustrates the capacity of small-scale fisheries to exert considerable cumulative pressure on marine systems, despite their seemingly benign gears. Finally, I conducted the first study on how depth refugia might help sustain coral-reef fish populations by combining fisheries modeling with IUCN Red List conservation assessments. Results emphasize the importance of locating some protected areas at depths inaccessible to small-scale fishers. Such tactics will contribute towards safeguarding biodiversity and food security in coral reef ecosystems. In summary, my thesis takes valuable steps towards finding solutions for marine biodiversity and fisheries crises; it uses pragmatic approaches to deduce key information that is highly relevant to the futures of coral-reef fish and small-scale fisheries.  iii  Preface  Some of the chapters of this thesis and ideas therein are already published. A version of Chapter 2 has been published. O’Donnell, K.P., Pajaro, M.G. and Vincent, A.C.J. (2010) How does the accuracy of fisher knowledge affect seahorse conservation status? Animal Conservation 13: 526-533. Chapter 3 has been provisionally accepted at a peer reviewed journal, Chapter 4 is presently in review, Chapter 5 is in preparation for submission, and some of the ideas in Chapter 6 have been published. O’Donnell, K.P., Pajaro, M.G. and Vincent, A.C.J. (2010) Improving conservation and fishery assessments with local knowledge: future directions. Animal Conservation 13: 539-540. I am the senior author on all papers and all have co-authors. I took primary responsibility for design, implementation, analysis, and writing of all chapters. Details of coauthorship contributions are outlined below. Chapter 2 – Dr. Amanda Vincent came up with the original idea to use existing, unpublished interview and logbook data to reconstruct the history of seahorse fishing. I embraced the opportunity to turn this idea into reality and built on it by linking to IUCN Red List assessments and evaluating the effects of uncertainty on extinction risk. I also developed the methodology to evaluate the effects of uncertainty. I expanded on the original idea, developed and carried out all analyses, and drafted the text. My other co-author, Marivic Pajaro, along with Amanda Vincent, provided the data. Chapter 3 – Dr. Amanda Vincent came up with the original idea to compare datacollection methods and provided fisher logbook and catch landings data. Co-author Dr. Phil Molloy provided guidance in learning R statistical software and also contributed edits,  iv  comments and suggestions for additional analyses on numerous iterations of the paper. I collected all interview data, carried out all data management and analyses, and drafted the text. Chapter 4 – I came up with the original idea to log fisher movements with GPS and depth loggers and to modify swept area method exploitation rate calculations to fit available data. I also devised the methodology for interpreting GPS and depth logger data. Co-authors Dr. Amanda Vincent and Dr. Jeremy Goldbogen contributed to the idea of this chapter through stimulating discussions and pointed comments on drafts of the manuscript. Analysis of diving behaviour was conducted by Jeremy Goldbogen and I conducted all other analyses. Logbook data were provided by Amanda Vincent. I collected all other information for the manuscript, carried out all data management, and drafted the text. Chapter 5 – The original idea of evaluating evidence for a depth refuge from lantern fishing was mine, and co-author Dr. Phil Molloy contributed the idea of constructing a model to evaluate the possible contribution of the refuge to shallow areas and provided edits and comments on numerous iterations of the paper. Both Phil Molloy and co-author Jon Schnute contributed to the methodology of the paper by providing guidance on constructing and validating the model. I collected all information for the manuscript, carried out all data management and analyses, and drafted the text. In addition to contributions described above, Dr. Amanda Vincent provided edits, comments and suggestions for improvement on drafts of each chapter. This research received Human Ethics (H06-80483 B06-0483) and Animal Care (A060318) approval from UBC.  v  Table of Contents Abstract.................................................................................................................................... ii Preface..................................................................................................................................... iv Table of Contents ................................................................................................................... vi List of Tables ........................................................................................................................ viii List of Figures......................................................................................................................... ix Acknowledgements ................................................................................................................ xi Chapter 1: Introduction ........................................................................................................ 1 Rationale ............................................................................................................................... 2 Fisheries ................................................................................................................................ 3 Fish conservation .................................................................................................................. 6 Reconciling small-scale fisheries and conservation ............................................................. 9 Approaches ...................................................................................................................................................9 Indicators of change ....................................................................................................................................14  Context and collaborations ................................................................................................. 17 Research questions.............................................................................................................. 21 Thesis outline ...................................................................................................................... 23 Chapter 2: How does the accuracy of fisher knowledge affect seahorse conservation status?..................................................................................................................................... 25 Introduction......................................................................................................................... 26 Methods............................................................................................................................... 28 Study area, fishery, and target species ........................................................................................................28 Analysis ......................................................................................................................................................30  Results................................................................................................................................. 34 Interview CPUE (1970s to 1994)................................................................................................................34 Logbook CPUE (1996 to 2003) ..................................................................................................................35 Entire fishery CPUE trend (1970s to 2003) ................................................................................................36 Evaluating extinction risk ...........................................................................................................................36  Discussion ........................................................................................................................... 37 Tables and figures ............................................................................................................... 42 Chapter 3: Comparing fisher interviews, logbooks, and catch landings estimates of extraction rates in a small-scale fishery .............................................................................. 45 Introduction......................................................................................................................... 46 Methods............................................................................................................................... 49 Study area, fishery, and species ..................................................................................................................49 Data collection ............................................................................................................................................51 Analysis ......................................................................................................................................................53 Trends in CPUE .....................................................................................................................................54 Absolute value of CPUE ........................................................................................................................55 Correlation – interview and catch landings ............................................................................................56  Results................................................................................................................................. 57 Data structure ..............................................................................................................................................57 Trends in CPUE ..........................................................................................................................................57 Absolute value of CPUE .............................................................................................................................58 Correlation – interview and catch landings.................................................................................................58  Discussion ........................................................................................................................... 59 Tables and Figures .............................................................................................................. 63 vi  Chapter 4: High fishing mortality rate in a small-scale fishery: digital tracking of Filipino lantern fishing. ........................................................................................................ 68 Introduction......................................................................................................................... 69 Methods............................................................................................................................... 72 Study site, fishery, and species ...................................................................................................................72 Spatial fishing behaviour - data collection and analysis .............................................................................73 Fishing mortality rate - analysis..................................................................................................................74  Results................................................................................................................................. 75 Spatial fishing behaviour ............................................................................................................................75 Fishing mortality rate ..................................................................................................................................77  Discussion ........................................................................................................................... 78 Tables and Figures .............................................................................................................. 83 Chapter 5: Out of reach - effects of a depth refuge on the sustainability of a small-scale fishery..................................................................................................................................... 87 Introduction......................................................................................................................... 88 Methods............................................................................................................................... 91 Study area, fishery, and species ..................................................................................................................91 Evidence for depth refuge ...........................................................................................................................94 Effects of depth refuge on sustainability.....................................................................................................96  Results............................................................................................................................... 104 Evidence for depth refuge .........................................................................................................................104 Fisher interviews ..................................................................................................................................104 Fishery independent survey..................................................................................................................104 Effects of depth refuge on sustainability...................................................................................................105 Default values ......................................................................................................................................105 Sensitivity analysis...............................................................................................................................106  Discussion ......................................................................................................................... 109 Tables and Figures ............................................................................................................ 114 Chapter 6: General discussion and conclusions.............................................................. 123 Summary of thesis and status of research questions......................................................... 124 Reconciling small-scale fishing and conservation ............................................................ 130 Seahorses on Danajon Bank......................................................................................................................130 Broader implications and research directions ...........................................................................................134  Literature Cited .................................................................................................................. 139 Appendices........................................................................................................................... 156 Appendix A: Example of a fisher’s ‘personal timeline’ visual aide................................. 157 Appendix B: Additional information about interviews and comparison between focal communities in Chapter 4 ................................................................................................. 158 Appendix C: Evaluation of the swept-area calculation assumption of randomly distributed fishing effort...................................................................................................................... 159 Data collection and analysis......................................................................................................................159 Results.......................................................................................................................................................160  Appendix D: R code to run depth refuge model and supplementary analyses ................ 162 Equations: .................................................................................................................................................162 R Code: .....................................................................................................................................................164 Supplementary Figures .............................................................................................................................167 Effect of the refuge on total catches..........................................................................................................169  vii  List of Tables Table 2.1. Interview schedule ................................................................................................ 42 Table 2.2. Extinction risk categorization for IUCN Red List criteria by time period and CPUE calculation: CR = Critically Endangered, EN = Endangered, VU = Vulnerable, LC = Least Concern ......................................................................................................................... 42 Table 3.1. Interview schedule. .............................................................................................. 63 Table 3.2. Sample sizes (n: total fishers) and best linear unbiased predictor model fit parameters to median CPUE date for 1996 to 2004. .............................................................. 63 Table 3.3. Sample mean (x) in seahorses fisher-1 night-1, Ln + 1 and confidence interval (2 x SE) values for RM-ANOVA analysis by data-collection method. ......................................... 63 Table 4.1. Transit, fishing, and dive statistics for each lantern fisher in Handumon and Batasan recorded by GPS and TDR........................................................................................ 83 Table 5.1. Interview schedule ............................................................................................. 114 Table 5.2. Symbols, definitions, values (and range used in sensitivity analysis), sources of indices and parameters used in model, and justification for default value choices. ............. 115 Table B.1. Interview schedule for discussions with lantern fishers about diving and fishing limits ..................................................................................................................................... 158 Table B.2. We conducted a Wilcoxon rank-sum test (non-normal data) to evaluate whether there were differences in transit, fishing, or dive parameters between our two focal communities of Handumon and Batasan. ............................................................................. 158  viii  List of Figures  Figure 2.1. Danajon Bank study region and our focal lantern fishing community of Handumon............................................................................................................................... 43 Figure 2.2. All CPUE trend calculations for interview and logbook data ............................ 44 Figure 2.3. Frequency and distribution of catch night-1 logbook data. ................................ 44 Figure 3.1. Danajon Bank study region and our focal communities of Handumon and Jandayan Norte........................................................................................................................ 64 Figure 3.2. Frequency of median CPUE for interviews, logbooks, logbooks without zeros, and catch landings 1996 to 2004............................................................................................. 65 Figure 3.3. Median CPUE used to assess trends from 1996 to 2004 for interview, catch landings, logbooks and logbooks without zeros data-collection methods.............................. 66 Figure 3.4. Comparison of pooled CPUE values and RM-ANOVA estimate of CPUE for fisher estimates of a good, bad, and normal night’s catch from interviews and median CPUE from logbooks without zeros, catch landings, and logbooks. ................................................. 67 Figure 4.1. Danajon Bank study region, focal communities of Handumon and Batasan, and illustrations of lantern fishing method. ................................................................................... 84 Figure 4.2. GPS tracks of Batasan and Hanudmon fishers’ night of fishing. ....................... 85 Figure 4.3. Depth profile for fisher #3. ................................................................................. 86 Figure 5.1. Danajon Bank study region, focal communities where interviews were conducted and UVC surey sites. ........................................................................................... 117 Figure 5.2. Conceptual representation of the model. .......................................................... 118 Figure 5.3. Frequency of interview mentions of fishing grounds where seahorses have been caught at particular depths and the range of depths fished by gear. ..................................... 119 Figure 5.4. Proportion of H. comes populations remaining after 10 years at default parameter values. .................................................................................................................. 119 Figure 5.5. Comparison of cumulative catches of H. comes after 10 years from the shallow and deep zones, and total (shallow+deep). ........................................................................... 120 Figure 5.6. Proportion of total H. comes population remaining after 10 years given a range of assumptions about strength of density dependence (rho), refuge size, natural mortality, and exploitation rate, and a range of uncertainty about the proportion of embryos that disperse to the deep. ................................................................................................................................ 120 Figure 5.7. Comparison of cumulative catches of H. comes after 10 years given a range of assumptions about density dependence (rho), refuge size, natural mortality, and exploitation rate......................................................................................................................................... 122 Figure A.1. Example of a fisher’s ‘personal timeline’ visual aide used during interviews about the history of seahorse catches in Chapter 3 ............................................................... 157 Figure C.2. Distribution of fishing effort at each fishing ground visited in 2001 from Handumon lantern fisher logbooks....................................................................................... 161 Figure D.1. Trajectory of a population at the combination of parameters that caused the most severe decline for comparison of amount of decline at 10 years and at equilibrium... 167 Figure D.2. Proportion of total, shallow, and deep H. comes populations remaining at equilibrium at default parameter values................................................................................ 168  ix  Figure D.3. Comparison of cumulative catches of H. comes at equilibrium from the shallow, deep, or total (shallow+deep) zones. .................................................................................... 168  x  Acknowledgements  My deepest and most sincere thank you goes to my supervisor Dr. Amanda Vincent: thank you for challenging me to push my boundaries, for teaching me how to write more effectively, and for your many contributions to this research. Thank you for providing me with the opportunity to work in the Philippines with such devoted and lovely people, and with such a captivating animal. You inspire me with your passion for conservation and drive to find practical solutions for our oceans. I look forward to a future of fruitful collaborations. It has been a pleasure to work with my supportive research committee, Dr. Leonora Angeles, Dr. Isabelle Côté, Dr. Steven Martell, and Dr. A.R.E. Sinclair. You have been an important component of my training as a scientist and I greatly enjoyed our discussions during and outside committee meetings. I believe my work is much richer thanks to your contributions. I am incredibly grateful to my Project Seahorse Post Doc mentor, Phil Molloy. Your enthusiasm for science and conservation, and fascination with the natural world fueled my work and you were so much fun that you reminded me to have some of that too. You also have a way of making the challenging seem possible and even fun that was vital to helping me finish. I promise to pay it forward. My research would not have been possible without generous financial support. For essential and much appreciated personal financial support I thank the National Science Foundation for awarding me a Graduate Research Fellowship. Thank you as well to Peoples Trust for Endangered Species, Toronto Zoo, and Ocean Park Conservation Foundation in Hong Kong for supporting my research and for your commitment to conservation.  xi  Additional financial support was provided by The University of British Columbia’s tuition waiver and international student entrance scholarship. Project Seahorse’s logistical help was facilitated by support from the John G. Shedd Aquarium through its partnership in marine conservation with Project Seahorse, and by Guylian Chocolates Belgium. Additional logistical support came from the Zoology Department graduate program coordinators Alice Liou and Marie Wallden, the UBC Fisheries Centre (especially Janice Doyle and Ann Tautz), and the University of British Columbia. I owe so much to the staff of Project Seahorse Foundation for Marine Conservation (PSF) in the Philippines. Thank you for making it possible for me to work in the fishing villages of Danajon Bank and for welcoming me into your work and lives. Thank you especially to Mia Apurado, Pert Auxilio, Wilson Barbon, Alfie Bartolo, Amado Blanco, Fermina Gensen, Lourdes Labrada, Angie Nellas, Hazel Panes, and Franco Villaruel. On a less formal note, thank you all for your openness which made me feel at home, though I was in a foreign land, and for your friendships that made me smile when I needed to most. I really enjoyed working with you and hope that I get the opportunity to do so again in the future. Daghang salamat! Thank you to my fearless research assistants Mia Comeros, JayR Dongallo, Lucio Minguito, and Gerry Sucano. As my eyes and ears during interviews and my dive buddies during night-time searches for seahorses, you literally were with me through every step of data collection. Thank you for your enthusiasm, dedication, hard work, and attention to detail – it has made for a fruitful research endeavor.  xii  There are no words to express my respect and gratitude to the fishers, fish buyers, and communities of Handumon, Jandayan Norte, Batasan, Guindacpan, Cataban, and Pinamgo. You took the time to share your stories with me and welcomed me into you communities for which I will forever be appreciative. I will not name individuals here so that respondents maintain anonymity, but your names, faces, and words will stay with me forever. Having spent most of my time in Handumon, I am grateful to the barangay captain Rowell Botero and people’s organization leader and Elvie Bohol for granting permission for me to work in your community and facilitating my research. Components of my research benefitted from intellectual input from colleagues whom I acknowledge, gratefully. Thank you to Sian Morgan for conducting excellent research on our favorite study species, H. comes, and for providing advice on everything from measuring and tagging seahorses to career options. A heart-felt thank you to Nick Hill for being a great listener and sounding board for research ideas, and for always being willing to help solve problems in the field that seemed to pop up at every turn. I am indebted to Jon Schnute for his mentorship and advice on the modeling in Chapter 5. Thank you for your interest in my work and your willingness to chat at a moment’s notice. I am very greatful to Jeremy Goldbogen for the creativity, fun, and interdisciplinary rigor he brought to my work. I also greatly appreciate the early mentorship provided by Nathan Taylor on stock assessment methods and research design and advice on statistical methods and analytical techniques from Laura Tremblay-Boyer, Rowan Barrett, Rob Ahrens, Marie Etienne, and Ruth Joy. Many of my chapters benefited greatly from the contributions of colleagues who kindly gave their time to read draft versions of my work. I gratefully acknowledge the support of the following people for their insights on draft proposals and chapters: Megan  xiii  Bailey, Natalie Ban, Iain Caldwell, William Cheung, Sarah Foster, Nick Hill, Ahmed Kahn, Janna Rist, Iain Taylor, and Mai Yasue. Thanks also to my co-authors of various chapters: Amanda Vincent, Phil Molloy, Marivic Pajaro, Jeremy Goldbogen, and Jon Schnute. I would also like to thank the anonymous reviewers, journal editors, and commentators Tim Daw, Shankar Aswani, and Barbara Paterson for improving and broadening my work. It has been amazing to work with Project Seahorse colleagues, friends, and volunteers. In addition to those already mentioned, thanks also to Jonathan Anticamara, Sarah Bartnick, Tarah Brachman, Gina Bestbier, Louise Blight, Mary Cashman, Shannon Charney, Janelle Curtis, Chrissy Czembor, Melissa Evanson, Eli Guieb, Geoff Gilliard, Lana Gunnlaugson, James Hrynyshyn, Bob Hunt, Heather Koldewey, Danika Kleiber, Sara Lourie, Anna Magera, Jean Marcus, Keith Martin-Smith, Christie Michailopoulos, Jorma Neuvonen, Tuya Ocher, Vincent Page, Nishan Perera, Candance Picco, Eve Robinson, Chloe Shen, Heidi Shuter, Marjorie Sorensen, Tyler Steim, and Lucy Woodall. A special thank you goes to Jennifer Selgrath for helping to unearth moldy, old, hard-copy data. To my friends who have journeyed through grad school with me – I couldn’t have made it through without you. You have enriched my life in Vancouver making it truly a home, and will always be in my heart. Thank you especially to Katie Kuker, Kristopher Perta, Meaghan and David Bryan, and Megan Bailey. You all helped me keep laughing. A warm thank you as well to Natalie Ban and Sarah Foster for challenging me to have a better work-life balance and being my supportive, and wise, older ‘lab sisters’ throughout this journey.  xiv  Thank you to my parents, Mark, Jean, Kevin, and especially my mother, Kathy, for your unfailing love and support and for encouraging me to follow my passion even though it may take me to places far from home. And last, but certainly not least, thank you to Stephane for being my anchor for the past few years of this journey. You are a source of strength and happiness that I can hardly put into words. Thank you for being there for all that is most important.  xv  Chapter 1: Introduction  1  Rationale  The capacity of humans to cause declines in wildlife species, especially from direct exploitation, is immense and growing. On land, hunting has caused extinction waves of megafauna (Alroy, 2001). In the sea, fishing has driven whales, manatees, dugongs, seacows, monk seals, sharks, rays and other large marine vertebrates to be functionally or entirely extinct in most coastal systems (Bertram & Bertram, 1973; Kenyon, 1977; Casey & Myers, 1998; Clapham, Young & Brownell Jr, 1999; Baum et al., 2003; Lotze & Worm, 2009). The largest predatory marine fishes have also largely succumbed to extensive, global declines due to chronic overfishing (Pauly et al., 1998; Jackson et al., 2001; Myers & Worm, 2003). It is generally agreed that conventional approaches to control fishing have failed to sustain fish populations or fisheries (Pitcher, Hart & Pauly, 1998; Pauly et al., 2002). Given the importance of fish to human food security (Delgado, 2003; Alder & Sumaila, 2004) and their roles as part of productive and functioning ocean systems (Field et al., 1998; Worm et al., 2006) action must be taken to prevent further declines and to restore the abundance of sea life. This dissertation addresses the major challenge of applying pragmatic approaches to assess and recover fish species threatened by overfishing. Getting this right for small-scale fisheries will be important for sustaining biodiversity and food security, and particularly challenging. I shall describe in more detail the challenges facing the world’s fisheries and fish conservation efforts. I shall then outline emerging approaches to reconcile the two paradigms, and describe metrics to measure progress toward healthy fish populations and fisheries.  2  Fisheries  Fisheries are globally important and in trouble. Fishing directly employs at least 51 million people (Berkes et al., 2001), and international trade in fisheries products generates more foreign exchange than the revenues earned from any other traded food commodity such as rice, cocoa, coffee or tea (FAO, 2003). The tremendous growth of fisheries in the 20th century reached a plateau by the 1980s (FAO, 2009) as new areas to exploit were exhausted (Swartz et al., 2010). The numerous examples of overexploitation causing population declines (Myers & Worm, 2003; Hutchings & Reynolds, 2004) and alteration of ecosystems (Pauly et al., 1998; Watling & Norse, 1998; Jackson et al., 2001; Worm et al., 2006), have led many to perceive a global fisheries crisis (Buckworth, 1998). Collapses of iconic species like Atlantic Cod ended centuries-old fishing traditions in New England and eastern Canada (Myers, Hutchings & Barrowman, 1997). There are examples of fishes recovering from overfishing (e.g. Worm et al., 2009), but in many cases recovery has not occurred, even after fishing ceased (Hutchings, 2000; Hutchings & Reynolds, 2004). While international attention to global fisheries problems has focused on industrial, mostly temperate, single-species fisheries, there is growing recognition that small-scale, multi-species fisheries, specifically those on coral reefs, need substantially more attention (Mahon, 1997; Bellwood et al., 2004; Castilla & Defeo, 2005). The distinction between industrial and small-scale fisheries tends to be by boat size, leading to spatial segregation: industrial fishing refers to large vessels that can operate in the open-ocean, while small-scale fishing refers to small, coastal boats (Chuenpagdee et al., 2006). One reason that attention has turned to small-scale fisheries is because they employ the vast majority of the world’s  3  fishers. It is estimated that fifty million of the world’s fifty-one million fishers are employed in small-scale fisheries (Berkes et al., 2001), though there is some debate about the exact number of people employed (reviewed in Béné, 2006). Small-scale fisheries occur globally, but are of particular concern in developing countries which overlap with coral reef systems (Pauly et al., 2002; Burke et al., 2011) that have been broadly degraded by overfishing (Jackson et al., 2001; Pandolfi et al., 2003). In addition, small-scale fisheries are poorly studied and rarely managed (Berkes et al., 2001; Parma et al., 2003; Chuenpagdee et al., 2006; McClanahan et al., 2009). Although they may catch half the world’s fish, basic information needed to assess the effects of small-scale fishing is largely unavailable - contributing to the challenge of management. Most countries’ official statistics do not specify catch from small-scale fisheries, thus implying that such catch is negligible. In fact, though, the few studies that estimated catch suggest that the few official statistics of small-scale catches are gross underestimates (Zeller et al., 2006a; Zeller et al., 2007; Jacquet et al., 2010). This underreporting tends to under-estimate the effects small-scale fishing on marine systems and their contribution to livelihoods (Zeller, Booth & Pauly, 2006b). Correcting for under-reporting suggests that globally small-scale fishers catch roughly the same amount (about 30 million tones year-1) as industrial fishers (Chuenpagdee et al., 2006; Pauly, 2006). In addition to poor documentation of small-scale catches, there is often no available biological information, even for the economically important species (Sadovy, 2005). Fish biological data are especially scarce on coral reefs because vast areas of tropical reefs have not been studied (Simon & Polunin, 1996). Data poverty has contributed to small-scale fisheries being poorly managed, if at all (Munro, 1996; Berkes et al., 2001). Most are open-  4  access fisheries and in many there is almost no economic constraint on effort; the possibility of a decent catch, however remote, is sufficient incentive for unremitting exploitation of resources (Munro, 1996). Small-scale fisheries have historically been considered ‘unmanageable’ (Orensanz et al., 2005) because conventional fisheries science and management lack methods to address the complexities of small-scale fisheries. While industrial fisheries specialize primarily on a single gear that catches a single species that is landed at few commercial ports, small-scale fisheries employ many gears catching many species with small landings in many places (Russ, 1991). This complexity renders it impractical to collect typical data on fishery status (e.g. catch rates) and implement command-and-control measures (e.g. quotas) for small-scale fisheries. Even if catch and effort data were available for small-scale fisheries, conventional assessment models (e.g. Stock Reconstruction Analysis; Walters, Martell & Korman, 2005) assume that fish populations are mobile, evenly distributed, and that large enough areas are considered to nullify effects of movement. These assumptions do not hold for small-scale fisheries (Parma et al., 2003) such as those on coral reefs where fish are relatively stationary, spatially heterogeneous, structured as metapopulations, and where movement may play an important role in population persistence (Gotelli, 2001). Overall, stock assessment-based fishery management has been too impractical, expensive, incomplete, and uncertain to address the needs of small-scale fisheries (Mahon, 1997; Berkes, 2003). Developing practical approaches to manage small-scale fishing pressure will be especially important for conserving the wealth of biodiversity in coral reefs. Small-scale fisheries tend to co-occur with some of the most biodiverse habitats on earth, coral reefs (Connell, 1978; Roberts et al., 2002). Although coral reefs account for only 0.1% of the  5  world’s ocean, their fisheries resources provide tens of millions of people with food and livelihood (Moberg & Folke, 1999; Pauly et al., 2002). Traditionally, reef fisheries provided tropical coastal communities with an important, sometimes sole, source of protein and livelihood (Sadovy, 2005). However, as early as 1900, these fisheries caused substantial degradation to reef systems (Jackson et al., 2001; Pandolfi et al., 2003). This degradation has been magnified more recently with increased demand for reef species on the international market, e.g. for the live reef food-fish trade (Sadovy, 2005). With both local and international pressures, fishing continues to be among the most pervasive threats to coral reefs (Burke et al., 2011), especially in Southeast Asia (Newton et al., 2007). Increases in reef fishing pressure are likely because globally 75% of coral reefs occur in developing countries where human populations are still increasing (Pauly et al., 2002).  Fish conservation  Fishes are heavily exploited wild animals, but have not been a conservation priority. Instead, global conservation efforts have focused on charismatic, largely terrestrial, megafauna (Clark & May, 2002). Among vertebrates, fishes are the least well-assessed for extinction risk on the IUCN Red List (www.iucnredlist.org) - the most universally accepted and comprehensive system for assessing species extinction risk (Colyvan et al., 1999; de Grammont & Cuaron, 2006; Rodrigues et al., 2006b). Just 11% of described fish species are assessed (as compared to 100% of mammals and birds; Vié, Hilton-Taylor & Stuart, 2009; Zamin et al., 2010).  6  Perhaps fishes have not received conservation attention because they have been assumed to be more resilient to exploitation than terrestrial species. Fish have been hunted for food for millennia (Jackson et al., 2001; Pauly, Watson & Alder, 2005) and were once bountiful which likely created assumptions that as a resource they were inexhaustible (Hutchings, 2001 and citations therein). As recently as 1996, when the first commercially exploited marine fish were listed on the IUCN Red List (www.redlist.org), fisheries experts argued that fishing could not drive fish to extinction (Vincent & Hall, 1996; Mace & Hudson, 1999). Red Listing aims to “convey the urgency and scale of conservation problems to the public and policy makers, and to motivate the global community to try to reduce species extinctions” (IUCN, 2008a). It achieves this by assessing species (e.g. based on population change) and then assigning them a threatened status, from least concern all the way to extinct. Despite the rigor of the Red Listing process, fisheries experts argued that higher fecundities in marine fishes and large scale dispersal would always allow recovery from low numbers and that high natural variability in population size rendered Red List decline criteria too conservative for marine fishes (Matsuda, Yahara & Uozumi, 1997; Musick, 1999). In response, Red List Criterion A that addresses population declines was revised to address their concerns (Dulvy, Sadovy & Reynolds, 2003). A comparison of these revised Red list decline criteria with fisheries stock assessments suggests that both methods provide consistent warnings of population collapse (Dulvy et al., 2005). Current evidence suggests that, in fact, fishes are just as vulnerable to extinction as terrestrial wildlife. Maximum reproductive rates (recruits per spawner/adult) for fishes are similar to those estimated for terrestrial mammals such as the grizzly bear (Ursus arctos), bighorn sheep (Ovis Canadensis) and white-tailed deer (Odocoileus virginiansus)  7  (Hutchings, 2001). Across fish species that vary widely in fecundity (number of gametes), reproductive rates are fairly consistent (Myers, Brown & Barrowman, 1999). Genetic evidence of juvenile fish returning to natal reefs, suggests that population distributions and sizes may be considerably more restricted than is assumed (reviewed by Dulvy et al., 2003). In addition, fish populations fluctuate no more from year to year than do populations of mammals and butterflies (Fagan et al., 2001). Therefore, there is little evidence to suggest that fishes warrant different at-risk criteria than other organisms (Hutchings, 2000; Dulvy et al., 2003; Dulvy et al., 2005). In fact, fishes may be more vulnerable to extinction than taxa like butterflies and birds (Fagan et al., 2001). Although efforts to quantify the global conservation status of marine fishes are in their infancy, continued pressures are cause for concern. Although only 32 populations and species of marine fish have been documented as experiencing local, regional, or global extinctions, our state of knowledge is so poor that many more extinctions may well have occurred without being noticed (Dulvy et al., 2003). Groupers (Serranidae) are the first fish family for which all species’ risk of extinction has been assessed using IUCN Red List criteria (IUCN, 2001). For this family, at least 12% of the world’s 161 species have been listed in threatened categories (Critically Endangered, Endangered, or Vulnerable), another 14% are Near Threatened, and 30% are considered to be Data Deficient (Polidoro et al., 2008). Similarly, the first regional assessment of all fish species in the Mediterranean showed that while only 8% of the 519 native fishes assessed were classified as threatened, about one-third were classified as Data Deficient. The good news was that more than half of the assessed species were not facing immediate risk of extinction (Least Concern). On the  8  other hand, with over half of species subject to fishing, and many more not protected or effectively managed, the true proportion of threatened species might be much higher.  Reconciling small-scale fisheries and conservation  There is a call for a pragmatic and effective system to maintain long-term viability of fish populations, benefiting both fisheries and conservation. This reconciliation of fisheries and conservation would contribute to sustainable fisheries - those that meet the needs of the present without compromising the ability of future generations to meet their own needs (UN, 1987). The global fishery crisis is a product of the conventional fisheries science and management paradigm (Buckworth, 1998). This conventional system relies on expert scientific knowledge to make top-down decisions, aims for the control of nature, seeks to maximize yield from wild species, and treats people as if they were separate from the environment (Pitcher et al., 1998; Berkes, 2003). Although entire books are devoted to ‘reinventing fisheries management’ (e.g. Pitcher et al., 1998; Pomeroy & Andrew, 2011), problems persist. I here outline five promising approaches and four indicators that can measure progress toward healthy fish populations in small-scale fisheries.  Approaches  Instead of the conventional approach to fishery assessment and management, a more pragmatic system would take five approaches, each of which I assess in turn:  9  1) tap into unconventional data-sources (Walters, 1998; Berkes, 2003) for rapid, costeffective, and reliable information on species abundance and exploitation; 2) consider a broader social-ecological system (Berkes, Folke & Colding, 2000) including the behaviour of fishers (Hilborn, 1985; Salas & Gaertner, 2004); 3) evaluate the effects of uncertainty on assessments and management options (Ludwig, Hilborn & Walters, 1993; Walters, 1998; Kinzig & Starrett, 2003); 4) reach conclusions and recommendations for action despite incomplete knowledge (Ludwig et al., 1993; Johannes, 1998; Robinson, 2006); and 5) err on the side of precaution – aiming to protect the viability of the resource (Ludwig et al., 1993).  Tap into unconventional data sources - local knowledge  Despite a growing interest in incorporating local knowledge into quantitative conservation (Danielsen, Burgess & Balmford, 2005) and fishery assessments (Neis et al., 1999b; Johannes, Freeman & Hamilton, 2000), there remain practical impediments to its use. Local knowledge can be defined as practical knowledge that has not necessarily been passed down through generations like traditional ecological knowledge (Berkes, 2003). Documenting the knowledge of resource users can be an inexpensive (Moore et al., 2010; Rist et al., 2010) and effective way of gathering data on species, catches, ecology, and habitat, especially to understand changes over time (Johannes, 1981; Mackinson & Nottestad, 1998; Neis et al., 1999b; Johannes et al., 2000; Scholz et al., 2004). Interviews with resource users (e.g. Sáenz-Arroyo et al., 2005) and their daily written logs (Baum et al., 2003;  10  Rosenberg et al., 2005) have even uncovered population declines that were previously undetected by western science (e.g. Gilchrist, Mallory & Merkel, 2005). Since the first studies that demonstrated the utility of local knowledge to quantitative fishery management (e.g. Neis et al., 1999b), there have been few conservation or fishery assessments based on local knowledge (Brook & McLachlan, 2008). Perhaps the most significant impediment to more widespread incorporation of local knowledge into quantitative assessments lies in appraising its reliability (in terms of accuracy and precision), especially because there are no formal guidelines for interpretation or analysis of local knowledge data. In addition, while biases in local knowledge have been acknowledged (Neis et al., 1999b; Ainsworth, Pitcher & Rotinsulu, 2008; Daw, 2008a), there is little understanding about how different assumptions of bias affect quantitative conservation and fishery assessments.  Understand fisher behaviour  Understanding the spatial dynamics of fishing behaviour can be a pragmatic way to improve assessment and management of industrial and small-scale fishing alike (Hutchings & Myers, 1994; Booth, 2000; Wilen et al., 2002). There is an increasing awareness that knowledge of the behaviour and decision-making process of fishers is essential for effective assessment and management (Hilborn, 1985; Salas & Gaertner, 2004; Walters & Martell, 2004; Branch et al., 2006). For example, knowledge of spatial fishing patterns allows for better interpretation of catch rates (catch per unit of effort – CPUE) because fish population declines can be masked by changes in spatial distribution of fish and/or fishing (Gillis,  11  Peterman & Tyler, 1993). Such misunderstandings may help explain the causes of the most severe fishery collapses (e.g. Atlantic cod, Peruvian anchovy; Hilborn & Walters, 1992). Although fisher behaviour affects trends in catch rates, little is known about fishing behaviour, especially for small-scale fisheries (Salas & Gaertner, 2004). Understanding where fishers fish also highlights areas where fishing effort is limited or non-existent (natural refugia), areas that could be important for fishery sustainability (Pauly et al., 2005). Low-tech, cost effective interviews with small-scale fishers about where they fish provides useful, but coarse, information (e.g. Daw, 2008b). Digital tracking technologies, more commonly applied to monitor animal movements (e.g. digital tags; Goldbogen et al., 2008) or industrial fisheries (e.g. GPS tracking; Bertrand et al., 2007) are becoming more affordable and demand less of fishers’ time than interviews. These highertech, spatial, data-collection methods could provide complementary detailed spatial information to interviews.  Evaluate effects of uncertainty on assessments and management options  The complex and dynamic nature of human interactions with ecosystems means that scientists and decision-makers must expect and learn to manage associated uncertainty (Ludwig et al., 1993; Myers, 1995; Kinzig, 2003; Robinson, 2006). Many important processes that structure fish populations, and therefore affect their resilience to fishing (e.g. population connectivity through dispersal, Jones et al., 2009) are still poorly understood (Sale et al., 2005). When urgent decisions need to be made about threatened species, modeling is one pragmatic tool that can combine available data from various sources and  12  systematically incorporate uncertainties to weigh management or recovery options. As a representation of how we think the world works, a model can allow for assumptions to be explicitly stated, uncertainties quantified, data gaps identified, and future data gathering efforts prioritized (Hilborn & Walters, 1992). Although classic fisheries modeling approaches are technically complex and relatively data-hungry, even simple or conceptual models can be useful (Milner-Gulland & Rowcliffe, 2007). Furthermore, it is becoming possible to populate models with data from unconventional sources (e.g. Ainsworth et al., 2008).  Reach conclusions and recommendations for action despite incomplete knowledge  Reaching conclusions in the absence of complete knowledge is necessary because conservation problems are often too complex for scientific methods to provide robust answers (Robinson, 2006). Problems such as conservation of world fish species are ‘wicked problems’ (Rittel & Webber, 1973) that involve many disciplines and have no test for a solution (Ludwig, Mangel & Haddad, 2001). Because of the complexity of these problems, controlled and replicated experiments are not possible in large-scale, marine systems such that optimum (maximize catch without threatening species viability) levels of exploitation can probably only be determined by trial and error (Walters, 1986; Kinzig & Starrett, 2003). A result is that consensus among scientists is unlikely, to a point where some have recommended against consensus as a requirement for action (Ludwig et al., 1993). Careful consideration of the management objectives and controls that are practical for threatened  13  marine resources for which few or no data are, or will be, available is overdue (Johannes, 1998). As Johannes so eloquently put it, “The key question should not be ‘what data are needed to make sound management decisions?’ but rather, ‘what are the best management decisions to make when such data are unattainable?” (Johannes, 1998).  Err on the side of precaution – aiming to protect the viability of the resource  Uncertainty mandates a more cautious approach to resource exploitation than is the norm (Ludwig et al., 1993). In reaction to uncertainty, fisheries stakeholders often prefer that governments prove negative effects of fishing before introducing more restrictive fishing regulations (Walters & Martell, 2004), but a number of international agreements pertaining to fisheries reject this notion. Both the Law of the Sea (UN, 1995) and the FAO Code of Conduct for Responsible Fisheries (FAO, 1995) reverse the ‘burden of proof’ in favor of the precautionary principle. According to this principle, management systems should err on the side of conservation, especially when there is chance of irreversible changes that may degrade the equity of future generations (Berkes et al., 2001). Erring on the side of conservation means that the ‘burden of proof’ shifts such that the onus is on fisheries stakeholders to show that odds of harm due to fishing are low (Charles, 1998).  Indicators of change  To understand the efficacy of the five approaches I outline, measurable indicators are required to track progress toward healthy fisheries and fish populations. These indicators are  14  also helpful because typical, more exhaustive metrics (e.g. direct counts of fish population size and changes in response to fishing) are rarely available for small-scale fisheries and threatened species. While some of the four indicators described below are commonly used in more data-rich situations, I will either use more pragmatic data to calculate them or revise the indicator itself to suit my limited data scenario.  Trend in CPUE  Trends in CPUE have been a key indicator of the effects of fishing on fish (despite flaws; Harley, Myers & Dunn, 2001), but tend to be collected by well-funded governments and are therefore largely unavailable for small-scale fisheries (FAO, 2009). I explore a more local, rapid approach to collecting CPUE data by using fisher-derived methods such as logbooks and interviews. In a short period of time, interviews can provide an historic baseline for fish abundance and fishing pressure, but little research had previously been done to examine the reliability (precision and accuracy) of collecting CPUE from fisher perceptions.  Extinction risk – IUCN Red Listing  Percent decline in CPUE, classically a fisheries metric, can also be used to evaluate species conservation status using IUCN Red List criteria. Red List assessments are most commonly based on ecological data (McClenachan, Ferretti & Baum, in review), but  15  exploring results of assessments based on alternative, more pragmatic sources like local knowledge could speed the assessment process for under-assessed taxa such as fishes.  Exploitation rate  The proportion of a population removed by fishing, exploitation rate, is an important indicator to assess overfishing. More specifically, exploitation rate quantifies the mortality associated with fishing relative to some pre-determined reference exploitation rate. However, typically it is too data hungry to be practical for use in small-scale fisheries. As a common fishery reference point, there are general guidelines for safe ranges of exploitation, based on life history. For example, it has been suggested that maintaining a fixed exploitation rate at a relatively small fraction (0.5-0.8) of the natural mortality rate will provide nearly optimum yields (Martell, Walters & Wallace, 2000). Although exploitation rate is usually calculated as the proportion of a population caught, population size is often not documented in data-poor small-scale fisheries (Orensanz et al., 2005). As an alternative, fishing mortality can be calculated as the proportion of a population’s distribution that is swept by fishing effort during a year (the 'swept area' method, Walters & Martell 2004). This fishing mortality (F) can be converted to exploitation rate (U) as U = 1 − e − F . If data on species distribution are unavailable, fisher interviews could fill the gap, and the area swept by fishing can be gathered relatively quickly and inexpensively using GPS technology as described above.  16  Area protected from fishing  The presence of either natural or managed refugia can be an important indicator of extinction risk and fishery sustainability. Natural refugia alone can promote sustainability because extinction from overexploitation is less likely if some of its range remains untouched (Clayton, Keeling & Milner-Gulland, 1997; Pauly et al., 2005). Natural refugia in the marine realm are disappearing (Morato et al., 2006; Swartz et al., 2010) and little is understood about how fished populations are linked to those in refugia (Tyler et al., 2009). Exploring fisher behaviour can help identify potential refugia by outlining constraints on fishing (inaccessible habitats, depths, etc). Conversely, if fishing behaviour changes in a way that expands the reach of exploitation into former refugia, then populations and fisheries may quickly become threatened. Managed refugia can be used to compensate for a lack of knowledge about the fishery, by taking the precaution of excluding the fishery from accessing a sufficiently large proportion of the resource (Mahon, 1997). The location and size of refugia required to protect species and maintain fisheries are still debated (reviewed by Gaines et al., 2010).  Context and collaborations  I have found a system that allows me to explore practical approaches to assessing and addressing a threatened fish exploited by a small-scale fishery. My research is focused on seahorses caught by lantern fishers in the Danajon Bank region of the Central Philippines.  17  Because of concerns over un-sustainable international trade in seahorses for use in traditional medicines, curiosities, and aquarium display, all species of seahorse (genus Hippocampus) were listed on Appendix II of CITES in 2002 (Foster & Vincent, 2005). This listing does not ban the trade of seahorses, but instead CITES member countries must prove any seahorse exports are not detrimental to survival of their wild populations. One difficulty is that we lack understanding of the status of most seahorse populations (25 of the 33 IUCN Red List assessed species are listed as Data Deficient). Many seahorses are caught in smallscale fisheries, which are globally poorly understood and ineffectively managed (Berkes et al., 2001; Parma et al., 2003; Chuenpagdee et al., 2006; McClanahan et al., 2009). Devising practical ways to assess and manage small-scale fisheries in the Philippines is a pressing issue. Reliance on fish for food (Fernandez, 2009), in combination with high human population pressure (human population growth rates have grown on average 5.4% per year from 1900 to 1977; Pauly, 2000) have resulted in intense pressure on marine resources in the Philippines. The vast majority of people (90%) who fish in the Philippines are engaged in small-scale fisheries (Luna et al., 2004). On paper, the country has laudable management goals (e.g. manage for MSY, Fernandez, 2009) and progressive policies (e.g. management of marine resources has been devolved to local governments) and certainly there have been some successes (e.g. establishment and successful recovery of fish and fishery resources at Apo island; Russ et al., 2004), but poor implementation of these rules helps explain the widespread and severe habitat degradation and collapse of fishery resources (Pauly, 2000; Green et al., 2003; Stobutzki et al., 2006). Historically the Philippines was a major exporter of seahorses (Vincent, 1996), but exports have officially ceased since the domestic Fisheries Code interpreted the CITES listing of seahorses as a complete ban on  18  extraction, despite the intended sustainable use provisions of CITES Appendix II (Morgan & Lourie, 2006). The Danajon Bank region of the Central Philippines is of particular conservation concern because high biodiversity, destructive fishing, and poverty are magnified here. Danajon Bank is a 145 km long double barrier reef that hugs the northern and western coasts of the province of Bohol and has an area in excess of 2350 km2 (Hansen et al., in press). This region has been shown to represent the peak of global marine biodiversity (Carpenter & Springer, 2005) and to contain the greatest diversity of seahorses (Lourie et al., 1999). Destructive fishing (e.g. trawling and blast fishing) has degraded a once-rich coral reef system to the point where marine life is grossly depleted and coastal communities are among the poorest in the Philippines (Green et al., 2003). Overfishing may also reach global peaks here (Lachica-Aliño, Wolff & David, 2006), raising conservation and food security concerns (Roberts et al., 2002). The lantern fishery in the Danajon Bank region, whose fishers are locally known as manuga’ay, is a typical small-scale fishery that is open access, spatially-dispersed, catches many species, and has never been formally assessed for sustainability. Lantern fishers sling a kerosene lantern across the bow of a small outrigger canoe at night to illuminate the reef below while they swim and free dive in search of fish, including seahorses, and invertebrates. Assessment of lantern fishing effects on seahorse populations of this region began in 1995 with the establishment of a community-based research and management programme (Martin-Smith et al., 2004; Vincent et al., 2007; O'Donnell, Pajaro & Vincent, 2010) by Project Seahorse. One focus of research has been on understanding the life history, ecology, and vulnerability of the most exploited seahorse species, the tiger tail seahorse, Hippocampus  19  comes Cantor 1850 (Perante, Vincent & Pajaro, 1998; Morgan & Lourie, 2006; Morgan & Vincent, 2007). As part of a larger effort to estimate the extent of the global trade in seahorses (Vincent, 1996), documentation of seahorse landings on Danajon Bank revealed very low CPUE (Vincent et al., 2007), especially compared with historical accounts (Pajaro et al., 1997). In addition, qualitative measures suggest that the fishery is unsustainable (Martin-Smith et al., 2004). Lantern fishers and their communities have taken the first steps to address declining seahorse catches. Consultations with fishers, managers, and scientists about preferred methods of addressing declines revealed that MPAs and Minimum Size Limits (MSLs) are the most preferred management options (Martin-Smith et al., 2004). Many of the communities of the Danajon Bank region have embraced community-based management of no-take marine protected areas (MPAs) as a remedial fishery management approach (Hansen et al., in press). While one community perceived their MPA to be improving fish populations and catches (Yasué, Kaufman & Vincent, 2010), the effect of MPAs on seahorses remains unclear and MPAs alone will certainly not be enough to recover depleted populations and catches. The first prediction of the effects of management options on H. comes viability suggested that slot sizes (smallest and largest individuals are not fished) would be the most precautionary tool to hedge against fishing driving H. comes extinct (Morgan, 2007). Although the lantern fishery is now arguably the best studied fishery for seahorses in the world, robust assessments of the effects of fishing on populations are required to inform prioritization of recovery/management options. The need to assess and address seahorse exploitation is especially urgent because fishing effort continues un-checked. Despite the  20  national ban on extraction of seahorses even for local use, fishing continues in some areas due to lack of enforcement of the ban (O’Donnell personal observation) and a persistent need for fishers to maintain a livelihood. To explore pragmatic ways of quantifying extinction risk and lantern fishery sustainability on seahorses, I collaborated closely with the Project Seahorse Foundation for Marine Conservation (PSF), a Filipino NGO. It has a strong tradition of facilitating community-based management of marine resources in Danajon Bank and its ties with communities afforded me the opportunity to live in the fishing villages where I conducted my research. My assessments of the characteristics of the lantern fishery, its effects on seahorses, and what results mean for the future can contribute to PSF’s ongoing efforts to assist communities in protecting seahorse populations and improving catches. In addition, my work can assist PSF in its advisory role to the Philippines federal fisheries agency that imposed the ban on seahorse fishing. This study system allows me to explore the five pragmatic approaches and four indicators of success I described earlier to evaluate the viability of fish populations and fisheries.  Research questions  My research applies unconventional approaches and indicators to assess risk of extinction and fishery sustainability. I also evaluate the advantages and limitations of the approaches and indicators, and use them to explore assumptions about the sustainability of small-scale fishing. Despite the uncertainty that will always accompany studies of such  21  complex systems, I strive to make reasonable recommendations to those responsible for ensuring the sustainability of populations and fisheries. The following five overarching questions map onto the pragmatic approaches and indicators described above: 1) How has CPUE changed over time according to fisher perception and what does that tell us about extinction risk? (approach: novel data; indicators: CPUE, Red Listing) 2) How do assumptions about the accuracy of fisher recall affect the assessment of extinction risk? (approach: novel data, consider uncertainty; indicator: Red Listing) 3) How comparable are three methods of quantifying CPUE (fisher interviews, fisher logbooks and ‘scientific’ catch landings)? (approaches: novel data, consider uncertainty; indicator: CPUE) 4) What are the spatial dynamics of small-scale fishing? And what does an estimate of exploitation rate say about the sustainability of small-scale fishing? (approach: consider fisher behaviour; indicator: exploitation rate) 5) What role does a depth refuge play in preventing short-term risk of species extinction and in supporting shallow catches given current exploitation rates? (approach: novel data, consider uncertainty, be precautionary; indicator: Red List, area protected) I explore these questions in the data chapters and a general discussion described in more detail below.  22  Thesis outline  This thesis has four data-based research chapters, followed by a general synthesis and discussion of implications of the work for conservation and management. In Chapter 2, I link fisher interviews and fisher logbooks to reconstruct the history of CPUE as perceived by fishers. I also quantify assumptions about how accurately fishers report past events (retrospective bias) and examine how the assumption I make about retrospective bias affects the characterization of changes in the fishery and extinction risk. In Chapter 3, I delve deeper into the reliability of data elicited from resource users, comparing them with data obtained using scientific approaches. I evaluate the degree of consensus among three methods of quantifying catch rates (CPUE): fisher interviews, logbooks, and catch landings. I compare trends and absolute values of CPUE among datacollection methods for the lantern fishery. I also test whether interview reports are correlated with recorded landings to evaluate how interchangeable the two methods might be. In Chapter 4, I quantify spatial fishing patterns and depth limits to lantern fishing, using GPS and depth tags. These data, superimposed onto satellite images of habitat and combined with direct observation and interviews, create a three-dimensional representation of spatial fishing behaviour that allows for estimation of exploitation rate – an important indicator of sustainability. Given steep declines in catch rate (Chapter 2) it is surprising that catches of seahorses persist (Chapter 3), but because fishing activities are restricted to shallow areas (Chapter 4), deeper waters may provide natural refugia for exploited species. Therefore, in Chapter 5, I examine whether seahorses have a depth refuge from lantern fishing and assess the  23  importance of the refuge for maintaining population size and catch of the targeted species, H. comes. I construct a deterministic population accounting model that integrates available life history information with estimates of current exploitation rate (Chapter 4) to predict population sizes, extinction risk, and fishery benefits under different assumptions about the size of the refuge, amount of dispersal between fished and refuge zones, density dependence, and natural mortality. Finally, I end with a general discussion of the findings presented in this thesis, and how those findings contribute to reconciling fisheries with conservation for small-scale, datapoor fisheries. I also draw upon my new understanding of the sustainability of the lantern fishery to make recommendations for future research, management, and conservation priority actions.  24  Chapter 2: How does the accuracy of fisher knowledge affect seahorse conservation status?  25  Introduction  Historic baselines for the world’s ocean fisheries are largely unknown (Pauly, 1995) and it is generally impractical to gather basic information on the many unstudied regions (Johannes, 1998). Formal scientific data required for managing many marine resources have not been gathered. Moreover, the large scale of most exploited areas means that controlled and replicated experiments are impossible (Ludwig et al., 1993). Catch rates (catch per unit of effort – CPUE) have been used in formal fishery assessments as proxies of changing fish abundance, but have only been systematically collected for commercially important industrial fisheries (FAO, 2009). Small-scale fisheries that employ 98% of the world’s fishers (Berkes et al., 2001) and land more than half the world’s annual marine catch (Davy, 2000) are poorly understood because there is little formal research or monitoring of exploitation. Careful accumulation of fishers’ knowledge is beginning to fill gaps in our quantitative understanding of the threatened status of marine species and ecosystems. We know from government reports, local knowledge and fisher logs that the Chinese bahaba (Bahaba taipingensis) and giant humphead parrot fish (Bolbometopon muricatum) have been fished to the brink of extinction (Sadovy & Cheung, 2003; Dulvy & Polunin, 2004) and current cod stocks (Gadus morhua) on Canada’s Scotian Shelf are a small fraction of their historic numbers (Rosenberg et al., 2005). In Indonesia, analysis of local knowledge suggests ecosystem-wide decline of exploited coral reef species (Ainsworth et al., 2008). Despite a growing interest in incorporating fisher knowledge into quantitative fisheries and conservation assessments, there remain practical impediments to its use. When  26  compared directly, fisher and scientific knowledge have been found to both agree (e.g. Neis et al., 1999b; Begossi, 2008; Lozano-Montes, Pitcher & Haggan, 2008) and disagree (e.g. Otero et al., 2005; Daw, 2008a), raising concerns about the accuracy of fisher knowledge. Work reviewed by Daw (2008a) illustrates that it is easier to recall events that are unusual or rare, especially if they were positive (e.g. a particularly lucrative fishing excursion), than other less striking circumstances (Bradburn, Rips & Shevell, 1987; Tourangeau, 2000). Rare, positive events may thus be overemphasized in retrospective interview datasets. ‘Retrospective bias’ has been recognized in recording historic fisher perceptions (Neis et al., 1999b; Ainsworth et al., 2008), but to our knowledge there are no reports on how different assumptions of bias affect quantitative assessments. We report our efforts to address retrospective bias in CPUE data for a small-scale, data poor seahorse fishery in the Philippines. Our work had three objectives: (1) reconstruct CPUE using fisher interviews and daily logbooks; (2) consider how assumptions about the accuracy of fisher recall affect the assessment of conservation status for that species; and (3) compare current and past conservation status. We do not test the reliability of interviews by directly comparing them with logbooks because the datasets did not overlap in time; instead, we explore simple calculations of CPUE for logbooks that might account for recall bias in interviews, allowing us to link interview and logbook data into one time series. Seahorses are collected as a cash commodity for export as non-food items, either dried for use in traditional medicines and as curiosities, or live for the aquarium trade. They have been traded internationally for decades, but we lack understanding of the status of most populations, hence we turn to fisher knowledge. Almost all seahorses are caught in smallscale fisheries that are not formally monitored or managed, but trade analyses led to  27  estimates of at least 20 million dried animals per year in 1995 (Vincent, 1996). Furthermore, all qualitative indicators of overfishing reveal that seahorses are overfished (Vincent, 1996; Martin-Smith et al., 2004). The IUCN Red List (IUCN, 2008a) classifies extinction risk for 33 seahorse species: seven as Vulnerable (including our study species Hippocampus comes), one as Endangered, and 25 as Data Deficient (i.e. having insufficient data to be classified). The Philippines has many unmanaged and unrecorded fisheries and has historically been a dominant source of seahorses in international trade. Almost all of the millions of seahorses exported (Vincent, 1996) are obtained in the small-scale fisheries that engage 90% of Filipino fishers (DA-BFAR, 2004). Exports have been recently curtailed by domestic legislation that has inadvertently (as a consequence of international trade restrictions) prohibited all seahorse exploitation since 2004. Nonetheless, domestically illegal fishing continues because there is little enforcement, seahorses command a high price on the internationally legal market, and fishers have few alternative sources of income. Quantifying the extent of declining seahorse catch rates is one way to estimate key parameters needed to evaluate effects of possible recovery options.  Methods Study area, fishery, and target species  Our study takes place in the Danajon Bank region of the Central Philippines (Fig. 2.1). Destructive fishing has degraded this once-rich coral reef system to the point where marine life is grossly depleted. The people living in the region’s coastal communities are  28  among the poorest in the Philippines. In 2002, 80% of fishing households in Northern Bohol fell below the poverty threshold (Green et al., 2003). Our focal community, Handumon, has the largest concentration of lantern fishers who catch the most seahorses in the region (Meeuwig et al., 2003). Project Seahorse has been working with the Handumon community since 1996 as part of a larger initiative to empower them to manage their marine resources sustainably. Although the lantern fishery is the most extensively studied seahorse fishery, it is understood solely through fisher-derived knowledge and trade records. Lantern fishers swim at night under a kerosene lantern, and free-dive for seahorses, other finfish and invertebrates. This open access, small-scale, multi-species fishery has not been formally monitored or managed, but fishers and traders have reported declines in seahorse availability (Vincent, 1996; Pajaro et al., 1997), increasing demand, price, and effort, suggesting overexploitation (Martin-Smith et al., 2004). Analysis of the recent fishery suggests current CPUE is seasonal and much lower than the anecdotal fisher reports of historic CPUE (Vincent et al., 2007). The historic anecdotes and magnitude of the decline have not yet been thoroughly assessed. The tiger-tail seahorse, Hippocampus comes Cantor 1850, is the most frequently caught seahorse species in the lantern fishery (Vincent et al., 2007). Hippocampus comes is found at night on coral reefs, sponge, and seaweed habitats (Perante et al., 2002). It is listed as Vulnerable on the IUCN Red List (IUCN, 2008a) and studies of its life history, abundance, and distribution suggest that it fares poorly under high levels of fishing pressure (Morgan & Lourie, 2006; Morgan & Vincent, 2007).  29  Analysis  To reconstruct the history of H. comes exploitation by lantern fishers we searched Project Seahorse unpublished data and reports. The only record of the earliest years of the fishery was contained in retrospective fisher interviews, whereas catch rates for more recent years were recorded in daily fisher logbooks. Because these two datasets do not overlap in time we cannot be sure how comparable they are. Interview data may be biased toward maximum catch nights, so we tried a number of calculations that could correct for retrospective bias linking interviews to logbooks as a single trend. Although CPUE data may not be a reliable measure of changes in exploited populations (Harley et al., 2001) we used it because: (1) it is the only metric of historic change available; (2) previous work suggested that it may be a robust indicator of H. comes population size (Vincent et al., 2007); and (3) CPUE is an accepted proxy for abundance for IUCN Red Listing (IUCN, 2008b).  Interview CPUE (1970s to 1994) data collection  From March to July 1995 Project Seahorse conducted extensive semi-structured interviews to document fisher understanding of the biology, fishery, trade, conservation and management of local seahorses. The second author, a Filipina, conducted interviews with 21 of the approximately 35 seahorse fishers in Handumon in the local Cebuano or Filipino language. All fishers were invited to be interviewed; all who were willing and available were  30  included. Their individual fishing experience ranged from six months to 24 years, covering the range of experience of all fishers. From these interviews, we extracted only respondent biographical information and answers to three structured questions focused on understanding fishers' perceptions of historic change in the seahorse fishery between 1969 and 1993 (Table 2.1). Fishers were asked if CPUE had increased or decreased in recent years (Q1) and completed trend diagrams (Deguit et al., 2004) representing relative change in CPUE over their career (Q2). Fishers were also asked to recall a typical catch night-1 for a few specific years (Q3). Years were chosen to correspond with events that might aide fishers’ recall, e.g. change of presidents, large typhoons, or something personal like the birth of a child (Means & Loftus, 1991). The number of responses to Q3 varied by time period (mean ± 1 S.D.: 12 ±7), with fewer fishers still alive and able to recall earliest catches than for more recent years. Our best estimate of the total number of seahorse fishers in the first decade of the fishery (roughly the 1970s) was between 5 and 10 individuals, hence the two fishers interviewed comprise between 20 and 40 % of all those who were fishing.  Interview CPUE analysis  To convert each fisher’s trend diagram (Q2) into a relative percent change in career CPUE we totaled the number of X’s for the initial year of a career (CPUE initial) and the most recent year of career (CPUE final). We then calculated change in career CPUE as (CPUE final – CPUE initial)/CPUE initial. We analyzed typical catch night-1 (Q3) raw interview transcripts  31  by standardizing responses, compressing years into bins, and then calculating two CPUE metrics corresponding to different assumptions about retrospective bias.  (i) Interview median: With this metric, we assumed that when fishers were asked about ‘typical’ CPUE, they reported a measure of centre, with some error. We choose a median, rather than mean, because distributions of counts (number of seahorses) tend to be skewed, making the median a more suitable measure of central tendency. Fishers differed in how they provided ‘typical’ CPUE: some gave a range while others a single value. To standardize responses, when fishers responded with a range, we took the median value; when they gave a single value, we used that. We then grouped years into bins that were larger for earlier time periods. Because our goal was to look at long-term trends, and we assumed fishers’ ability to recall precisely which year a CPUE occurred would be worse for dates farther back in time (Bradburn, 2000), we grouped responses into four categories: 1994, 1990, during the 1980s, during the 1970s. We then calculated the median of all fishers’ CPUEs in a given time period to produce a single CPUE value for each time period.  (ii) Interview max: With this metric, we assume that fishers consistently over-estimated CPUE. We chose only the highest CPUE for each time period as the single yearly maximum CPUE.  32  Logbook CPUE (1996 to 2003) data collection and analysis  From 1996 to 2003, fishers worked with Project Seahorse to record daily seahorses catch in logbooks. Logbooks were completed by 43 fishers between 1996 and 2003 with a mean participation rate in the logbook program of 16 ±5 (1 S.D.) fishers year−1 (Vincent et al., 2007). To minimize bias we present results only for years that had equal sampling during peak and lean seasons. We then did three calculations to explore how we might ‘correct for’ biases in interview data allowing us to link interview and logbook data into a single trend:  (i) Logbook median: For each fisher, we calculated the median catch night-1 for each year from logbook data. We then took the median of these catches per fisher to produce a single CPUE value for each year. (ii) Logbook median no zeros: We repeated the calculations used for Logbook median, but removed nights on which zero seahorses had been caught. (iii) Logbook max: We took the highest catch night-1 for each year as the single yearly maximum.  Evaluating extinction risk  Drawing on CPUE trends calculated from interview and logbook data we evaluated the extinction threat for H. comes using the IUCN Red List (www.redlist.org) criteria (IUCN, 2001). We calculated the percent decline in interview and logbook CPUE from the start to  33  the end of three time periods, the most recent of which (1990 to 2001) spanned interview and logbook datasets. We compared results to IUCN criteria to assign extinction risk. We focused on criterion A which measures the ‘observed, estimated, inferred, or suspected’ decline in the number of mature individuals over a ten year period or three generations, whichever is longer. Information needed to assess exact generation time as prescribed by the IUCN (2008b) is unavailable for H. comes. Therefore, we choose the default ten year time period because it is longer than three times the best estimate of maximum lifespan (Morgan, 2007) for this species. We were particularly concerned with criterion A2, in which the causes of decline may not have ceased, may not be understood or may not be reversible. In these contexts, a decline of >30% would see a species listed as Vulnerable, >50% as Endangered and >80% as Critically Endangered. Since we were unable to ascertain the number of ‘mature individuals’ over the entire history of the fishery, we used the tally of all fish. Limited to the years with data available for this calculation, our time periods were either slightly less or more than ten years and the final year of one time period became the starting point of the following time period.  Results Interview CPUE (1970s to 1994)  Fishers consistently reported declines in CPUE. All fishers said their CPUE in the past two years and past five years had declined. All changes in career CPUE were also  34  declines, ranging from 25-80% (mean 67%) with the maximum decline noted by a fisher who had been fishing since the 1970s. When asked specifically to recall numbers, fishers reported collecting a median CPUE of 150 seahorses fisher−1 night−1 at the outset of the market for seahorses in 1969. This had declined to ten seahorses fisher−1 night−1 by 1994, a 93% reduction. Max CPUE followed a similar trend falling from 200 to 50 seahorses fisher−1 night−1 (75% decline) over the same time period (Fig. 2.2). One fisher mentioned that he was motivated to start catching seahorses because two other fishers could catch more than 100 seahorses night-1 in the 1970s and early 1980s. Only five fishers reported having any zero seahorse catch nights.  Logbook CPUE (1996 to 2003)  From fisher logbooks we recorded a total of 31,381 H. comes caught over 11,179 nights of fishing from 1996 to 2003 with catches ranging from 0 – 80 seahorses night−1. We saw no clear trend in CPUE over the course of the logbook program (Fig. 2.2), although there was variability among years. This was true of all metrics, with or without zero catches, although absolute values varied according to the metric. Logbook data showed that catch night−1 was highly skewed, with no seahorses on nearly 70% of fishing nights (Fig. 2.3a). Medians were reduced as a result of the high number of low catch nights but there were also nights with large catches, in all years and for all fishers (Fig. 2.3b).  35  Entire fishery CPUE trend (1970s to 2003)  Overall reconstructed CPUE showed a fishery that experienced a severe historic decline with more recently low and variable CPUE. Our assumption about fisher recall bias, and therefore our choice of CPUE calculation, affected the inferred magnitude of decline in catch rate. Even the most optimistic calculation, using maxima, yielded a 77% decline in CPUE from 1970 to 2003. When we assumed that interviews overestimated historic CPUE we chose logbook calculations based on maxima or medians that excluded zero catch nights (Fig. 2.2). This choice revealed a fishery that was in decline from 1970 to 1993, but has been stabilizing at lower levels since 1996. When we assumed that interviews accurately portrayed historic CPUE we chose the logbook median CPUE calculation that included zero catch nights. This choice of CPUE calculation suggested that the fishery had collapsed rather comprehensively.  Evaluating extinction risk  Extinction risk classification depended on the assumed reliability of fisher recall and the time period of data used. Different assumptions about recall bias, calculated as a means of linking interview and logbook data, translated into a wide range of extinction risk classifications for the most recent decade (1990-2001: Table 2.2). For example, when we assumed that interviews overestimated historic CPUE and then calculated a metric that might make logbook comparisons valid (Logbook max) we concluded that CPUE was increasing  36  and therefore this species should be classified as Least Concern. On the other hand, if we assumed that interviews and logbooks were comparable without ‘correction’, and calculated a median to summarize their central tendencies (Interview median plus Logbook median), we concluded that CPUE had declined by 97% during the 1990s and that H. comes should be listed as Critically Endangered. When we compared historic with current extinction risk for H. comes, we found that for earlier decades extinction risk was consistently more worrying. Declines between 50 and 80% justified the classification of Endangered for both 1970-1980 and 1980-1990 (Table 2.2), regardless of how accurate we assumed interviews to be. In contrast, since 2002 H. comes has been listed only as Vulnerable on the IUCN Red List (2008a).  Discussion  Handumon lantern fisher interviews tell the story of a seahorse fishery that experienced severe historic decline. The severity of that decline (and therefore extinction risk) hinges on assumptions about the accuracy of fisher recall. Fisher interviews set an historic baseline which suggests that H. comes, and therefore other historically exploited species, may be more at risk than current Red Listing reflects. Fishers perceived historic declines in CPUE that were not apparent in more recent logbook trends. Fisher interviews provided the only record of historic catches and suggested that CPUE had declined 75 to 93% by 1994. Without fisher interviews we might have concluded that recent low CPUE from logbooks was also typical in the past. It is clear from  37  the consistent declines reported by all fishers and across triangulated questions that all fishers perceived a decline. Declining catch rates revealed by interviews are similar for other marine resources in Danajon Bank (Green et al., 2004) and other fishery declines in the Philippines (Stobutzki et al., 2006). Fishers’ concerns over dwindling seahorse CPUE catalyzed the establishment of a community-managed no fishing zone in the hopes of recovering seahorses and other target species (Pajaro et al., 1997). Certainly any management of seahorses should consider declines perceived by fishers to avoid the shifting baseline syndrome (Pauly, 1995). How perceived historic declines relate to recent logbook trends depended on assumptions we made about fisher recall bias. Nights on which fishers caught zero seahorses dominated logbook data (Fig. 3.3a), but were mostly absent from interviews, making us wonder why fishers did not report zero catch nights. It is possible that if seahorses were more abundant historically fishers caught them every night, but it is also likely that fisher reports were biased toward positive rare events (Bradburn et al., 1987; Tourangeau, 2000). Combined with the possibility that fishers may err toward pessimism (Ainsworth & Pitcher, 2005), fisher interviews likely tend to exaggerate the perceived decline. Although there are no standard methods for quantifying retrospective bias, our calculations show that if we assumed fishers overestimated historic CPUE, the decline to more recent CPUE in logbooks was less severe (Interview Max with Logbook Max: Fig. 2.2). On the other hand, if we assumed fisher historic interviews were accurate, the decline to present was more dramatic (Interview median with Logbook median: Fig. 2.2). In addition to recall bias, interviews and logbooks generated data with different characteristics, requiring careful consideration during analysis. Logbooks captured more variability than interviews, which influenced conservation assessments. Logbook data had  38  more samples per time than interview data: more nights in a year and more years in a row. It is then not surprising that logbooks captured a wider spectrum of events from common to rare, zero to many seahorses caught respectively. By extension, extinction risk calculated from logbooks (1990-2001) spanned a wider range than when calculated only from interview data (1970-1980 and 1980-1990 Table 2.2). Despite uncertainties about how best to link interviews and logbooks, our analysis suggests that our focal population of H. comes may be more at risk of extinction than its current listing of Vulnerable. Our uncertainty about the compatibility of interviews and logbooks placed our Red List assessment for the period 1990 - 2001 between Least Concern and Critically Endangered (Table 2.2). Least Concern is an unrealistic classification because it is based on a single max CPUE value per year, probably a rare value and possibly an outlier. Hippocampus comes should therefore be classified at least as Vulnerable, but possibly as at risk as Critically Endangered (CR). In situations where the spread of plausible values qualifies a taxon for two or more categories of threat, the IUCN recommends adopting a precautionary, but realistic, approach and choosing more threatened category (IUCN, 2008b). In our study, the more threatened choice of CR is based on data that assumes historic CPUE from interviews was not over estimated relative to logbooks. Clarification of this classification requires interviews that span the entire history of the fishery. The current listing of Vulnerable (IUCN, 2008a) is supported by our calculations that account for retrospective bias (Median interview and Median no zeros: Table 2.2), but may underestimate actual threat level, especially if the biggest declines occurred prior to 1990. For species like H. comes, that have undergone historic exploitation, our results suggest that relying on recent decades of data alone may underestimate extinction risk.  39  Ignoring historic data, the current IUCN Red List assessment is Vulnerable. In contrast, assessment of the earliest data qualifies this species as Endangered. While a similar pattern has been found for Gulf grouper (Mycteroperca jordani; Sáenz-Arroyo et al., 2005), the opposite pattern has been found for amphibians (Stuart et al., 2004). For marine species that have been impacted by fishing for at least hundreds of years (Jackson et al., 2001), assessment of contemporary conservation and fishery status must be conducted relative to historic baselines (Pauly, 1995). Our analysis highlights a potential limitation to the Red Listing process. Because the default time-frame for assessment is only a 10 year window, there is little capacity for incorporating long-term changes for relatively short-lived species. Capturing historic changes is especially important for species that have experienced historic declines. This type of retrospective threat classification is uncommon, but is becoming possible by incorporating fisher knowledge. We have focused on examining retrospective bias to improve methods for incorporating fisher knowledge into quantitative assessments, but other questions about the magnitude of decline have not yet been answered. It is perhaps surprising that despite severe perceived declines in CPUE, there are still seahorses to be caught. Certainly the flexible, multi-species nature of the fishery allows lantern fishers to continue taking seahorses whenever they see them (Vincent et al., 2007) and establishment of no fishing zones may allow for un-fished individuals to seed fishing grounds. In addition to retrospective bias, another way by which declines may be over-estimated is that some deep-water dwelling seahorses are less vulnerable to fishing because lantern fishers only fish in relatively shallow water. If deeper seahorses were not being fished, fishers would perceive their reachable stocks declining at a greater rate than the overall population. This bias applies to  40  scientifically gathered CPUE data as well and is known as hyperdepletion. We cannot yet assess this problem due to lack of data. Our study that links two historic fisher-derived datasets can inform management, prioritize future research, and provide insight into how we might tackle challenges of integrating fisher knowledge into quantitative assessments. Fishers perceive severe historic declines in seahorse catch rates, which means seahorses may be more endangered than we think because current Red listing doesn’t incorporate historic status. The success of future management and avoidance of the shifting baseline syndrome requires better understanding of fisher perceptions of the decline. Before our CPUE reconstruction can be used to evaluate effects of possible management options we will need to clarify the magnitude of interview bias (e.g. trying to confirm the magnitude of CPUE in the early 1970s that only two fishers could recall) and improve our understanding of how CPUE relates to H. comes abundance. Future studies that incorporate fisher knowledge into quantitative assessments require: 1) clearly stated assumptions about fisher knowledge bias; 2) clear criteria to compare fisher knowledge collected with different methods, and; 3) evaluation of the impact of assumptions on assessments.  41  Tables and figures  Table 2.1. Interview schedule Biographical information: name, age, number years fishing, number years fishing seahorses Q1. Has the number of seahorses you catch per night increased or decreased over the past two years? Over the past five years? Q2. On this timeline (where we have established the beginning, middle, and most recent year of your seahorse fishing career) please mark a number of X’s at each time period to represent relative change in CPUE. Q3. What was the usual number of seahorses you caught per night in: 1994, 1990, 1980s, 1970s?  Table 2.2. Extinction risk categorization for IUCN Red List criteria by time period and CPUE calculation: CR = Critically Endangered, EN = Endangered, VU = Vulnerable, LC = Least Concern CPUE Calculation  Time Period  Interview  1970-1980  1980-1990  Max  EN  EN  Median  EN  EN  Logbook  1990-2001  Max  Max  LC  Median  Median no zeros  VU*  Median  Median with zeros  CR  *VU indicates current Red List classification  42  Figure 2.1. Danajon Bank study region of the central Philippines and our focal lantern fishing community of Handumon  Reef Land  CEBU B  n j  a  n  k  o  a n a D  Handumon  BOHOL  43  Figure 2.2. All CPUE trend calculations for interview data (maximum: ▲, median: ■) and logbook data (maximum: ∆, median no zeros: □, median: ◊). Arrow indicates break between interview (1970 – 1994) and log book (1996 – 2003) datasets. All calculations are in units of seahorses fisher-1 night-1. Error bars are 25th and 75th percentiles for median values calculated for all medians, but large enough only to be visible on the graph for interview medians. 200  CPUE  150  100  50  0 1965  1970  1975  1980  1985  1990  1995  2000  2005  Figure 2.3. Frequency (a) and distribution (b) of catch night-1 logbook data. Box plot shows 25th, 75th percentile, and median; whiskers are 90th percentiles; hollow circles are outliers.  a)  b)  100  0.8  8000  0.7  7000  0.6  6000  0.5  5000  0.4  4000  0.3  3000 2000  0.2  1000  0.1  0  0.0 10 20 30 40 50 60 70 80 90 Catch night-1 TOTALSH_HAND  0  80  -1 Catchper night Catch night  9000  Proportion per Bar  Count  10000  60  40  20  0  916 19  927 19  938 19  949 19  050 20  061 20  072 20  083 20  44  Chapter 3: Comparing fisher interviews, logbooks, and catch landings estimates of extraction rates in a small-scale fishery  45  Introduction  Rapid, low-cost, and reliable assessment methods are needed to detect trends in the abundance and use of wildlife. Data needed for managing wildlife exploitation (e.g. population trends, hunting patterns) have traditionally been gathered using formal scientific methods (Mauro & Hardison, 2000). However in developing countries with high biodiversity, heavy reliance on wildlife resources, and rapid environmental change, these time-consuming and expensive methods (Anadón et al., 2009; Rist et al., 2010) are often impractical (Johannes, 1998; Danielsen et al., 2005). With the number of threatened species on the rise and the vast majority (>95%) of described species yet to be assessed (Vié et al., 2009), traditional scientific methods alone are unlikely to be able to inform species assessments – a critical first step toward halting declines. This issue is particularly concerning in marine systems, where our understanding of historic conditions is poor (Jackson et al., 2001). To understand changes in wildlife abundance and exploitation, researchers have begun to turn to alternative sources of data. Trade records (Baum & Vincent, 2005; Scales, Balmford & Manica, 2007), historic photographs (McClenachan, 2009), explorers’ records (Sáenz-Arroyo et al., 2006; McClenachan & Cooper, 2008), and unpublished government reports or grey literature (Sadovy & Cheung, 2003; Dulvy & Polunin, 2004) have all shed light on the extent of depletion in marine systems. Documenting the knowledge of resource users (variously known as local ecological knowledge (LEK), traditional knowledge, etc.), in particular, can be an inexpensive (Moore et al., 2010; Rist et al., 2010) and effective way of gathering data on species, catches, ecology, and habitat (Johannes, 1981; Neis et al., 1999b).  46  Interviews with resource users (e.g. Sáenz-Arroyo et al., 2005) and analyses of their daily written logs (Baum et al., 2003; Rosenberg et al., 2005) have uncovered population declines that were previously undetected by western science (e.g. Gilchrist et al., 2005). Despite its practicality, there are several reasons why resource user knowledge is not more widely incorporated into conservation and resource management. Although the use of LEK is increasing, only a small fraction (0.42%) of papers in 12 prominent ecological and conservation journals incorporated LEK between 1980 and 2004 (Brook & McLachlan, 2008). Perhaps the most significant impediment is the difficulty in appraising the reliability of LEK in terms of accuracy and precision. Information collected from resource users is subject to various and often unknown biases (reviewed by Daw, 2008a). Perceptions can be distorted through complex cognitive processes or vested interests (Bradburn et al., 1987) and little is known about the quality of the information reported regarding events in the distant past. The reliability of interview data as indicators of trends is therefore unclear (Jones et al., 2008). LEK assessments are also inter-disciplinary in nature and therefore either require appropriate training or collaborative work (Huntington, 2000). Furthermore, rigorous testing of LEK is necessary prior to its incorporation into management plans (Gilchrist et al., 2005; Anadón et al., 2009). Studies that have assessed the consistency of information from resource users (e.g. interview or logbook data) and scientific approaches (e.g. landings records) have produced variable findings. In the marine realm, fisher testimony of trends in the abundance of corvinas (Sciaenidae: Cynoscion spp.) in the upper Gulf of California was in agreement with 50 years of recorded catch data (Lozano-Montes et al., 2008). In contrast, northern British Columbia fishers reported steeper declines in multi-species abundance than were found in  47  stock assessments (Ainsworth & Pitcher, 2005). Compared to recorded landings, fishers may over- (Lunn & Dearden, 2006), under- (Rocha et al., 2004) or accurately estimate (Neis et al., 1999b) present catch or catch per unit of effort (CPUE) during interviews. Data in user logbooks led to inferences of lower catches (Walsh, Kleiber & McCracken, 2002) and higher effort (Rist et al., 2010) than did scientific data collection. There is, therefore, a continuing need to understand the relationships between user-generated knowledge (Gavin & Anderson, 2005; Gilchrist et al., 2005; White et al., 2005) and other datasets (Gagnon & Berteaux, 2009). Our objective was to evaluate the degree of consensus among three methods (fisher interviews, fisher logbooks, and “scientific” catch landings) of quantifying catch rates (catch per unit of effort or CPUE) from the same years for a small-scale, data-poor seahorse fishery in the Philippines. While other studies have evaluated a single resource-user based data collection method, we compare two data-collection methods that are based on concurrent resource user knowledge (interviews and logbooks) with scientific methods (catch landings) which is uncommon in the marine realm. Any differences or similarities among these modes of data collection will be important to consider when incorporating fisher knowledge into quantitative assessments (Neis et al., 1999b; Huntington et al., 2004; Gilchrist et al., 2005) and when selecting practical data collection methods for conservation and fishery assessment. Seahorses are globally exploited, but their fisheries remain data-poor. Almost 80 countries trade seahorses for use in traditional medicines, ornamental display, or curios (Vincent, 1996) yet we lack data to assess quantitatively the conservation status of 28 of the 36 seahorse species (IUCN, 2008a). Furthermore, traditional underwater survey methods are  48  time consuming and expensive for these small, cryptic, sparsely distributed fish (Foster & Vincent, 2004). There is, therefore, a pressing need for inexpensive rapid assessment methods that can detect trends in seahorse population size and exploitation rates to facilitate the management of these species.  Methods Study area, fishery, and species  In the Philippines, an historically important exporter of seahorses (Vincent, 1996), the authors and their Project Seahorse team have worked with fishing communities since 1995 to assess and manage seahorse fishing. Our work is focused on the Danajon Bank in the Central Philippines (Fig. 3.1) - a double-reef system that has been has been largely degraded by destructive fishing practices. Coastal communities in the region are among the poorest in the Philippines (Green et al., 2003). One of our focal communities, Handumon (municipality of Getafe, Bohol province), has the largest concentration of lantern fishers who catch the most seahorses in the region (Meeuwig et al., 2003). Another focal community, the neighboring village of Jandayan Norte, has only a handful of elder lantern fishers who fish the same waters as Handumon fishers. Seahorses are caught with many other species in a small-scale fishery that involves fishers free diving at night by the light of a lantern slung low over the bow of their outrigger boat. This open-access fishery has not been formally monitored or managed but fishers and traders have reported declines in seahorse CPUE (Vincent, 1996; Pajaro et al., 1997);  49  coupled with reports of increasing demand, price, and effort, all available evidence suggests overexploitation (Martin-Smith et al., 2004). Recent analysis of the fishery suggests current CPUE is extremely low and seasonal (Vincent et al., 2007). Furthermore, fishers perceive historic declines in fishing rate that cannot be detected in recent (1996 to 2004) data, although the extent of the historic decline (and therefore extinction risk) hinges on the assumptions we make about the accuracy of fisher memory (O'Donnell et al., 2010). The seahorse targeted by lantern fishers is the tiger-tail seahorse Hippocampus comes Cantor 1850. This species comprises the majority of seahorses caught in the lantern fishery and in the Danajon Bank region overall (Vincent et al., 2007). Although fishing seahorses has been illegal in the Philippines since 2004 - as a national reaction to international trade controls - fishers and buyers continue to trade openly. Project Seahorse has continued to obtain access to seahorse fishers by working through long-serving Filipino community organizers (who have engendered trust among villagers) and by maintaining open dialogue with fishing communities, emphasizing an interest in resource use rather than law enforcement. We also collected and retained our data in a way that coded for individuals, allowing them anonymity. As a result, most fishers and buyers spoke openly about seahorse catch and some volunteered to record catch in logbooks or have landings recorded. This allowed us to obtain and compare interview, logbook and landings data over the same time frame. We anchored our comparison in landings data, as representative of the scientific approach – whereas interview and logbook approaches both derive from local understanding – but this should not be taken to imply a greater confidence in that dataset.  50  Data collection Interviews  We conducted structured interviews with fishers from March to June 2008 in Handumon and Jandayan Norte (Fig. 3.1). Since there is no official register of fishers, we consulted village leaders for initial recommendations of individuals who might be willing to be interviewed and thereafter found respondents by word-of-mouth (Snowball sampling; Berg, 2001). Our goal was to gather information from all decades of the fishery. We achieved this by interviewing lantern fishers from neighboring communities that tend to fish the same grounds whose ages (24 – 66) and seahorse fishing experience (4 – 36 years), covered the range of all lantern fishers. Ultimately we interviewed 24 fishers from Handumon and 1 from Jandayan Norte (~50 and 30 % of the 2008 fleets, respectively). Each interview began by constructing a detailed visual timeline that mapped out the fisher’s personal history (Appendix A). This career history method (after Neis et al., 1999b) guided interviews and increased the accuracy of recalled dates (Means & Loftus, 1991). The ensuing visualization allowed us to link fishers’ important life events (e.g. first year fishing, marriage, birth of a child) to external events of known date (e.g. change of presidents, large typhoons). Subsequent questions asking fishers to recall CPUE were dated relative to the personal events mapped on the visualisation. We asked fishers to recall a range of CPUE values during time periods when they were most likely to have clear memories. Memories of rare, especially positive, events and those that are more recent tend to be more readily recalled (Bradburn et al., 1987; Tourangeau, 2000). We therefore asked individual lantern fishers about their catch on a  51  good, bad, and typical night of fishing during three time periods over their career: 1) the highest catch year, 2) the most recent year fishing, and 3) either a year early in their career or another year for which they had a clear memory of catch. Interviews were conducted in the local language using trained research assistants, and followed the interview schedule (Table 3.1). Fishers rarely reported zero seahorses caught per night, leaving ambiguity about whether they sought seahorses without success or never sought them at all.  Logbooks  From 1996 to 2003, Handumon fishers volunteered to record catch in logbooks. Logbooks were distributed by Filipino Project Seahorse team members who followed up with discussions and log collection at the end of each month. Of all the data recorded (refer to Vincent et al., 2007 for full details) we focused on fishers’ record of their nightly catch of seahorses. Logbooks were completed by 43 fishers between 1996 and 2003 with a mean participation rate in the logbook program of 16 (±5 S.D.) fishers year−1. This level of participation represented about 50% of the fishers in Handumon in a given year, with participation varying by month. Fishers seldom recorded whether they were trying to catch seahorses on a particular fishing expedition. Thus, nights or months with zero seahorse catch may have arisen either from an inability to find seahorses or from a decision to target other species. As logbooks were dominated by zero catch nights (Fig. 3.2) we also analysed a variant where we removed zeroes to make them more comparable to interviews and catch landings.  52  Catch landings  From 1996 to 1998 and 2002 to 2007 Project Seahorse collected biological data on seahorses caught by 42 lantern fishers in Handumon and five in Jandayan Norte with a mean participation rate of 29 (±5 S.D.) and 3 (±1 S.D.) fishers year−1, respectively. This level of participation represented about 80% of the fishers in Handumon and 60% in Jandayan Norte in a given year, with participation varying by month. Fishers are usually widely dispersed across their village and fishing grounds but predictably visit few buyers to sell their catch. Thus, to maximize sample sizes, data were collected at buyers’ houses by a trained local assistant. For each seahorse they recorded the fisher and the species, size, sex, reproductive status (sensu Vincent et al., 2007), number and location of capture. Since fishers only visited buyers when they had caught seahorses, we were unable to collect information on (or distinguish between) days when fishers did not fish or when they fished but caught no seahorses.  Analysis  We compared and contrasted the three sources of CPUE (seahorses-1fisher-1night-1) trends and overall absolute values: interviews, logbooks and landings. Because of the positively skewed nature of the datasets and presence of outliers (Fig. 3.2), we used the median rather than the mean CPUE. All analyses were conducted using R (R Core Development Team, 2009).  53  Trends in CPUE  We evaluated the change in CPUE from 1996 to 2004 using linear mixed-effects models separately for four datasets: 1) interviews, 2) logbooks, 3) logbooks without zeros, and 4) landings. To compare like with like, we contrasted typical CPUE reported in interviews with median CPUE from catch landings, logbooks and logbooks excluding zeroes specific to each year. Only 8 of the 25 fishers interviewed (representing ~20% of the 1996 to 2004 fishing fleet) provided data that fell between 1996 and 2004 – the range of years for which logbook and landings data were available; the ages (26 - 66) and years fishing experience (10 – 36) of these 8 fishers were comparable to our larger pool of respondents. To obtain fisher-specific medians from logbooks, logbooks without zeroes, and catch landings, we calculated median CPUE for each fisher within each year (henceforth, ‘median CPUE’). For logbooks and landings, the number of nights fished by a fisher in a given year varied widely among fishers (from 1 to 250). Therefore, we also weighted each fisher’s median CPUE year-1 by the number of nights fished in that year such that median CPUE values based on more nights fished were given more weight. We used top-down model minimization (Diggle et al., 2002) to choose the minimumadequate model for each dataset. We could not develop one all-encompassing model with method of data collection as a random component since the datasets differed in structure. We began with a linear mixed-effects model with CPUE as our response variable, year as a fixed effect and fisher as a random effect on slope and intercept (executed in R with the nlme package; Pinheiro et al., 2009). Random effects of fisher accounted for non-independence of CPUE data across years, and accounted for differences among fishers in their overall CPUE  54  (intercept) and in their change in catch through time (slope). Since a change can only be estimated with a minimum of two reports of CPUE, we excluded fishers for whom we had only a single year of data in logbooks and/or landings records. We then dropped random effects based on the results of likelihood ratio test (REML) that compared the fit of models with and without each term. Prior to analyses, all data were ln(x+1)-transformed to fulfill model assumptions, which were verified visually following Zuur et al. (2009). The optimal model varied by data-collection method (Table 3.2).  Absolute value of CPUE  Since slopes of trend in CPUE were not different among data sources (see results), we compared overall CPUE estimated using each data-collection method, including fisher interview estimates of good and bad CPUE in addition to the median CPUE from previous analyses. Here we compared CPUE between 1996 and 2004 for six datasets: 1) interview – good nights, 2) interview – bad nights, 3) interview – typical nights, 4) logbooks - median, 5) logbooks without zeros - median, and 6) landings – median. To compare the central tendencies of the data collection methods directly we tested the hypothesis that CPUE differed significantly among the six data-collection methods. Because no trend across years was detected (see results for median CPUE; interview good nights p = 0.31; interview bad nights p = 0.39) we pooled CPUE values for all fishers and all years within a data-collection method. We then conducted a repeated measures ANOVA (RM-ANOVA, Crawley, 2007) assuming a quasi-Poisson distribution. This test was  55  executed as a generalized linear mixed model with the lme4 package in R (Bates & Maechler, 2009). The response variable was CPUE, the categorical fixed explanatory variable was data collection method, and fisher was a random factor nested within method. Pairwise comparisons were made between landings and each of the other data collection methods. Since we were unable to find a follow-up multiple comparison test, we evaluated differences among model estimated means by using the ± 2 x SE criteria described by Payton et al. (2003). We considered non-overlapping confidence intervals to be significantly different.  Correlation – interview and catch landings  We plotted interview responses against recorded landings to understand the relationship between what a fisher reported catching and what someone else recorded him catching. Data consisted of a point for each of 14 fishers who both responded to an interview and had his catch landings recorded in the same year, for any year between 1996 and 2007 (longer than in other analyses). For interviews we used responses to a good, bad, and typical CPUE which we matched with the maximum, minimum, and median landings value for a particular fisher and year. Since data were not normally distributed, we used a Spearman’s rank correlation to test if the ranks of interviews co-varied with the ranks of landings.  56  Results Data structure  Median CPUE data provided from interviews, catch landings, and logbooks differed in terms of distribution (Fig. 3.2). While zeros were absent from interviews and catch landings, logbooks were dominated (66%) by nights with no catch. When zeros were removed from logbooks, data remained positively-skewed. The same was also true of landings data. Interview data were almost uniformly distributed between CPUE of 5 and 50.  Trends in CPUE  All metrics indicated no change in median CPUE with no slope parameters significantly different from zero (p>0.05, Table 3.2, Fig. 3.3). Quantile-quantile and plots of the standardized residuals vs. fitted values indicated that residuals and random effects were not normally distributed for logbook and logbook without zeros datasets, likely because model predicted values tended to over-estimate low observed values and under-estimate high observed values. Model residuals and random effects for catch landings were normally distributed, as were model residuals for interviews. We found no heteroscedasticity in residuals across years suggesting that models are equally good (or bad) at predicting CPUE across years.  57  Absolute value of CPUE  Interviews tended to give higher and more variable CPUE values than the other concurrent data-collection methods (Fig. 3.4a), but fisher reports of a bad night’s catch approximated median CPUE from other data collection methods. Fisher reports of a good night’s catch were significantly higher than other data collection methods (Table 3.3). Typical CPUE from interviews was on average ~ 16 times greater than median CPUE from logbooks and double the estimate from landings. Typical CPUE from interviews did not differ significantly from logbooks without zeros (Fig. 3.4b, Table 3.3). Interviews aside, CPUE from landings was only comparable to logbook estimates when zero-catch nights were not considered. Median CPUE recorded in logbooks was significantly lower than recorded in landings (Table 3.3), but when we removed the zeros from logbook data this difference disappeared. The assumption of homogeneity of variance was violated for this test because fishers provided CPUE values that varied more widely when responding to interviews (Levene’s test: F=39, d.f.=3, p<0.001) than were recorded in logbooks or landings.  Correlation – interview and catch landings  There was no consistent relationship between interview reports and landings. The tightest relationship, though not significant, was between what fishers reported as a good night’s catch and their maximum landings CPUE (r = -0.48, p = 0.09, n = 14 fishers). There was no correlation between fisher reports of bad night’s catch and recorded minimum CPUE  58  (r = -0.074, p = 0.80, n = 14 fishers) or a typical night’s catch and recorded median CPUE (r = -0.048, p = 0.87, n = 14 fishers). Generally minimum and median landings CPUE were roughly 5 seahorses night-1. In comparison, fisher reports of bad CPUE in the same year ranged from zero to 20 seahorses night-1 while typical CPUE ranged from 11 to over 40 seahorses night-1.  Discussion  Whether lantern fisher knowledge was comparable to other available CPUE data depended on which method (interview or logbook) was used to elicit knowledge and the variable (trends or absolute values) measured. While interviews and logbooks provided estimates of CPUE trends consistent with landings, interviews yielded higher and more variable estimates that were not correlated with landings. In spite of these discrepancies the lantern fishery, like other data-poor and depleted fisheries, is in need of recommendations for how to manage exploitation. We suggest two possible criteria to choose which methods might produce data appropriate for setting recovery or fishery targets. No data collection method identified a change in CPUE between 1996 and 2004, a finding that is supported by fishery independent data. Despite biases associated with recalling past events (reviewed by Daw, 2008a) and the illegality of the fishery, fisher interviews produced results not so different from other methods. Underwater visual census (UVC) surveys of the tiger tail seahorse conducted from 2001 to 2008 also found stable densities both inside Marine Protected Areas (MPAs) and in fishing grounds (Yasue et al. in  59  prep). These results suggest that interviews or logbooks may be a reasonable proxy for more costly landings or UVC surveys when seeking information about trends. We cannot rule out, however, that trends went undetected by all methods. Increasing the ‘grain’ of effort from catch per night to weeks or months could decrease variability in CPUE (Anadón et al., 2009), perhaps increasing the power of all data-collection methods to detect trends. The generally higher CPUE reported in interviews contradicted our expectation that these data would agree with logbook records, also collected by fishers. The discrepancy seemed to arise from the nights with zero seahorse catch, which were rare in interview recall but dominated logbooks; bad or typical CPUE reported in interviews produced the same results as data from logbooks without zeros. Fishers may more readily recall rare and/or positive events (e.g. a night when many seahorses were caught) than common occurrences (e.g. nights of zero seahorse catch) (Bradburn et al., 1987). Logbooks, in comparison, may be less biased if CPUE is recorded when it occurs, minimizing the reliance on memory. The fact that landings most closely approximated fisher reports of a bad nights’ catch supports an inference that interviews tend to over-estimate CPUE. Interview reports of higher CPUE than other datasets identify may represent an overreporting by the fisher or an under-recording in logbooks or landings data. There are three types of explanation for any over-reporting. First, seasonal variation in seahorse CPUE (Vincent et al., 2007) may make it challenging for lantern fishers to estimate a typical value for catch night-1 (Jones et al., 2008). Second, fishers may report peak-season CPUE as typical, using imperfect recollection (Tourangeau, 2000). Third, fishers may intentionally over-report catches to impress an interviewer, to comply with perceived interview attitudes,  60  or perceived opportunities to attract outside investment to their communities (reviewed by Yasué et al., 2010). The other explanation for the discrepancy between interview and landings data may lie in a possible under-estimate of both effort and catch. First, by focusing our sampling at buying stations where fishers congregate – the dispersed fishing activities of small-scale fishers are difficult to track – our landings data excluded nights when no seahorses were caught. Second, lantern fishers sometimes sold fish at sea before returning to shore (as in Young et al., 2006) gave fish as gifts before selling the bulk of the catch to the buyer or, on occasion, kept fish for personal use; in all cases, they would escape detection in landings data. The discrepancy between what fishers reported (interviews) and what they were documented as catching (landing records) has important consequences for conservation and fisheries analysis. Although other studies have found strong correlations between interview and formally collected data (Anadón et al., 2009; Rist et al., 2010), the overall correlation depends on the consistency of individual relationships between interview recall and other datasets; if all individual relationships have similar slopes, then the overall correlation should be obvious whereas a myriad different slopes would obscure any correlation. The latter could arise if, for example, fishers differed in their abilities to recall high catches, low catches or both. Therefore, even if individual fisher’s interview responses were correlated with official records, it may not be possible to quantify how fisher reports relate to scientifically collected data at a scale (the fishery) relevant to management/conservation. Despite discrepancy among data-collection methods, managers who need to know current CPUE to set recovery / fishery targets could choose based either on corroboration  61  among methods or precaution. In the lantern fishery, a bad night’s catch as reported in interviews, landings, and logbooks stripped of zeros all provided estimates of CPUE in the range of 1 to 10 seahorses fisher-1 night-1. The disadvantage of basing a decision on consensus, in this case, is that none of these methods accounted for fishing effort accurately; nights when no seahorses were caught were not incorporated. They therefore overestimate CPUE, potentially leading to higher catch limits or less area placed in MPAs if management is deemed necessary. A more precautionary manager would base management on logbook data because as the only method to account for all fishing effort, including nights when no seahorses were caught, they produced the lowest CPUE values (just 0 to 1 seahorses fisher-1 night-1). This choice would lead to lower catch limits or more area placed in MPAs than the consensus method. For researchers seeking rapid assessment methods for trends in the abundance and use of wildlife, our study provides two suggestions. First, if interview data are used to set recovery targets or catch limits, which are sensitive to the magnitude of reported values, it will be important to ask about extremes (e.g. good and bad catches) in addition to typical or average values. Second, due to the unpredictable relationship between resource user knowledge and other datasets, collecting and then comparing multiple datasets will produce the most robust information. Even if data elicited from resource users can not directly inform quantitative assessments, they will continue to be important to understand resource user perceptions to effectively engage communities in management and conservation.  62  Tables and Figures  Table 3.1. Interview schedule. Biographical information: name, age, number of years seahorse fishing Personal timeline questions (year): began fishing, married, children's births, purchased boats or engines CPUE questions: During this past year… How many seahorses could you catch on a good night? A typical night? A bad night? When was the time you remember catching the most seahorses in one night? How many did you catch on a good night back then? During that same time, what was a typical catch/night? A bad catch/night? At the beginning of your fishing career (or another year for which the respondent had a clear memory)… How many could you catch on a good night? A typical night? A bad night?  Table 3.2. Sample sizes (n: total fishers) and best linear unbiased predictor model fit parameters to median CPUE date for 1996 to 2004. Method Interview Logbook Logbook without zeros Catch Landings  Model Linear Linear mixed effects, random slope and intercept Linear mixed effects, random slope Linear mixed effects, random slope and intercept  n 8 43 39 42  Slope value p 0.062 0.64 0.007 0.91 -0.005 0.76 -0.018 0.23  Intercept value p 2.78 0.002 0.61 0.004 1.73 <0.001 1.66 <0.001  Table 3.3. Sample mean (x) in seahorses fisher-1 night-1, Ln + 1 and confidence interval (2 x SE) values for RM-ANOVA analysis by data-collection method.  Method Interview good nights Interview bad nights Interview typical nights Logbook without zeros Catch Landings Logbook  x 3.983 1.937 3.018 1.647 1.411 0.191  2 x SE 1.089 1.228 1.122 0.665 0.415 0.735  63  Figure 3.1. Danajon Bank study region of the central Philippines and our focal lantern fishing communities of Handumon and Jandayan Norte.  Reef Land  CEBU B  n j  a  n  k  o  a n a D  Handumon Jandayan Norte  BOHOL  64  Figure 3.2. Frequency of median CPUE (seahorses fisher-1 night-1) for interviews, logbooks, logbooks without zeros, and catch landings 1996 to 2004.  Logbook  80 100 60 20  40  Frequency  1.0  0  0.0  Frequency  0.5  Frequency  1.5  2.0  Interview  0  10  20  30  40  50  0  10  Median CPUE  30  40  50  40  50  Median CPUE  Landings  40 0  0  20  Frequency  60  10 20 30 40 50 60 70  Logbook no zeros zeros Logbook without  Frequency  20  0  10  20  30  Median CPUE  40  50  0  Median CPUE  10  20  30  Median CPUE  65  Figure 3.3. Median CPUE (seahorses fisher-1 night-1 , Ln + 1) used to assess trends from 1996 to 2004 for interview, catch landings, logbooks and logbooks without zeros data-collection methods. Each point represents an individual fisher.  3 2 1 0  0 0  2  4  6  8  0  2  4  6  8  3 2 1 0  1  2  3  4  Catch landings  4  Logbook without zeros  0  Median CPUE (Ln+1)  1  2  3  4  Logbook  4  Interview  96 19  98 19  00 20  02 20  96 19  04 20  98 19  00 20  02 20  04 20  Year  66  200  Figure 3.4. Comparison of pooled 1996 to 2004 CPUE (seahorses fisher-1 night-1) values (a) and RMANOVA estimate of CPUE (seahorses fisher-1 night-1, Ln + 1) ± 2 SE (b) for fisher estimates of a good, bad, and normal night’s catch from interviews and median CPUE from logbooks without zeros, catch landings, and logbooks.  60 0  20  40  CPUE  80  100  a)  b)  6  MAX.INT  MIN.INT  MED.INT  LBNO  CL  LB  5  a, b 4  a, b, c, d 3  b, c c  2  d 1  0  Good  Bad  Logbook w ithout zeros  Catch Landings  Logbook  w wi Lo th gb ou o t z ok er C os at ch la nd in gs Lo gb oo k  Int.med  In te rv ie  w  Int.min  In te rv ie  rv ie  w  Int. max  In te  M edian CPUE estimate  Median CPUE estimate  a  Typical / Median  67  Chapter 4: High fishing mortality rate in a small-scale fishery: digital tracking of Filipino lantern fishing.  68  Introduction  While international attention to global fishery problems has focused on industrial fisheries (e.g. Pauly et al., 2002; Myers & Worm, 2003; Castilla & Defeo, 2005), small-scale fisheries are increasingly suggested as a sustainable alternative, though they are poorly understood (Munro, 1996; Orensanz et al., 2005). The distinction between these two types of fisheries tends to be by boat size: industrial referring to large, open-ocean vessels and smallscale referring to small, coastal boats (Parma et al., 2003). Small-scale fisheries are praised because they catch roughly the same amount of fish while employing more people and consuming less fuel than industrial fisheries (Pauly, 2006; Jacquet et al., 2009). While these socio-economic benefits may be important for successful fisheries policy (Pauly, 2006), sustainability also requires that the biological productive capacity of the resource be maintained in the long-term. However, common indicators of sustainability, like trends in abundance (or their proxy, catch rates) or exploitation rates (the proportion of population removed by fishing annually) (Rochet & Trenkel, 2003), tend to be unavailable for smallscale fisheries. The assumption that small-scale fisheries can be more sustainable may be due to their seemingly benign fishing practices when compared to industrial fisheries. While only one million of the world’s 51 million fishers are considered industrial fishers (Berkes et al., 2001), they steer large boats capable of catching tonnes of fish at once using machines to operate nets large enough to encircle a dozen jumbo jets (Safina, 1995). Industrial fisheries also tend to be equipped with ever-improving technology, helping them locate fish and remain at sea for weeks or months at a time. On the other hand, despite having small boats  69  and limited technology small-scale fishers have power in numbers (50 million globally) which may allow them to mount strong pressure on marine resources. Perhaps the most important difference between industrial and small-scale fisheries is in the spatial extent of their reach. While industrial fisheries exploit from the coast far offshore to the deep seas, leaving few corners of the world’s oceans untouched (Morato et al., 2006; Swartz et al., 2010), small-scale fishers’ equipment limits them to shallower coastal areas. In fact, the mounting body of evidence suggests that small-scale fisheries have caused species declines and even extinction. For centuries, early artisanal and later colonial exploitation has caused declines in targeted species (Jackson et al., 2001; Pinnegar & Engelhard, 2008) and extinctions of marine mammals (e.g., Caribbean monk seal Monachus tropicalis; McClenachan & Cooper, 2008). Contemporary small-scale fishing has also caused target species declines (Sadovy et al., 2003; Dulvy & Polunin, 2004; Godoy et al., 2010) with cascading ecosystem effects (Ruttenberg, 2001; Campbell & Pardede, 2006), even at low levels of fishing pressure (Jennings & Polunin, 1996). In addition, one study that directly compared industrial and small-scale fishing gears (Bundy & Pauly, 2001) concluded that small-scale fishing had larger and more wide-ranging ecosystem effects than the industrial sector. Many of these examples are from coral reef fisheries, which may well be unsustainable on a global scale (e.g. Newton et al., 2007). To improve management of industrial and small-scale fishing alike, there is a call to understand the spatial patterns of fishing (Hutchings & Myers, 1994; Booth, 2000; Wilen et al., 2002). Fish population declines can be masked by changes in spatial distribution of fish and/or fishing (Gillis et al., 1993). Such misunderstandings may help explain some of the most severe fishery collapses (e.g. Atlantic cod, Peruvian anchovy; Hilborn & Walters,  70  1992). In fact, it has been suggested that the only sustainable fisheries are those unable to access a major portion of an exploited species’ distribution (Pauly et al., 2005). Our poor understanding of the spatial dynamics of small-scale fishing may be improved with the use of digital technology. The few studies conducted have gathered coarse spatial fishing information by asking fishers to identify their activities on maps (e.g. Daw, 2008b) or recording coordinates of fishing grounds (Bizzarro et al., 2009). Digital tracking technologies, more commonly applied to monitor animal movements or industrial fisheries (GPS tracking Bertrand et al., 2007; e.g. digital tags Goldbogen et al., 2008), could provide complementary detailed spatial information. This detailed spatial information could be used to calculate exploitation rate. Usually calculated as the proportion of a population caught (population size being especially illusive in data-poor small-scale fisheries; Orensanz et al., 2005), exploitation rate could also be calculated spatially from fishing mortality. Fishing mortality can be estimated as the proportion of all area over which species are distributed that is swept by fishing effort during a year (i.e. the ‘swept area’ method) (Walters & Martell, 2004). In this study we quantified the spatial dynamics and estimated fishing mortality rate for a small-scale, free-dive fishery in the Philippines, by combining Geographic Information Systems (GIS), time-depth recorders (TDR; previously used to study marine mammals; Goldbogen et al., 2008), satellite images of habitat, and ethnographic field methods. The lantern fishery we studied operates at night: a fisher holds onto an outrigger boat as he (only men are involved thus far) swims, dives, spears, or hand-collects fish and invertebrates by the light of a lantern slung low over the bow. Quantifying where fishing occurs allowed us to 1) visualize how lantern fishers exploit the marine environment, 2) calculate fishing mortality  71  rate, and 3) gain new insight into the sustainability of this small-scale fishery and the appropriateness of spatial management.  Methods Study site, fishery, and species  Our work is focused on the Danajon Bank in the Central Philippines, a degraded double-reef system (Fig. 4.1a). Hugging the northern and western shores of the province of Bohol, the inner and outer reefs are shallow (approximately <10 m) and composed of scattered and patchy coral reefs interspersed with Sargassum and seagrass (Meeuwig et al., 2003) habitats. The reefs are in poor condition because of over-fishing and destructive activities such as blast fishing (Marcus et al., 2007). One of our focal communities, Handumon (municipality of Getafe, Bohol province), has the largest concentration of lantern fishers in the region (Meeuwig et al., 2003). In the other focal community of Batasan (municipality of Tubigon), fishers used to breathe compressed air through a tube while diving deep to spear fish, but turned to lantern fishing when compressor diving was banned. At night, lantern fishers sling a shaded lantern on the front of an outrigger canoe to illuminate the benthos below (Fig. 4.1b, c). The kerosene lantern sheds a circle of light about 6 m in diameter (O’Donnell, personal observation). They walk on reefs, swim, and maneuver their boats through mazes of the ropes and poles of seaweed farms. They use a rubber-band-propelled spear to pin fish to the ocean floor or move seaweed aside to look for fish or collect seahorses or invertebrates by hand. With no modern navigational aides (they  72  do not use maps/charts), wearing wooden goggles (Fig 4.1c) and plywood fins, lantern fishers catch at least 20 families of reef fish (including seahorses - the focus of our longrunning community-based research and management programme (Martin-Smith et al., 2004; O'Donnell et al., 2010) and many species of invertebrates.  Spatial fishing behaviour - data collection and analysis  To quantify detailed fishing behaviour we recorded fishing trips with GPS and participant observation. From April through June of 2008, 12 fishers (7 in Handumon and 5 in Batasan) each volunteered to have one fishing trip recorded with GPS and temperaturedepth-recorder (TDR). A Garmin GPSMap 76 CSx, accurate to <10 m, recorded location (latitude and longitude) and time every 30 seconds. Participants ranged from 24 to 42 years of age, covering all but the oldest fishers. We did not sample the oldest fishers because they paddle to their fishing grounds and having us accompany them would be too much of a burden. We accompanied and observed fishers on half of the trips that were recorded and for consistency the other six trips recorded were also for boats with engines. In addition we observed, but did not GPS, 10 other trips between 2006 and 2008. To understand the depth profile of fishing, we used two methods. First we interviewed 13 Handumon expert lantern fishers about their diving limits, typical depths they go to while fishing, and constraints on dive depth (interview schedule in Appendix B). Second, we recorded fishing depths in real-time by asking fishers to wear an egg-sized (2.5 x  73  3.3 x 4.5 cm) TDR on a belt around their waist for an entire night’s fishing. We used the Sensus Ultra by Reefnet Inc., accurate to ± 0.33 m, which recorded data every second. We distinguished between travel and fishing behaviours by combining the GPS track with evidence of swimming from the TDR and notes from observed fishing trips.  We  defined travel as time periods when the fisher was out of the water (TDR read zero m) and the GPS recorded horizontal movement at a speed consistent with motoring or paddling. Any breaks in travel not accompanied by TDR evidence that fishers were in the water were considered travel. Fishing bouts were defined as time periods when the fisher was in the water (TDR read >0 m). We calculated the duration and distance travelled for each behaviour. We defined a dive as an excursion below the surface of the water of more than one meter; because the TDRs were attached at the waist, any depth greater than one meter would result in the fisher’s head being underwater. Typical dive statistics (Table 4.1) were obtained for foraging fishers, including dive duration and depth. We report means ± 1 S.D.  Fishing mortality rate - analysis  To estimate Danajon Bank-wide lantern fishing mortality rate we combined distance travelled while fishing from GPS tracks with published estimates of fishing effort and shallow-water area to calculate a ‘swept area’ estimate (Walters & Martell, 2004). To estimate a total instantaneous rate of exploitation (F), we multiplied catchability (q) by annual effort (E). Catchability is the area covered by a typical unit of fishing effort (a; average fishing distance travelled night-1 calculated from GPS tracks x 6 m width of lantern  74  light) divided by the total area available for lantern fishing in Danajon Bank (A; area in waters <5 m (see results): 712 Km2 (Hansen et al., in press). We calculated annual effort as the product of the 200 lantern fishers in the region and the 111 to 192 nights year-1 they fished (Meeuwig et al., 2003). The ‘swept-area’ calculation assumes that fished areas, a, are randomly distributed across the area available for fishing (A; analysis of fishing effort distribution in Appendix C). If we assume any prey item in a was captured, we can then calculate the maximum fishing mortality rate each year (U) as 1 – e –F. We report F and U as a range corresponding to the range of nights year-1 fished. Although these calculations are typically used to assess fishing mortality rates on individual species, we estimated F on a depth stratum because data on the spatial distribution of lantern fishery-exploited species were unavailable. .  Results Spatial fishing behaviour  Lantern fishers left shore between 8.30 pm and 2.00 am and returned between 5 and 8 am (Table 4.1) spending, on average, two hours travelling to go 14.4 ± 6.5 km round trip. The furthest straight-line distance fished or travelled was 11.8 km. Fishers bypassed nearby grounds to fish at more distant locations (Fig. 4.2a, b). Fishers spent an average of 4.6 ± 1.6 hours covering 3.5 ± 1.1 km horizontal distance while fishing. Although Batasan lantern fishers spent significantly more time fishing than those from Handumon (Wilcoxon Rank Sum: W = 32.5, p=0.02), there was no difference in distance covered during fishing between  75  the two communities (W = 24, p = 0.33). Once at a fishing ground, they followed ocean floor features like reef edges and shoals – or a sunken hill (Fig. 4.2c). Integrating information from GPS tracks, diving behavior, satellite habitat maps, and participant observation we generated a three-dimensional mapping of fisher behaviour. For example, the tracks at the beginning of fisher number 3’s night, show that he made two loops around a shoal (Fig. 4.2c: I, II). We linked depth and space by referring to his dive profile during these first 35 minutes. His dive profile during that time showed that the deepest dives were during the first 20 minutes, to about 4 m (Fig. 4.3 I and 4.2c I; the outer loop), followed by 15 minutes spent diving between 1 and 2 m (Fig. 4.3 II and 4.2c, II; the inner loop). This approach was then applied to the entire GPS and TDR record for each fisher. The depth logger reported ‘0 m’ when it is above the water, confirming that the fisher exited the water and returned to his boat three times (Fig. 4.2c III, V, VI and Fig. 4.3 III), paddling to a different fishing ground. Since we joined this trip, our observations also corroborate these behaviours. Lantern fishers exploited the shallow waters, making short dives. Two of the 11 fishers remained in such shallow water that their depth loggers never recorded depths exceeding one meter. For those who dove deeper than one meter, average dive depth according to the depth logger was 1.9 ± 0.68 m and maximum dive depth for any fisher was 4.7 m (Table 4.1). Average dive duration was 14.3 ± 4.73 seconds, up to a maximum of 20 seconds. There were no detectable differences in any of these parameters between Handumon and Batasan fishers (Appendix B). A more detailed analysis of fisher’s dive profiles suggests different phases or types of dives (Fig. 4.3), which participant observation confirms. For the deepest diving fisher (# 3;  76  Table 4.1), deeper dives were characterized by longer dive durations (Fig. 4.3 I; Spearman Rank Order Correlation: r = 0.73, p < 0.001, n= 46) and longer surface intervals (r = 0.65, p= 0.004, n=46). At intermediate depths a fisher can have one hand on the boat hull and reach the seafloor with the spear in his extended other hand, for example turning over loose macroalge with the spear (Fig. 4.3 II); he then dives closer to the bottom if he sees something to catch. Fishers may also spend extended periods in very shallow depths (Fig. 4.3, IV) to search the bottom while swimming at the surface or walking on the reef below them. Fishers told us that the depth in their fishing grounds was generally 5.4 ± 1.6 m and that although they were able to dive to 9.4 ± 3.4 m, they preferred to fish shallower because deeper dives exhausted them, shortening their fishing trip (3 responses) or deeper dives wasted time (1 response). Some fishers also said they were limited to shallower depths by the amount of time they could hold their breath (5 responses) or the depth the lantern could penetrate (6 responses).  Fishing mortality rate  Drawing on what we learned about lantern fishers’ spatial fishing behaviour, we were able to develop an overarching assessment of the pressure they exert on the shallows they fish. While a single lantern fisher could search only a tiny fraction of the fishable depths (< 5 m) of Danajon Bank (about 0.02 km2) per night, over the course of a year the entire fleet of 200 lantern fishers could cover between 55 and 95% of Danajon Bank (F) for an maximum  77  annual fishing mortality rate (U) of 0.4 – 0.6 (i.e. 40 – 60% of fish in waters <5 m are caught).  Discussion  Although lantern fishers, like other small-scale fishers, may be technologically limited, they remain capable of exerting considerable fishing pressure on the shallow marine environment. Digital tracking combined with direct observation and fisher reports enabled this estimate of fishing mortality rate for the lantern fishery and provided novel insight into spatial fishing behaviour useful for quantifying the sustainability of small-scale fishing. Digital tags worn by lantern fishers demonstrated that they are strictly shallow divers, which may be a strategy for continuous fishing throughout the night. Although lantern fishers reported being capable of deeper dives, the positive correlation between dive depth and post-dive surface time suggests that there is a cost associated with deeper dives. Similarly, free-diving fishers in Korea show patterns of fishing at shallower depths than the maximum depths they can reach (Hong et al., 1963). In addition to gear limitations, perhaps free-diving fishers remain in the shallows because they are subject to the same physiological trade-offs that are well understood in other air-breathing diving vertebrates (Nolet, Wansink & Kruuk, 1993; Goldbogen et al., 2008). That is, longer dives incur an oxygen debt and a build up of carbon dioxide requiring longer post-dive recovery times at the sea surface (Mori, 1998). Remaining at shallower depths minimizes diving costs, increasing the amount of time spent fishing (as opposed to recovering). Therefore, the shallow diving behavior exhibited  78  by lantern fishers may be a strategy that, as some fishers suggested, enables extended hours of fishing throughout the night. Despite a limited dive capacity, the use of low technology gear, and highly labour intensive methods, lantern fishers exert a high rate of fishing mortality on the shallow, marine environment. The regional community of 200 lantern fishers has the potential to exploit a surprisingly large area (55 to 95 % of fishable depths in our study region) each year. Lantern fishing effort may be even higher than we estimated. Another study suggested there were 25% more lantern fishers than we assumed (Green et al., 2004). If lantern fishers caught everything in their path, they would take about half of all prey items in waters <5 m (U = 0.4 to 0.6). This is a maximum annual fishing mortality because certainly lantern fishers miss some prey, e.g., the most cryptic seahorses. Fishing mortality rate will not be distributed equally across fishing grounds, as revealed by an analysis of fisher logbooks over the course of a year (Appendix C). The concentration of effort in preferred grounds (a common pattern for small-scale (Pet-Soede et al., 2001; Daw, 2008b) and industrial fisheries (Rijnsdorp et al., 1998; Jennings et al., 1999) alike) will result in higher fishing mortality rates, while other areas will remain almost untouched by lantern fishing pressure. If applied over a longer time period, the technique we developed of combining GPS, depth logger and habitat maps could be used to quantify fishing mortality rates in specific areas. These finer scale assessments are not possible with available data, but could be important for understanding the effect of fishing on sessile or sedentary species. For individual species, the accuracy of our fishing mortality rate estimates will depend on species-specific distributions. Although the calculation we used to estimate fishing mortality rate (Walters & Martell, 2004) is typically used to assess individual species,  79  we estimated pressure on a depth stratum because data on the spatial distribution of species exploited by lantern fishers were unavailable. For species that exist solely within the 5 m isobath we studied, our estimate of instantaneous fishing mortality rate represents a lower bound on possible values. If a species is distributed and targeted only in certain areas or habitats within those shallows, fishing mortality rate becomes more concentrated / higher than it would be if the species extended across the entire area. In contrast, for species distributed more broadly or deeper than lantern fishers can fish, our estimate of instantaneous fishing mortality rate represents an upper bound. Small-scale fishers may face physical or gear limitations that restrict their capacity to overexploit. Because of their shallow diving behavior, lantern fishers effectively forfeit prey in deeper waters, including potentially higher value prey or a greater density of prey. As a result, exploited species, including seahorses, may benefit from a deep sea refuge from fishing. Seahorse life history (e.g. small home range, limited mobility, complex mating system; Rowe & Hutchings, 2003; Foster & Vincent, 2004), declining catch rates (O'Donnell et al., 2010), and surging global demand for use in traditional medicines and aquaria (Vincent, 1996), all suggest they are over-exploited. Yet, thousands of individual seahorses are still caught each year in our focal communities (Vincent et al., 2007). The many marine reserves in the region (all in shallow water) protect some individuals. On the other hand, Outside the Philippines the tiger tail seahorse, Hippocampus comes, has been found in deeper waters (to at least 20 m; Kuiter, 2000) that would be inaccessible to lantern fishers. If there are inaccessible fish, they could seed the shallower, fished populations, providing some resilience to fishing.  80  Our quantification of spatial lantern fishing behaviour has implications for practical management of this small-scale fishery. Any conservation or fishery assessments based on fishery-dependent catch estimates may be biased toward underestimating population size because the fishers only ‘sample’ the shallows and preferred fishing grounds. In addition, communities and managers in the heavily exploited Danajon Bank should consider protecting deep-water areas that could be important sources of fish (Green et al., 2004). Although small-scale fisheries have the potential to be more sustainable than their industrial counterparts (Jacquet & Pauly, 2008; Jacquet et al., 2009), our study contributes to a growing literature suggesting small-scale fisheries can have big effects on the marine environment. While small-scale fisheries may never reach the intense coverage of the seafloor of industrial fisheries (areas of the North Sea and the northeastern shelf of the United States and Canada are on average completely covered by commercial fishing several times a year: de Groot, 1984; Churchill, 1989; Messieh et al., 1991; Auster et al., 1996), it is easy to see how the fishing mortality rate exerted by lantern fishers would lead to depletion in the shallows when put in a regional context. The effects of a few hundred lantern fishers are magnified by the effort of the almost 30,000 other small-scale fishers using 50 types of gear on Danajon Bank (Green et al., 2004). The multiplicative power of high fishing effort to result in large cumulative catches and therefore declines (as has occurred in recreational fisheries; Cooke & Cowx, 2004) is worrisome for the sustainability of small-scale fisheries, especially in developing countries where human populations are still increasing (Pauly et al., 2002). There is a clear need to temper the suggestion that small-scale fisheries are ‘our best hope for sustainable fisheries’ (Jacquet et al., 2009) to reflect cumulative pressures of many small-scale fishers operating  81  many types of gear on many days. While the shore-based, shallow habits of small-scale fishers certainly result in more localized effects than those of industrial fisheries, the capacity of un-fished areas to replenish highly exploited areas is poorly understood (Sale et al., 2005). Therefore, although small-scale fisheries remain largely un-regulated (Munro, 1996; Pauly, 2006), our study underscores the importance of developing management plans to ensure their sustainability.  82  Tables and Figures  Table 4.1. Transit, fishing, and dive statistics for each lantern fisher in Handumon and Batasan recorded by GPS and TDR. Transit  ID 1 2 3 4 5 6 7 8 9 10 11 12  Community Batasan Handumon Handumon Batasan Batasan Handumon Handumon Handumon Handumon Batasan Handumon Batasan  Start Time 20:27:53 0:50:05 23:22:05 20:35:58 21:49:23 0:07:05 23:48:53 2:13:31 0:26:35 23:52:04 1:28:34 21:47:05  End Time 5:44:53 5:53:05 5:54:05 4:53:28 6:37:23 6:00:00 5:32:23 5:42:45 8:01:05 5:21:34 6:30:34 5:55:05  Max Distance from Home (km) 11.8 5.2 8.7 2.4 9.4 10.3 6.2 4 7.6 3.7 5.1 8.2  Total Time (hr) 2.4 1.3 1.6 0.7 1.7 2.4 1.3 1 3.5 0.6 1.4 1.6  Total Distance Traveled (km) 22.9 12.3 17.7 3.5 18.4 21.2 12.7 7.7 23.0 6.4 11.8 15.2  Fishing  Mean Speed (km/hr) 9.5 9.5 11.1 5.0 10.8 8.8 9.8 7.7 6.6 10.7 8.4 9.5  Total Distance Total Traveled Time (hr) (km) 6.9 3.3 3.8 3.3 4.9 4.8 7.6 5.2 3.9 2.6 3.2 3.5 4.4 4.2 2.4 2.7 3.4 1.9 4.9 4.2 3.1 1.6 6.5 4.3  Dives  Mean Speed (km/hr) 0.5 0.9 1.0 0.7 0.7 1.1 1.0 1.1 0.6 0.9 0.5 0.7  No. of Dives >1m 27 12 47 0 0 7 2 4 78 42 23 89  Max. Dive Depth (m) 1.6 2.5 4.1 NA NA 3.5 3 2.7 4.1 3.9 3 4.7  Mean Max. Dive Dive Mean Dive Depth Duration Duration (m) (s) (s) 1.2 17 10 1.8 12 9 2.4 32 19 NA NA NA NA NA NA 1.8 19 14 2 12 9 2.2 26 21 2.4 31 20 1.9 19 12 1.9 24 17 1.3 24 10  * Did not qualify as ‘diving’ because spent all time in water equal to or shallower than 1 meter  83  Figure 4.1. Danajon Bank study region of the central Philippines and our focal fishing communities of Handumon and Batasan (a). Lantern fishing method (b) recreated from Vincent and Pajaro 1997 and (c) lantern fisher taking a breath at the surface.  a) Cebu  B  n o  a  n  k  j a n a Handumon  D  Batasan  b)  Province of Bohol  c)  84  Figure 4.2. GPS tracks (white dots) of a) Batasan and b) Handumon fishers’ night of fishing relative to land (terrestrial vegetation – red), shallow water reefs (light blue), and deeper water (dark blue). Section of fisher #3’s GPS track (yellow in b) is shown in detail in c) where fishing (I, II, IV; filled circles) and travel (III, V, VI; crosses) behaviours are described. Inset shows inner (pink) loop and outer (yellow) loop that are linked with dive depth profile (Fig. 4.3).  a)  b)  c) IV III  V VI I II  85  Figure 4.3. Dive profile for fisher #3, highlighting detailed dive behaviours (I, II, IV) and out-of-water travel time (III) that can be linked with ocean floor topography (Fig. 2b, c). I  I  II  III  II  IV  IV  Time (minutes)  86  Chapter 5: Out of reach - effects of a depth refuge on the sustainability of a small-scale fishery  87  Introduction  Natural refugia play an important role in protecting wildlife from overexploitation. Those species whose geographic ranges extend to inaccessible locations or even beyond where hunting is convenient or cost-effective have a spatial refuge from exploitation (e.g. Clayton et al., 1997). Under certain conditions animals may disperse out of refugia, bolstering or replenishing populations in hunted areas (Allen & Sargeant, 1993; Novaro, Redford & Bodmer, 2000). Refugia can be especially important if they contain critical life stages (breeding/nursery habitat) or are the last remaining intact habitat for otherwise highly threatened species (e.g. the Amur Tiger Panthera tigris altacaica, Novaro et al., 2000). Natural refugia from exploitation exist in the marine realm but are becoming scarce. Historically it was common for economic costs and technological limitations to create large refugia from fishing in the deep, open ocean. Contemporary studies of fisheries that cannot access an entire population (e.g. Kang et al., 1998; D’Onghia et al., 2009) suggest that historic natural refugia enabled fisheries to last for centuries although they were un-regulated (Walters, 2000; Pauly et al., 2005). Natural marine refugia began to disappear with improvements in engine power and ‘fish finding’ technology that came with increased industrialization of fishing in the 1950s (Dugan & Davis, 1993; Bohnsack, 1998; Walters, 1998; Swartz et al., 2010). These technological advances have not been acquired by most small-scale fisheries that predominate on tropical coral reefs, where fishing is limited to nearshore, shallow habitats (Munro, 1996). Refugia may be particularly important for the conservation of fishes on coral reefs, where there is high biodiversity, local species have been overexploited, and tens of millions  88  of people continue to depend on fish for food (Pauly et al., 2002; Bellwood et al., 2004; Burke et al., 2011). Although depth is a relatively unexplored dimension of the effects of fishing on reefs (Kahng et al., 2010), two recent studies have found strong evidence for depth refugia (> 7m and > 25 m respectively) in fished reef areas, as indicated by greater commercially exploited fish species richness and relative abundance with increasing depth (Tyler et al., 2009; Goetze et al., 2011). These results combined with the anecdotal evidence for depth gradients reported by previous reef fishery studies suggest the potential of a geographically widespread role of depth in mediating the effect of fishing (Tyler et al., 2009). The relatively shallow nature of refugia on coral reefs (starting at 10s of meters as compared to the 100s or 1000s of meters in temperate/industrial systems; e.g. Bailey et al., 2009) makes them more vulnerable to encroachment by technological improvements and other disturbances (e.g. dredging or sea filling). In addition, because many coral reef species tend move to deeper waters as they get older (Brokovich et al., 2007; Appeldoorn et al., 2009) and older fish tend to produce more and/or better quality young (Birkeland & Dayton, 2005), deep refugia may contribute disproportionately to shallower coral reef fish populations. Little is understood, however about how these deep reef areas may be important for replenishing their shallow counterparts. The science for understanding the roles of refugia in supporting coral reef fish populations comes largely from the burgeoning field of Marine Protected Areas (MPA) science. In theory, such refugia could support fisheries outside their boundaries if eggs or larvae dispersed out of the refuge or if increasing adult density inside the refuge caused adults to ‘spill over’ into fished areas (Dugan & Davis, 1993; Hilborn et al., 2004). Therefore, the distance, amount, and direction of fish movement between refuge and fished  89  areas (deemed ‘connectivity’), determines the capacity of refugia to enhance and stabilize marine fisheries. Unfortunately we have a limited understanding of the general characteristics of spatial dynamics in marine populations, especially at early life stages (Sale et al., 2005; Botsford et al., 2009). Therefore the effects of refugia, and any fishery benefits, remain uncertain (Hilborn et al., 2004; Sale et al., 2005). Modeling can be a pragmatic tool for evaluating the importance of natural refugia to species viability and small-scale fishery sustainability. Although small-scale fisheries employ the vast majority of fishers globally (Berkes et al., 2001), they are poorly understood with few data available on catches or the biology of exploited species (Orensanz et al., 2005). Those on coral reefs can be especially data poor (Munro, 1996; Sadovy, 2005). When urgent decisions need to be made about threatened species, modeling allows us to combine available data and systematically incorporate uncertainties to weigh management or recovery options. For example, models allow us to incorporate uncertainty in processes like dispersal (e.g. Jones et al., 2009) or density dependence (Hixon & Webster, 2002; White et al., 2010) of coral reef fish, critical to consider in the design of marine reserves (e.g. Gerber et al., 2005). While few modeling exercises in the marine reserve literature evaluate effects of refugia on conservation objectives (Gerber et al., 2003), model outputs can easily be used to assess extinction risk using criteria like the IUCN Red List (IUCN, 2001). This study casts the question about the importance of refugia in a slightly different light than other models of MPAs; we predict the risk (to shallow populations and catches) of beginning to fish an already established natural depth refuge. Other models of MPA effectiveness as a fishery management tool largely investigate whether establishment of a refuge can increase fishery yields despite the loss of fishing area to the refuge (Hastings &  90  Botsford, 1999; Gaylord et al., 2005; White & Kendall, 2007). We examined evidence for a depth refuge from a small-scale, free-dive fishery in the Philippines and assessed the importance of the refuge for maintaining population size and catch, and preventing extinction of a threatened target seahorse (listed as Vulnerable on the IUCN Red List; IUCN, 2008a) given current fishing pressure. We constructed a population accounting model to quantify the effects of the refuge and evaluated the sensitivity of model outcomes to assumptions about key parameters.  Methods Study area, fishery, and species  Our research is situated on Danajon Bank (Fig. 5.1), a double-reef system high in reef fish diversity, but threatened by overfishing and habitat destruction (Alcala & Russ, 2002; Green et al., 2003; Carpenter & Springer, 2005). Hugging the northern and western shores of the province of Bohol, the inner and outer reefs are shallow (approximately <10 m) and composed of scattered and patchy coral reefs interspersed with Sargassum and seagrass habitats (Meeuwig et al., 2003), but channels between reefs can reach 40 m (Hansen et al., in press). The shallow reefs are in poor condition because of over-fishing and destructive activities such as blast fishing (Marcus et al., 2007), which has spurred many communities to establish small MPAs (Marine Protected Areas – no fishing zones). The MPAs cover only a small-fraction (0.32%; Hansen et al. in press) of DB, are variably enforced (Samoilys et al., 2007), and are only in shallow waters (~<10 m; Hansen et al. in press).  91  Our work is focused on the threatened tiger-tail seahorse (Hippocampus comes Cantor 1850) which is one target of multi-species extraction by small-scale lantern fishers. Two of our focal communities, Handumon (municipality of Getafe) and Guindacpan (municipality of Talibon), are known for their expertise at lantern fishing to catch seahorses. Handumon has the largest concentration of lantern fishers in the region (Meeuwig et al., 2003) and catches the most seahorses, with Guindacpan a close second (Project Seahorse unpublished data). In the other two focal communities of Batasan (municipality of Tubigon) and Cataban (municipality of Talibon), fishers used to breathe compressed air through a tube while diving deep to fish (for sea cucumbers, seahorses, etc.), but turned to other gears when compressor diving was banned. Compressor fishing allowed access to species at depths beyond where most fishing activity occurs in this region and is known to have caused depletions in targeted species (Gallardo, Siar & Encena, 1995; Amar, Cheong & Cheong, 1996; Pet & Muljadi, 2001). At night, lantern fishers sling a shaded lantern on the front of an outrigger canoe to illuminate the benthos below. This allows them to free-dive, spear, and hand-collect fish and invertebrates, but lantern fishers are limited to the upper 5 m of water by their free diving (Chapter 4). While these fishing methods may seem benign, our research suggests that lantern fishers are capable of exerting considerable exploitation rates on the shallows and could take 40 – 60% of fish each year in waters <5 m (Chapter 4). The breath-hold nature of lantern fishing may create an important depth refuge for seahorses. Seahorse life history (e.g., small home range, limited mobility, complex mating system; Rowe & Hutchings, 2003; Foster & Vincent, 2004) and surging global demand for use in traditional medicines and aquaria (Vincent, 1996), put them at risk of over-  92  exploitation. With the Philippines as an important historic exporter of seahorses (Vincent, 1996), lantern fishers report that seahorse catch rates have declined dramatically (up to 93% by 1994; O'Donnell et al., 2010) since they began fishing in the early 1970s. Despite these declines in the shallows and other qualitative indications of over fishing (Martin-Smith et al., 2004), thousands of individual seahorses, primarily H. comes, are still caught by lantern fishers each year in our focal communities (Vincent et al., 2007), but we do not have tallies of total regional catches or population size. Shallow MPAs certainly protect some individuals (Yasue & Vincent, in review), but a depth refuge might provide an important contribution to resilience to fishing. Hippocampus comes has been best studied extensively in DB, as part of our longrunning community-based research and management programme (Martin-Smith et al., 2004; Vincent et al., 2007; Yasué et al., 2010), but only in the shallows. Overfishing and habitat damage contributed to this species’ being listed as Vulnerable to extinction on the IUCN Red List (IUCN, 2008a). This fish is a medium-sized smooth, dark seahorse often with yellowand-black-striped pattern on the tail, found around the South China Sea (Lourie et al., 2004), with most sightings in the Philippines. Our studies suggest that H. comes lives at least 2.5 years in the wild (Morgan, 2007) and like many seahorses, it tends to maintain prolonged, monogamous pairings for reproduction (Perante et al., 2002). Males give birth to live young in batches or ‘broods’ which can disperse 12 to 210 km (Morgan, 2007) before settling mostly in shallow sargassum habitats (Morgan & Vincent, 2007). Holdfasts grasped as anchor points by adults with their prehensile-like tails include a wider diversity of habitats (e.g. branching sponges and corals; Morgan & Vincent, 2007). Adults maintain small home ranges often no more than 1 m2 (Perante et al., 2002) and are patchily distributed at  93  extremely low densities (1.43 x 103 m-2; Morgan, 2007). Previous genetic analysis indicates that individuals from DB are likely to be part of a single interbreeding population (Casey, 1999). Although in DB H. comes has thus far only been studied in the shallows (< 5m), outside the Philippines H. comes has been found in deeper waters (to at least 20 m; Kuiter, 2000).  Evidence for depth refuge  We used two approaches to evaluate whether H. comes inhabits waters deeper (> 5 m) than lantern fishers can fish: fisher interviews and SCUBA surveys independent of any fisheries. We relied on fisher interviews to help prioritize locations to begin searching on SCUBA.  Fisher interviews  We consulted fishers to get a broad understanding of the depth distribution of seahorses and to identify locations for fishery independent surveys. During November and December of 2007 we conducted rapid, structured interviews with individuals in our four focal communities who fish with a variety of gears at a range of depths and catch seahorses. Since there is no official register of fishers, we consulted village leaders for initial recommendations of individuals who might be willing to be interviewed and thereafter found  94  respondents by word-of-mouth (Snowball sampling; Berg, 2001). For individuals who reported catching seahorses in the past 5 years, we asked for more detail about fishing ground names and locations, depths, and habitat types. Fishers generally reported depth in units of ‘dupa’ which on average converted to 1.68 m. Interviews were conducted in the local language using trained research assistants and followed an interview schedule (Table 5.1). Ultimately, we interviewed 16 fishers in Handumon, 20 in Guindacpan, 29 from Batasan, and 10 in Cataban. Numbers were determined by availability of respondents and their willingness to participate.  Fishery independent survey  We conducted an Underwater Visual Census for seahorses in waters deeper than lantern fishers can reach (from 8 to 14 m; Chapter 4) between February and June 2008. Although lantern fishers only fish to 5 m (Chapter 4), we began sampling at 8 m to be sure we were beyond their reach. We chose a total of 5 sites (Fig. 5.1) that fishers reported as having the most seahorses or the greatest vertical habitat that could serve as a seahorse holdfast. Four sites were directly offshore / down the reef slope from either current, shallow seahorse fishing grounds (Tajong tajong, Kangmungo, and Lubi) or an MPA (Lawis Toni). The fifth site (Takot pito) was in open water on a deep shoal that was a former seahorse fishing ground for compressor diving fishers. Since H. comes is most easily seen at night when it rises up on holdfasts (Morgan & Lourie, 2006), we searched for seahorses between 2100 and 0300 hours. In the late afternoon  95  we laid two 100 m transects haphazardly through areas with vertical habitat, seeking areas most likely to hold seahorses; such bias is acceptable in light of our objectives. At night, 2 divers trained to find seahorses searched an estimated 3 m on either side of the transect and timed their swims. The exact distance searched from the transect varied due to visibility and current in order to maintain safe diving conditions. We surveyed each site only once, searching for a total of 839 minutes and an estimated 8910 m-2. When a seahorse was sighted we removed it from its holdfast to record species, sex (male, female, juvenile), head length (HL) and trunk length (TrL) (Lourie, 2003) and then returned it to its holdfast. We report all seahorse sizes as estimated standard lengths (SL in mm; Lourie, 2003) based on the relationship SL = 10 log( HL +TrL ) 0.881+ 0.5531 (Project Seahorse unpublished data).  Effects of depth refuge on sustainability Model overview  We developed an age-structured model to explore the impact of a depth refuge on an exploited seahorse population and associated fisheries catches. The model outcomes are affected by rates of mortality (natural and fishing), reproduction and larval dispersal, and the relative size of the refuge. We began by constructing parameters to capture the population dynamics of H. comes. These parameters describe how H. comes of different ages survive, grow, become vulnerable to fishing, and reproduce (Fig. 5.2). We predicted population size without fishing until it reached equilibrium and then ran the model with and without a depth refuge for 120 model iterations (model indices described in Table 5.2), representing 10 years.  96  Fished areas were in waters < 5 m and the refuge/deep extended from 5 m to the limit of H. comes known distribution at 20 m. Each model iteration represented a month since this timeframe provided the best trade-off between computational speed and accurately capturing the growth and mortality schedules for this species. We chose a 10-year window because it is (a) the required timeframe to assess Red List status for H. comes (IUCN, 2001), and (b) a relevant horizon for policy makers and local stakeholders. Once the model had resolved we used predicted population size to explore the importance of the refuge on population persistence, benefit to fisheries, and conservation assessment (IUCN Red List categorization; IUCN, 2001). The model was constructed and run in R (R Core Development Team, 2009; code available in Appendix D).  Survival  We assumed that survivorship of seahorses at age a (Sa) declined with age in a negative exponential manner (Equation 1). M is instantaneous natural mortality rate, which is constant across age, zone (fishing vs. refuge zone), and time (Table 5.2). S a = e −( a −1)M  Equation 1  Growth  We assumed that the seahorses grew according to a von Bertalanffy relationship with age (Equation 2; Morgan, 2007 ). La is standard length (SL in mm; Lourie, 2003) at age a, L∞  is the modeled asymptotic length, K is the metabolic growth coefficient (months-1), and ao is  97  theoretical age at length 0 (Table 5.2). These parameter values yielded negative lengths for age 0 fish; therefore, we set all negative values to 0 (Equation 2).  [  ]  L∞ 1 − e − K ( a−a0 ) ; a > 0 La =  a=0 0;  Equation 2  Reproduction  As seahorses grew, we assumed that they became more fecund, both because of pouch capacity and because they could attract larger, more fecund females. Individual fecundity at age (fa) is based on male fecundity (mfa), because only those data are easily accessed. Males did not reach maturity until 93 mm SL, age 4 months in the model (Morgan, 2007), after which the number of embryos produced per month by a male at age a is a function of the average number of broods per month (b; Table 5.2) and the number of embryos per brood, e, (Equation 3a) which is itself a function of La (Equation 3b; Table 5.2). Sex ratios of H. comes in our study area approximate unity (Morgan & Lourie, 2006), reflecting their monogamous mating system. Overall individual fecundity at age was calculated as fa = 0.5mfa.  be; La ≥ 93 mf a =  0; La < 93  Equation 3a  e = α + βLa  Equation 3b  98  Vulnerability to fishing  We simulated a fishery to which seahorses recruit at 70 mm SL (Table 5.2) after which vulnerability to fishing increases linearly with length until L∞ (Equation 4; Morgan, 2007).  1 ; L ≥ 70  Va = 1 + e −a a 0; La < 70  Equation 4  We calculated catches after predicting population size at each age and time step (see below).  Designate fished and refuge zones  The population was divided into two zones i: fished (shallow, < 5 m) and refuge (deep, 5 to 20 m) (Fig. 5.2). We use this coarse, implicit spatial structure because it reflects our uncertainty about connectivity in our system. It has also been argued that such simple models of spatial structure can produce similar results to those that are more complex (White & Kendall, 2007). For our default simulations we assumed that the refuge harbored 25% of the population of seahorses (for simplicity we call this a small refuge), which is based on available information on our study species and region (Table 5.2). In the absence of better information, we assumed that (i) seahorses were evenly distributed in suitable habitat within and across fished and refuge areas and (ii) patchiness in suitable habitat was similar across the entire Bank for all depths considered here (0 to 20 m).  99  Recruitment and larval dispersal between fished and refuge zones  Monthly recruitment rates were based on the number of embryos (Et ) produced by individuals in all mature age classes and both zones N tai using Equation 5. 35  2  E t = ∑∑ f a N tai  Equation 5  a =1 i=1  Following embryo production, we simulated dispersal between the two zones. The number of embryos distributed to zone i during month t (Eti) was calculated as Eti = pi Et , where pi is the proportion of total embryos that moved to zone i, i = 1 or 2, and p1 + p2 = 1 . We explored the full range of possible p values, i.e., where both zones receive from 0 to 100% of the embryos. We assumed that (i) all movement between zones occurred within the first month of the juveniles’ life, because most dispersal occurs during the pelagic larval stage, which only lasts for 5-10 days (Morgan, 2007), and that (ii) adult seahorses were highly sedentary with home ranges often < 1m2 (Perante et al., 2002). Also based on the sedentary nature of adults, we assumed no adult migration between areas. Embryos from one month become age 1 month fish in the next according to a Goodyear compensation ratio, ρ , (Goodyear, 1993) form of the Beverton-Holt recruitment relationship (Walters et al., 2005) in Equation 6 N t +1,1i =  ρ (Ri′ / Ei′)Eti 1 + [(ρ − 1) / Ei′]Eti  Equation 6  where ρ is the relative improvement in survival rate as E approaches zero (Walters & Martell, 2004). Density dependence is given by ( ρ − 1) / Ei′ where Ei′ is average, un-fished, monthly embryo production in zone i for a population with recruitment Ri′ .  100  Model initialization  We initialized the model by seeding with 2 million fish and then allowed the total population to equilibrate without fishing. The starting population size is based on the most recent estimate of H. comes density in DB shallow waters (1.43 x 103 m-2; Morgan, 2007) and the area of reef on the Danajon Bank that falls within the 0 to 20 m depth range of H. comes (~1420 km2: Hansen et al. in press). To run this initialization, we first had to calculate the number of recruits required per month in zone i ( R′i where i = fished or refuge) needed to produce a total population of 2 million fish across both zones. We did this by solving the survival equation N1ai = Ri′S a for R′i , assuming that population sizes in both areas are equal and setting the number of fish at t=1 in zone i, N1ai, to 1 million ( Ri′ = 98,126). We then used R′i to calculate the average equilibrium monthly embryo production in zone i, Ei′ , as E ′i = φRi′ where φ was calculated with Equation 7. a =35  φ=  ∑f S a  a  Equation 7  a =1  We then predicted total population size forward in time until the total population size, Nt (as described below), had been stable for 10 months. This equilibrium un-fished number of fish at each age and zone, Ntai, then became our starting population in fishing scenarios.  Impacts of refugia on population persistence, fisheries, and conservation assessments  Once the model was initialized with the un-fished equilibrium population we explored three scenarios of increasing fishing pressure on population size and catches that  101  span the range of possible extremes: 1) no fishing in either area, 2) fishing only in the shallow (deep refuge), and 3) equal fishing pressure in deep and shallow (no refuge). In scenarios with fishing, we apply a constant U of 0.05 month-1, which is the upper bound of estimated current exploitation rates in the shallow waters of the Danajon Bank (Chapter 4). For age classes 2 through 35 months, in all months after the model had been initialized with equilibrium un-fished numbers (t = 2,...,120) we predicted numbers at age and zone according to the survival equation (Equation 8).  Nt +1, ai = Nt , a −1,i e − M (1 − Va −1U i ); a = 2,K,35  Equation 8  The total population before fishing each month (Nt) was calculated using Equation 9. 35  2  N t = ∑ ∑ N tai  Equation 9  a =1 i=1  We calculated monthly catches separately for each area (Cti) using Equation 10, where Ui is the constant monthly exploitation rate (proportion of Ntai removed per month) for each zone, and summed these to obtain total monthly catch (Ct). 35  Cti = U i ∑ Va N tai  Equation 10  a =1  Once simulations were complete, we calculated the decline in population size from unfished levels and used this decline to infer an IUCN Red List status. We compared the percent population decline predicted by the model with IUCN Red List (www.redlist.org) decline Criterion A2 (IUCN, 2001) sensu the method described in O’Donnell et al. (2010). We evaluated differences in extinction threat and fishery catch due to presence or absence of the depth refuge for H. comes.  102  Unknown parameters and sensitivity analysis  Several parameter values cannot be estimated based on available data (Table 5.2). To overcome this challenge we begin by assuming values for these parameters that reflect the information that is available. To abide by the precautionary principle (UN, 1972), however, we selected values that were slightly biased toward rendering the population less resilient to fishing pressure, hence yielding conservation management recommendations. Full justification for each of these default values is given in Table 5.2. We then explore the sensitivity of our conclusions to other values for each of these parameters. We explored the sensitivity of model outcomes across the spectrum of assumptions about 1) the strength of density dependence, 2) the amount of dispersal of embryos into the refuge, 3) refuge size, 4) the rate of natural mortality, and 5) the rate of exploitation (Table 5.2). Refuge size refers to the proportion of the total population inside the refuge before   R2′   and in addition to the small refuge used in default dispersal and fishing began   R1′ + R2′  calculations, we also tested the effect of having 50 and 75 % of total initial population in the refuge, which we refer to as medium and large refuges respectively (Table 5.2). We evaluated differences in percent population change, extinction threat, and fishery catch due to presence or absence of the depth refuge.  103  Results Evidence for depth refuge Fisher interviews  Fishers provided the first evidence of a depth refuge from lantern fishing and critical spatial information about likely spots to find seahorses. In total, eight types of fishing gear were used to catch seahorses in waters from ankle deep to 30 m (Fig. 5.3), far beyond the depth range of lantern fishers. The fishing grounds where most seahorses were caught were in waters shallower than 2 meters; these grounds were used by gleaners, seaweed farmers, fish weirs, and abalone divers. Crab pot fishers reported having seahorses attached to pots when they pulled them up from waters as deep as 14 m. Compressor divers, crab net and gillnet fishers reported catching seahorses across the widest range of depths (6 to 30 m). With the exception of abalone and compressor divers, seahorses were not targeted; rather they were caught as a consequence of attaching themselves to passive fishing gear and subsequently sold.  Fishery independent survey  SCUBA surveys confirmed the presence of seahorses in waters deeper than lantern fishers can fish. In waters from 8.3 to 14 m deep we found 29 tiger tail seahorses, comprising 6 juveniles (21 %), 11 males (38%) and 12 females (41%). The smallest recorded individual measured 72 mm SL, while the largest individual was 168 mm SL. The  104  overall sex ratio among adults was equal (Exact Binomial test, p = 1). We also found one female hedgehog seahorse (Hippocampus spinosissimus), for an overall density of 3 individuals per 1000 m2. Seahorses were found on sponge (60%) and gorgonian coral (40%) holdfasts.  Effects of depth refuge on sustainability Default values Population persistence and extinction risk  The presence of a refuge increased the size of the total population and decreased IUCN designated risk of extinction. However, even if 25% of the population was protected, the model predicts that current shallow exploitation rate (U = 0.05 month-1) could cause large declines in population size in just 10 years, rendering the population at least Vulnerable to extinction (Fig. 5.4a). Predictions were more concerning without the refuge, with just over 20% of the un-fished population remaining, triggering an Endangered Red Listing (Fig. 5.4b).  Benefit to fisheries  The refuge benefitted lantern fishers, who cannot fish deep waters, by increasing catches from the shallow areas. The magnitude of this improvement depended on dispersal. Maximum cumulative catch occurred when 10-20% of embryos dispersed to the refuge and  105  was 15% higher with than without a refuge (with: 2.34 million seahorses; without: 1.98 million seahorses; Fig. 5.5, dashed lines). In contrast, when 90% of the embryos dispersed to the refuge, cumulative catches could be as much as 52% higher with than without a refuge (with: 207,000 seahorses; without: 99,000 seahorses caught). The refuge imposed a cost to overall fishing: when a refuge was in place, fished area catch was always lower than total cumulative catch (shallow + deep) without a refuge (Fig. 5.5).  Sensitivity analysis  Model predictions regarding fished population sizes, extinction risk and fisheries benefits were most sensitive to the strength of density dependence and dispersal. Results were less sensitive to variation in refuge size, the rate of natural mortality, and exploitation rate.  Population persistence and extinction risk – density dependence (ρ) and dispersal (p)  At current exploitation rates, a population with low density dependence (ρ = 1.1) would need a refuge and the right amount of dispersal into the refuge to avoid collapse. With a refuge, the proportion of the population remaining increased exponentially with increased dispersal of embryos to the refuge (Fig. 5.6a, solid line). With less than about 45% dispersal to the refuge, a Critically Endangered listing would be triggered, but above 85% dispersal to the refuge, the population would not even be considered threatened. In contrast, with no  106  refuge the total population could not compensate for the universally heavy fishing pressure and effectively collapsed (classifying as Critically Endangered) at all levels of dispersal (Fig. 5.6b, solid line). With increasing density dependence, the refuge became less important for sustaining populations and preventing extinction – the proportion of the population remaining with a refuge began to approximate that with no refuge (Fig. 5.6a, b). When ρ = 7, the range of extinction risk classifications with a refuge was nearly the same as without a refuge (Not threatened to Vulnerable, depending on dispersal).  Population persistence – refuge size, natural mortality (M), and exploitation rate (U)  In general, smaller refugia led to more severe declines (Fig 5.6c, d). When the amount of the initial total population protected decreased from 75% (large refuge) to 25% (small refuge), the amount of decline increased only slightly from 28% to 39%, respectively (Fig. 5.6c). Across variation in natural mortality and exploitation rate, the refuge remained important for population persistence, and hence, decreased risk of extinction (Fig. 5.6 e – h).  Benefit to fisheries  The deep refuge provided the largest increases in shallow cumulative catch in scenarios that assumed low density dependence, a medium-sized refuge, low rates of natural mortality, and high exploitation rates. Benefits to shallow catches decreased with increasing  107  density dependence. With low density dependence (ρ = 1.1), shallow cumulative catches over 10 years could be as much as 64% higher with a refuge (Fig. 5.7a solid black line) than without (Fig. 5.7a solid grey line). In contrast, at higher levels of density dependence, although the magnitude of catches was higher, the improvement in catch with the refuge decreased (when ρ = 7, max improvement in catch was 19% with refuge; Fig 5.7a dashdotted lines). The size of the refuge did little to change the benefit of the refuge to shallow catches, but a medium-sized refuge did provide the largest overall improvement in catches. With a medium sized refuge (50% protected) a maximum of almost 616,000 more seahorses could be caught in the shallows than without a refuge (Fig. 5.7b dashed lines), compared with a maximum benefit of 541,000 and 488,000 with small (Fig. 5.7b solid lines) and large (Fig. 5.7b dotted lines) refugia. Low rates of natural mortality allowed for the largest improvements in catch with the refuge. At the lowest value of M (0.06 month-1), having a refuge increased catch by a maximum of over 1.2 million seahorses (Fig. 5.7c solid lines), compared to a maximum improvement of 541,000 seahorses with the highest value of M (0.15 month-1; Fig. 5.7c dotted lines). Finally, across variation in exploitation rate, the refuge increased shallow, cumulative catches (Fig. 5.7d). The refuge provided the biggest benefit, when exploitation rate was highest (U = 0.05-1 month, Fig. 5.7d dotted lines) increasing shallow catches by up to 541,000 over 10 years.  108  Discussion This study that considers how natural refugia contribute to coral reef fish population persistence and fisheries, highlights the benefits of maintaining/protecting such refugia from encroaching fishing pressure or other human-induced disturbances. Across a range of uncertainty in key parameter values, model predictions consistently suggested that the presence of a depth refuge increased H. comes population size, decreased extinction risk, and enhanced lantern fisher catches in the shallows. The refuge was particularly important when we assumed low density dependence, but the amount of protection and benefit to shallow catches it provided was affected both by the amount of embryos that dispersed to it and the assumed refuge size. Fisher interviews facilitated our first documentation of H. comes in waters deeper than lantern fishers can reach. Fisher knowledge about where seahorses in deep waters might be located, streamlined a challenging task of locating a small, cryptic fish that occurs at low densities. The small number of deep seahorses we found during SCUBA surveys appeared to share many characteristics with shallow H. comes populations in terms of density and sex ratio (Perante et al., 2002; Morgan & Lourie, 2006; Morgan & Vincent, 2007; Yasue & Vincent, in review). Gathering data allowing for more robust estimates of population characteristics at depth will be very challenging in coral reef systems because of the practicalities of diving in such remote locations, at depth, in search of a rare species, and in our case at night. While baited, remotely operated cameras used in other depth refuge surveys (Goetze et al., 2011) may be appropriate for more mobile fish species, they are unlikely to be an effective survey tool for lower trophic level or more sedentary fish like seahorses.  109  In the model, the deep refuge from lantern fishing played an important role in sustaining populations and shallow lantern fishery catches of H. comes, even when we admitted a wide range of uncertainty in key processes and parameters. As is common for most coral reef fish and small-scale fisheries we do not know how much of the population is protected by the refuge, how connected refuge and fished populations are through dispersal, or the strength of density dependence in regulating populations (Berkes et al., 2001; Sadovy, 2005; Tyler et al., 2009). By examining the role of the refuge across a broad range of uncertainty in these key processes and parameters, we find that although they do influence the amount of protection provided by the refuge, they do not change the result that having a refuge in place increased population sizes and shallow catches and decreased IUCN designated risk of extinction. Only at the highest values of density dependence did population sizes and catches with and without a refuge begin to approximate each other, suggesting the refuge was less important. A population with strong density dependence can bounce back from heavy fishing because juvenile survival will increase at low adult densities (Rose et al., 2001). Unfortunately, as we describe in Table 5.2, H. comes is unlikely to be this resistant to fishing pressure. Prohibiting destructive and efficient gears from the depth refuge could be a pragmatic approach to preventing further collapse of H. comes populations and catches. None of the areas in waters deeper than 10 m are protected in DB (Hansen et al., in press). Although small refugia have been established in shallow waters, the amount of area they protect (0.32% of DB; Hansen et al., in press) falls short of the law requiring municipalities in the Philippines to protect 15% of their waters (Fisheries Code: RA 8550). The currently protected shallow area is also dwarfed by even our conservative estimates of the amount of  110  area in waters deeper than lantern fishers can reach that could support seahorses (25% of DB). Given the large area in deeper waters and their distance away from shore and the watchful eyes of communities, it may be impractical to enforce or patrol deeper areas to prevent all fishing. Mobile fishing gears (e.g. trawling) that destroy habitat and compressor fishing that has high rates of depletion (Gallardo et al., 1995; Akamine, 2001; Morgan & Panes, 2004) are illegal on most of Danajon Bank. Perhaps the most pragmatic solution, therefore, would be to focus limited enforcement monies on excluding these most destructive and efficient gears from deep refugia. A refuge alone may not be enough to sustain H. comes at viable population sizes. Even with a refuge, most model scenarios suggest that after just ten years of lantern fishing at current levels, H. comes population declines put it at risk of extinction (classified as Vulnerable to Critically Endangered). Only at very high levels of density dependence and when a large refuge was supplied with very high dispersal was the amount of population decline small enough not to trigger a threatened classification. This suggests that, in addition to a refuge, other options for reducing exploitation rates may be necessary to sustain H. comes at viable population sizes (as has been suggested for other species; Hilborn et al., 2004). These declines may actually be underestimated because of some of the simplifying assumptions we made when constructing the model. First, the best life history data available were measured on wild, exploited populations (Morgan, 2007). Since growth and reproduction can increase in response to heavy exploitation (Heino & Godø, 2002), using these life history data to characterize equilibrium populations before fishing (and subsequent changes with fishing) may predict greater resilience than actually exists. Second, we did not incorporate any stochasticity into this model because we do not presently have an  111  understanding of inter-annual variation in the vital rates of H. comes (Morgan, 2007). Deterministic models tend to predict that populations are more robust to exploitation, relative to stochastic models (Pitchford, Codling & Psarra, 2007), meaning that declines estimated by our model are likely to be minimum possible estimates. Although we evaluated the importance of the refuge after just ten years, we believe that conclusions drawn after this relatively short time-frame are robust. For all parameter combinations, the ten-year time span captures most of the population change. For example, when we examined the population trajectory for the combination of parameters that would result in the most extreme declines, we found that by ten years (120 months) the population had already declined by 97%, which was within 2% of the total decline at equilibrium (99%, Appendix D). In addition, our conclusions about the importance of the refuge for sustaining shallow populations appear to apply in the long term. When we ran the model until populations came to equilibrium with fishing, conclusions about the importance of the refuge were the same as in the ten-year time frame (Appendix D). Policy-makers should consider, however, two differences in results from the equilibrium (long-term) and the ten-year (shortterm) analysis. In the long-term: 1) without a refuge, the population declines more and is more at risk of extinction (e.g. at default parameter values classifies as Critically Endangered as compared to Endangered; Appendix D, Fig. D.2), and 2) with a refuge, the improvement in shallow cumulative catch is smaller (e.g. at default parameter values cumulative catches were a maximum of 49% higher as compared to 52% higher; Appendix D, Fig. D.3). The nearly universal (across wide uncertainty in key parameters and processes) benefits of deep reef refugia suggested by our study, emphasizes the importance of protecting these areas just out of small-scale fishers’ reach. Although we are only beginning to quantify  112  depth refugia on coral reefs, there is evidence that they may be widespread (Tyler et al., 2009) and are an important component for sustainable fisheries (e.g. Karpov et al., 1998; D’Onghia et al., 2005). If other refugia support population persistence and shallow catches as broadly as in our study, protecting deep (>10 m) reef areas could help safeguard the food supply shallow reefs provide for millions of the world’s most poor (Moberg & Folke, 1999). Furthermore, if refugia in other locations are as large as we estimate they could be on Danajon Bank, protecting them could help reach global marine protected area goals (10% by 2020; CBD, 2006). More research is needed to quantify the contribution that un-fished deep reef areas make to the shallows for other species. Our work provides pragmatic methods for identifying refugia and evaluating their local importance.  113  Tables and Figures Table 5.1. Interview schedule Biographical information: name, gender, age Seahorse catch: In the past 5 years, have you ever caught seahorses? What kind of fishing gear were you using when you caught seahorses? What are the names of the fishing grounds where you have caught seahorses? Can you describe each fishing ground (location, depth, habitat)? How often do you visit this fishing ground? How often do you catch seahorses at this fishing ground? What do you do with the seahorses you catch? Other: How big is a dupa?  114  Table 5.2. Symbols, definitions, values (and range used in sensitivity analysis), sources of indices and parameters used in model, and justification for default value choices. Category  Parameter  Definition  Units  Value  Source  Justification  Indices  a  age  months  1…35  -  See text  i  depth zone  -  1: shallow, 2: deep  -  See text  n  maximum model iterations  months  120  -  See text  t  month  -  1…n  -  -  M  natural mortality  proportion -1 month  0.15 (0.06 – 0.15 month )  Martin-Smith et al. 2004, Morgan 2007  Chose highest estimated value  Sa  survivorship – the probability of surviving from age 1 to age a  a probability  S 1 = 1, S 35 = 0.006, Equation 1  Walters and Martell 2004  Captures the cumulative effect of natural mortality on fish as they age  a0  Von Bertalanffy theoretical age at which length is zero  months  0.36  Morgan 2007  Growth parameters estimated using mark-recapture data, fishery landings, and aquarium-reared juveniles of known age.  K  Von Bertalanffy metabolic growth coefficient  -  0.24 month  Morgan 2007  See above  mm  L 1 = 23.8 mm, L 35 = 167 mm, Equation 2  Morgan 2007  See above  Survival  Growth  La  Reproduction / Recruitment  length at age a  -1  -1  L∞  Von Bertalanffy length as a goes to infinity  mm  167 mm  Morgan 2007  See above  b  brood number – number of broods each male has in a month  broods  1.79 broods male month  Morgan 2007  Calculated from the average reproductive duration of 17 days (Morgan 2007).  e  number of embryos per brood  embryos  Ranges from 303 to 522 embryos -1 male , Equation 3b  Morgan 2007  Fecundity increases linearly as a function of body size. The relationship between size and fecundity was derived from an analysis of male brood size and lengths measured for 95 captured H. Comes.  E i′  total un-fished monthly embryo production in zone i for a population with recruitment R i ′  embryos  Walters and Martell 2004  See text  Et  total embryo production in month t (summed across depth zones i )  embryos  Equation 5  -  See text  E ti  The number of embryos distributed to zone i during month t  embryos  fa  fecundity at age a – monthly embryo production from one age a fish  mf a  male fecundity at age a  -1  φ R i′  -  embryos month 1  -  embryos month 1  -1  piEt  -  See text  f1 to 3 = 0, f4= 270, f35 = 466, Equation = 0.5mf a  Morgan 2007  The relationship between size and fecundity was derived from an analysis of male brood size and lengths measured for 95 captured H. Comes.  f1 to 3 = 0, f4= 540, f35 = 932, Equation 3a  Morgan 2007  Fecundity increases linearly as a function of body size. The relationship between size and fecundity was derived from an analysis of male brood size and lengths measured for 95 captured H. Comes.  R1′  total un-fished age-1 recruitment in zone 1 fished/shallow  fish  147189 fish, 75% of population protected [49063 (25%), 98126 (50%) protected]  -  Since the portion of the population in the refuge at un-fished equilibrium is unknown, we assumed that it was 25% of the total population. Certainly less than 50% of the total population currently resides in the deep, so we assume this was also true before fishing. Shallow and deep areas may have similar seahorse densities (see results and discussion) and the amount of area in the shallow (<5 m) and deep (5 to 20 m) is roughly equal (Hansen et al. in press), but there is less available vertical habitat at depth with vast areas as open sand or mud flats (O’Donnell personal observation). Since seahorses show strong preference for distinctive or artificial structures (Foster & Vincent 2004) and we only found individuals associated with vertical habitat, we assumed that the total number of individuals in the deep was less than in the shallow, choosing 25% as a best guess.  R′2  total un-fished age-1 recruitment in zone 2 deep/refuge  fish  49063 fish, 25% of population protected [147189 (75%), 98126 (50%) protected]  -  See above  115  Table 5.2. Continued Category Reproduction / Recruitment  Population size  Fishing  Parameter  Definition  pi  proportion of total embryos that settle or move to zone i , by age a =1, where p 1 + p 2 = 1  Value  Source  α  intercept of linear relationship between fecundity and size  0-1.0  -  This is the full range of possibilities.  -2.69  Morgan 2007  The relationship between size and fecundity was derived from an analysis of male brood size and lengths measured for 95 captured H. Comes.  -  β  slope of linear relationship between fecundity and size  -  3.14  Morgan 2007  The relationship between size and fecundity was derived from an analysis of male brood size and lengths measured for 95 captured H. Comes.  φ  Lifetime fecundity of one fish - the sum over ages of fecundity f a weighted by survivorship S a  embryos  1730 embryos, Equation 7  Walters and Martell, 2004 H. comes has ‘weak’ density dependence (ρ = 1.5) because, according to a version of life-history theory (Winemiller & Rose 1992), this species groups most closely with equilibrium strategists (Morgan 2007) which have relatively low fecundity, and, despite high early life stage survival, should result in low maximum reproductive rates a low population sizes (Rose et al. 2001).  -  Justification  ρ  Goodyear recruitment compensation ratio  -  1.5 (1.1 – 7)  Goodyear 1993, Myers et al. 1999  N tai  population at the start of month t (before fishing) for age a in zone i  fish  Equation 8  -  See text  Nt  total population at the start of month t (before fishing), summed across ages a and depth zones i  fish  Equation 9  -  See text  C ti  total catch in month t and zone i (summed across ages a )  fish  Equation 10  -  See text  -  See text  -1  Chapter 4  I chose the upper bound of my estimate of U from Chapter 4 as the default value and evaluated the sensitivity of the model to the full range of U that I estimated.  -1  2  Ct  total catch in month t (summed across ages a and depth zones i )  fish  ∑C  U1  monthly exploitation rate in zone 1  proportion of N removed  0.05 (0 – 0.05 month )  U2  monthly exploitation rate in zone 2  proportion of N removed  0.05 (0 – 0.05 month )  Chapter 4  See above  Va  vulnerability to fishing at age a  proportion of Na vulnerable  V 1 and 2 = 0, V 3 = 0.95, V 17 to 35 = 1, Equation 4  -  Catch landings data from 1996-1998 and 2002-2004 used in lengthconverted catch curve to show that extraction of H. comes began at ca. 70 mm SL, with full exploitation occurring above 150 mm SL (Morgan 2007).  ti  i =1  116  Figure 5.1. Danajon Bank study region of the central Philippines, our four focal communities where interviews were conducted, and UVC survey sites (red stars)  Cataban  Handumon  Guindacpan  Batasan  117  Figure 5.2. Conceptual representation of model where a meta-population (a shallow/fished and a deep/refuge population) is linked through embryo dispersal. Numbers of fish (N), survivorship (S), length (L), fecundity (F), and vulnerability to fishing (V) change over age (1-35 month - subscripts).  Proportion of embryos to shallow (Et1)  V1  V2,…,34  N1  N 2,…,34  N35  month  months  months  S2  V35  S35  L1  L2,…,34  L35  F1  F2,…,34  F35  }  SHALLOW & FISHED  Total embryos at time t (Et)  Proportion of embryos to deep (Et2)  F1 L1  F2,…,34 L2,…,34  F35 L35  N1  N 2,…,34  N35  month  months  months  V1  S2  V2,…,34  S35  }  DEEP REFUGE  V35  118  Figure 5.3. Frequency of interview mentions of fishing grounds where seahorses have been caught at particular depths and the range of depths fished by gear.  Gleaning  Total seahorse fishing grounds mentioned  140  Seaweed farming & fish weirs Lantern fishing & Abalone divers  120  Crab pot  100 Compressor diving, crab net & gillnet  80 60 40 20 0 2  4  6  8  10  12  14  16  18  20  22  24  26  28  Depth (m)  Figure 5.4. Proportion of total (solid line), shallow (dashed line), and deep (dotted line) H. comes populations remaining after 10 years at default parameter values and across a range of uncertainty about the proportion of embryos that disperse to the deep. We compare two scenarios: 1) a small, deep refuge where 25% of the un-fished equilibrium population was assigned to the refuge and 75% were fished at current exploitation rate (a), and 2) no refuge where 100% of the un-fished equilibrium population was fished at current exploitation rate (b).  Deep refuge  No refuge 1.2 0.8  1.0  Total population Fished/shallow population Refuge/deep population  0.6  0.6  0.8  1.0  1.2  b)  VU  0.4  0.4  VU  EN  0.2  0.2  EN  CR 0.0  CR 0.0  Proportion of un-fished population remaining after 10 years  a)  0.0  0.2  0.4  0.6  0.8  1.0  0.0  0.2  0.4  0.6  0.8  1.0  Proportion of embryos dispersing to refuge / deep 119  30  Figure 5.5. Comparison of cumulative catches of H. comes after 10 years from the shallow (dashed), deep (dotted), or total (shallow+deep; solid line) zones. We compare two scenarios: 1) a deep refuge (black lines) and 2) no refuge (grey lines).  2000000  Total Fished/shallow Refuge/deep  500000  1000000  Deep refuge No refuge  0  Cumulative catch (10 years)  Zone & Scenario  0.0  0.2  0.4  0.6  0.8  1.0  Proportion of embryos dispersing to refuge / deep  Figure 5.6. Proportion of total H. comes population remaining after 10 years given a range of assumptions about strength of density dependence (rho, a, b), refuge size (c, d), natural mortality (e, f), and exploitation rate (g, h) and a range of uncertainty about the proportion of embryos that disperse to the deep. For each parameter we compare two scenarios: 1) a small, deep refuge where 25% of the unfished equilibrium population was assigned to the refuge and 75% were fished at current exploitation rate (left column), and 2) no refuge where 100% of the un-fished equilibrium population was fished at current exploitation rate (right column).  120  Figure 5.6. No refuge  1.0 0.8  CR 0.0  d)  0.2 No 0.4refuge 0.6  0.8  1.0  Refuge size  0.8  1.0  Small Medium Large  0.6  0.6  0.8  1.0  1.2  1.0  1.2  0.2 Deep 0.4 refuge 0.6 0.8  VU  0.4  0.4  VU  0.2  CR  0.0  CR  0.0  EN  0.2 Deep 0.4 refuge 0.6 0.8  1.0  0.0  0.2 No 0.4refuge 0.6  0.8  1.0  Natural mortality  0.6  0.8  1.0  0.06 0.1 0.15  CR  0.0  CR  0.0  EN  0.2  0.2  EN  0.2 Deep 0.4  VU  0.4  0.4  VU  0.6 0.8 refuge  1.0  0.0  0.4refuge 0.6 0.8 No  1.0  Exploitation rate (U)  0.8  1.0  0.03 0.04 0.05  0.6  0.6  0.8  1.0  1.2  h)  0.2  1.2  0.6  0.8  1.0  1.2  f)  1.2  0.2  EN  0.4  0.6  CR  0.0  CR 0.2  EN  0.2  0.2  EN  0.0  VU  0.4  0.4  VU  0.0  Proportion of un-fished population remaining after 10 years  EN  0.0  0.0  CR  0.0  g)  VU  0.2  0.2  EN  0.0  e)  1.1 1.5 2 7  0.4  0.4  VU  0.0  c)  rho  0.6  0.6  0.8  1.0  1.2  b)  1.2  Deep refuge  a)  0.8  1.0  0.0  0.2  0.4  0.6  0.8  1.0  Proportion of embryos dispersing to refuge / deep 121  b)  3500000  a)  Deep refuge No refuge  2500000  Figure 5.7. Comparison of cumulative catches of H. comes after 10 years given a range of assumptions about (a) density dependence (rho), (b) refuge size, (c) natural mortality, and (d) exploitation rate. We compare two scenarios: 1) a deep refuge (black lines) and 2) no refuge (grey lines).  Deep refuge No refuge  1500000  Small Medium Large  0.0  0.2  0.4  0.6  0.8  1.0  4e+06  Deep refuge No refuge  d)  2500000  0  500000  500000  1500000  cum.Ct120  Refuge size  1.1 1.5 2 7  0  0.0  0.2  0.4  0.6 p2  Natural mortality  1.0  Exploitation rate (U) 0.03 0.04 0.05  0  0e+00  500000  1e+06  2e+06  3e+06  0.06 0.1 0.15  0.8  Deep refuge No refuge  1500000  c)  Cumulative catch  Cumulative shallow catch after 10 years  2500000  rho  0.0  0.2  0.4  0.6 p2  0.8  1.0  0.0  0.2  0.4  0.6  0.8  1.0  Proportion of embryos dispersing to refuge / deep  122  Chapter 6: General discussion and conclusions  123  Summary of thesis and status of research questions  By successfully applying pragmatic approaches I have cast new light on how we might reconcile fisheries and conservation for species threatened by small-scale fishing, with a particular emphasis on seahorses. Although the global trade for seahorses has operated for decades, I provided the first historic reconstruction of a seahorse fishery. More generally, I presented evidence that assumptions made when analyzing local knowledge affect conservation assessments for species at risk. Comparisons I made between data elicited from resource users and data obtained using scientific approaches will help to develop robust approaches for incorporating fisher knowledge into quantitative assessments. I next conducted a novel analysis of spatial fishing behaviour that illustrated the capacity of smallscale fisheries to exert considerable pressure on marine systems, raising questions about their inherent sustainability. I also contributed to marine conservation planning by quantifying the importance of deep reef refugia for sustaining shallow, overexploited reef fishes and fisheries. In this concluding chapter I summarize my collective research in applying pragmatic approaches and indicators to conservation problems and then highlight advances made toward reconciling fisheries and conservation. In addition, I discuss implications for future research both for my case study system and more broadly for threatened species and smallscale, data poor fisheries. I tapped into unconventional data sources in each of my chapters for rapid, cost-effective, information about species abundance and exploitation. Local knowledge featured prominently in Chapters 2, 3, and 5, and I directly evaluated the reliability of local knowledge in Chapter 3. I considered the behaviour of fishers in Chapter  124  4 and explicitly evaluated the effects of uncertainty on assessments in Chapters 2 and 5. Despite incomplete knowledge, I reached conclusions about extinction risk and fishery status in all chapters and made recommendations for action in Chapter 5. The precautionary principle featured prominently in my analysis in Chapter 5. Throughout my thesis, I combined indicators typically used in fisheries and conservation to quantify the approaches and to track the status of my case study fishery and threatened seahorse population. I quantified trends in catch rate (CPUE) in Chapters 2 and 3 and conducted IUCN Red Listing assessments in Chapters 2 and 5. I calculated exploitation rate in Chapter 4 and quantified the extent and potential contribution of a natural refuge from fishing in Chapter 5. I now describe how integrated application of five sets of approaches and indicators (as outlined in chapter 1) has allowed me to address my research questions with considerable success.  How has CPUE changed over time according to fisher perception and what does that tell us about extinction risk?  In the first reconstruction of a seahorse fishery (Chapter 2), I found that fishers perceived severe historic declines in CPUE that were not apparent in more recent logbook trends. Furthermore, the historic baseline set by interviews suggested that relying on recent decades of data alone may have underestimated extinction risk for the study species, and others that have been exploited historically. The current IUCN Red Listing of Vulnerable for  125  H. comes may underestimate the actual threat level, especially if the biggest declines occurred before 1990.  How do assumptions about the accuracy of fisher recall affect the assessment of extinction risk?  Also in Chapter 2 I developed a novel approach to correct for retrospective bias and explored its effects on conservation assessments. While local knowledge is increasingly being incorporated into conservation and fishery assessments, little is yet understood about its biases and there are no formal rules for how it should be analyzed or interpreted. For example, retrospective bias (how accurately fishers report past events) has been acknowledged in the literature (Neis et al., 1999b; Ainsworth et al., 2008), but there have been no attempts to quantify it, correct for it, or evaluate its effects on conservation assessments that use local knowledge. As is typical of small-scale fisheries, there were no multiple datasets that overlapped in time, so it was not possible to test the reliability of interviews by direct comparison with logbooks. Instead I developed simple calculations of CPUE for logbooks that might account for recall bias in interviews, allowing me to link interview and logbook data into one time series. My calculations showed that the relationship between perceived historic declines in interviews and recent logbook trends depended on assumptions I made about fisher recall bias. For example, assuming fisher historic interviews were accurate resulted in the most severe declines and the most worrying extinction risk classifications. To my knowledge, this  126  is the first study to show that assumptions made when analyzing local knowledge affect species conservation assessments. These findings highlight the importance of avoiding overly simplistic assumptions when employing resource user knowledge.  How comparable are three methods of quantifying CPUE (fisher interviews, fisher logbooks and ‘scientific’ catch landings)?  My results from Chapter 3 suggest that interviews or logbooks may be a reasonable proxy for more costly catch landings or underwater surveys (Johannes, 1998) when seeking information about trends, particularly given the need for rapid, low cost, and reliable assessment methods to detect trends in the abundance and use of wildlife (Danielsen et al., 2005). When compared directly, fisher and scientific knowledge have been found either to agree (Neis et al., 1999b; Begossi, 2008; Lozano-Montes et al., 2008) or to disagree (Otero et  al., 2005; Daw, 2008a). Given that my work in Chapter 2 showed that assumptions made about biases in fisher knowledge can affect conservation assessments, there is a continuing need to understand the relationships between user-generated knowledge and other datasets (Gavin & Anderson, 2005; Gilchrist et al., 2005; White et al., 2005; Gagnon & Berteaux, 2009). I found that fisher interviews and logbooks provided estimates of CPUE trends consistent with more formally collected catch landings. My results in this chapter also highlight two important limitations of local knowledge. First, interviews may not allow inferences about the absolute value of catch rates. When compared directly, interviews yielded higher and more variable estimates than other  127  methods, even than logbooks which were also fisher-derived. While this result may vary from fishery to fishery (other studies have found agreement e.g. Daw, 2008a), it highlights the importance of comparing data-collection methods when setting recovery targets or catch limits that are sensitive to the magnitude of reported values. Second, it may not be possible to quantify how fisher reports relate to scientifically collected data at a scale (the fishery) relevant to management/conservation because of variation among individual respondents. Having retrospective interview and landings data from the same individual fishers also allowed me a rare insight into the differences that may exist at the individual level. When I plotted interview responses against landings recorded for the same fisher in the same year, I hoped to find a consistent relationship (correlation) across fishers that might allow me to calculate a ‘correction factor’ that would adjust values provided in interviews to put them in the same scale as landings reports. In contrast to other studies (Anadón et al., 2009; Rist et  al., 2010), I found that interview responses were not correlated with recorded landings, making it impossible to calculate a correction factor.  What are the spatial dynamics of small-scale fishing? And what does an estimate of exploitation rate say about the sustainability of small-scale fishing?  Using a novel method of digitally tracking small-scale fishers at work, my results from Chapter 4 suggest that, despite technological limitations, small-scale fishing can be so intense as to challenge arguments that these fisheries are necessarily a sustainable alternative to their industrial counterparts (cf. Jacquet & Pauly, 2008; Jacquet et al., 2009). Information  128  about the spatial behaviour of small-scale fishing is largely unavailable (Salas & Gaertner, 2004). The few studies conducted have gathered coarse spatial fishing information by asking fishers to identify their activities on maps (e.g. Daw, 2008b) or recording coordinates of fishing grounds (Bizzarro et al., 2009). Detailed spatial information from GPS tracks and depth tags revealed that lantern fishers were strictly shallow divers and provided data to calculate exploitation rate. My results reveal the surprising capacity of small-scale fisheries to exert considerable pressure on the fishes within the reach of their limited-technology gear. Because the reach of small-scale fishers does not extend as deep or far offshore as industrial fisheries, there may be opportunity for exploited species to survive in natural refugia.  What role does a depth refuge play in preventing short-term risk of species extinction and in supporting shallow catches given current exploitation rates?  My findings in Chapter 5 highlight that although deep reef refugia are not a panacea, they can benefit both shallow reef fishes and fisheries. Interviews with fishers facilitated the first documentation of H. comes in waters deeper than lantern fishers can reach. While deep reef refugia have been identified as potentially important and under-studied (Tyler et al., 2009; Kahng et al., 2010; Goetze et al., 2011), this is the first consideration of how depth refugia might contribute to coral reef fish population persistence and fisheries. Combining fisheries modeling tools with IUCN Red List conservation assessments revealed that across a range of uncertainty in key parameter values, the presence of a depth refuge increased H.  comes population size, decreased extinction risk, and enhanced lantern fisher catches in the  129  shallows, given current exploitation rate (Chapter 4). Even with a refuge, however, model scenarios predict worrying declines in H. comes populations, suggesting that a depth refuge alone may not be enough to sustain viable populations of species exploited by small-scale fisheries.  Reconciling small-scale fishing and conservation  Having addressed my five original research questions, I now reflect on how my findings might make a material difference for seahorses, small-scale fisheries, and society. I also highlight how my work begins to reconcile fisheries with conservation and comment on areas that deserve further research.  Seahorses on Danajon Bank  Upon integrating the evidence from all of my chapters, I conclude that current exploitation by lantern fishers on Danajon Bank threatens seahorse populations and catches. The only record of historic baselines, fisher reports during interviews, suggests that catch rates have sharply declined. Hippocampus comes may be more at risk of extinction than its current Vulnerable listing because Red Listing does not consider long-term historic declines (Chapter 2). Based on the available data, it is difficult to be sure of the exact magnitude of these declines (Chapter 3), but corroboration across fishers that declines have occurred and  130  best estimates of minimum declines (75% from 1970 to 1994, Chapter 2) are cause for concern. That said, lantern fishers are restricted to shallow waters (Chapter 4) and H. comes occur well below their reach (Chapter 5), so it is possible that catch rates perceived by fishers declined more quickly than abundances for the entire population (hyperdepletion; Hilborn & Walters, 1992). Even when the total population (shallow and deep) is considered (Chapter5), however, current exploitation rates (Chapter 4) are still capable of causing alarming declines to a level that warrants threatened status on the Red List. Moreover, the depth refuge is currently at risk of being fished by highly efficient and destructive gears such as trawling and compressor fishing. This quantitative evidence of overexploitation complements qualitative indicators of overfishing that have already been deduced (e.g. juveniles captured; MartinSmith et al., 2004). In addition to the management/conservation actions that have already been recommended for this fishery (Martin-Smith et al., 2004), and seahorse fisheries generally (Foster & Vincent, 2005), my work suggests that protecting deep refugia could benefit both fisheries and conservation. No areas deeper than 10 m are protected in Danajon Bank (Hansen et al., in press). Although small refugia have been established in shallow waters, the amount of area they protect (0.32% of DB; Hansen et al., in press) falls short of national law requiring municipalities in the Philippines to protect 15% of their waters (Fisheries Code: RA 8550). The area of protected shallows is also dwarfed by my conservative estimates of depth refuge from lantern fishing area (25% of Danajon Bank, Chapter 5). Narrowing in on exactly which deep areas have seahorses will be very challenging because of the practicalities of diving in such remote locations, at depth, in search of a rare, nocturnal species. Protecting  131  deep areas identified by fishers as having seahorses or seahorse habitat would be one pragmatic approach. Prohibiting the most destructive and efficient gears from deeper waters (>10 m) might also help in preventing further collapse of H. comes populations. Given the large area in deeper waters and their distance away from shore and the watchful eyes of communities, it may be impractical to enforce or patrol deeper areas to prevent all fishing. Mobile fishing gears (e.g. trawling) that destroy habitat and compressor fishing that has high rates of depletion (Gallardo et al., 1995; Akamine, 2001; Pet & Muljadi, 2001; Morgan & Panes, 2004) are already illegal on most of Danajon Bank. Perhaps the most pragmatic solution, therefore, would be to focus limited enforcement resources on excluding these most destructive and efficient gears from deep refugia. Monitoring the response of seahorse populations and fisheries to conservation and management interventions will continue to be important. Further research on connectivity among populations would help refine recommendations for the placement and size of refugia (that this far appear not to bolster seahorse populations; Yasue & Vincent, in review) and may become more tractable with newly emerging tagging techniques (reviewed by Jones et  al., 2009). Results from tagging studies, combined with the modeling framework I developed (Chapter 5) may even allow predictions about how quickly refugia could help restore populations and fishery catches. In reality, of course, it is likely to remain more practical to monitor seahorse fisheries than seahorse populations. For example, monitoring spatial fishing behaviour could be an important way to detect the effect of technological improvements on fishing efficiency. The low-tech nature of the lantern fishery leaves it open for increased efficiency known to have  132  masked increased capacity in other fisheries (e.g. Atlantic cod; Hutchings & Myers, 1994). Since lantern fishers tend to rely on their engines to travel far from home to fish, changes in gas prices could have drastic effects on the amount of area they can access and exploit (Abernethy et al., 2010). Some fishers already report reverting to paddling because they cannot afford gasoline (O’Donnell personal observation). GPS tracking may be the most effective method for documenting spatial fishing patterns. It does not require fishers to read maps (a challenge for some), it can be groundtruthed easily, and it produces outputs that might re-ignite local interest in conservation. Moreover, GPS tracking technology is rapidly becoming more accurate, less expensive, and more broadly available now that even cell phones (omnipresent in the Philippines) are equipped with GPS tracking capacities. In general, overfishing on Danajon Bank will only be halted by addressing wider social and governance ills. Over-population, poverty, few income earning options, poor infrastructure and poor governance plague this region and are serious challenges to conservation and food security (Pauly, 2000; Green et al., 2003; Stobutzki et al., 2006; Fernandez, 2009). Fisheries research and management is shifting toward placing emphasis on interactions among people, power, external disturbance and uncertainty, and wider governance dynamics (Andrew et al., 2007). Recent research emphasizes the critical importance of understanding the socioeconomic context of small-scale fisheries for their effective management (Cinner, Sutton & Bond, 2007; McClanahan et al., 2009; Gutierrez, Hilborn & Defeo, 2011). The goal should be to move toward an adaptive management-type system where conservation/management actions are applied, monitored, re-evaluated, and  133  revised if necessary (Walters, 1986; McClanahan et al., 2009), but all this will need improved governance.  Broader implications and research directions  In addition to shedding light on the status of the lantern fishery and the threatened seahorse H. comes, my evaluation of approaches to reconcile fisheries and conservation contributes new insights that suggest future research. My research helps develop robust approaches to incorporating local knowledge into quantitative assessments. Although the uncertain reliability of local knowledge may be an impediment to its use (White et al., 2005; Brook & McLachlan, 2008), the existence of bias does not render local knowledge useless (Paterson, 2010). In fact, local knowledge should be probed with the same analytical rigor as scientifically collected data, and to do so we need methods to quantify and correct bias in local knowledge. Developing methods of quantifying these biases and/or procedures to correct for them, as I explored in Chapter 2, would allow researchers to interpret interview responses appropriately, improving conservation assessments. A range of existing techniques should be utilized to help us collect and apply local knowledge as accurately as possible. There is a growing body of literature about how to design studies, phrase questions, and choose appropriate respondents (e.g. Neis et al., 1999a; Davis & Wagner, 2003; Anadón et al., 2009). Quantitative assessment protocols are being developed that explicitly incorporate local knowledge (e.g. Walmsley, Medley & Howard,  134  2005). It is necessary to acknowledge memory biases that may be present (Daw, 2010) and consider the motivations of respondents to over- or under-report (reviewed by Yasué et al., 2010). Comparing multiple sources of local knowledge with scientific observations can help determine which biases are truly problematic (Chapter 3 and broader methods reviewed by Danielsen et al., 2005). Assumptions about biases should also be made explicit in research papers and researchers should experiment with metrics to account for biases and evaluate effects on results. The importance of understanding the perceptions of resource users extends beyond filling data gaps in single-species monitoring. Appropriate application of local knowledge requires a broader understanding of community, beliefs, and the economic context of that knowledge (Neis et al., 1999a; Aswani, 2010). For example, research in Chapter 2 is part of a larger project to evaluate options for seahorse conservation and management action. Project Seahorse has consulted with a range of stakeholders and experts from fishers to scientists about preferred management options (Martin-Smith et al., 2004). Understanding the spatial and behavioral dynamics of fishing (Chapter 4) will be part of a movement beyond managing for single species to a more comprehensive social-ecological system (Berkes et al., 2000). A better understanding of how humans perceive and recall environmental change can help us understand conflicts in conservation and resource governance (Daw, 2010). Conflict can arise when there is a gap between fishers’ perceptions and scientific knowledge used to inform management of fishing (e.g. Gray et al., 2008). From another study conducted in the focal community, Handumon, we know that locals perceive their Marine Protected Area (MPA) to be more successful than scientific data suggest (Yasué et al., 2010). Since local  135  support is a key determinant of ecological success in community-based MPAs, it will be important to acknowledge and then explore this discrepancy in order to ensure long-term viability of the MPA. A broader incorporation of resource user knowledge into IUCN Red Listing could facilitate rapid assessment of the many exploited and potentially threatened species. Assessing threats to species and evaluating their extinction risk are critical steps toward halting declines. The framework for conservation assessments developed by the IUCN has been thoroughly vetted and is considered the global gold standard (Colyvan et al., 1999; de Grammont & Cuaron, 2006; Rodrigues et al., 2006a). Red List assessments are most commonly based on ecological data (McClenachan et al., in review) and/or scientific ‘expert’ opinion. In developing countries with high biodiversity, heavy reliance on wildlife resources, and rapid environmental change, scientific data collection methods are often too time consuming or expensive (Anadón et al., 2009; Rist et al., 2010) and are therefore impractical (Johannes, 1998; Danielsen et al., 2005). With the number of threatened species on the rise and the vast majority (>95%) of described species yet to be assessed (Vié et al., 2009), traditional scientific methods alone are unlikely to be able to inform species assessments. Supplementing the use of scientifically collected data with fisher knowledge could rapidly increase understanding of the status of marine species. Retrospective threat classifications, made possible by incorporating local knowledge, highlight a potential limitation to the Red Listing process (Chapter 2). For marine species that have been affected by fishing for at least hundreds of years (Jackson et al., 2001), assessment of contemporary conservation and fishery status must be conducted relative to historic baselines (Pauly, 1995). Because the default time frame for Red List assessments is  136  only a 10-year window (IUCN, 2008b), there is little capacity for incorporating long-term changes for relatively short-lived species. Therefore, Red Listing may underestimate risk of extinction for historically exploited species (Vincent & Hall, 1996). This shifting baseline in Red Listing should be considered during current efforts to conduct first assessments for marine fishes (e.g. the Global Marine Species Assessment). The nearly universal (across wide uncertainty in key parameters and processes) benefits of deep reef refugia suggested by my study, emphasize the importance of protecting these areas just out of small-scale fishers’ reach. Although we are only beginning to quantify depth refugia on coral reefs, they may actually be widespread (Tyler et al., 2009). If other refugia support population persistence and shallow catches as broadly as in my study, protecting deep (>10 m) reef areas could help safeguard the food supply shallow reefs provide for millions of the world’s most poor (Moberg & Folke, 1999). Furthermore, if refugia in other locations are as large as we estimate they could be on Danajon Bank, protecting them could help reach global marine protected area goals (10% by 2020; CBD, 2006). More research is needed to quantify the contribution that un-fished deep reef areas make to the shallows for other species. My work provides pragmatic methods for identifying refugia and evaluating their local importance. Ultimately, if we are to stem the global decline of fish populations, especially those exploited by small-scale fisheries, we need ‘to be creative and innovative and to consider new ideas and approaches’ (Berkes et al., 2001). Tapping into unconventional data sources like local knowledge, improving understanding of fisher behaviour, and thinking broadly about how to make precautionary decisions despite uncertainty are pragmatic approaches that  137  provide key information highly relevant to the futures of coral reef fishes and small-scale fisheries alike.  138  Literature Cited  139  Abernethy, K. E., Trebilcock, P., Kebede, B., Allison, E. H. & Dulvy, N. K. (2010). Fuelling the decline in UK fishing communities? ICES Journal of Marine Science: Journal du Conseil, 67: 1076. Ainsworth, C. H. & Pitcher, T. J. (2005). Using local ecological knowledge in ecosystem models. In Fisheries Assessment and Management in Data-limited Situations: 289 – 321. Fairbanks: Alaska Sea Grant College Program, University of Alaska. Ainsworth, C. H., Pitcher, T. J. & Rotinsulu, C. (2008). Evidence of fishery depletions and shifting cognitive baselines in Eastern Indonesia. Biological Conservation, 141: 848859. Akamine, J. (2001). Holothurian exploitation in the Philippines: Continuities and discontinuities. Tropics, 10: 591-607. Alcala, A. C. & Russ, G. R. (2002). Status of Philippine Coral Reef Fisheries. Asian Fisheries Science, 15: 177-192. Alder, J. & Sumaila, U. R. (2004). Western Africa: a fish basket of Europe past and present. The Journal of Environment & Development, 13: 156. Allen, S. H. & Sargeant, A. B. (1993). Dispersal Patterns of Red Foxes Relative to Population Density. The Journal of Wildlife Management, 57: 526-533. Alroy, J. (2001). A Multispecies Overkill Simulation of the End-Pleistocene Megafaunal Mass Extinction. Science, 292: 1893-1896. Amar, E. C., Cheong, R. M. T. & Cheong, M. V. T. (1996). Small-scale fisheries of coral reefs and the need for community-based resource management in Malalison Island, Philippines. Fisheries Research, 25: 265-277. Anadón, J. D., Gimenez, A., Ballestar, R. & Perez, I. (2009). Evaluation of Local Ecological Knowledge as a Method for Collecting Extensive Data on Animal Abundance. Conservation Biology, 23: 617-625. Andrew, N. L., Béné, C., Hall, S. J., Allison, E. H., Heck, S. & Ratner, B. D. (2007). Diagnosis and management of small scale fisheries in developing countries. Fish and Fisheries, 8: 227-240. Appeldoorn, R. S., Aguilar-Perera, A., Bouwmeester, B. L. K., Dennis, G. D., Hill, R. L., Merten, W., Recksiek, C. W. & Williams, S. J. (2009). Movement of fishes (Grunts: Haemulidae) across the coral reef seascape: A review of scales, patterns and processes. Caribbean Journal of Science, 45: 304-316. Aswani, S. (2010). How does the accuracy of fisher knowledge affect seahorse conservation status? Animal Conservation, 13: 538. Auster, P. J., Malatesta, R. J., Langton, R. W., Watling, L., Valentine, P. C., Donaldson, C. L. S., Langton, E. W., Shepard, A. N., & Babb, I. G. (1996). The impact of mobile fishing gear on seafloor habitats in the Gulf of Maine (Northwest Atlantic): implications for conservation of fish populations. Reviews in Fisheries Science, 4: 185–202. Bailey, D. M., Collins, M. A., Gordon, J. D. M., Zuur, A. F. & Priede, I. G. (2009). Longterm changes in deep-water fish populations in the northeast Atlantic: a deeper reaching effect of fisheries? Proceedings of the Royal Society B: Biological Sciences, 276: 1965-1969. Bates, D. & Maechler, M. (2009). lme4: Linear mixed-effects models using S4 classes. R package version 0.999375-31. http://CRAN.R-project.org/package=lme4.  140  Baum, J. K., Myers, R. A., Kehler, D. G., Worm, B., Harley, S. J. & Doherty, P. A. (2003). Collapse and Conservation of Shark Populations in the Northwest Atlantic. Science, 299: 389-392. Baum, J. K. & Vincent, A. C. J. (2005). Magnitude and inferred impacts of the seahorse trade in Latin America. Environmental Conservation, 32: 305-319. Begon, M., Harper, J. L. & Townsend, C. R. (1996). Ecology: individuals, populations, and communities. Cambridge: Wiley-Blackwell. Begossi, A. (2008). Local knowledge and training towards management. Environment, Development and Sustainability, 10: 591-603. Bellwood, D., Hughes, T., Folke, C. & Nystroem, M. (2004). Confronting the coral reef crisis. Nature, 429: 827-833. Béné, C. (2006). Small-scale fisheries: Assessing their contribution to rural livelihoods in developing countries. In FAO fisheries circular No. 1008: 46. Berg, B. (2001). Qualitative research methods for the social sciences, 4th edition. Needham Heights, CA.: Allyn and Bacon. Berkes, F. (2003). Alternatives to conventional management: Lessons from small-scale fisheries. Environments, 31: 5-19. Berkes, F., Folke, C. & Colding, J. (2000). Linking social and ecological systems: management practices and social mechanisms for building resilience. Cambridge: Cambridge University Press. Berkes, F., Mahon, R., McConney, P., Pollnac, R. & Pomeroy, R. (2001). Managing Smallscale Fisheries: Alternative Directions and Methods. Ottawa: International Development Research Centre. Bertram, G. C. L. & Bertram, C. K. R. (1973). The modern Sirenia: their distribution and status. Biological Journal of the Linnean Society, 5: 297-338. Bertrand, S., Bertrand, A., Guevara-Carrasco, R. & Gerlotto, F. (2007). Scale-invariant movement of fishermen: the same foraging strategy as natural predators. Ecological Applications, 17: 331-337. Birkeland, C. & Dayton, P. (2005). The importance in fishery management of leaving the big ones. Trends in Ecology & Evolution, 20: 356-358. Bizzarro, J. J., Smith, W. D., Márquez-Farías, J. F., Tyminski, J. & Hueter, R. E. (2009). Temporal variation in the artisanal elasmobranch fishery of Sonora, Mexico. Fisheries Research, 97: 103-117. Bohnsack, J. A. (1998). Application of marine reserves to reef fisheries management. Australian Journal of Ecology, 23: 298-304. Booth, A. J. (2000). Incorporating the spatial component of fisheries data into stock assessment models. ICES J. Mar. Sci., 57: 858-865. Botsford, L., Brumbaugh, D., Grimes, C., Kellner, J., Largier, J., O’Farrell, M., Ralston, S., Soulanille, E. & Wespestad, V. (2009). Connectivity, sustainability, and yield: bridging the gap between conventional fisheries management and marine protected areas. Reviews in Fish Biology and Fisheries, 19: 69-95. Bradburn, N. M., Rips, L. J. & Shevell, S. K. (1987). Answering autobiographical questions: the impact of memory and inference on surveys. Science, 236: 157-161. Branch, T., Hilborn, R., Haynie, A., Fay, G., Flynn, L., Griffiths, J., Marshall, K., Randall, J., Scheuerell, J. & Ward, E. (2006). Fleet dynamics and fishermen behavior: lessons for  141  fisheries managers. Canadian Journal of Fisheries and Aquatic Sciences, 63: 16471668. Brokovich, E., Einbinder, S., Kark, S., Shashar, N. & Kiflawi, M. (2007). A deep nursery for juveniles of the zebra angelfish Genicanthus caudovittatus. Environmental biology of fishes, 80: 1-6. Brook, R. & McLachlan, S. (2008). Trends and prospects for local knowledge in ecological and conservation research and monitoring. Biodiversity and Conservation, 17: 35013512. Buckworth, R. C. (1998). World fisheries are in crisis? We must respond! In Reinventing Fisheries Management: 3-17. Pitcher, T. J., Hart, J. B. & Pauly, D. (Eds.). London: Kluwer Academic Publishers. Bundy, A. & Pauly, D. (2001). Selective harvesting by small-scale fisheries: ecosystem analysis of San Miguel Bay, Philippines. Fisheries Research, 53: 263-281. Burke, L., Reytar, K., Spalding, M. & Perry, A. (2011). Reefs at risk revisited. Washington, D.C.: World Resources Institute. Campbell, S. J. & Pardede, S. T. (2006). Reef fish structure and cascading effects in response to artisanal fishing pressure. Fisheries Research, 79: 75-83. Carpenter, K. E. & Springer, V. G. (2005). The center of the center of marine shore fish biodiversity: the Philippine Islands. Environmental biology of fishes, V72: 467-480. Casey, J. M. & Myers, R. A. (1998). Near extinction of a large, widely distributed fish. Science, 281: 690. Casey, S. P. (1999). Conservation genetics of seahorse (Hippocampus species). PhD Thesis, Queen Mary and Westfield College. University of London. Castilla, J. C. & Defeo, O. (2005). Paradigm shifts needed for world fisheries. Science, 309: 1324-1325. CBD (2006). Decisions Adopted by the Conference of the Parties to the Convention on Biological Diversity at its Eighth Meeting (Decision VIII/15, Annex IV). In Convention on Biological Diversity. Curitiba, Brazil. Charles, A. T. (1998). Beyond the status quo: rethinking fishery management. In Reinventing Fisheries Management: 111-112. Pitcher, T. J., Hart, P. J. B. & Pauly, D. (Eds.). London: Kluwer Academic Publishers. Chuenpagdee, R., Liguori, L., Palomares, M. L. D. & Pauly, D. (2006). Bottom-Up, Global Estimates of Small-Scale Marine Fisheries Catches. In Fisheries Centre Research Reports. Vancouver: UBC Fisheries Centre. Churchill, J.H. (1989). The effect of commercial trawling on sediment resuspension and transport over the Middle Atlantic Bight continental shelf. Continental Shelf Research, 9: 841-864. Cinner, J. E., Sutton, S. G. & Bond, T. G. (2007). Socioeconomic Thresholds That Affect Use of Customary Fisheries Management Tools. Conservation Biology, 21: 16031611. Clapham, P. J., Young, S. B. & Brownell Jr, R. L. (1999). Baleen whales: conservation issues and the status of the most endangered populations. Mammal Review, 29: 37-62. Clark, J. A. & May, R. M. (2002). Taxonomic Bias in Conservation Research. Science, 297: 191-192.  142  Clayton, L., Keeling, M. & Milner-Gulland, E. J. (1997). Bringing Home the Bacon: A Spatial Model of Wild Pig Hunting in Sulawesi, Indonesia. Ecological Applications, 7: 642-652. Colyvan, M., Burgman, M. A., Todd, C. R., Akcakaya, H. R. & Boek, C. B. (1999). The treatment of uncertainty and the structure of the IUCN threatened species categories. Biological Conservation, 89: 245-249. Connell, J. H. (1978). Diversity in tropical rain forests and coral reefs. Science, 199: 13021310. Cooke, S. J. & Cowx, I. G. (2004). The role of recreational fishing in global fish crises. Bioscience, 54: 857-859. Crawley, M. J. (2007). The R book. San Francisco: John Wiley & Sons Inc. DA-BFAR (2004). Department of Agriculture-Bureau of Fisheries and Aquatic Resources. In turbulent seas: The status of Philippine marine fisheries. In Coastal Resource Management Project of the Department of Environment and Natural Resources: 378. Cebu City. Danielsen, F., Burgess, N. & Balmford, A. (2005). Monitoring matters: examining the potential of locally-based approaches. Biodiversity and Conservation, 14: 2507-2542. Davis, A. & Wagner, J. (2003). Who Knows? On the Importance of Identifying “Experts” When Researching Local Ecological Knowledge. Human Ecology, 31: 463-489. Davy, B. (2000). Foreward. In Managing small-scale fisheries: vii - viii. Berkes, F., Mahon, P., McConney, P., Pollnac, R. & Pomeroy, R. (Eds.). Ottawa: IDRC-CRDI. Daw, T. (2010). Shifting baselines and memory illusions: what should we worry about when inferring trends from resource user interviews? Animal Conservation, 13: 534-535. Daw, T. M. (2008a). How fishers count: engaging with fishers' knowledge in fisheries science and management. PhD Thesis, School of Marine Sciences and Technology and School of Geography, Politics and Sociology. Newcastle University. Daw, T. M. (2008b). Spatial distribution of effort by artisanal fishers: Exploring economic factors affecting the lobster fisheries of the Corn Islands, Nicaragua. Fisheries Research, 90: 17-25. de Grammont, P. C. & Cuaron, A. D. (2006). An Evaluation of Threatened Species Categorization Systems Used on the American Continent. Conservation Biology, 20: 14-27. de Groot, S. J. (1984). The impact of bottom trawling on benthic fauna of the North Sea. Ocean Management, 9: 177–190. Deguit, E. T., Smith, R. P., Jatulan, W. P. & White, A. T. (2004). Participatory coastal resource assessment training guide. 0-96. Cebu City, Philippines: Coastal Resource Management Project of the Department of Environment and Natural Resources. Delgado, C. L. (2003). Outlook for fish to 2020: meeting global demand. Washington, D.C.: International Food Policy Research Institute. Diggle, P., Heagerty, P., Liang, K. Y. & Zeger, S. (2002). Analysis of longitudinal data. Oxford: Oxford University Press. D’Onghia, G., Maiorano, P., Capezzuto, F., Carlucci, R., Battista, D., Ciove, A., Sion, L. & Tursi, A. (2009). Further evidences of deep-sea recruitment of Aristeus antennatus (Crustacea: Decapoda) and its role in the population renewal on the exploited bottoms of the Mediterranean. Fisheries Research, 95: 236-245.  143  Dugan, J. E. & Davis, G. E. (1993). Applications of marine refugia to coastal fisheries management. Canadian Journal of Fisheries and Aquatic Sciences, 50: 2029-2042. Dulvy , N. K., Jennings , S., Goodwin , N. B., Grant, A. & Reynolds, J. D. (2005). Comparison of threat and exploitation status in North-East Atlantic marine populations. Journal of Applied Ecology, 42: 883-891. Dulvy, N. K. & Polunin, N. V. C. (2004). Using informal knowledge to infer human-induced rarity of a conspicuous reef fish. Animal Conservation, 7: 365-374. Dulvy, N. K., Sadovy, Y. & Reynolds, J. D. (2003). Extinction vulnerability in marine populations. Fish and Fisheries, 4: 25-64. ESRI (2010). Calculate distance to many points in a shapefile from a single point in another shapefile. http://resources.arcgis.com/content/kbase?fa=articleShow&d=33351. Fagan, W., Meir, E., Prendergast, J., Folarin, A. & Karieva, P. (2001). Characterizing population vulnerability for 758 species. Ecology Letters, 4: 132-138. FAO (1995). Code of conduct for responsible fisheries. Rome, 41pp: Food and Agriculture Organization of the United Nations. FAO (2003). Status and Important Recent Events Concerning International Trade in Fishery Products (Including the World Trade Organization). Committee on Fisheries: SubCommittee on Fish Trade: Ninth Session; Bremen, 10-14 February. FAO (2009). The state of the world's fisheries and aquaculture 2008. Rome: Food and Agriculture Organization of the United Nations, Fisheries and Aquaculture Department. Fernandez, P. R. (2009). The Sea Around the Philippines: Governance and Management for a Complex Coastal Ecosystem. Environment: Science and Policy for Sustainable Development, 51: 36-51. Field, C. B., Behrenfeld, M. J., Randerson, J. T. & Falkowski, P. (1998). Primary production of the biosphere: integrating terrestrial and oceanic components. Science, 281: 237. Foster, S. J. & Vincent, A. C. J. (2004). Life history and ecology of seahorses: implications for conservation and management. Journal of Fish Biology, 65: 1-61. Foster, S. J. & Vincent, A. C. J. (2005). Enhancing Sustainability of the International Trade in Seahorses with a Single Minimum Size Limit. Conservation Biology, 19: 10441050. Gagnon, C. A. & Berteaux, D. (2009). Integrating traditional ecological knowledge and ecological science: a question of scale. Ecology and Society, 14: 19. Gaines, S. D., White, C., Carr, M. H. & Palumbi, S. R. (2010). Designing marine reserve networks for both conservation and fisheries management. Proceedings of the National Academy of Sciences, 107: 18286-18293. Gallardo, W. G., Siar, S. V. & Encena, V. (1995). Exploitation of the window-pane shell Placuna placenta in the Philippines. Biological Conservation, 73: 33-38. Gavin, M. & Anderson, G. (2005). Testing a rapid quantitative ethnobiological technique: First steps towards developing a critical conservation tool. Economic Botany, 59: 112-121. Gaylord, B. Gaines, S. D., Siegel, D.A. & Carr, M. H. (2005). Marine reserves exploit population structure and life history in potentially improving fisheries yields. Ecological Applications, 15: 2180-2191.  144  Gerber, L. R., Botsford, L. W., Hastings, A., Possingham, H. P., Gaines, S. D., Palumbi, S. R. & Andelman, S. (2003). Population models for marine reserve design: a retrospective and prospective synthesis. Ecological Applications, 13: 47-64. Gerber, L. R., Heppell, S. S., Ballantyne, F. & Sala, E. (2005). The role of dispersal and demography in determining the efficacy of marine reserves. Canadian Journal of Fisheries and Aquatic Sciences, 62: 863-871. Gilchrist, G., Mallory, M. & Merkel, F. (2005). Can local ecological knowledge contribute to wildlife management? Case studies of migratory birds. Ecology and Society, 10: 20. Gillis, D. M., Peterman, R. M. & Tyler, A. V. (1993). Movement dynamics in a fishery: application of the ideal free distribution to spatial allocation of effort. Canadian Journal of Fisheries and Aquatic Sciences, 50: 323-333. Godoy, N., Gelcich, S., Vásquez, J. A. & Castilla, J. C. (2010). Spearfishing to depletion: evidence from temperate reef fishes in Chile. Ecological Applications, 20: 15041511. Goetze, J. S., Langlois, T. J., Egli, D. P. & Harvey, E. S. (2011). Evidence of artisanal fishing impacts and depth refuge in assemblages of Fijian reef fish. Coral Reefs, 30: 507-517. Goldbogen, J. A., Calambokidis, J., Croll, D. A., Harvey, J. T., Newton, K. M., Oleson, E. M., Schorr, G. & Shadwick, R. E. (2008). Foraging behavior of humpback whales: kinematic and respiratory patterns suggest a high cost for a lunge. Journal of Experimental Biology, 211: 3712-3719. Goodyear, P. C. (1993). Spawning stock biomass per recruit in fisheries management: foundation and current use. In Risk evaluation and biological reference points for fisheries management. : 67-81. Can. Spec. Pub. Fish. Aquat. Sci. 120. Gotelli, N. J. (2001). A Primer of Ecology. Sunderland, Massachusetts: Sinauer Associates, Inc. Gray, T., Hatchard, J., Daw, T. & Stead, S. (2008). New cod war of words: 'Cod is God' versus 'sod the cod'--Two opposed discourses on the North Sea Cod Recovery Programme. Fisheries Research, 93: 1-7. Green, S. J., Flores, J. O., Dizon-Corrales, J. Q., Martinez, R. T., Nunal, D. R. M., Armada, N. B. & White, A. T. (2004). The fisheries of the Central Visayas, Philippines: Status and trends. Cebu City, Philippines: Coastal Resource Management Project of the Department of Environment and Natural Resources and the Bureau of Fisheries and Aquatic Resource of the Department of Agriculture. Green, S. J., White, A. T., Flores, J. O., Carreon III, M. F. & Sia, A. E. (2003). Philippine fisheries in crisis: A framework for management. 77. Cebu City, Philippines: Coastal Resources Management Project of the Department of Environment and Natural Resources. Gutierrez, N. L., Hilborn, R. & Defeo, O. (2011). Leadership, social capital and incentives promote successful fisheries. Nature, 470: 386-389. Hansen, G. J. A. H., Ban, N. C., Jones, M. L., Kaufman, L., Panes, H. M., Yasué , M. & Vincent, A. C. J. (in press). Hindsight in marine protected area selection: a comparison of community-driven and biophysical approaches. Biological Conservation. Harley, S., Myers, R. & Dunn, A. (2001). Is catch-per-unit-effort proportional to abundance? Canadian Journal of Fisheries and Aquatic Sciences, 58: 1760-1772.  145  Hastings, A. & Botsford, L. W. (1999). Marine reserves and traditional fisheries management. Science, 284: 1537-1538. Heino, M. & Godø, O. R. (2002). Fisheries-induced selection pressures in context of sustainable fisheries. Bulletin of Marine Science, 70: 639-656. Hilborn, R. (1985). Fleet dynamics and individual variation: why some people catch more fish than others. Canadian Journal of Fisheries and Aquatic Sciences, 42: 2-13. Hilborn, R., Stokes, K., Maguire, J., Smith, T., Botsford, L., Mangel, M., Orensanz, J., Parma, A., Rice, J. & Bell, J. (2004). When can marine reserves improve fisheries management? Ocean & Coastal Management, 47: 197-205. Hilborn, R. & Walters, C. J. (1992). Quantitative Fisheries Stock Assessment: Choices, Dynamics and Uncertainties. New York: Chapman and Hall. Hixon, M. A. & Webster, M. S. (2002). Density dependence in marine fishes: coral-reef populations as model systems. In Coral reef fishes: dynamics and diversity in a complex ecosystem. : 303–325. Sale, P. (Ed.). San Diego: Academic Press. Hong, S. K., Rahn, H., Kang, D. H., Song, S. H. & Kang, B. S. (1963). Diving pattern, lung volumes, and alveolar gas of the Korean diving woman (ama). Journal of Applied Physiology, 18: 457-465. Huntington, H. (2000). Using Traditional Ecological Knowledge in Science: Methods and Applications. Ecological Applications, 10: 1270-1274. Huntington, H., Callaghan, T., Fox, S. & Krupnik, I. (2004). Matching Traditional and Scientific Observations to Detect Environmental Change: A Discussion on Arctic Terrestrial Ecosystems. In Ambio Special Report Number 13. The Royal Colloquium: Mountain Areas: A Global Resource: 18-23. Allen Press on behalf of Royal Swedish Academy of Sciences. Hutchings, J. (2000). Collapse and recovery of marine fishes. Nature, 406: 882-885. Hutchings, J. (2001). Conservation biology of marine fishes: perceptions and caveats regarding assignment of extinction risk. Canadian Journal of Fisheries and Aquatic Sciences, 58: 108-121. Hutchings, J. & Reynolds, J. D. (2004). Marine Fish Population Collapses: Consequences for Recovery and Extinction Risk. Bioscience, 54: 297-309. Hutchings, J. A. & Myers, R. A. (1994). What can be learned from the collapse of a renewable resource? Atlantic cod, Gadus morhua, of Newfoundland and Labrador. Canadian Journal of Fisheries and Aquatic Sciences, 51: 2126-2146. IUCN (2001). IUCN Red List Categories and Criteria: Version 3.1. Gland, Switzerland: IUCN. IUCN (2008a). 2008 IUCN Red List of Threatened Species. www.iucnredlist.org. Accessed May 2009. IUCN (2008b). Guidelines for Using the IUCN Red List Categories and Criteria. Version 7.0. Prepared by the Standards and Petitions Working Group of the IUCN SSC Biodiversity Assessments Sub-Committee in August 2008. Downloadable from http://intranet.iucn.org/webfiles/doc/SSC/RedList/RedListGuidelines.pdf. Jackson, J. B. C., Kirby, M. X., Berger, W. H., Bjorndal, K. A., Botsford, L. W., Bourque, B. J., Bradbury, R. H., Cooke, R., Erlandson, J., Estes, J. A., Hughes, T. P., Kidwell, S., Lange, C. B., Lenihan, H. S., Pandolfi, J. M., Peterson, C. H., Steneck, R. S., Tegner, M. J. & Warner, R. R. (2001). Historical Overfishing and the Recent Collapse of Coastal Ecosystems. Science, 293: 629-637. 146  Jacquet, J., Fox, H., Motta, H., Ngusaru, A. & Zeller, D. (2010). Few data but many fish: marine small-scale fisheries catches for Mozambique. African Journal of Marine Science, 32: 197-206. Jacquet, J., Hocevar, J., Lai, S., Majluf, P., Pelletier, N., Pitcher, T., Sala, E., Sumaila, R. & Pauly, D. (2009). Conserving wild fish in a sea of market-based efforts. Oryx, 44: 45 - 56. Jacquet, J. L. & Pauly, D. (2008). Funding Priorities: Big Barriers to Small-Scale Fisheries. Conservation Biology, 22: 832 - 835. Jennings, S., Alvsvåg, J., Cotter, A. J. R., Ehrich, S., Greenstreet, S. P. R., Jarre-Teichmann, A., Mergardt, N., Rijnsdorp, A. D. & Smedstad, O. (1999). Fishing effects in northeast Atlantic shelf seas: patterns in fishing effort, diversity and community structure. III. International trawling effort in the North Sea: an analysis of spatial and temporal trends. Fisheries Research, 40: 125-134. Jennings, S. & Polunin, N. V. C. (1996). Effects of fishing effort and catch rate upon the structure and biomass of Fijian reef fish communities. Journal of Applied Ecology, 33: 400-412. Johannes, R. (1981). Words of the Lagoon: Fishing and Marine Lore in the Palau District of Micronesia. Berkeley: University of California Press. Johannes, R., Freeman, M. & Hamilton, R. (2000). Ignore fishers' knowledge and miss the boat. Fish and Fisheries, 1: 257-271. Johannes, R. E. (1998). The case for data-less marine resource management: examples from tropical nearshore finfisheries. Trends in Ecology & Evolution, 13: 243-246. Jones, G., Almany, G., Russ, G., Sale, P., Steneck, R., van Oppen, M. & Willis, B. (2009). Larval retention and connectivity among populations of corals and reef fishes: history, advances and challenges. Coral Reefs, 28: 307-325. Jones, J. P. G., Andriamarovololona, M. M., Hockley, N., Gibbons, J. M. & Milner-Gulland, E. J. (2008). Testing the use of interviews as a tool for monitoring trends in the harvesting of wild species. Journal of Applied Ecology, 45: 1205-1212. Kahng, S., Garcia-Sais, J., Spalding, H., Brokovich, E., Wagner, D., Weil, E., Hinderstein, L. & Toonen, R. (2010). Community ecology of mesophotic coral reef ecosystems. Coral Reefs, 29: 255-275. Karpov, K. A., Haaker, P. L., Albin, D., Taniguchi, I. K. & Kushner, D. (1998). The red abalone, Haliotis rufescens, in California: Importance of depth refuge to abalone management. Journal of Shellfish Research, 17: 863-870. Kenyon, K. W. (1977). Caribbean monk seal extinct. Journal of Mammology, 58: 97-98. Kinzig, A. (2003). Uncertainty and the Scientist. Ambio, 32: 329. Kinzig, A. & Starrett, D. (2003). Coping with Uncertainty: A Call for a New Science-Policy Forum. Ambio, 32: 330-335. Kuiter, R. H. (2000). Seahorses, Pipefishes and their relatives: a comprehensive guide to Syngnathiformes. Chorleywood: TMC Publishing. Lachica-Aliño, L., Wolff, M. & David, L. (2006). Past and future fisheries modeling approaches in the Philippines. Reviews in Fish Biology and Fisheries, 16: 201-212. Lotze, H. K. & Worm, B. (2009). Historical baselines for large marine animals. Trends in Ecology & Evolution, 24: 254-262.  147  Lourie, S. A., Foster, S. J., Cooper, E. W. T. & Vincent, A. C. J. (2004). A Guide to the Identification of Seahorses. Project Seahorse and TRAFFIC North America. Washington D.C.: University of British Columbia and World Wildlife Fund. Lourie, S. A., Pritchard, J. C., Casey, S. P., Ky, T. S., Hall, H. J. & Vincent, A. C. J. (1999). The taxonomy of Vietnam’s exploited seahorses (family Syngnathidae). Biological Journal of the Linnean Society, 66: 231-256. Lozano-Montes, H. M., Pitcher, T. J. & Haggan, N. (2008). Shifting environmental and cognitive baselines in the upper Gulf of California. Frontiers in Ecology and the Environment, 6: 75-80. Ludwig, D., Hilborn, R. & Walters, C. (1993). Uncertainty, resource exploitation, and conservation: lessons from history. Science, 260: 11-12. Ludwig, D., Mangel, M. & Haddad, B. (2001). Ecology, conservation, and public policy. Annual Review of Ecology and Systematics, 32: 481-517. Luna, C. Z., Silvestre, G. T., Green, S. J., Carreon, M. F. I. & White, A. T. (2004). Profiling the status of Philippine marine fisheries: A general introduction and overview. In In turbulent seas: The status of Philippine marine fisheries. Silvestre, G., Green, S. J., White, A. T., Armada, N., Luna, C., Cruz-Trinidad, A. & Carreon, M. F. I. (Eds.). Cebu City: Coastal Resource Management Project: 3–11. DA-BFAR (Department of Agriculture Bureau of Fisheries and Aquatic Resources). Lunn, K. E. & Dearden, P. (2006). Monitoring small-scale marine fisheries: An example from Thailand’s Ko Chang archipelago. Fisheries Research, 77: 60-71. Mace, G. M. & Hudson, E. J. (1999). Attitudes toward Sustainability and Extinction. Conservation Biology, 13: 242-246. Mackinson, S. & Nottestad, L. (1998). Combining local and scientific knowledge. Reviews in Fish Biology and Fisheries, 8: 481-490. Mahon, R. (1997). Does fisheries science serve the needs of managers of small stocks in developing countries? Canadian Journal of Fisheries and Aquatic Sciences, 54: 2207-2213. Marcus, J. E., Samoilys, M. A., Meeuwig, J. J., Villongco, Z. A. D. & Vincent, A. C. J. (2007). Benthic status of near-shore fishing grounds in the central Philippines and associated seahorse densities. Marine Pollution Bulletin, 54: 1483-1494. Martell, S. J. D., Walters, C. J. & Wallace, S. S. (2000). The use of marine protected areas for conservation of lingcod (Ophiodon elongatus). Bulletin of Marine Science, 66: 729-743. Martin-Smith, K. M., Samoilys, M. A., Meeuwig, J. J. & Vincent, A. C. J. (2004). Collaborative development of management options for an artisanal fishery for seahorses in the central Philippines. Ocean Coast. Manage., 47: 165-193. Matsuda, H., Yahara, T. & Uozumi, Y. (1997). Is tuna critically endangered? Extinction risk of a large and overexploited population. Ecological Research, 12: 345-356. Mauro, F. & Hardison, P. D. (2000). Traditional knowledge of indigenous and local communities: international debate and policy initiatives. Ecological Applications, 10: 1263-1269. McClanahan, T., Castilla, J., White, A. & Defeo, O. (2009). Healing small-scale fisheries by facilitating complex socio-ecological systems. Reviews in Fish Biology and Fisheries, 19: 33-47.  148  McClenachan, L. (2009). Documenting loss of large trophy fish from the Florida Keys with historical photographs. Conservation Biology, 23: 636-643. McClenachan, L. & Cooper, A. B. (2008). Extinction rate, historical population structure and ecological role of the Caribbean monk seal. Proceedings of the Royal Society B: Biological Sciences, 275: 1351-1358. McClenachan, L., Ferretti, F. & Baum, J. K. (in review). From Archives to Conservation: Incorporating Historical Data into Quantitative Ecological Baselines. Conservation Letters. Means, B. & Loftus, E. F. (1991). When personal history repeats itself: decomposing memories for recurring events. Applied Cognitive Psychology, 5: 297-318. Meeuwig, J. J., Samoilys, M. A., Erediano, J. & Hall, H. (2003). Fishers’ perceptions on the seahorse fishery in central Philippines: interactive approaches and an evaluation of results. In Putting Fishers’ Knowledge to Work Conference. Vancouver, B.C., Canada. Messieh, S. N., Rowell, T. W., Peer, D. L. & Cranford, P. J. (1991). The effects of trawling, dredging, and ocean dumping on the Eastern Canadian Continental Shelf seabed. Continental Shelf Research, 11: 1237–1263. Milner-Gulland, E. J. & Rowcliffe, J. M. (2007). Conservation and sustainable use: a handbook of techniques. New York: Oxford University Press. Moberg, F. & Folke, C. (1999). Ecological goods and services of coral reef ecosystems. Ecological economics, 29: 215-233. Moore, J. E., Cox, T. M., Lewison, R. L., Read, A. J., Bjorkland, R., McDonald, S. L., Crowder, L. B., Aruna, E., Ayissi, I., Espeut, P., Joynson-Hicks, C., Pilcher, N., Poonian, C. N. S., Solarin, B. & Kiszka, J. (2010). An interview-based approach to assess marine mammal and sea turtle captures in artisanal fisheries. Biological Conservation, 143: 795-805. Morato, T., Watson, R., Pitcher, T. J. & Pauly, D. (2006). Fishing down the deep. Fish and Fisheries, 7: 24-34. Morgan, S. K. (2007). The ontogenetic ecology and conservation of exploited tropical seahorses. PhD Thesis, McGill University. Morgan, S. K. & Lourie, S. A. (2006). Threatened fishes of the world: Hippocampus comes Cantor 1850 (Syngnathidae). Environmental biology of fishes, 75: 311-313. Morgan, S. K. & Panes, H. M. (2004). Gold rush: the rise and fall of a sea cucumber fishery in the Central Philippines. In World Fisheries Congress. Vancouver, British Columbia. Morgan, S. K. & Vincent, A. C. J. (2007). The ontogeny of habitat associations in the tropical tiger tail seahorse Hippocampus comes Cantor, 1850. Journal of Fish Biology, 71: 701-724. Mori, Y. (1998). Optimal choice of foraging depth in divers. Journal of Zoology, 245: 279283. Munro, J. L. (1996). The scope of tropical reef fisheries and their management. In Reef Fisheries: 1-12. Polunin, N. V. C. & Roberts, C. (Eds.). London: Chapman and Hall. Musick, J. (1999). Criteria to define extinction risk in marine fishes. Fisheries, 24: 6-14. Myers, N. (1995). Environmental Unknowns. Science, 269: 358-360. Myers, R., Hutchings, J. & Barrowman, N. (1997). Why do Fish Stocks Collapse? The Example of Cod in Atlantic Canada. Ecological Applications, 7: 91-106. 149  Myers, R. A., Brown, K. G. & Barrowman, N. J. (1999). The maximum reproductive rate of fish at low population sizes. Canadian Journal of Fisheries and Aquatic Sciences, 56: 2404-2419. Myers, R. A. & Worm, B. (2003). Rapid worldwide depletion of predatory fish communities. Nature, 423: 280-283. Neis, B., Felt, L., Haedrich, R. & Schneider, D. (1999a). An interdisciplinary method for collection and integrating fishers' ecological knowledge into resource management. In Fishing places, fishing people: traditions and issues in Canadian small-scale fisheries: 217-238. Newell, D. & Ommer, R. E. (Eds.). Toronto: University of Toronto Press. Neis, B., Schneider, D., Felt, L., Haedrich, R., Fischer, J. & Hutchings, J. (1999b). Fisheries assessment: what can be learned from interviewing resource users? Canadian Journal of Aquatic and Fisheries Sciences, 56: 1949-1963. Newton, K., Côté, I. M., Pilling, G. M., Jennings, S. & Dulvy, N. K. (2007). Current and future sustainability of island coral reef fisheries. Current Biology, 17: 655-658. Nolet, B. A., Wansink, D. E. H. & Kruuk, H. (1993). Diving of otters (Lutra lutra) in a moare habitat: use of depths by a single-prey loader. Journal of Animal Ecology, 62: 22-32. Novaro, A. J., Redford, K. H. & Bodmer, R. E. (2000). Effect of Hunting in Source-Sink Systems in the Neotropics. Conservation Biology, 14: 713-721. O'Donnell, K. P., Pajaro, M. G. & Vincent, A. C. J. (2010). How does the accuracy of fisher knowledge affect seahorse conservation status? Animal Conservation, 13: 526-533. Orensanz, J. M., Parma, A. M., Jerez, G., Barahona, N., Montecino, M. & Elias, I. (2005). What are the key elements of sustainability of "s-fisheries"? Insights from South America. Bulletin of Marine Science, 76: 527-556. Otero, J., Rocha, F., Gonzáles, Á., Gracia, J. & Guerra, Á. (2005). Modeling artisanal coastal fisheries of Galicia (NW Spain) based on data obtained from fishers: the case of Octopus vulgaris. Scientia Marina, 69: 577-585. Pajaro, M., Vincent, A., Buhat, D. & Perante, N. (1997). The role of seahorse fishers in conservation and management. Proceedings of the 1st International Symposium in Marine Conservation. 118-126. Hong Kong. Pandolfi, J. M., Bradbury, R. H., Sala, E., Hughes, T. P., Bjorndal, K. A., Cooke, R. G., McArdle, D., McClenachan, L., Newman, M. J. H. & Paredes, G. (2003). Global trajectories of the long-term decline of coral reef ecosystems. Science, 301: 955-958. Parma, A. M., Orensanz, J. M., Elias, I. & Jerez, G. (2003). Diving for shellfish and data: incentives for the participation of fishers in the monitoring and management of artisanal fisheries around southern South America. In Towards Sustainability of DataLimited Multi-Sector Fisheries: 186. Newman, S. J., Gaughan, D. J., Jackson, G., Mackie, M. C., Molony, B., St John, J. & Kailola, P. (Eds.). Perth, Western Australia: Fisheries Occasional Publications No. 5 Department of Fisheries. Paterson, B. (2010). Integration fisher knowledge and scientific assessments. Animal Conservation, 13: 536-537. Pauly, D. (1995). Anecdotes and the shifting baseline syndrome of fisheries. Trends in Ecology & Evolution, 10: 430. Pauly, D. (2000). Fisheries in the Philippines and the world: an overview. In Tambuli:a publication for coastal management practitioners. No.6. 150  Pauly, D. (2006). Major trends in small-scale marine fisheries, with emphasis on developing countries, and some implications for the social sciences. Maritime Studies (MAST), 4: 7-22. Pauly, D., Christensen, V., Dalsgaard, J., Froese, R. & Torres, F. J. (1998). Fishing down marine food webs. Science, 279: 860-863. Pauly, D., Christensen, V., Guenette, S., Pitcher, T. J., Sumaila, U. R., Walters, C. J., Watson, R. & Zeller, D. (2002). Towards sustainability in world fisheries. Nature, 418: 689-695. Pauly, D., Watson, R. & Alder, J. (2005). Global trends in world fisheries: impacts on marine ecosystems and food security. Philosophical Transactions of the Royal Society B: Biological Sciences, 360: 5-12. Payton, M. E., Greenstone, M. H. & Schenker, N. (2003). Overlapping confidence intervals or standard error intervals: What do they mean in terms of statistical significance? Journal of Insect Science, 3: 1-6. Perante, N. C., Pajaro, M. G., Meeuwig, J. J. & Vincent, A. C. J. (2002). Biology of a seahorse species, Hippocampus comes in the central Philippines. Journal of Fish Biology, 60: 821-837. Perante, N. C., Vincent, A. C. J. & Pajaro, M. G. (1998). Demographics of the seahorse Hippocampus comes in the central Philippines. In Proceeding of the Third International Conference on the Marine Biology of the South China Sea: 439-448. Hong Kong: Hong Kong University Press. Pet-Soede, C., Van Densen, W. L. T., Hiddink, J. G., Kuyl, S. K. & Michaels, M. A. M. (2001). Can fishermen allocate their fishing effort in space and time on the basis of their catch rates? An example from Spermonde Archipelago, SW Sulawesi, Indonesia. Fisheries Management and Ecology, 8: 15-36. Pet, J. S. & Muljadi, A. H. (2001). Spawning and aggregations of groupers (Serranidae) and Napoleon wrasse (Labridae) in the Komodo National Park. Bali, Indonesia: TNC Coastal and Marine Conservation Center. Pinheiro, J., Bates, D., DebRoy, S., Sarkar, D. & team, R. C. (2009). nlme: Linear and Nonlinear Mixed Effects Models. R package version 3.1-93. Pinnegar, J. & Engelhard, G. (2008). The ‘shifting baseline’ phenomenon: a global perspective. Reviews in Fish Biology and Fisheries, 18: 1-16. Pitcher, T. J., Hart, P. J. B. & Pauly, D. (Eds.) (1998) Reinventing fisheries management, London, Kluwer Academic Publishers. Pitchford, J. W., Codling, E. A. & Psarra, D. (2007). Uncertainty and sustainability in fisheries and the benefit of marine protected areas. Ecological Modelling, 207: 286292. Polidoro, B. A., Livingstone, S. R., Carpenter, K. E., Hutchinson, B., Mast, R. B., Pilcher, N., Sadovy de Mitcheson, Y. & Valenti, S. (2008). Status of the World's Marine Species. In IUCN Redlist Report. Pomeroy, R. E. & Andrew, N. L. (Eds.) (2011) Managing Small Scale Fisheries: Frameworks and Approaches, Wollingford, CABI Publishing. R Core Development Team (2009). R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. ISBN 3-900051-07-0, URL http://www.R-project.org.  151  Rijnsdorp, A. D., Buys, A. M., Storbeck, F. & Visser, E. G. (1998). Micro-scale distribution of beam trawl effort in the southern North Sea between 1993 and 1996 in relation to the trawling frequency of the sea bed and the impact on benthic organisms. ICES Journal of Marine Science, 55: 403. Rist, J., Milner-Gulland, E. J., Cowlishaw, G. & Rowcliffe, M. (2010). Hunter Reporting of Catch per Unit Effort as a Monitoring Tool in a Bushmeat-Harvesting System. Conservation Biology, 24: 489-499. Rittel, H. & Webber, M. (1973). Dilemmas in a general theory of planning. Political Science, 4: 155-169. Roberts, C., McClean, C., Veron, J., Hawkins, J., Allen, G., McAllister, D., Mittermeier, C., Schueler, F., Spalding, M. & Wells, F. (2002). Marine Biodiversity Hotspots and Conservation Priorities for Tropical Reefs. Robinson, J. G. (2006). Conservation Biology and Real-World Conservation. Conservation Biology, 20: 658-669. Rocha, F., Gracia, J., Gonzalez, A. F., Jardon, C. M. & Guerra, A. (2004). Reliability of a model based on a short fishery statistics survey: application to the Northeast Atlantic monkfish fishery. ICES Journal of Marine Science, 61: 25-34. Rochet, M. J. & Trenkel, V. M. (2003). Which community indicators can measure the impact of fishing? A review and proposals. Canadian Journal of Fisheries and Aquatic Sciences, 60: 86-99. Rodrigues, A. S. L., Pilgrim, J. D., Lamoreux, J. F., Hoffman, M. & Brooks, T. M. (2006a). The value of the IUCN Red List for conservation. Trends in Ecology & Evolution, 21: 71-76. Rodrigues, A. S. L., Pilgrim, J. D., Lamoreux, J. F., Hoffmann, M. & Brooks, T. M. (2006b). The value of the IUCN Red List for conservation. Trends in Ecology & Evolution, 21: 71-76. Rose, K. A., Cowan, J. H., Winemiller, K. O., Myers, R. A. & Hilborn, R. (2001). Compensatory density dependence in fish populations: importance, controversy, understanding and prognosis. Fish and Fisheries, 2: 293-327. Rosenberg, A. A., Bolster, W. J., Alexander, K. E., Leavenworth, W. B., Cooper, A. B. & McKenzie, M. G. (2005). The history of ocean resources: modeling cod biomass using historical records. Frontiers in Ecology and the Environment, 3: 78-84. Rowe, S. & Hutchings, J. A. (2003). Mating systems and the conservation of commercially exploited marine fish. Trends in Ecology & Evolution, 18: 567-572. Russ, G. R. (1991). Coral reef fisheries: effects and yields. In The ecology of fishes on coral reefs: 601-635. Sale, P. F. (Ed.). Academic Press, Inc. Russ, G. R., Alcala, A. C., Maypa, A. P., Calumpong, H. P. & White, A. T. (2004). Marine reserve benefits local fisheries. Ecological Applications, 14: 597-606. Ruttenberg, B. I. (2001). Effects of artisanal fishing on marine communities in the Galapagos Islands. Conservation Biology, 15: 1691-1699. Sadovy, Y. (2005). Trouble on the reef: the imperative for managing vulnerable and valuable fisheries. Fish and Fisheries, 6: 167-185. Sadovy, Y. & Cheung, W. (2003). Near extinction of a highly fecund fish: the one that nearly got away. Fish and Fisheries, 4: 86-99. Sadovy, Y., Kulbicki, M., Labrosse, P., Letourneur, Y., Lokani, P. & Donaldson, T. J. (2003). The Humphead Wrasse, Cheilinus undulatus: Synopsis of a Threatened and 152  Poorly Known Giant Coral Reef Fish. Reviews in Fish Biology and Fisheries, 13: 327-364. Sáenz-Arroyo, A., Roberts, C. M., Torre, J. & Cariño-Olvera, M. (2005). Using fishers' anecdotes, naturalists' observations and grey literature to reassess marine species at risk: the case of the Gulf grouper in the Gulf of California, Mexico. Fish & Fisheries, 6: 121-133. Sáenz-Arroyo, A., Roberts, C. M., Torre, J., Cariño-Olvera, M. & Hawkins, J. P. (2006). The value of evidence about past abundance: marine fauna of the Gulf of California through the eyes of 16th to 19th century travellers. Fish & Fisheries, 7: 128-146. Safina, C. (1995). The world's imperiled fish. Scientific American, 273: 46-54. Salas, S. & Gaertner, D. (2004). The behavioural dynamics of fishers: management implications. Fish and Fisheries, 5: 153-167. Sale, P. F., Cowen, R. K., Danilowicz, B. S., Jones, G. P., Kritzer, J. P., Lindeman, K. C., Planes, S., Polunin, N. V. C., Russ, G. R. & Sadovy, Y. J. (2005). Critical science gaps impede use of no-take fishery reserves. Trends in Ecology & Evolution, 20: 7480. Samoilys, M. A., Martin-Smith, K. M., Giles, B. G., Cabrera, B., Anticamara, J. A., Brunio, E. O. & Vincent, A. C. J. (2007). Effectiveness of five small Philippines' coral reef reserves for fish populations depends on site-specific factors, particularly enforcement history. Biological Conservation, 136: 584-601. Scales, H., Balmford, A. & Manica, A. (2007). Impacts of the live reef fish trade on populations of coral reef fish off northern Borneo. Proceedings of the Royal Society B: Biological Sciences, 274: 989. Scholz, A., Bonzon, K., Fujita, R., Benjamin, N., Woodling, N., Black, P. & Steinback, C. (2004). Participatory socioeconomic analysis: drawing on fishermen's knowledge for marine protected area planning in California. Marine Policy, 28: 335-349. Simon, J. & Polunin, N. V. C. (1996). Impacts of Fishing on Tropical Reef Ecosystems. Ambio, 25: 44-49. Stobutzki, I. C., Silvestre, G. T., Abu Talib, A., Krongprom, A., Supongpan, M., Khemakorn, P., Armada, N. & Garces, L. R. (2006). Decline of demersal coastal fisheries resources in three developing Asian countries. Fisheries Research, 78: 130-142. Stuart, S. N., Chanson, J. S., Cox, N. A., Young, B. E., Rodrigues, A. S. L., Fischman, D. L. & Waller, R. W. (2004). Status and Trends of Amphibian Declines and Extinctions Worldwide. Science, 306: 1783-1786. Swartz, W., Sala, E., Tracey, S., Watson, R., Pauly, D. & Sandin, S. A. (2010). The Spatial Expansion and Ecological Footprint of Fisheries (1950 to Present). PLoS ONE, 5: 5995-6000. Tourangeau, R. (2000). Remembering what happened: memory errors and survey reports. In The Science of Self-Report: 29-47. Stone, A. A., Turkkan, J. S., Bachrach, C. A., Jobe, J. B., Kurtzman, H. S. & Cain, V. S. (Eds.). Mahwah: Lawrence Erlbaum Associates. Tyler, E. H. M., Speight, M. R., Henderson, P. & Manica, A. (2009). Evidence for a depth refuge effect in artisanal coral reef fisheries. Biological Conservation, 142: 652-667. UN (1972). Report of the United Nations Conference on the Human Environment. In Sales No. E.73.II.A.14 and corrigendum, chap. I. Stockholm: United Nations.  153  UN (1987). Report of the World Commission on Environment and Development: Our Common Future; Transmitted to the General Assembly as an Annex to document A/42/427 - Development and International Co-operation: Environment; Our Common Future, Chapter 2: Towards Sustainable Development; Paragraph 1"United Nations General Assembly. http://www.un-documents.net/ocf-02.htm. Retrieved 21 October, 2011. UN (1995). Agreement for the implementation of the provisions of the United Nations Convention on the Law of the Sea of 10 December 1982 relating to the conservation and management of straddling fish stocks and highly migratory fish stocks. In United Nations General Assembly. New York, A/RES151/35. Vié, J. C., Hilton-Taylor, C. & Stuart, S. N. (2009). Wildlife in a Changing World: An Analysis of the 2008 IUCN Red List of Threatened Species. World Conservation Union. Vincent, A. C. J. (1996). The International Trade in Seahorses. Cambridge, U.K: TRAFFIC International. Vincent, A. C. J. & Hall, H. J. (1996). The threatened status of marine fishes. Trends in Ecology & Evolution, 11: 360-361. Vincent, A. C. J., Meeuwig, J. J., Pajaro, M. G. & Perante, N. C. (2007). Characterizing an "S"-fishery: seahorses in the central Philippines. Fisheries Research, 86: 207-215. Walmsley, S. F., Medley, P. A. & Howard, C. A. (2005). Participatory Fisheries Stock Assessment (ParFish) Software Manual. London: Marine Resources Assessment Group Ltd. Walsh, W. A., Kleiber, P. & McCracken, M. (2002). Comparison of logbook reports of incidental blue shark catch rates by Hawaii-based longline vessels to fishery observer data by application of a generalized additive model. Fisheries Research, 58: 79-94. Walters, C. (1986). Adaptive Management of Natural Resources. New York: Mac-Millan. Walters, C. (2000). Impacts of dispersal, ecological interactions, and fishing effort dynamics on efficacy of marine protected areas: how large should protected areas be? Bulletin of Marine Science, 66: 745-757. Walters, C., Martell, S. J. D. & Korman, J. (2005). A stochastic approach to stock reduction analysis. Canadian Journal of Aquatic and Fisheries Sciences, 63: 212-223. Walters, C. J. (1998). Designing fisheries management systems that do not depend upon accurate stock assessment. In Reinventing Fisheries Management: 279-288. Pitcher, T. J., Hart, P. J. B. & Pauly, D. (Eds.). London: Kluwer Academic Publishers. Walters, C. J. & Martell, S. J. D. (2004). Fisheries Ecology and Management. Princeton, NJ: Princeton University Press. Watling, L. & Norse, E. A. (1998). Disturbance of the seabed by mobile fishing gear: a comparison to forest clearcutting. Conservation Biology, 12: 1180-1197. White, C. & Kendall, B. E. (2007). A reassessment of equivalence in yield from marine reserves and traditional fisheries management. Oikos,116: 2039-2043. White, J. W., Samhouri, J. F., Stier, A. C., Wormald, C. L., Hamilton, S. L. & Sandin, S. A. (2010). Synthesizing mechanisms of density dependence in reef fishes: behavior, habitat configuration, and observational scale. Ecology, 91: 1949-1961. White, P. C. L., Jenninge, N. V., Renwick, A. R. & Barker, N. H. L. (2005). Questionnaires in ecology: a review of past use and recommendations for best practice. Journal of Applied Ecology, 42: 421-430. 154  Wilen, J. E., Smith, M. D., Lockwood, D. & Botsford, L. W. (2002). Avoiding surprises: incorporating fisherman behavior into management models. Bulletin of Marine Science, 70: 553-575. Worm, B., Barbier, E. B., Beaumont, N., Duffey, J. E., Folk, C., Haplern, B. J., Jackson, J. B. C., Lotze, H. K., Micheli, F., Paulmbi, S. R., Sala, E., Seloke, K. A., Stachowicz, J. J. & Watson, R. (2006). Impacts of Biodiversity loss on ocean ecosystem services. Science, 314: 787-790. Worm, B., Hilborn, R., Baum, J. K., Branch, T. A., Collie, J. S., Costello, C., Fogarty, M. J., Fulton, E. A., Hutchings, J. A. & Jennings, S. (2009). Rebuilding global fisheries. Science, 325: 578 - 585. Yasué, M., Kaufman, L. & Vincent, A. C. J. (2010). Assessing ecological changes in and around marine reserves using community perceptions and biological surveys. Aquatic Conservation: Marine and Freshwater Ecosystems, 20: 407-418. Yasue, M. & Vincent, A. C. J. (in review). Seahorses helped drive creation of marine reserves, so what did the reserves do for the seahorses? Young, I. A. G., Pierce, G. J., Murphy, J., Daly, H. I. & Bailey, N. (2006). Application of the Gomez-Munoz model to estimate catch and effort in squid fisheries in Scotland. Fisheries Research Trends and Assessment of Cephalopod Fisheries. - Proceedings of the CEPHSTOCK Cephalopod Assessment Workshop, 78: 26-38. Zamin, T. J., Baillie, J. E. M., Miller, R. M., RodrÍGuez, J. O. N. P., Ardid, A. N. A. & Collen, B. E. N. (2010). National red listing beyond the 2010 target. Conservation Biology, 24: 1012-1020. Zeller, D., Booth, S., Craig, P. & Pauly, D. (2006a). Reconstruction of coral reef fisheries catches in American Samoa, 1950-2002. Coral Reefs, V25: 144-152. Zeller, D., Booth, S., Davis, G. & Pauly, D. (2007). Re-estimation of small-scale fishery catches for US flag-associated island areas in the western Pacific: the last 50 years. Fishery Bulletin, 105: 266. Zeller, D., Booth, S. & Pauly, D. (2006b). Fisheries Contributions to the Gross Domestic Product: Underestimating Small-scale Fisheries in the Pacific. Marine Resource Economics, 21: 355. Zuur, A. F., Ieno, E. N., Walker, N. J., Saveliev, A. A. & Smith, G. M. (2009). Mixed effects models and extensions in ecology with R. New York: Springer.  155  Appendices  156  Appendix A: Example of a fisher’s ‘personal timeline’ visual aide  Figure A.1. Example of a fisher’s ‘personal timeline’ visual aide used during interviews about the history of seahorse catches in Chapter 3. Fisher name blocked out for anonymity purposes. Small pictures representing events in a respondent’s life were tacked to the timeline at the start of each interview. In this example the fisher was born just before Marcos became president (~ 1962), began fishing on his own in 1977, was married in 1987, and then in the year he was married he also began lantern fishing, and had his 1st child (of 8). The highest number of seahorses he ever caught in one night (120) was in the same year his 4th child was born (1995), also the year Project Seahorse began working in his village and also around the same time as the typhoon ‘Ruping’. At that time he was putting the most effort (paningkamut) into lantern fishing of any time during his career.  157  Appendix B: Additional information about interviews and comparison between focal communities in Chapter 4 Table B.1. Interview schedule for discussions with lantern fishers about diving and fishing limits Biographical information: name, age, number years lantern fishing, number years fishing seahorses. Q1. What are the names of fishing grounds where you catch seahorses? Q2. How deep do you dive in each of these places? Q3. How big is a dupa*? Q4. What is the deepest you can dive? Q5. Do you go that deep while lantern fishing? Q5a. If yes - Do you find seahorses that deep? Q6. What limits how deep you dive while you are fishing? Q7. How deep do you think seahorses live and why? Q8. Do you think there are seahorses down deeper than you fish? *or 'person height' the local unit of depth measurement. We asked this question to be able to convert to meters.  Table B.2 We conducted a Wilcoxon rank-sum test (non-normal data) to evaluate whether there were differences in transit, fishing, or dive parameters between our two focal communities of Handumon and Batasan. Transit  Wilcoxon rank-sum test result W p  Total Time (hr) 15 0.74  Total Distance Traveled (km) 15 0.76  Fishing  Total Time (hr) 32.5 0.02  Total Distance Traveled (km) 24 0.33  Dives Max. Dive Depth (m) 12 0.82  Mean Mean Dive Max. Dive Dive Depth Duration Duration (m) (s) (s) 2.5 8 6 0.08 0.65 0.36  158  Appendix C: Evaluation of the swept-area calculation assumption of randomly distributed fishing effort  Data collection and analysis To evaluate the assumption that fishing effort was randomly distributed in space, we turned to the only data available - fisher logbooks completed for an entire year in Handumon. Logbooks were distributed by Filipino Project Seahorse team members who followed up with discussions and log collection at the end of each month (Vincent et al., 2007). Here, we focused on fishers’ record of the names of fishing grounds they visited. This program ran for many years in Handumon (though not in Batasan), but we chose to analyze data from 2001 because it was the only year when data were recorded in all months. Logbooks were completed by 12 fishers in 2001, representing about 40% of the lantern fishers in Handumon at that time. To place fishing ground names on a map we used two approaches. First, in November 2004 most (27) fishing grounds were GPS located with the aide of an experienced Handumon lantern fisher. If a fishing ground named in fisher logbooks had not been GPS located and was the name of a community or island (4 fishing grounds), we placed the fishing ground location just offshore from the center of that community. We were able to locate a total of 31 of the 50 fishing grounds mentioned in 2001 logbooks accounting for 89% of the total nights fished. We calculated straight-line distance to each fishing ground from a central point in Handumon using an ArcView script (ESRI, 2010) trends. We also calculated the coefficient of dispersion (variance / mean, Begon, Harper & Townsend, 1996) between fishing grounds to categorize the type of distribution.  159  Results  Lantern fishing effort from Handumon in 2001 was not randomly distributed, but aggregated at preferred fishing grounds. Handumon lantern fishers recorded 1,103 nights of fishing during 2001 on grounds that extended up to 12.5 km from Handumon, with the highest concentration of effort seen 3 to 6 km from shore (Fig. S1). The majority of effort was confined to a small proportion of the 31 fishing grounds visited. The five (16%) most visited fishing grounds (Fig. S1, largest black circles) accounted for more than 50% of the total nights fished, nearly 75% to total fishing effort applied. There were on average 34.5 visits per fishing ground with a dispersion coefficient of 55.9, which was significantly higher than 1, therefore, the overall effort allocation aggregated in space.  160  Figure C.1. Distribution of fishing effort (total nights) at each fishing ground visited in 2001 from Handumon lantern fisher logbooks.  Province of Bohol  161  Appendix D: R code to run depth refuge model and supplementary analyses Below are all the equations used in the model presented in Chapter 5, pseudocode, and the R code for running the model over the range of parameter combinations I evaluated. In addition, I present figures that provide evidence suggesting conclusions drawn from the model, even after running it for just ten years, are robust and I discuss possible explanations for the refuge causing lower total (shallow + deep) catches. More detailed definitions of indices and parameters are available in Chaper 5, Table 2. For ease of explanation, equations were presented in a slightly different order and numbered differently in the text than the equations below. The equation numbers in the R code, refer to these equation numbers, not those in the text of Chapter 5.  Equations: (1) S a = e − ( a −1) M  [  ]  L 1 − e − K ( a−a0 ) ; a > a0 (2) La =  ∞ a≤0 0;  1 ; L ≥ 70  (3) Va = 1 + e −a a 0; La < 70  b(− 2.69 + 3.14 × La ) ; La ≥ 93  (4) f a =  2 0; La < 93 35  (5) φ = ∑ f a S a a =1  (6) E′i = φRi′  162  Initialize for month 1for a = 1, K ,35 and i = 1, 2 and continue with no fishing until Nt reaches equilibrium: (7) N1ai = Ri′Sa Repeat for t = 1, K , n : 35  (8) N ti = ∑ N tai a =1 2  (9) N t = ∑ N ti i =1 35  2  (10) Et = ∑∑ f a N tai a =1 i =1 35  (11) Cti = U i ∑ Va N tai a =1 2  (12) Ct = ∑ Cti i =1  (13) Eti = pi Et ;  i = 1, 2  If t < n , for a = 1, K ,35 and i = 1, 2 :  (14) N t +1,ai   ρ (Ri′ / Ei′ )Eti ; a =1  = 1 + [(ρ − 1) / Ei′]Eti −M N  t ,a−1,i e (1 − Va−1U i ); a = 2, K ,35  Assign equilibrium no fishing Ntai to be the first month’s fish in fished simulation which will repeat equations (8) through (14) for ten years (120 months).  163  R Code: ##### R CODE: Set up depth refuge model to run across many different parameter ##### combinations. First dim parameters, then let model run ##### to equilibrium without fishing, then start fishing and let model run for ##### 10 years with fishing. #####LIST OF PARAMETERS TO CALCULATE ACROSS ##### param.options <- expand.grid(a0 = 0.36, b = 1.79, K = 0.24, Linf = 167, M = c(0.06, 0.1, 0.15), n = 120, p = seq(0,1,0.1), Rp1 = c(49063, 98162, 147189),Rp2 = c(147189, 98162, 49063), U1= c(0,0.03,0.04,0.05,0.06), U2= c(0,0.03,0.04,0.05,0.06), rho = c(1.1,1.5,2,5,7),.n=5000) #####  SET UP MODEL TO RUN SIMULATIONS  runSim2 <- function(param.options) {  ##### # dim or set up all variables  ##dim params in param.options a0 <- as.numeric(param.options[1]) # as.numeric () a0 is a single value which describes age at size 0 b <- as.numeric(param.options[2]) # b single value, slope in VB growth curve K <- as.numeric(param.options[3]) # K single value, VB growth parameter Linf <- as.numeric(param.options[4]) # Linf, single value, max size M <- as.numeric(param.options[5]) # as.numeric() M mortality rate, single value n <- as.integer(param.options[6]) # as.integer() n, number of time steps, single value p <- c(as.numeric(param.options[7]), 1-as.numeric(param.options[7])) # p, proportion of recruits going to shallow Rp <- as.numeric(param.options[8:9]) # Rp, vector of 2 values, number of recruits to shallow/deep at #equilibrium U <- as.numeric(param.options[10:11]) # U, exploitation rate (U1) shallow, (U2) deep rho <- as.numeric(param.options[12]) # rho, recruitment compensation parameter, single value .n <- as.numeric(param.options[13]) #n for no fishing equilib simulations ##set up params and seed model with starting population tt <- 1:n #time index and params not already 'dimmed' .tt <- 1:.n #for no fishing equilib simulations N <- array(data = NA, dim = c(n,35,2)) .N <- array(data = NA, dim = c(.n,35,2)) #for no fishing equilib simulations Nt <- rep(0,n) .Nt <- rep(0,.n) #for no fishing equilib simulations Eti <- array(data = NA, dim = c(n,2)) .Eti <- array(data=NA, dim =c(.n,2)) #for no fishing equilib simulations Cti <- Eti .Cti <- .Eti #for no fishing equilib simulations  164  Ct <- Nt; Et <- Nt; Nt1 <- Nt; Nt2 <- Nt; Cti1 <- Nt; Cti2 <-Nt .Ct <- .Nt; .Et <- .Nt; .Nt1 <- .Nt; .Nt2 <- .Nt; .Cti1 <- Nt; .Cti2 <-Nt a <- 1:35 Sa <- exp(-(a-1)*M) #(Eq. 1) these equations numbers refer to equations in the not equations #s in the text of Ch 5. La <- Linf*(1-exp(-K*(a-a0))); La[La<0] <-0 #(Eq. 2) Va <- 1/(1+exp(-a)); Va[La<70] <- 0 #(Eq. 3) fa <- floor(b*(-2.69+3.14*La)/2); fa[La<93] <-0 #(Eq. 4) phi <- (sum(fa*Sa)) #(Eq. 5) Ep <- phi *Rp #(Eq. 6) .N[1,,1] <- Sa * Rp[1] .N[1,,2] <- Sa * Rp[2]  #####  #appendix  #(Eq. 7) #(Eq. 7)  INITIALIZE MODEL BY PREDICTING POPULATIONS SIZE UNTIL COMES TO EQUILIBRIUM WITHOUT FISHING ##### equil.value <- 11 .tt = 1 Continue = TRUE while(Continue){ # tt = 12 .Nt1[.tt] <- sum(.N[.tt,,1], na.rm = TRUE) #(Eq. 8) .Nt2[.tt] <- sum(.N[.tt,,2], na.rm = TRUE) #(Eq. 8) .Nt[.tt] <- .Nt1[.tt] + .Nt2[.tt] #(Eq. 9) .Et[.tt] <- round(sum(fa * .N[.tt,,], na.rm = TRUE)) #(Eq. 10) .Cti1[.tt] <- sum(.N[.tt,,1] * 0 * Va, na.rm = TRUE) #(Eq. 11) .Cti2[.tt] <- sum(.N[.tt,,2] * 0 * Va, na.rm = TRUE) #(Eq. 11) .Ct[.tt] <- .Cti1[.tt] + .Cti2[.tt] #(Eq. 12) for (i in 1:2){ .Eti[.tt,i] <- p[i]*.Et[.tt] }  #(Eq. 13)  if (.tt<.n){ for (i in 1:2){ .N[.tt+1,1,i] <(rho*(Rp[i]/Ep[i])*.Eti[.tt,i])/(1+((rho1)/Ep[i])*.Eti[.tt,i]) #(Eq. 14) Nt.holder <- .N[.tt,-length(a),i]*exp(-M)*(1-Va[ length(a)]*0) #(Eq. 14) .N[.tt+1,-1,i] <- Nt.holder } } if(.tt> equil.value){ if(all(.Nt[.tt]==.Nt[(.tt-equil.value):(.tt-1)]) | .tt >= 5000) Continue = FALSE } .tt <- .tt + 1 } N[1,,1] <- .N[.tt,,1] N[1,,2] <- .N[.tt,,2] #Assign equilib. No fishing numbers at age and zone to be the first month’s fish in fished simulation below ##### START FISHING AND PREDICT POPULATIONS FOR 10 YEARS for (tt in 1:n){ Nt1[tt] <- sum(N[tt,,1], na.rm = TRUE) Nt2[tt] <- sum(N[tt,,2], na.rm = TRUE)  #####  #(Eq. 8) #(Eq. 8)  165  Nt[tt] <- Nt1[tt] + Nt2[tt] Et[tt] <- round(sum(fa * N[tt,,], na.rm = TRUE)) Cti1[tt] <- sum(N[tt,,1] * U[1] * Va, na.rm = TRUE) Cti2[tt] <- sum(N[tt,,2] * U[2] * Va, na.rm = TRUE) Ct[tt] <- Cti1[tt] + Cti2[tt]  #(Eq. #(Eq. #(Eq. #(Eq. #(Eq.  for (i in 1:2){ Eti[tt,i] <- p[i]*Et[tt] }  #(Eq. 13)  if (tt<n){ for (i in 1:2){  9) 10) 11) 11) 12)  # i = 1  N[tt+1,1,i] <-(rho*(Rp[i]/Ep[i])*Eti[tt,i])/(1+((rho1)/Ep[i])*Eti[tt,i]) #(Eq. 14) Nt.holder <- N[tt,-length(a),i]*exp(-M)*(1-Va[ length(a)]*U[i]) #(Eq. 14) N[tt+1,-1,i] <- Nt.holder } } } vals.2.return <- c(M = M, p = p, rho = rho, Rp = Rp, U = U) #return value of params in that model run data.2.return <- c(s.Nt120 = Nt[length(Nt)]/Nt[1], s.Nt120.1 = Nt1[120]/Nt1[1], s.Nt120.2 = Nt2[120]/Nt2[1],cum.Ct120 = sum(Ct), cum.Ct120.1 = sum(Cti1), cum.Ct120.2 = sum(Cti2)) #return proportion of pop remaining and cumulative catches return.these1 <- c(vals.2.return, data.2.return) return.these2 <- cbind(Nt = Nt, Nt1 = Nt1, Nt2 = Nt2,Ct =Ct, Cti1 = Cti1, Cti2 = Cti2) #return time series too return(list(return.these1,return.these2)) }  #####  # End of runSim2  RUN THE MODEL OVER RANGE OF PARAM VALUES  #####  model.results <- list() for(i in 1:nrow(param.options)){ model.results <- c(model.results, runSim2(param.options[i,])) } ##### CALL RESULTS OF THE MODEL AND REFORMAT (ex. Only for param.results so can use it for plotting ##### param.results <- model.results[seq(1, length(model.results), 2)] #pick all the odd ones out of model.results - these are params and data.2.return rem.names <- names(param.results[[1]]) #Remember names param.results.format <- unlist(param.results) names(param.results.format) <- NULL #ixnay the names dim(param.results.format) <- c(length(rem.names), length(param.results)) #get all the same values into the same ROWS param.results.format <- t(param.results.format) #Transpose the table so that the same values are now in Columns #more intuitive colnames(param.results.format) <- rem.names #name the columns param.results.format <- as.data.frame(param.results.format) #Convert to data.frame typeof(param.results.format) #check, should say list head(param.results.format) #check  166  Supplementary Figures  1500000 500000 0  Total population size  2500000  Figure D.1. Trajectory of a population at the combination of parameters that caused the most severe decline for comparison of amount of decline at 10 years (120 months, dashed line) and at equilibrium (758 months).  0  200  400  600  Time (Months)  167  Figure D.2. Proportion of total (solid line), shallow (dashed line), and deep (dotted line) H. comes populations remaining at equilibrium at default parameter values and across a range of uncertainty about the proportion of embryos that disperse to the deep. We compare two scenarios: 1) a small, deep refuge where 25% of the un-fished equilibrium population was assigned to the refuge and 75% were fished at current exploitation rate (a), and 2) no refuge where 100% of the un-fished equilibrium population was fished at current exploitation rate (b).  Deep refuge  No refuge 1.2 0.8  1.0  Total population Fished/shallow population Refuge/deep population  0.6  0.6  0.8  1.0  1.2  b)  VU  0.4  0.4  VU  EN  0.2  0.2  EN  CR 0.0  CR 0.0  Proportion of un-fished population remaining at equilibrium  a)  0.0  0.2  0.4  0.6  0.8  1.0  0.0  0.2  0.4  0.6  0.8  1.0  Proportion of embryos dispersing to refuge / deep  Zone & Scenario  1.5e+07  Total Fished/shallow Refuge/deep  5.0e+06  1.0e+07  Deep refuge No refuge  0.0e+00  Cumulative catch (equilibrium)  2.0e+07  Figure D.3. Comparison of cumulative catches of H. comes at equilibrium from the shallow (dashed), deep (dotted), or total (shallow+deep; solid line) zones. We compare two scenarios: 1) a deep refuge (black lines) and 2) no refuge (grey lines).  0.0  0.2  0.4  0.6  0.8  Proportion of embryos dispersing to refuge / deep  1.0  168  Effect of the refuge on total catches  This study was not designed to evaluate whether establishing a reserve can perform as well as other fishery management tools and increase fishery yield despite a loss of fishing area to the refuge. That said, it is interesting that in contrast to other models of MPA effectiveness that were designed for this purpose, we found that the presence of a refuge decreased total fishery catches (Fig. 5.5), even when the model was run until fishing came to equilibrium with the population (Fig. D.3). Other models similar in structure to ours suggest that when an MPA is established total catches can remain the same (Hastings & Botsford, 1999) or even increase (White & Kendall, 2007), despite fishing area being diminished. The life history of H. comes may have contributed to this difference. For example, it has been suggested that increases in yield using reserves are less likely for short-lived as compared to long-lived species (Gaylord et al., 2005).  169  

Cite

Citation Scheme:

    

Usage Statistics

Country Views Downloads
United States 32 1
China 12 0
Canada 10 0
Philippines 5 5
France 5 0
Mexico 2 0
Spain 2 0
Russia 2 0
Germany 2 3
Japan 2 0
Australia 1 0
United Kingdom 1 0
Chile 1 0
City Views Downloads
Washington 20 0
Unknown 19 8
Shenzhen 8 0
Ashburn 4 0
Burnaby 4 0
Vancouver 3 0
Beijing 3 0
North Vancouver 3 0
Coyoacan 2 0
Madrid 2 0
Kirkland 2 0
Tokyo 2 0
University Park 1 0

{[{ mDataHeader[type] }]} {[{ month[type] }]} {[{ tData[type] }]}
Download Stats

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.24.1-0072414/manifest

Comment

Related Items