"Science, Faculty of"@en . "Resources, Environment and Sustainability (IRES), Institute for"@en . "DSpace"@en . "UBCV"@en . "Dalsgaard, Anne Johanne Tang"@en . "2009-06-09T22:26:13Z"@en . "1999"@en . "Master of Science - MSc"@en . "University of British Columbia"@en . "An approach for modeling the trophic transfer of beta radioactivity within the marine food\r\nweb of Enewetak Atoll, Micronesia, Central Pacific is described. From 1948 to 1958 this atoll\r\nwas used by the US military for testing of nuclear weapons while monitoring the impact on the\r\necosystem. In parallel to these military operations, a marine laboratory was operating on the\r\natoll, hosting a wealth of scientists performing basic research. Probably the most renowned\r\nstudy was carried out by H.T. Odum and E.P. Odum in 1954, who examined the trophic\r\nstructure of the windward reef community and its productivity per unit area.\r\nBased on this study and on the vast amount of scientific literature on the atoll, a mass-balance\r\ntrophic model of the windward reef was constructed, based on the Ecopath modeling\r\nsoftware.\r\nEcopath uses as its basic inputs the biomass, production/biomass, and food consumption rates\r\nof the various functional groups in the ecosystem, along with a diet matrix. Based on these\r\ninputs it estimates the flow of biomass between the functional groups and presents the\r\ncorresponding predation mortalities in a matrix where the columns represent the intake of, and\r\nthe rows the losses of, biomass from the groups. A set of first-order differential equations,\r\nrelating the intake and loss of biomass to the amounts of radioactivity in the groups, was then\r\nset up. The equations were integrated over time and calibrated by minimizing the sum of\r\nsquared deviations between the observed and predicted levels of radioactivity, thus mapping\r\nthe transfer of radioactivity onto the transfer of biomass. The original food web / mass-balance\r\nmodel, which was constructed without reference to the data on radioactivity, was\r\nsubsequently re-calibrated to achieve a match between the food web and the radioactivity\r\ndata.\r\nThe results predict that there is a time lag between the observed maximum o f radioactivity and\r\nthe trophic position of the groups, and that beta radioactivity is not bioaccumulated up\r\nthrough the food web.\r\nFinally, suggestions on how to incorporate the approach as a general routine into the Ecopath\r\nsoftware are given."@en . "https://circle.library.ubc.ca/rest/handle/2429/8910?expand=metadata"@en . "7080353 bytes"@en . "application/pdf"@en . "M O D E L I N G T H E T R O P H I C T R A N S F E R O F B E T A R A D I O A C T I V I T Y I N T H E M A R I N E F O O D W E B O F E N E W E T A K A T O L L , M I C R O N E S I A by A N N E J O H A N N E T A N G D A L S G A A R D B . S c , The Univers i ty o f Copenhagen, 1995 A T H E S I S S U B M I T T E D I N P A R T I A L F U L F I L M E N T O F T H E R E Q U I R E M E N T S F O R T H E D E G R E E O F M A S T E R O F S C I E N C E i n T H E F A C U L T Y O F G R A D U A T E S T U D I E S (Department o f Resource Management and Environmenta l Studies; Fisheries Centre) W e accept this thesis as conforming to the^Fequ^ired^standard T H E U N I V E R S I T Y O F B R I T I S H C O L U M B I A January 1999 \u00C2\u00A9 A n n e Johanne T a n g 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 \u00C2\u00A3noy'/-o' nmen-cjoU c&u.oLCeS The University of British Columbia Vancouver, Canada DE-6 (2/88) Abstract A n approach for model ing the trophic transfer o f beta radioactivity w i t h i n the marine f o o d web o f E n e w e t a k A t o l l , M i c r o n e s i a , Central Paci f ic is described. F r o m 1948 to 1958 this atoll was used by the U S military for testing o f nuclear weapons whi le moni tor ing the impact on the ecosystem. In parallel to these military operations, a marine laboratory was operating o n the atoll , hosting a wealth o f scientists performing basic research. Probably the most renowned study was carried out by H . T . O d u m and E . P . O d u m in 1954, w h o examined the trophic structure o f the w i n d w a r d reef community and its product iv i ty per unit area. B a s e d on this study and o n the vast amount o f scientific literature o n the atol l , a mass-balance trophic model o f the w i n d w a r d reef was constructed, based o n the E c o p a t h model ing software. E c o p a t h uses as its basic inputs the biomass, production/biomass, and f o o d consumption rates o f the various functional groups in the ecosystem, along w i t h a diet matrix. B a s e d o n these inputs it estimates the f l o w o f biomass between the functional groups and presents the corresponding predation mortalities i n a matrix where the columns represent the intake of, and the r o w s the losses of, biomass f r o m the groups. A set o f f irst-order differential equations, relating the intake and loss o f biomass to the amounts o f radioactivity i n the groups, was then set up. T h e equations were integrated over time and calibrated by m i n i m i z i n g the sum o f squared deviations between the observed and predicted levels o f radioactivity, thus mapping the transfer o f radioactivity onto the transfer o f biomass. T h e original f o o d web / mass-balance model , w h i c h was constructed without reference to the data o n radioactivity, was subsequently re-calibrated to achieve a match between the f o o d web and the radioactivity data. T h e results predict that there is a t ime lag between the observed m a x i m u m o f radioactivity and the trophic posi t ion o f the groups, and that beta radioactivity is not bioaccumulated up through the f o o d web. Final ly , suggestions o n h o w to incorporate the approach as a general routine into the E c o p a t h software are given. n Table of Contents Abstract i i Table o f Contents i i i L i s t o f Tables v i L i s t o f Figures v i i i A c k n o w l e d g m e n t s ' x 1. Introduct ion 1 1.1 General introduct ion and objectives 1 1.2 E n e w e t a k A t o l l , locat ion and description 2 1.3 A t o l l format ion and D a r w i n ' s theory o f subsidence 6 1.4 H i s t o r i c a l events o n E n e w e t a k A t o l l 7 2. B a c k g r o u n d theory 13 2.1 T h e o r y o f radioactivity 13 2.1.1 N a t u r a l and artificial radioactivity 13 2.1.2 Radioact ive decay : 13 2.1.3 T h e b io log ica l effects o f radiation 16 2.1.4 M e a s u r i n g radioactivity 17 2.1.5 U n i t s o f radioactivity 18 2.2 Radioact iv i ty f r o m nuclear explosions 18 2.2.1 Types o f nuclear explosions 18 2.2.2 D e c a y o f mixed fission products 21 2.2.3 U p t a k e o f radioactivity by marine organisms 21 2.3 Compartment model ing 24 2.3.1 E c o p a t h 24 2.3.2 E c o r a n g e r 26 3. M e t h o d s 2 8 3.1 Def in ing the modeled area 28 3.1.1 F o r e ree f 3 0 3.1.2 A l g a l ridge 30 3.1.3 R e e f flat 3 1 3.1.4 C o r a l head zone 3 1 3.1.5 S a n d / s h i n g l e 3 > 3.2 V a l i d a t i n g the E c o p a t h model 31 3.2.1 N o n - f i s h groups 32 3.2.1.1 Detr i tus 3 2 3.2.1.2 B e n t h i c P r i m a r y Producers 3 2 3.2.1.3 P h y t o p l a n k t o n 3 3 3 . 2 . 1 . 4 Z o o p l a n k t o n 3 3 3.2.1.5 C o r a l s and sea anemones (Class anthozoa) 34 3.2.1.6 Foraminiferans and other protozoans 34 i i i 3.2.1.7 Gastropods 35 3.2.1.8 B i v a l v e s 35 3.2.1.9 Shrimps and lobsters 36 3.2.1.10 Stomatopods 36 3.2.1.11 Misce l laneous crustaceans 37 3.2.1.12 E c h i n o d e r m s - not including holothurians 37 3.2.1.13 Holothur ians 37 3.2.1.14 Polychaetes and other w o r m - l i k e invertebrates 38 3.2.1.15 Sessile invertebrates 38 3.2.1.16 Cephalopods 38 3.2.2 B i o m a s s , P / B , and Q/B values o f non-fish groups 39 3.2.2.1 R e m a r k s to Table 3.3 39 3.2.3 F i s h groups 49 3.2.3.1 T h e distribution and abundance o f fish 49 3.3 T h e or ig in and incorporat ion o f the radioactivity data 52 3.3.1 T h e or ig in o f the radioactivity data 52 3.3.2 Observed trends in radioactivity i n various organisms 53 3.3.3 Radioact iv i ty in whole organisms 55 3.3.4 Simulat ing the observed trends i n beta radioactivity 55 4. Results 59 4.1 B a l a n c i n g the E c o p a t h model 59 4.1.1 First run w i t h Ecoranger using initial input parameters 59 4.1.1.1 M o d i f y i n g the predation mortality experienced by surgeonfish 60 4.1.1.2 M o d i f y i n g the predation mortality experienced by shrimps, miscellaneous crustaceans, and gastropods 60 4.1.2 Second run w i t h Ecoranger using modif ied input parameters 61 4.2 T h e fate o f beta radioactivity 64 4.2.1 M a p p i n g the fate o f beta radioactivity 64 4.2.2 Simulat ing the fate o f beta radioactivity 64 4.2.3 Re-cal ibrat ing the E c o p a t h model 68 4.2.4 B e t a radioactivity and trophic levels 70 4.3 Parameter estimation and network analysis 71 4.3.1 Summary statistics 71 4.3.2 Transfer efficiencies 71 4.3.3 M i x e d trophic impact 72 5. D i s c u s s i o n 74 5.1 M o d e l input parameters 74 5.1.1 T h e t ime span covered by the model 74 5.1.2 F i s h biomass and abundance estimates 75 5.1.2.1 V i s u a l census and rotenone sampling 75 5.1.2.2 F i s h biomass estimates at E n e w e t a k A t o l l 76 5.1.2.3 C o m p a r i n g the standing stock o f coral reef fish 77 5.1.2.4 The abundance and role o f herbivorous fish 77 5.1.2.5 T h e fish fauna and z o o p l a n k t o n o n the fore reef 80 iv 5.2 Outputs o f the model 82 5.2.1 T h e role o f benthic primary producers 82 5.2.2 T h e role o f detritus 82 5.2.3 C o m p a r i n g w i t h other models 83 5.2.3.1 T r o p h i c transfer efficiencies 83 5.2.3.2 B i o m a s s by trophic level 84 5.2.3.3 E c o s y s t e m maturity 85 5.3 Simulat ing radioactivity 86 5.3.1 T h e re-calibrated E c o p a t h model 86 5.3.2 T r o p h i c transfer o f radioactivity and the ' f o o d web time l a g ' 87 5.3.3 D i l u t i o n effects and the additional mortality (IvT) 87 5.3.4 B i o - d i m i n u t i o n o f beta radioactivity 88 5.3.5 Potentials o f the approach 89 References 91 Appendices 104 A p p e n d i x 1. D i e t matrix o f the seventeen non-fish groups included in the model 104 A p p e n d i x 2. D i e t matrix o f the fish species included i n the m o d e l 105 A p p e n d i x 3. Scientific and c o m m o n names o f the fish species included in the model 112 A p p e n d i x 4. D e r i v i n g the Q/B values o f the ten fish groups 117 A p p e n d i x 5. F i s h biomass estimates o f O d u m and O d u m (1955) 123 A p p e n d i x 6. L i s t o f the ecosystem maturity attributes, defined by O d u m (1969), that are quantified i n E c o p a t h 125 List of Tables Table 1.1. Stratigraphic subdivisions o f bore holes dri l led at E n e w e t a k A t o l l 7 Table 2.1. N a t u r a l l y occurr ing radioisotopes in sea water 14 Table 2.2. Radioact ive decay f rom a nuclear detonation 15 Table 2.3. Dates and locat ion o f the military nuclear detonations at E n e w e t a k A t o l l 19 Table 2.4. A r t i f i c i a l radioisotopes originating f r o m a nuclear explos ion 20 Table 2.5. Radioisotopes in marine organisms at E n e w e t a k A t o l l 23 Table 3.1. T h e areal extent o f the five zones across the w i n d w a r d reef. 29 Table 3.2. M a r i n e benthic algae at E n e w e t a k A t o l l 33 Table 3.3. Summary table o f the biomass, P / B , and Q/B values o f the non-fish groups 39 Table 3.4. B i o m a s s estimates o f detritus 41 Table 3.5. B i o m a s s estimates o f benthic primary producers 41 Table 3.6. C o r a l biomass estimates 42 Table 3.7. T h e biomass o f foraminiferans 42 Table 3.8. B i o m a s s estimates o f gastropods 43 Table 3.9. B i o m a s s estimates o f shrimps and lobsters 43 Table 3.10. B i o m a s s estimates o f crabs and other crustaceans 44 Table 3.11. B i o m a s s estimates o f echinoderms 44 Table 3.12. B i o m a s s estimates o f holothurians 45 Table 3.13. B i o m a s s estimates o f polychaetes and other w o r m - l i k e invertebrates 45 Table 3.14. Rate constants for some holothurians at E n e w e t a k A t o l l 48 Table 3.15. B i o m a s s , P/B and Q/B values for infaunal polychaetes 48 Table 3.16. E x a m p l e o f the stomach contents o f Neoniphon sammara 50 Table 3.17. Summary table o f the biomass, P / B , and Q/B values o f the ten fish groups 51 Table 3.18. Sample sizes o n w h i c h the data o n beta radioactivity were derived 54 Table 3.19. Relat ive weight o f the different body parts o f fish, bivalves, holothurians, and gastropods 55 Table 4.1. B a s i c estimates o f the 'best m o d e l ' 62 Table 4.2. Scal ing factors generated by Solver 69 v i Table 4.3. ' A d d i t i o n a l mortalit ies ' ( M ) 70 Table 4.4. Summary statistics 72 Table 4.5. Transfer efficiency (%) by trophic level 72 Table 5-1. Standing stock o f fish o n coral reefs in different regions 78 Table 5-2. T r o p h i c transfer efficiencies (%) for four coral reef ecosystem models 84 Table 5-3. B i o m a s s at discrete trophic levels 84 v i i List of Figures Figure 1-1. Western Paci f ic and M i c r o n e s i a 3 F i g u r e 1-2. E n e w e t a k A t o l l , R e p u b l i c o f the M a r s h a l l Islands 5 F i g u r e 1-3. T i m e arrow showing the major human events o n E n e w e t a k A t o l l 9 F i g u r e 1-4. T h e physiographic zones o f the reef i n surface and cross section v i e w 10 F i g u r e 2-1. T h e uranium series 15 F igure 2-2. Processes taking place once radioactive fallout reaches the ocean surface 22 Figure 2-3. Schematic representation o f an E c o p a t h model 27 F i g u r e 3-1. Cross-reef currents and channel currents 28 F i g u r e 3-2. Z o n a t i o n across the w i n d w a r d reef as defined i n the present study 29 Figure 3-3. T o t a l beta radioactivity in corals (Acropora) after the ' N e c t a r ' shot 54 Figure 3-4. Transfer o f radioactivity between compartments o f an ecosystem 56 Figure 3-5. Radioact iv i ty in the benthic primary producers 57 F igure 4-1. Simpli f ied trophic f l o w diagram o f the w i n d w a r d reef o f E n e w e t a k A t o l l 63 Figure 4-2. M a p p i n g the fate o f radioactivity 65 Figure 4-3. Trends in beta radioactivity i n the functional groups 66 Figure 4-4. M o d i f y i n g the columns or the r o w s o f the predation mortal i ty matrix 68 Figure 4-5. T r o p h i c levels and days required to reach m a x i m u m levels o f beta radioactivity. 71 F igure 4-6. M a x i m u m level o f beta radioactivity and trophic levels 71 F igure 4-7. M i x e d trophic impact diagram 73 F i g u r e 5-1. L e s l i e plots 76 Figure 5-2. B i o m a s s pyramids 84 Figure 5-3. N e t w o r k analysis 86 v i i i Acknowledgments I w o u l d l ike to express my sincere gratitude to m y advisor and mentor, D r . D a n i e l Pauly, for his unique guidance and never failing trust. F o r g iv ing me the opportunity to study at U B C and for his ideas and inputs without w h i c h this fruitful project w o u l d never have happened. A l s o a very special thanks to D r . C a r l Walters for his invaluable inputs, to D r . L e s s L a v k u l i c h for his tremendously w a r m and very helpful support throughout m y time as a student in R M E S , and to D r . V i l l y Christensen for helpful discussions and pleasant interactions. I thank J . L . M u n r o o f I C L A R M , M . D u k e o f the U n i v e r s i t y o f Washington, and R. E . F o r e m a n o f the U n i v e r s i t y o f B r i t i s h C o l u m b i a for k indly lending me hard to access data and literature. I also w i s h to thank D r . R E . F o r e m a n for helpful comments o n the manuscript. Thanks to my fe l low students and friends at the Fisheries Centre, and to D r . T o n y Pitcher, A n n Tautz , and Ingrid R o s s w h o helped realize my stay at the Fisheries Centre. A n d finally, thanks to m y dear family and close friends for always being there for me. ix 1. Introduction 1.1 General introduction and objectives. In December 1942, w h e n Italian physicist E n r i c o F e r m i produced the first nuclear fission reaction in a secret underground military laboratory i n C h i c a g o (Lenssen 1991), artificial radioactivity became an environmental reality. Since the m i d 1940s, radioactivity has been released into the marine environment f r o m various anthropogenic sources inc luding nuclear weapons testing, radioactive waste disposal (both c iv i l ian and mil i tary sources), and effluent f r o m p o w e r and fuel reprocessing plants as w e l l as accidental releases (Kennish 1998, Osterberg et al. 1964, R o w a n and Rasmussen 1994). O n c e i n the marine environment this radioactivity is o f serious human health concern because o f its potential to distribute itself throughout diffuse f o o d webs ( C l a r k 1989, Lenssen 1991, K e n n i s h 1998). Essential to understanding the contaminant pathways and ultimate fate o f radioactivity in marine ecosystems is the knowledge o f trophic relationships (Jarman et al. 1996). H o w e v e r , incomplete or thermodynamical ly unbalanced f o o d webs have often been used to describe the fate o f radioactivity. Indeed, laboratory experiments structured around simplified f o o d chains are probably among the main reasons for contradictory reports concerning the relative importance o f transfer w i t h i n f o o d webs versus direct uptake (adsorption and absorption) o f radioactivity by aquatic organisms (Ophel and J u d d 1966, P o l i k a r p o v 1966, T h o m a n n 1981). Similarly, investigations based on field observations have suffered f r o m difficulties in adequately representing and quantifying the trophic posi t ion o f the organisms ( K i r i l u k et al. 1995, Zanden and Rasmussen 1996). This problem has impeded studies f r o m examining the importance and quantifying the effects o f trophic transfer and f o o d web dynamics in explaining observed patterns o f contaminant bioaccumulat ion ( K i r i l u k et al. 1995, Z a n d e n and Rasmussen 1996). Recently , the study o f the enrichment o f stable isotopes (particularly 5 1 5 N / 5 1 4 N ratios) through aquatic f o o d webs has shown to be a promising measure o f the organism's (fractional) trophic posit ion, taking into account the importance o f omnivory and complexity , characteristics o f aquatic f o o d webs (Cabana and Rasmussen 1994, K i r i l u k et al. 1995, Z a n d e n and Rasmussen 1 1996). It has further been demonstrated that the enrichment o f 5 1 5 N is correlated w i t h the contaminant levels o f certain persistent pollutants, suggesting that trophic transfer o f contaminants can be significant ( K i r i l u k et al. 1995, Z a n d e n and Rasmussen 1996). A n o t h e r approach for determining trophic positions o f organisms is through the use o f mass-balance f o o d web models constructed w i t h the E c o p a t h approach and software, initiated by P o l o v i n a (1984) and further developed by Christensen and P a u l y (1992a, 1995). Recently, K l i n e and P a u l y (1998) examined the relation between trophic posit ions estimated by 5 1 5 N enrichment and by E c o p a t h and found an extremely high correlat ion (r = 0.986). In this study, the E c o p a t h approach is taken one step further and its potential for predicting the fate o f radioactivity w i t h i n a marine f o o d web is examined. T h e study proceeds by mapping the fate o f beta radioactivity onto an E c o p a t h generated f o o d web o f the marine ecosystem o f E n e w e t a k A t o l l , M i c r o n e s i a , Centra l Pacif ic . T h i s mapping involves re-calibration o f a preliminary model , initially constructed without reference to the data on radioactivity, and subsequent modif icat ion o f some o f the m o d e l inputs, unti l a match is achieved between the f o o d web and the pollutant data. T h e dissemination o f radioactivity is then simulated, using the trophic fluxes determined f r o m the model . Thus the objectives o f this study are: \u00E2\u0080\u00A2 T o simulate the trophic interactions among the functional groups o f the marine ecosystem o f E n e w e t a k A t o l l using the E c o p a t h model ing software; and \u00E2\u0080\u00A2 T o modify the E c o p a t h model to simulate the fate o f beta radioactivity, originating f rom a nuclear detonation, w i t h i n the marine ecosystem o f E n e w e t a k A t o l l . 1.2 Enewetak Atoll, location and description. E n e w e t a k A t o l l belongs to the R e p u b l i c o f the M a r s h a l l Islands. This is a y o u n g republic formed in 1987 as one o f the easternmost states o f M i c r o n e s i a in the W e s t Paci f ic Ocean (Figure 1.1). T h e M a r s h a l l Islands are situated on t w o subparallel chains o f extinct volcanoes (Henry and W a r d l a w 1990), the R a t a k and R a l i k meaning \" S u n r i s e \" and \"Sunset\", respectively, in the Marshal lese language (Karo l le 1993). E n e w e t a k itself means \" is land w h i c h points to the east\" (Hel f r ich and R a y 1987), and is situated o n the northwestern extreme o f the R a l i k C h a i n at 1 1 \u00C2\u00B0 3 0 ' N latitude and 1 6 2 \u00C2\u00B0 1 5 ' E longitude. There has been some confusion 2 regarding place names o f the atoll. P r i o r to 1973 the E n g l i s h spell ing o f E n e w e t a k was ' E n i w e t o k ' ( A n o n . 1979), but w i t h the gradual recognit ion o f the native people and the movement o f the M a r s h a l l Islands towards independence (see F i g u r e 1.3), the spelling was changed to acknowledge the native pronunciat ion ( A n o n . 1979). In the case o f island names, up to four different spellings might be found ( D a w s o n 1957) as both native names and the E n g l i s h spelling has changed over time. O n top o f this are the mil i tary code names that were assigned to the islands w h e n the atoll was used for nuclear testing (Henry and W a r d l a w 1990). The atoll is oval-shaped and dominated by a 40 k m long north-south by 32 k m w i d e east-west oriented lagoon w i t h a mean depth o f 48 m and a m a x i m u m depth o f 64 m ( A t k i n s o n et al. 1981). Surrounding the lagoon is a necklace o f small islands and submerged coral reefs (Figure 1.2). \"Surpris ingly , it is difficult to determine the exact number o f islands. D u e to the effects o f storms, small islands are ephemeral, and t w o islands and part o f a third were obliterated by nuclear explosions. Current ly [1987] there are 39 recognizable islands . . . \" (Reese 1987). T h e atoll is situated w i t h i n the belt o f the northeast trade w i n d s and the N o r t h E q u a t o r i a l Current w h i c h moves westward at a speed o f 20 - 50 c m s \" 1 . F o r nine months o f the year the w i n d w a r d side must therefore withstand a constant w a v e attack ( L a d d 1973, A t k i n s o n 1987), w h i c h i n turn brings about a fresh supply o f oxygen, nutrients and f o o d to the aquatic ecosystem. T h e w i n d w a r d side is thus the most v i ta l part o f the reef, concentrating l iv ing organisms and inducing active reef bui lding ( L a d d 1961, 1973). T h e islands consist o f reef debris that is formed o n the reef front and pi led up by the currents, waves, and winds ( L a d d 1973). Hence , the majority o f the islands are concentrated on the w i n d w a r d side compris ing the northeastern and eastern reef perimeter. T h e remainder o f the reef may be divided into three parts w i t h distinct morphologies related to their posit ion relative to the prevail ing winds. These are; the leeward reef o n the southwest, a transitional reef o n the northwest, and a transitional reef o n the southeast (Ristvet 1987) (Figure 1.2). T h e m a x i m u m elevation above sea level o f any island is approximately 4 meter. The total land area is 6.5 k m 2 whi le the lagoon covers an area o f 932 k m 2 ( A t k i n s o n et al. 1981). Because o f the l o w elevation and little land mass, the weather conditions o n the atoll are dictated by the surrounding ocean ( C o l i n 1987b). T h e air temperatures range f r o m 28.5\u00C2\u00B0C in the dry season to 30.0\u00C2\u00B0C in the wet season. T h e wet season stretches f r o m A p r i l to m i d - N o v e m b e r and 4 162\u00C2\u00B0 162\u00C2\u00B010' 162\u00C2\u00B020' 1 \u00E2\u0080\u0094 162\u00C2\u00B030' I I I I I I I I I I I I I I I I I 1 NAUTICAL MILES FEET KILOMETERS I I I Figure 1-2. Enewetak Atol l , Republic of the Marshall Islands, with locations relevant for the present study. Modified from Wardlaw et al. (1991). w i t h i n this per iod, the atoll receives about 8 5 % o f an average yearly rainfall o f 1470 m m ( M e r r i l l and D u c e 1987). T h e atoll is traversed by three channels; the East D e e p Channel w h i c h is 1.5 k m w i d e and 55 m deep, the South Passage w h i c h is 9.7 k m w i d e and 11-22 m deep, and the 4.2 k m w i d e and 2 m deep Southwest Passage consisting o f a network o f small passages rather than a single channel ( A t k i n s o n et al. 1981) (Figure 1.2). 5 W a v e s breaking on the w i n d w a r d reef constantly drive water across the reef flat into the lagoon making this surf-driven inf low the major water input. T h e speed o f the cross-reef current depends o n the height o f the sur f and the tide and ranges between 10 to 150 c m s \" 1 ( A t k i n s o n et al. 1981, A t k i n s o n 1987). T h e tides are semi-diurnal w i t h a m a x i m u m range o f approximately 1.8 m (Wel ls 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 whi le the currents in the D e e p Channel and the Southwest Passage reverse w i t h the tide (every 6.2 hours) ( A t k i n s o n 1987). Three current levels exist in the lagoon w h i c h can be distinguished by their speed and direction. A t the top is a wind-dr iven surface current, 5-15 m thick, m o v i n g southwest at a speed o f 10 cm-s\" 1 ( W a r d l a w et al. 1991). B e l o w this, at 10 to 30 m depth, is a mid-depth current f lowing northeast at a speed o f 2 to 4 c m s \" 1 . F inal ly , at 30 to 50 m depth, a deep current f lows southward at a speed o f 0.5 to 1.5 cm-s\" 1. Despi te this three-layer c irculat ion system, the water i n the lagoon is w e l l mixed w i t h an average salinity o f 34.4 %o and an average temperature o f 2 7 - 2 9 \u00C2\u00B0 C ( W a r d l a w et al. 1991). The average residence time o f the water i n the lagoon is 33 days but may be up to four times longer for water entering across the northern perimeter o f the atoll and somewhat shorter for water entering across the southern perimeter ( A t k i n s o n 1987). 1.3 Atoll formation andDarwins theory of subsidence. D u r i n g his voyage w i t h the H.M.S. Beagle (1831 - 1836) Charles D a r w i n conceived his theory o f reef formation. H e had noticed the existence o f three basic types o f coral reefs: fringing reefs along the shoreline, barrier reefs separated f rom land by a w i d e lagoon, and atolls w h i c h are reefs encircl ing a central lagoon (Blanchon 1997). A t o l l s , he hypothesized, are the last step i n a geological process o f a subsiding volcanic islands ( L a l l i and Parsons 1994). A s a volcanic island fringed by coral reefs s lowly sinks (a process that may happen w h e n a 'newly ' formed volcano presses d o w n on a thin oceanic plate) the coral organisms g r o w upwards t o w a r d the light and outwards t o w a r d the fresh supply o f oxygen, nutrients, and food. I f the coral organisms are successful in keeping up w i t h the speed o f subsidence, a barrier reef w i l l gradually f o r m as the corals closest to the island suffocate i n the debris formed o n the front 6 reef. A s the subsidence continues the central island w i l l eventually disappear leaving a lagoon w i t h a perimeter o f coral reefs (an atoll) ( M a r a g o s et al. 1996). In 1951, D a r w i n ' s theory was confirmed when t w o holes were dri l led at E n e w e t a k A t o l l , penetrating through the limestone cap and reaching the volcanic r o c k basement at depths o f 1267 m and 1405 m, respectively (see Table 1.1). \" T h e limestones recovered were all o f shal low water or ig in demonstrating both subsidence o f the atol l and the u p w a r d g r o w t h o f shal low water corals since E o c e n e time, approximately 49 m i l l i o n years B . P . . . \" ( G r i g g 1982). The limestone was characterized by thick intervals o f unleached aragonite-rich carbonate sediment w i t h w e l l preserved aragonitic fossils alternating w i t h layers where the aragonite had been dissolved. T h e latter represent periods when the atoll s tood above water level and was subjected to the local weather conditions ( L a d d 1973, H e n r y and W a r d l a w 1990). Table 1.1. Stratigraphic subdivisions recognized in holes drilled at Enewetak Atol l . Modified from Ladd (1973). Stratigraphic divisions Depth (m) Post-Miocene 0 - 2 0 0 Upper Miocene 200 - 3 0 0 Lower Miocene 300 - 9 0 0 Upper Eocene 900 - 1400 1.4 Historical events on Enewetak Atoll. E n e w e t a k A t o l l is no doubt one o f the most intensively studied and most abused atolls in the w o r l d . It was a major battle site at the end o f W o r l d W a r II between the U . S . armed forces and Japan w h o had occupied the M a r s h a l l Islands (and the majority o f M i c r o n e s i a ) f rom 1914 (see F i g u r e 1.3). T h e battle damage was augmented w i t h the grounding o f a fully loaded o i l tanker on the w i n d w a r d reef during the A m e r i c a n invasion, possibly causing the death o f long sections o f the s u r f zone ( E m e r y et al. 1954, L a d d 1973). W h i l e Japan's interests i n M i c r o n e s i a had been mainly economic, the A m e r i c a n interests were purely strategic. T h e former Japanese areas were placed under administration by the U n i t e d N a t i o n s and i n 1947 a Trusteeship Agreement was signed by President H . S . T r u m a n establishing the Trust Terr i tory o f the Paci f ic Islands ( K a r o l l e 1993). In the same year, the native people o f E n e w e t a k A t o l l were removed, as officials i n W a s h i n g t o n D . C . announced that the atoll was g o i n g to be used for nuclear testing ( V a n D y k e 1991). 7 F r o m 1948 to 1958, the atoll was part o f the U . S . Paci f ic P r o v i n g G r o u n d s and test site for 43 nuclear detonations (Figure 1.3 and Table 2.3). T h e largest o f these was the M i k e test i n 1952 (part o f ' O p e r a t i o n Ivy ' , F igure 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 H i r o s h i m a bomb, resulting i n the vaporizat ion o f an island ( A n o n . 1998). T h e testing o f nuclear devises damaged or destroyed the vegetation o n all but t w o o f E n e w e t a k ' s islands ( L a d d 1973). T h e military activities o n the atol l also included the construct ion o f buildings, runways, and causeways, the latter connecting islets to facilitate transportation. \"These structures barred cross-reef c irculat ion o f ocean water i n certain areas and radically changed ecological conditions on the reef and i n parts o f the l a g o o n \" ( L a d d 1973). E n e w e t a k also received fallout f r o m the nuclear testing o n B i k i n i A t o l l . This atol l , located u p w i n d and up-current f r o m E n e w e t a k (see F i g u r e 1.1), was part o f the U . S . Paci f ic P r o v i n g G r o u n d s as w e l l , and was the first atoll used for nuclear testing (Operat ion Crossroads in 1946, compris ing t w o tests). M e a n w h i l e , between tests, the atoll was studied intensively by biologists, geologists, oceanographers, and geophysicists ( L a d d 1973). O n e o f the most remarkable, and probably the most cited study o n the atol l , was conducted by H . T . O d u m and E . P . O d u m i n 1954 ( O d u m and O d u m 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 F igure 1.2) o n the w i n d w a r d side o f the atol l i n an area which , at that time, was yet not seriously affected by the nuclear tests. T h e objectives o f the study were t w o f o l d . The first objective was to determine the relationship between the standing stock o f organisms and their product iv i ty per area. This w o u l d give the investigators a rough idea o f the proport ional i ty between the product iv i ty o f 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 w o u l d \" a i d 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 Q S f l 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. T i m e o f C h r i s t . 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 v ic inity o f these inherently stable reef communities, a unique opportunity is prov ided for crit ical assays o f the effects o f radiations due to fission products on whole populations and entire ecological systems i n the f ie ld\" ( O d u m and O d u m 1955). T h e transect was div ided into six zones, as il lustrated in F i g u r e 1.4: the w i n d w a r d buttress zone, coral-algal ridge, encrusting zone, zone o f small coral heads, zone o f large coral heads, and zone o f sand and shingle. LARGE SMALL ENCRUSTING BUTTRES SAND-SHINGLE HEADS HEADS ZONE RIDGE Z 0 N E 1 1 1 r \u00E2\u0080\u0094 i LOW TIDE 100 M M/SEC ' .11 .16 \u00E2\u0080\u00A2 \u00E2\u0080\u00A232 \u00E2\u0080\u00A236 AVERAGE TIDE \u00E2\u0080\u00A240 ...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). D u r i n g a 6 week per iod the organisms w i t h i n each zone, ranging f r o m zooxanthellae w i t h i n coral polyps to sharks i n pelagic waters, were described and grouped into discrete trophic levels compris ing primary producers, herbivores, carnivores, and decomposers. A rough estimate o f the biomass per area for each group was obtained using a variety o f methods. T h e 10 ambitiousness o f the study and the l imited time available meant that \" fewer replications were made than w o u l d be required to obtain m a x i m u m accuracy f r o m each method. Therefore it is the orders o f magnitude w h i c h emerge, but care is taken to base conclusions only on large, probably significant differences\" ( O d u m and O d u m 1955). T h e results o f the groupings and quantification was presented as biomass pyramids, one for each zone. Despite the different taxonomic composit ions i n the six zones and the various errors accompanying the crude biomass estimates, the general shape o f the pyramids was the same and the ratio o f standing crop between trophic levels was estimated as: herbivores/primary producers 18.9%; carnivores/herbivores 8 . 3 % (see F igure 5.2). T h e study moreover showed that the reef was highly product ive compared to other systems (with more than 24 g glucose-m^-day\" 1 or 8760 g glucose-m\" 2-year _ 1 (gross primary production)) . O n a yearly basis the p r o d u c t i o n o f the reef seemed to match the respiration, indicating that the community was at ecological cl imax. G i v e n a standing stock o f about 850 g dry weight-m\" 2 , the production/biomass ratio o f the reef was approximately 12.5 year\" 1. F r o m 1954 to 1986, 1028 scientists visited Enewetak, many returning several times to f o l l o w up o n their field w o r k (Hel fr ich and R a y 1987). T h e scientists were stationed at E n i w e t o k M a r i n e B i o l o g i c a l L a b o r a t o r y ( E M B L ) w h i c h was established i n 1954 and managed by the U n i v e r s i t y o f H a w a i i . The laboratory was initially sponsored by the U . S . A t o m i c E n e r g y C o m m i s s i o n ( A E C ) and later by the U . S . Department o f E n e r g y ( D O E ) (Hel f r ich and R a y 1987, W e l l s and Jenkins 1988). It was run part-time unti l 1974 w i t h the research focusing on increasing the knowledge o f the atoll ecosystem. In 1974 it was upgraded to a full-time laboratory and renamed the M i d - P a c i f i c M a r i n e L a b o r a t o r y ( M P M L ) . T h e focus o f the research changed to include mainly lagoon oceanography, groundwater dynamics, and ciguatera fish poisoning. In 1979 it was decided to phase d o w n the laboratory. A major cleanup / rehabilitation program, initiated in 1977 by the U . S . Defense N u c l e a r A g e n c y ( D N A ) w i t h the objective o f preparing the atoll for the return o f its native people, was about to finish, and funding for the laboratory was running l o w . T h e name was changed for the second time to the M i d - P a c i f i c Research L a b o r a t o r y ( M P R L ) to reflect that the research no longer was confined to the marine environment. In 1982 the laboratory was finally terminated (Hel fr ich and R a y 1987). 11 One o f the results o f the research on the atoll is a col lect ion o f more than 200 reprints o f scientific publications. T h e col lect ion was issued as three vo lumes i n 1976 and a fourth vo lume in 1979 ( A n o n . 1976a-c, A n o n . 1979) 1 . Furthermore, i n 1987, t w o volumes synthesizing the research o f the laboratory's entire history was published (Devaney et al. 1987, H e l f r i c h and R a y 1987). These t w o publications together w i t h the study by O d u m and O d u m (1955) and recently declassified material o n the level o f beta radioactivity i n the b iota o f the a t o l l 2 f o r m the bulk o f material u p o n w h i c h 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 f o l l o w i n g chapter provides the background theory for the thesis, and is divided into three main parts. T h e first part deals w i t h the theory o f radioactivity, describing the processes o f radioactive decay and the effects o f radioactivity o n l iv ing organisms. T h e second part deals w i t h the different types o f nuclear explosions at E n e w e t a k A t o l l , the decay o f the radioactive material produced by these tests, and the uptake o f radioisotopes by marine organisms. Final ly , the E c o p a t h compartment model ing software, w h i c h a l lows for the construction, analysis, and comparison o f mass-balance trophic models, is described. 2.1 Theory of radioactivity. 2.1.1 N a t u r a l and artificial radioactivity. Radioact iv i ty has always been a natural component o f the environment. W h e n the E a r t h was formed, m u c h o f its constituent matter was radioactive. O v e r t ime, this radioactive material has decayed, leaving behind only the isotopes w i t h the longest h a l f l i fe 3 and their decay products. C o s m i c radiation is another source o f natural radioactivity that continuously adds small quantities o f 1 4 C a r b o n , T r i t i u m , and other radioisotopes to the upper atmosphere (Table 2.1). F r o m here they reach the E a r t h ' s surface as part o f the rain ( M a u c h l i n e and Templeton 1964, Bablet and Perrault 1987b, Smith 1994). Radioact iv i ty f r o m various anthropogenetic activities such as f r o m nuclear testing and nuclear waste is, o n the other hand, an artificial source o f radioactivity that has added more and new radioisotopes to the environment (Seymore 1960). 2.1.2 Radioact ive decay. Radioact iv i ty is the spontaneous emission o f excess energy f r o m atoms. A t o m s consist o f a nucleus orbited by electrons. T h e nucleus contains protons and neutrons, and whi le the number o f protons are fixed for every element i n the periodical system, the number o f neutrons may vary. A t o m s w i t h the same number o f protons but a different number o f neutrons are k n o w n as isotopes. Radioisotopes are isotopes in w h i c h the ratio o f 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. T o become stable the nucleus must give o f f energy, a process w h i c h is k n o w n as radioactive decay. Table 2.1. Naturally occurring radioisotopes in sea water. Modified from Seymore (1960) and Bablet and Perrault (1987b). Radioisotope H a l f life 3 Amount Activity (years) (g-1\"1) ( B q - l ' ) b Terrestrial origin: 8 7 Rubidium 4.70-10'\u00C2\u00B0 3.4-10 - 5 0.2 2 3 2 T h o r i u m 1.42-10 1 0 1.0-10 1 0 <0.1 2 3 8 U r a n i u m 4.50-10 9 3.0-10 s 0.1 ^Potassium 1.25-109 4.7-IO\"5 12.3 2 3 5 U r a n i u m 7.13-108 2.1-10\"8 <0.1 2 3 4 U r a n i u m 2.48-10 5 1.9-10''\u00C2\u00B0 -2 3 0 T h o r i u m 7.52-104 <3.0-10 1 3 -2 3 'Protactinium 3.43-10 4 2.0-10' 1 2 -2 2 6 R a d i u m 16.22-102 1.0-10 1 3 <0.1 2 2 7 A c t i n i u m 21.60-10\u00C2\u00B0 <1.0-10 1 5 -2 '\u00C2\u00B0Lead 19.40-10\u00C2\u00B0 1.1-10\"'5 -2 2 8 R a d i u m 6.70-10\u00C2\u00B0 1.4-10\"'6 -2 1 0 Polonium 0.38-10\u00C2\u00B0 2.2-IO - ' 7 -2 3 4 T h o r i u m 0.07-10\u00C2\u00B0 4.3-10\"'7 -Cosmic origin: 1 4 Carbon 55.70-10 2 (2to3)-10- 1 4 <0.1 Trit ium 12.26-10\u00C2\u00B0 1.7-10\"'8 <0.1 a. The radioactive 'half live' is the time it takes for 5 0 % of the radioisotopes to decay; b. Becquerel-litef 1 (becquerel (Bq) = disintegrations per second). Dashes indicate that the activity in sea water is insignificant. Three types o f radiation products are typical ly generated i n a nuclear detonation: fissile non-fission products, fission products, and activation products. Fissi le non-f ission products consist o f P l u t o n i u m and U r a n i u m atoms that d id not undergo fission, i.e., d i d not split apart i n the detonation. T h e y are essentially alpha emitters (see Table 2.2). F i s s i o n products, o n the other hand, are the radioisotopes that are formed w h e n U r a n i u m and P l u t o n i u m atoms do undergo fission. A c t i v a t i o n products are formed w h e n elements f r o m the surrounding environment, or f r o m the nuclear bomb itself, captures neutrons produced i n the detonation. F iss ion and activation products mainly give o f f 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 H e ) from the nucleus: 2 ^U-> 2 3 9 4 0 Th+^He +energy Emission of an ordinary electron (negatron) as a neutron transforms into a proton: 2\u00C2\u00B0 Co\u00E2\u0080\u0094^gNi* + negatron + uncharg ed particle or the emission of an anti-electron (positron) as a proton transforms into a neutron: 2 2 Na\u00E2\u0080\u0094\u00C2\u00BB 2 2 ,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 N i atom produced in the negatron emission example above, is marked by a star to indicate that it is in an excited state, and wi l l emit gamma rays (y): 28 N i * -> 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 ^ / 2 3 0 T n 8.0x10*1 y 2 3 4 T h 24.1 d 2 2 6 R a 210p Q \"* .. ^ r 3.05 mir 2i\u00C2\u00B0Bi / 2 1 4 Bi / , I\u00C2\u00A3bw f^'^ 2 1 2 2 P 3 y V Z\u00E2\u0084\u00A2\u00E2\u0084\u00A2 2 1 4 P b >C r . ' r 26.8 min 2 1 8 0At omv. 2 2 2 R n 3.82 d / J a Decay N 206J| 4.20 min 2 0 6 H g 8.1 min \u00E2\u0080\u0094 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 D e c a y product are themselves often radioactive, forming so-called decay chains that ends w h e n a stable element is formed. A typical example is the U r a n i u m series s h o w n i n F igure 2.1. 2.1.3 The b io log ica l effects o f radiation. A s neutrons, photons, alpha and beta particles travel through l iv ing tissue, they interact w i t h its constituent atoms. D u r i n g this interaction, energy is transferred f r o m the particles to the incident atoms, i n most cases to the orbit ing electrons. I f the gain i n energy is large enough, the electrons w i l l be k n o c k e d away f r o m the atoms, i.e., the atoms are ionized. I f not enough energy is transferred the result w i l l be an excitation o f the atom. A l p h a particles are too large to penetrate the epidermis and do not constitute any serious external risk. O n c e inside the organism, however, they are extremely hazardous. Their size and l o w veloci ty means that for a g iven level o f energy, they are more l ike ly to cause ionizat ion and excitation than, for example, beta particles. A l p h a particles are thus characterized by a rapid loss o f energy and dense ionizat ion o f the nearby tissue. B e t a particles, o n 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 o f the eye and to the epidermis. Photons (gamma-rays), depending o n their energy, may either cause excitation o f atoms, k n o c k away loosely b o u n d electrons, or convert into matter i n the f o r m o f a posi tron and a negatron that, i n turn, may ionize surrounding .atoms. Gamma-rays are the most penetrating type o f radiat ion and may cause severe harm to the whole body. Ionized and excited atoms / molecules are thus the principal result o f absorbed radiation i n organic tissues. This especially applies for water molecules w h i c h constitute the greater part o f the tissue. H e r e , the most important process is generally considered to be: T h e ionized water molecule just formed reacts w i t h another water molecule to give a hydroxyl radical ( O H * ) : H 2 0 ionization > H 2 0 + + e (2.1) H 2 O + + H 2 O H > H ; O + O H * (2.2) 16 The electron formed i n (2.1) readily reacts w i t h water molecules and hydrogen ions to f o r m hydrogen radicals (FT): e ~ + H 2 0 - > H 2 C T - * O r T + H ' (2.3) e - + H + H \u00C2\u00BB H * (2.4) Radicals are extremely reactive. H y d r o x y l radicals are strong o x i d i z i n g 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 i n the tissue, direct alterations o f molecules other than water also occur. A n example is the rupture o f the phosphate sugar backbone i n D N A 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 o f the organism as long as its normal repair mechanisms can keep pace. It is the cumulative effect o f cellular damage that makes radioactivity dangerous. D i v i d i n g 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 M e a s u r i n g radioactivity. The radioisotopes originating f r o m a nuclear detonation do not decay immediately, but have a characteristic probabil i ty o f decay per unit o f time. F o r a large populat ion o f similar radioisotopes, the number that decay per unit t ime (the activity) can be described as the product o f the number o f nuclei (N) and the decay constant (A,) ( K i n n e 1984): d N / d t = - A . - N (2.5) Integrating this equation results i n a negative exponential function o f the form: N = N 0 - e - X t (2.6) where N 0 is the number o f radioisotopes present at t ime zero. Besides X, radioactive decay is often measured by the physical hal f life ( t 1 / 2 ) , w h i c h is the t ime it takes for hal f o f the radioisotopes to decay: 17 N 0 / 2 = N 0 e 1 / 2 => t 1 / 2 = l n 2 / X (2.7) Radioact ive hal f l ives may range f rom fractions o f a second to tr i l l ions o f years (Seymore 1960). 2.1.5 U n i t s o f radioactivity. Radioact iv i ty is measured in number o f disintegrations per unit o f time. T h e SI unit is the becquerel ( B q ) , defined as disintegrations per second. A c t i v i t y is solely a measure o f the energy that is released f r o m the atom and does not distinguish between alpha or beta particles, gamma rays, or neutrons. N o r does it tell anything about the quantity o f energy that is absorbed i n the irradiated matter. This quantity, however, is important w h e n determining the bio logica l effects o f irradiation, because the different forms o f radioactivity differ in their penetration p o w e r 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 ( G y ) , and is defined as an absorbed radiation dose o f one joule per k g . 2.2 Radioactivity from nuclear explosions. 2.2.1 Types o f nuclear explosions. F o r t y three nuclear devices were detonated at E n e w e t a k A t o l l (see Table 2.3). F o r t y one were detonated f r o m towers, barges, the ocean surface, or dropped f r o m aircraft 's whi le t w o were detonated underwater. \" E a c h o f the explosions exerted its o w n special effects on the atolls. The underwater shots released large amounts o f radioactive materials into the water to be absorbed, retained, or passed o n i n the complex marine b io log ica l web, but deposited only minimal amounts o f radioactive substances o n the land areas or in the atmosphere. The tower shots caused spectacular physical disturbance to the islets o f the atolls, often completely obliterating life i n the immediate area, and released radioactive material w h i c h contaminated nearby land and water areas. The third general type o f shot, the high-altitude detonation, probably had little or no effects o n the atolls. A n y radioactive materials released w o u l d have been quickly m o v e d out o f the area by w i n d s \" (Welander et al. 1966). 18 Table 2.3. Dates and location of the military nuclear detonations at Enewetak Atol l . 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\u00C2\u00B0 Enjebi YOKE Apr. 30, 1948 Tower, 63 m 49.0-10\u00C2\u00B0 Aomon ZEBRA May 14, 1948 Tower, 63 m 18.0-10\u00C2\u00B0 Runit DOG Apr. 7, 1951 Tower, 94 m Class. Runit EASY Apr. 20, 1951 Tower, 94 m 47.0-10\u00C2\u00B0 Enjebi GEORGE May 8, 1951 Tower, 63 m Class. Eleleron ITEM May 24, 1951 Tower, 63 m Class. Enjebi MIKE 3 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\u00C2\u00B0 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\u00C2\u00B0 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\u00C2\u00B0 Runit BUTTERNUT May 11, 1958 Barge Class. Runit KOA a 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 OAK a 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. Of f bottom in approximately 50 m of water; c. kt = kiloton, Class. = classified. 19 The nuclear detonations at E n e w e t a k A t o l l were either f ission or fusion (thermonuclear) types. 239 * * F i s s i o n bombs contain U r a n i u m and P l u t o n i u m atoms that split i n hal f under the detonation forming t w o isotopes o f approximately hal f the size o f the original atom. Since the neutron to p r o t o n ratio i n light elements is l o w e r than i n heavier elements, most o f the ' f ission daughter elements' are unstable and give o f f radioactivity. A p p r o x i m a t e l y 2 0 0 isotopes o f 35 elements are created w i t h the detonation o f a fission bomb (see Table 2.4). M o s t have very short hal f lives, but a few have hal f lives o f up to 30 years (Seymore 1960), and U r a n i u m and P l u t o n i u m atoms that d id not split i n the detonation (fissile non-f ission products, see section 2.1.2) have hal f l ives o f up to 4.5-10 9 years (see Table 2.4). In fusion bombs, isotopes o f H y d r o g e n (Deuter ium and T r i t i u m ) fuse under very high 6 233 239 * temperatures (on the order o f 10 \u00C2\u00B0 C , triggered by the fission o f U r a n i u m or Plutonium). F u s i o n bombs have also been termed \"clean devices\" or \"clean b o m b s \" since relatively few radioisotopes are formed (Seymore 1960). Those that are formed include T r i t i u m , D e u t e r i u m and a variety o f induced radioisotopes (activation products, see section 2.1.2 and Table 2.4). Table 2.4. Artif icial 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 1 2 5 Antimony 98.6-101 Tritium 44.5-102 235Uranium 71.0-102 147Promethium 91.3-101 6\u00C2\u00B0Cobalt 19.3-102 24\u00C2\u00B0Plutonium 66.0-102 155Europium 62.1-101 55Iron 94.9-101 2 4 1Americium 45.8-101 Ruthenium-106Rhodium 36.5-101 54Manganese 31.2-101 238Plutonium 86.4-10\u00C2\u00B0 Cesium-144Praseodymium 8.4-101 57Cobalt 27.1-101 Zirconium-95Niobium 65.0-10\u00C2\u00B0 1 1 0 mSilver a 25.0-101 91Yttrium 58.8-10\u00C2\u00B0 6 5Zink 24.5-101 89Strontium 52.7-10\u00C2\u00B0 58Cobalt 70.0-10\u00C2\u00B0 Ruthenium-103Rhodium 40.0-10\u00C2\u00B0 59Iron 45.0-10\u00C2\u00B0 Cesium-141 Praseodymium 32.5-10\u00C2\u00B0 1 4 0Barium 13.0-103 1 3 1 Iodine 80.0-10\"1 140Lanthanum 45.7-10\"4 90Yttrium 73.1-10\"4 a. T h e ' m ' indicates that the silver atom is an isomer, i.e., that it is in a long-lived excited state but eventually w i l l 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. H e r e , the high temperatures f r o m a fission process triggers a fusion process that i n turn releases a flux o f neutrons that can fission more material (Seymore 1960). 2.2.2 D e c a y o f m i x e d fission products. The many radioisotopes created i n a nuclear detonation have different ha l f lives, some longer and some shorter than the average. T h e overal l rate o f decay o f all f ission products combined therefore decreases over time (Seymore 1960): \" F o r the mixture o f all f ission products, radioactivity decreases tenfold for each sevenfold increase i n t ime f o l l o w i n g the detonation i n w h i c h the isotopes were produced. A t this rate the decrease i n activity f r o m one hour after to 343 hours after (approximately t w o weeks) is a thousandfold.\" T h e theoretical gross beta-decay o f mixed s low-neutron initiated f ission products o f 2 3 5 U r a n i u m was estimated by H u n t e r and B a l l o u (1951). O v e r a per iod o f 1 to 1000 days the decay is approximated by a straight line o n a l o g - l o g scale, w i t h an average slope o f -1.2 ( B o n h a m 1958). This f o r m o f decay can be described by a p o w e r funct ion D = a \u00E2\u0080\u00A2 t \" 1 2 , where D is the amount o f radioactivity at time t i n days after the detonation and a is the intercept. A dist inction should be made between decay and decline. D e c a y refers to the decrease in activity determined f r o m a sample kept and measured repeatedly i n the laboratory. Decl ine , o n the other hand, refers to the rate o f change in activity determined f r o m samples collected over time in the same locality. \" I f decline is more rapid than decay a reduct ion o f activity i n the environment beyond that caused solely by physical decay is suggested, and conversely, a steeper decay than decline suggests either an increase i n availability i n the environment or an accumulat ion or concentration o f radioactivity by the organism. E q u a l i t y o f decay and decline suggests that uptake and excretion o f radioisotopes have reached an equi l ibr ium w i t h the environment\" ( B o n h a m 1958). 2.2.3 U p t a k e o f radioactivity by marine organisms. T h e most abundant naturally occurr ing radioisotopes i n sea water ( 'background radioact ivi ty ' ) are 4 0 P o t a s s i u m and 8 7 R u b i d i u m , compris ing approximately 9 0 % and 10%, respectively (see Table 2.1). 4 0 P o t a s s i u m is also the principal radioisotope found i n marine organisms, f o l l o w e d by 1 4 C a r b o n , 2 3 2 T h o r i u m , 2 3 4 T h o r i u m and 2 2 6 R a d i u m (Bablet and Perrault 1987b). 21 W i t h the nuclear explosions at E n e w e t a k A t o l l , a huge number o f artificial radioisotopes were added to the marine environment, initially i n the f o r m o f loca l fallout o n the ocean surface. F r o m here, the radioisotopes were subjected to oceanic di lut ion, dispersion, concentration, and transport ( M a u c h l i n e and Templeton 1964) as illustrated i n F i g u r e 2.2, where the thicker a r r o w indicates the route f o l l o w e d i n 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). B i o l o g i c a l concentration refers to the uptake o f radioisotopes by marine organisms. Depending o n 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 w i t h water engulfed during the formation o f f o o d vacuoles (Sanders and G i l m o u r 1994). H i g h e r organisms may acquire them through the f o o d f o l l o w e d by an absorption in the gut, or they may absorb them directly f r o m the water through g i l l surfaces or other external epithelia ( M a u c h l i n e and Templeton 1964). O n c e w i t h i n the organism, the radioisotopes are treated the same w a y as their stable isotopes, or other similar elements, and are accumulated 22 as non-radioactive elements. Radio iodine , for example, is concentrated i n the thyro id because it is similar to stable Iodine. Stront ium is concentrated i n the skeleton o f vertebrates because o f its similarity w i t h C a l c i u m , and C e s i u m is concentrated i n muscles and flesh due to its similarity w i t h Potass ium (Bablet and Perrault 1987b). The first marine organisms to concentrate fallout material f o l l o w i n g a nuclear test are the primary producers including phytoplankton and algae. W i t h i n a few hours, these organisms can concentrate radioisotopes a thousand fo ld (Donaldson 1959, Seymore 1960, Bablet and Perrault 1987b). P l a n k t o n seems to have no preference but contains most o f the fallout radioisotopes found in sea water (Seymore 1960). T h e most c o m m o n radioisotopes detected i n primary producers w i t h i n a few weeks after a nuclear detonation at E n e w e t a k A t o l l are s h o w n i n Table 2.5. F r o m primary producers, the radioisotopes are disseminated to the rest o f the f o o d web as determined by the prey / predator relationships and the selective uptake o f isotopes by the organisms (see Table 2.5). Table 2.5. Radioisotopes typically detected in marine organisms at Enewetak A t o l l following a nuclear detonation. Based on Bablet and Perrault (1987b), and Donaldson (1959). Type or organisms' Dominant radioisotopes Primary producers 9 5 Zirconium - 9 5 N i o b i u m , 5 7 ' 5 8 ' 6 0 Cobalt , 6 5 Z i n c , 5 5 ' 5 9 Iron, 1 4 U 4 4 C e r i u m (rare earth), 1 0 3 ' 1 0 6 Rufhenium, 5 4Manganese Herbivores (except fish) Cerium (rare earth), Ruthenium-Rhodium Herbivorous fish (e.g., surgeonfish) 6 5 Z i n c , 5 7 ' 5 8 ' 6 0 Cobalt First order carnivores (except fish) 1 4 4 C e r i u m , 1 0 3 , 1 0 6 Ruthenium, 5 7 , 5 8 ' 6 0 Cobal t , 6 5 Z i n c , Zirconium - 9 5 N i o b i u m , n 0 m S i l v e r b , 5 4Manganese, 1 3 7 C e s i u m (only as trace element) First order carnivorous fish 6 5 Z i n c , 5 5 ' 5 9 Iron, 5 7 ' 5 8 ' 6 0 Cobalt , 1 3 7 C e s i u m Second and higher order carnivores 6 5 Z i n c , 1 4 1 , 1 4 4 C e r i u m (rare earth), 5 7 , 5 8 ' 6 0 Cobal t , 1 0 3 , 1 0 6 Ruthenium, traces of 5 4Manganese, Zirconium -9 5 N i o b i u m , and 1 3 7 C e s i u m 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. T h e ' m ' indicates that the silver atom is an isomer, i.e., that it is in a long-lived excited state but eventually w i l l give off gamma rays (usually gamma rays are emitted instantly following an alpha or beta decay). 23 Stront ium and C e s i u m , w h i c h are both considered hazardous to humans because o f their long hal f lives and similarity to C a l c i u m and Potass ium, 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 o f radioactivity f o l l o w e d by the liver, skin, bone, and muscles (Donaldson 1959). 2.3 Compartment modeling. A convenient w a y o f dealing w i t h whole ecosystems and the transfer o f material amongst their components is through the use o f compartment models. H e r e the ecosystem under investigation is d iv ided into a number o f distinct functional groups ( 'compartments ') compris ing either single species or groups o f similar species, and the transport o f material between the groups is described as a flux per unit time. Mathemat ica l ly , the compartments are connected by a set o f either linear or non-linear equations, one for the balance o f each group ( O ' N e i l l 1979). Furthermore, depending o n whether the approach is dynamic or static, the equations may be either differential or non-differential. Compartment models have been used extensively i n the study o f tracers dynamics (see for example the review by O ' N e i l l (1979)). 2.3.1 E c o p a t h . One o f the latest approaches i n compartment model ing is the E c o p a t h software based o n a concept original ly proposed by P o l o v i n a and co-workers ( P o l o v i n a 1984, 1993). It has since been further developed by V . Christensen and D . P a u l y (Christensen and P a u l y 1992a, 1992b, 1995) to incorporate routines for network analysis and system maturity indices based o n the theory o f R E . U l a n o w i c z , H . T . O d u m and E . P . O d u m . It has furthermore been turned into an periodical ly updated software distributed by the International Center for L i v i n g A q u a t i c Resources Management ( I C L A R M ; see http://www.ecopath.org). E c o p a t h is a model ing software that al lows for the straightforward construct ion, analysis, and comparison o f mass-balance trophic models (Vasconcel los et al. 1997) (see F i g u r e 2.3). It is applicable for w e l l defined ecosystems in either 'steady-state' or in w h i c h biomass changes do occur. A n important constraint i n E c o p a t h is that during the time per iod considered, the energy entering any functional group must balance the energy leaving the group plus whatever energy 24 is accumulated w i t h i n the group (the mass-balance concept). Thus , E c o p a t h can be compared to a b o o k keeping system where every flux must be accounted for. A s s u m i n g similar condit ions over the time per iod covered by the model , the trophic interactions among the functional groups o f the ecosystem can be described by a set o f linear mass-balance equations wherein P r o d u c t i o n by (i) = all predation o n (i) + non-predation losses o f (i) + export o f (i) (2.8) w h i c h may also be wri t ten P 1 - M 2 i - P I ( l - E E 1 ) - E X 1 = 0 (2.9) where; P i is the p r o d u c t i o n o f i ; M 2 ; is the predation mortal ity o n i ; E E ; is the ecotrophic efficiency o f i , or the fraction o f the product ion o f i that is consumed w i t h i n the system and exported or harvested ( E E ; is usually left as the u n k n o w n to be estimated when solving E q u a t i o n (2.9)); 1-EEj is the 'other mortal i ty ' , i.e., the non-predation losses o f i , or the fraction o f the product ion o f i that f lows to detritus; and E X ; is the export o f i . E q u a t i o n (2.9) can be re-expressed as: B , ( P / B ) , - \u00C2\u00B1 B 1 \u00E2\u0080\u00A2 ( Q / B ) j \u00E2\u0080\u00A2 D C j i - B , \u00E2\u0080\u00A2 ( P / B ) i ( l - E E ^ - E X , = 0 (2.10) J=I or B , .(P/B). E E , - I B J (Q/BV D C j j - E X , = 0 (2.11) j=i where; B ; is the biomass o f i during the per iod considered; P/B; is the production/biomass ratio o f i w h i c h , under the assumption o f equi l ibrium, is equal to the total mortal ity rate (Z;) (Christensen and P a u l y 1992b); Q/Bj is the consumption/biomass ratio o f i ; and D C j i is the fraction o f prey i i n the average diet o f predator j. B a s e d on E q u a t i o n (2.11), for an ecosystem w i t h n functional groups, a system o f n linear equations can be set up: 25 B, (P/B)j EE, - B j (Q/B), DC;, - B , (Q/B) 2 \u00E2\u0080\u00A2DQ-. . . -B n -(Q/B) n DC;, - E X , =0 (2.12) EL \u00E2\u0080\u00A2 (P/B)2 \u00E2\u0080\u00A2 EE, - f i , \u00E2\u0080\u00A2 (Q/EI), \u00E2\u0080\u00A2 DC, 2 - E L \u00E2\u0080\u00A2 (Q/E!)2 \u00E2\u0080\u00A2 D C ^ - . - B , , \u00E2\u0080\u00A2 (Q/E!)n \u00E2\u0080\u00A2 DC^ - E X , =0 B n \u00E2\u0080\u00A2 (P/B)n \u00E2\u0080\u00A2 E E , - B , \u00E2\u0080\u00A2 (Q/B), \u00E2\u0080\u00A2 DC,n - B , \u00E2\u0080\u00A2 (Q/B)2 \u00E2\u0080\u00A2 DC,,,-. - B , \u00E2\u0080\u00A2 (Q/E!)n\u00E2\u0080\u00A2 D C m - E ^ , =0 This system o f linear equations can be solved using standard matrix algebra (Christensen and P a u l y 1992a, 1992b). O n l y one o f the input parameters: B ; , ( P / B ) i ; (Q/B); or E E ; , may i n general be left u n k n o w n , whi le the diet composi t ion matrix, exports, and harvests always must be provided. The solution o f E q u a t i o n (2.11) al lows calculation o f the energy balance o f each compartment, using C o n s u m p t i o n by (i) = product ion by (i) + respiration by (i) + unassimilated f o o d by (i) ( 2 1 3 ) Rearranging the equation, respiration can be quantified given the other flows: Respirat ion by (i) = consumption by (i) - product ion by (i) - unassimilated f o o d by (i) ( 2 1 4 ) In E c o p a t h , the mass-balance concept implies that Equat ions (2.8) through (2.13) applies for all compartments o f the ecosystem, i.e., that the estimated E E ; range between 0 and 1 (a diagnostic for mass-balance). 2.3.2 E c o r a n g e r The majority o f E c o p a t h models so far have been created w i t h a single set o f mean input parameters for the per iod under consideration, and the researchers were unable to take into account the large uncertainties that tend to accompany b io log ica l data. This rather crit ical point has n o w been solved w i t h the introduct ion o f Ecoranger (Christensen and P a u l y 1995), an E c o p a t h routine that a l lows one to enter, for each input parameter, a mean or mode value, a range, and a distribution. T h e shape o f the distribution depends o n one's pr ior knowledge o f the data and may be either uni form, triangular or normal . O n c e the routine is running, input variables for each parameter type are d r a w n randomly f r o m the specified distributions and the resulting models are evaluated. O n l y models that pass the constraints o f mass-balance and 26 w h i c h are thermodynamical ly possible are accepted. T h e process is repeated i n a M o n t e - C a r l o fashion where the user specifies the number o f realizations and desired successful runs (accepted models). O f the accepted models the 'best m o d e l ' i n a least-square sense (i.e., that w i t h the least square deviat ion f r o m 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 o f a coral reef in the Virg in 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 T h e f o l l o w i n g chapter describes the methodology applied i n the study and is divided into three main parts. In the first part, the modeled section o f the atoll perimeter is defined and the zonat ion across the reef section is described. T h e second part is devoted to the process o f deriving the model input parameters for the seventeen non-fish and ten fish groups included in the model . Last ly , the or ig in and 'processing ' o f the radioactivity data is described, and the theory o f combining the model outputs w i t h the radioactivity data to simulate the observed trends i n radioactivity over time is explained. 3.1 Defining the modeled area. A s mentioned in section 1.2, E n e w e t a k atoll may be div ided into four parts: the w i n d w a r d reef, the leeward reef, and t w o transitional reefs. T h e majority o f the b io log ica l research has been conducted o n the w i n d w a r d reef inc luding the study by O d u m and O d u m (1955) summarized i n section 1.4. The E c o p a t h model was therefore restricted to this area stretching f rom, but not including, E n e w e t a k Island in the south up to, and inc luding, B o g o n Island in the north as shown in F igure 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 T h e w i n d w a r d reef i tself may be divided into distinct zones, perpendicular to the prevail ing westward m o v i n g N o r t h E q u a t o r i a l Current, each zone w i t h a characteristic f lora and fauna ( L a d d 1973). G d u m and O d u m (1955) distinguished between six zones o n their transect (see Figure 1.4), a zonat ion that is fairly typical o f the central and southeastern part o f the w i n d w a r d reef ( C o l i n 1987a). In this study, however, only five zones were distinguished, beginning f r o m the oceanic side: fore reef, algal ridge, reef flat, cora l 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). T h e area o f each zone was determined f r o m a digit ized bathymetric map o f the atol l (see Table 3.1). A short description o f the five zones fo l lows 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) (km 2) Fore reef 0 - 2 0 10.00 0.83 Algal ridge a 0 0.25 d 1.15 Reefflat b 0 0.60 d 3.45 Coral heads 0 - 4 2.00 2.24 Sand / shingle 4 - 2 0 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. A s identified from a bathymetric map (Anon. 1944). 29 3.1.1 F o r e reef. The fore reef area, w h i c h is located seaward o f the algal ridge, has never been accurately determined, but C o l i n (1987a) suggested that it is about 3 to 4 times smaller than the reef flat. It ranges i n w i d t h f r o m about 300 m in the south to less than 100 m i n the north ( C o l i n 1987a). A typical spur and groove system, formed by encrusting corall ine algae ( M a r s h 1970), characterizes the zone immediately seaward o f the algal ridge (Wiens 1962, C o l i n 1987a, Ristvet 1987). Invertebrates such as sea urchins are abundant o n the sides o f the spurs whi le the grooves are f loored w i t h boulders and cobbles that are m o v e d by the currents, preventing organisms f r o m settling. F i s h ( 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 ( C o l i n 1987a). Seaward o f the spur and groove system, the b o t t o m slopes gently d o w n to about 18-23 m depth where a sharp drop-of f begins. A few corals (primarily the vasi form Acropora cytherd) can be found i n this area, however, the b o t t o m is mostly covered by rocks , many w i t h signs o f bor ing c l inoid sponges (Smith and H a r r i s o n 1977, C o l i n 1987a). 3.1.2 A l g a l ridge. T h e spur and groove system leads up to a marginal algal ridge located w i t h i n the surf zone. Some o f the grooves continue beneath the algal ridge and reef flat, forming large surge channels. L o n g sections o f the algal ridge are dead 4 , w h i c h is l ike ly a result o f wart ime o i l pol lut ion ( L a d d 1973) (see section 1.4). In general, the algal ridge is poor ly developed, consisting o f a narrow band o f corals and (mostly soft) algae. O d u m and O d u m (1955) found that the encrusting y e l l o w coral Acropora palmerae, P o c i l l o p o r a , and Millepora platyphylla covered up to 5 0 % o f the area, whi le Smith and M a r s h (1973), o n a transect close to O d u m s ' , found that the coral cover was m u c h less. D o m i n a n t algae o n the algal ridge are the fleshy algae Dictyosphaeria intermedia, Zonaria variegate, Caulerpa elongata, C e r a m i u m , and D i c t y o t a as w e l l as the calcareous red algae Porolithon onkodes ( O d u m and O d u m 1955, Smith and M a r s h 1973). 4 Assumed to be the situation throughout the.period covered by the model. 30 3.1.3 R e e f flat. The reef flat varies i n w i d t h f r o m 90 - 160 m, and is most ly covered by water. It consist o f sandy areas and smooth rocks that slope gradually towards the l a g o o n ( L a d d 1973). T h e zone is paved w i t h corall ine algae such as Jania capillacea and P o r o l i t h o n ( S m i t h 1973b, Smith and M a r s h 1973). C o r a l s are sparse and cover m u c h less than hal f o f the area. The most conspicuous corals are Acropora and Millepora. F i lamentous red, - b r o w n , -and green algae, cyanobacteria and foraminiferans f o r m heavy mats throughout the area ( O d u m and O d u m 1955, L a d d 1973). 3.1.4 C o r a l head zone. The coral head zone is strictly subtidal and located towards the lagoon. It is r i c h o n corals such as rounded heads o f Favia pallida and Cyphastrea serailia, microatol ls (colonies where the central part is dead but the sides are still thriving) o f Porites lutea, branching forms o f Acropora gemmifera, A. cymbicyathus, P o c i l l o p o r a , Stylophora, and the blue coral Turbinaria mesenterina. Strips o f sand, shingle, and cobble runs between the corals. The area varies i n w i d t h f r o m about 200 m i n O d u m s ' study area to about 1 k m i n the north. F i s h are abundant ( O d u m and O d u m 1955, Smith and M a r s h 1973, Johannes and. Gerber 1974, C o l i n 1987a). 3.1.5 Sand / shingle. T h e central l a g o o n is bordered by a terrace dotted w i t h numerous patch reefs. The terrace varies i n w i d t h f r o m a few hundred meters i n the south to more than 1 k m i n the north (Ristvet 1987). T h e depth at the outer edge o f the terrace is about 15-22 m ( W a r d l a w et al. 1991). The patch reefs were ignored i n this study. Instead the area was considered to be uniformly covered by sand and shingle (produced upstream). Foraminiferans and filamentous algae, the latter l iv ing w i t h i n the coral shingle, are very abundant ( O d u m and O d u m 1955). 3.2 Validating the Ecopath model. T h e data used for the construct ion o f the E c o p a t h m o d e l were al l f r o m the published literature. T h e y represent more than 30 years o f research, ranging f r o m the study by O d u m and O d u m (1955) in 1954 to studies w e l l into the 1980s. This rather large time span was 31 justif ied as coral reefs are k n o w n to be very stable systems changing little over time ( O d u m and O d u m 1955). In a few cases, data were imported f r o m similar systems i n other parts o f the w o r l d . Seventeen non-fish and ten fish groups were identified, and biomass estimates for each group and for each zone were obtained. T h e estimates were subsequently averaged into a single weighted biomass estimate for each group. 3.2.1 N o n - f i s h groups. A qualitative description o f the non-fish groups is given below, whi le the E c o p a t h parameters ( B , P / B , and Q/B) are explained in Table 3.3 and the remarks f o l l o w i n g the table. A diet matrix can be found in A p p e n d i x 1. 3.2.1.1 Detritus. Detri tus consist o f dissolved and particulate organic matter ( D O M and P O M ) . D O M stems f r o m phytoplankton, benthic algae, and corals that excrete large fractions o f their primary product ion directly into the water. It is an important source o f energy for filter feeding organisms including bacteria, zooplantkon, bivalves, sponges, polychaetes, tunicates and corals. P O M consist o f dead organic matter including excrements, feces, non-assimilated food, etc. It is co lonized by bacteria and algae, and is consumed by a variety o f filter-feeders and fish ( S o r o k i n 1990). 3.2.1.2 Benthic Primary Producers. This group consist o f all primary producers associated w i t h 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-l iving small algae, boring red and green algae, and cyanobacteria. A l l in al l , 238 species o f benthic algae have been identified at E n e w e t a k A t o l l (see Table 3.2). Benthic algae fragments f r o m the reef front constitute a large fraction o f the plankton over the reef and is ut i l ized by many herbivores and detritivores (Wiebe et al. 1975). 32 Table 3.2. Marine benthic algae at Enewetak Atol l . 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 A u s t i n (1949) measured an extremely l o w concentration o f phytoplankton in the lagoon o f E n e w e t a k A t o l l , supporting the general bel ief that phytoplankton is o f no significance in coral reef ecosystems. S o r o k i n (1993), however, has recently provided several examples o f phytoplankton b looms i n atoll lagoons, and C o l i n (1987a) has o n several occasions observed large phytoplankton b looms i n the lagoon o f E n e w e t a k A t o l l (perhaps an artifact o f the nuclear testing). This inconsistency between observations, and a general lack o f data f r o m reef zones other than the lagoon, meant that the biomass o f phytoplankton was left as the u n k n o w n to be estimated by E c o p a t h (see section 2.3.1). 3.2.1.4 Zooplankton. This group consist o f meroplankton ( ' temporary' z o o p l a n k t o n such as fish -and invertebrate larvae) and holoplankton ( ' ful l-t ime' zooplankton). In a study o f the fish and zooplankton at E n e w e t a k A t o l l , H o b s o n and Chess (1978) discovered that p lankt ivorous fish concentrate in areas o f strong current during the day, where they feed o n z o o p l a n k t o n o f oceanic origin. In contrast, nocturnal planktivorous fish concentrate in areas o f weak currents where they feed o n resident z o o p l a n k t o n that enter the water c o l u m n at night i n concentrations 2-3 times higher than the day time concentrations. \" . . . z o o p l a n k t o n is extremely abundant in reef waters and so is an important component o f the coral reef ecosystem.. . S u c h information as is available shows that the largest z o o p l a n k t o n biomass c o u l d be found at night up the shallow reef areas w i t h patch reefs covered w i t h l iv ing corals or w i t h r u b b l e . . . , i.e., i n places where in accordance w i t h earlier data. . . it ought to be lowest, being depleted by b o t t o m sessile predators, and especially corals\" ( S o r o k i n 1993). There is a general lack o f z o o p l a n k t o n data (and other data) f r o m the fore reef w h i c h is physically very difficult to monitor. H a m n e r et al. (1988), however, was able to sample the w i n d w a r d side o f D a v i e s Reef, Austra l ia , and found that z o o p l a n k t o n is a major source o f 33 energy to the reef. P lankt ivorous fish l iv ing o n the fore reef f o r m a \" w a l l o f mouths\" that effectively removes the z o o p l a n k t o n before the water hits the reef. A s most investigations thus have underestimated the biomass o f zooplankton, it was left to be estimated by E c o p a t h . 3.2.1.5 Corals and sea anemones (Class anthozoa). Thirty eight species o f octocorals (Octocoral l ia) and 169 species o f stony corals (Scleractinia) have been identified at E n e w e t a k A t o l l ( B u r c h 1987, Devaney and L a n g 1987). Sea anemones are m u c h less abundant than corals, and no taxonomic w o r k has been published (Cutress and A r n e s o n 1987). It has been estimated that there is about three times as m u c h primary producer biomass i n corals as there is animal biomass ( O d u m and O d u m 1955). T h e primary producers consisted o f 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 w i t h i n the coral skeleton, it does not receive enough light to contribute significantly to the reefs primary product ion ( L e w i s 1981, M a r s h 1987). A l t h o u g h some corals are capable o f obtaining all their energy f r o m the zooxanthellae ( M a r s h 1987), most hermatypic corals feed autotrophically, as predators, and as filter feeders all at the same time. Cora ls primari ly feed at night, but some also feed dur ing day or at dusk and d a w n ( S o r o k i n 1990, 1993). Scleractinian corals have been s h o w n to feed o n \"copepods, ostracods, mysids, chaetognaths, appendicularians, nematods, polychaetes, small je l ly fish and salps. T h e dominating components i n the gut contents were zoea and copepods. T h e suspended organic material ingested by corals . . . included bacteria, protozoa , detritus, and dead z o o p l a n k t e r s . . . \" ( S o r o k i n 1993). 3.2.1.6 Foraminiferans and other protozoans. A p p r o x i m a t e l y 2 8 0 species o f foraminifera and nonplanktonic protozoans have been identified at E n e w e t a k A t o l l (Chave and Devaney 1987). T h e p r o t o z o a n fauna is fairly typical o f the Western Paci f ic though they are particular scarce o n some o f the northern islands. This , however, may very w e l l be a result o f the nuclear testing that t o o k place i n the area (Hirshfield et al. 1968). 34 Foraminiferans are an important f o o d source for many benthic invertebrates including holothurians, sea urchins, polychaetes and shrimps as w e l l as for fish that graze and scrape the coral surfaces and sandy substrates ( S o r o k i n 1993). L i p p s and D e l a c a (1980) identified approximately 200 shal low water species. A large number o f suspension feeding foraminiferans, many containing zooxanthellae, were found i n cryptic habitats where they l ive protected f r o m their predators. Fi lamentous and mat-l ike types, o n the other hand, have adapted to the high predation pressure through a high turnover rate. Symbiot ic foraminiferans are amongst the most important primary producers i n the sand / shingle zone ( S o r o k i n 1993). Foraminiferans feed either autotrophical ly or o n bacteria, algae, other benthic protozoans, and eggs and larvae o f meiobenthic organisms ( S o r o k i n 1993). 3.2.1.7 Gastropods. A total o f 1116 species o f marine mol lusks have been identified at E n e w e t a k A t o l l , 994 o f w h i c h are gastropods ( K a y and Johnson 1987). Conus , M o u r l a , D r u p a , Thais, and Cypraea are particular abundant o n the reef flat (Renaud 1976, K o h n 1980, M i l l e r 1982). \"Gastropods are m u c h more abundant o n intertidal benches than o n more complex and benign subtidal coral reefs i n the same regions, although species diversity is considerably lower. . . w i t h particular attention to the genus C o n u s \" ( K o h n and L e v i t e n 1976). M a n y gastropods are important predators o n other gastropods and o n polychaetes. The diet o f Drupa morum, e.g., consisted o f 4 4 % vermetids, 4 2 % nereids, 5 % other polychaetes, and 9 % crustaceans ( K o h n 1987). 3.2.1.8 Bivalves. O n e hundred and fifteen different species o f bivalves have been identified at E n e w e t a k A t o l l ( K a y and Johnson 1987), and t w o groups were distinguished i n the E c o p a t h model : tridacnids (giant clams) and 'other bivalves ' . Giant clams, l ike 'other bivalves ' , are filter feeders but, in addit ion, contain symbiotic zooxanthellae w h i c h under favorable condit ions may supply up to 1 0 0 % o f the giant clams energy requirement (Hesl inga and Fit t 1987). I assumed that giant 35 clams, as a group, obtain 7 5 % o f their energy f r o m symbiotic zooxanthellae and the remaining 2 5 % f r o m feeding o n phytoplankton 5 . F e w records o f bivalves other than giant clams were found i n the literature, but S o r o k i n (1993) mentioned that bivalves ( including giant clams) \"comprise 1 0 - 3 0 % o f the r e e f s malacofauna and about the same part o f its total biomass\". M a n y bivalves bore into corals and rocks w h i c h might explain that they have been somewhat over looked. B iva lves are numerous i n the sand / shingle zone (Riddle et al. 1990). 3.2.1.9 Shrimps and lobsters. A p p r o x i m a t e l y 150 species o f decapod shrimps and lobsters (infraorders: Penaeidea, Stenopodidea, Caridea, and Palinura) have been identified at E n e w e t a k A t o l l (Devaney and B r u c e 1987). N o t included in this number are the callianassid shrimp (ghost / b u r r o w i n g shrimp) l iv ing i n the sand / shingle zone. T h o u g h they represent some o f the most abundant infauna in this area, no biomass estimate has been derived (Suchanek and C o l i n 1986, Suchanek et al. 1986). \"Call ianassids are amongst the most elusive o f coral reef animals. Their high density (indicated by the frequency o f their feeding mounds) and high sediment-turnover rates.. . suggest they are major consumers. N o satisfactory technique has yet been developed to quantify these animals or their contribution to total community metabol ism\" (Riddle et al. 1990). Devaney and B r u c e (1987) discovered four species o f lobster o n the w i n d w a r d reef. Panulirus penicillatus, w h i c h lives on the outer reef slope during day and moves onto the reef flat at night, was particularly c o m m o n (Ebert and F o r d 1986). 3.2.1.10 Stomatopods. Stomatopods were included as a group because they often occur i n the diet o f other groups. T w e l v e species o f stomatopods, considerably smaller than the same or related species found in other parts o f the Indo-West Paci f ic area, have been identified at E n e w e t a k A t o l l (Reaka and M a n n i n g 1987). T h e biomass o f the group was left to be estimated by E c o p a t h . 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 primari ly o f crabs f r o m the infraorders B r a c h y u r a and A n o m u r a , but also includes other similar sized crustaceans, as w e l l as amphipods and isopods. Seventy six species o f A n o m u r a n crabs and 291 species o f Brachyuran crabs ( 5 3 % xanthid) have been identified at E n e w e t a k A t o l l (Garth et al. 1987). A c c o r d i n g to K o h n (1987), seven out o f eight o f the most c o m m o n xanthid species studied were herbivores. O n 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 o f Ophiuroidea (basket stars / brittle stars), A s t e r o i d e a (sea stars), E c h i n o i d e a (sea urchins and heart urchins), and C r i n o i d e a (sea lilies). N i n e t y seven species have been identified at E n e w e t a k A t o l l (Devaney 1987a). Br i t t le stars o f the genus Ophiocoma can be found i n all zones (Chartock 1983a), however, open sandy areas are dominated by irregular herbivores sea urchins ( C o l i n 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 shel ls . . . \" (Chartock 1983a). 3.2.1.13 Holothurians. Sea cucumbers have few i f any predators and are very abundant i n the atol l environment. Holothuria atra is, according to K o h n (1987), \"the most conspicuous deposit-feeding invertebrate on interisland platforms.\" T w e n t y species o f sea cucumbers f r o m five genera have been identified at E n e w e t a k A t o l l ( B u r c h 1987, Cutress and R o w e 1987). L a w r e n c e (1980) studied eight o f the most conspicuous species and found a distinct distr ibution o n the reef flat, w i t h only t w o co-occurr ing species. B a k u s (1968) reported an average density o f Holothuria difficilis o f 1 to 32 individuals per 900 c m 2 i n daytime, but up to 200 individuals per 900 c m 2 at night o n the tops o f slab rocks in certain areas. Bacter ia and foraminiferans are major sources o f f o o d for holothurians ( B a k u s 1973), though the foraminiferans probably pass through the digestive tract wi thout m u c h effect, leaving bacteria and organic detritus as the main sources o f energy. A n o t h e r important source o f energy is dissolved organic matter that the holothurians obtain directly f r o m the water. The 37 feces o f H. difficilis contains \" l i v i n g and dead filamentous blue-green and red algae, fish eggs, unidentified detritus, sponge spicules, copepod exuvia, foraminiferans, fragments o f sea urchin spines, holothurian ossicles, gastropods, fish teeth and calcareous fragments\" ( B a k u s 1968). W e b b et al. (1977) estimated an assimilation efficiency for// , atra o f 4 0 % . 3.2.1.14 Polychaetes and other worm like invertebrates. A total o f 132 species o f polychaetes f r o m 34 families have been identified at E n e w e t a k A t o l l (Devaney and B a i l e y - B r o c k 1987). Polychaetes play an important role as bioeroders and as f o o d for a variety o f fish and invertebrates, particularly mol lusks ( S o r o k i n 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 l iv ing w i t h i n the coral heads ( O d u m and O d u m 1955). Polychaetes are suspension feeders, deposit feeders and carnivores, preying o n encrusting invertebrates such as corals ( K o h n 1987). A c c o r d i n g to S o r o k i n (1993), the shal low parts o f a coral reef contains approximately 3 0 % filter feeding polychaetes, 4 0 % detritophages and omnivorous polychaetes, and 3 0 % predatory polychaetes. 3.2.1.15 Sessile invertebrates. Besides sponges, this group comprises hydrozoa, chordates, hemichordates, and other sessile invertebrates. F o r t y species o f sponges have been identified at E n e w e t a k A t o l l . B o r i n g sponges are \"the most c o m m o n infaunal associates o f the corals studied, w i t h 8 6 % o f the corals showing sponge bioerosion effects\" (Devaney 1987b). \" B o r i n g sponges can contribute up to 2 5 % o f the total erosion o f the substratum o n E n e w e t a k . . . and are considered major eroders o n most coral reefs . . . \" (Russo 1980). 3.2.1.16 Cephalopods. Cephalopods were included as a group because they often occur i n the diet o f other groups. N o parameters pertaining particularly to cephalopods f r o m E n e w e t a k A t o l l were found i n the literature. 38 3.2.2 B i o m a s s , P / B , and Q/B values o f 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 i n the E c o p a t h model . Table 3.3. Summary table of the biomass 6, P/B, and Q/B values o f 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 M i s c . 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 o f detritus ( P O M and D O M ) i n the different reef zones. T h e weighted mean for the reef as a w h o l e was 185 t w w - k m \" 2 2) O d u m and O d u m (1955) estimated the biomass o f benthic pr imary producers, including zooxanthellae, i n all o f their zones. I grouped zooxanthellae w i t h their symbiotic counterparts (corals, foraminiferans, and giant clams) and used a ratio o f 16:1 plant biomass to zooxanthellae ( O d u m and O d u m 1955) to derive the zooxanthellae biomass i n corals (see section 3.2.1.5) and subtract it f r o m the O d u m s ' 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. T h e 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. T h e weighted mean for the reef as a w h o l e was 212 t ww-km\" 2 . 4) Table 3.7 summarizes the foraminiferan biomass estimates i n 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. M i l l e r (1982) found a density o f detritus-feeding vermetids (sessile w o r m snails) o n the reef flat o f 151-1084 per m 2 whi le the biomass estimate in Table 3.8 only includes species o f the genus Thais. 6) O d u m and O d u m (1955) estimated the biomass o f small giant clams and small herbivorous mol lusks (gastropods) in the coral head zone. B a s e d o n their comments, I assumed that the small herbivorous mol lusks consisted o f 1/3 giant clams and 2/3 gastropods. This resulted i n a biomass estimate o f giant clams o f 35.94 t w w - k m \" 2 in the coral head zone (conversion: d w = 1 0 % w w ( O p i t z 1996)), and a weighted mean for the reef as a w h o l e o f 6.2 t ww-km\" 2 . 7) R i d d l e et al. (1990) estimated a biomass o f bivalves o f 37 t w w - k m \" 2 i n the lagoon o f Davies Reef, Austra l ia (conversion: 1 g C = 2 g organic matter (ash) ( R i d d l e et al. 1990), and ash = 8 % w w (Sambilay 1993)). A s s u m i n g that this estimate also applies for the lagoon o f E n e w e t a k A t o l l , a weighted mean o f 21 .11 w w - k m \" 2 for the reef as a w h o l e was derived. 8) Ebert and F o r d (1986) estimated a total populat ion o f 7800 lobsters o n the w i n d w a r d reef (reef flat and fore reef zone). W i t h a mean carapace length o f 91.6 m m for males and 81.2 m m for females, assuming a sex ration o f 1:1, and applying a weight- length relationship o f W = 0 . 0 0 2 1 - L 2 7 7 3 , a total biomass o f 0.716 t w w - k m \" 2 was derived (for the fore reef, algal ridge and reef flat). O d u m and O d u m (1955) estimated a biomass o f shrimps o f ~1 g dw-m\" 2 i n the coral head zone. Table 3.9 summarizes the biomass estimates o f shrimps and lobsters. The weighted mean for the reef as a whole was 3.1 w w t-km\" 2 . 9) H e r m i t crabs have been found o n the reef flat in densities ranging f r o m 3 to 65 m\" 2 ( K o h n 1987). Table 3.10 summarizes the biomass estimates for the group. T h e weighted mean for the reef as a w h o l e was 6 t ww-km\" 2 . 40 Table 3.4. Biomass estimates of detritus ( P O M and D O M ) in the different reef zones. Zone P O M a D O M a b P O M + D O M d Biomass 6 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 P O M estimates plus the D O M 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 producers a (t ww-km\"2) Fore reef 3869 c Algal ridge 3520 Reef flat 5640 Coral head zone b 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 polyps 3 (t ww-km\"2) Total biomass (t ww-km\"2) Fore reef 37 467 504 Algal-coral ridge 106 700 806 Reef flat 47 333 b 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 Biomass 3 / mats containing small forams large forams (t ww-km\"2) foraminifera (%) (0.01 cm) (0.1 cm) Fore reef - - - 9.70 b Algal ridge - - - 0.03 c 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 ( A F D W ) 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 o f 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 A F D W of small foraminiferans of 1.33- 10\"7g (conversion: A F D W = 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 25.3 a Odum and Odum (1955) Algal ridge - -Reef flat 5.5 b ' c Odum and Odum (1955) Coral heads 2 5 . 3 M Odum and Odum (1955) Sand / shingle 19.0 e 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\u00C2\u00B0 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 3.93 a Odum and Odum (1955) Sand / shingle 3.93 b -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 9.6 a -Algal ridge 46.8 b Odum and Odum (1955) Reef flat 2.0 b Odum and Odum (1955) Coral heads 9.6 b Odum and Odum (1955) Sand / shingle 1.7C Riddle eta l . (1990) Weighted mean 6.0 -a. Biomass assumed equal to the coral head zone; b. Conversions: dw = 2 5 % 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 Biomass 3 Source (t ww-km\"2) Fore reef 53.6 b Odum and Odum (1955) Algal ridge 157.6 Odum and Odum (1955) 240.0 C Chartock (1983a) Reef flat 13.6 Odum and Odum (1955) Coral heads 53.6 Odum and Odum (1955) Sand / shingle 120.0 d Colin (1987a) Weighted mean 93.1 -a. Conversion: dw = 2 5 % 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 W e b b e t a l . (1977) Sand / shingle 2.6 Riddle eta l . (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 o f polychaetes and other worm-like invertebrates in the different reef zones. Zone Biomass a Source (t ww-km\"2) Fore reef 65 b 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\u00C2\u00B0 Riddle eta l . (1990) Weighted mean 29 -a. Conversion: dw = 2 0 % 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. T h e weighted mean for the reef as a whole was 93.1 t ww-km\" 2 . 11) W e b b et al. (1977) observed an average density o f Holothuria atra o f 3.03 per m 2 in an area similar to O d u m s ' coral head zone. O d u m and O d u m (1955) further estimated the density o f holothurians o n the reef flat and in the coral head zone. Table 3.12 summarizes the various biomass estimates. T h e weighted mean for the reef as a w h o l e was 42 t w w - k m \" 2 . 12) Table 3.13 summarizes the biomass o f polychaetes and other w o r m l ike invertebrates in the different zones. The weighted mean for the reef as a w h o l e was 29 t w w - k m \" 2 . 13) V e r y few biomass estimates were found for the sessile invertebrate group. Bas i le et al. (1984) found that sponges occur in the fore reef zone, and K o h n (1987) found them o n the reef flat as w e l l : \" C l i n o i d sponges that excavate chambers in the hermatypic coral Porites lutea o n interisland platforms are ecological ly the most important P o r i f e r a o f the Enewetak intertidal and shal low subtidal z o n e s . . . \" O d u m and O d u m (1955) estimated a biomass o f 34 g dw-m\" 2 i n the coral head zone. E x c e p t for the sand-shingle zone, I applied this estimate for all zones, w h i c h lead to a weighted mean o f 37 t w w - k m \" 2 (conversion: d w = 3 9 % organic w w (Opitz 1996)). 14) M a r s h (1970) estimated a product iv i ty o f reef-building calcareous red algae o n the algal ridge and i n the spur and groove system o f 4008 g ww-m^-yr\" 1 (conversion: 1 g C = 16.7 g w w ( O p i t z 1996)). U s i n g the biomass estimate f r o m the algal ridge (Table 3.5), a P/B ratio o f 2.2 year\" 1 was derived. B a k u s (1967), using exclosure experiments, measured the primary product iv i ty o f cyanobacteria o n the reef flat and found that it ranged between 0.65 - 2.15 gC-m^-day\" 1 or 3958 - 13111 g w w - m ^ y r \" 1 (conversion: 1 g C = 16.7 g w w ( O p i t z 1996)). A p p l y i n g the biomass estimate for this zone (Table 3.5) lead to a P/B value ranging f rom 0.8 to 2.7 year\" 1. C o m b i n i n g the t w o P/B estimates resulted i n a mean P/B value for the group as a whole o f 2 year\" 1. 15) Gerber and M a r s h a l l (1982) measured a phytoplankton concentration behind the reef o f 1.23 m g C-m\" 3 , and Sargent and A u s t i n (1949) measured a primary p r o d u c t i o n o f 2 m g C-m\" 3-day\"\ F r o m here a P/B ratio o f 593 year\" 1 was derived. 16) F r o m S o r o k i n (1993). 17) F r o m O p i t z (1996). 46 18) A P/B ratio o f 9.1 year\" 1 for zooxanthellae and 1.1 year\" 1 for coral animal polyps ( S o r o k i n 1993) was combined w i t h the biomass estimates f r o m Table 3.6 to give a weighted P/B value o f 2 year\" 1. 19) Cora ls obtain approximately 7 0 % o f their energy f r o m symbiotic zooxanthellae and 30%> f r o m other external sources ( S o r o k i n 1993, example f r o m H e r o n Island, Austral ia) . A s s u m i n g that corals consume 2993 t ww-km^-year\" 1 ( O p i t z 1996) (conversion: 1 k c a l = 1 g w w ) , and using the weighted mean coral biomass estimate f r o m Table 3.6, a Q/B value o f 14 year\" 1 was derived. The value was reduced to 4 year\" 1 to take into account the 70%> internal feeding o n zooxanthellae. 20) H a l l o c k (1981) estimated a turnover rate (P/B ratio) o f 11-16 year\" 1 for three species o f foraminifereans (Amphistegine lessoni, A. lobifera, and Calcarina spengleri) i n the Phil ippines. Since the same species occur at E n e w e t a k A t o l l (Chave and Devaney 1987), an average P/B value o f 14 year\" 1 was applied for the group as a whole. 21) A Q/B value o f 30 year\" 1 (Opi tz 1996) for foraminiferans was l o w e r e d to 21 year\" 1 to take into account that 30%> o f the diet comes f r o m internal feeding o n zooxanthellae (see A p p e n d i x 1). 22) F r o m R i d d l e et al. (1990) . 23) R i d d l e et al. (1990) estimated a yearly consumption rate for large gastropods (> 2 mm) o f 277 k J m\" 2 . T h e y also estimated a biomass o f 31 kJ-m\" 2 (conversion: 1 g C = 42 k J , R i d d l e et al. (1990)). This lead to a Q/B value o f 9 year\" 1. 24) F r o m L e w i s (1981). 25) A r i a s - G o n z a l e s (1993) used a Q/B value o f 10 year\" 1 for a group o f bivalves including Tridacna maxima. I lowered the value to 3 year\" 1 to take into account 7 5 % internal feeding on zooxanthellae (based o n H e s l i n g a and Fit t (1987)). 26) Ebert and F o r d (1986) estimated a natural mortal ity ( M ) for Panulirus penicillatus (spiny lobster) o f 0.284 year\" 1 for males and 0.244 year\" 1 for females. W i t h no fishing, and w i t h a seemingly stable age structured populat ion, M can be assumed to equal the total mortal ity (Z) w h i c h again equals P/B (Christensen and Pauly 1992b). A sex ratio o f 1:1, therefore, resulted in an average P/B value o f 0.264 year\" 1. A P/B value o f 5.34 year\" 1 for shrimps was obtained 47 f r o m A r i a s - G o n z a l e s (1993). A p p l y i n g the biomass estimates f r o m Table 3.9, a weighted P/B value o f 4.6 year\" 1, for the group as a whole , was derived. 27) P a u l y et al. (1993) estimated a Q/B value o f 29 year\" 1 for t w o penaid shrimps, and O p i t z (1996) estimated a Q/B value for spiny lobsters o f 7.4 year\" 1. A p p l y i n g the biomass estimates f r o m Table 3.9, a weighted Q/B value o f 27 year\" 1, for the group as a whole , was derived. 28) F r o m O p i t z (1996), f r o m a group compris ing shrimps, hermit crabs, and stomatopods. 29) F r o m A r i a s - G o n z a l e s (1993), f rom a group dominated by xanthid crabs. 30) F r o m P a u l y et al. (1993). 31) A c c o r d i n g to P a u l y et al. (1993), the natural mortal i ty ( M ) for low-metabol ism echinoderms is approximately equal to the rate constant K (time\" 1 ) o f the v o n Bertalanffy g r o w t h function. K-va lues for some holothurians at E n e w e t a k A t o l l are presented i n Table 3.14. Since there was no harvest o f holothurians i n the per iod considered for the model , the total mortal ity (Z) can be expressed i n terms o f M . Z , however, also equals P/B (Christensen and P a u l y 1992b), and hence a P/B value o f 0.227 year\" 1 was derived f r o m Table 3.14. 32) R i d d l e et al. (1990) found that the infauna o f the D a v i e s R e e f lagoon, Austra l ia , was dominated by polychaetes (Table 3.13). A P/B value o f 5.8 year\" 1 and a Q/B value o f 24 year\" 1 was derived f r o m Table 3.15. 33) F r o m A r i a s - G o n z a l e s (1993). Table 3.14. Rate constants for some holothurians at Enewetak Atol l . Modified from Pauly et al. (1993). Species K (year 1 ) 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 \u00C2\u00B1 9 5 % confidence limits. Feeding type Size class Biomass P/B a Q/B b (mm) (mg C-m\"2) (year 1 ) Macrophagous > 2.0 124 \u00C2\u00B1 5 5 2.6 12 Microphagous > 2.0 217 \u00C2\u00B1 80 3.3 15 Macrophagous 0.5 - 2.0 71 \u00C2\u00B1 16 10.3 37 Microphagous 0.5 - 2.0 224 \u00C2\u00B1 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 F i s h groups. 3.2.3.1 The distribution and abundance of fish. F i s h are very abundant in all o f the M a r s h a l l Islands. F r o m 1953 to 1966 Schultz and collaborators (1953, 1960, 1966) identified and described 543 species, and a checklist o f 817 species (338 genera and 97 families) was recently assembled by R a n d a l l and R a n d a l l (1987). In a comprehensive study, Hiat t and Strasburg (1960) examined the f o o d and feeding habits and ecological relationship o f 223 fish species (56 families and 127 genera) o f the M a r s h a l l Islands. B y comparing this w o r k w i t h that o f Schultz and col laborators (1953, 1960, 1966) and R a n d a l l and R a n d a l l (1987), 190 o f H i a t t and Strasburgs 223 species were found to occur at E n e w e t a k A t o l l . I grouped the 190 species into ten functional groups based on: 1) size: small < 30 c m T L 7 and large > 30 c m T L ; 2) feeding type: herbivorous (parrotfish and surgeonfish), omnivorous (> 1 0 % o f diet consist o f plant material), carnivorous or piscivorous; 3) a data set o n the radioactivity in reef fishes o f B e l l e Island (Figure 1.3), E n e w e t a k A t o l l (Welander 1957) (see section 3.3.1); and 4) t w o E c o p a t h models by A r i a s -Gonzales (1993). T h e ten fish groups were: miscellaneous pisc ivorous fish (mainly sharks and jacks) , small carnivorous fish, large carnivorous fish, small o m n i v o r o u s fish, large omnivorous fish, snappers / groupers, butterflyfish, surgeonfish, parrotfish, and herring (see A p p e n d i x 2 and 3). The diet composit ions ( A p p e n d i x 2) were, i n most cases, derived f r o m H i a t t and Strasburg (1960), w h o examined 2051 fish stomachs, and identified the types o f f o o d consumed. Table 3.16 shows an example o f the stomach context o f Neoniphon sammara (Holocentrus sammara i n Hiat t and Strassburg ( I 9 6 0 ) ) as presented by H i a t t and Strasburg (1960). T o convert the percent c o l u m n in Table 3.16 into a diet c o m p o s i t i o n as required i n E c o p a t h , I counted the number o f different f o o d items and assigned an equal weight to each. In the example f r o m Table 3.16, 12 different items were consumed and thus, were assigned a weight o f 100/12 = 8.3%o each. T h e items were then grouped into the appropriate functional group (as identified i n the E c o p a t h model) and the 'weights ' added to derive the diet composit ion. 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 o f fish containing the item Crustacea Crab fragments 36 Parthenopiid crab 27 Thalamita sp. 18 Pachygrapsus plicatus 9 Portunic crab 9 M a i i d 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 f o o d items were grouped into the f o l l o w i n g diet composit ion: miscellaneous crustaceans 6 6 . 7 % , corals 8 . 3 % , polychaetes 8.3%>, gastropods 8.3%>, and benthic primary producers 8 . 3 % . In a few cases, the diet composi t ions were obtained directly f rom A r i a s - G o n z a l e s (1993) and H o b s o n and Chess (1978). Despite numerous studies o f fish at Enewetak A t o l l , few have been quantitative and then mostly concentrating o n a few species or families occurr ing in a specific habitat (Bakus 1967, M i l l e r 1982). O d u m and O d u m (1955) used visual census to roughly estimate the biomass o f small and large herbivorous and carnivorous fish in each o f their six reef zones (see A p p e n d i x 5). Smal l fish were counted in 36 m 2 quadrates, and their numbers converted into biomass assuming a mean dry weight (dw) o f 2.42 g per individual , as determined f r o m a rotenone sample. L a r g e fish were \"rapidly counted w i t h 360\u00C2\u00B0 underwater v i s i o n \" , and converted to dry weight based on a sample o f 12 speared fish o f the same size (120 g d w per individual) . Table 3.17 summarizes the biomass estimates that were converted f r o m dry weight (dw) to wet weight ( w w ) assuming that for fish except sharks, d w = 26%> w w (based o n Sambilay (1993)). T h e conversion factor used for sharks and derived f r o m O d u m and O d u m (1955) was d w = 2 0 % w w 50 The P/B values for the 10 fish groups (Table 3.17) were obtained f r o m published E c o p a t h models ( A l i n o et al. 1993, A r i a s - G o n z a l e s 1993, Silvestre et al. 1993, O p i t z 1996), while the Q/B values were estimated using the empirical regression by P a u l y et al. (Pauly et al. 1990, Christensen and P a u l y 1992b, Table 3.17, and A p p e n d i x 4): Q / B = 106 3 7 \u00E2\u0080\u00A2 0.0313^ \u00E2\u0080\u00A2 W , - 0 1 6 8 \u00E2\u0080\u00A2 1.38p f \u00E2\u0080\u00A2 1.89H d (3.1) where; W o o is the asymptotic or m a x i m u m weight o f the fish in g r a m wet weight; Tk is the mean annual habitat temperature expressed as 1000/(T\u00C2\u00B0C + 273.1) (an annual mean temperature o f 27.5 was used in all cases based o n A t k i n s o n (1987)); P f is one for apex predators, pelagic predators, and z o o p l a n k t o n feeders, and zero for all other feeding types; and H d characterizes the f o o d 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 Biomass 3 P/B 8 Q/B b (t ww-km\"2-year\"') (year\"1) (year\"1) M i s c . piscivorous fish 6.3 0.3 C 6 Herring 0.4 3.5 d 30 Small carnivorous fish - 2.4 f 14 Large carnivorous fish - 0.6 d 6 Small omnivorous fish - 2.5 d 24 Large omnivorous fish - 2.2 e 9 Snappers / groupers - 0.8 C 6 Butterflyfish 5.7 2.6C 14 Surgeonfish 9.4 1.2C 13 Parrotfish 3.9 2.1 c 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 = 2 6 % ww (Sambilay 1993), except for sharks where dw = 2 0 % 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 T h e or ig in o f the radioactivity data. F r o m shortly before and up to t w o years after the ' N e c t a r ' shot o n M a y 14, 1954 at E n e w e t a k A t o l l (Table 2.3), the level o f beta radioactivity i n the most c o m m o n aquatic organisms was measured, and the results prepared in three reports by the A p p l i e d Fisheries Laboratory , Univers i ty o f Washington, Seattle ( in contract w i t h the U n i t e d States A t o m i c E n e r g y C o m m i s s i o n ) : B o n h a m (1958) studied the radioactivity i n invertebrates; P a l u m b o (1959) reported on the level o f radioactivity in algae; and Welander (1957) l o o k e d at the radioactivity in reef fish. T h e ' N e c t a r ' shot t o o k place about 4.3 k m east-northeast o f B e l l e Island (see Figure 3.1), w h i c h received a greater amount o f fallout than the rest o f the islands (Welander 1957). It therefore became center for subsequent investigations (as w e l l as the focus o f this study). Invertebrates, inc luding bivalves (Tridacna crocea), sea cucumbers (Holothuria atra), gastropods (Lambis), and corals (Acropora, Porites, Pocillopora, and Heliopora), were collected along the seaward side o f B e l l e Island. A total o f 693 specimens o f fish, representing 57 species and 22 families, were collected i n the same area using rotenone, h o o k and line, or spear in water f r o m about 5 c m to about 4 m depth. A l g a e were col lected f r o m the intertidal zone all around the island, while plankton and water samples were col lected o n the lagoon side. A l l samples, except for plankton and water, were immediately put on ice and kept in freezers at the E n e w e t a k field laboratory unti l further processed. F i s h and invertebrates were dissected as to tissue (Welander 1957, B o n h a m 1958): bivalves were dissected into mantle, adductor muscle, g i l l , kidney, visceral mass, and shell; gastropods were dissected into mantle, foot muscle, terminal port ions o f l iver and gut, v iceral mass, and shell; sea cucumbers into gonads, gut w i t h context, and body w a l l ( B o n h a m 1958); and fish were dissected into skin, muscle, bone, l iver and viscera (Welander 1957). Similar tissues f r o m small fish were pooled, or in some cases the entire fish was used (Welander 1957). T h e samples were sent to the Univers i ty o f W a s h i n g t o n laboratory, Seattle, i n insulated containers w i t h dry ice. 52 F i v e mil l i l i ter water samples and filtered plankton samples were placed o n 3.8 c m stainless steel plates and dried and ashed before they were sent to Seattle ( B o n h a m 1958). A t the Univers i ty o f W a s h i n g t o n laboratory, samples o f approximately one gram were placed on pre-weighed 3.8 c m stainless steel plates, weighed and dried at 97\u00C2\u00B0-99\u00C2\u00B0C for 12 to 24 hours. They were subsequently ashed overnight at 500\u00C2\u00B0-550\u00C2\u00B0C. Af ter c o o l i n g and weighing, they were slurried w i t h ethyl a lcohol , spread evenly o n the plates w i t h a glass r o d , dried, and affixed to the plates w i t h a few drops o f 0.5 per cent F o r m v a r i n ethylene dichloride. The samples were counted in Nuc leometer internal gas-flow (methane) count ing chambers, and were corrected back to the date o f col lect ion based o n the decay o f a soil sample collected at B e l l e Island o n M a y 15, 1954 (Donaldson 1953, Welander 1957, B o n h a m 1958, 1959). 3.3.2 Observed trends in radioactivity i n various organisms. V a r y i n g w i t h the organisms, the field samples indicated that a m a x i m u m level o f radioactivity, in the organisms, was obtained between 1 - 1 0 days after the detonation, f o l l o w e d by an approximately linear decline on a log- log scale o f beta decay versus time (up to t w o years) after the detonation ( B o n h a m 1958, P a l u m b o 1959, Welander 1957). F i g u r e 3.3 shows an example o f the observed beta radioactivity i n corals (Acropora) f r o m B e l l e Island as presented by B o n h a m (1958). The figure shows that f r o m day 36 to day 710 after the nuclear detonation, the decline was approximately linear, w i t h a correlat ion o f -0.976, according to the (anti-logged) regression Y = 2.7 \u00E2\u0080\u00A2 108 - T 2 2 3 . The ' N e c t a r ' shot was not the first nuclear detonation in the area, and residual long-l ived radioisotopes f r o m earlier explosions rendered the decline curves less steep than i f they had been a result o f the ' N e c t a r ' shot alone ( B o n h a m 1958). The observed data o n beta radioactivity in the various organisms were based on variable sample sizes as s h o w n in Table 3.18. 53 oTt > 1 O 1000 100 10 1 \u00C2\u00B0 Not used in regression \u00E2\u0080\u00A2 Used in regression 6 pre-detonation level 3 o.i 1 10 100 1000 Days after nuclear detonation on M a y 14, 1954 Figure 3-3. Total beta radioactivity in corals (Acropora) in disintegrations-mm^g'wet-lO 3 after the 'Nectar' shot on M a y 14, 1954 (Bonham 1958). From day 36 to 710 the decline approximates a straight line. Table 3.18. Mean number o f 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 M i s c . pisciv. fish Halimeda, Dictyota, Caulerpa, Lyngbya, Spyridia, Udotea, 17.0 Codium, Microdictyon Acropora 3 . 6 \u00C2\u00B1 0 . 5 b Tridacna crocea 2.1 \u00C2\u00B10.2 Lambis 2.3 \u00C2\u00B1 0.5 Holothuria atra 3.0 Scarus purpureus 2.9 \u00C2\u00B1 2.1 Acanthurus triostegus 2.8 \u00C2\u00B1 1.6 Chaetodon auriga 2.1 \u00C2\u00B1 1.4 Damseffish (Abudefdufbiocellatus), blennies 3.9 \u00C2\u00B1 2.5 30.0 Cardinalfish, squirrefish (Holocentrus sp.), wrasses 3.0 \u00C2\u00B1 1.7 (Halichoeres trimaculatus) Mullet (Neomyxis chaptalli), triggerfish, goatfish 5.4 + 5.7 (Mulloidichthys samoensis) Epinephelus merra 2.8 \u00C2\u00B1 1.7 Jacks, sharks 1.7 + 0.8 a. Mean values \u00C2\u00B1 standard deviation; b. The value refers to the number of plates counted in the gas-flow counting chamber. 54 3.3.3 Radioact iv i ty in whole organisms. T h e data o n radioactivity i n bivalves, gastropods, holothurians and fish were reported as the activity i n various tissues (section 3.3.1) and not as the activity i n the organisms as a whole. T h e relative weight o f the different body parts o f bivalves, gastropods, holothurans, and fish was therefore obtained f r o m 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, 6 3 % muscle, 18% bone, 2 % liver, 9 % viscera 3. Bivalves {Tridacna crocedf 9 5 % 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, 5 4 % body wall / integument (excluding body fluids and respiratory tree)c. Gastropods 7 0 % shell and the remaining 3 0 % equally divided between liver, gut, mantle and muscle d. 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 Simulat ing the observed trends i n beta radioactivity. One o f the outputs o f a balanced E c o p a t h model is the estimated fluxes o f biomass among the functional groups as presented i n the ' f o o d intake matr ix ' . These fluxes were used to simulate the fate o f beta radioactivity w i t h i n the marine ecosystem o f E n e w e t a k A t o l l . A s s u m i n g that the radioactivity is mixed evenly w i t h i n a functional group, one may think o f it as ' tagged' biomass (T) that flows f r o m one functional group to another according to the overal l flux o f biomass between the groups as illustrated i n Figure 3.4, where; B ; and B j are the biomasses (t-km\" 2) o f group i and j , respectively; T; and Tj are the tagged biomasses (t-km\" 2) i n group i and j , respectively; and Q y is the flux o f biomass (t-km\" 2-year\"') f r o m 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. T h e transfer o f radioactivity per unit t ime f r o m group i to j , py-, is proport ional to the fraction o f ' t a g g e d ' biomass to total biomass i n group i , T / B ; , and the flux ( Q y ) o f biomass f r o m group i to j : T <=> Pij = V B; (3.2) where; M y = Qy/B;. M y is the transfer coefficient (year\" 1) f r o m group i to group j , i.e., that part o f the natural mortal i ty o f i that is due to j . The M s are calculated in E c o p a t h and presented i n a predation mortal ity matrix. W h e n dealing w i t h radioactivity, there is an additional loss besides predation, 8, w i t h i n each group, resulting f r o m the physical decay o f the radioisotopes. T h e total or gross beta radiation emitted by fission products, produced i n a process where a number o f 2 3 5 U r a m i u m atoms undergo fission simultaneous, e.g., i n a nuclear detonation, was presented by H u n t e r and B a l l o u (1951) (see also section 2.2.2). They found that the gross decay curve o f the fission products can be described by a power function D = a \u00E2\u0080\u00A2 t b , where D is the amount o f radioactivity at time t i n days after the detonation; a is the intercept; and b is the slope / decay rate (equal to -1.2). W h e n differentiated, the equation may be re-expressed as: d D dt = a - b - t b - ' = b - ( a - t b ) - t _ 1 = b - D - t _ 1 = (b/t)-D = 5 - D (3.3) where; 5 = (b/t). C o m b i n i n g the income, loss and decay terms, the trend in radioactivity in the functional groups may be described by a linear differential equation system o f the form: loss dT ; decay - L = Z T 1 . M 1 J - T J . X M J I - 5 . T J dt i = . (3.4) i=l w h i c h can be integrated over time. 56 The Solver routine in M i c r o s o f t E x c e l was applied to minimize the total sum o f squared deviations between the observed and predicted levels o f beta radioactivity [__(ln(obs/pred)2)] by modifying the predation mortality matrix f r o m the original E c o p a t h model , thus simulating a trend that was more consistent w i t h the observed data: the columns o f the predation mortal ity matrix (i.e., the mortalities exerted by a predator o n its various prey groups) were scaled up / d o w n by mult iply ing them w i t h a factor that Solver was preprogrammed to vary w i t h i n a certain range (0.25-1.75), while m i n i m i z i n g the residuals defined above. T h e r o w s o f the predation mortal ity matrix (i.e., the mortalities experienced by a g iven group) were also modif ied, by a l lowing Solver to add an additional mortal ity ( M + ) to the sum o f the mortalities (sum o f the rows) . T h e coefficient o f determination, R 2 ( = l - S S r e s / S S t o t ) , was used to evaluate the fit o f different runs. T h e modif ied predation mortality matrix was subsequently used to re-calibrate the original E c o p a t h model by modify ing the input biomass, i.e., the inputs directly proport ional to the predation mortalities, used for the next iteration (see E q u a t i o n (2.11)). T o initiate the simulation it was necessary to k n o w the levels o f beta radioactivity at the ' b o t t o m ' o f the f o o d web, i.e., at trophic level one, w h i c h comprises the benthic primary producers, phytoplankton and detritus. T o derive the radioactivity in the benthic primary producers, a linear regression was performed o n the observed concentrations o f beta radioactivity over t ime (Palumbo 1959) (Figure 3.5), and the result used in the simulation. Benthic primary producers S cr m o 8 0 1 2 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 o f the levels o f beta radioactivity in phytoplankton and detritus. It was therefore assumed that phytoplankton incorporates beta radioactivity similar to benthic primary producers, and l ikewise that detritus contains levels similar to benthic primary producers w h i c h are the main contributors to detritus (see section 4.2.1). Radioact iv i ty f r o m earlier nuclear tests i n the area was ignored under the assumption that it had no influence o n the overal l trends. It was further assumed that the organisms do not take up radioisotopes selectively. 58 4. Results T h e f o l l o w i n g chapter is divided into three main parts later to be discussed in Chapter 5. First the process o f balancing the E c o p a t h model using E c o r a n g e r is explained, including some necessary modifications o f the initial input parameters. T h e results are presented in f o r m o f a table o f the basic estimates and as a f o o d web diagram. N e x t , the radioactivity data are mapped onto the f o o d web diagram, and the simulation process inc luding the modifications o f the predation mortal i ty matrix and re-calibration o f the original E c o p a t h model is explained. T h e results are shown graphically and in t w o tables. Final ly , the summary statistics o f the model are presented together w i t h the results o f the network analysis, inc luding the trophic transfer efficiencies estimated by E c o p a t h and a mixed trophic impact diagram. 4.1 Balancing the Ecopath model. 4.1.1 First run w i t h E c o r a n g e r using initial input parameters. T h e E c o p a t h model was balanced w i t h Ecoranger using uni form distributions and a variability o f 9 9 % around the initial input parameters (see Table 3.3, 3.17, A p p e n d i x 1, 2, and 4). The majority o f the data came f r o m the O d u m s ' study (1955) and the high variabil ity assumed for the data was based o n their comment that too few replicates were made to obtain m a x i m u m accuracy f r o m the methods applied, and therefore, all they estimated were orders o f magnitude (see also section 1.4). Ecoranger w i l l not balance a model i f the input are grossly erroneous. In the present case, g iven the initial input parameters, Ecoranger failed to find any solutions (balanced models, see section 2.3.2), and instead displayed a list o f the first 50 runs that failed. In all cases the problem was the ecotrophic efficiency ( E E ) o f surgeonfish, shrimps, miscellaneous crustaceans, and gastropods. Their E E values were m u c h higher than one, suggesting that they c o u l d not sustain the predation pressure. The initial E c o p a t h input parameters were therefore modif ied, w i t h i n the ranges given in the literature (see be low) , unt i l E c o r a n g e r was able to balance the model . 59 4.1.1.1 Modifying the predation mortality experienced by surgeonfish. Surgeonfish initially accounted for more than 5 0 % o f the diet o f miscellaneous piscivorous fish. T h e particular diet composi t ion, however, had been derived f r o m just nine species w i t h detailed diet information (see A p p e n d i x 2). B y incorporat ing less detailed information f rom species such as Triaenodon obesus (Whitetip reef shark), Nebrius ferrugineus (Nurse shark), Sphyraena qenie ( B l a c k f i n barracuda), Carangoides orthogrammus (Island trevally), and Gymnosarda unicolor ( D o g t o o t h tuna) all k n o w n to consume (small) reef fish and planktivorous fish (Lieske and M y e r s 1994), and by consult ing other models ( P o l o v i n a 1984, O p i t z 1993), the f o l l o w i n g modif ied diet composi t ion for miscellaneous pisc ivorous fish was derived, reducing their consumption o f 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. M o r e o v e r , to make the diet composi t ion 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. Smal l omnivorous and carnivorous fish i n particular, caused a high mortal i ty o f shrimps, miscellaneous crustaceans, and gastropods. Af ter comparing the Q/B values o f these t w o fish groups (estimated as arithmetic means rather than weighted arithmetic means as no biomass information was available, see A p p e n d i x 4) , w i t h other models ( A l i n o et al. 1993, O p i t z 1996), the Q/B value was lowered f r o m 14 to 10 year\" 1 for small carnivorous fish and f rom 24 to 15 year\" 1 for small omnivorous fish. T o reduce the mortal ity o f the three invertebrate groups even further, the diet o f butterflyfish was modif ied to include 6 instead o f 12%> shrimps w i t h the difference assigned to corals w h i c h are the only f o o d source for some butterflyfish ( S o r o k i n 1993, L i e s k e and M y e r s 1994). The diet o f echinoderms was modif ied to include less miscellaneous crustaceans (1 instead o f 2%>) and more benthic primary producers w h i c h are the most important f o o d i tem for sea urchins 60 (Ruppert and Barnes 1994) . Cannibal ism by miscellaneous crustaceans was lowered f r o m 10 to 1% by transferring the difference to detritus, as many crabs are scavengers and detritus feeders (Ruppert and Barnes 1994). Final ly , cannibalism by small carnivorous fish was completely removed, whi le it was reduced to 1% and 3 % for z o o p l a n k t o n and cephalopods, respectively, by rescaling their diet composit ions to one. 4.1.2 Second run w i t h Ecoranger using modif ied input parameters. U s i n g the modif ied input parameters, Ecoranger (with similar constraints as outl ined i n section 4.1.1 and a l lowing a m a x i m u m number o f total runs o f 100,000) was n o w able to find balanced models. O f 50 successful runs (restricted by l imited computer capacity) the basic estimates o f the 'best m o d e l ' are shown i n Table 4.1 and graphically in F i g u r e 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 l i S TJ 03 > CO OS CsJ OS cu CU _3 > o a s o 3 - r t c\3 ts o TJ .a CD \u00E2\u0080\u00A2a .3 > (D > J3 CD 03 o o W > o M r3 S o co f s 2 \u00C2\u00B0 CD (D \u00E2\u0080\u009Es TJ +3 1 If --3 cn I i o d O CD co CD CD 3 & S TJ +3 C co c j D ^ o o o o o o o o o o o o o o o o o o o o o e9\" \u00C2\u00A39'53 O O O ^ f ^ ] r S ^ ^ ^ n ^ 0 0 Q ^ 0 ^ 0 0 O ^ ^ ) 0 0 v 0 ^ ^ 0 ^ O ^ t C 0 0 0 ' v n ^ c o i n r ^ i x i ^ o o r ^ q c p O T r ^ 0\ fN r i Q : ~\"\" ~ ; , X '\"< ~ ^ \"\", ~~ ON O N m v o 9 5 r--r- VO o Os i n o CN I/-I i n o i n CN 00 Os CN o o O v d 00 a s v d r n CN VO r-v d v d r- 5-00 OS CN v d CN m Os CN l-H OS O o i r~ o o vo os o C N c N C N C ^ r S C s l r s l T r c N V O C N C N C N C N C N C N C N C N C N C N C N o o o o o o o o o o o o o o o o o s f C S (N o o o o o t ^ c ^ v o o s t ^ c N o o c o s r ^ i n o s c s J < ^ r n c o c N v p v o c \u00C2\u00BB v o o s c > o o o O i - i O H H T r r O C N O H H O m O c n c N H H C N C N O H H C N - H O O O o o o o o o o o o o o o o o o o o o W V D t s M c s l r O c n O S r H H - o ^ r s H H O S O s c N O s r - O O ^ H - O S C N V O o o o o o o o o o o o o o o o o o r\u00E2\u0080\u0094 m m H Os cn o cN 3 H - ^ r i S c N s d r ^ ^ r v b o ^ r ^ CN CN HH c-t \u00E2\u0080\u00A2\u00E2\u0080\u0094i \u00E2\u0080\u00A2 \u00C2\u00AB > l H CO . rH CD 5 rH CD .i= II CN 3 o 3 03 j o TJ 3 a a a TJ CD a. CD CD \" i 1 6 S \u00E2\u0080\u0094 ' O 8 a \u00E2\u0080\u0094 3 CD .>; o i -5 H 3 o 3 TJ \u00E2\u0080\u00A2\u00C2\u00A7 O \"rrt O a M 3 II .a o Cs] VO _ : II : II: : ditL 1 I 5 X o _ _ 111111111 CO. QO; : \u00E2\u0080\u00A2 I ; ; \u00E2\u0080\u00A2\u00C2\u00BB : S-2 m Ml H ill: : I* a. m ::: _V Oi:0i: : : \u00C2\u00A3 : : : C 0 : : : i O : : | 9 A 9 | OjLjCJOJJ, ft \u00C2\u00AB - i 8 s * T 3 -3 3 ^ T3 *- \u00E2\u0080\u0094' o r2 S S 4.2 The fate of beta radioactivity. 4.2.1 M a p p i n g the fate o f beta radioactivity. In F igure 4.2 the fate o f beta radioactivity has been mapped onto the f l o w chart f rom Figure 4.1, omitt ing the f lows for simplicity o f presentation. T h e level o f beta radioactivity was monitored i n 15 o f the 27 functional groups defined in the model . T h e 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 o f detritus (see Table 4.1). T h e level in phytoplankton, the other, but less important primary producer group in the system, was derived in a similar manner. 4.2.2 Simulat ing the fate o f beta radioactivity. T h e results o f the simulations are shown i n Figure 4.3 and 4.4. F i g u r e 4.3 illustrates the simulated trends o f beta radioactivity predicted f r o m the original and the re-calibrated E c o p a t h model , respectively, whi le F igure 4.4 features the results o f modi fy ing only the columns or the r o w s o f the predation mortal ity matrix in the minimizat ion process (see section 3.3.4). M o d i f y i n g the columns thus moves the curves vertical ly up or d o w n (Figure 4.4a), whi le modifying the r o w s changes the slope o f the curves (Figure 4.4b). T o simulate the uptake o f beta radioactivity by the organisms in the first few days after the nuclear detonation, E q u a t i o n (3.4) was integrated over time steps o f 0.1 days. After day nine, daily t ime steps were used. In the minimizat ion process, Solver was preprogrammed to vary the scaling factors (for modifying the columns o f the predation mortality matrix, see section 3.3.4) w i t h i n the range o f 0.25 to 1.75. I f a wider range was used, the resulting predation mortal i ty matrix could subsequently not be used for re-calibration o f the original E c o p a t h m o d e l without seriously v iolat ing the concept o f mass-balance. T h e model is parameterized f rom the t o p - d o w n so that the biomass f lows at the lower trophic levels are estimated to match the f o o d demand o f the upper levels. A s a consequence, the predation mortalities generated by miscellaneous pisc ivorous 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 \u00C2\u00A7 .2 G T 3 W rt ^ ^ \u00E2\u0080\u00A2 \u00C2\u00B0 O U-i <+H >r> H \u00E2\u0080\u0094 <^ \"H E 2 CD o * I o \u00C2\u00A7 CD cf \u00C2\u00A3 T 3 O o OH 03 J D \"o 3 fl CD .as ft* o S CD toO C CN T3 O Parrotfish y = 2.564 + 0.121x R2 = 0.506 Giant clams 12 Time (days) Figure 4-5. Regression o f 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 1 2 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 o f the original and re-calibrated model , respectively, computed by E c o p a t h and useful for comparing the models w i t h each other and w i t h other reef or non-reef ecosystems. Several o f the parameters are quantifications o f O d u m s (1969) 24 ecosystem attributes for assessing ecosystem development and maturity. 4.3.2 Transfer efficiencies. A trophic aggregation routine in E c o p a t h reverses the calculat ion o f fractional t rophic levels and quantifies the trophic f lows on discrete trophic levels sensu L i n d e m a n (1942) (Christensen and P a u l y 1992a). It hereby becomes possible to estimate the transfer efficiency between successive trophic levels and the result o f the analysis for the re-calibrated m o d e l is shown i n Table 4.5. T h e transfer efficiencies may be further split into f lows originating f r o m primary producers and f r o m 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 1 5Finn's cycling index (% of total throughput) FCI - 12.47 18.97 16Finn's mean path length PL - 3.23 3.81 2 1 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 \ T L I II III I V V V I 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 A l l flows 12.9 10.8 14.4 6.2 4.6 3.0 4.3.3 M i x e d trophic impact. F igure 4.7 shows the direct and indirect impact, i n relative terms, that the groups in the rows have o n the groups in the columns. A positive impact is indicated by a sol id bar pointing upwards, whi le a negative impact is indicated by a gray bar point ing downwards . T h e figure is f r o m the re-calibrated model . 72 IMPACTED GROUP ^ \u00E2\u0080\u00A2\u00C2\u00AB -s -s W CO CO CO l3 C+H c+_, > > \u00C2\u00A3 : \u00C2\u00A3 : 6 6 g 9 op o c _ fc \u00C2\u00AB eg g on ca \u00E2\u0080\u00A2J oo o ~ I S 2 >\u00E2\u0080\u00A2 C L *H c/3 CQ C .2 o >*3 u -g 3 ca 00 CL, o -o o ca CD b 3 5 O q CM W C L O I O C L o T3 o cn c^a co ~ ca O CQ O 55 T3 o ca 6 -a S O PH O S CO C O C L C L 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 \u00E2\u0080\u00A2 ' II II I I II | M | I I I Figure 4-7. M i x e d 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 f o l l o w i n g discussion is structured into three parts. F irst some o f the model input parameters are discussed, including the time span covered by the m o d e l and the fish biomass estimates. A l s o discussed here is the role o f herbivorous and plankt ivorous fish o n the reef. T h e second part deals w i t h model outputs, i n particular the role o f detritus recycl ing and ecosystem maturity. Final ly , the re-calibration process and the use o f radioactivity data are discussed, as w e l l as the result o f the simulation and some potentials o f the approach. 5.1 Model input parameters. 5.1.1 T h e t ime span covered by the model . T h e E c o p a t h model was constructed w i t h data f r o m a variety o f sources, representing w o r k spanning several decades. W i t h i n 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 ( A n o n . 1991, D e v a n e y et al. 1987). U n d e r these circumstances it seems appropriate to ask whether it was legitimate to use the data col lected w i t h i n this per iod for the synthesized o f a steady-state / mass-balance model . F e w signs o f destruction, however, were observed in the marine environment. In atmospheric tests (41 o f the 43 test at Enewetak, Table 2.3), the marine organisms were partly protected by the water. W h i l e the explosions at l o w tide ki l led exposed organisms, e.g., those l iv ing in the intertidal zone, organisms l iv ing beyond the algal ridge or hiding under rocks and corals were hardly damaged at all (Bablet and Perrault 1987a). D e a d fish were observed repeatedly i n the v ic inity o f the detonation sites immediately after a nuclear test (Bablet and Perrault 1987a), but were not observed for extended periods o f time, as might have been expected f r o m the elevated levels o f radioactivity (Seymore 1960). Weakened fish w o u l d quickly have been removed by predators, and therefore, w o u l d not have been observed. H o w e v e r , it was still believed that the increase i n radioactivity was not enough to k i l l the fish directly (Seymore 1960). In addit ion, o f the thousands o f fish that were examined, none showed apparent signs o f internal damages, the only exception being the thyroids where damages ranged f r o m zero to a hundred percent ( G o r b m a n and James 1963). 74 Superficially, the damaged fish appeared normal (Seymore 1960). Genet ic effects, w h i c h should show up in the progeny, are not easily studied in situ, but requires that the natural variabil ity o f the populations is perfectly k n o w n (Bablet and Perrault 1987b). Regarding l o w e r trophic level organisms, K n u t s o n and B u d d e m e i r (1973) found that \"results to date indicate that the macroscopic g r o w t h rates and patterns o f [massive] corals are relatively unaffected by uptake o f the observed amounts o f r a d i o a c t i v i t y . . . \" M o r e o v e r , the absorbed lethal radiation dose has been shown to be inversely related to the evolutionary level o f the organism, imply ing that less derived species are less sensitive to radiation (Bablet and Perrault 1987b). B a s e d o n these observations I therefore conclude that it was legitimate to use a w i d e range o f data for the construct ion o f the model , and thus assuming that the reef ecosystem did not change w i t h i n the modeled period. It is very l ikely that the reef ecosystem has been more harmfully affected by physical damages than by radioactivity. A constant leakage o f silt f r o m the craters w o u l d c l o g up the coral polyps and reduce water visibil ity, reducing photosynthesis. H o w e v e r , this problem has to my knowledge not been reported for the part o f the reef modeled here. 5.1.2 F i s h biomass and abundance estimates. 5.1.2.1 Visual census and rotenone sampling. T h e fish biomass estimates (Table 3.17) were either estimated by the p r o g r a m or derived f r o m the O d u m s ' study (1955). T h e latter used rotenone sampling and visual census based o n a method by B r o c k (1954) (see section 3.2.3.1 and A p p e n d i x 5). A c c o r d i n g to R a n d a l l (1963) and Smith (1973 a), v isual census, however, is o f l imited value for quantitative studies o n coral reefs. H u m a n s errors are involved i n estimating the numbers and sizes o f fish and secretive, hole-dwel l ing, and nocturnal species are not observed. Later , i n a rev iew o f the method himself, B r o c k (1982) found that it \"underestimates both the most cryptic as w e l l as the most abundant fish species.. . [and that] only diurnally exposed fish species are censused w i t h any accuracy using the visual census technique.\" H e then compared the results obtained w i t h visual census w i t h results obtained using rotenone, and found the f o l l o w i n g power-75 relationship: Y = 0.74 \u00E2\u0080\u00A2 X 1 1 5 , where X is the number o f 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 populat ion estimate (Smith 1973a). A n idea o f the real populat ion size can be obtained w i t h a Les l ie plot, as s h o w n by P a u l y (1984) for t w o coral reef species (Figure 5.1). H e r e , the catch/effort (where effort is the appl icat ion o f poison) has been plotted as a function o f the cumulative catch, and the intercept o f the regression w i t h the abscissa is an estimate o f the real populat ion number ( N 0 ) . T h e regressions indicate that only about one third o f the populations are sampled w i t h the first appl icat ion o f poison. 8 c \u00E2\u0080\u00A2 Gramma loreto of effort 6 < \u00E2\u0080\u00A2 Kaupichthys hyoproroides Catch/uit 4 2 N 0 1 0 0 5 10 15 20 25 Cumulat ive catch Figure 5-1. Leslie plots for reef eels {Kaupichthys hyoproroides) and a Fairy 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 (N 0 ) . 5.1.2.2 Fish biomass estimates at Enewetak A toll. It can be inferred f r o m the discussion above that O d u m s ' fish biomass estimates o f 6 g d w m \" 2 or 26 g ww-m\" 2 (Table 3.17, A p p e n d i x 5) are highly conservative. T h e real stock size is more l ikely in the area o f 80 g w w - m \" 2 or even higher. The problem was recognized by the authors w h o wrote that \" m o r a y eels were estimated f r o m the rotenone samples o n the surely underestimating assumption that all the morays had cl imbed out into the channels to die .\" B o t h the original and the re-calibrated E c o p a t h model , however, only estimated a fish biomass o f 35 g w w - m \" 2 (Table 4.1). This l o w estimate c o u l d be the result o f using the O d u m s ' 76 estimates as input, and balancing the model w i t h E c o r a n g e r a l lowing for a variabil ity o f 9 9 % . T h o u g h this might seem as a high variability, it is not enough to capture the ' rea l ' stock size. Furthermore, as the model is parameterized f r o m the t o p - d o w n , the rest o f the system is balanced so that the biomass f lows at the l o w e r trophic levels match the f o o d demands o f the upper levels. L o w input values in the top therefore 'scale d o w n ' the w h o l e system. M o r e o v e r , i n the re-calibration process, the biomass estimates were only a l lowed to vary w i t h i n 75%o, and for many o f the l o w e r trophic level groups they were not a l lowed to vary at all. The overal l result, given the constraints, was therefore that the ' rea l ' stock size was never simulated. 5.1.2.3 Comparing the standing stock of coral reef fish. Table 5.1 lists the standing stock o f fish on various reefs. A g a i n , these estimates are probably too l o w . H o w e v e r , assuming that they are too l o w by the same factor, it is still possible to compare among them. E x c e p t for L o o e K e y , the standing stock varies by an order o f magnitude, f r o m 26 to 243 g ww-m\" 2 , w i t h E n e w e t a k A t o l l at the l o w e r end o f the scale. S o r o k i n (1993) presents similar results. R a n d a l l (1963) listed four reasons explaining w h y the standing stock might differ between reefs. T h e first o f these was that the standing stock is largely determined by the amount o f cover / shelter afforded by the reef. A l s o , since reefs typical ly differ substantially in the degree o f sculpturing, measuring only the horizontal plane introduces considerable error when comparing among reefs. The type o f benthic g r o w t h is a third factor affecting the stock size; finally, the fishing effort w i l l tend to reduce biomasses. 5.1.2.4 The abundance and role of herbivorous fish. T h e shal low water fish fauna at E n w e t a k A t o l l , as w e l l as o n many other reefs, is dominated by herbivorous surgeonfish (Acanthuridae) and omnivorous parrotfish (Scaridae) that invade the reef flat w i t h the incoming tide [present study; O d u m and O d u m 1955, B a k u s 1967, W i e b e et al. 1975, O g d e n and L o b e s 1978, L e w i s 1981, M i l l e r 1982, C h a r t o c k 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 Ato l l , (Ecopath) 35 Present study Windward reef, Enewetak Atol l 26 (Odum and Odum 1955) Barrier reef, Moorea (Ecopath) 243 (Arias-Gonzales 1993) Fringing reef, Moorea (Ecopath) 146 (Arias-Gonzales 1993) French Frigate Shoals, Hawai i (Ecopath) b ' c 24 (Polovina 1984) Keahole Pt., Kona, Hawai i 0 185 (Brock 1954) Hawai i , 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, Virg in 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 o f parrotfish is somewhat controversial (Randal l 1963, 1974, B a k u s 1967, Smith and P a u l s o n 1974). Contrary to most studies, Hiat t and Strasburg (1960) found an abundance o f coral polyps in the stomach o f parrotfish f r o m the M a r s h a l l Islands, and characterized them as grazing omnivores. In a comment, Randal l (1967) wrote: \" T h e greater ut i l izat ion o f coral by scarids i n the M a r s h a l l Islands noted by Hiatt & Strasburg may be related to the high coral cover o f the reefs.\" After having dived at Enewetak, R a n d a l l (1974) later wrote: \" W i t h i n the last t w o years the author has dived at both H e r o n Island o n the Great B a r r i e r R e e f and E n i w e t o k in the M a r s h a l l Islands and finds it difficult to explain the apparent difference in the amount o f coral in the diet o f scarids at these t w o localities. M o s t o f the scarid species are c o m m o n to both islands, and there appears to be no notable overal l difference in the amount o f coral cover. T h e coral g r o w t h can vary enormously, o f course, among the marine environments o f the islands. Poss ibly Hiat t and Strasburg's specimens were collected mainly f r o m a zone o f heavy coral but little algal g r o w t h \" (Randal l 1974). In retrospect, the diet composi t ion o f parrotfish used in the present study ( 7 0 % corals, 3 0 % benthic primary producers, based o n Hiat t and Strasburg (1960), see A p p e n d i x 2) is therefore questionable. T h e diet composi t ion is also partly responsible for the high trophic level o f parrotfish (3.46, see Table 4.1) estimated by E c o p a t h . A n o t h e r reason for this high trophic level is that, in the 78 vers ion o f E c o p a t h used here, trophic levels were computed as: 1 + mean trophic level o f prey. Since autotrophy is not included in the calculation, the t rophic level o f partly autotrophic organisms l ike giant clams and corals is in fact too high, w h i c h i n turn affects the trophic level o f the organisms feeding o n them, e.g., parrotfish. A n o t h e r controversy is the dominance o f herbivorous fish o n coral reefs i n general ( O d u m and O d u m 1955, B a r d a c h 1959, Randal l 1963, B a k u s 1966, O g d e n and L o b e s 1978, S o r o k i n 1993). B o u c h o n - N a v a r o and H a r m e l i n - V i v i e n (1981) found that, i n contrast to O d u m s ' study, most authors have estimated that herbivorous fish constitute between 10 and 15% o f the fish fauna o n coral reefs, both in terms o f numbers and biomass. L i k e w i s e , B a k u s (1966) found that coral reef fish generally consist o f roughly 25%> herbivores and 65%> carnivores. H e based this o n the three studies described below, w h i c h I have consulted wi thout being able to reach the same conclusion: \u00E2\u0080\u00A2 B a r d a c h (1959), w o r k i n g on a patch reef in B e r m u d a , found that the weight o f carnivores (many feeding o n a close by seagrass bed) was almost twice that o f omnivores, but o n a large reef found that herbivorous / omnivorous fish outweighed the carnivores by about nine to one; \u00E2\u0080\u00A2 R a n d a l l (1963, 1967) found that Scaridae appear to be the largest family by weight i n most tropical reef areas; and \u00E2\u0080\u00A2 Hiat t and Strasburg (1960) found that carnivores in the M a r s h a l l Islands dominated by number, however, both the present study and the study by O d u m and O d u m (1955) estimated that herbivorous / omnivorous fish dominate by weight. Support ing the findings at Enewetak, S o r o k i n (1993) reported that herbivorous \"reef fish f o r m one o f the most important trophic guilds, w h i c h includes some 10 - 2 0 % o f the total species and 15 - 50%> o f total fish biomass. . . Sometimes their biomass comprises over 50%> even o f the t o t a l . . . \" Final ly , C o l i n (1987a) wrote that the \"general lack o f herbivores as significant as fishes at E n e w e t a k presents an interesting contrast to reefs in some other areas o f the w o r l d . \" There is little doubt that herbivorous fish are m u c h more abundant o n coral reefs than they are i n temperate waters ( S o r o k i n 1993), and play the important role o f channeling primary product iv i ty up the f o o d web (Ogden and L o b e s 1978). Surgeonfish are, according to Hiat t and Strasburg (1960), the most important group o n the reef i n convert ing primary productivi ty 79 into animal tissue. The model predicted that o f the benthic pr imary p r o d u c t i o n that is grazed directly, 6 4 % is grazed by invertebrates and 3 6 % by fish. O f the latter, 6 7 % is grazed by surgeonfish alone. T h e assimilation efficiency o f surgeonfish, however, is quite l o w (Chartock 1983b). M o s t o f the algae they eat are recycled into detritus before they, in the f o r m o f D O M and P O M (see section 3.2.1.1), are ut i l ized by the various suspension and sediment feeders (Chartock 1983b) that are so abundant on the reef (see Table 4.1). Interestingly, the ecotrophic efficiency ( E E ) o f the herbivorous / omnivorous surgeonfish, parrotfish, and butterflyfish are particularly l o w (0.12 - 0.14), imply ing that the majority o f their product ion is recycled to the detritus p o o l . Similar observations have been made on other coral reefs ( O p i t z 1996, Parr ish et al. 1986 (cited i n Opitz)) , and the explanation is not clear. It c o u l d perhaps have something to do w i t h the behavior o f these fish. H e r b i v o r o u s fish are most active during daytime w h e n predators are also more visible and easier to escape. Furthermore, herbivorous fish tend to concentrate i n the shallower parts o f the reef where the primary product iv i ty is highest (Ogden and L o b e s 1978), and it c o u l d be that their predators are not w e l l adapted to the physical conditions at these shal low depths. F o r example, studies o f some coral reefs in the R e d Sea have shown that 7 1 % o f the herbivorous fish concentrated at 0 to 5 m o f depth, whi le the remaining 2 9 % occurred between 10 and 4 0 m ( B o u c h o n -N a v a r o and H a r m e l i n - V i v i e n 1981). 5.1.2.5 The fish fauna and zooplankton on the fore reef. A s i d e f r o m herbivorous fish, coral reefs are characterized by a high abundance o f planktivorous fish feeding o n resident and oceanic z o o p l a n k t o n ( H o b s o n and Chess 1978, S o r o k i n 1993). T h e oceanic z o o p l a n k t o n provides an external input o f energy and nutrients to the reef, and the feces o f the fish feeding o n them provides f o o d for corals and other benthic filter feeders ( S o r o k i n 1993). A c c o r d i n g to W i l l i a m s (1991), the \"tradit ional v i e w o f coral reefs as energetically self-contained ecosystems occurr ing only in clear oceanic.waters ( O d u m and O d u m , 1955) suggests little environmental variabil ity and relatively little variat ion in the structure o f fish communities among reefs. This v i e w is grossly i n error. A s far as fishes are concerned, cora l reefs are not energetically self-contained. A major p r o p o r t i o n o f the fish 80 biomass feeds on z o o p l a n k t o n derived f r o m an external source - the waters surrounding the reef.\" This was clearly s h o w n i n a study by H a m n e r et al. (1988) o n the fore reef o f D a v i e s Reef, central Great B a r r i e r Reef. H e r e they found that oceanic z o o p l a n k t o n was consumed by a \" w a l l o f m o u t h \" formed by the many planktivorous fish on the fore reef. \" M o s t o f the z o o p l a n k t o n in this water is captured and eaten by planktivorous fish w h i c h i n turn defecate onto the reef surface, a process w h i c h enhances the g r o w t h o f corals and benthic algae. B r e a k i n g waves tear benthic algae of f the reef crest and the f loating assemblage o f algal fragments and fecal material w h i c h has not yet settled f lows across the reef top. It is this admixture o f algae, feces, and even sand w h i c h most previous investigators [e.g., at Enewetak A t o l l ] have treated as i f it were oceanic in or ig in (Johannes and Gerber 1974)\" (Hamner et al. 1988). That the situation is probably the same at E n e w e t a k A t o l l is supported by some observations by H o b s o n and Chess (1978, 1986) (see also section 3.2.1.4). In places o f strong currents at the lagoonward side o f the w i n d w a r d reef they observed that the abundance o f oceanic z o o p l a n k t o n increases dramatically at night, and explained it as a consequence o f a general rise o f z o o p l a n k t o n towards the surface waters at night i n the open sea. Inferring f rom the study by H a m n e r et al. (1988) it could , however, also be the consequence o f reduced predation o f z o o p l a n k t o n o n the fore reef at night, as most planktivores fish are diurnal (Hamner et al. 1988). I f this is the case, it w o u l d i n turn explain the high f o o d consumption, estimated by the program, o f zooplankton by corals w h i c h 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 ( H a r m e l i n - V i v i e n 1977). H o w e v e r , H a r r y (1953, cited f r o m Stevenson and M a r s h a l l (1974)), e.g., \"offered the general impression that the outer reef at R a r o i a A t o l l supported fifty per cent o f the fish populat ion o f the entire a t o l l . . . \" 81 5.2 Outputs of the model. 5.2.1 T h e role o f benthic primary producers. U n l i k e most aquatic ecosystem where the secondary p r o d u c t i o n is typical ly phytoplankton driven, benthic pr imary producers are the chief supporter (directly or indirectly through detritus, see be low) o f the secondary product ion o n coral reefs ( L e w i s 1981, P o l u n i n and K l u m p p 1992b, P o l u n i n 1996). Therefore, despite the visible dominance o f corals, benthic primary producers are the most abundant organisms in terms o f l iv ing biomass (Table 4.1). The primary p r o d u c t i o n o n coral reefs is generally very high, and they rank among the most productive ecosystems in the w o r l d ( O d u m and O d u m 1955, L e w i s 1977, S o r o k i n 1990). A t E n e w e t a k A t o l l , as o n many other reefs, the bulk o f the benthic pr imary product ion ( 9 3 % ) is not consumed directly, but is recycled to the detritus p o o l . T h e 7 % that is consumed / grazed directly is a relatively l o w fraction compared to other coral reef systems, e.g., 22 to 3 3 % o n Tiahura reef, M o o r e a Island, F r e n c h Polynesia (Ar ias-Gonza les et al. 1997), 3 6 % at a V i r g i n Island reef ( O p i t z 1996), 3 3 % in the L o o e K e y ' s (Venier 1997), and 43 and 6 5 % on the reef crest o f D a v i e s Reef, central Great B a r r i e r Reef, (Po lunin and K l u m p p 1992a). The 7 % f r o m Enewetak , however, is a weighted mean o f all the reef zones, inc luding the sand / shingle zone w h i c h comprises 5 7 % o f the total area included in the E c o p a t h model (see Table 3.1). T h e intensity o f grazing probably varies considerably between the zones, highest on the outer reef flat ( B a k u s 1967, M i l l e r 1982, C h a r t o c k 1983b, P o l u n i n 1996) and lowest i n the sand / shingle zone. 5.2.2 T h e role o f detritus. T h e ecotrophic efficiency o f most groups i n the system is relatively l o w (Table 4.1), suggesting that m u c h o f their product ion is recycled directly to the detritus p o o l . In turn, approximately 9 2 % o f the detritus is recycled, and 7 1 % o f all f lows i n the system originates f r o m there (Table 4.1 and 4.4). T h e secondary product ion is thus largely dependent o n detritus, a characteristic o f coral reefs in general ( S o r o k i n 1990, A r i a s - G o n z a l e s et al. 1997, O p i t z 1996, V e n i e r 1997). A high degree o f recycl ing is also a characteristic o f mature ecosystems according to O d u m (1969), w h o hypothesized that \" f o o d chains become complex 82 webs in mature stages, w i t h the bulk o f b io logica l energy f l o w f o l l o w i n g detritus pathways. . . heterotrophic ut i l izat ion o f primary product ion i n mature ecosystems involves largely a delayed consumption o f detritus\" Some o f 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). T h e bulk, however, is channeled v ia bacteria into a microbial loop ( A r i a s - G o n z a l e s et al. 1997, A z a m 1998) whereupon it again becomes accessible to the abundance o f benthic filter-feeders inc luding polychaetes, sessile invertebrates (e.g., sponges), miscellaneous crustaceans, bivalves, zooplankton, foraminiferans, corals, and others, that characterizes the reef. T h e importance o f detritus can also be seen in F igure 4.7 w h i c h shows that detritus has a positive impact o n most groups i n the system. F e w fish consume detritus directly, but many o f their prey do and detritus therefore still has a positive impact o n fish, though indirectly. 5.2.3 C o m p a r i n g w i t h other models. 5.2.3. J Trophic transfer efficiencies. T h e transfer efficiencies between successive trophic levels were presented i n Table 4.5. The program estimated relatively high efficiencies at trophic levels II, III, and I V , higher for flows originating f r o m detritus than f r o m primary producers. This indicates that the energy context o f the organic material is higher and more accessible to the organisms after it has been processed / enriched by bacteria. A t higher trophic levels ( V , V I , V I I ) , the efficiencies are considerably lower , and overall decreasing. There is no longer any difference between flows that originated f r o m detritus and f rom primary producers. A s organisms tend to be more mobi le at higher trophic levels, increasingly more energy is lost through respiration at the expense o f being transferred up the system. O p i t z (1996) reasoned that \"the r e e f s 'strategy' is not to achieve high transfer efficiencies between trophic levels but to bui ld up biomass through maintenance o f short cycles for an effective recycl ing o f matter back to its base, the detrital p o o l . . . \" In Table 5.2, the transfer efficiencies f rom E n e w e t a k A t o l l are compared w i t h those o f three other non-fished coral reef ecosystems. There seems to be no clear overal l 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, G E (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, Hawaii b 10.1 4.0 4.1 3.3 - - - -Virgin Islands0 13.1 16.6 9.9 9.0 10.9 11.0 - -Looe Key d 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 w h i c h is no surprise, as many o f the input data for the E c o p a t h model were based o n the O d u m s ' data. T h e biomass i n the present study, however, is distributed over seven trophic levels as compared to three i n O d u m s ' study. T h e more detailed trophic structure is the result o f estimating the trophic levels instead o f assigning them. In the latter case, the complex f o o d web interactions that characterizes most ecosystems (see Figure 4.1) are not taken into account, w h i c h in turn has important implicat ions for the predictability o f such models (see section 1.1). 5.2.3.3 Ecosystem maturity. Quantitative measures are very useful for assessing and compar ing the state and performance o f ecosystems (Dalsgaard and Of ic ia l 1998). O d u m (1969) identified 24 attributes o f ecosystem maturity, hypothesizing h o w ecosystems develop over time. O n the basis o f network analysis, Christensen (1992) quantified several o f O d u m s ecosystem attributes ( A p p e n d i x 6) and used a selection o f them to ranked 41 steady state f o o d web models compris ing ponds, lakes, rivers, temperate and tropica l coastal areas, cora l reefs, tropical shelves and upwell ings. O f the 41 systems, cora l reefs were found to rank intermediate in maturity w i t h lakes / rivers ranking lowest and coastal areas highest. A c c o r d i n g to O d u m (1969), the capacity o f an ecosystem to entrap, w i t h h o l d , and cycle nutrients increases w i t h maturity. The degree o f recycl ing can be measured by F i n n ' s cycl ing index ( F C I , see A p p e n d i x 6), w h i c h expresses the fraction o f the total system throughput that is recycled (F inn 1976, Christensen and P a u l y 1992a, 1993). W h e n Christensen and P a u l y (1993) ranked the 41 systems, mentioned above, after this index, they found a strong correlation w i t h the maturity ranking by Christensen (1992). Similarly , i n a study o f ecosystem stability, Vasconce l los et al. (1997) showed that recycl ing plays an important role in the maintenance o f ecosystem stability as does path length, the average number o f groups that a unit o f f l o w passes through (Christensen and P a u l y 1993, V a s c o n c e l l o s et al. 1997). In F igure 5.3, E n e w e t a k A t o l l is compared w i t h three other coral reef ecosystems. T h e figure shows that; a) coral reefs w i t h l o w Pp/R (net primary production/respiration) ratio display high degree o f recycl ing; and b) recycl ing is posit ively correlated w i t h path length. B o t h trends are consistent w i t h O d u m s theory o f ecosystem maturity (Christensen and P a u l y 1993, O d u m 85 1969), and it can be inferred f rom the figure that E n e w e t a k A t o l l is more 'mature ' or stable (Vasconcel los 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 o T3 O ft O g CD O cj I o \"o o CD CD > .3 CJ CO T 3 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). 3 0 % 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 3 0 % , 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 9 0 % ; 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). 7 5 % 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\u00C2\u00A7JBT UBTJ snojOAruiBD rpuis SSJOAlDSld 3STJV l^uauiSag irst J S3]Rjq3U3AUT 3ffSS3S SajaEipAJOd suii3pounio3 spodonsEo S3ATBAIR pdurmig SUB33B18TU3 OSIW r spodoiBiuois spodoieqdao (^ Xuois / yos) sfoioo uoi^uEidoo^ 3B\u00C2\u00A7rv o o o o o d \u00C2\u00A9 ei m o o n n o o \u00C2\u00A9 \u00C2\u00A9 \u00C2\u00A9 \u00C2\u00A9 vo s\u00C2\u00A3> d d \u00E2\u0080\u0094 o o \u00E2\u0080\u0094 o \u00E2\u0080\u0094 \u00E2\u0080\u0094 d d d d vo \u00C2\u00A9 \u00E2\u0080\u0094' r> o\ r-o o o o d d -a a n v ^puaurSBij usi^ satEjqauaAui snsssg pduruirs EUB33B1STU3 '3SnAJ spodotEuiois spodoisqdao J A U O J S / U O S ) SpjJOQ uoi^iiEidoo^l s R R O H o f X '-5 c CU o. o. O T3 CD c+4 0 0 43 \u00E2\u0080\u00A2 - i P H H 00 C O OH 3 > X A p p e n d i x 2, Table 2. D i e t matrix o f the ten fish groups included in the model . B a s e d o n A p p e n d i x 2 Table 1. tH \u00E2\u0080\u0094 \u00C2\u00AB CJ X J H X I rv X X trt ^ ' f g W \u00C2\u00AB Prey \ Predator \u00E2\u0080\u00A2\u00C2\u00A7 'I ' 5 '3 '\u00C2\u00A7 P n O O C J i n g o t g O W i-l c o c o 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 - - - - - - - -I l l A p p e n d i x 3. Scientific and c o m m o n names o f the fish species inc luded i n the m o d e l 3 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 wrasse d Threespot wrasse Pink-belly wrasse Blakspotted wrasse Cortez rainbow wrasse Sixbar wrasse Snooty wrasse Silver sweeper Arc-eye hawkfish 1 1 2 A p p e n d i x 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 L a r g e 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 Yel low 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 B i r d 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 A p p e n d i x 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 A p p e n d i x 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 ? L a r g e 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? Yel low 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 A p p e n d i x 3 (continued). Scientific name b 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 Parrotf ish 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. B y comparing this work with that o f Schultz and collaborators (1953, 1960, 1966) and Randall and Randall (1987), 190 of Hiatt and Strasburgs 223 species were found to occur at Enewetak Atol l . 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 A p p e n d i x 4. D e r i v i n g the Q/B values o f the ten fish groups. T h e Q/B values for the fish species included in the model were estimated using the empirical regression by P a u l y et al. (Christensen and P a u l y 1992b, P a u l y et al. 1990): Q / B = 1 0 6 3 7 \u00E2\u0080\u00A2 0.0313^ \u00E2\u0080\u00A2 W ^ \" 0 1 6 8 \u00E2\u0080\u00A2 1.38p f \u00E2\u0080\u00A2 1.89H d where; W r o is the asymptotic or m a x i m u m weight o f the fish i n gram wet weight; T k is the mean annual habitat temperature expressed as 1000/(T\u00C2\u00B0C + 273.1) (an annual mean temperature o f 27.5 was used i n all cases, based o n A t k i n s o n (1987)); Pf is one for apex predators, pelagic predators, and z o o p l a n k t o n feeders, and zero for all other feeding types; and H d characterizes the f o o d type and is set to one for herbivores and zero for carnivores. Estimates o f W m a x were obtained f r o m FishBase (1998) for as many species as possible. In many cases W m a x was given directly, but i n some cases it had to be estimated using the weight-length relationship: W m a x - a - L b ; where a and b are t w o constants and L is the total length ( T L ) (see Table 1). W m a x was converted to W B assuming that W o o = W m a x / 0 . 9 5 (see Table 2). A p p e n d i x 4, Table 1. D e r i v i n g the m a x i m u m weight o f fish species i n the model . B a s e d o n FishBase (1998). Species LnjJ3 & t / Place of origin W w J (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 A p p e n d i x 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 A p p e n d i x 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. W m a x 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-L b 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. A l l values are from FishBase (1998). T L = total length, S L = standard length, F L = fork length; W m a x is estimated using T L , and SLs and F L s were therefore converted to T L 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 A p p e n d i x 4, Table 2. Q/B values o f the fish groups included i n 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 A p p e n d i x 4, Table 2 (continued). Species a W m a x (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 A p p e n d i x 4, Table 2 (continued). Species a W m a x (g) wj 5 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. W m a x 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\u00C2\u00BB = W m a x / 0 . 9 5 ; c. Characterizes the food type: Value of one for herbivores (in this study when 3 0 % 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 = 10 6 3 7 -0.0313 T k -W c o \" 0 1 6 8 -1 .38 p f -1 .89 H d where; W\u00E2\u0080\u009Ec is the asymptotic or maximum weight of the fish in gram wet weight; T k is the mean annual habitat temperature expressed as 1000/(T\u00C2\u00B0C + 273.1) (an annual mean temperature of 27.5 was used in al l cases based on Atkinson (1987)); P f is one for apex predators, pelagic predators, and zooplankton feeders, and zero for all other feeding types; and H d 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 A p p e n d i x 5. F i s h biomass estimates in O d u m and O d u m (1955). A p p e n d i x 5, Table 1. D r y weight estimates o f fish for each o f the zones included i n O d u m and O d u m (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% w w (Vinogradov 1953). 123 A p p e n d i x 5, Table 2. W e i g h t e d mean fish biomass estimate across the w i n d w a r d reef as derived f r o m O d u m and O d u m (1955) (see also A p p e n d i x 5, Table 1). Reef area \ fish group Herbivores Carnivores Total Total 3 (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 = 2 6 % ww (Sambilay 1993). 124 A p p e n d i x 6. L i s t o f the ecosystem maturity attributes defined by O d u m (1969) that are quantified i n E c o p a t h (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 "@en . "Thesis/Dissertation"@en . "1999-05"@en . "10.14288/1.0074838"@en . "eng"@en . "Resource Management and Environmental Studies"@en . "Vancouver : University of British Columbia Library"@en . "University of British Columbia"@en . "For non-commercial purposes only, such as research, private study and education. Additional conditions apply, see Terms of Use https://open.library.ubc.ca/terms_of_use."@en . "Graduate"@en . "Modeling the trophic transfer of beta radioactivity in the marine food web of Enewetak atoll, Micronesia"@en . "Text"@en . "http://hdl.handle.net/2429/8910"@en .