UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Modeling the trophic transfer of beta radioactivity in the marine food web of Enewetak atoll, Micronesia Dalsgaard, Anne Johanne Tang 1999-06-09

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

Item Metadata

Download

Media
831-ubc_1999-0024.pdf [ 6.75MB ]
Metadata
JSON: 831-1.0074838.json
JSON-LD: 831-1.0074838-ld.json
RDF/XML (Pretty): 831-1.0074838-rdf.xml
RDF/JSON: 831-1.0074838-rdf.json
Turtle: 831-1.0074838-turtle.txt
N-Triples: 831-1.0074838-rdf-ntriples.txt
Original Record: 831-1.0074838-source.json
Full Text
831-1.0074838-fulltext.txt
Citation
831-1.0074838.ris

Full Text

MODELING THE TROPHIC TRANSFER OF BETA RADIOACTIVITY IN THE MARINE FOOD WEB OF ENEWETAK ATOLL, MICRONESIA by ANNE JOHANNE TANG DALSGAARD B.Sc, The University of Copenhagen, 1995 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE in THE FACULTY OF GRADUATE STUDIES (Department of Resource Management and Environmental Studies; Fisheries Centre) We accept this thesis as conforming to the^Fequ^ired^standard THE UNIVERSITY OF BRITISH COLUMBIA January 1999 © Anne Johanne Tang Dalsgaard, 1999 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of ^JZJQQTCC Manaae/nerrL ctnct £noy'/-o' nmen-cjoU c&u.oLCeS The University of British Columbia Vancouver, Canada DE-6 (2/88) Abstract An approach for modeling the trophic transfer of beta radioactivity within the marine food web of Enewetak Atoll, Micronesia, Central Pacific is described. From 1948 to 1958 this atoll was used by the US military for testing of nuclear weapons while monitoring the impact on the ecosystem. In parallel to these military operations, a marine laboratory was operating on the atoll, hosting a wealth of scientists performing basic research. Probably the most renowned study was carried out by H.T. Odum and E.P. Odum in 1954, who examined the trophic structure of the windward reef community and its productivity per unit area. Based on this study and on the vast amount of scientific literature on the atoll, a mass-balance trophic model of the windward reef was constructed, based on the Ecopath modeling software. Ecopath uses as its basic inputs the biomass, production/biomass, and food consumption rates of the various functional groups in the ecosystem, along with a diet matrix. Based on these inputs it estimates the flow of biomass between the functional groups and presents the corresponding predation mortalities in a matrix where the columns represent the intake of, and the rows the losses of, biomass from the groups. A set of first-order differential equations, relating the intake and loss of biomass to the amounts of radioactivity in the groups, was then set up. The equations were integrated over time and calibrated by minimizing the sum of squared deviations between the observed and predicted levels of radioactivity, thus mapping the transfer of radioactivity onto the transfer of biomass. The original food web / mass-balance model, which was constructed without reference to the data on radioactivity, was subsequently re-calibrated to achieve a match between the food web and the radioactivity data. The results predict that there is a time lag between the observed maximum of radioactivity and the trophic position of the groups, and that beta radioactivity is not bioaccumulated up through the food web. Finally, suggestions on how to incorporate the approach as a general routine into the Ecopath software are given. n Table of Contents Abstract ii Table of Contents iiList of Tables vList of Figures viii Acknowledgments 'x 1. Introduction 1 1.1 General introduction and objectives 1 1.2 Enewetak Atoll, location and description 2 1.3 Atoll formation and Darwin's theory of subsidence 6 1.4 Historical events on Enewetak Atoll 7 2. Background theory 13 2.1 Theory of radioactivity2.1.1 Natural and artificial radioactivity 12.1.2 Radioactive decay : 13 2.1.3 The biological effects of radiation 6 2.1.4 Measuring radioactivity 17 2.1.5 Units of radioactivity 8 2.2 Radioactivity from nuclear explosions 12.2.1 Types of nuclear explosions2.2.2 Decay of mixed fission products 21 2.2.3 Uptake of radioactivity by marine organisms 22.3 Compartment modeling 24 2.3.1 Ecopath2.3.2 Ecoranger 6 3. Methods 28 3.1 Defining the modeled area 23.1.1 Fore reef 30 3.1.2 Algal ridge3.1.3 Reef flat 31 3.1.4 Coral head zone3.1.5 Sand/shingle 3> 3.2 Validating the Ecopath model 33.2.1 Non-fish groups 2 3.2.1.1 Detritus3.2.1.2 Benthic Primary Producers3.2.1.3 Phytoplankton 33 3.2.1.4Zooplankton3.2.1.5 Corals and sea anemones (Class anthozoa) 34 3.2.1.6 Foraminiferans and other protozoansiii 3.2.1.7 Gastropods 35 3.2.1.8 Bivalves3.2.1.9 Shrimps and lobsters 6 3.2.1.10 Stomatopods3.2.1.11 Miscellaneous crustaceans 37 3.2.1.12 Echinoderms - not including holothurians 33.2.1.13 Holothurians 33.2.1.14 Polychaetes and other worm-like invertebrates 38 3.2.1.15 Sessile invertebrates3.2.1.16 Cephalopods3.2.2 Biomass, P/B, and Q/B values of non-fish groups 39 3.2.2.1 Remarks to Table 3.3 33.2.3 Fish groups 43.2.3.1 The distribution and abundance of fish 49 3.3 The origin and incorporation of the radioactivity data 52 3.3.1 The origin of the radioactivity data 53.3.2 Observed trends in radioactivity in various organisms 53 3.3.3 Radioactivity in whole organisms 5 3.3.4 Simulating the observed trends in beta radioactivity 54. Results 59 4.1 Balancing the Ecopath model 54.1.1 First run with Ecoranger using initial input parameters 59 4.1.1.1 Modifying the predation mortality experienced by surgeonfish 60 4.1.1.2 Modifying the predation mortality experienced by shrimps, miscellaneous crustaceans, and gastropods 64.1.2 Second run with Ecoranger using modified input parameters 61 4.2 The fate of beta radioactivity 4 4.2.1 Mapping the fate of beta radioactivity 64.2.2 Simulating the fate of beta radioactivity4.2.3 Re-calibrating the Ecopath model 8 4.2.4 Beta radioactivity and trophic levels 70 4.3 Parameter estimation and network analysis 1 4.3.1 Summary statistics 74.3.2 Transfer efficiencies4.3.3 Mixed trophic impact 2 5. Discussion 74 5.1 Model input parameters 75.1.1 The time span covered by the model 75.1.2 Fish biomass and abundance estimates 5 5.1.2.1 Visual census and rotenone sampling5.1.2.2 Fish biomass estimates at Enewetak Atoll 76 5.1.2.3 Comparing the standing stock of coral reef fish 77 5.1.2.4 The abundance and role of herbivorous fish5.1.2.5 The fish fauna and zooplankton on the fore reef 80 iv 5.2 Outputs of the model 82 5.2.1 The role of benthic primary producers 85.2.2 The role of detritus5.2.3 Comparing with other models 3 5.2.3.1 Trophic transfer efficiencies 85.2.3.2 Biomass by trophic level 4 5.2.3.3 Ecosystem maturity 5 5.3 Simulating radioactivity 86 5.3.1 The re-calibrated Ecopath model 85.3.2 Trophic transfer of radioactivity and the 'food web time lag' 87 5.3.3 Dilution effects and the additional mortality (IvT) 85.3.4 Bio-diminution of beta radioactivity 88 5.3.5 Potentials of the approach 89 References 91 Appendices 104 Appendix 1. Diet matrix of the seventeen non-fish groups included in the model 104 Appendix 2. Diet matrix of the fish species included in the model 105 Appendix 3. Scientific and common names of the fish species included in the model 112 Appendix 4. Deriving the Q/B values of the ten fish groups 117 Appendix 5. Fish biomass estimates of Odum and Odum (1955) 123 Appendix 6. List of the ecosystem maturity attributes, defined by Odum (1969), that are quantified in Ecopath 125 List of Tables Table 1.1. Stratigraphic subdivisions of bore holes drilled at Enewetak Atoll 7 Table 2.1. Naturally occurring radioisotopes in sea water 14 Table 2.2. Radioactive decay from a nuclear detonation 5 Table 2.3. Dates and location of the military nuclear detonations at Enewetak Atoll 19 Table 2.4. Artificial radioisotopes originating from a nuclear explosion 20 Table 2.5. Radioisotopes in marine organisms at Enewetak Atoll 23 Table 3.1. The areal extent of the five zones across the windward reef. 29 Table 3.2. Marine benthic algae at Enewetak Atoll 33 Table 3.3. Summary table of the biomass, P/B, and Q/B values of the non-fish groups 39 Table 3.4. Biomass estimates of detritus 41 Table 3.5. Biomass estimates of benthic primary producers 4Table 3.6. Coral biomass estimates 2 Table 3.7. The biomass of foraminiferans 4Table 3.8. Biomass estimates of gastropods 3 Table 3.9. Biomass estimates of shrimps and lobsters 4Table 3.10. Biomass estimates of crabs and other crustaceans 44 Table 3.11. Biomass estimates of echinoderms 4Table 3.12. Biomass estimates of holothurians 5 Table 3.13. Biomass estimates of polychaetes and other worm-like invertebrates 45 Table 3.14. Rate constants for some holothurians at Enewetak Atoll 48 Table 3.15. Biomass, P/B and Q/B values for infaunal polychaetesTable 3.16. Example of the stomach contents of Neoniphon sammara 50 Table 3.17. Summary table of the biomass, P/B, and Q/B values of the ten fish groups 51 Table 3.18. Sample sizes on which the data on beta radioactivity were derived 54 Table 3.19. Relative weight of the different body parts of fish, bivalves, holothurians, and gastropods 55 Table 4.1. Basic estimates of the 'best model' 62 Table 4.2. Scaling factors generated by Solver 9 vi Table 4.3. 'Additional mortalities' (M) 70 Table 4.4. Summary statistics 2 Table 4.5. Transfer efficiency (%) by trophic level 7Table 5-1. Standing stock of fish on coral reefs in different regions 78 Table 5-2. Trophic transfer efficiencies (%) for four coral reef ecosystem models 84 Table 5-3. Biomass at discrete trophic levels 8vii List of Figures Figure 1-1. Western Pacific and Micronesia 3 Figure 1-2. Enewetak Atoll, Republic of the Marshall Islands 5 Figure 1-3. Time arrow showing the major human events on Enewetak Atoll 9 Figure 1-4. The physiographic zones of the reef in surface and cross section view 10 Figure 2-1. The uranium series 15 Figure 2-2. Processes taking place once radioactive fallout reaches the ocean surface 22 Figure 2-3. Schematic representation of an Ecopath model 27 Figure 3-1. Cross-reef currents and channel currents 8 Figure 3-2. Zonation across the windward reef as defined in the present study 29 Figure 3-3. Total beta radioactivity in corals (Acropora) after the 'Nectar' shot 54 Figure 3-4. Transfer of radioactivity between compartments of an ecosystem 56 Figure 3-5. Radioactivity in the benthic primary producers 57 Figure 4-1. Simplified trophic flow diagram of the windward reef of Enewetak Atoll 63 Figure 4-2. Mapping the fate of radioactivity 65 Figure 4-3. Trends in beta radioactivity in the functional groups 66 Figure 4-4. Modifying the columns or the rows of the predation mortality matrix 68 Figure 4-5. Trophic levels and days required to reach maximum levels of beta radioactivity. 71 Figure 4-6. Maximum level of beta radioactivity and trophic levels 7Figure 4-7. Mixed trophic impact diagram 73 Figure 5-1. Leslie plots 76 Figure 5-2. Biomass pyramids 84 Figure 5-3. Network analysisviii Acknowledgments I would like to express my sincere gratitude to my advisor and mentor, Dr. Daniel Pauly, for his unique guidance and never failing trust. For giving me the opportunity to study at UBC and for his ideas and inputs without which this fruitful project would never have happened. Also a very special thanks to Dr. Carl Walters for his invaluable inputs, to Dr. Less Lavkulich for his tremendously warm and very helpful support throughout my time as a student in RMES, and to Dr. Villy Christensen for helpful discussions and pleasant interactions. I thank J.L. Munro of ICLARM, M. Duke of the University of Washington, and R. E. Foreman of the University of British Columbia for kindly lending me hard to access data and literature. I also wish to thank Dr. RE. Foreman for helpful comments on the manuscript. Thanks to my fellow students and friends at the Fisheries Centre, and to Dr. Tony Pitcher, Ann Tautz, and Ingrid Ross who helped realize my stay at the Fisheries Centre. And finally, thanks to my dear family and close friends for always being there for me. ix 1. Introduction 1.1 General introduction and objectives. In December 1942, when Italian physicist Enrico Fermi produced the first nuclear fission reaction in a secret underground military laboratory in Chicago (Lenssen 1991), artificial radioactivity became an environmental reality. Since the mid 1940s, radioactivity has been released into the marine environment from various anthropogenic sources including nuclear weapons testing, radioactive waste disposal (both civilian and military sources), and effluent from power and fuel reprocessing plants as well as accidental releases (Kennish 1998, Osterberg et al. 1964, Rowan and Rasmussen 1994). Once in the marine environment this radioactivity is of serious human health concern because of its potential to distribute itself throughout diffuse food webs (Clark 1989, Lenssen 1991, Kennish 1998). Essential to understanding the contaminant pathways and ultimate fate of radioactivity in marine ecosystems is the knowledge of trophic relationships (Jarman et al. 1996). However, incomplete or thermodynamically unbalanced food webs have often been used to describe the fate of radioactivity. Indeed, laboratory experiments structured around simplified food chains are probably among the main reasons for contradictory reports concerning the relative importance of transfer within food webs versus direct uptake (adsorption and absorption) of radioactivity by aquatic organisms (Ophel and Judd 1966, Polikarpov 1966, Thomann 1981). Similarly, investigations based on field observations have suffered from difficulties in adequately representing and quantifying the trophic position of the organisms (Kiriluk et al. 1995, Zanden and Rasmussen 1996). This problem has impeded studies from examining the importance and quantifying the effects of trophic transfer and food web dynamics in explaining observed patterns of contaminant bioaccumulation (Kiriluk et al. 1995, Zanden and Rasmussen 1996). Recently, the study of the enrichment of stable isotopes (particularly 515N/514N ratios) through aquatic food webs has shown to be a promising measure of the organism's (fractional) trophic position, taking into account the importance of omnivory and complexity, characteristics of aquatic food webs (Cabana and Rasmussen 1994, Kiriluk et al. 1995, Zanden and Rasmussen 1 1996). It has further been demonstrated that the enrichment of 515N is correlated with the contaminant levels of certain persistent pollutants, suggesting that trophic transfer of contaminants can be significant (Kiriluk et al. 1995, Zanden and Rasmussen 1996). Another approach for determining trophic positions of organisms is through the use of mass-balance food web models constructed with the Ecopath approach and software, initiated by Polovina (1984) and further developed by Christensen and Pauly (1992a, 1995). Recently, Kline and Pauly (1998) examined the relation between trophic positions estimated by 515N enrichment and by Ecopath and found an extremely high correlation (r = 0.986). In this study, the Ecopath approach is taken one step further and its potential for predicting the fate of radioactivity within a marine food web is examined. The study proceeds by mapping the fate of beta radioactivity onto an Ecopath generated food web of the marine ecosystem of Enewetak Atoll, Micronesia, Central Pacific. This mapping involves re-calibration of a preliminary model, initially constructed without reference to the data on radioactivity, and subsequent modification of some of the model inputs, until a match is achieved between the food web and the pollutant data. The dissemination of radioactivity is then simulated, using the trophic fluxes determined from the model. Thus the objectives of this study are: • To simulate the trophic interactions among the functional groups of the marine ecosystem of Enewetak Atoll using the Ecopath modeling software; and • To modify the Ecopath model to simulate the fate of beta radioactivity, originating from a nuclear detonation, within the marine ecosystem of Enewetak Atoll. 1.2 Enewetak Atoll, location and description. Enewetak Atoll belongs to the Republic of the Marshall Islands. This is a young republic formed in 1987 as one of the easternmost states of Micronesia in the West Pacific Ocean (Figure 1.1). The Marshall Islands are situated on two subparallel chains of extinct volcanoes (Henry and Wardlaw 1990), the Ratak and Ralik meaning "Sunrise" and "Sunset", respectively, in the Marshallese language (Karolle 1993). Enewetak itself means "island which points to the east" (Helfrich and Ray 1987), and is situated on the northwestern extreme of the Ralik Chain at 11°30'N latitude and 162°15'E longitude. There has been some confusion 2 regarding place names of the atoll. Prior to 1973 the English spelling of Enewetak was 'Eniwetok' (Anon. 1979), but with the gradual recognition of the native people and the movement of the Marshall Islands towards independence (see Figure 1.3), the spelling was changed to acknowledge the native pronunciation (Anon. 1979). In the case of island names, up to four different spellings might be found (Dawson 1957) as both native names and the English spelling has changed over time. On top of this are the military code names that were assigned to the islands when the atoll was used for nuclear testing (Henry and Wardlaw 1990). The atoll is oval-shaped and dominated by a 40 km long north-south by 32 km wide east-west oriented lagoon with a mean depth of 48 m and a maximum depth of 64 m (Atkinson et al. 1981). Surrounding the lagoon is a necklace of small islands and submerged coral reefs (Figure 1.2). "Surprisingly, it is difficult to determine the exact number of islands. Due to the effects of storms, small islands are ephemeral, and two islands and part of a third were obliterated by nuclear explosions. Currently [1987] there are 39 recognizable islands ..." (Reese 1987). The atoll is situated within the belt of the northeast trade winds and the North Equatorial Current which moves westward at a speed of 20 - 50 cms"1. For nine months of the year the windward side must therefore withstand a constant wave attack (Ladd 1973, Atkinson 1987), which in turn brings about a fresh supply of oxygen, nutrients and food to the aquatic ecosystem. The windward side is thus the most vital part of the reef, concentrating living organisms and inducing active reef building (Ladd 1961, 1973). The islands consist of reef debris that is formed on the reef front and piled up by the currents, waves, and winds (Ladd 1973). Hence, the majority of the islands are concentrated on the windward side comprising the northeastern and eastern reef perimeter. The remainder of the reef may be divided into three parts with distinct morphologies related to their position relative to the prevailing winds. These are; the leeward reef on the southwest, a transitional reef on the northwest, and a transitional reef on the southeast (Ristvet 1987) (Figure 1.2). The maximum elevation above sea level of any island is approximately 4 meter. The total land area is 6.5 km2 while the lagoon covers an area of 932 km2 (Atkinson et al. 1981). Because of the low elevation and little land mass, the weather conditions on the atoll are dictated by the surrounding ocean (Colin 1987b). The air temperatures range from 28.5°C in the dry season to 30.0°C in the wet season. The wet season stretches from April to mid-November and 4 162° 162°10' 162°20' 1— 162°30' I I I I I I I I I I I I I I I I I 1 NAUTICAL MILES FEET KILOMETERS I  I Figure 1-2. Enewetak Atoll, Republic of the Marshall Islands, with locations relevant for the present study. Modified from Wardlaw et al. (1991). within this period, the atoll receives about 85% of an average yearly rainfall of 1470 mm (Merrill and Duce 1987). The atoll is traversed by three channels; the East Deep Channel which is 1.5 km wide and 55 m deep, the South Passage which is 9.7 km wide and 11-22 m deep, and the 4.2 km wide and 2 m deep Southwest Passage consisting of a network of small passages rather than a single channel (Atkinson et al. 1981) (Figure 1.2). 5 Waves breaking on the windward reef constantly drive water across the reef flat into the lagoon making this surf-driven inflow the major water input. The speed of the cross-reef current depends on the height of the surf and the tide and ranges between 10 to 150 cms"1 (Atkinson et al. 1981, Atkinson 1987). The tides are semi-diurnal with a maximum range of approximately 1.8 m (Wells and Jenkins 1988). Water also enters and leaves the lagoon across the leeward sections and through the three channels. The South Passage is the main exit while the currents in the Deep Channel and the Southwest Passage reverse with the tide (every 6.2 hours) (Atkinson 1987). Three current levels exist in the lagoon which can be distinguished by their speed and direction. At the top is a wind-driven surface current, 5-15 m thick, moving southwest at a speed of 10 cm-s"1 (Wardlaw et al. 1991). Below this, at 10 to 30 m depth, is a mid-depth current flowing northeast at a speed of 2 to 4 cms"1. Finally, at 30 to 50 m depth, a deep current flows southward at a speed of 0.5 to 1.5 cm-s"1. Despite this three-layer circulation system, the water in the lagoon is well mixed with an average salinity of 34.4 %o and an average temperature of 27-29°C (Wardlaw et al. 1991). The average residence time of the water in the lagoon is 33 days but may be up to four times longer for water entering across the northern perimeter of the atoll and somewhat shorter for water entering across the southern perimeter (Atkinson 1987). 1.3 Atoll formation andDarwins theory of subsidence. During his voyage with the H.M.S. Beagle (1831 - 1836) Charles Darwin conceived his theory of reef formation. He had noticed the existence of three basic types of coral reefs: fringing reefs along the shoreline, barrier reefs separated from land by a wide lagoon, and atolls which are reefs encircling a central lagoon (Blanchon 1997). Atolls, he hypothesized, are the last step in a geological process of a subsiding volcanic islands (Lalli and Parsons 1994). As a volcanic island fringed by coral reefs slowly sinks (a process that may happen when a 'newly' formed volcano presses down on a thin oceanic plate) the coral organisms grow upwards toward the light and outwards toward the fresh supply of oxygen, nutrients, and food. If the coral organisms are successful in keeping up with the speed of subsidence, a barrier reef will gradually form as the corals closest to the island suffocate in the debris formed on the front 6 reef. As the subsidence continues the central island will eventually disappear leaving a lagoon with a perimeter of coral reefs (an atoll) (Maragos et al. 1996). In 1951, Darwin's theory was confirmed when two holes were drilled at Enewetak Atoll, penetrating through the limestone cap and reaching the volcanic rock basement at depths of 1267 m and 1405 m, respectively (see Table 1.1). "The limestones recovered were all of shallow water origin demonstrating both subsidence of the atoll and the upward growth of shallow water corals since Eocene time, approximately 49 million years B.P ..." (Grigg 1982). The limestone was characterized by thick intervals of unleached aragonite-rich carbonate sediment with well preserved aragonitic fossils alternating with layers where the aragonite had been dissolved. The latter represent periods when the atoll stood above water level and was subjected to the local weather conditions (Ladd 1973, Henry and Wardlaw 1990). Table 1.1. Stratigraphic subdivisions recognized in holes drilled at Enewetak Atoll. Modified from Ladd (1973). Stratigraphic divisions Depth (m) Post-Miocene 0 -200 Upper Miocene 200 -300 Lower Miocene 300 -900 Upper Eocene 900 - 1400 1.4 Historical events on Enewetak Atoll. Enewetak Atoll is no doubt one of the most intensively studied and most abused atolls in the world. It was a major battle site at the end of World War II between the U.S. armed forces and Japan who had occupied the Marshall Islands (and the majority of Micronesia) from 1914 (see Figure 1.3). The battle damage was augmented with the grounding of a fully loaded oil tanker on the windward reef during the American invasion, possibly causing the death of long sections of the surf zone (Emery et al. 1954, Ladd 1973). While Japan's interests in Micronesia had been mainly economic, the American interests were purely strategic. The former Japanese areas were placed under administration by the United Nations and in 1947 a Trusteeship Agreement was signed by President H.S. Truman establishing the Trust Territory of the Pacific Islands (Karolle 1993). In the same year, the native people of Enewetak Atoll were removed, as officials in Washington D.C. announced that the atoll was going to be used for nuclear testing (Van Dyke 1991). 7 From 1948 to 1958, the atoll was part of the U.S. Pacific Proving Grounds and test site for 43 nuclear detonations (Figure 1.3 and Table 2.3). The largest of these was the Mike test in 1952 (part of'Operation Ivy', Figure 1.3 and Table 2.3). This was the first hydrogen devise ever to be tested, and the blast was estimated at 10.4 megatons or 750 times the Hiroshima bomb, resulting in the vaporization of an island (Anon. 1998). The testing of nuclear devises damaged or destroyed the vegetation on all but two of Enewetak's islands (Ladd 1973). The military activities on the atoll also included the construction of buildings, runways, and causeways, the latter connecting islets to facilitate transportation. "These structures barred cross-reef circulation of ocean water in certain areas and radically changed ecological conditions on the reef and in parts of the lagoon" (Ladd 1973). Enewetak also received fallout from the nuclear testing on Bikini Atoll. This atoll, located upwind and up-current from Enewetak (see Figure 1.1), was part of the U.S. Pacific Proving Grounds as well, and was the first atoll used for nuclear testing (Operation Crossroads in 1946, comprising two tests). Meanwhile, between tests, the atoll was studied intensively by biologists, geologists, oceanographers, and geophysicists (Ladd 1973). One of the most remarkable, and probably the most cited study on the atoll, was conducted by H.T. Odum and E.P. Odum in 1954 (Odum and Odum 1955). Since this study also serves as an important background for the present study, it is here briefly summarized. It was conducted along a transect on a typical inter-island reef (Japtan, see Figure 1.2) on the windward side of the atoll in an area which, at that time, was yet not seriously affected by the nuclear tests. The objectives of the study were twofold. The first objective was to determine the relationship between the standing stock of organisms and their productivity per area. This would give the investigators a rough idea of the proportionality between the productivity of the coral reef community and its standing crop (i.e., turnover rate, or production/biomass ratio). The second objective was to provide for a reference point that would "aid future comparisons between the normal and the irradiated reef 8 'Nuclear events' Historical events 1990 1987: The Republic of the Marshall Islands is formed. 1983: MPRL is closed down. 1 QSfl 1^80: Cleanup 'complete' - Enewetak officially IVoU returned to the people of Enewetak. MPML chan ges to Mid-Pacific Research Laboratory (MPRL). 1977: The US Defense Nuclear Agency (DNA) begins the cleanup of Enewetak. 1974: EMBL is renamed 'Mid-Pacific Marine 1970 Laboratory' (MPML). 1958: Operation Hardtack; 22 detonations. 1956: Operation Redwing; 11 detonations. 1954: Operation Castle; 1 detonation. 1952: Operation Ivy; 2 detonations. 1951: Operation Greenhouse; 4 detonations. 1948: Operation Sandstorm; 3 detonations._ 1960 _ 1950 < < 1940 } 1963: Limited Test-Ban Treaty: All future testing should be underground. 1958: Joint moratorium on atmospheric testing of nuclear weapons by the US and Soviet (violated in 1961 by the Soviet and by the US in 1962). 1954: Establishment of Enewetak Marine Biolo gical Laboratory (EMBL). 1948: The Pacific Proving Grounds are formed. 1947: Officials in Washington DC announces that Enewetak will be used for nuclear testing. The native people removed to Ujelang Atoll. 1944: American landing. 1914 -1944: The Marshall Islands part of the Pacific Territories controlled by Japan. 1900 1890s The Marshall Islands declared a German protectorate. 1790s Europeans 're-visit' Enewetak. 1529: Enewetak 'discovered' by the Spanish explorer Alvaro de Saavedra. Time of Christ. No archeological research has been conducted at Enewetak, but other area in the Marshall Islands were occupied at this time. According to the Enewetakese they "were there from the beginning" (Kiste, 1987). Figure 1-3. Time arrow showing the major human events (civil and military) on Enewetak Atoll (Helfrich and Ray 1987, Kiste 1987, Wells and Jenkins 1988, Karolle 1993). 9 ecosystem... Since nuclear explosion tests are being conducted in the vicinity of these inherently stable reef communities, a unique opportunity is provided for critical assays of the effects of radiations due to fission products on whole populations and entire ecological systems in the field" (Odum and Odum 1955). The transect was divided into six zones, as illustrated in Figure 1.4: the windward buttress zone, coral-algal ridge, encrusting zone, zone of small coral heads, zone of large coral heads, and zone of sand and shingle. LARGE SMALL ENCRUSTING BUTTRES SAND-SHINGLE HEADS HEADS ZONE RIDGE Z0NE 1 1 1 r—i LOW TIDE 100 M M/SEC ' .11 .16 • •32 •36 AVERAGE TIDE •40 ...39 .01 .02^S***^ LOW SPRING TIDE .09 .15 .37 .26 Figure 1-4. Diagram showing the physiographic zones of the reef in surface and cross section view, and the average current velocities in m/sec. The approximate location of the 6 quadrats is indicated in the upper diagram. Redrawn from Odum and Odum (1955). During a 6 week period the organisms within each zone, ranging from zooxanthellae within coral polyps to sharks in pelagic waters, were described and grouped into discrete trophic levels comprising primary producers, herbivores, carnivores, and decomposers. A rough estimate of the biomass per area for each group was obtained using a variety of methods. The 10 ambitiousness of the study and the limited time available meant that "fewer replications were made than would be required to obtain maximum accuracy from each method. Therefore it is the orders of magnitude which emerge, but care is taken to base conclusions only on large, probably significant differences" (Odum and Odum 1955). The results of the groupings and quantification was presented as biomass pyramids, one for each zone. Despite the different taxonomic compositions in the six zones and the various errors accompanying the crude biomass estimates, the general shape of the pyramids was the same and the ratio of standing crop between trophic levels was estimated as: herbivores/primary producers 18.9%; carnivores/herbivores 8.3% (see Figure 5.2). The study moreover showed that the reef was highly productive compared to other systems (with more than 24 g glucose-m^-day"1 or 8760 g glucose-m"2-year_1 (gross primary production)). On a yearly basis the production of the reef seemed to match the respiration, indicating that the community was at ecological climax. Given a standing stock of about 850 g dry weight-m"2, the production/biomass ratio of the reef was approximately 12.5 year"1. From 1954 to 1986, 1028 scientists visited Enewetak, many returning several times to follow up on their field work (Helfrich and Ray 1987). The scientists were stationed at Eniwetok Marine Biological Laboratory (EMBL) which was established in 1954 and managed by the University of Hawaii. The laboratory was initially sponsored by the U.S. Atomic Energy Commission (AEC) and later by the U.S. Department of Energy (DOE) (Helfrich and Ray 1987, Wells and Jenkins 1988). It was run part-time until 1974 with the research focusing on increasing the knowledge of the atoll ecosystem. In 1974 it was upgraded to a full-time laboratory and renamed the Mid-Pacific Marine Laboratory (MPML). The focus of the research changed to include mainly lagoon oceanography, groundwater dynamics, and ciguatera fish poisoning. In 1979 it was decided to phase down the laboratory. A major cleanup / rehabilitation program, initiated in 1977 by the U.S. Defense Nuclear Agency (DNA) with the objective of preparing the atoll for the return of its native people, was about to finish, and funding for the laboratory was running low. The name was changed for the second time to the Mid-Pacific Research Laboratory (MPRL) to reflect that the research no longer was confined to the marine environment. In 1982 the laboratory was finally terminated (Helfrich and Ray 1987). 11 One of the results of the research on the atoll is a collection of more than 200 reprints of scientific publications. The collection was issued as three volumes in 1976 and a fourth volume in 1979 (Anon. 1976a-c, Anon. 1979)1. Furthermore, in 1987, two volumes synthesizing the research of the laboratory's entire history was published (Devaney et al. 1987, Helfrich and Ray 1987). These two publications together with the study by Odum and Odum (1955) and recently declassified material on the level of beta radioactivity in the biota of the atoll2 form the bulk of material upon which this study is based. 1 The four volumes (Anon 1976a-c; Anon. 1979), which are hard to access, were kindly made available by J.L. Munroof ICLARM. 2 Kindly provided by Marcus Duke from the University of Washington, Seattle. The material consisted of reports prepared by the Laboratory of Radiation Biology, University of Washington, in contract with the United States Atomic Energy Commission. 12 2. Background theory The following chapter provides the background theory for the thesis, and is divided into three main parts. The first part deals with the theory of radioactivity, describing the processes of radioactive decay and the effects of radioactivity on living organisms. The second part deals with the different types of nuclear explosions at Enewetak Atoll, the decay of the radioactive material produced by these tests, and the uptake of radioisotopes by marine organisms. Finally, the Ecopath compartment modeling software, which allows for the construction, analysis, and comparison of mass-balance trophic models, is described. 2.1 Theory of radioactivity. 2.1.1 Natural and artificial radioactivity. Radioactivity has always been a natural component of the environment. When the Earth was formed, much of its constituent matter was radioactive. Over time, this radioactive material has decayed, leaving behind only the isotopes with the longest half life3 and their decay products. Cosmic radiation is another source of natural radioactivity that continuously adds small quantities of 14Carbon, Tritium, and other radioisotopes to the upper atmosphere (Table 2.1). From here they reach the Earth's surface as part of the rain (Mauchline and Templeton 1964, Bablet and Perrault 1987b, Smith 1994). Radioactivity from various anthropogenetic activities such as from nuclear testing and nuclear waste is, on the other hand, an artificial source of radioactivity that has added more and new radioisotopes to the environment (Seymore 1960). 2.1.2 Radioactive decay. Radioactivity is the spontaneous emission of excess energy from atoms. Atoms consist of a nucleus orbited by electrons. The nucleus contains protons and neutrons, and while the number of protons are fixed for every element in the periodical system, the number of neutrons may vary. Atoms with the same number of protons but a different number of neutrons are known as isotopes. Radioisotopes are isotopes in which the ratio of neutrons to 3 The time it takes for half of the radioisotopes to decay. See also section 2.1.2 on radioactive decay. 13 protons makes the nucleus unstable. To become stable the nucleus must give off energy, a process which is known as radioactive decay. Table 2.1. Naturally occurring radioisotopes in sea water. Modified from Seymore (1960) and Bablet and Perrault (1987b). Radioisotope Half life3 Amount Activity (years) (g-1"1) (Bq-l')b Terrestrial origin: 87Rubidium 4.70-10'° 3.4-10-5 0.2 232Thorium 1.42-1010 1.0-1010 <0.1 238Uranium 4.50-109 3.0-10s 0.1 ^Potassium 1.25-109 4.7-IO"5 12.3 235Uranium 7.13-108 2.1-10"8 <0.1 234Uranium 2.48-105 1.9-10''° -230Thorium 7.52-104 <3.0-1013 -23'Protactinium 3.43-104 2.0-10'12 -226Radium 16.22-102 1.0-1013 <0.1 227Actinium 21.60-10° <1.0-1015 -2'°Lead 19.40-10° 1.1-10"'5 -228Radium 6.70-10° 1.4-10"'6 -210Polonium 0.38-10° 2.2-IO-'7 -234Thorium 0.07-10° 4.3-10"'7 -Cosmic origin: 14Carbon 55.70-102 (2to3)-10-14 <0.1 Tritium 12.26-10° 1.7-10"'8 <0.1 a. The radioactive 'half live' is the time it takes for 50% of the radioisotopes to decay; b. Becquerel-litef1 (becquerel (Bq) = disintegrations per second). Dashes indicate that the activity in sea water is insignificant. Three types of radiation products are typically generated in a nuclear detonation: fissile non-fission products, fission products, and activation products. Fissile non-fission products consist of Plutonium and Uranium atoms that did not undergo fission, i.e., did not split apart in the detonation. They are essentially alpha emitters (see Table 2.2). Fission products, on the other hand, are the radioisotopes that are formed when Uranium and Plutonium atoms do undergo fission. Activation products are formed when elements from the surrounding environment, or from the nuclear bomb itself, captures neutrons produced in the detonation. Fission and activation products mainly give off beta and gamma radiation (see Table 2.2) (Bablet and Perrault 1987b). 14 Table 2.2. Types of radioactive decay from a nuclear detonation and the processes that lead to them. Based on Smith (1994) and Skarsgard (1997). Radioactivity Process Alpha Beta Gamma Neutrons Emission of a Helium particle (2 He) from the nucleus: 2^U->23940Th+^He +energy Emission of an ordinary electron (negatron) as a neutron transforms into a proton: 2° Co—^gNi* + negatron + uncharg ed particle or the emission of an anti-electron (positron) as a proton transforms into a neutron: 22Na—»22,Ne + positron + uncharg ed particle The result of beta decay is that the atomic number changes, and a new element is formed. Emission of photons from a nucleus going from a high energy excited state to a low energy stable state, e.g., the Ni atom produced in the negatron emission example above, is marked by a star to indicate that it is in an excited state, and will emit gamma rays (y): 28 Ni* -> 2y Formed during fission of Uranium and Plutonium. When the atoms split, excess neutrons are emitted that in turn may ionize other atoms. 90 85 80 234|J 2.45x1_q'y ^ / 230Tn 8.0x10*1 y 234Th 24.1 d 226 R a 210pQ "* .. ^ r 3.05 mir 2i°Bi / 214Bi /, I£bw f^'^ 2122P3y V Z™™ 214Pb >C r . ' r 26.8 min 2180At omv. 222Rn 3.82 d / J a Decay N 206J| 4.20 min 206Hg 8.1 min —i A 210T| / 2x 10*% 1 .3 t p Decay Denotes Major Branch 125 130 135 N 140 145 Figure 2-1. The uranium series. Redrawn and modified from Skarsgaard (1997). N is the number of neutrons in the nucleus, while Z is the atomic number equal to the number of protons in the nucleus. 15 Decay product are themselves often radioactive, forming so-called decay chains that ends when a stable element is formed. A typical example is the Uranium series shown in Figure 2.1. 2.1.3 The biological effects of radiation. As neutrons, photons, alpha and beta particles travel through living tissue, they interact with its constituent atoms. During this interaction, energy is transferred from the particles to the incident atoms, in most cases to the orbiting electrons. If the gain in energy is large enough, the electrons will be knocked away from the atoms, i.e., the atoms are ionized. If not enough energy is transferred the result will be an excitation of the atom. Alpha particles are too large to penetrate the epidermis and do not constitute any serious external risk. Once inside the organism, however, they are extremely hazardous. Their size and low velocity means that for a given level of energy, they are more likely to cause ionization and excitation than, for example, beta particles. Alpha particles are thus characterized by a rapid loss of energy and dense ionization of the nearby tissue. Beta particles, on the other hand, are smaller, faster, and travel further before they lose their energy, and are more penetrating than alpha particles. Besides internal damage, they can cause severe external damage to the lenses of the eye and to the epidermis. Photons (gamma-rays), depending on their energy, may either cause excitation of atoms, knock away loosely bound electrons, or convert into matter in the form of a positron and a negatron that, in turn, may ionize surrounding .atoms. Gamma-rays are the most penetrating type of radiation and may cause severe harm to the whole body. Ionized and excited atoms / molecules are thus the principal result of absorbed radiation in organic tissues. This especially applies for water molecules which constitute the greater part of the tissue. Here, the most important process is generally considered to be: The ionized water molecule just formed reacts with another water molecule to give a hydroxyl radical (OH*): H20 ionization >H20++e (2.1) H2O++H2OH>H;O+OH* (2.2) 16 The electron formed in (2.1) readily reacts with water molecules and hydrogen ions to form hydrogen radicals (FT): e~+H20->H2CT -*OrT+H' (2.3) e-+H+H»H* (2.4Radicals are extremely reactive. Hydroxyl radicals are strong oxidizing agents and hydrogen radicals and electrons are strong reducing agents. Besides the radiation induced processes that affects the water molecules, and hence indirectly affects the other molecules in the tissue, direct alterations of molecules other than water also occur. An example is the rupture of the phosphate sugar backbone in DNA molecules, but many other alternations may also take place (Kinne 1984, Smith 1994, Skarsgard 1997). A single or a few radiation events do not pose any risk to the health of the organism as long as its normal repair mechanisms can keep pace. It is the cumulative effect of cellular damage that makes radioactivity dangerous. Dividing cells are particular prone to radiation damage and so fetuses are more susceptible than fully developed organisms. Besides burns and skin ulcers, radiation may cause cancer, degenerative diseases, mental retardation, chromosome aberrations, and genetic disorders (Lenssen 1991). 2.1.4 Measuring radioactivity. The radioisotopes originating from a nuclear detonation do not decay immediately, but have a characteristic probability of decay per unit of time. For a large population of similar radioisotopes, the number that decay per unit time (the activity) can be described as the product of the number of nuclei (N) and the decay constant (A,) (Kinne 1984): dN/dt = -A.-N (2.5) Integrating this equation results in a negative exponential function of the form: N = N0-e-Xt (2.6where N0 is the number of radioisotopes present at time zero. Besides X, radioactive decay is often measured by the physical half life (t1/2), which is the time it takes for half of the radioisotopes to decay: 17 N0/2 = N0e 1/2 => t1/2=ln2/X (2.7) Radioactive half lives may range from fractions of a second to trillions of years (Seymore 1960). 2.1.5 Units of radioactivity. Radioactivity is measured in number of disintegrations per unit of time. The SI unit is the becquerel (Bq), defined as disintegrations per second. Activity is solely a measure of the energy that is released from the atom and does not distinguish between alpha or beta particles, gamma rays, or neutrons. Nor does it tell anything about the quantity of energy that is absorbed in the irradiated matter. This quantity, however, is important when determining the biological effects of irradiation, because the different forms of radioactivity differ in their penetration power and their ability to ionize matter (see section 2.1.3) (Seymore 1960). The SI unit for the absorbed dose or energy deposited in matter is the gray (Gy), and is defined as an absorbed radiation dose of one joule per kg. 2.2 Radioactivity from nuclear explosions. 2.2.1 Types of nuclear explosions. Forty three nuclear devices were detonated at Enewetak Atoll (see Table 2.3). Forty one were detonated from towers, barges, the ocean surface, or dropped from aircraft's while two were detonated underwater. "Each of the explosions exerted its own special effects on the atolls. The underwater shots released large amounts of radioactive materials into the water to be absorbed, retained, or passed on in the complex marine biological web, but deposited only minimal amounts of radioactive substances on the land areas or in the atmosphere. The tower shots caused spectacular physical disturbance to the islets of the atolls, often completely obliterating life in the immediate area, and released radioactive material which contaminated nearby land and water areas. The third general type of shot, the high-altitude detonation, probably had little or no effects on the atolls. Any radioactive materials released would have been quickly moved out of the area by winds" (Welander et al. 1966). 18 Table 2.3. Dates and location of the military nuclear detonations at Enewetak Atoll. Modified from Henry and Wardlaw (1991) and Helfrich and Ray (1987). The origin of the beta radioactivity traced in this study is mainly from the Nectar shot (1954), which is marked in bold. Military name of event Date Burst type/height Yield (ktf Location X-RAY Apr. 14, 1948 Tower, 63 m 37.0-10° Enjebi YOKE Apr. 30, 1948 Tower, 63 m 49.0-10° Aomon ZEBRA May 14, 1948 Tower, 63 m 18.0-10° Runit DOG Apr. 7, 1951 Tower, 94 m Class. Runit EASY Apr. 20, 1951 Tower, 94 m 47.0-10° Enjebi GEORGE May 8, 1951 Tower, 63 m Class. Eleleron ITEM May 24, 1951 Tower, 63 m Class. Enjebi MIKE3 Oct. 31, 1952 Surface 10.4-103 Elugelab KING Nov. 15, 1952 Airdrop, 471 m 50.0-101 Runit NECTAR May 13, 1954 Barge 16.9-102 Elugelab LACROSSE May 4, 1956 Surface 40.0-10° Runit YUMA May 27, 1956 Tower, 63 m Class. Aomon ERIE May 30, 1956 Tower, 94 m Class. Runit SEMINOLE June 6, 1956 Surface 13.7-10° Boken BLACKFOOT June 11, 1956 Tower, 63 m Class. Runit KICKAPOO June 13, 1956 Tower, 94 m Class. Aomon OSAGA June 16, 1956 Airdrop Class. Runit INCA June 21, 1956 Tower, 63 m Class. Lujor MOHAWK July 2, 1956 Tower, 94 m Class. Eleleron APACHE July 8, 1956 Barge Class. Elugelab HURON July 21, 1956 Barge Class. Elugelab CACTUS May 5, 1958 Surface 18.0-10° Runit BUTTERNUT May 11, 1958 Barge Class. Runit KOAa May 12, 1958 Surface 13.7-102 Teiteiripucchi WAHOO May 16, 1958 Underwater, 157 Class. Mut HOLLY May 20, 1958 Barge Class. Runit YELLOWWOOD May 26, 1958 Barge Class. Enjebi MAGNOLIA May 26, 1958 Barge Class. Runit TOBACCO May 30, 1958 Barge Class. Enjebi ROSE June 2, 1958 Barge Class. Runit UMBRELLA June 8, 1958 Underwaterb Class. Ikuren WALNUT June 14, 1958 Barge Class. Enjebi LINDEN June 18, 1958 Barge Class. Runit ELDER June 27, 1958 Barge Class. Enjebi OAKa June 28, 1958 Barge 8.9-103 Bokoluo SEQUOIA July 1, 1958 Barge Class. Runit DOGWOOD July 5, 1958 Barge Class. Enjebi SCAEVOLA July 14, 1958 Barge Class. Runit PISONIA July 17, 1958 Barge Class. Runit OLIVE July 22, 1958 Barge Class. Enjebi PINE July 26, 1958 Barge Class. Enjebi QUINCE Aug. 6, 1958 Surface Class. Runit FIG Aug. 18, 1958 Surface Class. Runit a. Thermonuclear bomb; b. Off bottom in approximately 50 m of water; c. kt = kiloton, Class. = classified. 19 The nuclear detonations at Enewetak Atoll were either fission or fusion (thermonuclear) types. 239 * * Fission bombs contain Uranium and Plutonium atoms that split in half under the detonation forming two isotopes of approximately half the size of the original atom. Since the neutron to proton ratio in light elements is lower than in heavier elements, most of the 'fission daughter elements' are unstable and give off radioactivity. Approximately 200 isotopes of 35 elements are created with the detonation of a fission bomb (see Table 2.4). Most have very short half lives, but a few have half lives of up to 30 years (Seymore 1960), and Uranium and Plutonium atoms that did not split in the detonation (fissile non-fission products, see section 2.1.2) have half lives of up to 4.5-109 years (see Table 2.4). In fusion bombs, isotopes of Hydrogen (Deuterium and Tritium) fuse under very high 6 233 239 * temperatures (on the order of 10 °C, triggered by the fission of Uranium or Plutonium). Fusion bombs have also been termed "clean devices" or "clean bombs" since relatively few radioisotopes are formed (Seymore 1960). Those that are formed include Tritium, Deuterium and a variety of induced radioisotopes (activation products, see section 2.1.2 and Table 2.4). Table 2.4. Artificial radioisotopes originating from a nuclear explosion. Modified from Seymore (1960) and Bablet and Perrault (1987b). Fission products Activation products Non-fission fissile isotopes Radioisotopes Half life Radioisotopes Half life Radioisotopes Half life (days) (days) (days) 137Cesium 11.0-103 1 "Carbon 20.3-105 238Uranium 45.0-108 Strontium-90 Yttrium 10.1-103 207Bismuth 10.2-103 239Plutonium 24.4-103 125 Antimony 98.6-101 Tritium 44.5-102 235Uranium 71.0-102 147Promethium 91.3-101 6°Cobalt 19.3-102 24°Plutonium 66.0-102 155Europium 62.1-101 55Iron 94.9-101 241Americium 45.8-101 Ruthenium-106Rhodium 36.5-101 54Manganese 31.2-101 238Plutonium 86.4-10° Cesium-144Praseodymium 8.4-101 57Cobalt 27.1-101 Zirconium-95Niobium 65.0-10° 110mSilvera 25.0-101 91Yttrium 58.8-10° 65Zink 24.5-101 89Strontium 52.7-10° 58Cobalt 70.0-10° Ruthenium-103Rhodium 40.0-10° 59Iron 45.0-10° Cesium-141 Praseodymium 32.5-10° 140Barium 13.0-103 131 Iodine 80.0-10"1 140Lanthanum 45.7-10"4 90Yttrium 73.1-10"4 a. The'm' indicates that the silver atom is an isomer, i.e., that it is in a long-lived excited state but eventually will give off gamma rays (usually gamma rays are emitted instantly following an alpha or beta decay). 20 In the largest nuclear devices a so-called fission-fusion-fission process takes place. Here, the high temperatures from a fission process triggers a fusion process that in turn releases a flux of neutrons that can fission more material (Seymore 1960). 2.2.2 Decay of mixed fission products. The many radioisotopes created in a nuclear detonation have different half lives, some longer and some shorter than the average. The overall rate of decay of all fission products combined therefore decreases over time (Seymore 1960): "For the mixture of all fission products, radioactivity decreases tenfold for each sevenfold increase in time following the detonation in which the isotopes were produced. At this rate the decrease in activity from one hour after to 343 hours after (approximately two weeks) is a thousandfold." The theoretical gross beta-decay of mixed slow-neutron initiated fission products of 235Uranium was estimated by Hunter and Ballou (1951). Over a period of 1 to 1000 days the decay is approximated by a straight line on a log-log scale, with an average slope of -1.2 (Bonham 1958). This form of decay can be described by a power function D = a • t"12, where D is the amount of radioactivity at time t in days after the detonation and a is the intercept. A distinction should be made between decay and decline. Decay refers to the decrease in activity determined from a sample kept and measured repeatedly in the laboratory. Decline, on the other hand, refers to the rate of change in activity determined from samples collected over time in the same locality. "If decline is more rapid than decay a reduction of activity in the environment beyond that caused solely by physical decay is suggested, and conversely, a steeper decay than decline suggests either an increase in availability in the environment or an accumulation or concentration of radioactivity by the organism. Equality of decay and decline suggests that uptake and excretion of radioisotopes have reached an equilibrium with the environment" (Bonham 1958). 2.2.3 Uptake of radioactivity by marine organisms. The most abundant naturally occurring radioisotopes in sea water ('background radioactivity') are 40Potassium and 87Rubidium, comprising approximately 90% and 10%, respectively (see Table 2.1). 40Potassium is also the principal radioisotope found in marine organisms, followed by 14Carbon, 232Thorium, 234Thorium and 226Radium (Bablet and Perrault 1987b). 21 With the nuclear explosions at Enewetak Atoll, a huge number of artificial radioisotopes were added to the marine environment, initially in the form of local fallout on the ocean surface. From here, the radioisotopes were subjected to oceanic dilution, dispersion, concentration, and transport (Mauchline and Templeton 1964) as illustrated in Figure 2.2, where the thicker arrow indicates the route followed in the present study. Dilution and Transport Concentration Retentive dispersion processes Advection Turbulent diffusion Currents Moving organisms Physical-chemical Biological Deposition of silt Static organisms Adsorption Co-precipitation Ion-exchange Flocculation and sedimentation Figure 2-2. A schematic illustration of the various processes taking place once the radioactive fallout reaches the ocean surface. The thicker arrows indicate the pathway followed in this study. Adapted and modified from Mauchline and Templeton (1964). Biological concentration refers to the uptake of radioisotopes by marine organisms. Depending on the organism, this happens in different ways. Single cell organisms may acquire the radioisotopes through passive diffusion, active uptake across membranes, adsorbed to the surface, or with water engulfed during the formation of food vacuoles (Sanders and Gilmour 1994). Higher organisms may acquire them through the food followed by an absorption in the gut, or they may absorb them directly from the water through gill surfaces or other external epithelia (Mauchline and Templeton 1964). Once within the organism, the radioisotopes are treated the same way as their stable isotopes, or other similar elements, and are accumulated 22 as non-radioactive elements. Radioiodine, for example, is concentrated in the thyroid because it is similar to stable Iodine. Strontium is concentrated in the skeleton of vertebrates because of its similarity with Calcium, and Cesium is concentrated in muscles and flesh due to its similarity with Potassium (Bablet and Perrault 1987b). The first marine organisms to concentrate fallout material following a nuclear test are the primary producers including phytoplankton and algae. Within a few hours, these organisms can concentrate radioisotopes a thousand fold (Donaldson 1959, Seymore 1960, Bablet and Perrault 1987b). Plankton seems to have no preference but contains most of the fallout radioisotopes found in sea water (Seymore 1960). The most common radioisotopes detected in primary producers within a few weeks after a nuclear detonation at Enewetak Atoll are shown in Table 2.5. From primary producers, the radioisotopes are disseminated to the rest of the food web as determined by the prey / predator relationships and the selective uptake of isotopes by the organisms (see Table 2.5). Table 2.5. Radioisotopes typically detected in marine organisms at Enewetak Atoll following a nuclear detonation. Based on Bablet and Perrault (1987b), and Donaldson (1959). Type or organisms' Dominant radioisotopes Primary producers 95Zirconium - 95Niobium, 57'58'60Cobalt, 65Zinc, 55'59Iron, 14U44Cerium (rare earth), 103'106Rufhenium, 54Manganese Herbivores (except fish) Cerium (rare earth), Ruthenium-Rhodium Herbivorous fish (e.g., surgeonfish) 65Zinc, 57'58'60Cobalt First order carnivores (except fish) 144Cerium, 103,106Ruthenium, 57,58'60Cobalt, 65Zinc, Zirconium - 95Niobium, n0mSilverb, 54Manganese, 137Cesium (only as trace element) First order carnivorous fish 65Zinc, 55'59Iron, 57'58'60Cobalt, 137Cesium Second and higher order carnivores 65Zinc, 141,144Cerium (rare earth), 57,58'60Cobalt, 103,106Ruthenium, traces of 54Manganese, Zirconium -95Niobium, and 137Cesium a. The organisms are arranged after trophic level with lower trophic level organisms mentioned first and higher trophic level organisms mentioned last in the column; b. The'm' indicates that the silver atom is an isomer, i.e., that it is in a long-lived excited state but eventually will give off gamma rays (usually gamma rays are emitted instantly following an alpha or beta decay). 23 Strontium and Cesium, which are both considered hazardous to humans because of their long half lives and similarity to Calcium and Potassium, respectively, are not concentrated to any high extent in marine organisms (Bablet and Perrault 1987b). In fish, the alimentary tract generally shows the greatest amount of radioactivity followed by the liver, skin, bone, and muscles (Donaldson 1959). 2.3 Compartment modeling. A convenient way of dealing with whole ecosystems and the transfer of material amongst their components is through the use of compartment models. Here the ecosystem under investigation is divided into a number of distinct functional groups ('compartments') comprising either single species or groups of similar species, and the transport of material between the groups is described as a flux per unit time. Mathematically, the compartments are connected by a set of either linear or non-linear equations, one for the balance of each group (O'Neill 1979). Furthermore, depending on whether the approach is dynamic or static, the equations may be either differential or non-differential. Compartment models have been used extensively in the study of tracers dynamics (see for example the review by O'Neill (1979)). 2.3.1 Ecopath. One of the latest approaches in compartment modeling is the Ecopath software based on a concept originally proposed by Polovina and co-workers (Polovina 1984, 1993). It has since been further developed by V. Christensen and D. Pauly (Christensen and Pauly 1992a, 1992b, 1995) to incorporate routines for network analysis and system maturity indices based on the theory of RE. Ulanowicz, H.T. Odum and E.P. Odum. It has furthermore been turned into an periodically updated software distributed by the International Center for Living Aquatic Resources Management (ICLARM; see http://www.ecopath.org). Ecopath is a modeling software that allows for the straightforward construction, analysis, and comparison of mass-balance trophic models (Vasconcellos et al. 1997) (see Figure 2.3). It is applicable for well defined ecosystems in either 'steady-state' or in which biomass changes do occur. An important constraint in Ecopath is that during the time period considered, the energy entering any functional group must balance the energy leaving the group plus whatever energy 24 is accumulated within the group (the mass-balance concept). Thus, Ecopath can be compared to a book keeping system where every flux must be accounted for. Assuming similar conditions over the time period covered by the model, the trophic interactions among the functional groups of the ecosystem can be described by a set of linear mass-balance equations wherein Production by (i) = all predation on (i) + non-predation losses of (i) + export of (i) (2.8) which may also be written P1-M2i-PI(l-EE1)-EX1 = 0 (2.9) where; Pi is the production of i; M2; is the predation mortality on i; EE; is the ecotrophic efficiency of i, or the fraction of the production of i that is consumed within the system and exported or harvested (EE; is usually left as the unknown to be estimated when solving Equation (2.9)); 1-EEj is the 'other mortality', i.e., the non-predation losses of i, or the fraction of the production of i that flows to detritus; and EX; is the export of i. Equation (2.9) can be re-expressed as: B, (P/B),-±B1 • (Q/B)j • DCji -B, •(P/B)i (l-EE^-EX, = 0 (2.10) J=I or B, .(P/B). EE, -IBJ (Q/BV DCjj -EX, = 0 (2.11) j=i where; B; is the biomass of i during the period considered; P/B; is the production/biomass ratio of i which, under the assumption of equilibrium, is equal to the total mortality rate (Z;) (Christensen and Pauly 1992b); Q/Bj is the consumption/biomass ratio of i; and DCji is the fraction of prey i in the average diet of predator j. Based on Equation (2.11), for an ecosystem with n functional groups, a system of n linear equations can be set up: 25 B, (P/B)j EE, -Bj (Q/B), DC;, -B, (Q/B)2•DQ-...-Bn-(Q/B)n DC;, -EX, =0 (2.12) EL • (P/B)2 • EE, -fi, • (Q/EI), • DC,2 -EL • (Q/E!)2 • DC^-.-B,, • (Q/E!)n • DC^ -EX, =0 Bn • (P/B)n • EE,-B, • (Q/B), • DC,n -B, • (Q/B)2 • DC,,,-. -B,• (Q/E!)n• DCm -E^, =0 This system of linear equations can be solved using standard matrix algebra (Christensen and Pauly 1992a, 1992b). Only one of the input parameters: B;, (P/B)i; (Q/B); or EE;, may in general be left unknown, while the diet composition matrix, exports, and harvests always must be provided. The solution of Equation (2.11) allows calculation of the energy balance of each compartment, using Consumption by (i) = production by (i) + respiration by (i) + unassimilated food by (i) (213) Rearranging the equation, respiration can be quantified given the other flows: Respiration by (i) = consumption by (i) - production by (i) - unassimilated food by (i) (214) In Ecopath, the mass-balance concept implies that Equations (2.8) through (2.13) applies for all compartments of the ecosystem, i.e., that the estimated EE; range between 0 and 1 (a diagnostic for mass-balance). 2.3.2 Ecoranger The majority of Ecopath models so far have been created with a single set of mean input parameters for the period under consideration, and the researchers were unable to take into account the large uncertainties that tend to accompany biological data. This rather critical point has now been solved with the introduction of Ecoranger (Christensen and Pauly 1995), an Ecopath routine that allows one to enter, for each input parameter, a mean or mode value, a range, and a distribution. The shape of the distribution depends on one's prior knowledge of the data and may be either uniform, triangular or normal. Once the routine is running, input variables for each parameter type are drawn randomly from the specified distributions and the resulting models are evaluated. Only models that pass the constraints of mass-balance and 26 which are thermodynamically possible are accepted. The process is repeated in a Monte-Carlo fashion where the user specifies the number of realizations and desired successful runs (accepted models). Of the accepted models the 'best model' in a least-square sense (i.e., that with the least square deviation from the modes or means) is saved and used for further analysis. I Apex predators B-2.5 Q. O Benthic producers B = 1,300 Detritus B = 2,000 Figure 2-3. Schematic representation of an Ecopath model of a coral reef in the Virgin Islands, Caribbean (Opitz 1996). The functional groups are arranged along the vertical axis according to their trophic level. The area of each box is proportional to the logarithm of the biomass of the corresponding functional group. Flows exit a box from the upper half and enters a box in the bottom. Flows cannot be divided, but can merge with flows from other boxes. 27 3. Methods The following chapter describes the methodology applied in the study and is divided into three main parts. In the first part, the modeled section of the atoll perimeter is defined and the zonation across the reef section is described. The second part is devoted to the process of deriving the model input parameters for the seventeen non-fish and ten fish groups included in the model. Lastly, the origin and 'processing' of the radioactivity data is described, and the theory of combining the model outputs with the radioactivity data to simulate the observed trends in radioactivity over time is explained. 3.1 Defining the modeled area. As mentioned in section 1.2, Enewetak atoll may be divided into four parts: the windward reef, the leeward reef, and two transitional reefs. The majority of the biological research has been conducted on the windward reef including the study by Odum and Odum (1955) summarized in section 1.4. The Ecopath model was therefore restricted to this area stretching from, but not including, Enewetak Island in the south up to, and including, Bogon Island in the north as shown in Figure 3.1. Figure 3-1. Cross-reef currents and channel currents. Redrawn and modified from Atkinson et al. BOGON is. TRADES ENEWETAK IS. JAPTAN IS. (1981). 28 The windward reef itself may be divided into distinct zones, perpendicular to the prevailing westward moving North Equatorial Current, each zone with a characteristic flora and fauna (Ladd 1973). Gdum and Odum (1955) distinguished between six zones on their transect (see Figure 1.4), a zonation that is fairly typical of the central and southeastern part of the windward reef (Colin 1987a). In this study, however, only five zones were distinguished, beginning from the oceanic side: fore reef, algal ridge, reef flat, coral head zone, and sand / shingle zone (Figure 3.2). Sand / shingle Coral head Reef Algal Fore reef zone zone flat ridge zone Figure 3-2. Zonation across the windward reef as defined in the present study. Redrawn and modified from Johannes and Gerber (1974). The area of each zone was determined from a digitized bathymetric map of the atoll (see Table 3.1). A short description of the five zones follows below. Table 3.1. The area of the five zones across the windward reef as determined from a digitized map of the atoll (Anon. 1944). Islands were included under the reef flat area. Zone Depth rangee Mean depth Area (m) jm) (km2) Fore reef 0-20 10.00 0.83 Algal ridgea 0 0.25d 1.15 Reefflatb 0 0.60d 3.45 Coral heads 0-4 2.00 2.24 Sand / shingle 4-20 12.00 10.18 a. Defined as lA of the 0 m depth zone; b. Defined as 3A of the 0 m depth zone; c. Defined as V5 of the area down to 100m depth; d. From Figure 2 in Buddemeier (1975), the depth at mean tide; e. As identified from a bathymetric map (Anon. 1944). 29 3.1.1 Fore reef. The fore reef area, which is located seaward of the algal ridge, has never been accurately determined, but Colin (1987a) suggested that it is about 3 to 4 times smaller than the reef flat. It ranges in width from about 300 m in the south to less than 100 m in the north (Colin 1987a). A typical spur and groove system, formed by encrusting coralline algae (Marsh 1970), characterizes the zone immediately seaward of the algal ridge (Wiens 1962, Colin 1987a, Ristvet 1987). Invertebrates such as sea urchins are abundant on the sides of the spurs while the grooves are floored with boulders and cobbles that are moved by the currents, preventing organisms from settling. Fish (including herbivores parrotfish and surgeonfish that move onto the algal ridge and reef flat at high tide, see section 5.1.2.4) are numerous (Colin 1987a). Seaward of the spur and groove system, the bottom slopes gently down to about 18-23 m depth where a sharp drop-off begins. A few corals (primarily the vasiform Acropora cytherd) can be found in this area, however, the bottom is mostly covered by rocks, many with signs of boring clinoid sponges (Smith and Harrison 1977, Colin 1987a). 3.1.2 Algal ridge. The spur and groove system leads up to a marginal algal ridge located within the surf zone. Some of the grooves continue beneath the algal ridge and reef flat, forming large surge channels. Long sections of the algal ridge are dead4, which is likely a result of wartime oil pollution (Ladd 1973) (see section 1.4). In general, the algal ridge is poorly developed, consisting of a narrow band of corals and (mostly soft) algae. Odum and Odum (1955) found that the encrusting yellow coral Acropora palmerae, Pocillopora, and Millepora platyphylla covered up to 50% of the area, while Smith and Marsh (1973), on a transect close to Odums', found that the coral cover was much less. Dominant algae on the algal ridge are the fleshy algae Dictyosphaeria intermedia, Zonaria variegate, Caulerpa elongata, Ceramium, and Dictyota as well as the calcareous red algae Porolithon onkodes (Odum and Odum 1955, Smith and Marsh 1973). 4 Assumed to be the situation throughout the.period covered by the model. 30 3.1.3 Reef flat. The reef flat varies in width from 90 - 160 m, and is mostly covered by water. It consist of sandy areas and smooth rocks that slope gradually towards the lagoon (Ladd 1973). The zone is paved with coralline algae such as Jania capillacea and Porolithon (Smith 1973b, Smith and Marsh 1973). Corals are sparse and cover much less than half of the area. The most conspicuous corals are Acropora and Millepora. Filamentous red, -brown, -and green algae, cyanobacteria and foraminiferans form heavy mats throughout the area (Odum and Odum 1955, Ladd 1973). 3.1.4 Coral head zone. The coral head zone is strictly subtidal and located towards the lagoon. It is rich on corals such as rounded heads of Favia pallida and Cyphastrea serailia, microatolls (colonies where the central part is dead but the sides are still thriving) of Porites lutea, branching forms of Acropora gemmifera, A. cymbicyathus, Pocillopora, Stylophora, and the blue coral Turbinaria mesenterina. Strips of sand, shingle, and cobble runs between the corals. The area varies in width from about 200 m in Odums' study area to about 1 km in the north. Fish are abundant (Odum and Odum 1955, Smith and Marsh 1973, Johannes and. Gerber 1974, Colin 1987a). 3.1.5 Sand / shingle. The central lagoon is bordered by a terrace dotted with numerous patch reefs. The terrace varies in width from a few hundred meters in the south to more than 1 km in the north (Ristvet 1987). The depth at the outer edge of the terrace is about 15-22 m (Wardlaw et al. 1991). The patch reefs were ignored in this study. Instead the area was considered to be uniformly covered by sand and shingle (produced upstream). Foraminiferans and filamentous algae, the latter living within the coral shingle, are very abundant (Odum and Odum 1955). 3.2 Validating the Ecopath model. The data used for the construction of the Ecopath model were all from the published literature. They represent more than 30 years of research, ranging from the study by Odum and Odum (1955) in 1954 to studies well into the 1980s. This rather large time span was 31 justified as coral reefs are known to be very stable systems changing little over time (Odum and Odum 1955). In a few cases, data were imported from similar systems in other parts of the world. Seventeen non-fish and ten fish groups were identified, and biomass estimates for each group and for each zone were obtained. The estimates were subsequently averaged into a single weighted biomass estimate for each group. 3.2.1 Non-fish groups. A qualitative description of the non-fish groups is given below, while the Ecopath parameters (B, P/B, and Q/B) are explained in Table 3.3 and the remarks following the table. A diet matrix can be found in Appendix 1. 3.2.1.1 Detritus. Detritus consist of dissolved and particulate organic matter (DOM and POM). DOM stems from phytoplankton, benthic algae, and corals that excrete large fractions of their primary production directly into the water. It is an important source of energy for filter feeding organisms including bacteria, zooplantkon, bivalves, sponges, polychaetes, tunicates and corals. POM consist of dead organic matter including excrements, feces, non-assimilated food, etc. It is colonized by bacteria and algae, and is consumed by a variety of filter-feeders and fish (Sorokin 1990). 3.2.1.2 Benthic Primary Producers. This group consist of all primary producers associated with the benthic environment such as encrusting, matted, and fleshy green algae, calcareous red algae, large branching algae attached to dead coral fragments / heads (mainly Halimedd), free-living small algae, boring red and green algae, and cyanobacteria. All in all, 238 species of benthic algae have been identified at Enewetak Atoll (see Table 3.2). Benthic algae fragments from the reef front constitute a large fraction of the plankton over the reef and is utilized by many herbivores and detritivores (Wiebe et al. 1975). 32 Table 3.2. Marine benthic algae at Enewetak Atoll. From Tsuda (1987). Division Number of species Cyanophyta (cyanobacteria) Chlorophyta (green algae) Phaeophyta (brown algae) Rhodophyta (red algae) 16 89 24 109 3.2.1.3 Phytoplankton. Sargent and Austin (1949) measured an extremely low concentration of phytoplankton in the lagoon of Enewetak Atoll, supporting the general belief that phytoplankton is of no significance in coral reef ecosystems. Sorokin (1993), however, has recently provided several examples of phytoplankton blooms in atoll lagoons, and Colin (1987a) has on several occasions observed large phytoplankton blooms in the lagoon of Enewetak Atoll (perhaps an artifact of the nuclear testing). This inconsistency between observations, and a general lack of data from reef zones other than the lagoon, meant that the biomass of phytoplankton was left as the unknown to be estimated by Ecopath (see section 2.3.1). 3.2.1.4 Zooplankton. This group consist of meroplankton ('temporary' zooplankton such as fish -and invertebrate larvae) and holoplankton ('full-time' zooplankton). In a study of the fish and zooplankton at Enewetak Atoll, Hobson and Chess (1978) discovered that planktivorous fish concentrate in areas of strong current during the day, where they feed on zooplankton of oceanic origin. In contrast, nocturnal planktivorous fish concentrate in areas of weak currents where they feed on resident zooplankton that enter the water column at night in concentrations 2-3 times higher than the day time concentrations. "... zooplankton is extremely abundant in reef waters and so is an important component of the coral reef ecosystem... Such information as is available shows that the largest zooplankton biomass could be found at night up the shallow reef areas with patch reefs covered with living corals or with rubble..., i.e., in places where in accordance with earlier data... it ought to be lowest, being depleted by bottom sessile predators, and especially corals" (Sorokin 1993). There is a general lack of zooplankton data (and other data) from the fore reef which is physically very difficult to monitor. Hamner et al. (1988), however, was able to sample the windward side of Davies Reef, Australia, and found that zooplankton is a major source of 33 energy to the reef. Planktivorous fish living on the fore reef form a "wall of mouths" that effectively removes the zooplankton before the water hits the reef. As most investigations thus have underestimated the biomass of zooplankton, it was left to be estimated by Ecopath. 3.2.1.5 Corals and sea anemones (Class anthozoa). Thirty eight species of octocorals (Octocorallia) and 169 species of stony corals (Scleractinia) have been identified at Enewetak Atoll (Burch 1987, Devaney and Lang 1987). Sea anemones are much less abundant than corals, and no taxonomic work has been published (Cutress and Arneson 1987). It has been estimated that there is about three times as much primary producer biomass in corals as there is animal biomass (Odum and Odum 1955). The primary producers consisted of boring filamentous algae and zooxanthellae (in the ratio 16:1). In this study, however, zooxanthellae were included in the coral biomass. Since the plant biomass is located within the coral skeleton, it does not receive enough light to contribute significantly to the reefs primary production (Lewis 1981, Marsh 1987). Although some corals are capable of obtaining all their energy from the zooxanthellae (Marsh 1987), most hermatypic corals feed autotrophically, as predators, and as filter feeders all at the same time. Corals primarily feed at night, but some also feed during day or at dusk and dawn (Sorokin 1990, 1993). Scleractinian corals have been shown to feed on "copepods, ostracods, mysids, chaetognaths, appendicularians, nematods, polychaetes, small jelly fish and salps. The dominating components in the gut contents were zoea and copepods. The suspended organic material ingested by corals... included bacteria, protozoa, detritus, and dead zooplankters..." (Sorokin 1993). 3.2.1.6 Foraminiferans and other protozoans. Approximately 280 species of foraminifera and nonplanktonic protozoans have been identified at Enewetak Atoll (Chave and Devaney 1987). The protozoan fauna is fairly typical of the Western Pacific though they are particular scarce on some of the northern islands. This, however, may very well be a result of the nuclear testing that took place in the area (Hirshfield et al. 1968). 34 Foraminiferans are an important food source for many benthic invertebrates including holothurians, sea urchins, polychaetes and shrimps as well as for fish that graze and scrape the coral surfaces and sandy substrates (Sorokin 1993). Lipps and Delaca (1980) identified approximately 200 shallow water species. A large number of suspension feeding foraminiferans, many containing zooxanthellae, were found in cryptic habitats where they live protected from their predators. Filamentous and mat-like types, on the other hand, have adapted to the high predation pressure through a high turnover rate. Symbiotic foraminiferans are amongst the most important primary producers in the sand / shingle zone (Sorokin 1993). Foraminiferans feed either autotrophically or on bacteria, algae, other benthic protozoans, and eggs and larvae of meiobenthic organisms (Sorokin 1993). 3.2.1.7 Gastropods. A total of 1116 species of marine mollusks have been identified at Enewetak Atoll, 994 of which are gastropods (Kay and Johnson 1987). Conus, Mourla, Drupa, Thais, and Cypraea are particular abundant on the reef flat (Renaud 1976, Kohn 1980, Miller 1982). "Gastropods are much more abundant on intertidal benches than on more complex and benign subtidal coral reefs in the same regions, although species diversity is considerably lower... with particular attention to the genus Conus" (Kohn and Leviten 1976). Many gastropods are important predators on other gastropods and on polychaetes. The diet of Drupa morum, e.g., consisted of 44% vermetids, 42% nereids, 5% other polychaetes, and 9% crustaceans (Kohn 1987). 3.2.1.8 Bivalves. One hundred and fifteen different species of bivalves have been identified at Enewetak Atoll (Kay and Johnson 1987), and two groups were distinguished in the Ecopath model: tridacnids (giant clams) and 'other bivalves'. Giant clams, like 'other bivalves', are filter feeders but, in addition, contain symbiotic zooxanthellae which under favorable conditions may supply up to 100% of the giant clams energy requirement (Heslinga and Fitt 1987). I assumed that giant 35 clams, as a group, obtain 75% of their energy from symbiotic zooxanthellae and the remaining 25% from feeding on phytoplankton5. Few records of bivalves other than giant clams were found in the literature, but Sorokin (1993) mentioned that bivalves (including giant clams) "comprise 10-30% of the reefs malacofauna and about the same part of its total biomass". Many bivalves bore into corals and rocks which might explain that they have been somewhat overlooked. Bivalves are numerous in the sand / shingle zone (Riddle et al. 1990). 3.2.1.9 Shrimps and lobsters. Approximately 150 species of decapod shrimps and lobsters (infraorders: Penaeidea, Stenopodidea, Caridea, and Palinura) have been identified at Enewetak Atoll (Devaney and Bruce 1987). Not included in this number are the callianassid shrimp (ghost / burrowing shrimp) living in the sand / shingle zone. Though they represent some of the most abundant infauna in this area, no biomass estimate has been derived (Suchanek and Colin 1986, Suchanek et al. 1986). "Callianassids are amongst the most elusive of coral reef animals. Their high density (indicated by the frequency of their feeding mounds) and high sediment-turnover rates... suggest they are major consumers. No satisfactory technique has yet been developed to quantify these animals or their contribution to total community metabolism" (Riddle et al. 1990). Devaney and Bruce (1987) discovered four species of lobster on the windward reef. Panulirus penicillatus, which lives on the outer reef slope during day and moves onto the reef flat at night, was particularly common (Ebert and Ford 1986). 3.2.1.10 Stomatopods. Stomatopods were included as a group because they often occur in the diet of other groups. Twelve species of stomatopods, considerably smaller than the same or related species found in other parts of the Indo-West Pacific area, have been identified at Enewetak Atoll (Reaka and Manning 1987). The biomass of the group was left to be estimated by Ecopath. 5 After completion of this study, Dr. R.E. Foreman (pers. com.) has later noted that giant clams also feed extensively on DOM. 36 3.2.1.11 Miscellaneous crustaceans. This group consist primarily of crabs from the infraorders Brachyura and Anomura, but also includes other similar sized crustaceans, as well as amphipods and isopods. Seventy six species of Anomuran crabs and 291 species of Brachyuran crabs (53% xanthid) have been identified at Enewetak Atoll (Garth et al. 1987). According to Kohn (1987), seven out of eight of the most common xanthid species studied were herbivores. On the reef flat, however, several xanthid species have been found to be important carnivores. 3.2.1.12 Echinoderms - not including holothurians. This group consist of Ophiuroidea (basket stars / brittle stars), Asteroidea (sea stars), Echinoidea (sea urchins and heart urchins), and Crinoidea (sea lilies). Ninety seven species have been identified at Enewetak Atoll (Devaney 1987a). Brittle stars of the genus Ophiocoma can be found in all zones (Chartock 1983a), however, open sandy areas are dominated by irregular herbivores sea urchins (Colin 1987a). Ophiocoma are suspension and deposit feeders eating algae, and detritus. Some "specimens inhabiting the reef floor occasionally contained foraminiferans, sponge spicules, crustacean (e.g., isopod) skeletal parts, nematodes, and juvenile snail shells..." (Chartock 1983a). 3.2.1.13 Holothurians. Sea cucumbers have few if any predators and are very abundant in the atoll environment. Holothuria atra is, according to Kohn (1987), "the most conspicuous deposit-feeding invertebrate on interisland platforms." Twenty species of sea cucumbers from five genera have been identified at Enewetak Atoll (Burch 1987, Cutress and Rowe 1987). Lawrence (1980) studied eight of the most conspicuous species and found a distinct distribution on the reef flat, with only two co-occurring species. Bakus (1968) reported an average density of Holothuria difficilis of 1 to 32 individuals per 900 cm2 in daytime, but up to 200 individuals per 900 cm2 at night on the tops of slab rocks in certain areas. Bacteria and foraminiferans are major sources of food for holothurians (Bakus 1973), though the foraminiferans probably pass through the digestive tract without much effect, leaving bacteria and organic detritus as the main sources of energy. Another important source of energy is dissolved organic matter that the holothurians obtain directly from the water. The 37 feces of H. difficilis contains "living and dead filamentous blue-green and red algae, fish eggs, unidentified detritus, sponge spicules, copepod exuvia, foraminiferans, fragments of sea urchin spines, holothurian ossicles, gastropods, fish teeth and calcareous fragments" (Bakus 1968). Webb et al. (1977) estimated an assimilation efficiency for//, atra of 40%. 3.2.1.14 Polychaetes and other worm like invertebrates. A total of 132 species of polychaetes from 34 families have been identified at Enewetak Atoll (Devaney and Bailey-Brock 1987). Polychaetes play an important role as bioeroders and as food for a variety of fish and invertebrates, particularly mollusks (Sorokin 1993). The algal ridge is dominated by carnivorous nereid type annelids, the reef flat by sedentary annelids and nereid type annelids, and the coral head zone by sedentary species living within the coral heads (Odum and Odum 1955). Polychaetes are suspension feeders, deposit feeders and carnivores, preying on encrusting invertebrates such as corals (Kohn 1987). According to Sorokin (1993), the shallow parts of a coral reef contains approximately 30% filter feeding polychaetes, 40% detritophages and omnivorous polychaetes, and 30% predatory polychaetes. 3.2.1.15 Sessile invertebrates. Besides sponges, this group comprises hydrozoa, chordates, hemichordates, and other sessile invertebrates. Forty species of sponges have been identified at Enewetak Atoll. Boring sponges are "the most common infaunal associates of the corals studied, with 86% of the corals showing sponge bioerosion effects" (Devaney 1987b). "Boring sponges can contribute up to 25% of the total erosion of the substratum on Enewetak... and are considered major eroders on most coral reefs..." (Russo 1980). 3.2.1.16 Cephalopods. Cephalopods were included as a group because they often occur in the diet of other groups. No parameters pertaining particularly to cephalopods from Enewetak Atoll were found in the literature. 38 3.2.2 Biomass, P/B, and Q/B values of non-fish groups. Table 3.3 summarizes the biomass, production/biomass (P/B), and consumption/biomass (Q/B) values for the seventeen non-fish groups included in the Ecopath model. Table 3.3. Summary table of the biomass6, P/B, and Q/B values of the non-fish groups included in the Ecopath model.3 Functional group Biomass Remark P/B Remark Q/B Remark (t wwkm"2-year"!) (year"1) (year"1) Detritus 185 1 n.a. - n.a. -Benthic primary prod. 3255 2 2.0 14 n.a. -Phytoplankton ? - 593.0 15 n.a. -Zooplankton ? - 55.0 16 165 17 Corals 212 3 2.0 18 4 19 Foraminiferans 33 4 14.0 20 21 21 Gastropods 16 5 2.2 22 9 23 Giant clams 6 6 0.2 24 3 25 Bivalves 21 7 2.3 22 10 17 Shrimp and lobster 3 8 4.6 26 27 27 Stomatopods ? - 1.8 28 27 28 Misc. crustaceans 6 9 4.3 29 30 29 Echinoderms 93 10 1.2 17 4 30 Holothurians 42 11 0.2 31 4 30 Polychaetes 29 12 5.8 32 24 32 Sessile invertebrates 37 13 2.3 33 29 33 Cephalopods ? - 2.1 33 7 33 a. The remarks are explained in the text that follows the table. Dashes indicate that no source was found, and question marks indsicate that the value was estimated by Ecopath. n.a. = not applicable. 3.2.2.1 Remarks to Table 3.3. 1) Table 3.4 summarizes the biomass estimates of detritus (POM and DOM) in the different reef zones. The weighted mean for the reef as a whole was 185 t ww-km"2 2) Odum and Odum (1955) estimated the biomass of benthic primary producers, including zooxanthellae, in all of their zones. I grouped zooxanthellae with their symbiotic counterparts (corals, foraminiferans, and giant clams) and used a ratio of 16:1 plant biomass to zooxanthellae (Odum and Odum 1955) to derive the zooxanthellae biomass in corals (see section 3.2.1.5) and subtract it from the Odums' benthic primary producer estimate. 6 Wet weight was used as the model 'currency', however, Dr. R.E. Foreman (pers.com.) has later noted that organic weights (C), especially when working with radioisotopes, are preferrable. 39 Table 3.5 summarizes the benthic primary producer estimates in the different reef zones. The weighted mean for the reef as a whole was 3255 t ww-km"2. 3) Table 3.6 summarizes the coral biomass estimates in the different reef zones. The weighted mean for the reef as a whole was 212 t ww-km"2. 4) Table 3.7 summarizes the foraminiferan biomass estimates in the different reef zones. The weighted mean for the reef as a whole was 33 t ww-km"2. 5) Table 3.8 summarizes the gastropod biomass estimates for the different reef zones. The weighted mean for the reef was 16 t ww-km"2. This is a very conservative estimate. Miller (1982) found a density of detritus-feeding vermetids (sessile worm snails) on the reef flat of 151-1084 per m2 while the biomass estimate in Table 3.8 only includes species of the genus Thais. 6) Odum and Odum (1955) estimated the biomass of small giant clams and small herbivorous mollusks (gastropods) in the coral head zone. Based on their comments, I assumed that the small herbivorous mollusks consisted of 1/3 giant clams and 2/3 gastropods. This resulted in a biomass estimate of giant clams of 35.94 t ww-km"2 in the coral head zone (conversion: dw = 10% ww (Opitz 1996)), and a weighted mean for the reef as a whole of 6.2 t ww-km"2. 7) Riddle et al. (1990) estimated a biomass of bivalves of 37 t ww-km"2 in the lagoon of Davies Reef, Australia (conversion: 1 g C = 2 g organic matter (ash) (Riddle et al. 1990), and ash = 8% ww (Sambilay 1993)). Assuming that this estimate also applies for the lagoon of Enewetak Atoll, a weighted mean of 21.11 ww-km"2 for the reef as a whole was derived. 8) Ebert and Ford (1986) estimated a total population of 7800 lobsters on the windward reef (reef flat and fore reef zone). With a mean carapace length of 91.6 mm for males and 81.2 mm for females, assuming a sex ration of 1:1, and applying a weight-length relationship of W = 0.0021-L2773, a total biomass of 0.716 t ww-km"2 was derived (for the fore reef, algal ridge and reef flat). Odum and Odum (1955) estimated a biomass of shrimps of ~1 g dw-m"2 in the coral head zone. Table 3.9 summarizes the biomass estimates of shrimps and lobsters. The weighted mean for the reef as a whole was 3.1 ww t-km"2. 9) Hermit crabs have been found on the reef flat in densities ranging from 3 to 65 m"2 (Kohn 1987). Table 3.10 summarizes the biomass estimates for the group. The weighted mean for the reef as a whole was 6 t ww-km"2. 40 Table 3.4. Biomass estimates of detritus (POM and DOM) in the different reef zones. Zone POMa DOMab POM+DOMd Biomass6 Source (g ww-m"3) (g ww-m'3) (g ww-m"3) (t ww-km"2) Fore reef 0.12 21.6 21.60 216.0 Marshall et al. (1975) II 0.28 - - - Johannes (1967) Algal ridge 0.25 26.2 26.45 6.6 Marshall et al. (1975) Reef flat0 0.28 22.0 22.27 13.4 Marshall et al. (1975) II 0.26 - - - Gerber and Marshall (1974) Coral heads 0.28 22.0 22.35 44.7 Marshall et al. (1975) 0.26 _ _ _ Gerber and Marshall (1974) II 0.24 _ - - Gerber and Marshall (1982) Sand / shingle 0.36 24.0 24.30 291.6 Marshall etal. (1975) 0.34 _ _ _ Johannes (1967) it 0.20 - Gerber and Marshall (1974) Weighted mean - - - 185.0 -a. Assuming C = 10% ww; b. The estimates were derived by wet combustion (Marshall et al. 1975) and therefore, according to Sorokin (1993), were underestimated 1.5-2 times. To account for this all values were multiplied by 2; c. Biomass assumed equal to the coral head zone; d. Calculated as the average of the POM estimates plus the DOM estimate; e. Estimated using Table 3.1. Table 3.5. Biomass estimates of benthic primary producers (free living algae and boring filamentous algae) in the different reef zones. Modified from Odum and Odum (1955). Zone / group Benthic primary producersa (t ww-km"2) Fore reef 3869c Algal ridge 3520 Reef flat 564Coral head zoneb 3869 Sand / shingle zone 2233 Weighted mean 3255 a. Conversion: 1 g dw = 5.71 g ww (Opitz 1996); b. Average between the Odums' zone of small and large coral heads; c. Biomass assumed equal to the coral head zone. 41 Table 3.6. Coral biomass estimates (not including inorganic skeleton). From Odum and Odum (1955). Zone Zooxanthellae (t ww-km"2) Animal polyps3 (t ww-km"2) Total biomass (t ww-km"2) Fore reef 37 467 504 Algal-coral ridge 106 700 806 Reef flat 47 333b 380 Coral heads 37 467 504 Sand / shingle 0 0 0 Weighted mean 22 190 212 a. Conversion: dw = 15% ww (Vinogradov 1953, Odum and Odum 1955); b. This is the only zone where sea anemones were mentioned (4.3 gm"2) (Odum and Odum 1955); c. Biomass assumed equal to the coral head zone. Table 3.7. The biomass of foraminiferans in the different reef zones. Modified from Odum and Odum (1955). Zone Coverage of sand Counts-cm"2 of Counts-cm"2 of Biomass3 / mats containing small forams large forams (t ww-km"2) foraminifera (%) (0.01 cm) (0.1 cm) Fore reef - - - 9.70b Algal ridge - - - 0.03c Reef flat 70 25 0 0.03 Coral heads 34 2 19 9.70 Sand / shingle 67 3 54 55.63 Weighted mean - - - 33.40 a. Odum and Odum (1955) estimated an ash-free dry weight (AFDW) of large foraminiferans of 1.33-10"4 g. Assuming that foraminiferans are spherical, their volume is equal to 4/37tr3 and the volume-ratio of a small foraminiferan with a diameter of 0.01 cm to a large foraminiferan with a diameter of 0.1 cm is 1:1000. This gives an AFDW of small foraminiferans of 1.33- 10"7g (conversion: AFDW = 86.5% ww (Odum and Odum 1955)); b. Biomass assumed equal to the coral head zone; c. Biomass assumed equal to the reef flat. 42 Table 3.8. Biomass estimates of gastropods in the different reef zones/ Zone Biomass Source (t ww-km"2) Fore reef 25.3a Odum and Odum (1955) Algal ridge - -Reef flat 5.5b'c Odum and Odum (1955) Coral heads 25.3M Odum and Odum (1955) Sand / shingle 19.0e Riddle et al. (1990) Weighted mean 16.0 -a. Biomass assumed equal to the coral head zone; b. Conversion: dw = 18% ww (Arias-Gonzales et al. 1993); c. The estimate only includes species of the genus Thais; d. The estimate includes species of the genera Thais, Coury, and Conus; e. From Davies Reef lagoon, Australia, where Riddle et al. (1990) estimated a biomass of 765 mg C-m"2 (conversion: 1 g C = 2 g ash (Riddle et al. 1990), and ash = 8% ww (Sambilay 1993)); f. Dashes indicate that no biomass estimate was found and that gastropods probably do not occur in the zone. Table 3.9. Biomass estimates of shrimps and lobsters in the different reef zones. Zone Biomass Source (t ww-km"2) Fore reef 4.65° Ebert and Ford (1986) Odum and Odum (1955) Algal ridge 0.72 Ebert and Ford (1986) Reef flat 0.72 Ebert and Ford (1986) Coral heads 3.93a Odum and Odum (1955) Sand / shingle 3.93b -Weighted mean 3.10 -a. Conversion: dw = 26.7% ww (Opitz 1996); b. Biomass assumed equal to the coral head zone; c. Assuming that both lobsters and shrimps occur in this zone in densities similar to the reef flat and coral head zone. 43 Table 3.10. Biomass of crabs and other crustaceans in different reef zones. Zone Biomass Source (t ww-km"2) Fore reef 9.6a -Algal ridge 46.8b Odum and Odum (1955) Reef flat 2.0b Odum and Odum (1955) Coral heads 9.6b Odum and Odum (1955) Sand / shingle 1.7C Riddle etal. (1990) Weighted mean 6.0 -a. Biomass assumed equal to the coral head zone; b. Conversions: dw = 25% ww (Opitz 1996); c. The crustacean infauna biomass from Davies Reef lagoon, Australia. Conversion: 1 g C = 2.2 g dw. Table 3.11. Biomass of echinoderms in the different reef zones. Zone Biomass3 Source (t ww-km"2) Fore reef 53.6b Odum and Odum (1955) Algal ridge 157.6 Odum and Odum (1955) 240.0C Chartock (1983a) Reef flat 13.6 Odum and Odum (1955) Coral heads 53.6 Odum and Odum (1955) Sand / shingle 120.0d Colin (1987a) Weighted mean 93.1 -a. Conversion: dw = 25% ww (Vinogradov 1953, Arias-Gonzales 1993); b. Biomass assumed equal to the coral head zone; c. Biomass of Ophiocoma anaglyptica, the dominant benthic invertebrate in this zone according to Chartock (1983a); d. A density of >50 urchins-m"2 (Colin 1987a) was converted into a biomass estimate assuming an average wet weight of irregular sea urchins of 2.4 g-ind"1 (Odum and Odum 1955). 44 Table 3.12. Biomass estimates of holothurians in the different reef zones. Zone Biomass Source (t ww-km"2) Fore reef 233.5 b -Algal ridge 1.4 c -Reef flat 1.4 a Odum and Odum (1955) Coral heads 204.3 a Odum and Odum (1955) 262.7 Webbetal. (1977) Sand / shingle 2.6 Riddle etal. (1990) Weighted mean 42.0 -a. Conversion: dw = 14% ww ((Bakus 1968) based on specimens of Holothuria difficilis); b. Biomass assumed equal to the mean of the coral head zone; c. Biomass assumed equal to the reef flat. Table 3.13. Biomass of polychaetes and other worm-like invertebrates in the different reef zones. Zone Biomassa Source (t ww-km"2) Fore reef 65b Odum and Odum (1955) Algal ridge 81 Odum and Odum (1955) M 105 Bailey-Brock et al. (1980) Reef flat 40 Odum and Odum (1955) II 41 Bailey-Brock et al. (1980) Coral heads 65 Odum and Odum (1955) Sand / shingle 7° Riddle etal. (1990) Weighted mean 29 -a. Conversion: dw = 20% ww (Arias-Gonzales 1993); b. Biomass assumed equal to the coral head zone; c. From Table 3.15, conversion: 1 g C = 11 g ww (Riddle et al. 1990, Opitz 1996). 45 10) Table 3.11 summarizes the biomass estimates for echinoderms other than holothurians in the different zones. The weighted mean for the reef as a whole was 93.1 t ww-km"2. 11) Webb et al. (1977) observed an average density of Holothuria atra of 3.03 per m2 in an area similar to Odums' coral head zone. Odum and Odum (1955) further estimated the density of holothurians on the reef flat and in the coral head zone. Table 3.12 summarizes the various biomass estimates. The weighted mean for the reef as a whole was 42 t ww-km"2. 12) Table 3.13 summarizes the biomass of polychaetes and other worm like invertebrates in the different zones. The weighted mean for the reef as a whole was 29 t ww-km"2. 13) Very few biomass estimates were found for the sessile invertebrate group. Basile et al. (1984) found that sponges occur in the fore reef zone, and Kohn (1987) found them on the reef flat as well: "Clinoid sponges that excavate chambers in the hermatypic coral Porites lutea on interisland platforms are ecologically the most important Porifera of the Enewetak intertidal and shallow subtidal zones..." Odum and Odum (1955) estimated a biomass of 34 g dw-m"2 in the coral head zone. Except for the sand-shingle zone, I applied this estimate for all zones, which lead to a weighted mean of 37 t ww-km"2 (conversion: dw = 39% organic ww (Opitz 1996)). 14) Marsh (1970) estimated a productivity of reef-building calcareous red algae on the algal ridge and in the spur and groove system of 4008 g ww-m^-yr"1 (conversion: 1 g C = 16.7 g ww (Opitz 1996)). Using the biomass estimate from the algal ridge (Table 3.5), a P/B ratio of 2.2 year"1 was derived. Bakus (1967), using exclosure experiments, measured the primary productivity of cyanobacteria on the reef flat and found that it ranged between 0.65 - 2.15 gC-m^-day"1 or 3958 - 13111 g ww-m^yr"1 (conversion: 1 gC = 16.7 g ww (Opitz 1996)). Applying the biomass estimate for this zone (Table 3.5) lead to a P/B value ranging from 0.8 to 2.7 year"1. Combining the two P/B estimates resulted in a mean P/B value for the group as a whole of 2 year"1. 15) Gerber and Marshall (1982) measured a phytoplankton concentration behind the reef of 1.23 mg C-m"3, and Sargent and Austin (1949) measured a primary production of 2 mg C-m" 3-day"\ From here a P/B ratio of 593 year"1 was derived. 16) From Sorokin (1993). 17) From Opitz (1996). 46 18) A P/B ratio of 9.1 year"1 for zooxanthellae and 1.1 year"1 for coral animal polyps (Sorokin 1993) was combined with the biomass estimates from Table 3.6 to give a weighted P/B value of 2 year"1. 19) Corals obtain approximately 70% of their energy from symbiotic zooxanthellae and 30%> from other external sources (Sorokin 1993, example from Heron Island, Australia). Assuming that corals consume 2993 t ww-km^-year"1 (Opitz 1996) (conversion: 1 kcal = 1 g ww), and using the weighted mean coral biomass estimate from Table 3.6, a Q/B value of 14 year"1 was derived. The value was reduced to 4 year"1 to take into account the 70%> internal feeding on zooxanthellae. 20) Hallock (1981) estimated a turnover rate (P/B ratio) of 11-16 year"1 for three species of foraminifereans (Amphistegine lessoni, A. lobifera, and Calcarina spengleri) in the Philippines. Since the same species occur at Enewetak Atoll (Chave and Devaney 1987), an average P/B value of 14 year"1 was applied for the group as a whole. 21) A Q/B value of 30 year"1 (Opitz 1996) for foraminiferans was lowered to 21 year"1 to take into account that 30%> of the diet comes from internal feeding on zooxanthellae (see Appendix 1). 22) From Riddle et al. (1990) . 23) Riddle et al. (1990) estimated a yearly consumption rate for large gastropods (> 2 mm) of 277 kJ m"2. They also estimated a biomass of 31 kJ-m"2 (conversion: 1 g C = 42 kJ, Riddle et al. (1990)). This lead to a Q/B value of 9 year"1. 24) From Lewis (1981). 25) Arias-Gonzales (1993) used a Q/B value of 10 year"1 for a group of bivalves including Tridacna maxima. I lowered the value to 3 year"1 to take into account 75% internal feeding on zooxanthellae (based on Heslinga and Fitt (1987)). 26) Ebert and Ford (1986) estimated a natural mortality (M) for Panulirus penicillatus (spiny lobster) of 0.284 year"1 for males and 0.244 year"1 for females. With no fishing, and with a seemingly stable age structured population, M can be assumed to equal the total mortality (Z) which again equals P/B (Christensen and Pauly 1992b). A sex ratio of 1:1, therefore, resulted in an average P/B value of 0.264 year"1. A P/B value of 5.34 year"1 for shrimps was obtained 47 from Arias-Gonzales (1993). Applying the biomass estimates from Table 3.9, a weighted P/B value of 4.6 year"1, for the group as a whole, was derived. 27) Pauly et al. (1993) estimated a Q/B value of 29 year"1 for two penaid shrimps, and Opitz (1996) estimated a Q/B value for spiny lobsters of 7.4 year"1. Applying the biomass estimates from Table 3.9, a weighted Q/B value of 27 year"1, for the group as a whole, was derived. 28) From Opitz (1996), from a group comprising shrimps, hermit crabs, and stomatopods. 29) From Arias-Gonzales (1993), from a group dominated by xanthid crabs. 30) From Pauly et al. (1993). 31) According to Pauly et al. (1993), the natural mortality (M) for low-metabolism echinoderms is approximately equal to the rate constant K (time"1) of the von Bertalanffy growth function. K-values for some holothurians at Enewetak Atoll are presented in Table 3.14. Since there was no harvest of holothurians in the period considered for the model, the total mortality (Z) can be expressed in terms of M. Z, however, also equals P/B (Christensen and Pauly 1992b), and hence a P/B value of 0.227 year"1 was derived from Table 3.14. 32) Riddle et al. (1990) found that the infauna of the Davies Reef lagoon, Australia, was dominated by polychaetes (Table 3.13). A P/B value of 5.8 year"1 and a Q/B value of 24 year"1 was derived from Table 3.15. 33) From Arias-Gonzales (1993). Table 3.14. Rate constants for some holothurians at Enewetak Atoll. Modified from Pauly et al. (1993). Species K (year1) Holothuria atra 0.110 Actinopyga mauritana 0.120 Stichopus chloronotus 0.450 Mean 0.227 Table 3.15. Bimass, P/B and Q/B values for infaunal polychaetes in the lagoon sediments of Davies Reef, Australia. Modified from Riddle et al. (1990). Biomass values are ± 95% confidence limits. Feeding type Size class Biomass P/Ba Q/Bb (mm) (mg C-m"2) (year1) Macrophagous > 2.0 124 ±55 2.6 12 Microphagous > 2.0 217 ± 80 3.3 15 Macrophagous 0.5 - 2.0 71 ± 16 10.3 37 Microphagous 0.5 - 2.0 224 ± 24 8.5 35 a. Average value from Table 5 in Riddle et al. (1990); b. Conversion: 1 g C = 48 kJ (Opitz 1996). 48 3.2.3 Fish groups. 3.2.3.1 The distribution and abundance of fish. Fish are very abundant in all of the Marshall Islands. From 1953 to 1966 Schultz and collaborators (1953, 1960, 1966) identified and described 543 species, and a checklist of 817 species (338 genera and 97 families) was recently assembled by Randall and Randall (1987). In a comprehensive study, Hiatt and Strasburg (1960) examined the food and feeding habits and ecological relationship of 223 fish species (56 families and 127 genera) of the Marshall Islands. By comparing this work with that of Schultz and collaborators (1953, 1960, 1966) and Randall and Randall (1987), 190 of Hiatt and Strasburgs 223 species were found to occur at Enewetak Atoll. I grouped the 190 species into ten functional groups based on: 1) size: small < 30 cm TL7 and large > 30 cm TL; 2) feeding type: herbivorous (parrotfish and surgeonfish), omnivorous (> 10% of diet consist of plant material), carnivorous or piscivorous; 3) a data set on the radioactivity in reef fishes of Belle Island (Figure 1.3), Enewetak Atoll (Welander 1957) (see section 3.3.1); and 4) two Ecopath models by Arias-Gonzales (1993). The ten fish groups were: miscellaneous piscivorous fish (mainly sharks and jacks), small carnivorous fish, large carnivorous fish, small omnivorous fish, large omnivorous fish, snappers / groupers, butterflyfish, surgeonfish, parrotfish, and herring (see Appendix 2 and 3). The diet compositions (Appendix 2) were, in most cases, derived from Hiatt and Strasburg (1960), who examined 2051 fish stomachs, and identified the types of food consumed. Table 3.16 shows an example of the stomach context of Neoniphon sammara (Holocentrus sammara in Hiatt and Strassburg (I960)) as presented by Hiatt and Strasburg (1960). To convert the percent column in Table 3.16 into a diet composition as required in Ecopath, I counted the number of different food items and assigned an equal weight to each. In the example from Table 3.16, 12 different items were consumed and thus, were assigned a weight of 100/12 = 8.3%o each. The items were then grouped into the appropriate functional group (as identified in the Ecopath model) and the 'weights' added to derive the diet composition. 7 "The length of a fish measured from the front of the jaw which is most anterior to the end of the longest caudal ray, but excluding the caudal filaments" (FishBase 1998). 49 Table 3.16. Example of the stomach context of Neoniphon sammara as presented by Hiatt and Strasburg (1960). Food item Percentage of fish containing the item Crustacea Crab fragments 36 Parthenopiid crab 27 Thalamita sp. 18 Pachygrapsus plicatus 9 Portunic crab 9 Maiid crab 9 Unidentified crustacean fragments 18 Copepods 9 Coelenterata Pieces of unidentified coral, partly digested 18 Polychaeta Unidentified polychaetes 9 Gastropods Cerithium sp. 9 Algae Algal frond, bitten off 9 In this example, the 12 food items were grouped into the following diet composition: miscellaneous crustaceans 66.7%, corals 8.3%, polychaetes 8.3%>, gastropods 8.3%>, and benthic primary producers 8.3%. In a few cases, the diet compositions were obtained directly from Arias-Gonzales (1993) and Hobson and Chess (1978). Despite numerous studies of fish at Enewetak Atoll, few have been quantitative and then mostly concentrating on a few species or families occurring in a specific habitat (Bakus 1967, Miller 1982). Odum and Odum (1955) used visual census to roughly estimate the biomass of small and large herbivorous and carnivorous fish in each of their six reef zones (see Appendix 5). Small fish were counted in 36 m2 quadrates, and their numbers converted into biomass assuming a mean dry weight (dw) of 2.42 g per individual, as determined from a rotenone sample. Large fish were "rapidly counted with 360° underwater vision", and converted to dry weight based on a sample of 12 speared fish of the same size (120 g dw per individual). Table 3.17 summarizes the biomass estimates that were converted from dry weight (dw) to wet weight (ww) assuming that for fish except sharks, dw = 26%> ww (based on Sambilay (1993)). The conversion factor used for sharks and derived from Odum and Odum (1955) was dw = 20% ww 50 The P/B values for the 10 fish groups (Table 3.17) were obtained from published Ecopath models (Alino et al. 1993, Arias-Gonzales 1993, Silvestre et al. 1993, Opitz 1996), while the Q/B values were estimated using the empirical regression by Pauly et al. (Pauly et al. 1990, Christensen and Pauly 1992b, Table 3.17, and Appendix 4): Q / B = 106 37 • 0.0313^ • W,-0168 • 1.38pf • 1.89Hd (3.1) where; Woo is the asymptotic or maximum weight of the fish in gram wet weight; Tk is the mean annual habitat temperature expressed as 1000/(T°C + 273.1) (an annual mean temperature of 27.5 was used in all cases based on Atkinson (1987)); Pf is one for apex predators, pelagic predators, and zooplankton feeders, and zero for all other feeding types; and Hd characterizes the food type and is set to one for herbivores and zero for carnivores. Table 3.17. Parameter estimates of the ten fish groups in the Ecopath model. Functional group Biomass3 P/B8 Q/Bb (t ww-km"2-year"') (year"1) (year"1) Misc. piscivorous fish 6.3 0.3C 6 Herring 0.4 3.5d 30 Small carnivorous fish - 2.4f 14 Large carnivorous fish - 0.6d 6 Small omnivorous fish - 2.5d 24 Large omnivorous fish - 2.2e 9 Snappers / groupers - 0.8C 6 Butterflyfish 5.7 2.6C 14 Surgeonfish 9.4 1.2C 13 Parrotfish 3.9 2.1c 13 a. Derived from Odum and Odum (1955) assuming that all large fish, other than sharks, in the sand / shingle zone could be grouped as miscellaneous piscivorous fish. Conversion: dw = 26% ww (Sambilay 1993), except for sharks where dw = 20% ww (Odum and Odum 1955). Dashes indicate that no biomass estimate was derived, but was left to be estimated by Ecopath; b. Q/B = consumption/biomass. Estimated using the empirical equation derived by Pauly et al. (1990) (see Equation (3.1) and Appendix 4); c. Derived from Arias-Gonzales (1993); d. Derived from Opitz (1996); e. Derived from Silvestre et al. (1993); f. Derived from Alino et al. (1993); g. P/B = production / biomass ratio. 51 3.3 The origin and incorporation of the radioactivity data. 3.3.1 The origin of the radioactivity data. From shortly before and up to two years after the 'Nectar' shot on May 14, 1954 at Enewetak Atoll (Table 2.3), the level of beta radioactivity in the most common aquatic organisms was measured, and the results prepared in three reports by the Applied Fisheries Laboratory, University of Washington, Seattle (in contract with the United States Atomic Energy Commission): Bonham (1958) studied the radioactivity in invertebrates; Palumbo (1959) reported on the level of radioactivity in algae; and Welander (1957) looked at the radioactivity in reef fish. The 'Nectar' shot took place about 4.3 km east-northeast of Belle Island (see Figure 3.1), which received a greater amount of fallout than the rest of the islands (Welander 1957). It therefore became center for subsequent investigations (as well as the focus of this study). Invertebrates, including bivalves (Tridacna crocea), sea cucumbers (Holothuria atra), gastropods (Lambis), and corals (Acropora, Porites, Pocillopora, and Heliopora), were collected along the seaward side of Belle Island. A total of 693 specimens of fish, representing 57 species and 22 families, were collected in the same area using rotenone, hook and line, or spear in water from about 5 cm to about 4 m depth. Algae were collected from the intertidal zone all around the island, while plankton and water samples were collected on the lagoon side. All samples, except for plankton and water, were immediately put on ice and kept in freezers at the Enewetak field laboratory until further processed. Fish and invertebrates were dissected as to tissue (Welander 1957, Bonham 1958): bivalves were dissected into mantle, adductor muscle, gill, kidney, visceral mass, and shell; gastropods were dissected into mantle, foot muscle, terminal portions of liver and gut, viceral mass, and shell; sea cucumbers into gonads, gut with context, and body wall (Bonham 1958); and fish were dissected into skin, muscle, bone, liver and viscera (Welander 1957). Similar tissues from small fish were pooled, or in some cases the entire fish was used (Welander 1957). The samples were sent to the University of Washington laboratory, Seattle, in insulated containers with dry ice. 52 Five milliliter water samples and filtered plankton samples were placed on 3.8 cm stainless steel plates and dried and ashed before they were sent to Seattle (Bonham 1958). At the University of Washington laboratory, samples of approximately one gram were placed on pre-weighed 3.8 cm stainless steel plates, weighed and dried at 97°-99°C for 12 to 24 hours. They were subsequently ashed overnight at 500°-550°C. After cooling and weighing, they were slurried with ethyl alcohol, spread evenly on the plates with a glass rod, dried, and affixed to the plates with a few drops of 0.5 per cent Formvar in ethylene dichloride. The samples were counted in Nucleometer internal gas-flow (methane) counting chambers, and were corrected back to the date of collection based on the decay of a soil sample collected at Belle Island on May 15, 1954 (Donaldson 1953, Welander 1957, Bonham 1958, 1959). 3.3.2 Observed trends in radioactivity in various organisms. Varying with the organisms, the field samples indicated that a maximum level of radioactivity, in the organisms, was obtained between 1-10 days after the detonation, followed by an approximately linear decline on a log-log scale of beta decay versus time (up to two years) after the detonation (Bonham 1958, Palumbo 1959, Welander 1957). Figure 3.3 shows an example of the observed beta radioactivity in corals (Acropora) from Belle Island as presented by Bonham (1958). The figure shows that from day 36 to day 710 after the nuclear detonation, the decline was approximately linear, with a correlation of -0.976, according to the (anti-logged) regression Y = 2.7 • 108 -T2 23. The 'Nectar' shot was not the first nuclear detonation in the area, and residual long-lived radioisotopes from earlier explosions rendered the decline curves less steep than if they had been a result of the 'Nectar' shot alone (Bonham 1958). The observed data on beta radioactivity in the various organisms were based on variable sample sizes as shown in Table 3.18. 53 oTt > 1 O 1000 100 10 1 ° Not used in regression • Used in regression 6 pre-detonation level 3 o.i 1 10 100 1000 Days after nuclear detonation on May 14, 1954 Figure 3-3. Total beta radioactivity in corals (Acropora) in disintegrations-mm^g'wet-lO3 after the 'Nectar' shot on May 14, 1954 (Bonham 1958). From day 36 to 710 the decline approximates a straight line. Table 3.18. Mean number of specimens from which the observed data on beta radioactivity were derived. Based on Bonham (1958), Palumbo (1959), and Welander (1957). Functional group Dominant organisms Mean number of specimens3 Bent. prim. prod. Corals Giant clams Gastropods Holothurians Parrotfish Surgeonfish Butterflyfish Small omniv. fish Herring Small carniv. fish Large carniv. fish Snappers / groupers Misc. pisciv. fish Halimeda, Dictyota, Caulerpa, Lyngbya, Spyridia, Udotea, 17.0 Codium, Microdictyon Acropora 3.6±0.5b Tridacna crocea 2.1 ±0.2 Lambis 2.3 ± 0.5 Holothuria atra 3.0 Scarus purpureus 2.9 ± 2.1 Acanthurus triostegus 2.8 ± 1.6 Chaetodon auriga 1 ± 1.4 Damseffish (Abudefdufbiocellatus), blennies 3.9 ± 2.5 30.0 Cardinalfish, squirrefish (Holocentrus sp.), wrasses 3.0 ± 1.7 (Halichoeres trimaculatus) Mullet (Neomyxis chaptalli), triggerfish, goatfish 5.4 + 5.7 (Mulloidichthys samoensis) Epinephelus merra 2.8 ± 1.7 Jacks, sharks 1.7 + 0.8 a. Mean values ± standard deviation; b. The value refers to the number of plates counted in the gas-flow counting chamber. 54 3.3.3 Radioactivity in whole organisms. The data on radioactivity in bivalves, gastropods, holothurians and fish were reported as the activity in various tissues (section 3.3.1) and not as the activity in the organisms as a whole. The relative weight of the different body parts of bivalves, gastropods, holothurans, and fish was therefore obtained from the literature (see Table 3.19), and the activity for whole organisms derived. Table 3.19. Relative weight of the different body parts of fish, bivalves, holothurians, and gastropods as determined from the literature. Organism Relative weight of various tissues Fish 8% skin, 63% muscle, 18% bone, 2% liver, 9% viscera3. Bivalves {Tridacna crocedf 95% shell, 0.75% adductor muscle (or 15% of the tissue excluding shell, based on Heslinga and Watson (1985)), and the remaining 4.25% equally divided into gills, viscera, mantle, and kidney. Holothurians (Parastichopus calufornicus) 19% gut, 10% gonads, 17% muscle, 54% body wall / integument (excluding body fluids and respiratory tree)c. Gastropods 70% shell and the remaining 30% equally divided between liver, gut, mantle and muscled. a. Based on Welander (1957); b. Reaches about 20 cm (Sorokin 1993); c. Based on Giese (1966); d. Based on Hammen (1980). 3.3.4 Simulating the observed trends in beta radioactivity. One of the outputs of a balanced Ecopath model is the estimated fluxes of biomass among the functional groups as presented in the 'food intake matrix'. These fluxes were used to simulate the fate of beta radioactivity within the marine ecosystem of Enewetak Atoll. Assuming that the radioactivity is mixed evenly within a functional group, one may think of it as 'tagged' biomass (T) that flows from one functional group to another according to the overall flux of biomass between the groups as illustrated in Figure 3.4, where; B; and Bj are the biomasses (t-km"2) of group i and j, respectively; T; and Tj are the tagged biomasses (t-km"2) in group i and j, respectively; and Qy is the flux of biomass (t-km"2-year"') from group i to j. 55 Group i Group j Figure 3-4. A schematic representation of the transfer of radioactivity between compartments of an ecosystem. The radioactivity can be thought of as 'tagged' biomass (T) that flows from group i to j according to the flux of biomass (Qy) where Bj and Bj are the biomasses of group i and j, respectively. The transfer of radioactivity per unit time from group i to j, py-, is proportional to the fraction of'tagged' biomass to total biomass in group i, T/B;, and the flux (Qy) of biomass from group i to j: T <=> Pij = V B; (3.2) where; My = Qy/B;. My is the transfer coefficient (year"1) from group i to group j, i.e., that part of the natural mortality of i that is due to j. The Ms are calculated in Ecopath and presented in a predation mortality matrix. When dealing with radioactivity, there is an additional loss besides predation, 8, within each group, resulting from the physical decay of the radioisotopes. The total or gross beta radiation emitted by fission products, produced in a process where a number of 235Uramium atoms undergo fission simultaneous, e.g., in a nuclear detonation, was presented by Hunter and Ballou (1951) (see also section 2.2.2). They found that the gross decay curve of the fission products can be described by a power function D = a • tb, where D is the amount of radioactivity at time t in days after the detonation; a is the intercept; and b is the slope / decay rate (equal to -1.2). When differentiated, the equation may be re-expressed as: dD dt = a-b-tb-' = b-(a-tb)-t_1 = b-D-t_1 = (b/t)-D = 5-D (3.3) where; 5 = (b/t). Combining the income, loss and decay terms, the trend in radioactivity in the functional groups may be described by a linear differential equation system of the form: loss dT; decay -L = ZT1.M1J-TJ.XMJI-5.TJ dt i=. (3.4) i=l which can be integrated over time. 56 The Solver routine in Microsoft Excel was applied to minimize the total sum of squared deviations between the observed and predicted levels of beta radioactivity [__(ln(obs/pred)2)] by modifying the predation mortality matrix from the original Ecopath model, thus simulating a trend that was more consistent with the observed data: the columns of the predation mortality matrix (i.e., the mortalities exerted by a predator on its various prey groups) were scaled up / down by multiplying them with a factor that Solver was preprogrammed to vary within a certain range (0.25-1.75), while minimizing the residuals defined above. The rows of the predation mortality matrix (i.e., the mortalities experienced by a given group) were also modified, by allowing Solver to add an additional mortality (M+) to the sum of the mortalities (sum of the rows). The coefficient of determination, R2 (= l-SSres/SStot), was used to evaluate the fit of different runs. The modified predation mortality matrix was subsequently used to re-calibrate the original Ecopath model by modifying the input biomass, i.e., the inputs directly proportional to the predation mortalities, used for the next iteration (see Equation (2.11)). To initiate the simulation it was necessary to know the levels of beta radioactivity at the 'bottom' of the food web, i.e., at trophic level one, which comprises the benthic primary producers, phytoplankton and detritus. To derive the radioactivity in the benthic primary producers, a linear regression was performed on the observed concentrations of beta radioactivity over time (Palumbo 1959) (Figure 3.5), and the result used in the simulation. Benthic primary producers S cr m o 8 0 12 3 4 Time after detonation (log day) Figure 3-5. Regression of the observed concentrations of beta radioactivity over time in the benthic primary producers. Based on Palumbo (1959). 57 There was no observations of the levels of beta radioactivity in phytoplankton and detritus. It was therefore assumed that phytoplankton incorporates beta radioactivity similar to benthic primary producers, and likewise that detritus contains levels similar to benthic primary producers which are the main contributors to detritus (see section 4.2.1). Radioactivity from earlier nuclear tests in the area was ignored under the assumption that it had no influence on the overall trends. It was further assumed that the organisms do not take up radioisotopes selectively. 58 4. Results The following chapter is divided into three main parts later to be discussed in Chapter 5. First the process of balancing the Ecopath model using Ecoranger is explained, including some necessary modifications of the initial input parameters. The results are presented in form of a table of the basic estimates and as a food web diagram. Next, the radioactivity data are mapped onto the food web diagram, and the simulation process including the modifications of the predation mortality matrix and re-calibration of the original Ecopath model is explained. The results are shown graphically and in two tables. Finally, the summary statistics of the model are presented together with the results of the network analysis, including the trophic transfer efficiencies estimated by Ecopath and a mixed trophic impact diagram. 4.1 Balancing the Ecopath model. 4.1.1 First run with Ecoranger using initial input parameters. The Ecopath model was balanced with Ecoranger using uniform distributions and a variability of 99% around the initial input parameters (see Table 3.3, 3.17, Appendix 1, 2, and 4). The majority of the data came from the Odums' study (1955) and the high variability assumed for the data was based on their comment that too few replicates were made to obtain maximum accuracy from the methods applied, and therefore, all they estimated were orders of magnitude (see also section 1.4). Ecoranger will not balance a model if the input are grossly erroneous. In the present case, given the initial input parameters, Ecoranger failed to find any solutions (balanced models, see section 2.3.2), and instead displayed a list of the first 50 runs that failed. In all cases the problem was the ecotrophic efficiency (EE) of surgeonfish, shrimps, miscellaneous crustaceans, and gastropods. Their EE values were much higher than one, suggesting that they could not sustain the predation pressure. The initial Ecopath input parameters were therefore modified, within the ranges given in the literature (see below), until Ecoranger was able to balance the model. 59 4.1.1.1 Modifying the predation mortality experienced by surgeonfish. Surgeonfish initially accounted for more than 50% of the diet of miscellaneous piscivorous fish. The particular diet composition, however, had been derived from just nine species with detailed diet information (see Appendix 2). By incorporating less detailed information from species such as Triaenodon obesus (Whitetip reef shark), Nebrius ferrugineus (Nurse shark), Sphyraena qenie (Blackfin barracuda), Carangoides orthogrammus (Island trevally), and Gymnosarda unicolor (Dogtooth tuna) all known to consume (small) reef fish and planktivorous fish (Lieske and Myers 1994), and by consulting other models (Polovina 1984, Opitz 1993), the following modified diet composition for miscellaneous piscivorous fish was derived, reducing their consumption of surgeonfish: 20%> surgeonfish, 19.1% small carnivorous fish, 12.5% butterflyfish, 12.5% parrotfish, 12.5% large omnivorous fish, 4.6% small omnivorous fish, 2.8% large carnivorous fish, 1%> snappers / groupers, 1%> herring, 6%> miscellaneous crustaceans, 4.6%> shrimps, 0.5%> cephalopods, and 0.4%> stomatopods. Moreover, to make the diet composition add up to one, cannibalism was increased to 2.5%. 4.1.1.2 Modifying the predation mortality experienced by shrimps, miscellaneous crustaceans, and gastropods. Small omnivorous and carnivorous fish in particular, caused a high mortality of shrimps, miscellaneous crustaceans, and gastropods. After comparing the Q/B values of these two fish groups (estimated as arithmetic means rather than weighted arithmetic means as no biomass information was available, see Appendix 4), with other models (Alino et al. 1993, Opitz 1996), the Q/B value was lowered from 14 to 10 year"1 for small carnivorous fish and from 24 to 15 year"1 for small omnivorous fish. To reduce the mortality of the three invertebrate groups even further, the diet of butterflyfish was modified to include 6 instead of 12%> shrimps with the difference assigned to corals which are the only food source for some butterflyfish (Sorokin 1993, Lieske and Myers 1994). The diet of echinoderms was modified to include less miscellaneous crustaceans (1 instead of 2%>) and more benthic primary producers which are the most important food item for sea urchins 60 (Ruppert and Barnes 1994) . Cannibalism by miscellaneous crustaceans was lowered from 10 to 1% by transferring the difference to detritus, as many crabs are scavengers and detritus feeders (Ruppert and Barnes 1994). Finally, cannibalism by small carnivorous fish was completely removed, while it was reduced to 1% and 3% for zooplankton and cephalopods, respectively, by rescaling their diet compositions to one. 4.1.2 Second run with Ecoranger using modified input parameters. Using the modified input parameters, Ecoranger (with similar constraints as outlined in section 4.1.1 and allowing a maximum number of total runs of 100,000) was now able to find balanced models. Of 50 successful runs (restricted by limited computer capacity) the basic estimates of the 'best model' are shown in Table 4.1 and graphically in Figure 4.1. 8 After completion of this study, Dr. R.E. Foreman (pers. com.) has later noted that many echinoderms, including sea urchins, can survive solely on DOM. 61 TJ s 3 O li S TJ 03 > CO OS CsJ OS cu CU _3 > o a s o 3 -rt c\3 ts o TJ .a CD •a .3 > (D > J3 CD 03 o o W > o M r3 S o co f s 2 ° CD (D „s TJ +3 1 If --3 cn I i o d O CD co CD CD 3 & S TJ +3 C co cj <D o " oo g '—1 VH ^ a. CD ^ W & H .a r—< <N oo ro oo ro --"ttsc^oo^ocN^IsoorsTrisrsrH OOOOO°HHr-H,-Hr-Hcslr0V0in'-»'q-v0inv0Oi— O HH <%l CN CO ooooooooooooooooooooo© a'I in 00 VO oo 3 ON ro 3 r- Os o CN VO ro in CN os r« CN in O o s° m CO CN 00 ON r- ro os ro VO HH vo m 00 •«T ro Os CN s o in os in 3 Os r^ S S ro 8 r—i CN in S CN r—i o in o CN CN o CN 00 o ro ' 1 VO r- ro VO VO VO vXir^^^o^vn«nro(NCOrvirn^a\^^oo^\t^^(>D^ oooooooooooooo o o o o o o o e9" £9'53 OOO^f^]rS^^^n^00Q^0^00O^^)00v0^^0^O^tC000 ' vn^coinr^ixi^oor^qcpOTr^ 0\ fN ri Q : ~"" ~; ,X '"< ~ ^ "", ~~ ON ON m vo 9 5 r--r- VO o Os in o CN I/-I in o in CN 00 Os CN o o O vd 00 as vd rn CN VO r- vd vd r- 5- 00 OS CN vd CN m Os CN l-H OS O o i r~ o o vo os o CNcNCNC^rSCslrslTrcNVOCNCNCNCNCNCNCNCNCNCNCN ooooooooooooooooo sf CS (N o o o oot^c^voost^cNoocosr^inoscsJ<^rncocNvpvoc»voosc>ooo Oi-iOHHTrrOCNOHHOmOcncNHHCNCNOHHCN-HOOO oooooooooooooooooo WVDtsMcslrOcnOSrHH-o^rs HHOSOscNOsr-OO^H-OSCNVO ooooooooooooooooo r— m m H Os cn o cN 3 H-^riScNsdr^^rvbo^r ^ CN CN HH c-t •—i <N Svdcnos^i/ii/iodrn 1H- r-H oo .-I -^f r-i r-H in m c*l SO CN O r-H \o VO ^rcs\<Nr~-r-oocNmr<-iooo3-^-OscOi-HCNOOOs^rcNOsoOiq-W-l r-H r-H C*-, r-H r— CN — H- o —i O O vo r- , , OS m in r- Os CO 00 00 3 OS OS r— o o 00 •n (N CN o O co O vo o CO M" ro r- o 00 CO O0 r-H in o OS S vd o O vd 00 Os CN o vd o CO CO CO r-H «n" o CN o r—i Os O CN in ' 1 CN in cil in m CN o CO 00 vo 00 CN CO m !0 CO 00 CO o ro CO vo V0 o CO in OS S 00 o vd 00 Os CN 00 vd o OS 1—1 CN CN CO in CN CN HVOcsr-cicnrjo roinrooovocNOO r-H VO i-H O O CN ro o O m o in in OOOOOr-Hr-Hr-Hr-H^^inr^l^l^MMOSi-HCNrOrO'^^Oq HHHcNCNCNCNCNHHtNHcNCNCNCNCNH HH O O O O O O O O O OOOOOOOOOOOCN II b" &J l-l > • « > lH CO . rH CD 5 rH CD .i= II CN 3 o 3 03 jo TJ 3 a a a TJ CD a. CD CD " i 16 S —' O 8 a — 3 CD .>; o i-5 H 3 o 3 TJ •§ O <D 3 ,H CD 3 <D "fe B 2 CD +3 •° s co" « II a, 11 o -3 CO J3 & § -a S cB c3 O TJ II « o ^ e ^ 3 u to CD 3 T3 CO CD <D a t-1 CO . csi o a CD -3 VH II «S 03 S O' TJ co O c3 a a o .2 OH d 2 .2 o 3 TJ o CD M .tin CO o 3 CD ^ 3 ^ O M g S sD 'a ;§ 3 U ° I co > "rrt O a M 3 II .a o Cs] VO _ : II : II: : ditL 1 I 5 X o _ _ 111111111 CO. QO; : • I ; ; •» : S-2 m Ml H ill: : I* a. m ::: _V Oi:0i: ::£:::C0:::iO:: |9A9| OjLjCJOJJ, ft « -i 8 s * T3 -3 3 ^ T3 *- —' o r2 S S 4.2 The fate of beta radioactivity. 4.2.1 Mapping the fate of beta radioactivity. In Figure 4.2 the fate of beta radioactivity has been mapped onto the flow chart from Figure 4.1, omitting the flows for simplicity of presentation. The level of beta radioactivity was monitored in 15 of the 27 functional groups defined in the model. The radioactivity in detritus was assumed to be equal to the level in benthic primary producers (taking the biomass into account) as the latter is the main source of detritus (see Table 4.1). The level in phytoplankton, the other, but less important primary producer group in the system, was derived in a similar manner. 4.2.2 Simulating the fate of beta radioactivity. The results of the simulations are shown in Figure 4.3 and 4.4. Figure 4.3 illustrates the simulated trends of beta radioactivity predicted from the original and the re-calibrated Ecopath model, respectively, while Figure 4.4 features the results of modifying only the columns or the rows of the predation mortality matrix in the minimization process (see section 3.3.4). Modifying the columns thus moves the curves vertically up or down (Figure 4.4a), while modifying the rows changes the slope of the curves (Figure 4.4b). To simulate the uptake of beta radioactivity by the organisms in the first few days after the nuclear detonation, Equation (3.4) was integrated over time steps of 0.1 days. After day nine, daily time steps were used. In the minimization process, Solver was preprogrammed to vary the scaling factors (for modifying the columns of the predation mortality matrix, see section 3.3.4) within the range of 0.25 to 1.75. If a wider range was used, the resulting predation mortality matrix could subsequently not be used for re-calibration of the original Ecopath model without seriously violating the concept of mass-balance. The model is parameterized from the top-down so that the biomass flows at the lower trophic levels are estimated to match the food demand of the upper levels. As a consequence, the predation mortalities generated by miscellaneous piscivorous fish (ultimate top predator) on their prey was restricted to either stay the same or decrease (scaling factor < 1). 64 m o p. o 1 o Vt OI o : ::*""*:: ;::;;;;:o:: H I 1-1 CD o > rt § .2 G T3 W rt ^ ^ • ° O U-i <+H >r> H — <^ "H E 2 CD o * I o § CD cf £ T3 O o OH 03 JD "o 3 fl CD .as ft* o S CD toO C CN T3 <D 23 O 3 3fl bO O o E ab S in 4.2.3 Re-calibrating the Ecopath model. Table 4.2 summarizes the changes to the predation mortality matrix required for minimizing the residuals (see section 3.3.4) and re-calibrating the Ecopath model by modifying the input biomasses. The modifications were justifiable as the biomasses were associated with high uncertainties, most of them derived from the study by Odum and Odum (1955) (see section 1.4, 3.2.2, and 3.2.3) and based on visual censuses (the fish groups, see section 3.2.3.1 and 5.1.2.1) or generated by the Ecopath program (zooplankton). The re-calibrated Ecopath model was very close to balancing. However, it was necessary to modify the scaling factors for miscellaneous piscivorous fish, snappers / groupers, and small carnivorous fish from 1 to 0.5, 0.25 to 0.5, and 0.67 to 0.8 respectively (see Table 4.2). As a consequence the sum of squared deviations (SSQ) between the observed and predicted levels of beta radioactivity (see section 3.3.4) increased from 1.02-102 to 1.14-102. A slightly better fit (reducing SSQ to 1.13 102) was obtained when Solver was allowed to also modify the 'detritus mortality' (the part of the mortality of a group caused by non-predation losses such as diseases, starvation, etc. See section 5.3.3), increasing it for all groups by a common factor of 1.18. This, however, was not used for the re-calibration of the model (see section 5.3.3). 100000 0.1 1 10 100 1000 Days after nuclear detonation 1E+13 (b) 'J 1E+11 & & 1E+09 o ca o il 1E-KJ7 ca o 100000 0.1 1 10 100 1000 Days after nuclear detonation Figure 4-4. Result of modifying either the columns or the rows of the predation mortality matrix, respectively, in the minimization process, here illustrated for surgeonfish. The dots are the observed levels of beta radioactivity, thin lines the simulated trends using the predation mortality matrix from the original model, and thick lines the simulated trends obtained by modifying only the columns (a) or the rows (b) of the original predation mortality matrix. 68 Table 4.2. Scaling factors generated by Solver for modifying the predation mortalities, i.e., columns of the predation mortality matrix (and re-calibrating the original Ecopath model), resulting in the best fit between the observed and predicted levels of beta radioactivity. Functional group Scaling factor Miscellaneous piscivorous fish 1.00a Snappers / groupers 0.25a Parrotfish 1.75 Large carnivorous fish 1.75 Small carnivorous fish 0.67a Herring 0.25 Gastropods 1.75 Butterflyfish 0.55 Small omnivorous fish 1.75 Corals 1.75 Holothurians 1.75 Zooplankton 1.75 Giant clams 0.91 a. Later modified to 0.5 for miscellaneous piscivorous fish and snappers / groupers, and 0.8 for small carnivorous fish. Finally, the biomass of phytoplankton, which was a model estimate (section 3.2.1.3), was increased by a factor of 1.3 to accommodate an increased predation pressure from zooplankton. The adjusted biomass was still within the range of phytoplankton biomass estimates in other models of similar reef ecosystems (Arias-Gonzales 1993, Opitz 1996). The basic estimates of the re-calibrated model are shown in Table 4.1 together with the estimates of the original model. The 'additional mortality' (IVT) (added to the summed predation mortality of a groups, see section 3.3.4), set by Solver in the course of minimization and shown in Table 4.3, cannot readily be related to the Ecopath outputs, a point to which I return in the Discussion (section 5.3.3). 69 Table 4.3. 'Additional mortalities' (M+) added to the summed predation mortality of the groups. Group 'Additional mortality' (M+) Phytoplankton 0.000 Benthic primary producers 0.003 Sessile invertebrates 0.047 Bivalves 0.053 Giant clams 0.023 Zooplankton 0.000 Foraminiferans 0.038 Echinoderms 0.066 Miscellaneous crustaceans 0.052 Surgeonfish 0.394 Shrimps 0.046 Holothurians 0.015 Polychaetes 0.027 Stomatopods 0.026 Corals 0.153 Small omnivorous fish 0.602 Butterflyfish 0.267 Large omnivorous fish 0.010 Gastropods 0.003 Herring 0.078 Small carnivorous fish 0.285 Cephalopods 0.125 Large carnivorous fish 0.064 Parrotfish 0.603 Snappers / groupers 0.082 Miscellaneous piscivorous fish 0.049 4.2.4 Beta radioactivity and trophic levels. The result of the simulation (Figure 4.3) indicates that groups at higher trophic levels are delayed in reaching their maximum level of measured beta radioactivity compared with groups at lower trophic levels. The relationship is shown explicitly in Figure 4.5, which is a regression of the trophic level of the functional groups as a function of the number of days required for them to reach their maximum level of radioactivity. The result of the simulation (Figure 4.3) also indicates that the maximum level of beta radioactivity in the various groups diminishes with higher trophic levels. This is illustrated in Figure 4.6, which is a regression of the maximum level of radioactivity in the functional groups as a function of their trophic level. 70 CD > O Parrotfish y = 2.564 + 0.121x R2 = 0.506 Giant clams 12 Time (days) Figure 4-5. Regression of the trophic level of the 14 functional groups as a function of days required for them to reach their maximum level of beta radioactivity. Note that parrotfish and giant clams are outliers (see text). 60 m > o CO o '-3 Pi 6 r 4 V > 2 y = 6.304 - 1.017x 1^ = 0.545 12 3 4 Trophic level Figure 4-6. Regression of the maximum level of beta radioactivity (Bq-g'ww) in the functional groups as a function of their trophic level. 4.3 Parameter estimation and network analysis. 4.3.1 Summary statistics. Table 4.4 lists selected summary statistics of the original and re-calibrated model, respectively, computed by Ecopath and useful for comparing the models with each other and with other reef or non-reef ecosystems. Several of the parameters are quantifications of Odums (1969) 24 ecosystem attributes for assessing ecosystem development and maturity. 4.3.2 Transfer efficiencies. A trophic aggregation routine in Ecopath reverses the calculation of fractional trophic levels and quantifies the trophic flows on discrete trophic levels sensu Lindeman (1942) (Christensen and Pauly 1992a). It hereby becomes possible to estimate the transfer efficiency between successive trophic levels and the result of the analysis for the re-calibrated model is shown in Table 4.5. The transfer efficiencies may be further split into flows originating from primary producers and from detritus. 71 Table 4.4. Summary statistics computed by Ecopath and useful for comparison of the original and the re-calibrated model. Dashes indicate that the corresponding parameter has no dimension. Parameter3 Symbol Unit Original model Re-calibrated model Sum of all consumption - t-km"2-year"1 10187 15468 Sum of all exports EXP t-km"2-year"1 3681 1082 Sum of all respiratory flows R t-km"2-year"1 5938 9008 Sum of all flows to detritus - t-km"2-year"1 11228 12837 Total system throughput T - 31035 38394 Total biomass (excluding detritus) B t-km"2 3718 3949 Total net primary production Pp t-km"2-year_1 9430 9793 Net system production P t-km"2-year"1 3492 785 'Total primary production/total respiration Pp/R - 1.59 1.09 2Total primary production/total biomass Pp/B year"1 2.54 2.48 3Total biomass/total throughput B/T year 0.12 0.10 5System omnivory index SOI - 0.24 0.25 5Fraction of total flow originating from detritus Dom.Det. - 0.66 0.71 15Finn's cycling index (% of total throughput) FCI - 12.47 18.97 16Finn's mean path length PL - 3.23 3.81 21 Nutrient conservation Oex - 0.12 0.18 a. Numbers to the left of the parameters refer to the corresponding ecosystem attribute in Table 1 in Odum (1969). Table 4.5. The transfer efficiency (%) by trophic level. Source \ TL I II III IV V VI VII From producers 10.3 9.5 14.1 6.3 4.7 3.0 From detritus 13.2 10.9 14.5 6.2 4.6 3.0 All flows 12.9 10.8 14.4 6.2 4.6 3.0 4.3.3 Mixed trophic impact. Figure 4.7 shows the direct and indirect impact, in relative terms, that the groups in the rows have on the groups in the columns. A positive impact is indicated by a solid bar pointing upwards, while a negative impact is indicated by a gray bar pointing downwards. The figure is from the re-calibrated model. 72 IMPACTED GROUP ^ •« -s -s W CO CO CO l3 C+H c+_, >>£:£: 6 6 g 9 op o c _ fc « eg g on ca •J oo o ~ IS 2 >• CL *H c/3 CQ C .2 o >*3 u -g 3 ca 00 CL, o -o o ca CD b3 5 O q CM W CL O I O CL o T3 o cn ^ca co ~ ca O CQ O 55 T3 o ca 6 -a S O PH O S CO C O CL CL o Misc. pisciv. fish Herring Small carniv. fish Large carniv. fish Small omniv. fish Large omniv. fish Snappers / groupers... Butterflyfish Surgeonfish Parrotfish Polychaetes Echinoderms Holothurians Sessile invertebrates... Shrimps Misc. crustaceans Gastropods Bivalves Giant clams Stomatopods Zooplankton Foraminiferans Corals Phytoplankton Benthic primary prod. Cephalopods Detritus •' II II I I II |M| I I I Figure 4-7. Mixed trophic impact diagram from the re-calibrated Ecopath model. Solid bars pointing upwards indicate the relative positive impact and gray bars pointing downwards the relative negative impact that the groups in the rows have on groups in the columns. 73 5. Discussion The following discussion is structured into three parts. First some of the model input parameters are discussed, including the time span covered by the model and the fish biomass estimates. Also discussed here is the role of herbivorous and planktivorous fish on the reef. The second part deals with model outputs, in particular the role of detritus recycling and ecosystem maturity. Finally, the re-calibration process and the use of radioactivity data are discussed, as well as the result of the simulation and some potentials of the approach. 5.1 Model input parameters. 5.1.1 The time span covered by the model. The Ecopath model was constructed with data from a variety of sources, representing work spanning several decades. Within the same period, the atoll was used for nuclear testing and whole islands disappeared, others were burned to the ground, and yet others were completely restructured to accommodate military facilities (Anon. 1991, Devaney et al. 1987). Under these circumstances it seems appropriate to ask whether it was legitimate to use the data collected within this period for the synthesized of a steady-state / mass-balance model. Few signs of destruction, however, were observed in the marine environment. In atmospheric tests (41 of the 43 test at Enewetak, Table 2.3), the marine organisms were partly protected by the water. While the explosions at low tide killed exposed organisms, e.g., those living in the intertidal zone, organisms living beyond the algal ridge or hiding under rocks and corals were hardly damaged at all (Bablet and Perrault 1987a). Dead fish were observed repeatedly in the vicinity of the detonation sites immediately after a nuclear test (Bablet and Perrault 1987a), but were not observed for extended periods of time, as might have been expected from the elevated levels of radioactivity (Seymore 1960). Weakened fish would quickly have been removed by predators, and therefore, would not have been observed. However, it was still believed that the increase in radioactivity was not enough to kill the fish directly (Seymore 1960). In addition, of the thousands of fish that were examined, none showed apparent signs of internal damages, the only exception being the thyroids where damages ranged from zero to a hundred percent (Gorbman and James 1963). 74 Superficially, the damaged fish appeared normal (Seymore 1960). Genetic effects, which should show up in the progeny, are not easily studied in situ, but requires that the natural variability of the populations is perfectly known (Bablet and Perrault 1987b). Regarding lower trophic level organisms, Knutson and Buddemeir (1973) found that "results to date indicate that the macroscopic growth rates and patterns of [massive] corals are relatively unaffected by uptake of the observed amounts of radioactivity..." Moreover, the absorbed lethal radiation dose has been shown to be inversely related to the evolutionary level of the organism, implying that less derived species are less sensitive to radiation (Bablet and Perrault 1987b). Based on these observations I therefore conclude that it was legitimate to use a wide range of data for the construction of the model, and thus assuming that the reef ecosystem did not change within the modeled period. It is very likely that the reef ecosystem has been more harmfully affected by physical damages than by radioactivity. A constant leakage of silt from the craters would clog up the coral polyps and reduce water visibility, reducing photosynthesis. However, this problem has to my knowledge not been reported for the part of the reef modeled here. 5.1.2 Fish biomass and abundance estimates. 5.1.2.1 Visual census and rotenone sampling. The fish biomass estimates (Table 3.17) were either estimated by the program or derived from the Odums' study (1955). The latter used rotenone sampling and visual census based on a method by Brock (1954) (see section 3.2.3.1 and Appendix 5). According to Randall (1963) and Smith (1973 a), visual census, however, is of limited value for quantitative studies on coral reefs. Humans errors are involved in estimating the numbers and sizes of fish and secretive, hole-dwelling, and nocturnal species are not observed. Later, in a review of the method himself, Brock (1982) found that it "underestimates both the most cryptic as well as the most abundant fish species... [and that] only diurnally exposed fish species are censused with any accuracy using the visual census technique." He then compared the results obtained with visual census with results obtained using rotenone, and found the following power-75 relationship: Y = 0.74 • X115, where X is the number of individuals visually assessed and Y is the number removed by rotenone. Rotenone sampling, however, is not without problems either, and repeated sampling is necessary to obtain a reliable population estimate (Smith 1973a). An idea of the real population size can be obtained with a Leslie plot, as shown by Pauly (1984) for two coral reef species (Figure 5.1). Here, the catch/effort (where effort is the application of poison) has been plotted as a function of the cumulative catch, and the intercept of the regression with the abscissa is an estimate of the real population number (N0). The regressions indicate that only about one third of the populations are sampled with the first application of poison. 8 c • Gramma loreto of effort 6 < • Kaupichthys hyoproroides Catch/uit 4 2 N0 1 0 0 5 10 15 20 25 Cumulative catch Figure 5-1. Leslie plots for reef eels {Kaupichthys hyoproroides) and aFairy basslet (Gramma loreto) from an isolated Bahamian reef patch, with estimates of virgin population sizes (redrawn and modified from Figure 6.1 in Pauly (1984); amis-labeled Thalassoma bifasciatum (Bluehead wrasse) in Pauly (1984)). The arrows identify the original population sizes (N0). 5.1.2.2 Fish biomass estimates at Enewetak A toll. It can be inferred from the discussion above that Odums' fish biomass estimates of 6 g dwm"2 or 26 g ww-m"2 (Table 3.17, Appendix 5) are highly conservative. The real stock size is more likely in the area of 80 g ww-m"2 or even higher. The problem was recognized by the authors who wrote that "moray eels were estimated from the rotenone samples on the surely underestimating assumption that all the morays had climbed out into the channels to die." Both the original and the re-calibrated Ecopath model, however, only estimated a fish biomass of 35 g ww-m"2 (Table 4.1). This low estimate could be the result of using the Odums' 76 estimates as input, and balancing the model with Ecoranger allowing for a variability of 99%. Though this might seem as a high variability, it is not enough to capture the 'real' stock size. Furthermore, as the model is parameterized from the top-down, the rest of the system is balanced so that the biomass flows at the lower trophic levels match the food demands of the upper levels. Low input values in the top therefore 'scale down' the whole system. Moreover, in the re-calibration process, the biomass estimates were only allowed to vary within 75%o, and for many of the lower trophic level groups they were not allowed to vary at all. The overall result, given the constraints, was therefore that the 'real' stock size was never simulated. 5.1.2.3 Comparing the standing stock of coral reef fish. Table 5.1 lists the standing stock of fish on various reefs. Again, these estimates are probably too low. However, assuming that they are too low by the same factor, it is still possible to compare among them. Except for Looe Key, the standing stock varies by an order of magnitude, from 26 to 243 g ww-m"2, with Enewetak Atoll at the lower end of the scale. Sorokin (1993) presents similar results. Randall (1963) listed four reasons explaining why the standing stock might differ between reefs. The first of these was that the standing stock is largely determined by the amount of cover / shelter afforded by the reef. Also, since reefs typically differ substantially in the degree of sculpturing, measuring only the horizontal plane introduces considerable error when comparing among reefs. The type of benthic growth is a third factor affecting the stock size; finally, the fishing effort will tend to reduce biomasses. 5.1.2.4 The abundance and role of herbivorous fish. The shallow water fish fauna at Enwetak Atoll, as well as on many other reefs, is dominated by herbivorous surgeonfish (Acanthuridae) and omnivorous parrotfish (Scaridae) that invade the reef flat with the incoming tide [present study; Odum and Odum 1955, Bakus 1967, Wiebe et al. 1975, Ogden and Lobes 1978, Lewis 1981, Miller 1982, Chartock 1983b). 77 Table 5-1. Standing stock of fish on coral reefs in different regions. Location Standing stock (g wwm"2) Source Windward reef, Enewetak Atoll, (Ecopath) 35 Present study Windward reef, Enewetak Atoll 26 (Odum and Odum 1955) Barrier reef, Moorea (Ecopath) 243 (Arias-Gonzales 1993) Fringing reef, Moorea (Ecopath) 146 (Arias-Gonzales 1993) French Frigate Shoals, Hawaii (Ecopath)b'c 24 (Polovina 1984) Keahole Pt., Kona, Hawaii0 185 (Brock 1954) Hawaii, mean of Table 1 in Brock (1954) 20 (Brock 1954) (not incl. Keahole and Rabbit Island)0 Fringing Reef, St. John, Virgin Islands3'0 160 (Randall 1963) 158 (Randall 1963) Fringing reef, Virgin Islands (Ecopath) 104 (Opitz 1996) Looe Key, Florida (Ecopath) 785 (Venier 1997) Patch reef, Bermuda 49 (Bardach 1959) a. Not including small pelagics and large apex predators; b. Only including reef fishes; c. Fished. The diet of parrotfish is somewhat controversial (Randall 1963, 1974, Bakus 1967, Smith and Paulson 1974). Contrary to most studies, Hiatt and Strasburg (1960) found an abundance of coral polyps in the stomach of parrotfish from the Marshall Islands, and characterized them as grazing omnivores. In a comment, Randall (1967) wrote: "The greater utilization of coral by scarids in the Marshall Islands noted by Hiatt & Strasburg may be related to the high coral cover of the reefs." After having dived at Enewetak, Randall (1974) later wrote: "Within the last two years the author has dived at both Heron Island on the Great Barrier Reef and Eniwetok in the Marshall Islands and finds it difficult to explain the apparent difference in the amount of coral in the diet of scarids at these two localities. Most of the scarid species are common to both islands, and there appears to be no notable overall difference in the amount of coral cover. The coral growth can vary enormously, of course, among the marine environments of the islands. Possibly Hiatt and Strasburg's specimens were collected mainly from a zone of heavy coral but little algal growth" (Randall 1974). In retrospect, the diet composition of parrotfish used in the present study (70% corals, 30% benthic primary producers, based on Hiatt and Strasburg (1960), see Appendix 2) is therefore questionable. The diet composition is also partly responsible for the high trophic level of parrotfish (3.46, see Table 4.1) estimated by Ecopath. Another reason for this high trophic level is that, in the 78 version of Ecopath used here, trophic levels were computed as: 1 + mean trophic level of prey. Since autotrophy is not included in the calculation, the trophic level of partly autotrophic organisms like giant clams and corals is in fact too high, which in turn affects the trophic level of the organisms feeding on them, e.g., parrotfish. Another controversy is the dominance of herbivorous fish on coral reefs in general (Odum and Odum 1955, Bardach 1959, Randall 1963, Bakus 1966, Ogden and Lobes 1978, Sorokin 1993). Bouchon-Navaro and Harmelin-Vivien (1981) found that, in contrast to Odums' study, most authors have estimated that herbivorous fish constitute between 10 and 15% of the fish fauna on coral reefs, both in terms of numbers and biomass. Likewise, Bakus (1966) found that coral reef fish generally consist of roughly 25%> herbivores and 65%> carnivores. He based this on the three studies described below, which I have consulted without being able to reach the same conclusion: • Bardach (1959), working on a patch reef in Bermuda, found that the weight of carnivores (many feeding on a close by seagrass bed) was almost twice that of omnivores, but on a large reef found that herbivorous / omnivorous fish outweighed the carnivores by about nine to one; • Randall (1963, 1967) found that Scaridae appear to be the largest family by weight in most tropical reef areas; and • Hiatt and Strasburg (1960) found that carnivores in the Marshall Islands dominated by number, however, both the present study and the study by Odum and Odum (1955) estimated that herbivorous / omnivorous fish dominate by weight. Supporting the findings at Enewetak, Sorokin (1993) reported that herbivorous "reef fish form one of the most important trophic guilds, which includes some 10 - 20% of the total species and 15 - 50%> of total fish biomass... Sometimes their biomass comprises over 50%> even of the total..." Finally, Colin (1987a) wrote that the "general lack of herbivores as significant as fishes at Enewetak presents an interesting contrast to reefs in some other areas of the world." There is little doubt that herbivorous fish are much more abundant on coral reefs than they are in temperate waters (Sorokin 1993), and play the important role of channeling primary productivity up the food web (Ogden and Lobes 1978). Surgeonfish are, according to Hiatt and Strasburg (1960), the most important group on the reef in converting primary productivity 79 into animal tissue. The model predicted that of the benthic primary production that is grazed directly, 64% is grazed by invertebrates and 36% by fish. Of the latter, 67% is grazed by surgeonfish alone. The assimilation efficiency of surgeonfish, however, is quite low (Chartock 1983b). Most of the algae they eat are recycled into detritus before they, in the form of DOM and POM (see section 3.2.1.1), are utilized by the various suspension and sediment feeders (Chartock 1983b) that are so abundant on the reef (see Table 4.1). Interestingly, the ecotrophic efficiency (EE) of the herbivorous / omnivorous surgeonfish, parrotfish, and butterflyfish are particularly low (0.12 - 0.14), implying that the majority of their production is recycled to the detritus pool. Similar observations have been made on other coral reefs (Opitz 1996, Parrish et al. 1986 (cited in Opitz)), and the explanation is not clear. It could perhaps have something to do with the behavior of these fish. Herbivorous fish are most active during daytime when predators are also more visible and easier to escape. Furthermore, herbivorous fish tend to concentrate in the shallower parts of the reef where the primary productivity is highest (Ogden and Lobes 1978), and it could be that their predators are not well adapted to the physical conditions at these shallow depths. For example, studies of some coral reefs in the Red Sea have shown that 71% of the herbivorous fish concentrated at 0 to 5 m of depth, while the remaining 29% occurred between 10 and 40 m (Bouchon-Navaro and Harmelin-Vivien 1981). 5.1.2.5 The fish fauna and zooplankton on the fore reef. Aside from herbivorous fish, coral reefs are characterized by a high abundance of planktivorous fish feeding on resident and oceanic zooplankton (Hobson and Chess 1978, Sorokin 1993). The oceanic zooplankton provides an external input of energy and nutrients to the reef, and the feces of the fish feeding on them provides food for corals and other benthic filter feeders (Sorokin 1993). According to Williams (1991), the "traditional view of coral reefs as energetically self-contained ecosystems occurring only in clear oceanic.waters (Odum and Odum, 1955) suggests little environmental variability and relatively little variation in the structure of fish communities among reefs. This view is grossly in error. As far as fishes are concerned, coral reefs are not energetically self-contained. A major proportion of the fish 80 biomass feeds on zooplankton derived from an external source - the waters surrounding the reef." This was clearly shown in a study by Hamner et al. (1988) on the fore reef of Davies Reef, central Great Barrier Reef. Here they found that oceanic zooplankton was consumed by a "wall of mouth" formed by the many planktivorous fish on the fore reef. "Most of the zooplankton in this water is captured and eaten by planktivorous fish which in turn defecate onto the reef surface, a process which enhances the growth of corals and benthic algae. Breaking waves tear benthic algae off the reef crest and the floating assemblage of algal fragments and fecal material which has not yet settled flows across the reef top. It is this admixture of algae, feces, and even sand which most previous investigators [e.g., at Enewetak Atoll] have treated as if it were oceanic in origin (Johannes and Gerber 1974)" (Hamner et al. 1988). That the situation is probably the same at Enewetak Atoll is supported by some observations by Hobson and Chess (1978, 1986) (see also section 3.2.1.4). In places of strong currents at the lagoonward side of the windward reef they observed that the abundance of oceanic zooplankton increases dramatically at night, and explained it as a consequence of a general rise of zooplankton towards the surface waters at night in the open sea. Inferring from the study by Hamner et al. (1988) it could, however, also be the consequence of reduced predation of zooplankton on the fore reef at night, as most planktivores fish are diurnal (Hamner et al. 1988). If this is the case, it would in turn explain the high food consumption, estimated by the program, of zooplankton by corals which mainly takes place at night (see section 3.2.1.5) Generally, the fore reef is physically very difficult to monitor and no reliable fish biomass estimates are available (Harmelin-Vivien 1977). However, Harry (1953, cited from Stevenson and Marshall (1974)), e.g., "offered the general impression that the outer reef at Raroia Atoll supported fifty per cent of the fish population of the entire atoll..." 81 5.2 Outputs of the model. 5.2.1 The role of benthic primary producers. Unlike most aquatic ecosystem where the secondary production is typically phytoplankton driven, benthic primary producers are the chief supporter (directly or indirectly through detritus, see below) of the secondary production on coral reefs (Lewis 1981, Polunin and Klumpp 1992b, Polunin 1996). Therefore, despite the visible dominance of corals, benthic primary producers are the most abundant organisms in terms of living biomass (Table 4.1). The primary production on coral reefs is generally very high, and they rank among the most productive ecosystems in the world (Odum and Odum 1955, Lewis 1977, Sorokin 1990). At Enewetak Atoll, as on many other reefs, the bulk of the benthic primary production (93%) is not consumed directly, but is recycled to the detritus pool. The 7% that is consumed / grazed directly is a relatively low fraction compared to other coral reef systems, e.g., 22 to 33% on Tiahura reef, Moorea Island, French Polynesia (Arias-Gonzales et al. 1997), 36% at a Virgin Island reef (Opitz 1996), 33% in the Looe Key's (Venier 1997), and 43 and 65% on the reef crest of Davies Reef, central Great Barrier Reef, (Polunin and Klumpp 1992a). The 7% from Enewetak, however, is a weighted mean of all the reef zones, including the sand / shingle zone which comprises 57% of the total area included in the Ecopath model (see Table 3.1). The intensity of grazing probably varies considerably between the zones, highest on the outer reef flat (Bakus 1967, Miller 1982, Chartock 1983b, Polunin 1996) and lowest in the sand / shingle zone. 5.2.2 The role of detritus. The ecotrophic efficiency of most groups in the system is relatively low (Table 4.1), suggesting that much of their production is recycled directly to the detritus pool. In turn, approximately 92% of the detritus is recycled, and 71% of all flows in the system originates from there (Table 4.1 and 4.4). The secondary production is thus largely dependent on detritus, a characteristic of coral reefs in general (Sorokin 1990, Arias-Gonzales et al. 1997, Opitz 1996, Venier 1997). A high degree of recycling is also a characteristic of mature ecosystems according to Odum (1969), who hypothesized that "food chains become complex 82 webs in mature stages, with the bulk of biological energy flow following detritus pathways... heterotrophic utilization of primary production in mature ecosystems involves largely a delayed consumption of detritus" Some of the detritus, such as benthic algal fragments torn loose at the fore reef, is consumed directly by various herbivores and detritivores (including fish) (Wiebe et al. 1975). The bulk, however, is channeled via bacteria into a microbial loop (Arias-Gonzales et al. 1997, Azam 1998) whereupon it again becomes accessible to the abundance of benthic filter-feeders including polychaetes, sessile invertebrates (e.g., sponges), miscellaneous crustaceans, bivalves, zooplankton, foraminiferans, corals, and others, that characterizes the reef. The importance of detritus can also be seen in Figure 4.7 which shows that detritus has a positive impact on most groups in the system. Few fish consume detritus directly, but many of their prey do and detritus therefore still has a positive impact on fish, though indirectly. 5.2.3 Comparing with other models. 5.2.3. J Trophic transfer efficiencies. The transfer efficiencies between successive trophic levels were presented in Table 4.5. The program estimated relatively high efficiencies at trophic levels II, III, and IV, higher for flows originating from detritus than from primary producers. This indicates that the energy context of the organic material is higher and more accessible to the organisms after it has been processed / enriched by bacteria. At higher trophic levels (V, VI, VII), the efficiencies are considerably lower, and overall decreasing. There is no longer any difference between flows that originated from detritus and from primary producers. As organisms tend to be more mobile at higher trophic levels, increasingly more energy is lost through respiration at the expense of being transferred up the system. Opitz (1996) reasoned that "the reefs 'strategy' is not to achieve high transfer efficiencies between trophic levels but to build up biomass through maintenance of short cycles for an effective recycling of matter back to its base, the detrital pool ..." In Table 5.2, the transfer efficiencies from Enewetak Atoll are compared with those of three other non-fished coral reef ecosystems. There seems to be no clear overall trend, except that the efficiencies always are higher at lower trophic levels. Since the transfer efficiencies depend 83 strongly on the gross food conversion efficiencies, GE (i.e. the fraction of the consumption of a group that is channeled toward production), and since these are not standardized for similar functional groups in different ecosystem models, the variation among systems tends to be high (Christensen and Pauly 1993). Table 5-2. Trophic transfer efficiencies (%) of four coral reef ecosystem models. Non of the systems are fishes to any extent. System II III IV V VI VII VIII IX Enewetak3 12.9 10.8 14.4 6.2 4.6 3.0 - -FFS, Hawaiib 10.1 4.0 4.1 3.3 - - - -Virgin Islands0 13.1 16.6 9.9 9.0 10.9 11.0 - -Looe Keyd 25.7 29.4 14.9 7.6 8.8 9.0 7.4 4.8 a. Present study; b. French Frigate Shoals (Christensen and Pauly 1993); c. Opitz (1996); d. Florida (Venier 1997). 5.2.3.2 Biomass by trophic level. The biomass at successive trophic levels, estimated in the present study and by Odum and Odum (1955), are compared in Table 5.3 and graphically in Figure 5.2. Table 5-3. Biomass at discrete trophic levels for the windward reef ecosystem at Enewetak Atoll as estimated by Odum and Odum (1955) and in the present study. Trophic level I II III IV V VI VII Total Odum and Odum (1955)3 2812 528 44 - - - - 3384 Ecopath (this study) 3198 372 356 18 6 0.4 0.02 3950 a. Assuming dw = 25% ww. @ 0.5 Wet weight (t/km2) U C -11 H-132 PRODUCERS - 703 Figure 5-2. Biomass pyramids of the marine food web of the windward reef at Enewetak Atoll. The pyramid to the left is redrawn and modified from Odum and Odum (1955), and shows the average biomass (in g dry weight) for the reef where H = herbivores and C = carnivores. The pyramid to the right was created using a routine in Ecopath and is scaled such that the volume of each layer is proportional to the biomass on the corresponding discrete trophic level. 84 The total biomass is very similar which is no surprise, as many of the input data for the Ecopath model were based on the Odums' data. The biomass in the present study, however, is distributed over seven trophic levels as compared to three in Odums' study. The more detailed trophic structure is the result of estimating the trophic levels instead of assigning them. In the latter case, the complex food web interactions that characterizes most ecosystems (see Figure 4.1) are not taken into account, which in turn has important implications for the predictability of such models (see section 1.1). 5.2.3.3 Ecosystem maturity. Quantitative measures are very useful for assessing and comparing the state and performance of ecosystems (Dalsgaard and Oficial 1998). Odum (1969) identified 24 attributes of ecosystem maturity, hypothesizing how ecosystems develop over time. On the basis of network analysis, Christensen (1992) quantified several of Odums ecosystem attributes (Appendix 6) and used a selection of them to ranked 41 steady state food web models comprising ponds, lakes, rivers, temperate and tropical coastal areas, coral reefs, tropical shelves and upwellings. Of the 41 systems, coral reefs were found to rank intermediate in maturity with lakes / rivers ranking lowest and coastal areas highest. According to Odum (1969), the capacity of an ecosystem to entrap, withhold, and cycle nutrients increases with maturity. The degree of recycling can be measured by Finn's cycling index (FCI, see Appendix 6), which expresses the fraction of the total system throughput that is recycled (Finn 1976, Christensen and Pauly 1992a, 1993). When Christensen and Pauly (1993) ranked the 41 systems, mentioned above, after this index, they found a strong correlation with the maturity ranking by Christensen (1992). Similarly, in a study of ecosystem stability, Vasconcellos et al. (1997) showed that recycling plays an important role in the maintenance of ecosystem stability as does path length, the average number of groups that a unit of flow passes through (Christensen and Pauly 1993, Vasconcellos et al. 1997). In Figure 5.3, Enewetak Atoll is compared with three other coral reef ecosystems. The figure shows that; a) coral reefs with low Pp/R (net primary production/respiration) ratio display high degree of recycling; and b) recycling is positively correlated with path length. Both trends are consistent with Odums theory of ecosystem maturity (Christensen and Pauly 1993, Odum 85 1969), and it can be inferred from the figure that Enewetak Atoll is more 'mature' or stable (Vasconcellos et al. 1997) than the three other reef systems. (a) 20 Enewetak (b) 20 Enewetak^ g 15 Looe Key g 15 Looe Key # X rr-t X <a Finn's cycling inc 10 B M 10 -Finn's cycling inc 5 0 FFS • # Virgin Is. i i i i Finn's cyclir 5 0 FFS Virgin Is.^ 1 i 01234 0246 Primary productivity /respiration Path length Figure 5-3. a) Finn's cycling index versus primary productivity/respiration ratio and; b) Finn's cycling index versus average path length for four coral reef ecosystems. Based on data in the present study, Christensen (1992), and Venier (1997). FFS is the French Frigate Shoals in Hawaii. Looe Key and the Virgin Islands are both in the Caribbean. 5.3 Simulating radioactivity. 5.3.1 The re-calibrated Ecopath model. Re-calibrating the Ecopath model, using observed data on beta radioactivity, resulted in a more 'mature' ecosystem as seen by comparing the summary statistics of the original and re calibrated model presented in Table 4.4. Several of the statistics are quantification's of Odums 24 attributes of ecosystem maturity (Odum 1969, Christensen 1992, 1995) (see also section 5.2.3.3 and Appendix 6), and the general trend supports a move towards a more mature ecosystem. For example, the degree of recycling and nutrient conservation is higher, Pp/R moves toward unity, and the average path length gets longer, all properties that are ascribed to mature ecosystems such as a coral reef. The re-calibrated model thus represents a more 'realistic' system than the original, implying that the incorporation of radioactivity has added valuable information to the model. This is an important finding that might help improve the trophic flow assessments by Ecopath models once the approach is incorporated as a general routine in the program. 86 5.3.2 Trophic transfer of radioactivity and the 'food web time lag'. Trophic transfer of persistent pollutants, such as radioactivity, within aquatic food webs should in theory imply a time lag between the observed maximum and the trophic position of the functional group, reflecting the time required for the pollutant to be moving up the food web. Maximum values should be observed first in the primary producers, then in herbivores, then in higher trophic levels, with top predators reaching their maximum concentrations last (Elliott et al. 1992). A clear example has recently been presented for the Central Baltic Sea, following the Chernobyl accident, by Dalsgaard et al. (1998). The scenario was also observed at Enewetak Atoll (Figure 4.3), though the partly autotrophic giant clams and parrotfish differed from the expected trend (see Figure 4.3 and 4.5). The reason for this is that, in the version of Ecopath used here, autotrophy was not accounted for in the calculation of trophic levels (see also section 5.1.2.4). Partly autotrophic organisms are therefore assigned a higher trophic level than they should, which unfortunately also has implication for all the predators feeding on them. A detailed analysis of the 'food web time lag' will therefore have to await until autotrophy is explicitly included into Ecopath, and thus into the computation of trophic levels. 5.3.3 Dilution effects and the additional mortality (M+). Most laboratory experiments have indicated that direct uptake of radioisotopes from the water by the organisms is more important than is food web transfer (Ophel and Judd 1966, Polikarpov 1966, Thomann 1981). Laboratory experiments, however, can only simulate simple food chain relationships, and typically place the organisms under highly unnatural conditions, such as constant concentrations of radioisotopes in the surrounding media. In their natural settings, reef organisms are part of complex food webs (Figure 4.1) and many of them are highly mobile, moving in and out of the radiation contaminated area. In addition, currents, winds, tides, and rain work to dilute the radioactivity (Welander 1957) so that its concentration is anything but uniform. Under these circumstances, "radionuclide concentration factors [the ratio of the concentration of the isotope in the organism and in the water] lose their meaning, and the operative factors become the elimination of radionuclides from the organisms and the assimilation of radioactive food by predators" (Polikarpov 1966). 87 An necessary assumption in Ecopath is that foraging takes place only within the modeled area (even when 'imports' are included as food items), as defined by the diet composition of the functional groups explicitly included in the model. Violations of this assumption undoubtedly accounts for at least some of the differences between the observed and predicted trends (Figure 4.3) both before and after re-calibration of the model. This problem of 'dilution' / 'migration' was solved by treating it as an additional mortality (M+), on top of the total mortality (Z, i.e., P/B) to which a functional group is subjected (section 3.3.4, 4.2.3, and Table 4.3). Contrary to what might have been expected, knowing that higher trophic level organisms tend to be more mobile, no correlation was found between M4" and the trophic level of the functional groups. This suggests that the physical dilution effects discussed above were more important at the windward reef than the effects of migration. This is not surprising as the reef is small and swept by relatively strong currents (Atkinson et al. 1981, Atkinson 1987). Moreover, most of the organisms are either sessile, or have restricted ranges, with fish showing strong territorial behavior. As mentioned in section 4.2.3, increasing slightly the detritus mortality / 'other mortality' (organisms dying because of disease, starvation, etc., and entered as EE, i.e., as (1 - other mortality)) of all groups by a common factor, resulted in a slightly better fit between the observed and predicted trends. This result was not used for the re-calibration of the model, but implies that all EE values should be lowered, i.e., the non-predation losses of all groups should be higher. The interpretation is not straightforward, but perhaps indicates that the organisms experienced at slightly higher non-predation mortality as a result of the increased radiation. 5.3.4 Bio-diminution of beta radioactivity. Overall, the result in Figure 4.3 shows good agreement between the model predictions and the observed trends, R2 ranging between 0.57 - 0.98. The regression in Figure 4.6 further suggests that beta radioactivity is not accumulated in the food web but diminishes with higher trophic levels. A similar result was observed in a study at Enewetak Atoll of radiation damages in fish thyroids, caused by accumulation of 131Iodine (Gorbman and James 1963). It was found that, aside from the distance from the explosion site, diet was the most important factor paralleling 88 the degree of thyroid damage. Fish at lower trophic levels were more severely affected than fish at higher trophic levels. A possible explanation, in both cases, is that the majority of isotopes produced during a nuclear detonation have very short half lives (131Iodine has a half life of approximately eight days; see also section 2.1.4, and 2.2.1) and, combined with the time lag discussed above, this leads to a reduced amount of radioactivity being transferred up the food web. This, however, does not rule out the possibility that certain isotopes are bioaccumulated with higher trophic levels. As described in section 2.2.3, selective uptake of beta emitting radioisotopes by aquatic organisms is known to occur. Another explanation, related to the former, could be that bioaccumulation is related to the way that contamination takes place, i.e., as a pulse release versus a constant release / equilibrated system. The latter would very likely give a different result (Rowan and Rasmussen 1994). 5.3.5 Potentials of the approach. Besides simulating the levels of beta radioactivity in the monitored groups, the simulation was extrapolated to also make predictions about the levels of contamination in the non-monitored groups (Figure 4.2 and 4.3). This is an important outcome as, in most cases of marine pollution, there is not enough time and / or money to monitor all groups in a system, or there is uncertainty as to which groups to monitor. For example, it was recently found that flounders, soles and mussels in Britain's largest estuary contain levels of radioactive Tritium hundreds of times higher than expected (Edwards 1998). The expected levels were based on the assumption that one kilo fish would acquire the same concentration as one liter of sea water. Scientist now suspect that the organisms accumulated the radioactivity through consumption (Edwards 1998). In a case like this it would be very helpful, prior to initiating a comprehensive field study, if one could use the data obtained so far to simulate the fate of Tritium within the whole food web, making predictions about the level of contamination in the non-monitored groups. The result could then be used as a guide as to which organisms might have accumulated critical levels of contamination and should be sampled for further testing. For this to be realized, however, the approach presented here of mapping available measurements onto Ecopath-generated food webs must first be generalized to any type of contaminant, and the minimization routine should be improved. Regarding the first point, 89 Equation (3.4) could be modified to include, for example, 'affinity factors' in cases where one is dealing with fat-soluble contaminants. The second point refers to the use of scaling factors and fixed relationships between the elements of the predation mortality matrix (Figure 4.4). Ideally, to optimize the fit between the observed and predicted levels of contamination, it should be possible to modify all the parameters of the predation mortality matrix independently of each other, to avoid the situation presented in Figure 4.4. A possible solution, suggested by Dr. C. Walters and Dr. D. Pauly (pers.com), is to add a minimization routine to the Ecoranger module of Ecopath. As described in section 2.3.2, this module already includes a Monte-Carlo resampling routine (Christensen and Pauly 1995) which could be used as an 'importance-sampling' routine to combine prior distributions from Ecopath (which are given by the uncertainties of the input data), with a likelihood function given by the fits of the observed to the predicted pollutants series. This would allow generating posterior distributions of key parameters, thus allowing their interpretation in a Bayesian context (Walters 1996). 90 References Alino, P. M., L. T. McManus, J. W. McManus, C. L. J. Nanola, M. D. Fortes, G. C J. Trono, and G. S. Jacinto. 1993. Initial parameter estimations of a coral reef flat ecosystem in Bolinao, Pangasinan, Northwestern Philippines, p. 252 - 258. In V. Christensen and D. Pauly (eds.) Trophic models of aquatic ecosystems. ICLARJVI Conf. Proc. 26. 390 p. Anon. 1944. Contour chart of Eniwetok lagoon. United States Department of the Interior Geological Survey. Professional Paper 260, chart 5. Anon. 1976a. Eniwetok Marine Biological Laboratory Contributions 1955-1974. Vol. 1. EMCL's Successor, The Mid-Pacific Marine Laboratory. 658 p. Anon. 1976b. Eniwetok Marine Biological Laboratory Contributions 1955-1974. Vol. 2. EMCL's Successor, The Mid-Pacific Marine Laboratory. 683 p. Anon. 1976c. Eniwetok Marine Biological Laboratory Contributions 1955-1974. Vol. 3. EMCL's Successor, The Mid-Pacific Marine Laboratory. 632 p. Anon. 1979. Mid-Pacific Marine Laboratory (formerly Eniwetok Marine Biological Laboratory) contributions 1975-1979. Vol. 4. The Mid-Pacific Marine Laboratory of the University of Hawaii. 703 p. Anon. 1991. Radioactive heaven and earth: the health and environmental effects of nuclear weapons testing in, on, and above the earth. Apex Press, New York. 193 p. Anon. 1998. RMI Online. Internet guide to the Republic of the Marshall Islands. Nuclear testing in the Marshall Islands: a chronology of events. http://www.rmiembassyus.org/nucchron.html. Arias-Gonzales, J. E. 1993. Fonctionnement trophique d'un ecosystem recifal: secteur de Tiahura, He de Moorea, Polynesie Francaise. Ph.D. Thesis. Ecole pratique des hautes etudes, Perpignan. 358 p. Arias-Gonzales, B. Delesalle, B. Salvat, and R. Galzin. 1997. Trophic functioning of the Tiahura reef sector, Moorea Island, French Polynesia. Coral Reefs 16: 231-246. Atkinson, M., S. V. Smith, and E. D. Stroup. 1981. Circulation in Enewetak Atoll lagoon. Limnology and Oceanography 26(6): 1074-1083. Atkinson, M. J. 1987. Oceanography of Enewetak Atoll, p. 57-70. In D. M. Devaney, E. S. Reese, B. L. Burch, and P. Helfrich (eds) The natural history of Enewetak Atoll I. The ecosystem: environments, biotas, and processes. U.S. Department of Energy, Oak Ridge, Tennessee. 228 p. Azam, F. 1998. Microbial control of oceanic carbon flux: the plot thickens. Science 280: 694-696. Bablet, J. P., and G. H. Perrault. 1987a. Effects on a coral environment of a nuclear detonation, p. 151-163. In B. Salvat (ed.) Human impacts on coral reefs: facts and recommendations. Antenne Museum E.P.H.E., French Polynesia. 253 p. 91 Bablet, J. P., and G. H. Perrault. 1987b. Radioactivity in a coral ecosystem, p. 135-149. In B. Salvat (ed.) Human impacts on coral reefs: facts and recommendations. Antenne Museum E.P.HE., French Polynesia. 253 p. Bailey-Brock, J. H, J. K. White, and L. A. Ward. 1980. Effects of algal turf and depressions as refuges on polychaete assemblages of a windward reef bench at Enewetak Atoll. Micronesica 16(1): 43-58. Bakus, G. J. 1966. Some relationships of fishes to benthic organisms on coral reefs. Nature 210: 280-284. Bakus, G. J. 1967. The feeding habits of fishes and primary production at Eniwetok, Marshall Islands. Micronesica 3: 135-149. Bakus, G. J. 1968. Defensive mechanisms and ecology of some tropical holothurians. Marine Biology 2(1): 23-32. Bakus, G. J. 1973. The biology and ecology of tropical Holothurians, p. 325-367. In O. A. Jones and R. Endean (eds.) Biology and Geology of Coral Reefs II: Biology 1. Academic Press, New York. 480 p. Bardach, J. E. 1959. The summer standing crop of fish on a shallow Bermuda reef. Limnology and Oceanography 4(1): 77-85. Basile, L. L., R. J. Cuffey, and D. F. Kosich. 1984. Sclerosponges, Pharetronids, and Sphinctozoans (relict cryptic hard-bodied porifera) in the modern reefs of Enewetak Atoll. Journal of Paleontology 58(3): 636-650. Blanchon, P. 1997. Historical development of reef study, http://www.indiana.edu/~reefpage/. Bonham, D. 1959. Further contributions on gross beta radioactivity of biological and related samples at the Eniwetok proving ground, 1952-1958. UWFL-63. Laboratory of Radiation Biology, University of Washington, Seattle, Washington. 41 p. Bonham, K. 1958. Radioactivity of invertebrates and other organisms at Eniwetok Atoll during 1954-5. UWFL-53. Applied Fisheries Laboratory, University of Washington, Seattle, Washington. 52 p. Bouchon-Navaro, Y., and M. L. Harmelin-Vivien. 1981. Quantitative distribution of herbivorous reef fishes in the Gulf of Aqaba (Red Sea). Marine Biology 63: 79-86. Brock, R. E. 1982. A critique of the visual census method for assessing coral reef fish populations. Bulletin of Marine Science 32(1): 269-276. Brock, V. E. 1954. A preliminary report on a method of estimating reef fish populations. Journal of Wildlife Management 18: 297-308. Buddemeier, R. W., S. V. Smith, and R. A. Kinzie. 1975. Holocene windward reef-flat history, Enewetak Atoll. Geological Society of America Bulletin 86: 1581-1584. Burch, B. L. 1987. Introduction, p. xvii-xviii. In D. M. Devaney, E. S. Reese, B. L. Burch, and P. Helfrich (eds.) The natural history of Enewetak Atoll II. Biogeography and systematics. U.S. Department of Energy, Oak Ridge, Tennessee. 348 p. 92 Cabana, G, and J. B. Rasmussen. 1994. Modelling food chain structure and contaminant bioaccumulation using stable nitrogen isotopes. Nature 372: 255-257. Chartock, M. A. 1983a. Habitat and feeding observations on species of Ophiocoma (Ophiocomidae) at Enewetak. Micronesica 19(1-2): 131-149. Chartock, M. A. 1983b. The role of Acanthurus guttatus (Bloch and Schneider 1801) in cycling algal production to detritus. Biotropica 15(2): 117-121. Chave, E. H., and D. M. Devaney. 1987. Recent foraminifera and nonplanktonic protozoans, p. 37-47. In D. M. Devaney, E. S. Reese, B. L. Burch, and P. Helfrich (eds.) The natural history of Enewetak Atoll II. Biogeography and systematics. U.S. Department of Energy, Oak Ridge, Tennessee. 348 p. Christensen, V. 1992. Network analysis of trophic interactions in aquatic ecosystems. Ph.D. Thesis. Royal Danish School of Pharmacy, Copenhagen. 55 p. Christensen, V. 1995. Ecosystem maturity - towards quantification. Ecological Modelling 77: 3-32. Christensen, V., and D. Pauly. 1992a. ECOPATH II - a software for balancing steady-state ecosystem models and calculating network characteristics. Ecological Modelling 61: 169-185. Christensen, V., and D. Pauly. 1992b. A guide to the ECOPATH 11 software system (version 2.1). ICLARM software 6. 72 p. Christensen, V., and D. Pauly. 1993. Flow characteristics of aquatic ecosystems, p. 338-352. In V. Christensen and D. Pauly (eds.) Trophic models of aquatic ecosystems. ICLARM Conf. Proc. 26. 390 p. Christensen, V., and D. Pauly. 1995. Fish production, catches and the carrying capacity of the world oceans. Naga, the ICLARM Quarterly 18(3): 34-40. Clark, R. B. 1989. Marine pollution. Clarendon Press, Oxford. 220 p. Colin, P. L. 1987a. Subtidal environments and ecology of Enewetak Atoll, p. 91-138. In D. M. Devaney, E. S. Reese, B. L. Burch, and P. Helfrich (eds.) The natural history of Enewetak Atoll I. The ecosystem: environments, biotas, and processes. U.S. Department of Energy., Oak Ridge, Tennessee. 228 p. Colin, P. L. 1987b. Physiography of Enewetak Atoll, p. 27-35. In D. M. Devaney, E. S. Reese, B. L. Burch, and P. Helfrich (eds.) The natural history of Enewetak Atoll I. The ecosystem: environments, biotas, and processes. U.S. Department of Energy, Oak Ridge, Tennessee. 228 p. Cutress, B. M., and F. W. E. Rowe. 1987. Holothurians of Enewetak Atoll, p. 263-275. In D. M. Devaney, E. S. Reese, B. L. Burch, and P. Helfrich (eds.) The natural history of Enewetak Atoll II. Biogeography and systematics. U.S. Department of Energy, Oak Ridge, Tennessee. 348 p. Cutress, C. E., and A. C. Arneson. 1987. Sea anemones of Enewetak Atoll, p. 53-62. In D. M. Devaney, E. S. Reese, B. L. Burch, and P. Helfrich (eds.) The natural history of 93 Enewetak Atoll II. Biogeography and systematics. U.S. Department of Energy, Oak Ridge, Tennessee. 348 p. Dalsgaard, J., Jarre-Teichman, A., Walters, C, and D. Pauly. 1998. An approach to the modelling of persistent pollutants in marine ecosystems. ICES CM. 1998/V:10, 16 p. Dalsgaard, J. P. T., and R. T. Oficial. 1998. Modeling and analyzing the agroecological performance of farms with ECOPATH. ICLARM Tech. Rep. 53. 54 p. Dawson, E. Y. 1957. An annotated list of marine algae from Eniwetok Atoll, Marshall Islands. Pacific Science XI: 92-132. Devaney, D. M. 1987a. Echinodermata other than Holothuroidea of Enewetak Atoll, p. 277-285. In D. M. Devaney, E. S. Reese, B. L. Burch, and P. Helfrich (eds.) The natural history of Enewetak Atoll II. Biogeography and systematics. U.S. Department of Energy, Oak Ridge, Tennessee. 348 p. Devaney, D. M. 1987b. Porifera of Enewetak Atoll, p. 49-51. In D. M. Devaney, E. S. Reese, B. L. Burch, and P. Helfrich (eds.) The natural history of Enewetak Atoll II. Biogeography and systematics. U.S. Department of Energy, Oak Ridge, Tennessee. 348 p. Devaney, D. M., and J. H. Bailey-Brock. 1987. Polychaetes of Enewetak Atoll, p. 97-103. In D. M. Devaney, E. S. Reese, B. L. Burch, and P. Helfrich (eds.) The natural history of Enewetak Atoll II. Biogeography and systematics. U.S. Department of Energy, Oak Ridge, Tennessee. 348 p. Devaney, D. M., and A. J. Bruce. 1987. Crustacea Decapoda (Penaeidea, Stenopodidea, Caridea, and Palinura) of Enewetak Atoll, p. 221-233. In D. M. Devaney, E. S. Reese, B. L. Burch, and P. Helfrich (eds.) The natural history of Enewetak Atoll II. Biogeography and systematics. U.S. Department of Energy, Oak Ridge, Tennessee. 348 p. Devaney, D. M., and J. C. Lang. 1987. Scleractinia (stony corals) of Enewetak Atoll, p. 67-76. In D. M. Devaney, E. S. Reese, B. L. Burch, and P. Helfrich (eds.) The natural history of Enewetak Atoll II. Biogeography and systematics. U.S. Department of Energy, Oak Ridge, Tennessee. 348 p. Devaney, D. M., E. S. Reese, B. L. Burch, and P. Helfrich, editors. 1987. The natural history of Enewetak Atoll II. Biogeography and systematics. U.S. Department of Energy, Oak Ridge, Tennessee. 348 p. Donaldson, L. R. 1953. Radiobiological studies at Eniwetok atoll before and following the Mike shot of the November 1952 testing program. UWFL-33. Applied Fisheries Laboratory, University of Washington, Seattle, Washington. 126 p. Donaldson, L. R. 1959. Radiobiological studies at the Eniwetok test site and adjacent areas of the Western Pacific. Laboratory of Radiation Biology, University of Washington, Seattle, Washington. 18 p. 94 Ebert, T. A., and R. F. Ford. 1986. Population ecology and fishery potential of the spiny lobster Panulirus penicillatus at Enewetak Atoll, Marshall Islands. Bulletin of Marine Science 38(1): 56-67. Edwards, R. 1998. Hot seafood. New Scientists 160(2): 10. Elliott, J. M., J. Hilton, E. Rigg, P. A. Tullett, D. J. Swift, and D. R. P. Leonard. 1992. Sources of variation in post-Chernobyl radiocaesium in fish from two Cumbrian lakes (north-west England). Journal of Applied Ecology 29: 108-119. Emery, K. O, J. I. J. Tracey, and H. S. Ladd. 1954. Geology of Bikini and nearby atolls. United States Government Printing Office, Washington. 265 p. Finn, J. T. 1976. Measures of ecosystem structure and function derived from analysis of flows. Journal of Theoretical Biology 56: 363-380. FishBase. 1998. FishBase 98 CD-ROM. ICLARM (see also www.fishbase.org). Garth, J. S., J. Gaig, and J. W. Knudsen. 1987. Crustacea Decapoda (Brachyura and Anomura) of Enewetak Atoll, p. 235-261. In D. M. Devaney, E. S. Reese, B. L. Burch, and P. Helfrich (eds.) The natural history of Enewetak Atoll II. Biogeography and systematics. U.S. Department of Energy, Oak Ridge, Tennessee. 348 p. Gerber, R. P., and N. Marshall. 1974. Ingestion of detritus by the lagoon pelagic community at Eniwetok Atoll. Limnology and Oceanography 19(5): 815-824. Gerber, R. P., and N. Marshall. 1982. Characterization of the suspended particulate organic matter and feeding by the lagoon zooplankton at Enewetak Atoll. Bulletin of Marine Science 32(1): 290-300. Giese, A. C, editor. 1966. On the biochemical constitution of some echinoderms, p. 757-796. In Boolootian, R.A. (ed.) Physiology of echinodermata. John Wiley & Sons, New York. 822 p. Gorbman, A., and M. S. James. 1963. An exploratory study of radiation damage in the thyroids of coral reef fishes from the Eniwetok Atoll, p. 385-399. In V. Schultz and A. W. Jr. Klement (eds.) Radioecology, Proc. First National Symp. Radioecol., Colorado State University, Fort Collins, Colorado. 746 p. Grigg, R. W. 1982. Darwin Point: a treshold for atoll formation. Coral Reefs 1: 29-34. Hallock, P. 1981. Production of carbonate sediments by selected large benthic foraminifera on two Pacific coral reefs. Journal of Sedimentary Petrology 51(2): 467-474. Hammen. 1980. Marine invertebrates: comparative physiology. University Press of New England, Hanover, England. 127 p. Hamner, W. M., M. S. Jones, J. H. Carleton, I. R. Hauri, and D. M. Williams. 1988. Zooplankton, planktivorous fish, and water currents on a windward reef face: Great Barrier Reef, Australia. Bulletin of Marine Science 42(3): 459-479. Harmelin-Vivien, M. L. 1977. Ecological distribution of fishes on the outer slope of Tulear Reef (Madagascar). Proc. Third Int. Coral Reef Symp. 1: 289-295. 95 Harry, R. R. 1953. Ichthyological field data of Raroia Atoll, Tuamotu Archipelago. Atoll Research Bulletin 18: 190. Helfrich, P., and R. Ray. 1987. Research at Enewetak Atoll: a historical perspecitve, p. 1-16. In D. M. Devaney, E. S. Reese, B. L. Burch, and P. Helfrich (eds.) The natural history of Enewetak Atoll I. The ecosystem: environments, biotas, and processes. U.S. Department of Energy, Oak Ridge, Tennessee. 228 p. Henry, T. W., and B. R. Wardlaw. 1990. Introduction: Enewetak Atoll and the PEACE Program. U.S. Geological Survey professional paper; 1513-A. United States Government Printing Office, Washington. 29 p. Heslinga, G. A., and W. K. Fitt. 1987. The domestication of reef-dwelling clams. Bioscience 37(5): 332-339. Heslinga, G. A., and T. C. Watson. 1985. Recent advances in giant clam mariculture. Proc. Int. Congr. Coral Reefs 5: 531-537. Hiatt, R. W., and D. W. Strasburg. 1960. Ecological relationships of the fish fauna on coral reefs of the Marshall Islands. Ecological Monographs 30: 65-127. Hirshfield, H. I, R. Charmatz, and L. Helson. 1968. Foraminifera in samples taken mainly from Eniwetok Atoll in 1965. Journal of Protozoology 15(3): 497-502. Hobson, E. S., and J. R. Chess. 1978. Trophic relationships among fishes and plankton in the lagoon at Enewetak Atoll, Marshall Islands. Fishery Bulletin 76(1): 133-153. Hobson, E. S., and J. R. Chess. 1986. Diel movements of resident and transient zooplankters above lagoon reefs at Enewetak Atoll, Marshall Islands. Pacific Science 40: 7-26. Hunter, H. F., and N. E. Ballou. 1951. Fission-product decay rates. Nucleonics 9(5): c2-c7. Jarman, W. M., K. A. Hobson, W. J. Sydeman, C. E. Bacon, and E. B. Mclaren. 1996. Influence of trophic position and feeding location on contaminant levels in the Gulf of the Farallones food web revealed by stable isotope analysis. Environmental Science and Technology 30(2): 654-660. Johannes, R. E. 1967. Ecology of organic aggregates in the vicinity of a coral reef. Limnology and Oceangraphy 12(2): 189-195. Johannes, R. E., and R. Gerber. 1974. Import and export of net plankton by an Eniwetok coral reef community. Proc. Sec. Int. Coral Reef Symp. 1: 97-104. Karolle, B. G. 1993. Atlas of Micronesia. Bess Press, Honolulu, Hawaii. 122 p. Kay, E. A., and S. Johnson. 1987. Mollusca of Enewetak Atoll, p. 105-146. In D. M. Devaney, E. S. Reese, B. L. Burch, and P. Helfrich (eds.) The natural history of Enewetak Atoll II. Biogeography and systematics. U.S. Department of Energy, Oak Ridge, Tennessee. 348 p. Kennish, M. J. 1998. Pollution impacts on marine biotic communities. CRC Press, New York. 310p. Kinne, O., editor. 1984. Marine ecology 5: Ocean management 3: pollution and protection of the seas - radioactive materials, heavy metals and oil. John Wiley & Sons, Chichester. 96 Kiriluk, R. M., M. R. Servos, D. M. Whittle, G. Cabana, and J. B. Rasmussen. 1995. Using ratios of stable nitrogen and carbon isotopes to characterize the biomagnification of DDE, mirex, and PCB in a Lake Ontario pelagic food web. Canadian Journal of Fisheries and Aquatic Sciences 52: 2660-2674. Kiste, R. C. 1987. History of the people of Enewetak Atoll, p. 17-25. In D. M. Devaney, E. S. Reese, B. L. Burch, and P. Helfrich (eds.) The natural history of Enewetak Atoll I. The ecosystem: environments, biotas, and processes. U.S. Department of Energy, Oak Ridge, Tennessee. 228 p. Kline, T. C, and D. Pauly. 1998. Cross-validation of trophic level estimates from a mass-balance model and 15N/14N data from Prince William Sound. Proc. 15th Lowell Wakefield Symp. (inpress). Knutson, D. W., and R. W. Buddemeier. 1973. Distributions of radionuclides in reef corals: opportunity for data retrieval and study of effects. International Atomic Energy Agency IAEA/SM-158/49. 14 p. (Also in Anon. 1976b). Kohn, A. J., and P. J. Leviten. 1976. Effect of habitat complexity on population density and species richness in tropical intertidal predatory gastropod assemblages. Oecologia 25: 199-210. Kohn, A. 1980. Populations of tropical intertidal gastropods before and after a typhoon. Micronesica 16(2): 215-228. Kohn, A. J. 1987. Intertidal ecology of Enewetak Atoll, p. 139-157. In D. M. Devaney, E. S. Reese, B. L. Burch, and P. Helfrich (eds.) The natural history of Enewetak Atoll I. The ecosystem: environments, biotas, and processes. U.S. Department of Energy, Oak Ridge, Tennessee. 228 p. Ladd, H. S. 1961. Reef building. Science 134: 703-715. Ladd, H. S. 1973. Bikini and Eniwetak Atolls, Marshall Islands, p. 93-112. In O. A. Jones and R. Endean (eds.) Biology and Geology of Coral Reefs 1: Geology 1. Academic Press, New York. Lalli, C. M., and T. R. Parsons. 1994. Biological oceanography: an introduction. Elsevier Science Ltd., Oxford, U.K. 301 p. Lawrence, J. M. 1980. Numbers and biomass of the common Holothuroids on the windward reef flat at Enewetak atoll, Marshall Islands. Echinoderms: present and past. Proceedings of the European Colloquium on Echinoderms: 201-204. 408 p. Lenssen, N. 1991. Worldwatch paper 106: Nuclear waste; the problem that won't go away. Worldwatch Institute. 62 p. Lewis, J. B. 1977. Processes of organic production on coral reefs. Biological Reviews 52: 305-347. Lewis, J. B. 1981. Coral Reef Ecosystems, p. 127-158. In A. R. Longhurst (ed.) Analysis of Marine Ecosystems. Academic Press, London. 741 p. Lieske, E., and R. Myers. 1994. Coral reef fishes: Indo-Pacific & Caribbean. Harper Collins Publishers, London. 400 p. 97 Lindeman, R. L. 1942. The trophic-dynamic aspect of ecology. Ecology 23: 399-418. Lipps, J. H., and T. E. Delaca. 1980. Shallow-water foraminiferal ecology, Pacific Ocean, p. 325-340. In M. E. Field, A. H. Bouma, I. P. Colburn, R. G. Couglas, and J. C. Ingle (eds.) Quaternary depositional environments of the Pacific coast. Pacific Coast Paleogeography Symposium 4. 355 p. Maragos, J. E., M. P. Crosby, and J. W. McManus. 1996. Coral reefs and biodiversity: A critical and threatened relationship. Oceanography 9(1): 83-99. Marsh, J. A. J. 1970. Primary productivity of reef-building calcareous red algae. Ecology 51(2): 255-263. Marsh, J. A. J. 1987. Reef processes: energy and materials flux, p. 159-179. In D. M. Devaney, E. S. Reese, B. L. Burch, and P. Helfrich, (eds.) The natural history of Enewetak Atoll I. The ecosystem: environments, biotas, and processes. U.S. Department of Energy, Oak Ridge, Tennessee. 228 p. Marshall, N., A. G. Durbin, R. Gerber, and G. Telek. 1975. Observations on particulate and dissolved organic matter in coral reef areas. Internationale Revue der gesamten Hydrobiologie 60(3): 335-345. Mauchline, J., and W. L. Templeton. 1964. Artificial and natural radioisotopes in the marine environment. Oceanographical Marine Biological Annual Review 2: 229-279. Merrill, J. T., and R. A. Duce. 1987. Meteorology and atmospheric chemistry of Enewetak Atoll, p. 71 - 90. In D. M. Devaney, E. S. Reese, B. L. Burch, and P. Helfrich (eds.) The natural history of Enewetak Atoll I. The ecosystem: environments, biotas, and processes. US Department of Energy, Oak Ridge, Tennessee. 228 p. Miller, A. C. 1982. Effects of differential fish grazing on the community structure of an intertidal reef flat at Enewetak Atoll, Marshall Islands. Pacific Science 36(4): 467-482. Odum, H. T., and E. P. Odum. 1955. Trophic structure and productivity of a windward coral reef community on Eniwetok Atoll. Ecological Monographs 25(3): 291-320. Odum, E. P. 1969. The strategy of ecosystem development. Science 164: 162-270. Ogden, J. C, and P. S. Lobes. 1978. The role of herbivorous fishes and urchins in coral reef communities. Environmental biology of fishes 3(1): 49-63. O'Neill, R. V. 1979. A review of linear compartmental analysis in ecosystem science, p. 3-28. In J. H. Matis, B. C. Patten, and G. C. White (eds.) Compartmental analysis of ecosystem models. International Co-operative Publishing House, Fairland, Maryland USA. 368 p. Ophel, I. L., and J. M. Judd. 1966. Experimental studies of radiostrontium accumulation by freshwater fish from food and water, p. 859-865. In B. Aberg and F. P. Hungate (eds.) Radioecological concentration processes. Proceedings of an International Symposium held in Stockholm. Pergamon Press, Oxford. 1040 p. Opitz, S. 1993. A quantitative model of the trophic interactions in a Caribbean coral reef system, p.259-267. In V. Christensen and D. Pauly (eds.) Trophic models of aquatic ecosystems. ICLARM Conf. Proc. 26. 390 p. 98 Opitz, S. 1996. Trophic interactions in Caribbean coral reefs. ICLARM Tech. Rep. 43. 341 p. Osterberg, C, W. G. Pearcy, and H. J. Curl. 1964. Radioactivity and its relatioship to oceanic food chains. Journal of Marine Research 22(1): 2-12. Palumbo, R. F. 1959. Gross beta radioactivity of the algae at Eniwetok Atoll, 1954-1956. UWFL-61. Laboratory of Radiation Biology, University of Washington, Seattle, Washington. 32 p. Parrish, J. D., J. E. Norris, M. W. Callahan, J. K. Callahan, E. J. Magarifuji, and R. E. Schroeder. 1986. Piscivory in a coral reef fish community, p. 285-297. In C. A. Simenstad and G. M. Cailliet (eds.) Contemporary studies on fish feeding. Dr. W. Junk Publishers, The Netherlands. Pauly, D. 1984. Fish population dynamics in tropical waters: a manual for use with programmable calculators. ICLARM Stud. Rev. 8. 325 p. Pauly, D., V. Christensen, and V. Sambilay. 1990. Some features of fish food consumption estimates used by ecosystem modellers. ICES CM 1990/G:17. 8 p. Pauly, D., V. J. Sambilay, and S. Opitz. 1993. Estimates of relative food consumption by fish and invertebrate populations required for modelling the Bolinao Reef ecosystem, p. 236-251. In V. Christensen and D. Pauly (eds.) Trophic models of aquatic ecosystems. ICLARM Conf. Proc. 26. 390 p. Polikarpov, G. G. 1966. Radioecology of aquatic organisms; the accumulation and biological effect of radioactive substances. Reinhold Book Division, New York. 314 p. Polovina, J. J. 1984. Model of a coral reef ecosystem. I. The ECOPATH model and its application to French Frigate Shoals. Coral Reefs 3: 1-11. Polovina, J. J. 1993. The first ECOPATH, p. vii-viii. In V. Christensen and D. Pauly (eds.) Trophic models of aquatic ecosystems. ICLARM Conf. Proc. 26. 390 p. Polunin, N. V. C. 1996. Trophodynamics of reef fisheries productivity, p. 113-135. In N. V. C. Polunin and C. M. Roberts (eds.) Reef fisheries. Chapman & Hall, London. 477 p. Polunin, N. V. C, and D. W. Klumpp. 1992a. Algal food supply and grazer demand in a very productive coral-reef zone. Journal of Experimental Marine Biology and Ecology 164: 1-15. Polunin, N. V. C, and D. W. Klumpp. 1992b. A trophodynamic model of fish production on a windward reef tract, p. 213-233. In D. M. John, S. J. Hawkins, and J. H. Price (eds.) Plant-animal interactions in the marine benthos. Clarendon Press, Oxford. 570 p. Randall, J. E. 1963. An analysis of the fish populations of artificial and natural reefs in the Virgin Islands. Caribbean Journal of Science 3(1): 31-47. Randall, J. E. 1967. Food habits of reef fishes of the West Indies. Studies in Tropical Oceanography 5: 665-847. Randall, J. E., and H. A. Randall. 1987. Annotated checklist of the fishes of Enewetak Atoll and other Marshall Islands, p. 289-324. In D. M. Devaney, E. S. Reese, B. L. Burch, 99 and P. Helfrich (eds.) The natural history of Enewetak Atoll II. Biogeography and systematics. U.S. Department of Energy, Oak Ridge, Tennessee. 348 p. Randall, R. E. 1974. The effects of fishes on coral reefs. Proc. Sec. Int. Coral Reef Symp. 1: 159-166. Reaka, M. L., and R. B. Manning. 1987. Stomatopod Crustacea of Enewetak Atoll, p. 181-190. In D. M. Devaney, E. S. Reese, B. L. Burch, and P. Helfrich (eds.) The natural history of Enewetak Atoll II. Biogeography and systematics. U.S. Department of Energy, Oak Ridge, Tennessee. 348 p. Reese, E. S. 1987. Preface, p. xi-xii. In D. M. Devaney, E. S. Reese, B. L. Burch, and P. Helfrich (eds.) The natural history of Enewetak Atoll I. The ecosystem: environments, biotas, and processes. U.S. Department of Energy, Oak Ridge, Tennessee. 228 p. Reiswig, H. M. 1971. Particle feeding in natural populations of three marine desmosponges. Biological Bulletin 141: 568-591. Renaud, M. L. 1976. Observations on the behavior and shell types of Cypraea moneta (Molluca, Gastropoda) at Enewetak, Marshall Islands. Pacific Science 30(2): 147-158. Riddle, M. J., D. M. Alongi, P. K. Dayton, J. A. Hansen, and D. W. Klumpp. 1990. Detrital pathways in a coral reef lagoon. Marine Biology 104: 109-118. Ristvet, B. L. 1987. Geology and geohydrology of Enewetak Atoll, p. 37 - 56. In D. M. Devaney, E. S. Reese, B. L. Burch, and P. Helfrich (eds.) The natural history of Enewetak Atoll I. The ecosystem: environments, biotas, and processes. U.S. Department of Energy, Oak Ridge, Tennessee. 228 p. Rowan, D. J., and J. B. Rasmussen. 1994. Bioaccumulation of radiocesium by fish: the influence of physicochemical factors and trophic structure. Canadian Journal of Fisheries and Aquatic Sciences 51: 2388-2410. Ruppert, E. E., and R. D. Barnes. 1994. Invertebrate zoology. Saunders College Publishing, Fort Worth. 1056 p. Russo, A. R. 1980. Bioerosion by two rock boring echinoids (Echinometra mathaei and Echinostrephys aciculatus) on Enewetak Atoll, Marshall Islands. Journal of Marine Research 38(1): 99-110. Sambilay, V. J. 1993. Proximate analysis of selected organisms important for constructing trophic models of ecosystems, p. 353-355. In V. Christensen and D. Pauly (eds.) Trophic models of aquatic ecosystems. ICLARM Conf. Proc. 26. 390 p. Sanders, R. W., and C. C. Gilmour. 1994. Accumulation of Selenium in a model freshwater microbial food web. Applied and Environmental Microbiology 60(8): 2677-2683. Sargent, M. C, and T. S. Austin. 1949. Organic productivity of an atoll. Transactions, American Geophysical Union 30(2): 245-249. Schultz, L. P. and collaborators 1953. Fishes of the Marshall and Marinas Islands. I Smithsonian Institute United States National Museum Bulletin 202. United States Government Printing Office, Washington. 685 p. 100 Schultz, L. P. and collaborators. 1960. Fishes of the Marshall and Marinas Islands. II Smithsonian Institute United States National Museum Bulletin 202. United States Government Office, Washington. 438 p. Schultz, L. P. and collaborators. 1966. Fishes of the Marshall and Marinas Islands. Ill Smithsonian Institute United States National Museum Bulletin 202. United States Government Printing Office, Washington. 176 p. Seymore, A. H. 1960. Fish and radioactivity. Laboratory of Radiation Biology, University of Washington, Seattle, Washington. 83 p. Silvestre, G, S. Selvanathan, and A. H. M. Salleh. 1993. Preliminary trophic model of the coastal fisheries resources of Brunei Darussalam, South China Sea, p. 300-306. In V. Christensen and D. Pauly (eds.) Trophic models of aquatic ecosystems. ICLARM Conf. Proc. 26. 390 p. Skarsgard, L. D. 1997. Radiation biophysics. Lecture notes. University of British Columbia. Department of Physics. Smith, C. L. 1973a. Small rotenone stations: a tool for studying coral reef fish communities. American Museum Novitates 2512: 1-21. Smith, C. M. 1994. Radionuclide safety and methodology. Reference manual University of British Columbia. Department of health, safety and environment, Vancouver, Canada. 40 p. Smith, R. L., and A. C. Paulson. 1974. Food transit times and gut pH in two Pacific parrotfishes. Copeia3: 796-799. Smith, S. V. 1973b. Carbon dioxide dynamics: a record of organic carbon production, respiration, and calcification in the Eniwetok reef flat community. Limnology and Oceanography 18(1): 106-120. Smith, S. V., and J. A. Marsh. 1973. Organic carbon production on the windward reef flat of Eniwetok Atoll. Limnology and Oceanography 18(6): 953-961. Smith, S. V., and J. T. Harrison. 1977. Calcium carbonate production of the Mare Incognitum, the upper windward reef slope, at Enewetak Atoll. Science 197: 556-559. Sorokin, Y. I. 1990. Aspects of trophic relations, productivity and energy balance in coral ecosystems, p. 401-410. In Z. Dubinsky (ed.) Ecosystems of the World 25: coral reefs. Elsevier Science Publishers, Amsterdam. 550 p. Sorokin, Y. I. 1993. Coral reef ecology. Springer-Verlag, Berlin. 465 p. Stevenson, D. K., and N. Marshall. 1974. Generalizations on the fisheries potential of coral reefs and adjacent shallow-water environments. Proc. Sec. Int. Coral Reef Symp. 1: 147-156. Suchanek, T. H, and P. L. Colin. 1986. Rates and effects of bioturbation by invertebrates and fishes at Enewetak and Bikini Atolls. Bulletin of Marine Science 38(1): 25-34. 101 Suchanek, T. H., P. L. Colin, G. M. McMurtry, and C. S. Suchanek. 1986. Bioturbation and redistribution of sediment radionuclides in Enewetak Atoll lagoon by callianassid shrimp: biological aspects. Bulletin of Marine Science 38(1): 144-154. Thomann, R. V. 1981. Equilibrium model of fate of microcontaminants in diverse aquatic food chains. Canadian Journal of Fisheries and Aquatic Sciences 38: 280-296. Tsuda, R. 1987. Marine benthic algae of Enewetak Atoll, p. 1-9. In D. M. Devaney, E. S. Reese, B. L. Burch, and P. Helfrich (eds.) The natural history of Enewetak Atoll II. Biogeography and systematics. U.S. Department of Energy, Oak Ridge, Tennessee. 348 p. Van Dyke, J. M. 1991. Protected marine areas and low-lying atolls. Ocean & Shoreline Management 16: 87-160. Vasconcellos, M., S. Mackinson, K. Sloman, and D. Pauly. 1997. The stability of trophic mass-balance models of marine ecosystems: a comparative analysis. Ecological Modelling 100: 125-134. Venier, J. M. 1997. Seasonal ecosystem models of the Looe Key National Marine Sanctuary, Florida. M.Sc. Thesis. Department of Zoology. University of British Columbia, Vancouver. 109 p. Vinogradov, A. P. 1953. The elementary chemical composition of marine organisms. Sears Foundation for Marine Research, Yale University, Connecticut. 647 p. Walters, C. J. 1996. Suggested improvements for Ecopath modeling, p. 82-86. In D. Pauly and V. Christensen (eds.) Mass-balance models of North-Eastern Pacific ecosystems. Fisheries Centre Research Report 4(1). Wardlaw, B. R., T. W. Henry, and W. E. Martin. 1991. Sediment facies of Enewetak Atoll lagoon. U.S. Geological Survey Professional Paper; 1513-B. United States Government Printing Office, Washington. 60 p. Webb, K. L., D. W.D., and C. F. D'Elia. 1977. Biomass and nutrient flux measurements of Holothuria atra populations on windward reef flat at Enewetak, Marshall Islands. Proc. Third Int. Coral Reef Symp.: 409-415. Welander, A. D. 1957. Radioactivity in the reef fishes of Belle Island Eniwetok Atoll April 1954 to November 1955. UWFL-49. Applied Fisheries Laboratory, University of Washington, Seattle, Oak Ridge, Tennessee. 37 p. Welander, A. D., K. Bonham, L. R. Donaldson, R. F. Palumbo, S. P. Gessel, F. G. Lowman, and W. B. Jackson. 1966. Bikini-Eniwetok studies, 1964. UWFL-93 (Pt.I). Laboratory of Radiation Biology, University of Washington, Seattle, Washington. 277 p. Wells, S. M., and M. D. Jenkins. 1988. Coral reefs of the world. 3: Central and Western Pacific. UNEP and JUCN, Cambridge, U.K. 329 p. Wiebe, W. J., R. E. Johannes, and K. L. Webb. 1975. Nitrogen fixation in a coral community. Science 188: 257-259. Wiens, H. J. 1962. Atoll environment and ecology. Yale University Press, London. 532 p. 102 Williams, D. M. 1991. Patterns and processes in the distribution of coral reef fishes, p. 437-474. In P. F. Sale (ed.) The ecology of fishes on coral reefs. Academic Press Inc., San Diego. 754 p. Zanden, M. J. V., and J. B. Rasmussen. 1996. A trophic position model of pelagic food webs: impact on contaminant bioaccumulation in lake trout. Ecological Monographs 66(4): 451-477. 103 Appendices Appendix 1. Diet matrix of the seventeen non-fish groups included in the model.' Prey \ predators o o o o o -a o a, o CJ > o T3 O ft O g CD O cj I o "o o CD CD > .3 CJ CO T3 o a, _o 73 X! PH CJ o Detritus Benthic prim prod. Phytoplankton Zooplankton Corals Foraminiferans Gastropods Giant clams Bivalves Shrimps / lobsters Stomatopods Misc. crustaceans Echinoderms Holothurians Polychaetes Sessile invertebrates Cephalopods 0.740 0.400 0.430 0.450 - 0.750 0.851 0.500 0.777 0.450 0.500 0.700 1.000 -- 0.430 - - - 0.011 - 0.010 0.450 0.050 -0.160 - - - 1.000 0.250 --------0.100 0.600 0.010 ------- 0.010 -- 0.012 0.070 0.020 - - 0.300 -- 0.130 ------ 0.020 0.400 -- 0.100 - - 0.020 0.004 - 0.020 0.010 - - 0.050 0.005 0.001 - 0.010 0.010 0.450 0.022 0.035 0.099 0.020 -0.007 - - - 0.010 - 0.005 - - 0.010 0.054 0.375 0.006 0.020 -0.018 - 0.088 0.020 0.010 0.420 0.460 0.060 a. Diet composition based on Gerber and Marshall (1982). Cannibalism was reduced to 10% by scaling the predation pressure exerted on the other groups, b. From Heron Island, Australia (Sorokin 1993); c. Diet composition based on Sorokin (1993). 30% of the diet is assumed to be covered by symbiotic zooxanthellae; d. Diet composition based on Kohn (1987); e. From Arias-Gonzales (1993); f. A weighted mean between a diet composition for spiny lobster from Kohn (1987): gastropods 30%, bivalves 7%, crustaceans 32%, polychaetes 12%, echinoderms 10%, corals 4%, algae 5%, and a diet composition for shrimps from Arias-Gonzales (1993): polychaetes 1%, corals 1%, algae 1%, sessile invertebrates 2%, meiobenthos 5%, and detritus 90%; g. Diet composition based on Chartock (1983b); h. Diet composition based on Bakus (1968, 1973); i. Diet composition based on Sorokin (1993); j. Diet composition based on Reiswig (1971); k. Diet composition based on Helsinga and Fitt (1987). 75% of the diet is assumed to be contributed by symbiotic zooxanthellae; 1. Dashes indicate that there is no interactions between the two groups. 104 irsguos^jns irsgAoJannB sjsdnojS / sjaddEus ijsg snoJOAiuuio neuis qsy snojOAtLUG3 3§JBT UBTJ snojOAruiBD rpuis SSJOAlDSld 3STJV ^luauiSag irst J S3]Rjq3U3AUT 3ffSS3S SajaEipAJOd suii3pounio3 spodonsEo S3ATBAIR pdurmig SUB33B18TU3 OSIW r spodoiBiuois spodoieqdao (^Xuois / yos) sfoioo uoi^uEidoo^ 3B§rv o o o o o d © ei m o o n n o o © © © © vo s£> d d — o o — o <ri © —; <N odd M ^ O vi d d d d — — o o o © © © o o © Csl fl © © © © goo© d d d d © © vo © dodo o © © O — VO VO T © *T <*"> — — d d d d vo © —' r> o\ r-o o o odd -a a <u CL a < irsiiuoGSjns iisgXTjjgrms siadnorS / sjaddEus ITSTJ snOJOATUUIO [fBUIS •qsg snojoAnireo 3§JBT| qsy snojoATUJE3 TTEUIS sgjoApsrd osnA] ssiaiqaysAui arissas spodoiisEO ^S3ArBAig ^duruiis spodoiBuiojs, spodoiEudao ^umajnraiiBJOj ^Xuow / yos) SJBJCO uoi^uEidocrz fN O d d o o ooo o o o vo d d 1*1 o oo o d d ooo o g d d o o d d ooo — VO d d — o d d o r-o vo d d o vo d d <M O VO >n <N o o r- vo W « « -T "1 o o d d d d d d d VO O OOO © O <N odd CN K3 a <u CL CL TlSTJUOaSjTlS iisijXiiJaiing usy snojOAnnuo news iisy snojOAruiE3 32IBI qsy snojoAjuiBO [TBUIS S3JOApSld ^uaurSaij IJSTJ S3]Ejq3y3AUT 3[ISS3S ^suuapounpg spodoxisBO S3ATBAIQ ^duruiis spodo"fBuiois spodoTBLjdao JAUOIS i yos) sraioo ^uopruBidooz ooo © «1 "1 odd ooo © © — ooo O pi d d — — as odd ©ooo odd© o © © © W M M n M -© © o o o © o o o o © o O CN t- <0 t- — — © OOOOOOOO© — CN © rn © odd©©© I i 8 -S I i I Hi •S .2 3 .. 11 3 -S 11 o-i % • n -a KooooooS 3 3 3=1 a is S. s. ^ ° ^ *o *° i i i i i " " muOOcaOo-ocjUfJUD-Q, 3 I 3 1 P OD oil o 2 2 , PH W 60 ( 11 ll 'fl T3 C cu o. D-irsiiuoaSjns HsyXijJswng sjadnojS / siaddeus 1[SIJ SnOJOATUUIO HEUIS qsy snojOArujBO aftreq usu snojOApjjBa rreiug sajOAiosid *OSIJ\ ^JUSIUSEJI USI-J sajRiqayaAui anssag sujjapounpg spodoijsEfj saATEAig ^duiinrs SUE33EJSIU0 'OSI|^ spodoiBiuois spodoreiids^ |UEJ3jnnurcjcr} (Auois i yos) srejoQ uoj^uEidoo2| 8 8 — CO O O 8 8 d d o o o o o m \o r- o S5888 dooodd — do o — d d 3 9 o o o d CM 00 O CM d d odd d d 00 O (N ^ P dodo ^ CN o do--iisyuoaSjns ijsy^uj3)]ne| sjadnarS / siaddEus usy snoJOAnruio neuis trsy snojOAiuiBO aSiBi qsy snojoAiuiBD TIBIUS i sajoApsid osn>v ^puaurSBij usi^ satEjqauaAui snsssg pduruirs EUB33B1STU3 '3SnAJ spodotEuiois spodoisqdao JAUOJS/UOS) SpjJOQ uoi^iiEidoo^l s R R O H of X '-5 c CU o. o. <! o o © d o © © © © © — o © © — — I 1 1 1 1 ! I - 1 » ' a s S .a '6 i a • ilia is sit B | 8 o ON o cv) _C> O T3 CD c+4 00 43 • -i PH H 00 C O OH 3 > X Appendix 2, Table 2. Diet matrix of the ten fish groups included in the model. Based on Appendix 2 Table 1. tH — « CJ X JH XI rv X X trt ^ 'f g W « Prey \ Predator •§ 'I '5 '3 '§ PnOOCJingotgOW i-l coco O m co j DH J x Stony / soft corals - 0.027 0.027 - 0.183 0.040 0.078 0.700 0.042 -Detritus - - 0.037 - - - 0.007 - -Zooplankton - 0.258 0.315 0.173 0.042 0.030 0.057 - 0.018 0.830 Benthic prim prod. - 0.010 0.256 - 0.454 0.900 0.007 0.300 0.382 -Foraminiferans - 0.023 0.054 - 0.021 0.030 0.050 - 0.090 -Misc. crustaceans 0.060 0.209 0.080 0.287 - - 0.193 - 0.135 -Stomatopods 0.004 0.009 - 0.016 - - 0.007 - -Shrimps 0.046 0.165 0.030 0.096 0.125 - 0.131 - 0.027 -Cephalopods 0.005 0.002 - 0.022 - - 0.031 - -Bivalves - 0.008 0.012 0.017 - - 0.057 - 0.101 -Gastropods - 0.071 0.046 0.027 0.021 - 0.100 - 0.054 -Echinoderms - 0.002 0.008 0.039 - - 0.030 - 0.019 -Sessile invertebrates - 0.003 0.011 0.021 - - 0.040 - 0.020 -Polychaetes - 0.058 0.051 0.046 0.154 - 0.053 - 0.064 0.170 Small omniv. fish 0.046 0.061 0.027 0.037 - 0.000 0.099 - 0.047 -Small camiv. fish 0.199 0.073 0.046 0.117 - 0.000 0.050 - -Surgeonfish 0.521 - - - - - 0.005 - -Herring - - - 0.043 - - - - -Parrotfish - 0.021 - - - - - - -Groupers / snappers 0.010 - - - - - - - -Butterflyfish 0.074 - - - - - - - -Large camiv. fish 0.028 - - 0.060 - - 0.005 - -Misc. pisciv. fish 0.007 - - - - - - - -Ill Appendix 3. Scientific and common names of the fish species included in the model3 Scientific name Common name0 Miscellaenous piscivorous fish Carcharhinus melanopterus Carcharhinus amblyrhynchos Triaenodon obesus Nebrius ferrugineus Sphyraena qenie Selar crumenophthalmus (all juveniles) Trachinotus baillonii Carangoides orthogrammus Caranx melampygus Caranx ignobilis Elagatis bipinnulata Synodus variegatus Saurida gracilis Fistularia petimba Gymnosarda unicolor Euthynnus affmis Katsuwonus pelamis Tylosurus crocodilus crocodilus Aulostomus chinensis Herring Spratelloides delicatulus Small carnivorous fish Rhabdamia gracilis Apogon novaeguineae Apogon fuscus Apogon novemfasciatus Apogon kallopterus Apogon nigrofasciatus Apogon erythrinus Cheilodipterus quinquelineatus Pseudocheilinus hexataenia Labroides dimidiatus Novaculichthys taeniourus Stethojulis balteata Stethojulis sp. Halichoeres trimaculatus Halichoeres margaritaceus Macropharyngodon meleagris Thalassoma lucasanum Thalassoma hardwickii Cheilinus oxycephalus Pempheris oualensis Paracirrhites arcatus Blacktip reef shark Grey reef shark Whitetip reef shark Tawny nurse shark Blackfin barracuda Bigeye scad Smallspotted dart Island trevally Bluefin trevally Giant trevally Rainbow runner Variegated lizardfish Gracile lizardfish Red cornetfish Dogtooth tuna Kawakawa Skipjack tuna Hound needlefish Chinese trumpetfish Delicate round herring Luminous cardinalfish Samoan cardinalfish Sevenstriped cardinalfish Iridescent cardinalfish Blackstripe cardinalfish Fivelined cardinalfish Pyjama Bluestreak cleaner wrasse Rockmover wrasse Belted wrasse wrassed Threespot wrasse Pink-belly wrasse Blakspotted wrasse Cortez rainbow wrasse Sixbar wrasse Snooty wrasse Silver sweeper Arc-eye hawkfish 112 Appendix 3 (continued). Scientific name Common name0 Cirrhitus pinnulatus Scorpaenopsis gibbosa Pterois radiata Dinematichthys iluocoeteoides Valenciennea strigata Corythoichthys jlavofasciatus Hypoatherina temminckii Plesiops melas Plesiops nigricans Synanceia verrucosa Caracanthus unipinnus (unipinna ?) Thysanophrys sp. Parapercis cephalopunctata Myripristis pralinia Myripristis murdjan Myripristis violacea Neoniphon sammara Sargocentron microstoma Holocentrus laeris ? Sargocentron diadema Gymnothorax margaritophorus Oxymonacanthus longirostris Gomphosus varius Halichoeres melanurus Halichoeres marginatus Halichoeres hortulanus Thalassoma quinquevittata Thalassoma lutescens Large carnivorous fish Thalassoma purpureum Epibulus insidiator Cheilinus chlorourus Cheilinus trilobatus Coris aygula Coris gaimard gaimard Hemigymnus melapterus Mulloidichthys flavolineatus Parupeneus trifasciatus Parupeneus barberinus Parupeneus cyclostomus Upeneus arge Balistoides viridescens Pseudobalistes flavimarginatus Melichthys vidua Monotaxis grandoculis Stocky hawkfish Humpbacked scorpionfish Radial firefish Yellow pigmy brotula Blueband goby Network pipefish Samoan silverside Crimsontip longfin Whitespotted longfin Stonefish Pygmy coral croucher ? flathead" Scarlet soldierfish Pinecone soldierfish Lattice soldierfish Sammara squirrelfish Smallmouth squirrelfish Crown squirrelfish Blotched-neck moray Longnose fielfish Bird wrasse Tail-spot wrasse Splendid rainbow wrasse Checkerboard wrasse Fivestripe wrasse Sunset wrasse Surge wrasse Sling-jaw wrasse Floral wrasse Tripletail wrasse Clown wrasse Yellowtail coris Blackeye thicklip Yellowtail goatfish Dash-and-dog goatfish Gold-saddle goatfish Band-tail goatfish Titan triggerfish Yellowmargin triggerfish Pinktail triggerfish Humpnose big-eye bream 113 Appendix 3 (continued). Scientific name Common name0 Hyporhamphus dussumieri Bothus mancus Polydactylies sexfilis Heteropriacanthus cruentatus Pterois volitans Crenimugil crenilabis Echeneis naucrates Leiuranus semicinctus Branchysomophis sauropsis ? Echidna polyzona Siderea picta Gymnothorax flavimarginatus Gymnothorax buroensis Gymnothorax undulatus Gymnothorax rueppelliae Gymnothorax fimbriatus Conger noordzieki ? Moringua macrochir (microchir ?) Sargocentron spiniferum Arothron meleagris Arothron nigropunctatus Arothron sp. Small omnivorous fish lstiblennius chrysospilos Blenniella periophthalmus Cirripectes variolosus Gnatholepis anjerensis Gobis ornatus ? Bathygobius fuscus fuscus Gobiodon rivulatus Paragobiodon echinocephalus Chromis agilis Chromis caerulea Chromis lepidolepis Chromis margaritifer Chromis atripectoralis Pomacentrus coelestus (coelestis ?) Pomacentrus vaiuli Pomacentrus pavo Stegastes nigricans Stegastes fasciolatus Dascyllus reticulatus Dascyllus aruanus Amblyglyphidodon curacao Abudefduf sordidus Dussumier's halfbeak Tropical flounder Sixfinger threadfin Glasseye Lionfish Fringelip mullet Live sharksucker Saddled snake-eel Barred moray Peppered moray Yellow-edged moray Vagrant moray Undulated moray Banded moray Fimbriated moray Sabre squirrelfish Guineafowl puffer Blackspotted puffer pufferd Redspotted blenny Blue-dashed rockskipper Red-speckled blenny Dusky frill-goby Rippled coralgoby Redhead goby Agile chromis Green chromis Scaly chromis Bicolor chromis Black-axil chromis (Neon damselfish ?) Ocellate damselfish Sapphire damselfish Dusky farmerfish Pacific gregory Reticulate dascyllus Whitetail dascyllus Staghorn damselfish Blackspot sergeant 114 Appendix 3 (continued). Scientific name Common name0 Abudefduf septemfasciatus Abudefduf leucopomus (Chrysiptera leucopoma ?) Abudefduf amabilis ? Abudefduf glaucus (Chrysiptera galuca ?) Plectroglyphidodon lacrymatus Abudefduf biocelatus (Chrysiptera leucopoma ?) Centropyge flavissimus Canthigaster solandri Amanses carolae ? Large omnivorous fish Ostracion cubicus Arothron hispidus Siganus argenteus Kyphosus cinerascens Rhinecanthus rectangulus Rhinecanthus aculeatus Balistapus undulatus Snappers / groupers Epinephelus merra Epinephelus fuscoguttatus Epinephelus cyanopodus Epinephelus macrospilos Epinephelus howlandi Variola louti Cephalopholis urodeta Cephalopholis miniata Cephalopholis argus Anyperodon leucogrammicus Plectropomus areolatus Pseudanthias pascalus Lutjanus monostigma Lutjanus vitta Lutjanus gibbus Gymnocranius griseus Lethrinus microdon Aprion virescens Gnathodentex aureolineatus Pterocaesio tile Butterflyfish Chaetodon lunula Chaetodon citrinellus Chaetodon ephippium Banded sergeant Surge damselfish ? Grey demoiselle ? Whitespotted devil Surge damselfish ? Lemonpeel angelfish Spotted sharpnose filefishd? Yellow boxfish White-spotted puffer Streamlined spinefoot Blue sea chub Wedge-tail triggerfish White-banded triggerfish Orange-lined triggerfish Honeycomb grouper Brown-marbled grouper Speckled blue grouper Snubnose grouper Blacksaddle grouper Yellow-edged lyretail Darkfin hind Coral hind Peacock hind Slender grouper Squaretail coralgrouper Amethyst anthias One-spotted snapper Brownstripe red snapper Humpback snapper Grey large-eye bream Smalltooth emperor Green jobfish Striped large-eye bream Dark-banned fusilier Raccoon butterflyfish Speckled butterflyfish Saddle butterflyfish 115 Appendix 3 (continued). Scientific nameb Common name0 Chaetodon reticulatus Mailed butterflyfish Chaetodon trifascialis Chevron butterflyfish Chaetodon auriga Threadfin butterflyfish Surgeonfish Acanthurus mata Blue-lined surgeonfish Acanthurus xanthopterus Yellowfin surgeonfish Acanthurus gahhm -Acanthurus olivaceus Orangespot surgeonfish Acanthurus triostegus Convict surgeonfish Acanthurus achilles Achilles tang Acanthurus nigricans Whitecheek surgeonfish Acanthurus nigroris Bluelined surgeonfish Acanthurus guttatus Whitespotted surgeonfish Acanthurus lineatus Lined surgeonfish Ctenochaetus striatus Striated surgeonfish Naso lituratus Orangespine unicornfish Naso unicornis Bluespine unicornfish Zebrasoma veliferum Sailfin tang Parrotfish Cetoscarus bicolor Bicolour parrotfish Scarus sordidus Daisy parrotfish a. Hiatt and Strasburg (1960) examined the food and feeding habits and ecological relationship of 223 fish species of the Marshall Islands. By comparing this work with that of Schultz and collaborators (1953, 1960, 1966) and Randall and Randall (1987), 190 of Hiatt and Strasburgs 223 species were found to occur at Enewetak Atoll. The 190 species were grouped into the ten functional groups: miscellaneous piscivorous fish, snappers / groupers, herring, large carnivorous fish, small carnivorous fish, large omnivorous fish, small omnivorous fish, parrotfish, surgeonfish, and butterflyfish. Dashes indicate that FishBase (1998) does not have a common name for the species; Question marks either indicate that the species name was not found in FishBase, or indicate an inconsistency in spelling in which case the closest 'match' / synonym in FishBase (1998) is given in brackets; b. The scientific names were cross-checked with FishBase 98 (1998); c. FishBase (1998) common name; d. Family. 116 Appendix 4. Deriving the Q/B values of the ten fish groups. The Q/B values for the fish species included in the model were estimated using the empirical regression by Pauly et al. (Christensen and Pauly 1992b, Pauly et al. 1990): Q / B = 10637 • 0.0313^ • W^"0168 • 1.38pf • 1.89Hd where; Wro is the asymptotic or maximum weight of the fish in gram wet weight; Tk is the mean annual habitat temperature expressed as 1000/(T°C + 273.1) (an annual mean temperature of 27.5 was used in all cases, based on Atkinson (1987)); Pf is one for apex predators, pelagic predators, and zooplankton feeders, and zero for all other feeding types; and Hd characterizes the food type and is set to one for herbivores and zero for carnivores. Estimates of Wmax were obtained from FishBase (1998) for as many species as possible. In many cases Wmax was given directly, but in some cases it had to be estimated using the weight-length relationship: Wmax - a-Lb; where a and b are two constants and L is the total length (TL) (see Table 1). Wmax was converted to WB assuming that Woo = Wmax/0.95 (see Table 2). Appendix 4, Table 1. Deriving the maximum weight of fish species in the model. Based on FishBase (1998). Species LnjJ3 & t/ Place of origin WwJ (TL, SL.orFL) of a and b (g) Msc. piscivorous fish Carcharhinus melanopterus 300 TL? 0.003 3.649 Australia 3610185 Caranx melampygus 100 TL 0.024 2.980 Philippines 22253 Caranx ignobilis 165 TL 0.028 2.940 Philippines 91267 Elagatis bipinnulata 150 TL? 0.014 2.920 Philippines 30515 Saurida gracilis 36/32 TIVSL 0.005 3.150 New Caledonia 367 Fistularia petimba 200 TL? 0.000 3.160 New Caledonia 3735 Carcharhinus amblyrhynchos 120 TL 0.009 3.050 New Caledonia 19319 Synodus variegatus 40 TL 0.003 3.300 New Caledonia 523 " 40 TL 0.003 3.310 New Caledonia 502 Average S. variegatus: 512 117 Appendix 4, Table 1 (continued). Species Lmax a a ba Place of origin a Wmax (TL, SL, or FL) of a and b (g) Gymnosarda unicolor 224/206 Tl/FL 0.011 3.070 N. Marianas 172366 " 224 TL 0.041 2.800 Vanuatu 155746 Average G. unicolor: 164056 Herring Spratelloides delicatulus 7 TL? 0.002 3.290 New Caledonia 1 Small carnivorous fish Apogon fuscus 10/7.7 TL/SL 0.012 2.600 New Caledonia 5 Apogon kallopterus 15/12.2 TL/SL 0.009 3.180 New Caledonia 48 Cheilodipterus quinquelineatus 12 TL 0.014 3.040 New Caledonia 26 Labroides dimidiatus 11.5 TL 0.004 3.180 New Caledonia 10 Halichoeres trimaculatus 26/22 TL/SL 0.048 2.740 New Caledonia 362 Myripristis pralinia 21/17 TL/SL 0.021 3.070 New Caledonia 235 Myripristis violacea 20 TL? 0.051 2.900 New Caledonia 305 Neoniphon sammara 30 0.049 2.820 New Caledonia 710 Sargocentron microstoma 19/16 TL/SL 0.002 3.850 Micronesia 151 Sargocentron diadema 23 TL? 0.037 2.890 New Caledonia 321 Thalassoma lutescens 22/19 TL/SL 0.010 3.080 New Caledonia 140 Small omnivorous fish Chromis caerulea 8/6.5 TL/SL 0.030 2.410 New Caledonia 4 Pomacentrus vaiuli 11/10 TL/FL 0.037 2.890 New Caledonia 38 Pomacentrus pavo 11/8.5 TL/SL 0.068. 2.750 New Caledonia 49 Stegastes nigricans 14/11.5 TL/SL 0.081 2.350 New Caledonia 40 Dascyllus aruanus 8/6.5 TL/SL 0.014 2.690 New Caledonia 4 Amblyglyphidodon curacao 12/9 TL/SL 0.054 2.890 New Caledonia 71 Groupers / snappers Epinephelus merra 35/30 TL/SL 0.026 2.890 New Caledonia 745 Epinephelus fuscoguttatus 120 TL? 0.016 3.000 Philippines 27648 Epinephelus cyanopodus 122 TL 0.012 3.050 New Caledonia 28630 Epinephelus macrospilos 51 TL 0.015 3.000 New Caledonia 1963 Cephalopholis miniata 40 TL 0.066 2.760 New Caledonia 1730 Cephalopholis argus 55 TL 0.016 3.020 New Caledonia 2794 Lutjanus vitta 40 TL 0.010 3.090 Australia 892 Lutjanus gibbus 50 TL 0.021 3.000 New Caledonia 2625 Aprion virescens 96/80 TL/SL 0.005 3.260 N. Marianas 15073 Gnathodentex aureolineatus 30 TL 0.009 3.290 Micronisia 652 Variola louti 81 TL 0.018 2.970 N. Marianas 8524 II 81 TL 0.013 3.040 New Caledonia 8490 Average V. louti: 8507 Butterflyfish Chaetodon citrinellus 12.5 TL 0.034 2.950 New Caledonia 59 Chaetodon auriga 23 TL 0.023 3.040 New Caledonia 317 Surgeonfish Acanthurus mata 50 TL 0.040 2.950 New Caledonia 4071 Acanthurus xanthopterus 70 TL 0.009 2.770 New Caledonia 1110 Acanthurus olivaceus 35 TL 0.007 3.400 Micronesia 1244 Acanthurus nigricans 21 TL 0.067 2.670 Micronesia 227 118 Appendix 4, Table 1 {continued). Species Lmax a a ba Place of origin Wmax (TL, SL,orFL) of a and b (g) Acanthurus lineatus 38 TL 0.019 3.070 Micronesia 1359 Naso lituratus 45 TL 0.050 2.840 Micronesia 2463 Zebrasoma veliferum 40 TL 0.047 2.860 New Caledonia 1799 Acanthurus lineatus 38 TL 0.019 3.070 Micronesia 1359 Naso lituratus 45 TL 0.050 2.840 Micronesia 2463 Zebrasoma veliferum 40 TL 0.047 2.860 New Caledonia 1799 Acanthurus triostegus 27 TL 0.016 3.140 Micronesia 512 27 TL 0.052 2.390 New Caledonia 137 Average A. triostegus: 325 Ctenochaetus striatus 26 TL 0.021 3.040 Micronesia 420 " 26 TL 0.028 3.000 New Caledonia 489 Average C. striatus: 455 Naso unicornis 70 TL 0.023 2.920 Micronesia 5567 70 TL 0.022 2.990 New Caledonia 7298 Average N. unicornis'. 6432 Large carnivorous fish Cheilinus chlorourus 43/36 TL/SL 0.009 3.150 New Caledonia 1258 Parupeneus barberinus 62/50 TIVSL 0.012 3.080 New Caledonia 4078 Melichthys vidua 35 TL 0.006 3.550 Micronesia 1757 Heteropriacanthus cruentatus 34 TL 0.019 3.000 USA 739 Echeneis naucrates 110 TL? 0.005 3.300 New Caledonia 25627 Sargocentron spiniferum 45 TL 0.017 3.060 New Caledonia 1947 Monotaxis grandoculis 60 TL 0.036 2.850 Micronesia 4208 II 60 TL 0.026 2.990 New Caledonia 5370 Average M. grandoculis: 4789 Parrotfish Scarus sordidus 40 TL 0.013 3.140 Micronesia 1362 Large omnivorous fish Ostracion cubicus 45 TL 0.026 2.590 New Caledonia 501 Arothron hispidus 54/45 TL/SL 0.009 2.800 New Caledonia 638 Rhinecanthus rectangulus 25 TL 0.036 2.880 Micronesia 377 Rhinecanthus aculeatus 25 TL 0.018 3.100 Micronesia 386 Siganus argenteus 40 TL 0.025 2.880 Micrenesia 1028 it 40 TL 0.011 3.100 New Caledonia 981 Average S. argenteus: 1004 a. Wmax is the maximum reported weight of the fish. The values were obtained either directly from FishBase (1998) or estimated using the weight-length relationship: W = a-Lb where; L is the maximum length reported; and a and b are two constants (all values from FishBase (1998)); b. The maximum reported length of the fish. All values are from FishBase (1998). TL = total length, SL = standard length, FL = fork length; Wmax is estimated using TL, and SLs and FLs were therefore converted to TL by measuring, on a picture of the species, the ratio between the two lengths. Question marks indicate that the length is not specified in FishBase (1998), in which case it is assumed to be the total length (TL). 119 Appendix 4, Table 2. Q/B values of the fish groups included in the model. Species a Wmax (g) w00b HdC Pfd Q/B6 Msc. piscivorous fish Carcharhinus melanopterus 3610185 3800194 0 1 2.51 Caranx melampygus 22253 23424 0 1 5.90 Caranx ignobilis 91267 96070 0 1 4.66 Elagatis bipinnulata 30515 32122 0 1 5.60 Synodus variegatus 512 539 0 0 8.06 Saurida gracilis 367 387 0 0 8.52 Fistularia petimba 3735 3932 0 1 7.96 Gymnosarda unicolor 164056 172691 0 1 4.22 Triaenodon obesus 18250 19211 0 1 6.10 Sphyraena qenie 6800 7158 0 1 7.20 Trachinotus baillonii 900 947 0 1 10.12 Carangoides orthogrammus 4300 4526 0 1 7.78 Euthynnus affinis 13600 14316 0 1 6.41 Katsuwonus pelamis 34500 36316 0 1 5.48 Average 6.47 Herring Spratelloides delicatulus 1 1 0 1 30.48 Small carnivorous fish Apogon juscus 5 5 0 1 24.25 Apogon kallopterus 48 50 0 0 12.00 Cheilodipterus quinquelineatus 26 27 0 0 13.30 Labroides dimidiatus 10 11 0 1 21.49 Halichoeres trimaculatus 362 381 0 0 8.54 Myripristis pralinia 235 247 0 1 12.68 Myripristis violacea 305 321 0 1 12.13 Neoniphon sammara 710 747 0 0 7.63 Sargocentron microstoma 151 159 0 0 9.90 Sargocentron diadema 321 338 0 0 8.71 Thalassoma hardwickii - - - 17.28 f Pempheris oualensis - - - 16.90 g Paracirrhites arcatus - - - 16.90 h Thalassoma lutescens 140 148 0 0 10.01 Gomphosus varius - - - 15.73 f Halichoeres marginatus - - - 14.16 f Halichoeres hortulanus - - - 10.97 f Average 13.68 Small omnivorous fish Chromis caerulea 4 5 0 1 24.65 Pomacentrus vaiuli 38 40 1 0 23.56 Pomacentrus pavo 49 52 1 0 22.56 Stegastes nigricans 40 42 1 0 23.39 Dascyllus aruanus 4 4 0 1 25.51 Amblyglyphidodon curacao 71 74 0 1 15.51 Gnatholepis anjerensis - - - - 39.10 i 120 Appendix 4, Table 2 (continued). Species a Wmax (g) w b HdC Pfd Q/Be Bathygobius fiiscus fuscus - - - 9.50 j Centropyge flavissimus - - - - 38.08 f Canthigaster solandri - - - - 15.00 k Average 23.69 Groupers / snappers Epinephelus merra 745 784 0 0 7.57 Epinephelus fuscoguttatus 27648 29103 0 0 4.12 Epinephelus cyanopodus 28630 30137 0 0 4.10 Epinephelus macrospilos 1963 2067 0 0 6.43 Variola louti 8507 8955 0 0 5.03 Cephalopholis miniata 1730 1821 0 0 6.57 Cephalopholis argus 2794 2941 0 0 6.06 Lutjanus vitta 892 939 0 0 7.34 Lutjanus gibbus 2625 2763 0 0 6.12 Aprion virescens 15073 15867 0 0 4.57 Gnathodentex aureolineatus 652 686 0 0 7.74 Average 5.97 Butterflyfish Chaetodon citrinellus 59 62 0 0 11.59 Chaetodon lunula - - - - 12.23 f Chaetodon auriga 317 334 0 0 8.73 Chaetodon ephippium - - - - 23.32 f Chaetodon reticulatus - - - - 14.25 f Average 14.02 Surgeonfish Acanthurus mata 4071 4285 1 0 10.75 Acanthurus xanthopterus 1110 1169 1 0 13.37 Acanthurus olivaceus 1244 1310 1 0 13.12 Acanthurus triostegus 325 342 1 0 16.44 Acanthurus nigricans 227 239 1 0 17.46 Acanthurus lineatus 1359 1431 1 0 12.93 Ctenochaetus striatus 455 478 1 0 15.54 Naso lituratus 2463 2593 1 0 11.70 Naso unicornis 6432 6771 1 0 9.96 Zebrasoma veliferum 1799 1893 1 0 12.33 Average 13.36 Large carnivorous fish Cheilinus chlorourus 1258 1324 0 0 6.93 Parupeneus barberinus 4078 4293 0 0 5.69 Melichthys vidua 1757 1850 0 0 6.55 Monotaxis grandoculis 4789 5041 0 0 5.54 Heteropriacanthus cruentatus 739 778 0 0 7.58 Echeneis naucrates 25627 26976 0 0 4.18 121 Appendix 4, Table 2 (continued). Species a Wmax (g) wj5 HdC Pfd Q/B6 Sargocentron spiniferum 1947 2049 0 0 6.44 Epibulus insidiator - - - - 12.53 f Cheilinus trilobatus - - - - 9.10 f Bothus mancus - - - - 4.90 1 Echidna polyzona - - - - 5.40 m Gymnothorax jlavimarginatus - - - - 4.50 n Gymnothorax buroensis - - - - 4.50 n Gymnothorax undulatus - - - - 4.50 n Gymnothorax rueppelliae - - - - 4.50 n Gymnothorax fimbriatus - - - - 4.50 n Average 6.08 Parrotfish Scants sordidus 1362 1547 1 0 12.76 Large omnivorous fish Ostracion cubicus 501 528 0 0 8.09 Arothron hispidus 638 672 0 0 7.77 Siganus argenteus 1004 1057 1 0 13.60 Rhinecanthus rectangulus 377 397 0 0 8.48 Rhinecanthus aculeatus 386 406 0 0 8.45 Average 9.28 a. Wmax is the maximum reported weight of the fish. The values were obtained either directly from FishBase (1998), or from Appendix 4, Table 1; b. Estimated assuming that W» = Wmax/0.95; c. Characterizes the food type: Value of one for herbivores (in this study when 30% or more of the diet comes from primary producers) and zero for carnivores; d. Value of one for apex predators, pelagic predators, and zooplankton feeders, and zero for all other feeding types; e. Unless otherwise noted, Q/B was estimated using the empirical regression derived from Pauly et al. (1990): Q/B = 10637-0.0313Tk-Wco"0168-1.38pf-1.89Hd where; W„c is the asymptotic or maximum weight of the fish in gram wet weight; Tk is the mean annual habitat temperature expressed as 1000/(T°C + 273.1) (an annual mean temperature of 27.5 was used in all cases based on Atkinson (1987)); Pf is one for apex predators, pelagic predators, and zooplankton feeders, and zero for all other feeding types; and Hd characterizes the food type and is set to one for herbivores and zero for carnivores. f. Obtained from Arias-Gonzales (1993) for same species; g. From Opitz (1996), Pempheris poeyi (a small carnivorous reef fish); h. From Opitz (1996), Amblycirrhitus pinos (a small carnivorous reef fish); i. From Opitz (1996), Gnatholepis thompsoni (a small omnivorous reef fish); j. From Opitz (1996), Bathygobius soporator (a small carnivorous reef fish); k. From Opitz (1996), Canthigaster rostrata (a small omnivorous reef fish); 1. From Opitz (1996), Bothus lunatus (intermediate carnivorous reef fish); m. From Opitz (1996), Echidna catenata (intermediate carnivorous reef fish); n. From Opitz (1996). Average value for three Gymnothorax species (G funebris, G. vicinus, G. miliaris). 122 Appendix 5. Fish biomass estimates in Odum and Odum (1955). Appendix 5, Table 1. Dry weight estimates of fish for each of the zones included in Odum and Odum (1955). Zone Fish category Quantity measured and basis for calculation Biomass (g dwm"2)a Algal ridge Parrotfish Visual count: 0.4 fish-28 m"2; 9.3 g loss on ignition-individual"1 0.10 Reef flat Parrotfish Visual count: 0.4 fish-28 m"2; 9.3 g loss on ignition-individual"1 0.10 Small heads Small herbivores Visual count: 25 fish-36 m"2; 2.42 g dwfish"1 and 61% herbivores based on rotenone sampling 1.00 Large herbivores Visual count: 52 fish-692 m"2 of horizontal visibility in all directions. 120 g dwfish"1; 90% herbivorous; large fishes absent from area 1/3 of time during maximum currents 5.00 Small carnivores 39% of fish counted (see herbivorous fish above) 0.65 Large carnivores 10% of fish counted (see herbivorous fish above) 1 stone fish-36m"2 (100 g dw) 0.70 2.80 Large heads Small herbivores Visual count: 71 fish-36 m"2; 2.42 g dwfish"1 4.80 Large herbivores Visual count: 30 fish-600 m"2 horizontally visible area; estimated V* herbivorous; 120 g dwfish"1 4.50 Small carnivores 5.3 fish-36 m"2; 2.42 g dwfish"1 0.34 Large carnivores 'A of fishes counted in visible horizontal area (see herbivorous above) 1.50 Sand / shingle Small herbivores Visual count: 23 fish-36 m"2; 2.42 g dwfish"1 1.50 sardine/herring Count of schools: 1.2-600 m"2 horizontal visible area; About 100 fish-school"1; 1 g dwfish"1 0.20 Large herbivores Visual count: 16 fish-600 m"2 horizontal visible area; 240 g dwfish"1 6.40 Large fish not including sharks Visual count: 3.2 fish-600 m"2 horizontal visible area; 240 g dwfish"1 1.30 Sharks Counts per 20 min observation: 1.6 sharks-600 m"2 visible area; 90 degrees visibility at one time; each individual in sight about 30 sec; Weight per shark about 4540 g dw (20% of wet) (Vinogradov 1953). 1.20 a. The dry weights (dw) were converted to wet weights (ww) assuming that, except for sharks, dw = 26% ww (Odum and Odum 1955, Sambilay 1993). For sharks, dw = 20% ww (Vinogradov 1953). 123 Appendix 5, Table 2. Weighted mean fish biomass estimate across the windward reef as derived from Odum and Odum (1955) (see also Appendix 5, Table 1). Reef area \ fish group Herbivores Carnivores Total Total3 (g dwm"2) (g dwm"2) (g dwm"2) (g ww-m"2) Not including sand / shingle zone 3.88 1.50 5.37 20.66 Including sand / shingle zone 4.68 1.74 6.42 24.69 a. Assuming that dw = 26% ww (Sambilay 1993). 124 Appendix 6. List of the ecosystem maturity attributes defined by Odum (1969) that are quantified in Ecopath (Christensen 1992, 1995). Odums ecosystem attribute3 Corresponding Ecopath output Developmental stage Mature stage 1. Gross production / community respiration (P/R ratio) Deviation of Pp/R (Teta) Greater or less than 1 Approaches 1 2. Gross production / standing crop biomass (P/B ratio) Production / biomass (Pp/B) High Low 3. Biomass supported / unit energy flow (B/E ratio) Biomass supported (B/T) Low High 5. Food chains Connectance (C) System omnivory index (SOI) Linear, predom. grazing II Web-like, predom. detritus M Dominance of detritus (Dom.Det.) It II 8. Species diversity - variety component Flow diversity (H) Low High 9. Species diversity - equitability component Flow diversity (H) Low High 13. Size of organism Average organism size (B/P) Small Large 15. Mineral cycles Finn's cycling index (FCI) Open Closed 16. Nutrient exchange rate, betw. organisms and environment Path length (PL) Rapid Slow 18. Growth form Residence time (B/(R+EXP)) Rapid growth ("r-selection") Feedback control ("K-selection") 21. Nutrient conservation Nutrient conservation (Oex) Poor Good 23. Entropy Schrddinger ratio (R/B) High Low 24. Information Information content of flows (I) Low High a. The numbers in the left column correspond to the numbers in Table 1 in Odum (1969). 125 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

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.831.1-0074838/manifest

Comment

Related Items