SYMPOSIA SERIES FOR UNDERSEA RESEARCH * NOAAS UNDERSEA RESEARCH PROGRAM VOL. 1 NO. 1, 1983 3yr*p.W. V«M No I The Ecology of Deep and Shallow Coral Reefs WHOI DOCUMENT COLLECTION 5% 5 1 9 $3 U.S. DEPARTMENT OF COMMERCE National Oceanic and Atmospheric Administrat ^"HE^Of^ Oi i =0 : n- , □ o D ! D I CD SYMPOSIA SERIES FOR UNDERSEA RESEARCH NOAA'S UNDERSEA RESEARCH PROGRAM. VOL. 1 NO 1. 19S3 The Ecology of Deep and Shallow Coral Reefs Results of a Workshop on Coral Reef Ecology held by the American Society of Zoologists, Philadelphia, Pennsylvania, December 1983 Edited by Marjorie L.Reaka Department of Zoology University of Maryland Washington, D.C. December 1983 > c o l-» rt> o z U.S. DEPARTMENT OF COMMERCE Malcolm Baldrige, Secretary National Oceanic and Atmospheric Administration John V. Byrne, Administrator Oceanic and Atmospheric Research Ned A. Ostenso, Assistant Administrator Office of Undersea Research Elliott Finkle, Director Symposia Series for Undersea Research The National Oceanic and Atmospheric Administration's Office of Undersea Research, through the NOAA Undersea Laboratory System (NULS) begun in 1977, has provided manned underwater facilities and other research support for investigations of coastal marine environments including biological, geological, and ecological problems. There are currently five National Undersea Research Facilities which operate under cooperative agreements between universities and NOAA. The facilities are located at: the West Indies Laboratory of Fairleigh Dickinson University, the University of North Carolina at Wilmington, the University of Connecticut at Avery Point, the University of Southern California at the Catalina Marine Science Center, and the University of Hawaii at Makapuu Point on Oahu . NOAA's Undersea Research Program funds scientists to conduct research at these facilities which is in line with NOAA's research needs and priorities in the areas of: (1) fisheries, (2) marine pollution, (3) sea floor properties and processes, and (4) ocean services. Since its inception, NOAA also has encouraged and supported the use of submersibles to perform in-situ underwater observations and data gathering. Many shallow-water submersible missions have been supported by the Undersea Research Program, including the Johnson S e a - L i n k , the Nekton-Gamma , and the Mermai d , as well as deep-water missions using the A 1 v i n . Submersible program goals are to support research from NOAA's major program elements, and the submersible needs of the Sea Grant College system. The Symposi a Series for U ndersea Research has been developed specifically to provide a publishing medium for national symposia whose contents have dealt with major questions, problems, or syntheses resulting from NOAA's Undersea Research Program. Information about the Symposi a Series for Undersea Research, other results of NOAA's Undersea Research Program, the National U nde rsea Resea rch Facilities, or NOAA's submersible program can be obta i ned from : NOAA's Undersea Research Program R/SE2 6010 Executive Boulevard Rockville, MD 20852 Photo Credits: All photographs were taken by Bruce Nyden and Dale Anderson of the Hydrolab Facility unless otherwi se stated . ii TABLE OF CONTENTS CHAPTER I. INTRODUCTION. Marjorie L. Reaka 1 CHAPTER II. GROWTH AND LIFE HISTORY PATTERNS OF CORAL REEF ORGANISMS 7 Size structure and growth rates in populations of colonial and solitary invertebrates. Kenneth P. Sebens 9 Life histories and growth of corals over a depth gradient. Terence Hughes 17 Depth-related changes in the colony form of the reef coral Pontes astreoides. WillemH. Brakel 21 CHAPTER III. THE DYNAMICS OF RECRUITMENT IN CORAL REEF ORGANISMS 27 Reef fishes at sea: ocean currents and the advection of larvae. Phillip S. Lobel and Allan R. Robinson 29 On the possibility of kin groups in coral reef fishes. Douglas Y. Shapiro 39 Settlement and larval metamorphosis produce distinct marks on the otoliths of the slippery dick, Halichoeres bivittatus. Benjamin C. Victor 47 CHAPTER IV. THE ORGANIZATION OF CORAL REEF COMMUNITIES 53 Sponges as important space competitors in deep Caribbean coral reef communities. Thomas H. Suchanek, Robert C. Carpenter, Jon D. Whitman, and C. Drew Harvell 55 Distribution of sweeper tentacles on Montastraea cavernosa. Elizabeth A. Chornesky and Susan L. Williams 61 Relationships between fishes and mobile benthic invertebrates on coral reefs. Nancy G. Wolf, Eldredge B. Bermingham, and Marjorie L. Reaka 69 Fish grazing and community structure of reef corals and algae: a synthesis of recent studies. Mark A. Hixon 79 Coral recruitment at moderate depths: the influence of grazing. H. Carl Fitz, Marjorie L. Reaka, Eldredge Bermingham, and Nancy G. Wolf 89 Between-habitat differences in herbivore impact on Caribbean coral reefs. Mark E. Hay and Tim Goertemiller 97 Quantifying herbivory on coral reefs: just scratching the surface and biting off more than we can chew. Robert S. Steneck 103 in Differential effects of coral reef herbivores on algal community structure and function. Robert C. Carpenter 113 Nearshore and shelf-edge Ocul ina coral reefs: the effects of upwelling on coral growth and on the associated faunal communities. John K. Reed 119 CHAPTER V. THE ORGANIZATION OF CORAL REEF ECOSYSTEMS 125 Net production of coral reef ecosystems. S. V. Smith 127 Functional aspects of nutrient cycling on coral reefs. Alina Szmant Froel ich 133 Contrasts in benthic ecosystem response to nutrient subsidy: community structure and function at Sand Island, Hawaii. S. J. Dollar 141 Metabolism of interreef sediment communities. John T. Harrison, III 145 IV CHAPTER I INTRODUCTION Ma rj ori e L . Rea ka Department of Zoology The University of Maryland College Park, Maryland 20742 This v in a Wo rk Meet i ng s December Ecol ogy o The go to debate rel ati ons d i rect i on for our d four topi and life of recrui structure of coral or not d i habitats woul d be working i g roup of group and in which empha ses i ty for a For ex Goreau , m coral ree research However , Hyd rol ab D i c k i nson extensive of the de through t the Under Hawa i i ha en v i ronme Island. Fou nda t i o the Ocul i These many of t rapidly g ecosystem these pro olume includes re shop on Coral Ree of the Ameri can S 1983. The worksh f the ASZ. a 1 of the worksho what we currentl hips on coral ree s for future rese iscussion of the cs were chosen fo history patterns tment in coral re coral reef commu reef ecosystems, fferent processes and in different especially fruitf n different labor researchers , i nf 1 by the particula t hey have worked, i n resea rch outl o more compl ete ex ample, following uch of our early fs (e.g., 30-100 group at the Disc NOAA's Undersea R program at the We University in St observations and ep reef commun i ty he use of its sha sea Research Faci s made extensive nts possible in H Supported by NOAA n's submersible, na reefs off the research efforts hem completed wit rowi ng body of k n s in different pa grams, especially search contributions from participants f Ecology that is being held at the ociety of Zoologists in Philadelphia, op is sponsored by the Division of p is to serve as a forum for researchers y do and do not know about functional fs and to identify the most significant arch. To provide a conceptual framework current status of coral reef ecology, r detailed consideration: 1) growth of coral reef organisms, 2) the dynamics ef organisms, 3) the processes that nities, and 4) the structure and function Of particular interest also was whether predominate in shallow v_s. deep reef geographical regions. I felt that it ul to bring together investigators atories and in different oceans. Each uenced by the unique history of their r aspects of the reef environment(s) has its own strengths and unique ok. The workshop presented an opportun- change of ideas among these groups, the dynamic leadership of Dr. Thomas knowledge of the biology of deeper m) came through the efforts of an active overy Bay Marine Laboratory in Jamaica, esearch Facilities, initiated with the st Indies Laboratory of Fairleigh . Croix in 1977, have allowed more even experimentation with components than was previously feasible. Also, 1 1 ow water submersible, the M a k a 1 i ' i , lity operated by the University of observations and sampling in deep reef awaii, Enewetak Atoll, and Johnston 's submersible program, Harbor Branch the Johnson Sea-Link, has investigated east coast of Florida, on the ecology of deep coral hin the last 5 years, complem owledge on shallow reef commu rts of the world. Because ma those on deeper reefs, have reefs , en t a nities ny of been and 1 initiated recently, some of th long term experimental manipul available to the scientific pu date exchange of ideas and inf gators seemed highly desirable and the ensuing volume of T he Reefs include contributions fr each of the institutional site contributions from invest igato a variety of other geographica The papers are organized in addresses the currently import patterns in coral reef environ overview of the factors that c in populations of different ha these causal factors. Using 1 evaluate life history patterns Hughes arrives at the interest and injury-mortality (but not shallower (10 m) than deeper ( Brakel provides an interesting that influence the morphology He shows that light and water but instead place phenotypic c of corals in deep and shallow Providing a conceptual link patterns and the following cha organisms, the second set of p which we probably know the lea factors control recruitment? overview of how offshore curre eddies can influence recruitme fishes. The widespread occurr organisms, particularly in tro consequences for both ecologic One consequence is that kin se marine organisms than in taxa widespread dispersal (e.g., ma water environments). Shapiro analysis of the processes of p mechanisms that may prevent di that prevent dispersion, and s fishes that would be most like kin selection. Also, as point knowledge of the dynamics of r key role in our understanding In order to understand the str know how and when recruits col experiments on settling and me describes bands that identify to an indistinctly banded zone then a period of benthic metam amorphic bands. These marks o tool for determining the preci that is independent of the fir e results (e.g., those involving ations) are not yet generally blic. An opportunity for an up-to- ormation among groups of invest i- at this time. Consequently, this E col ogy of Deep and Shal 1 ow Coral om researchers who have worked at s mentioned above, as well as rs who have studied coral reefs in 1 1 oca t i ons . to four chapters. The first chapter ant topic of growth and life history ments. Sebens provides a general an cause different size distributions bitats and suggests ways of isolating ong term monitoring techniques to of corals over a depth gradient, ing conclusion that both recruitment necessarily growth) are higher in 35 m) sites on a Jamaican reef. biophysical analysis of the factors of corals over a 27 m depth gradient, movement do not directly determine, onstraints, on the possible shapes environments, respectively. between the chapter on life history pter on community structure of reef apers addresses the issue about st in coral reef ecology: what Lobel and Robinson provide an nts and particularly mesoscale nt of the planktonic larvae of reef ence of dispersing larvae in marine pical species, has many important al and evolutionary processes, lection should be less likely in which are characterized by less ny species in terrestrial or fresh addresses this question with an assive dispersion _y_s. the biological spersion. He evaluates mechanisms uggests particular taxa of reef ly to show the characteristics of ed out by Lobel and Robinson, our ecruitment undoubtedly will play a of how communities are organized, ucture of a community, we must onize particular sites. Using tamorphosing fish larvae, Victor the planktonic stage, a transition that corresponds to settling and orphosis, and distinct postmet- n the otolith provide a valuable se time of settling and metamorphosis st visual observations of recruits. One cora 1 specie papers g rowth Suchan i n a 1 occupy wh i 1 e ( « 3 were t f reque i n f orm mod i f y Using Montas of the proc reef communi s , g rowt h i s in this sec and 1 i f e hi ek , et al . , i ttl e studi e open reef h a gorgonacea m) water , in he most sign n 1 1 y ove rg ro ation on the ove rg rowt h field experi t rea ca verno with o rel ati of swe of pas Con ties , wa ter . substr crypti f i s h e s and th zi ng s pol ych of mob effect of the shows t ri but f i shes drama t of the i nd i re of the the ab f ramew Fitz, are kn thel es reefs . ci al r vary i is red predom Tha 1 as shal 1 o on dee tance than o ther species vely persist epers often t interactio sumers also an effect th Using expe ates to vary c fauna at 2 is i nf 1 uenc at fish pred toma topods , aetes , howev ile crypt i c s of fish pr effect s of that , while ion and abun (especial 1 y i c . These e entire cora ctl y determi reef frarnew undance of b o rk ; and the et a 1 . , show lown to dec 1 i s control be Experiment eefs) demons n the differ uced inside inate) compa s i a bi oa ssay w reef si ope p reef slope to herb ivory n reef flats esses ties the t i on story exami d but abi ta n wa s deep i f i c a wn t a phen rel at ments sa de "Tm. ent , obser ns . clear at ap r i men the 0 m, ed by ators the 1 er , t i river edati fish demon dance terr f feet 1 ree ne 1 o ork ; ent h i rec r that ne at nt hi c a 1 ma t rate ent e herb i red t tech s (1- s (30 is g ( whe thought to structure apparently saturated is competition for space. In many sessile primary mechanism of competition, and are equally relevant to the chapter on patterns in coral reef organisms, ne the significance of overgrowth interactions important assemblage of sponges that ts. Interestingly, their data show that, the most important aggressor in shallow er water (> 10 m) four genera of demosponges nt aggressors, while corals were the most xa. Chornesky and Williams provide new omenon of sweeper tentacles, known to ionships among some species of corals. , they show that sweeper tentacles on velop in response to competitive encounters annularis), and that these changes are so that the apparently anomalous distribution ved in this species may reflect a history ly i nf pears ta 1 ma number Wol f , the p in tu a rgest he sec tebrat on . H grazer strati of CO i tori a s have f c omm cal a 1 n i t r o g c mi c r ui tmen , wher g rea t c ommu n i p u 1 a that , x p e r i m vore e o cont n i que 10 m) -40 m) reater re al g 1 uenc to i n n i pul s of et a]_ resen rn in and reti v es ap i x on s upo ons t ral s 1 dam ma jo un i ty gal a en f i of a un t , gr ea s n er de ni ty ti ons wh i 1 ental xcl os rol s . tosh than . Th on s ae oc e t he crea s ation fish . , sh ce of f 1 uen most e hab pears p r o v i n ree hat f have selfi r con , sin bunda xa t i o a ; b i owt h , umber pt hs , struc ( exc e cor cond ures Hay ow t h on ve ey a 1 hallo cupy st r e f r s of pred ow t cry ce t mob i its to des f CO i she been shes sequ ce f nee n by oero and s of the t ure 1 us i al r i t i o (whe and at h ry s sod w an a s p uct u om m a rt ator hat pt i c he d 1 e p of t prot a n e ral s s i n few ) up ence i she a nd blu s i on sur her se h on on c ecru ns, re h Goe erb i hal 1 emon d de at i a re of oderat i f i c i a s and col oni i nver ensi t i rey. he rem ect t h xcel 1 e and a f 1 uenc , the on ree s for s dire divers e- g ree of th v i v a 1 b i v o r o erb i v o deeper ages o i tmen t s u r v i v ea vy t rtm i 1 1 v o ry i ow ree st rate ep ree 1 ref u reef e to 1 re of t za t i tebr es o Exce a i n i em f nt r 1 gae e th effe f al the ctly ity ; n a 1 e re of c us f res (20 ver doe al o urf s er u s h i f fl tha f si ge f com dee ef hem on b ates f co pt f ng t rom ev i e . H e d i ct s gae stru o r ace gae; ef ora 1 i she neve m) a rt i s no f re of sea gher at s t re opes rom mum - P ob i 1 e y » 1 on i - or axa the w e s- of a re ct ure ret i on s r- fi- t c r u i t s al gae on o r s1 s- herbivores). Steneck uses six the Tha 1 ass i a bioassay) to ass several functional groups. He forereef sites (1-2 m) and dec backreef, and deep wall reef h of the Tha 1 ass i a bioassay do n techniques in all cases. Carp mental analysis of the effects algal community on coral reefs composition of the algal commu in primary production. Carpen connection between herbivores, and community metabolism. His the present chapter on communi which addresses the processes coral reef ecosystems. Taking a slightly di coral reef communities, variations in the chara reefs off of eastern Fl Growth rates of the Ocu different techniques (including ess the intensity of herbivory for rb ivory is most intense on shallow reases in the shallow algal ridge, abitats. Interestingly, the results ot agree with those of the other enter also provides a nice experi- of different herbivores upon the Grazing regimes influence species nity, leading to subsequent changes ter's study makes an important turnover rates of consumed species, study thus provides a link between ty structure and the last chapter, of nutrient flow that integrate mollusk and decapod fau reefs (which are episod year) than on shallower are cooler and the cora reefs . Si nee i t addres exogenous nutrients, th and life history patter system processes ) . The last chapter add communities of high bio nutrients in tropical w from several sites in t the Indian Ocean, Smith requirements of reefs d in the surrounding ocea ecosystems do respond t within the reef ecosyst the nutritional require low. The key to the hi system is efficient tra nents of the system. S vs . recycled nutrients coral reefs. On the ba concludes that efficien sediments and feces tha and caves in the reef f community to recycle nu Harrison examines anoth reef ecosystem, a lagoo (3-55 m) in Enewetak. become more efficient w heterotrophic, represen in the atoll ecosystem. ress mass a ter he C con o no n . o nu em d ment gh b ns f e zman to si s t re tar rame trie er i n fl Prod i th ting Ha est are s a r entr cl ud t di He s trie o ex s of i oma r of t Fr expl of h gene e co work nts nter oor uct i dept a s rri s he ques ma i nta e 1 ow. al Paci es that f f er gr uggests nt load h i b i t h the en ss and nutri e oel i ch a i n the er stud ration ncentra and t h with a est i ng commun i on and h. The ink for on also t i on i ned Dra fie, the eat! tha ing, igh tire prod nts exam hig i es of n ted , en r high subs ty, resp sed car des of whi wi ng Wes ove y fr t , w and meta ree ucti amon i nes h pr i n t utri emi euse deg ect i al on i rat i men bon crib how su 1 e sou upon t Aust rail n om tho h i 1 e m while bol i c f ecos v i ty o g diff sourc oduct i he Car ents , tted f d , may ree of on of g a de ion de t comm produc es how ch d i v rces o i nf o rm ral i a , u t r i t i se of any re compo acti v i ystem f the erent es of vi ty o i b b e a n perhap rom po al 1 ow effic the co pth gr c 1 i n e unity ed el s parti erse f at i on and onal pi ankton ef nents ties, a re eco- compo- new f , she s via res the i ency . ral adi ent but i s ewhere cul ar 4 macrob shrimp bi otur organi paper , of nut trophi metabo at 10 of a s were n deeper to nut of t hi reef e The fronts knowl e assemb Spe McFarl Szmant part i c Ronal d ideas , extra p u bl i c for th who pr prepa r of Ma r assist I am h time a were s I also their vol ume vol ume and re gratef reserv Al exei during cation ent hi c s i n E bat i on z a t i o n Dol la r i e n t c p o s i 1 i sm ( and 70 ewage 0 t e d i out fa ri ent s appr cosyst se stu , many dge of 1 ages ci al t and, P Froel ul a'rly Karl s conta assist at ion. e publ o v i d e d a t i o n ci a Co ance i appy t nd eff ubstan would pat i en avail and o search u 1 1 y a ed for , for the f spec newet . Al and r pre s u b s i ti on , rates m de out fa n the 11 . subsi oach ems . dies , of t the we kn hanks hilli ich f wou 1 on , a cted ance NOA i cat i val u of th 11 ie, n al 1 o ack ort t tial 1 like ce an able f the cont cknow my h their i nal ies (a ak ) ca so pro addres sents dy upo diver of nu pt hs o 11 upo shal 1 Commun dy at for de 1 phei n alt v i d i n sing an in n c om si ty , t r i e n ff of n the ow si i ty m the d t e c t i d sh ere g a some tere muni abu t fl Haw str te, etab eep ng r r l mps ommun 1 ink of t sting ty st ndanc uxes ) a i i . uct ur they ol i sm site, es pon in Haw i ty met with t h he i ss u compa r ruct ure e , and on ben Wherea e of th were no , howev i nd i ca ses to ail an abol i s e chap es ra i i s o n o (spec bi oma s t h i c r s d ram e bent t dete er , re ting t pert ur d ca m t h te r sed f th ies s ) _v eef at i c h i c ctab spon he s bat i Ilia roug on c i n S e ef comp s^. c comm eff comm le a ded en si ons nassi d h ommun i ty mi t h ' s f ects o s i t i o n , ommun i ty unities ects uni ty t the st rongl y t i v i ty i n the hem proc ow a a re p Lo ort d 1 i nd S addi duri A's on . able e vo Sta of nowl 0 th y im to d co i n t wor ri bu 1 edg usba eve stag refo new esse S CO due bel , heir ke t teph t i on ng P Of f i I t sug 1 ume ff A the edge ese prov expr oper i me ksho ti on ed. nd , n mo es o re , repr and many s that g ral reef Ronal d Mark Hi primary o expres en Smith al contr r e p a r a t i ce of Un hank Ala g e s t i o n s , and I s si stant final st the rev proceed i ed by th ess my a a t i o n , w for the p depend s of the Lastly , Stephen , re than f prepa r esen of over s . Karl xon , rol s my , wh i but ons ders n Hu and grat , fo ages i ewe ngs; e i r ppre hi ch wo rk s , a se i esp and us ua ati o t a series of advancing them interconnected, in our n the marvelously diverse son , Step es in appr 0 con ors , for t ea Re 1 bert assi eful 1 r her of p rs wh many cri ti c i ati alio shop . fter n vest e c i a 1 my 1 1 sup n of Kenneth hen Smi the wo eci at io tri bute and pro he work search , Sc i en stance y ac kno expert repa rat o contr parts ques an on to t wed u s The s all, up i g a t o r s ly warm 1/2-ye port an this vo Sebe th , a rksho n to d val vi ded shop prov i ce Di t hrou wl edg edit ion o i bute of th d sug he au to ha ucces on t h , and than ar-ol d und 1 ume ns , W nd Al P. I Mark uabl e subs and t ded s recto ghout e the or i al f the d val e vol gesti t hors ve th s of e ide they ks ar d son ersta for p i 1 1 i am i na Hi xon , tant i al he upport r, hel p vol ume uabl e ume on s . for i s thi s as a re e » n d i n g ubl i- CHAPTER II: GROWTH AND LIFE HISTORY PATTERNS OF CORAL REEF ORGANISMS SIZE STRUCTURE AND GROWTH RATES IN POPULATIONS OF COLONIAL AND SOLITARY INVERTEBRATES Kenneth P. Sebens Biological Laboratories and Museum of Comparative Zoology Harvard University, Cambridge, MA, USA 02138 ABSTRACT Benthic invertebrate populations on coral reefs and in other marine communities often occupy several habitat types and display very different size structures in those habitats. The explanation for such size differences may lie in recruitment, mortality, or growth rate differences. These possible causal factors can be separated by carefully monitoring populations of marked individuals or colonies. INTRODUCTION Even a brief examination of a coral reef indicates that such a community is easily divided into obvious zones or habitat types. Certain coral species, other sessile invertebrates, and many mobile ones, have distributions that span a number of distinct habitats. In such cases, the sizes of individual organisms or colonies can have radically different distributions as can the maximum sizes attained. The same is true on temperate rocky shores (Fig. 1B,C,D). However, a single period of samplinq, resulting in a series of size-frequency histograms, is insufficient to explain the inter-habitat size differences and can serve only to illustrate the pattern. Conclusions arrived at by linking particular distribution types to size or aae-dependent mortality patterns (e.q., Grassle and Sanders 1973) can be hypotheses at most, especially when qrowth is indeterminate or colonial. A size-frequency histogram skewed to the right (Fig. 1A) reflects a population dominated by large or old individuals and may result from: 1.) constant high rates of juvenile mortality, 2.) infrequent recruitment followed by good juvenile survivorship and arowth that slows and approaches an asymptotic size or 3.) rapid juvenile nrowth with low mortality, growth cessation at a size asymptote, and high rates of mortality only for large individuals, just the opposite of the first expla- nation. A skew to the left, as in a population dominated by small or young indivi- duals, could result from either constant high mortality at all sizes or from size- selective predation on large individuals (e.g., Grassle and Sanders 1973). Such a distribution could also be found shortly after an infrequent high recruitment event, and be followed by high juvenile mortality and an eventual switch to a distribution skewed toward large individuals. When clonal or colonial organisms are considered, another explanation arises; periods of colony or individual fragmentation (Highsmith 1980, Hughes and Jackson 1980, Sebens 1983) could also produce such a pattern. If individuals have very indeterminate growth, reaching different size maxima under various microhabitat conditions, a left-skewed distribution might result if a population spans microhabitats, within one of the distinct zones or habitats, that are mostly of poor quality (producing small individuals) with a few patches of better quality (thus larger individuals). This pattern would result even if mortality and recruitment rates were equal in all microhabitats and if mortality was size-independent. Distributions approaching a Gaussian curve could also result from any of the above processes. Finally, size-frequency histograms sometimes exhibit strona mul timodol ity. A bimodal distribution (Fig. 1A) is often taken as evidence of distinct age classes. Yet, the same distribution would be produced by indeter- minate growth and a habitat comprising two microhabitat types 'poor' and 'good'. It should thus be clear that, the shapes of size-freouency histograms do little more than illustrate existing patterns. Causal interpretations are impossible from these alone, especially when size and age are at least partially uncoupled. Once a particular inter-habitat size pattern is determined, it is looical to attempt to find an explanation for the pattern. Assume that a sampling study shows that mean sizes of individuals (or colonies) change across habitats of increasing wave exposure (e.g., Birkeland 1973, Ebert 1982, Sebens and Paine 1978, Sebens ms). The following hypotheses (and probably others) could be proposed: 1) mortality rates are higher in habitats with smaller individuals; mortality may or may not be size-specific (greater for larger individuals). 2) growth rates and thus maximum or asymptotic sizes are greater in habitats with the larger mean individual sizes. 3) recruitment rates of juveniles are greater in habitats with smaller individuals, thus skewing the distribution and reducing the mean size. 4) fission or colony fragmentation rates are higher where the individuals or colo- nies are generally smaller. Biomass growth could still be as high or higher than in other habitats, but must be considered as 'clonal' growth rate. A variety of experimental and monitoring approaches must be undertaken to test such hypotheses. Furthermore, the above hypotheses are not mutually exclusive nor are they comprehensive. All of the above causes could influence the size pattern in the same direction. The following discussion will explore the second hypothesis. Figure 1 A. Shapes of hypothetical size-frequency histograms. B. Diameters of the octocoral Alcyonium in three habitats (Sebens Ms.) C. Mean sizes of Alcyonium versus wave exposure index (elevation of intertidal Mean sizes of Alcyonium versus wave exposure index barnacle zone) (Sebens MS) , DIA = average colony diameter D. Sizes of 10 laroest sea anemones (Phymactis clematis) as exposure index (Sebens and Paine 1978) , DIA = basal a function of wave diameter, mm. 20 10 80 60 DIA 40 cm 20 A SKEWED M LEFT GAUSSIAN SfZE r> PROTECTED 10 J^ SKEWED RIGHT J n MODERATELY EXPOSED £ 10 20 40 60 80 20 DIA cm C 10 n BIMODAL EXPOSED n^lr^rL, 40 60 80 20 40 60 100 M 10 1.5 2.0 2.5 12 3 4 5 6 T EXPOSURE INDEX 10 Fi aure 2 A. where c R, C. Energy intake (I) < C£ (see text). As^in A, but where Ci and cost (0) as functions of individual biomass (W) W *. = optimum size, E$ = 1-0, as in colony qrowth. °R, Enemy surplus (Es) as a function of W where Ci : C2> Habitats increase in food availability or decrease in physiological cost. Hj, H2, H- D. As in C but where c^ = C2, M = maximum size (non-asymptotic), set by non- energetic causal factor such as size-dependent mortality. E E, A B w w Growth metabolic Extrinsic potential getic cos example, dators in behavior) the diffe to reprod GROWTH RATES AMD SIZE ASYMPTOTES rates depend on intrinsic factors such as the all activity and to reproduction, or genetically dete factors that influence growth include local food energy intake) and the physical nature of the hab t). Intertidal habitats with high temperature or generally increase metabolism integrated over time a particular habitat can affect their prey's ener and intake (time available for foraging). Growth Ci c rence between energy intake (kiW ) and cost U2W uction (g(W)) (Sebens 1979, 1982) ocation of energy to rmined limits to size, availability (determining itat (determining ener- desiccation stress, for The frequency of pre- gy loss (e.g. , escape rate (dW/dt) depends on ) minus the allocation dW/dt kjW ■1 k2W g(W) assuming a simple power function relationship between intake or cost and body mass (W) (Fig. 2). Under most conditions, reproductive output (g(W)) (and thus offspring number = fitness) will be maximized when W is chosen such that when dW/dt = 0 (growth ceases), k^W * - k2W ^ is at its greatest. In other words, this hypo- thetical organism should stop growing at some asymptotic size WQpt where Wopt = (C2k2/c1k1)1/(cl"c2). The parameter Co equals 0.7-1.0 and cj £0.7 for a variety of organisms, thus producing the predicted size asymptote because energy intake scales at a lower power of mass than does cost. Colonial organisms need not conform to 11 this pattern. If a planar colony doubles in area and mass, both its energy intake and metabolic costs could double as well. If Cj >_ c2 there is no size asymptote predicted on energetic arounds (Fig. 2B). A similar argument can, however, be used to predict the size of units within a colony (polyps, zooids, etc.) which often show guite determinate growth and a size asymptote (Sebens 1979). On energetic arounds, individuals or colonies (where c^ < C2) might thus grow to an asymptotic size that maximizes their reproduction under a given set of habitat conditions. This asymp- tote will be higher in habitats that are 'better' for either prey availability or physical conditions affecting metabolic rate (Fig. 2C). Mortality rates may be high enough, however, in some habitats that the energetically predicted asymptote is never reached. Mechanisms other than energetics will lead to a habitat-dependent maximum size that is non-asymptotic (growth does not slow and cease). Size-dependent fission (e.g. Sebens 1982) could have such an effect. Size-dependent mortality will also produce a maximum size that can differ between habitats (Fig. 2D). Birkeland (1973) found that seafans in Panama have smaller maximum sizes in habitats with higher wave energy because storm waves tear off large colonies. Octocoral colonies (Alcyonium siderium) at wave exposed sites in New England are larger than at calmer sites, but there is no obvious growth cessation in large colonies at the most exposed sites; storm waves may also set the maximum size in this case. Size-selective predation would produce a similar pattern. In fact, Paine (1976) showed that a bimodal distribution of mussel sizes can arise because of constant seastar predation on most size classes; an escape in size occurs for the few mussels that, by chance, grow large enough that they cannot be consumed by seastars. Figure 3 A. Growth rate differences in two habitats, Hj and H2, with equal asymptotic sizes (SA). SI = initial size, S2 = size after one time interval. B. Similar growth rates in two habitats, Hj and H2, with unequal asymptotic sizes (SA1, SA2). C. Ford-Walford Plot of A. D. Ford-Walford Plot of B. E,F. Comparison of mean size increments by size class in two habitats (see text). Bars along abscissa indicate size classes 1-6. 12 Figure 4 A. Growth models for individuals: 1. parabolic asymptotic (Von Bertalanffy), 2. sigmoid asymptotic (Gompertz), 3. linear, non-asymptotic, T = Time. B. Ford-Walford Plot. Numbers refer to models as in A. C. Size-specific orowth vs. size (see text). Numbers refers to models as in A. P. Ford-Walford Plot illustrating determinate growth (D) and relatively indeterminate growth (I), as envelopes around clouds of data points. COMPARING GROWTH RATES Testino the hypothesis that growth rates differ across habitats involves some means of tagging, mapping, or otherwise identifying individuals covering the full range of size classes. It is perfectly possible to have similar asymptotic or maximum sizes in two habitats yet to have significantly different juvenile growth rates, for example. Alternatively, statistically indistinguishable growth rates at all sizes could still lead to asymptotic sizes that differ significantly (Fig. 3 A- D). One of the most widely used means of comparing growth increments is the Ford- Walford plot (Ricker 1975; Ebert 1980, 1982; Sebens 1983) which graphs initial size on the abcissa and size after some time interval on the ordinate (Figs. 3C,D;4B,D). This presentation does not show any single individual's growth through all size classes, but instead presents a 'snapshot' of growth in the entire population over a single interval. Variations of the Ford-Walford plot can also be used to fit growth models (von Bertalanffy, Gompertz, Richards functions; Ricker 1975, Yamaguchi 1975, Ebert 1980) (Fig. 4 A, B). A regression line through the points on a Ford-Walford plot (either or both axes can be transformed to produce a linear plot) can be tested for significant differences in slope or intercept, thus testing the hypothesis that differences in growth rates exist between habitats. If a growth -model is to be fit to the data, Ebert (1980, 1982) suggests that the Richards function be used because it has the ability to incorporate deviations from linearity which are especially common in the small size classes. . . Growth increments can also be plotted as the specific growth rate (aS*a t^'S-1) on the ordinate vs. size (S) at the middle of the time interval At on the abscissa 13 (reviewed in Kaufmann 1981) (Fin. 4 C). This method has the advantaae of allowina growth to he measured over different time periods (At). Such plots can also be subjected to reoressions and these lines can then be used to fit various growth models. This technique has the disadvantaoe that size (S) is incorporated into both parameters of the regression, making statistical interpretation difficult by violat- ing the assumption of independence (Ricker 1975). The Ford-Walford plot avoids this problem, and can incorporate a small degree of variability in At values by assuming short-term linear orowth and normalizino all data to a constant time interval. The Ford-Walford plot allows direct examination of growth and 'degrowth', growth rate as a function of size, and size asymptotes or maxima. The degree of 'indeter- minateness' is also evident as the spread around the regression line (Fig. 4D), as is the fraction of individuals below the zero growth line (showing shrinkage or degrowth). Coral growth, measured as density bands in the skeleton, could be compared in this manner, as could whole colony size increase. Growth can also be compared statistically without resorting to either regression lines (assuming linearity) or any particular growth model. If the data are broken up into size classes (Fig. 3E, F) such that there is a reasonably large number of points in each size class, then the mean growth increments in each size class can be compared across habitats (e.g., by analysis of variance, Sebens 1983). This method allows comparison across several habitats (via a multiple comparisons test) and determines where in the size distribution the significant differences actually occur. This is the only method that would pick up differences between habitats which have equal maximum individual sizes and differ only in early growth rate. Growth by fission and fragmentation cannot be incorporated directly into a Ford- Walford plot. A good descriptor of such growth is the rate of biomass change (e.g., exponential function) within a clone, colony, or group of colony fragments. However, some clones or colonies may have reached limits imposed by available space. There- fore, choosing experimental subjects not apparently space limited, or removing such limits experimentally (e.g., clearinq space around the subject, transplanting) may be necessary to compare potential growth rates (Sebens 1983). SUGGESTIONS AND LIMITATIONS There is no shortcut method that allows an investigator to read causation from static samples of size distributions across populations. Mortality schedules can be arrived at only by following individuals or cohorts over time. It is theoretically possible to calculate an average overall mortality from a size distribution (e.g. Ebert 1982), but this approach assumes constant, mortality and recruitment rates. It is the latter assumption that causes real problems because many sessile inverte- brates have successful recruitment only rarely and sometimes many years apart. In colonial organisms, partial mortality (e.g. corals, Hughes and Jackson 1980) pre- sents another problem for interhabitat comparisons. This process is best treated as 'degrowth' or 'shrinkage' and can be compared across habitats directly. Similarly, binary fission does not constitute 'mortality' of the original individual, even though that individual no longer exists as such. Fission rates can also be compared directly either numerically or by using biomass (Sebens 1983). Growth rate differences, resulting from habitat or microhabitat variability, can produce some of the observed size gradients within populations. There are several methods for comparing growth and for fitting growth models; however, it is not necessary to choose a growth model if the goal of a study is only to determine whether or not growth rates differ across habitats. Ford-Walford plots of growth increments can be used for this directly, and can also illustrate growth rate variability and the extent of degrowth or shrinkage in the population. An even simpler qrowth comparison between habitats would be to compare only maximum size (mean size of largest N individuals) and nrowth at some earlier staqe, for example at 1/2 the maximum size in the habitat with the smallest maximum size. This would 14 necessitate fewer data points but would answer the basic question of whether or not growth rate differences could be causinn the observed pattern. However, even if growth rate differences can be established, either mortality, recruitment, or fragmentation could still be major factors affectino size distributions across habitats, and must be investigated. LITERATURE CITED Birkeland, C. 1973. The effect of wave action on the population dynamics of Gorgonia ventalina Linnaeus. Stud. Trop. Oceanogr.: 115-126. Ebert, T.A. 1980. Estimating parameters in a flexible growth eouation, the Richards function. Can. J. Fish. Aquat. Sci . 37: 687-692. Ebert, T.A. 1982. Longevity, life history, and relative body size in sea urchins. Ecol. Monoqr. 52: 353-394. Grassle, J.F., and H.W. Sanders. 1973. Life histories and the role of disturbance. Deep Sea Res. 20: 643-659. Highsmith, R. 1980. Passive colonization and asexual colony multiplication in the massive coral Porites lutea Milne Edwards and Haime. J. Exp. Mar. Biol. Ecol. 47: 55-67: Hughes, T., and J. B.C. Jackson. 1980. Do corals lie about their age? Some demo- graphic consequences of partial mortality, fission, and growth. Science 209: 713-715. Kaufmann, K.W. 1981. Fitting and using growth curves. Oecologia 49: 293-299. Paine, R.T. 1976. Size limited predation: an observational and experimental approach with the Pisaster Mytilus interaction. Ecology 57: 858-873. Ricker, W.E. 1975. Computation and interpretation of biological statistics of fish populations. Bull. Fish. Res. Bd. Can. 1°1: 382 pp. Sebens, K.P. 1979. The enerqetics of asexual reproduction and colony formation in benthic marine invertebrates. Amer. Zool . 19: 683-697. Sebens, K.P. 1982. The limits to indeterminate nrowth: An optimal size model applied to passive suspension feeders. Ecology 63: 209-222. Sebens, K.P. 1983. Population dynamics and habitat suitability of the intertidal sea anemones Anthopleura elegantissima and A. xanthogrammica. Ecol. Monogr. (in Press) . ' Sebens, K.P. MS. Water flow determines colony size in a temperate octocoral. (ms. submitted). Sebens, K.P., and R.T. Paine. 1978. Piogeography of anthozoans alone the west coast of South America: habitat, disturbance and prey availability, pp. 219-238. In : Proc. Int. Symp. on Marine Biogeography and Evolution in the Southern Hemisphere, Auckland, N.Z., N.Z. D.S.i.R. Information Series Mo. 137. Yamaguchi, M. 1975. Estimating growth parameters from growth rate data. Problems with marine sedentary invertebrates. Oecologia 20: 321-332. 15 LIFE HISTORIES AND GROWTH OF CORALS OVER A DEPTH GRADIENT Terence Hughes Dept. of Earth & Planetary Sciences, The Johns Hopkins University, Baltimore, Maryland 21218, and Discovery Bay Marine Laboratory, Discovery Bay P.O. Box 35, St. Ann, Jamaica. ABSTRACT Results from long term monitoring of coral populations from -10m to -35m show that several important life history parameters are strongly correlated with depth. Typically, coral colonies tend to settle in greater numbers in shallow water, where they are also more likely to be injured or killed compared to deeper sites. Colony extension rates were found to be very weakly dependent on depth, with some of the fastest growing colonies at -55m and -35m growing faster than many -10m specimens. The net growth rate of shallow corals may be limited by higher rates of injuries. INTRODUCTION One of the best studied features of coral reefs is the variation in species composition and area associated with depth gradients (see review in Stoddart 1969). Less well understood is how such changes in relative abundance are brought about. The number of colonies in a coral population is determined by the balance between sexual (and, in some species, asexual) recruitment and colony mortality, while colony size is a function of time since settlement and rates of growth and injury. In the present study, these aspects of the population dynamics of five species of foliaceous corals (Agaricia agari cites, A. lamarcki , Leptoseris cucullata, Po rites astreoides , and Montastrea annularis) were studied on the north coast of Jamaica over a wide depth range (-10m to -55m). I present here some of the major findings; a fuller account, including species comparisons, will be presented elsewhere (Hughes and Jackson, in review). METHODS 2 Twelve lm quadrats were tied permanently to the reefs off the Discovery Bay Marine Laboratory between -10m and -20m ("shallow") and at -35m ("deep"). The quadrats, which enclosed more than 1,000 foliaceous coral colonies were photo- graphed each year from 1977 to 1980. Analysis of the photographs yielded popul- ation structures, as well as rates of mortality and sexual recruitment. Coral growth was measured in situ using fixed reference points (nails) and a plastic tape measure, and at -55m by collection of Alizarin stained colonies. All colonies in the quadrats were measured regardless of colony condition, size or ease of access. RESULTS Many important population parameters were strongly dependent on depth. The size-frequency distribution of colonies of all species were dominated by small corals over their entire depth range, but especially in shallow water. Large colonies greater than 200cm in area constituted only 8.9% of the total colony counts from -10m to -20m, compared to 17.4% at -35m. Depending on species, the mortality rate of whole colonies for all size classes 17 combined was 2-6 times higher in shallow (-10m to -20m) compared to deep water (-35m) (Table 1). This reflects, in part, depth-related changes in the proportion of small colonies in the populations, since large corals are much less likely to be killed than small (e.g. Connell 1973, Loya 1976, Hughes and Jackson 1980). How- ever, within-size class mortality rates were also invariably higher in shallow water, particularly for the largest corals. Not a single colony out of 32 greater than 200cm2 was killed at -35m in three years, compared to the loss of 8 out of 39 colonies of the same size at -10m to -20m. Further more, total tissue losses (from both injuries and whole colony mortal- ity) were greater in shallow water (Table 1). The relative importance of injuries and whole mortality also varied with depth. At -35m fully 80% of the coral tissue destroyed was due to injuries, particularly to large colonies, while in shallow populations losses from injuries and whole mortality were almost equal. The higher rates of colony and tissue losses in shallow water were compensated somewhat by enhanced amounts of larval recruitment, which was more than twice the levels of the deep (-35m) quadrats (Table 1). Table 1. Various aspects of coral population dynamics as a function of depth for the three species common at all depths. (A) Percentage mortality of coral colonies monitored photographically from 1977 to 1980. The total number of 1977 colonies was 425 shallow (-10m to -20m), and 301 deep (-35m). (B) Total percentage area of coral tissue destroyed by injuries and whole colony mortality. (C) Number of new recruits observed settling within the permanent quadrats (equal areas shallow and deep) from 1977 to 1980. A. aqari cites A. lamarcki L. cucullata (A) % of colonies Killed Shallow Deep 41% 21% 39% 6% 63% 34% (B) Amount of tissue killed (% of '77 area) Shallow Deep 47% 33% 57% 9% 103% 56% (C) Larval Settlement (no./12m2/3yrs) Shallow Deep 60 25 1 3 39 16 Coral Growth Wi thin-species growth reoides and M. annularis rates of foliaceous A. agaricites, A. lamarcki were only weakly dependent on depth, and many , P. ast- individual colonies of corals at -35m and -55m exceeded the extension rates of specimens in shallow water (Fig. 1), Deeper foliaceous corals were less likely to have their growth interrupted by partial mortality, and also had thinner skeletons (Hughes 1982), i.e., they could increase their surface area with a smaller addition of cal- cium carbonate, so that they often grew surprisingly fast relative to shallow foliaceous corals. DISCUSSION Several previous studies have shown only a very weak relationship between lateral coral growth and depth because of considerable within-depth variation between colonies (see Highsmith 1979 for review). This variability is often artificially eliminated by collecting biased samples of undamaged colonies, which probably grow faster than colonies that have recently suffered injuries. Coral injuries were found in the present study to be common events; of 883 colonies 18 2 2 HI I- < DC < < N=187 N=146 A. agaricites N=50 P. astreoides 10m 20m 35m A. lamarcki CC N=27 N = 50 N = 43 ° 15i 20m 35m 55m DEPTH N = 71 N=22 10m 20m M. annularis 15 N=56 N = 63 N=63 20m 35m DEPTH 55m Figure 1. Frequency distribution of annual radial qrowth rates versus depth for A. agaricites, A. lamarcki, P. astreoides and M. annularis. Measurements at each depth were categorized into 5mm growth classes and the frequency distrib- ution (shaded areas) standardized by dividing by the sample size. The sample sizes (N) indicate the number of colonies measured in situ or after staining and collection. The total number of colonies measured or stained over the three years was 778. Arrows signify means. surviving from 1977 to 1980, over 75% were injured at least once. Although deeper colonies tend to be larger, they are not necessarily much older than smaller, shallow colonies, whose history of growth and injury will often be more complex. Whether similarly sized corals of different past histories will have comparable expectations of survival and fecundity remains an interesting, but untested, question. LITERATURE CITED Connell, J.H. 1973. Population ecology of reef building corals, p. 205-245. In: O.A. Jones and R. Endean (eds.), Biology and Geology of Coral Reefs, Vol. 2. Academic Press, New York. Highsmith, R.C. 1979. Coral growth rates and environmental control of density banding. J. Exp. Mar. Biol. Ecol . 37: 105-125. 19 Hughes, T.P. 1982. On the relationship between coral growth and calcification rates. Geo!. Soc. Amer. 14(7): 519-520 (Abstract). Hughes, T.P., and J. B.C. Jackson. 1980. Do corals lie about their age ? Some demographic consequences of partial mortality, fission and fusion. Science 209: 713-715. Hughes, T.P., and J. B.C. Jackson. (In review). Population dynamics and life histories of foliaceous corals. Submitted to Ecol . Monogr. Loya, Y. 1974. Settlement, mortality and recruitment of a Red Sea scleractinian population, p. 89-100. jji: G.0. Mackie (ed.), Coelenterate Ecology and Behaviour. Plenum Press, New York. Stoddart, D.R. 1969. Ecology and morphology of recent coral reefs. Biol. Rev. 44: 433-498. 20 DEPTH-RELATED CHANGES IN THE COLONY FORM OF THE REEF CORAL PORITES ASTREOIDES Will em H. Brakel Department of Biology Loyola College Baltimore, MD 21210 ABSTRACT The depth-related morphological transition of Porites astreoides colonies from near-hemispherical to flattened forms is well known, but requires closer exam- ination. For this reason, colonies of _P. astreoides were surveyed at 7 sites spanning a depth range of 27 m at Discovery Bay, Jamaica. The depth, substrate slope, and degree of exposure of each colony were assessed prior to measurements of colony shape and corallite structure. Regression analysis of quantum irradiance measurements taken throughout the study area provided equations to estimate the light available to each colony as a function of its depth, substrate slope, and exposure. Predicted wave heights and periods from wind statistics were used in conjunction with theoretical equations for the attenuation of wave motion with depth to estimate the relative water movement about each colony. It was found that availability of light does not directly determine the shape of the colony, but only sets an upper limit to colony height, a limit that decreases with decreasing illumination. Similarly, water movement places only an upper constraint on morphological variation, with only very flat colonies possible in high energy environments. Consequently, flattened colonies were found at all depths, but were exclusively present in very shallow water (due to high turbulence) and in very deep water (due to low light). These morphological constraints were shown to be primarily the result of a phenotypic response of the colony to its environment. INTRODUCTION Coral reefs typically exhibit sharp gradients in environmental factors such as illumination and water movement over relatively short vertical or horizontal distances. Physical gradients in relation to depth and their effects on reef organisms are of particular interest at this workshop on deep and shallow reefs. One approach to investigating the impact of depth and depth-related variables on reefs is to examine those organisms that have somehow managed to overcome the physiological stresses imposed by the depth gradient and that have, as a con- sequence, achieved a relatively wide depth range. By studying the adaptations of these eurytopic reef organisms we can better understand the effects of depth on the entire biota. Among the reef-building scleractinian corals of the Caribbean region, several species that are primarily massive in their colonial growth form exhibit broad depth ranges for light-dependent organisms. They include Montastrea annularis and Porites astreoides, both of which are common from shallow lagoons to depths of 27 m or more (Roos, 1964; Goreau & Goreau, 1973). These species, along with others such as Meandrina meandrites, Stephanocoenia michel inii , Colpophyllia natans, C_. breviserialis, Dichocoenia stokesi , and Montastrea cavernosa, show a morpho- logical transition from roughly hemispherical colonies in shallow water to flattened plates at greater depth. The consensus has been that this change in colony shape is a phenotypic response to maximize light interception by the polyps (Goreau, 1959, 1963; Roos, 1964; Macintyre & Smith, 1974). This view was rein- forced by Roos (1967), who was able to relate morphological change in Porites 21 astreoides colonies on Curacao reefs to measured changes in the distribution of underwater radiance, and who observed the expected morphological changes in transplanted Montastrea annularis in computer simulations of coral growth based on empirical measurements of light distribution and on simple assumptions about the effects of liqht on calcification. As part of a larger study of the ecology and variation of the genus Porites (Brakel, 1976), I collected detailed information on the colony shape and distri- bution of Porites astreoides on Jamaican reefs. These data were supplemented by relatively simple, but objective and repeatable, estimates of two important depth- related variables, water movement and irradiance. The results presented below reveal that the generally recognized depth-related morphological trend is more complex than expected. METHODS Study area Observations and collections were made on the north coast of Jamaica at Discovery Bay. The structure and zonation of the reefs of this region have been described by Goreau (1959) and Goreau & Goreau (1973). I selected 7 different reef habitats as study sites, encompassing a depth range of 27 m (Fig. 1). Four sites were located on the exposed seaward side of the reef crest; 3 additional were situated in the sheltered, more turbid waters leeward of the fringing At each study site the coral fauna was surveyed in quadrats along transect sites reef, lines perpendicular to the slope of the following features were recorded: (1) colony was growing; (2) the exposure of from 0 (fully exposed, open to water on crevice); and (3) the shape of the colony, measurements described below. Irradiance estimates for each colony Light measurements were made at all reef. For every Porites encountered, the the slope of the substrate on which the the colony, measured on an arbitrary scale all sides) to 3 (in a deep hollow or based on a series of morphological Fig. 1. Study sites at Discovery Bay, Jamaica. Depth contours are in meters sites using a diver-operated Lambda Instru- ments L192S underwater quantum sensor coupled to an LI 85 quantum/radiometer/ photometer mounted in an Ikelite 5910 underwater housing. In addition to down- ward irradiance taken with the cosine sensor pointed straight up, measurements were also taken at various angles from the vertical in the four cardinal directions, providing data on the 3-dimensional angular irradiance distribution. Results relevant to this paper are summarized below; complete details of the methods and statistical analyses are provided else- where (Brakel, 1979). Regression equations were obtained to describe the attenuation of light with depth. Two equations were necessary, since the water column in the bay (sites 1-3) had a light extinction coefficient significantly different from that of the fore reef (sites 4-7): In I = 4.31 - 0.11 d (bay) In I = 4.20 - 0.06 d (fore reef), 22 derived from theoretical winds blow from the E to NE where d is the depth in meters. The equations predict the midday irradiance at various depths on clear, sunny days as a percentage of the surface irradiance. These expressions for I were modified to take into account the slope, s, of the substrate on which the coral was growing, using a relationship also developed from regression analysis: F = exp[(4.61 - 0.02 s)/100] The light score, L, for each coral was further modified by the exposure, E, of the coral (a measure of the amount of shading by overarching branches of neighboring corals and other reef features described above): L ■ I F [1 - (E/6)]. Estimates of water movement A relative water movement score for each coral was calculations. In the belt of the NE trade winds, the 88% of the time with a maximum velocity of 35 km/hr (U.S. Weather Bureau, 1962). According to the wave forecasting curves of Bretschneider (1966), winds of this direction and magnitude for 8 hours would generate waves with a period, T, of 5.9 sec, and a significant wave height, H, of 159 cm on the north coast of Jamaica. Empirical observations of wave height (Reiswig, 1 971 :114; U.S. Weather Bureau, 1959) fall in the same range as the Dredictions. Under the maximum trade wind conditions specified, the fore reef (sites 4-7) would be under the influence of the surface conditions given above. On the leeward side of the reef crest (sites 1-3), the maximum fetch is 1 km, limiting wave conditions to H = 61 cm, T = 2.5 sec according to the wave prediction curves. The effect of surface waves diminishes with increasing depth, d, below the surface, in such a manner that the major axis, A, of the ellipse traced out by a particle as the wave passes over is approximated by the following equation, modified from Eagleson and Dean (1966): A = H exp[(-47T2d)/(gT2)], where g is the universal gravitational constant. Simplifying all the constants, this reduces to: A = H exp[(-4.062 d)/T2]. The wave movement score for each colony was then defined as: W = A/T, to give an arbitrary, but reproducible measure of the average expected wave- induced water movement. Measurement of colony shape Although an elaborate multivariate analysis of various aspects of colony morphology can be done (see Brakel , 1976), for the purposes of this discussion a simple measure of the degree of flattening of the colony will suffice. I used the relative colony height, defined as the average diameter of the colony in the plane of the substrate divided by its height measured perpendicular to the substrate. For a perfectly hemispherical colony the relative colony height would be 0.5; for a very flat colony it could 1 10 20 30 be as little as 0.05. To avoid the DEPTH (m) allometric complications inherent in such a ratio of morphological measurements, it was found necessary to exclude all very small colonies (less than 9 cm in diameter) i eg LU I > Z o _i o o 0.8 0.6 0.4 0.2 0.0 Fig. 2. height as Graph of relative colony a function of depth. 23 from the analysis. Those few colonies with diameters greater than 30 cm were also omitted. The above procedures yielded data on 27 P_. astreoides colonies. For each colony the relative colony height was measured and the irradiance and water move- ment in the immediate vicinity of the colony (based on its depth, location, expos- ure, and substrate slope) were estimated. RESULTS The mean relative colony height of the corals sampled was 0.24 with a range of 0.08 to 0.69. A graph of colony height as a function of depth is shown in Fig. 2. Depth is plotted on a logarithmic scale to emphasize differences among the shallower sites where physical gradients are most pronounced. Surprisingly, the expected monotonic relationship between depth and colony morphology is not seen. When irradiance, L, is plotted against colony height (Fig. 3), a linear or curvilinear relationship is still not apparent. Instead, it appears that irrad- iance acts only to set an upper limit to colony height: under high illumination all colony heights are possible, but as available light diminishes only increasingly flattened colonies are seen to occur. The line drawn on the graph marks the limit to colony height imposed by light regime. The graph of colony height as a function of water movement, W, (Fig. 4) shows that this parameter also exercises a constraining effect on morphological variation without directly determining the height of the colony. At low wave energies the full range of coral 1 urn variability is seen; in more turbulent micro-environments only the more flattened forms are found. DISCUSSION The data on Porites astreoides illustrate that the response of organisms such as corals to changes in depth can be complex. As depth increases, different bio- logically-relevant physical and chemical variables such as illumination and water 2 3 IRRADIANCE Fig. 3. Plot of relative colony height as a function of irradiance. x (3 LU I > o _l o o 4 8 12 WATER MOVEMENT Fig. 4. Graph of relative colony height against water movement. 16 24 movement, as well as salinity, temperature, and sedimentation rates, may vary in disparate ways. When all these factors act at once, individually or synergistic- ally, on the physiology and morphology of the benthos, the outcome is not easily foreseen. Even when considered in isolation, an ecological variable such as irradiance or water movement does not have the expected biological effects. One might have anti- cipated some linear relationship between irradiance or wave energy and relative colony height, a relationship amenable to correlation or regression analysis. In actuality, the roles of light and water movement are quite different: they do not directly determine the shape of the colony, but rather set limits to the possible range of morphological variation. A detailed discussion of the physiology of coral colony growth is beyond the scope of this paper, but one question that should be addressed is whether the observed environmental constraints on Porites colony morphology are environmentally- induced or the result of natural selection at different depths for genotypes with the requisite inherent shape characteristics. Examination of fine skeletal struc- tures of the individual coral lites known to be unaffected by environmental condi- tions (Brakel, 1977) showed that those colonies most flattened or constrained with respect to colony height are not genetically distinct, but constitute a random subsample of the population. This implies that their modified colony shape is the result of a direct physiological response to environmental stimuli. The observed morphological transition of £_. astreoides in relation to irradiance and water move- ment is therefore principally due to phenotypic plasticity, not selection. The mechanism by which this morphological transition occurs is not under- stood. Goreau (1963) and Barnes (1973) suggested that the form of the corallum is the result of two separate processes; skeletal accretion (dependent on light) and tissue growth (independent of light), so that at low light intensities cal- cification cannot keep up with tissue growth, resulting in the lateral prolifera- tion of excess tissue to form a flattened colony. The adaptive significance of a flat colony profile in environments with high wave action has been discussed by Graus, et aj. (1977), but again the mechanism by which water movement controls colony height is not clear. Jokiel (1978) has suggested that water movement influences corals by controlling the exchange of materials between the polyps and the surrounding sea water. Whatever the nature of the physiological controls on coral growth and form, it is evident that, with changes in depth, environmental factors act on the colony in intricate ways and that the adaptive response of the coral can be subtle and complex. ACKNOWLEDGMENTS This research was made possible with the support of the Biology Department, Yale University, and with the assistance and facilities provided by the Discovery Bay Marine Laboratory. LITERATURE CITED Barnes, D. J. 1973. Growth in colonial scleractinians. Bull. Mar. Sci . 23: 280-298. Brakel, W. H. 1976. The ecology of coral shape: microhabitat variation in the colony form and corallite structure of Porites on a Jamaican reef. Ph.D. Diss., Yale Univ. 246 p. Brakel, W. H. 1977. Corallite variation in Porites and the species problem in corals. Proc. Third Int. Coral Reef Symp. 1: 457-462. Brakel, W. H. 1979. Small-scale spatial variation in light available to coral reef benthos. Bull. Mar. Sci. 29: 406-413. 25 Bretschneider, C. L. 1966. Wave generation by wind, deep and shallow water. In: A. T. Ippen (ed.), Estuary and coastline hydrodynamics, p. 133-196. McGraw-Hill. New York, N.Y. Eagleson, P. S., & R. G. Dean. 1966. Small amplitude wave theory. In: A. T. Ippen (ed.), Estuary and coastline hydrodynamics, p. 1-92. McGraw-Hill, New York, N.Y. Goreau, T. F. 1959. The ecology of Jamaican coral reefs I. Species composition and zonation. Ecology 40: 67-90. Goreau, T. F. 1963. Calcium carbonate deposition by coralline algae and corals in relation to their roles as reef builders. Ann. N. Y. Acad. Sci. 109: 127-167. Goreau, T. F.,and N. I. Goreau. 1973. The ecology of Jamaican coral reefs II. Geomorphology, zonation, and sedimentary phases. Bull. Mar. Sci. 23: 399-464. Graus, R. R., J. A. Chamberlain, Jr., & A. M. Boker. 1977. Structural modifi- cation of corals in relation to waves and currents. Stud. Geol . 4: 135-153. Graus, R. R.,and I. G. Macintyre. 1976. Light control of growth form in colonial reef corals: computer simulation. Science 193: 895-897. Jokiel, P. L. 1978. Effects of water motion on reef corals. J. Exp. Mar. Biol. Ecol. 35: 87-97. Macintvre, I. G., & S. V. Smith. 1974. X-radioqraphic studies of skeletal deve- lopment in coral colonies. Proc. Second Int, Coral Reef Symp. 2:_ 277-287. Reiswig, H. M. 1971. The physiological ecology of Porifera: a comparative study of three species of tropical marine Demospongiae. Ph.D. Diss., Yale Univ. Roos, P. J. 1964. The distribution of reef corals in Curacao. Stud. Fauna Curacao 20: 1-108. Roos, P. J. 1967. Growth and occurrence of the reef coral Porites astreoides Lamarck in relation to submarine radiance distribution. Ph.D. Diss., Univ. of Amsterdam. 72 p. U.S. Weather Bureau. 1959. CI imatological and oceanographic atlas for mariners. I. North Atlantic Ocean. 182 charts + 7 p. Washington, D.C. U.S. Weather Bureau. 1962. Summary of hourly observations; San Juan, P.R., 1951- 1960. Decennial census of United States climate. Climatography of the U.S. No. 82-52. Washington, D.C. 26 CHAPTER III: THE DYNAMICS OF RECRUITMENT IN CORAL REEF ORGANISMS 27 REEF FISHES AT SEA: OCEAN CURRENTS AND THE ADVECTION OF LARVAE Phillip S. Lobel and Allan R. Robinson Center for Earth and Planetary Physics Harvard University, Cambridge, MA. 02138 ABSTRACT This paper presents an overview of an interdisciplinary study of the kinematics and dynamics of reef fish larvae in offshore currents, especially mesoscale quasigeostrophic eddies. We develop herein the conceptual framework and background infor- mation for our on-going studies, the preliminary results of which will be presented at the 1983 ASZ meetings. The focus of this research has been to define the appropriate scales in ocean cir- culation relative to the developmental biology and reproductive ecology of coastal marine fishes whose larvae are planktonic. INTRODUCTION Interest in the topic of larval dispersal by ocean currents has been longstanding and its broad implications to ecology, evo- lution and fisheries are well known. Yet, the factors that ulti- mately limit the distribution of oceanic planktonic organisms have not been clearly specified (Wiebe 1976) . The constraints upon the transport of planktonic larvae along coasts and across oceans are few. Only the direction and velocity of the prevail- ing currents, the timing of reproduction, and the length of the planktonic stage place limits upon where, when and how far species are transported (Scheltema 1972, 1977). One critical variable in the survival of certain shore-dwelling species may be the phase and quantity of larvae of one species relative to the phasing and quantity of another in contest for spaces sporadi- cally open to occupancy (Sale 1977, 1978). Determining the fate of fish larvae as plankton in open ocean currents is not just interesting with regard to the life cycle of fishes. It is also crucial to the resolution of key ecological hypotheses concerning species distributions and diver- sity in reef and shore communities (Helfman 1978, McFarland 1982) . Furthermore, resolution of these important questions leads directly into a discussion about the evolutionary stra- tegies of island species, one notable feature of which is the origin and maintenance of endemic species. Other aspects of this work are basic to applied fisheries and the development of effec- tive fisheries management plans, since a majority of food, game and aquarium fishes possess planktonic larvae and/or may seek larval fishes as food (e.g., Huntsman e_t al. 1982). The following discussion highlights the key questions and outlines significant results and references pertaining to the study of planktonic larvae of coastal marine fishes. An overall approach for the study of physical oceanic processes affecting ichthyoplankton distributions ideally would include: 1) mapping current patterns and confirming the appropri- ate scales of variability, 2) strategic sampling of plankton in 29 and out of specified currents, 3) relating the time scales of current variability to the duration of pelagic lifespans of lar- vae by aging techniques (especially otolith ring counts) , 4) mon- itoring spawning and recruitment of shore fishes in relation to the dynamics of the offshore current field. Some other aspects for such an interdisciplinary study also have been discussed by Richards (1982) . OCEAN EDDIES AND PLANKTON Mesoscale eddies, in general, are the features of ocean cir- culation which most likely entrain planktonic organisms, thereby affecting their transport (Wiebe et aj,. 1976, Cox and Wiebe 1979, Angel and Fasham, 1983). In particular, recent findings on the variable nature of ocean circulation around islands and along coasts (for review see Chopra 1973, Hogg e_t a_l. 1978) and the increasing understanding of eddy entrainment and advective processes provide a potential solution to the anomolous spawning seasons of tropical marine fishes at some locations. The vari- able occurrence of mesoscale ocean eddies (on the order of monthly and perhaps seasonal periods) and the peak reproductive season of coastal marine animals (with planktonic larvae) may be synoptic. We are studying such a case in the Hawaiian Islands. We will discuss the potential role of eddies in marine biol- ogy as a general phenomenon. Among the specific details yet to be carefully considered are any potential differences that may exist according to the type of eddy involved (Angel and Fashman 1983) . Two types of eddies are known: warm-core or anticyclonic eddies and cold-core or cyclonic eddies. In the northern hemi- sphere anticyclonic eddies rotate clockwise and cyclonic eddies rotate counterclockwise. Eddies develop by a variety of physical mechanisms including wake phenomena in the lee of islands and by extreme meanders of a strong stream current. Whether or not the way in which an eddy is generated has a subsequent impact on its biological function is unknown. Robinson (1983) recently has compiled a comprehensive review of the role of eddies in marine science. A more general physical feature relating to the distribution of pelagic fishes and ichthyoplankton are thermal fronts. Ther- mal fronts are sharp gradients of temperature with distance. The expression of fronts can result from several sources, including the edges of eddies, upwellings, current shears, convergence and divergence zones, jet streams and, vertically, at the thermo- cline. To what degree pelagic larvae of shore fishes are able to select locations within currents by temperature or other cues is unknown. The overall correlation of fish distributions across fronts in general has yet to be determined. The possibility also exists that various fronts may differ in some physical way that is detectable by fishes but which we have not yet recognized. There are, however, good examples of the distribution of mycto- phid fishes consistent with thermal patterns across fronts and eddies (Brandt 1981, 1983; Brandt and Wadley 1981). Myctophids are mesopelagic fishes whose entire lifespans are spent in the open sea. 30 It is not our intent here to review past works on plankton communities. The open ocean dynamics of plankton migrations and global distributions are well known and have been extensively discussed by others (e.g., Reid e_t aj,. 1978, Wiebe e_t al. 1976, Cox and Wiebe 1979, Wiebe and Boyd 1978, Boyd et a_l. 1978) . Among the remaining questions is how the physical dynamics interact with the biological dynamics (e.g., reproductive pat- terns, swimming and energetic abilities of the plankton, etc. ) to form and maintain a "patch" (Haury e_t al. 1978) . Evidence for the role of ocean eddies in trapping planktonic organisms can be obtained by sampling zooplankton densities inside and outside of eddies. In the few cases where such discrete sampling has been done, the general result has shown higher abundances of zooplank- ton in eddies than in surrounding waters (Uda 1957, Uda and Ishino 1958, Wiebe ejt al. 1976, Ortner £t al. 1978). Other planktonic populations can become trapped inside eddies and tran- sported out of the species' normal range as the eddies move (Wiebe et al. 1976, Ortner e_t a_l. 1978, Boyd et al. 1978, Wiebe and Boyd 1979, Cox and Wiebe 1979) . Loeb (1979) presented data on larvae and mesopelagic fishes which accumulate inside the North Pacific Gyre (also, Reid e_t al. 1978) . REEF FISHES SPAWNING STRATEGIES The longstanding belief that tropical marine animals spawn continuously throughout the year without seasonal variation no longer appears generally valid. Distinct peaks of reproduction have now been documented for marine fishes in several tropical localities (Munro e_t al. 1973, Watson and Leis 1974, Johannes 1978, Lobel 1978, Nzioka 1979). In the absence of strong and recognizable seasonal fluctuations characteristic of the tem- perate latitudes, annual periods of peak reproduction by tropical coastal marine species are difficult to explain. We have collected data and examined evidence which suggests that seasonal reproduction by certain tropical species may be in phase with variable offshore quasigeostrophic mesoscale circula- tion. This circulation is a major environmental factor determin- ing the fate of the planktonic larvae of coastal species. The model species is one which lives its adult life in coastal marine habitats but whose larvae are planktonic in offshore waters. This is typical for a majority of reef fishes. Such species may spawn seasonally in response to natural selection acting on the survival of planktonic offspring. These offspring float with ocean currents which advect and disperse them. Past emphasis has been on the idea that widespread transport of planktonic larvae across long distances is an evolutionary adaptation reducing the susceptibility of a population or lineage to extinction by local catastrophes (Vermeij 1978) . The ecology of some shore fishes, however, suggests the possibility that transport of offspring far from the site of origin or native habitat may not always be favored by Natural Selection. The "lottery" hypothesis, described by Sale (1977, 1978, 1982) , is based on experimental field evidence showing that the availabil- ity of living sites limits the numbers of pomacentrid fishes and 31 that similar species utilize the same kind of space. Priority of arrival as recruits, rather than subtle differences in ecological requirements or competitive abilities as adults, appears to determine which species occupy each site (Sale 1978) . Thus, Natural Selection should favor those individuals of a species who maximize the return of their offspring to home reefs. Sale's lottery hypothesis has stirred considerable debate (Smith 1978, Dale 1978, Anderson £t al. 1981) . It is clear that the resolu- tion of whether to accept or reject it as a viable hypothesis lies, in part, in determining the fate of the planktonic larvae in the ocean currents (Helfman 1978, McFarland 1982). Even though an extensive literature deals with the genetics and evolutionary consequences of dispersing offspring (Gadgil 1971) , little strategic sampling has been done to ascertain the frequence of dispersal in natural animal populations in general, and in marine populations in particular (Leis 1983) . Recent advances in knowledge of the ocean circulation make this just now a feasible scientific undertaking. The Hawaiian fauna has a high percentage of endemic species, emphasizing that many species have clearly delineated and limited distributions. Conversely, the life history strategy of some species may be to colonize distant habitats. A mechanism whereby such larVae are transported en masse would increase the chance of successful colonization elsewhere. Hawaiian eddies appear to remain stationary near the islands up until several months and then, perhaps, to move off into the open ocean. Thus, eddy mechanisms could be responsible for transporting fish larvae through the open ocean as well as maintaining populations near shore. Any feature of the ocean circulation which would accumu- late larvae is likely to accumulate other planktonic particles also. Larvae in eddies are transported in microcosm ecological communities (Wiebe e_t al. 1976, Wiebe 1976, Ortner £t al. 1978). The duration of existence and extent of movements of eddies are important factors determining whether or not eddies function to retain plankton near islands or to transport plankton away. An especially important relationship which needs to be examined is the comparative lifetimes of mesoscale eddies and the planktonic phase of shore fishes. FISH LARVAE IN EDDIES We have investigated whether or not offshore ocean eddies near Hawaii play a key role in the life cycle of coastal marine species by functioning to retain planktonic larvae near the islands until such larvae metamorphosize and return to inshore habitats. In general, the peak period of eddy formation and movement appears to coincide with the peak season of reproduction by Hawaiian shore fishes. In the following sections we will develop the ideas, discuss the mechanisms related to the proto- type model, and present relevant biological and physical evi- dence. We will present preliminary results of our field investi- gations at the December 1983 meeting of the ASZ. The possibility that ocean eddies near islands might func- tion as reservoirs for planktonic larvae of coastal species was 32 indicated early by Boden (1952) . He commented that animals in Bermuda breed during months when wind driven circulation is at a minimum. At this time, a convergence between warm and cold water currents occur, and anticyclonic eddies form. Boden discovered that plankton accumulated in this region of the convergence. Emery (1972) reported a similar situation for the island of Bar- bados, West Indies, and provided evidence for the existence of eddies in the island's lee. Planktonic larvae in these eddies ostensibly avoid being swept away from the island (Emery 1972) . Eddies discovered in Hawaiian waters have been similarly impli- cated by Jones (1968) . Evidence for the function of eddies in preventing loss of larvae from Hawaiian waters and maintaining fish larvae of the family Acanthuridae near shores was later obtained by Sale (1970) . He presented data suggesting that the surface eddies were effective in trapping planktonic larvae which then were revolved past the island of Oahu (about 25 to 50km from shore) every five to six days (see also Leis and Miller 1976) . A key feature of the Hawaiian eddies is that some remain in the vicinity of the islands for at least 65 days, (Patzert 1967, Lobel and Robinson, in prep.), which is sufficient time for development of some larvae into a stage capable of migrating back to the inshore habitat (eg., Acanthurus triostegus sandvicensus, Randall, 1961; Chaetodon milaris. Ralston 1976) . The biological importance of retaining planktonic eggs and larvae near shore to the maintenance of island populations is obvious. However, the behavioral and physical mechanisms by which planktonic larval fishes return have not been elucidated and the potential role of eddies has not been widely recognized. It is well known, for example, that the larvae of fishes which dominate Hawaiian inshore habitats (e.g., labrids, scarids, acanthurids, and chaetodontids) are nearly absent from inshore waters, but instead are found offshore many kilometers away (Miller 1973) . Many Hawaiian fishes display a collective spawn- ing peak in the spring (Watson and Leis 1974, Lobel 1978). Wat- son and Leis (1974) suggested that the spring spawning peak was an adaptation to local currents. A general shift in the prevail- ing large scale currents around the Hawaiian Islands occurs in late spring and again in the fall (Barkley e_t al. 1964) . Watson and Leis (1974) proposed "These shifts, which should be associ- ated with weaker currents, occur with spring and fall spawning peaks. Synchronization of spawning with periods of reduced current flow would allow development and metamorphosis of the pelagic larvae before they were swept out to sea" (see also Johannes 1978) . Additional evidence suggests that it is not the shifts in prevailing currents, per se, but the offshore eddies and other variable mesoscale currents which form during those times which may be the important factor involved (Lobel 1978) . We have suggested a relationship between the occurrence of ocean eddies and the distribution and abundance of coastal marine larvae. If this relationship is approximately true, then we expect the following, given two alternative environmental cir- cumstances: 33 I. If ocean eddies or other currents acting to reduce dispersal are predictable in time and space and meet other basic criteria (i.e., the eddies persist at least 2.5 months and remain near islands) then: A. Spawning of coastal marine species with planktonic larvae is expected to be synoptic with the time that eddies are most probably present. B. Recruitment of pre- juveniles to the reef also will be concurrent with eddies in time and space. II. If eddies occur but are unpredictable, dissolve sooner than 2.5 months, or move far away, then: A. Spawning will be independent of eddy occurrence. B. Recruitment probably will be greatest when eddies occur near shore but also will be unpredictable. The complex process of offshore transport and return of lar- vae to coasts involves quasi-continuous exchange of recently spawned eggs being swept offshore and older larvae being brought back. Ocean currents affecting the transportation of larvae must not only bring fish to hospitable coasts but also must do so within time scales appropriate to larval developmental periods. Larvae must not merely be brought back nearshore but must be returned at a time when they can undergo metamorphosis and settle onto reefs. Among the significant questions remaining are: What is the mechanism by which larvae return to coastal habitats? Are pre- juveniles (post-larvae) able to "home-in" on some cue and actively swim some distance from sea to shore or is passive drift the sole mechanism? A CONCEPTUAL MODEL FOR THE EFFECT OF MESOSCALE OCEAN CURRENTS ON THE LIFE CYCLE OF REEF FISHES Our research to date has led us to formulate a working con- ceptual model for the processes discussed above. In summary, reproduction and recruitment of coastal fishes occurs to some degree all year but with peaks during the spring-summer months in Hawaii, with a phase lag of a few months between first reproduc- tion and first recruitment. The offshore currents are variable in both space and time (fairly rapid to quasi-steady) , and typi- cally consist of one or more eddies and/or fronts and currents. These mesoscale features will, in certain places, make contact with sections of coasts on islands in the Hawaiian Archipelago. This creates locations where offshore water is swept onto the reef and locations where water is swept off. Thus, reef habitats may be either near the source of steady incoming offshore water or be situated further down stream along the same path. While an eddy is quasistationary in adjacent offshore waters, all nearby coastal currents are dominated by the eddy flow field, and drift- ing particles are likely to be entrained. At other places or 34 times, the mesoscale features are absent, the currents are weak and simple tidal oscillations occur (except during episodes of occasional oceanic events, e.g., storms, tsunamis, errant eddies, etc.). Under these conditions, oceanic and coastal waters mix mainly in regions of flood divergence and ebb convergence flows associated with tidal processes. The overall picture will change with time and location especially when a large eddy is present and moves along or away from the coast. An eddy is expected to act as a major entrapment and near-island retention mechanism. Thus, depending on the location of a reef relative to offshore mesoscale features, the larvae produced by the residing fauna may be swept over stretches of reef before moving offshore, move quickly offshore, or be trapped nearshore. Once of f shore, larvae may be carried out to sea and lost or trapped in an eddy which, if it remains near an island, will enable larvae to return to suitable adult habitats. Recruitment may depend, in part, on those mesoscale eddies and currents bringing larvae near to shore. If the seasonality of such currents is predictable in time and space, then potential exists for species to adapt by developing a peak in reproduction at times when the offshore mesoscale field most favors nearshore retention of larvae. Research supported by National Science Foundation grants No. OCE-80-09554 and No. OCE-81-17891 . LITERATURE CITED Anderson, G.R.V. , A.H. Ehrlich, P.R. Ehrlich, J.D. Roughgarden, B.C. Russell, and F.H. Talbot. 1981. The community structure of coral reef fishes. Amer. Nat. 117 (4) : 476-495 . Angel, M. V. , and M.J.R. Fasham. 1983. Eddies and biological processes, p. 492-524. In: A.R. Robinson (ed.), Eddies in Marine Science. Springer Verlag, Berlin. Barkley, R.A. , B.M. Ito, and R.P. Brown. 1964. Releases and recoveries of drift bottles and cards in the Central Pacific. U.S. Fish Wildlf. Serv. Spec. Sci. Rpt. 492:32 P. Boden, B.P. 1952. Natural conservation of insular plankton. Nature 169:697-699. Boyd, S.H., P.H. Wiebe, and J.L. Cox. 1978. Limits of Nematoscelis megalops in the Northwestern Atlantic in relation to Gulf Stream cold core rings. II. Physiological and biochemical effects of expatriation. J. Mar. Res. 36:143-159. Brandt, S.B. 1981. Effects of a warm-core eddy on fish distributions in the Tasman sea off East Australia. Mar. Ecol. Prog. Ser. 6:19-33. Brandt, S.B. 1983. Temporal and spatial patterns of lanternfish (family Myctophidae) communities associated with a warm-core eddy. Mar. Biol. 74:231-244. Brandt, S.B. , and V.A. Wadley. 1981. Thermal fronts" as ecotones and zoogeographic barriers in marine and freshwater systems. Proc. Ecol. Soc. Aust. 11:13-26. Chopra, K.P. 1973. Atmospheric and oceanic flow problems 35 introduced by islands. Adv. Geophys. 16:297-421. Cox, J., and P.H. Wiebe. 1979. Origins of oceanic plankton in the middle Atlantic Bight. Est. Coastal Mar. Sci. 9:509-527. Dale, G. 1978. Money-in-the-bank: A model for coral reef fish coexistence. Env. Biol. Fish. 3:103-108. Emery, A.R. 1972. Eddy formation from an oceanic island: Ecological effects. Carib. J. Sci. 12:121-128. Gadgil, M. 1971. Dispersal: Population consequences and evolution. Ecology 52:253-261. Haury, L.R. , J. A. McGowan and P.H. Wiebe. 1978. Patterns and processes in the time-space scales of plankton distributions, p. 277-327. In: J. Steele (ed.) Spatial Patterns in Planktonic Communities. Plenum Press, N.Y. Helfman, G.S. 1978. Patterns of community structure in fishes: Summary and overview. Env. Biol. Fish. 3:129-148. Hogg, N.G., E.J. Katz, and T.B. Sanford. 1978. Eddies, islands and mixing. J. Geophys. Res. 83:2921-2938. Huntsman, G.R. , W.R. Nicholson, and W.W. Fox, Jr. 1982. The biological basis for reef fishery management. NOAA Tech. Mem. NMFS-SEFC-80. Johannes, R.E. 1978. Reproductive strategies of coastal marine fishes in the tropics. Env. Biol. Fish. 3:65-84. Jones, R.S. 1968. Ecological relationships in Hawaiian and Johnston Island Acanthuridae (surgeonf ishes) . Micronesica 4:309-361. Leis, J.M., and J.M. Miller. 1976. Offshore distributional patterns of Hawaiian fish larvae. Mar. Biol. 36:359-367. Leis, J.M. 1983. Coral reef fish larvae (Labridae) in the East Pacific barrier. Copeia 1983 (3) :826-828. Lobel, P.S. 1978. Diel, lunar and seasonal periodicity in the reproductive behavior of the Pomacanthid fish, Centropyge potteri. and some other reef fishes in Hawaii. Pac. Sci. 32:193-207. Loeb, V.J. 1979. Larval fishes in the zooplankton community of the North Pacific central gyre. Mar. Biol. 53:173-191. McFarland, W.N. 1982. Recruitment patterns in tropical reef fishes, p. 83-91. In: G.R. Huntsman, W.R. Nicholson, and W.W. Fox, Jr. (eds.), The Biological Basis for Reef Fishery Management. NOAA Tech. Mem. NMFS-SEFC-80. Miller, J.M. 1973. Nearshore distribution of Hawaiian marine fish larvae : Effects of water quality, turbidity and currents, p. 217-231. In: J.H.S. Blaxter (ed.), The Early Life History of Fish. Springer- Verlag, Berlin. Munro, J.L., V.C. Gaunt, R. Thompson, and P.H. Reeson. 1973. The spawning seasons of Caribbean reef fishes. J. Fish Biol. 5:69-84. Nzioka, R.M. 1979. Observations on the spawning seasons of East African reef fishes. J. Fish Biol. 14:329-342. Ortner, P.B. , P.H. Wiebe, L. Haury, and S.H. Boyd. 1978. Variability in the zooplankton biomass distribution in the Northern Sargasso Sea: The contribution of Gulf Stream cold core rings. Fish. Bull. U.S. 76:323-334. 36 Patzert, W.C. 1969. Eddies in Hawaiian waters. Hawaii Inst. Geophys. Rpt. No. 69-8. Ralston, S. 1976. Age determination of a tropical reef butterf lyf ish utilizing daily growth rings of otoliths. Fish. Bull. U.S. 74: 990-994. Randall, J.E. 1961. A contribution to the biology of the convict surgeon- fish of the Hawaiian Islands. Acanthurus triostegus sandvicensis. Pac. Sci. 15:215-272. Reid, J.L., E. Brinton, A. Flemings, E.L. Venrick, and J. A. McGowan. 1978. Ocean circulation and marine life, p. 65-130. In: H. Charnock and G. Deacon (eds.) , Advances in Oceanography. Plenum, Press, New York. Richards, W.J. 1982. Planktonic processes affecting establishment and maintenance of reef fish stocks, p. 92-100. In: G.R. Huntsman, W.R. Nicholson, and W.W. Fox (eds.), The Biological Basis for Reef Fishery Management. NOAA Tech. Mem. NMFS-SEFC-80. Robinson, A.R. (ed.). 1983. Eddies in Marine Science. Springer- Verlag, Berlin. Sale, P.F. 1970. Distribution of larval Acanthuridae of Hawaii. Copeia 1970:765-766. Sale, P.F. 1977. Maintenance of high diversity in coral reef fish communities. Amer. Nat. 111:337-359. Sale, P.F. 1982. Stock-recruitment relationships and regional coexistence in a lottery competitive system: A simulation study. Amer. Nat. 120:139-159. Scheltema, R.S. 1972. Dispersal of larvae as a means of genetic exchange between widely separated populations of shoal-water benthic invertebrate species, p. 101-114. In: B. Battaglia (ed.), Fifth Eur. Mar. Biol. Symp. Piccin Editore, Padua. Scheltema, R.S. 1977. Dispersal of marine invertebrate organisms: Paleo- biogeographic and biostratigraphic implications, p. 73-108. In: E.G. Rauffman and J.E. Hazel (eds.), Concepts and Methods of Biostratigraphy . Dowden, Hutchinson and Ross, Inc., Stroudsburg, PA. Smith, C.L. 1978. Coral reef communities: A compromise view. Env. Biol. Fish. 3:109-128. Uda, M. 1957. Enrichment patterns resulting from eddy systems. Proc. Ninth Pac. Sci. Congress 16:91-93. Uda, M. , and M. Ishino. 1958. Enrichment patterns resulting from eddy systems in relation to fishing grounds. J. Tokyo Univ. Fish. 44:102-129. Watson, W. , and J.M. Leis. 1974. Ichthyoplankton of Kaneohe Bay, Hawaii: A one-year study of fish eggs and larvae. Unhi-Seagrant-TR-75-01, Honolulu. 178 pp. Wiebe, P.H. 1976. The biology of cold-core rings. Oceanus 19:69-76. Wiebe, P.H., E.M. Hulburt, E.J. Carpenter, A.E. Jahn, G.P. Knapp, S.H. Boyd, P.B. Ortner, and J.L. Cox. 1976. Gulf Stream cold core rings: Large scale interaction sites for open ocean plankton communities. Deep Sea Res. 23:695-710. Wiebe, P.H., and S.H. Boyd. 1978. Limits of Nematoscelis megalops in the Northwestern Atlantic in relation to Gulf 37 Stream cold core rings. I. Horizontal and vertical distributions. J. Mar. Res. 36:119-142. 38 ON THE POSSIBILITY OF KIN GROUPS IN CORAL REEF FISHES Douglas Y. Shapiro Department of Marine Sciences, University of Puerto Rico Mayaguez, Puerto Rico 00708 ABSTRACT This paper concerns the widely-held assumption that members of schools of coral reef fishes are unlikely to be related because of pelagic dispersion of their eggs or larvae. The most direct way for adult schools to contain kin would be for siblings to remain together throughout the pelagic period and then to settle together as juveniles. For pelagic sibling groups to endure, mixing and dispersion by turbulent diffusion must be counter- balanced by biological processes for the formation, maintenance or selective survival of discrete egg or larval aggregations. I argue that there are sufficient hints of the existence of such processes to warrant examination of the possibility of kin groups in coral reef fishes. For example, demersal spawners may reduce the period of passive dispersion almost to zero by laying benthic eggs that release developmentally advanced larvae with the sensory and motor ability to aggregate actively soon after hatching. Other likely places to look for kin groups are among mouthbrooders , livebearers and pelagic spawners like lionfishes or species like Anthias whose young juveniles occupy large, sedentary groups. With rare exception, eggs and/or larvae of coral reef fishes undergo a pelagic phase during which many are thought to travel large distances in open ocean. It is generally assumed that pelagic eggs and larvae are thoroughly mixed, so that re- cruits at any particular site on a reef represent a random sample of the reproduc- tive products entering the local gene pool. If mixing occurs, kin are unlikely to settle together or find one another after metamorphosis. Hence, it is widely assumed that kinship plays no role in the evolution of coral reef fish behavior, schools or social groups (e.g., Thresher , 1977). However, at the beginning of, during, and at the end of the pelagic phase, eggs and larvae are clumped. Eggs are initially released in the sea by pelagic spawners or laid on the substrate by demersal spawners in a small, tightly-clumped mass. Thus, at the outset sibling eggs are closely aggregated. Little is known about patterns of pelagic dispersion for coral reef fishes, but among commercial temperate water fishes pelagic eggs and larvae are often found in patches (Hunter, 1980). Among these species patchiness is sufficiently prevalent that Hewitt (1981) concluded that patchiness is an important characteristic of larval existence with strongly positive survival value. At the end of the pelagic phase, coral reef fish larvae metamorphose and settle on the reef. Larvae of at least 23 species in several families settle in aggregates (Luckhurst & Luckhurst, 1977; Tal bot,£t aj . , 1978), often in bursts of recruitment over short periods of time (Victor, 1983). Thus, one finds patches in all pelagic stages. The most direct way for adult social groups, colonies or schools to contain kin would be for siblings to remain together throughout the pelagic period and then to settle together as juveniles. The aim of this paper is to outline some of the factors determining whether or not sibling aggregations will persist through the pelagic phase. To place dimensions on the problem, I will first treat eggs as 'though they were a passive, inert substance, such as dye, that is dispersed by diffusion. Assume that 2,000-40,000 neutral or positively buoyant eggs are released in the water col- umn at a depth of 10 m in a single spawn as a sphere 10-20 cm in diameter. If (1) the eggs are non-adherent, (2) horizontal turbulent diffusion is radially symmetric, 39 and (3) vertical diffusion is Fickian with a constant vertical diffusivity (Okubo, 1974), how large a volume will the eggs occupy 24 hours later? Calculations (Okubo, pers. comm.) are based on diffusion diagrams that plot horizontal variance a^ (i.e., the mean square distance from the center of mass) of dye distribution as i function of time since point release of dye in the ocean. Horizontal diameter of the egg patch is considered to be 2ax, i.e., twice the standard deviation (Okubo, 1971), and vertical height of the patch is 2az. After 24 hours, eggs would be spread 3-9 m vertically and 400-1000 m horizontally. Mean separation between neighboring eggs would be 24-71 cm vertically and 28-70 m horizontally for a spawn of 2,000 eggs with zero mortality, and 9-26 cm vertically, 10-26 m horizon- tally for 40,000 eggs. The average density of 40,000 eggs spread horizontally in a circle 700 m in diameter, and neglecting vertical spread, would be 0.10 eggs/m2. Assuming the eggs are dispersed at random within the circle according to a Poisson distribution, the probability of two or more eggs occupying a 1 m? portion of the circle is 0.004679. We could expect 1800 units, each 1 m? in size, within this circle each to contain >_ 2 eggs. What factors operate to make aggregation of sibling eggs more or less likely than this simple model suggests? Firstly, eggs probably do not behave like dye in the ocean. While the chorionic surface of fish eggs may often be smooth (Hemple, 1979), 27 tropical species (e.g., Robertson, 1981) produce eggs with external chorionic hooks, tendrils, spikes, or raised hexagonal ridges. In some cases these structures hold fertilized eggs in compact sheets. In others, chorionic sculpturing increases drag (Robertson, 1981). Female lionfishes of the family Pteroidae produce transparent mucous balls containing 2,000-15,000 eggs that remain in the mass until hatching 36 hours later (Fishelson, 1975). Turbulent motion is customarily broken into Fourier components with different length scales. Components with length scales smaller than or equal to the egg patch tend to break the patch and mix and spread the eggs. Components with length scale larger than the patch move the patch as a whole (Okubo, 1971). "As a result, the apparent power of the mixing which acts on a diffusing patch increases with patch size or, to put it another way, with the time elapsed since the patch first began to mix." (Okubo, 1971: 91). Empirically, horizontal variance is a function of diffusion time raised to the power 2.34 (Okubo, 1974). Thus, any tendency for newly spawned eggs to adhere to each other or otherwise remain in an aggregated cloud would delay the time at which any particular eddy length became capable of dispersing it and produce an inordinately large reduction of egg patch dimensions 24 hours later. Species in many families lay demersal eggs that have adhesive structures or properties that maintain their benthic attachments (Thresher, 1980). Dispersion does not begin for demersal spawners, for mouthbrooders such as cardinal fish and jawfish, or for livebearers such as some clinids, until after hatching or live birth. The period of completely passive egg dispersion is by-passed in these species. The simple model predicted an egg density 24 hours after spawning of 0.10 eggs per m?. Much higher egg densities have been recorded: up to 46,000 eggs/m2 in temperate fishes and up to 12 eggs/m? in tropical fishes (Hunter, 1980; Watson & Leis, 1974). Inter-egg distances may be as low as 1-2 cm at spawning and 15-20 cm several days later (Hunter, 1980). These distances are an order of magnitude lower than those in the above model. While these field observations suggest that egg clumping may be more likely than expected by the above model, very high egg densities almost certainly result in part from superposition of egg products from multiple spawns. I will return to the effect of multiple spawns on sibling aggregation later. Unlike an assumption of the model, dispersion is not uniform throughout the egg patch, but is Gaussian with densities greater at the center than at the periphery 40 (Hunter, 1980; Walsh, et al., 1981). This distribution will accentuate aggregation. On the other hand, passive dispersion will continue longer than 24 hours, at least among pelagic spawners. The actual duration will depend on the time of hatching, the onset of larval swimming, and perhaps on the overall length of the pelagic period. In cold-water fishes, the egg stage may endure from 2-4 days to 2 years (Hempel , 1979), but the precise time is species- and temperature-dependent. In coral reef fishes, where incubation temperatures are high, pelagic eggs hatch generally after 15-36 hours (Thresher, 1980). Demersal eggs usually require longer to hatch, up to 12 days in coral reef fish, but, since dispersion does not begin in these species until emergent larvae enter the water column, the duration of the egg stage does not influence spread. Once young larvae begin actively swimming, the possibility exists for behavioral mechanisms to influence dispersion. Swimming begins between 0.5-3 days post- hatching in coral reef fishes (Thresher, 1980) and 1-3 days after hatching in anchovies and Pacific mackerel (e.g., Hunter & Kimbrell , 1981). Larvae hatch from demersal eggs at a more advanced stage of development than larvae from pelagic eggs. Sensory and motor systems are more fully functional in the former and new larvae swim and catch prey very soon after hatching (Iwai, 1980). The short time interval, for these larvae, between hatching and onset of swimming provides short purchase for turbulent diffusion to act. Hence, we should expect dispersion to be far less in larvae from demersal than from pelagic eggs. Finally, very little is known about the total length of the pelagic period. Estimates vary from 3 days to 10 weeks in some coral reef species (Thresher, 1980; Brothers & McFarland, 1981). The density of eggs and larvae, whether aggregated or not, will clearly vary with mortality. During the pelagic egg stage, mortality varies between 2-95% per day, with larval mortality running 2-15% per day (Jones & Hall, 1975; Hempel , 1979). In spite of the large area covered by eggs in the model, eggs did aggregate within some 1 m? units simply from random processes. The probability of finding eggs or larvae in aggregates would be substantially increased if selective forces favored their survival over that of isolated eggs or larvae (Hewitt, 1981). On theoretical grounds, the probability that a predator will detect aggregated eggs may be less than the corresponding probability for dispersed eggs (Rubinstein, 1978). The same phenomenon will favor aggregated larvae as well. Larval aggrega- tion may also increase the difficulty of prey capture for a predator (Milinski, 1977). The larger the larval aggregation, the less likely it is that any particu- lar larva will be eaten, provided the predator is incapable of eating the entire patch. It is not known whether more active anti-predator behavioral adaptations occur in fish larvae or not, e.g., increased total vigilance as aggregate size increases, but such processes are well documented for adult social groups of terrestrial animals and for some marine insects (Treherne & Foster, 1980). Food densities necessary for larval survival and growth tend to be substantially higher than mean densities of food in the open sea. Consequently, there should be strong selective pressures for larvae to locate and remain in high-density food clumps (Hunter, 1980). One way to achieve this is through local enhancement: larvae in clumps should find food patches faster than isolated individuals because each larva could watch its neighbor's behavior as well as look for food. When a larva begins feeding, the observant neighbor could swim to the feeding site and itself begin feeding. Since (1) young larvae feed largely through visual mechanisms of perceiving prey (Iwai, 1980), (2) perceptive distance increases with the size of the perceived object (Hunter, 1980), and (3) neighbors are much larger than food items, a larva should see neighbors at greater distances than food. The effect will be a substantial increase in the search volume for food by each aggregated larva. Search volume, currently estimated at 0.1-1.0 liter/hour for individual larvae (Hunter, 1980), will be determined by larval swimming speed. In general, early 41 larval cruising speeds range between 1-3 body lengths per second and speed increases as larvae grow (Hunter, 1980; Hunter & Kimbrell, 1981). These figures provide a starting point for evaluating how far away a larval neighbor might be and still provide positive local enhancement. Larval food patches commonly occur in the open sea (Owen, 1981) and fish larvae do find food patches (Sherman, et al.. , 1981). When turbulent sea conditions disrupted food aggregations and diluted potential food items to below densities needed for larval survival, first-year anchovy recruitment declined markedly (Lasker, 1981). Once in a food patch, larvae tend to remain there by employing non-random search patterns. When larval anchovies entered dense food patches they decreased swimming speed and time spent swimming, and increased their turning probability by a factor of 5-6 (Hunter, 1980). The result was a substantial increase in the probability of staying within the food patch. The onset time of these oriented search patterns influences dispersion. The sooner they begin in larval development, the less widely dispersed larvae will be at the moment mechanisms for maintaining patchiness become operative. This leads to a consideration of the time of first feeding by young larvae. Generally, first feeding precedes or coincides with onset of swimming (Hunter & Kimbrell, 1981). Temperate fishes begin to feed 2-4 days after hatching. Since development is faster in tropical than in temperate waters, and armed with the above estimates of onset of swimming, we can estimate that feeding in coral reef fishes begins 0.5-3 days after hatching. We see, then, that there should be strong selective pressures for eggs and larvae to aggregate, as an anti-predator device, for local enhancement of food- patch searching, or simply to remain with food patches once they are found. In order to form or remain in patches, larvae must have appropriate motor and percep- tual skills and we have seen that swimming ability and feeding emerge early in larval development. Another mechanism by which larvae may limit dispersion is by diurnal vertical migration. Surface waters in the mixed layer suffer greater turbulent diffusion than deeper waters (Okubo, 1974). Particulate organic matter and many planktonic organisms concentrate at the boundary between layers. While vertical stratifica- tion of coral reef fish larvae has scarcely been studied (Watson & Leis, 1974), larvae may find a rich source of food if they migrated to the boundary, and there they would undergo less turbulent diffusion than if they remained near the surface. Regardless of the mechanisms, dispersion of fish larvae apparently does not continue throughout the entire pelagic period. Temperate larvae disperse for a limited period and then re-aggregate (Hunter, 1980; Hewitt, 1981). Thus, the duration of passive dispersion may be relatively short. Thus far, discussion has been directed to processes influencing dispersion of a single spawn of eggs or larvae. Spawning, however, generally involves all or part of a local population reproducing during the same period. Turbulent diffusion operating on multiple egg masses will mix offspring from adjacent spawns. The larger the number of eqq masses that are mixed and the greater the thoroughness of mixing, the lower will be the probability that small clumps of eggs or larvae late in the pelagic phase will contain kin. The amount of mixing will depend on slightly different factors for pelagic and demersal spawners. For pelagic spawners, critical factors will be the time and distance separating successive spawning events, and current speed. For example, if current speed over a reef is a modest 6.2 m/min (0.2 knots), pelagic spawnings 1-4 hours apart, as in some parrotfish (Clavijo, 1982), would initially produce egg masses separated by 0.4-1.5 km due to current plus whatever distance separated the spawning sites. With eggs concentrated 24 hours later in the center of a circle 700 m in diameter in a Gaussian distribution, eggs from early and late spawns would 42 mix relatively little. Similarly, eggs from spawning sites 500 m apart will mix substantially less than those from closer sites. Coral reef fishes often spawn at sites openly exposed to current that quickly sweeps eggs off the reef (Johannes, 1978; Barlow, 1981) and away from later egg masses. These current conditions could restrict genetic mixing. In demersal spawners, degree of genetic mixing will be influenced by spatial separation between benthic egg masses, current speed, and the degree of asynchrony of hatching within and between egg masses. If all eggs within one mass hatch synchronously, then young larvae will form a tight clump as they enter the water column. If intra-spawn hatching is asynchronous, sibs will enter the water column at intervals and will be more widely spaced by the currents. If adjacent egg masses hatch hours or days apart, currents will space emergent larvae with relatively 1 i ttle mixinq . In summary, the eggs of coral reef fishes begin their pelagic existence in small, discrete patches. Egg and larval patches subsequently are found at various stages throughout pelagic life and newly recruiting juveniles or post-larvae may settle in or form aggregations. There are tantalizing hints of strong selection pressures for pelagic eggs and larvae to aggregate, either to increase protection from predation or to enhance the search for food patches. The perceptual and motor apparatus to support active aggregation are available within a few days of hatching. Since coral reef fish eggs hatch rapidly in tropical seas, the total time for which eggs and young larvae are susceptible to passive turbulent diffusion, prior to onset of the ability to form or remain in patches, is short, on the order of 2-5 days. This time is substantially reduced, perhaps to zero, for demersal spawners, mouthbrooders, livebearers, and lionfishes, whose eggs remain together throughout egg development and whose larvae hatch at a large size and at a relatively advanced stage. Mixing of eggs released by separate spawning pairs within a local popula- tion will be limited to spawning episodes occurring close together in time and space. The stronger the current at the time of spawning, the closer in space and time successive spawns can be without substantial mixing. All of these factors render it possible, for at least some members of some species, that eggs will remain together throughout pelagic life and release larvae that settle aggregately to form kin groups or groups containing kin on the reef. Few of the factors influencing the possibility of kin groups have been thorough- ly examined in coral reef fishes. However, genetic relatedness among individuals on a reef can be studied with existing electrophoretic techniques. Likely places to look for kin groups are among colony-forming demersal spawners, lionfishes, mouthbrooders, livebearers, and fish like Anthias whose young juveniles occupy large, sedentary groups. ACKNOWLEDGMENTS I wish to thank A. Okubo, A. Mercado and P. Yoshioka for calculations, and R. Appeldoorn, J.Corredor, D. Hensley, M. Leighton-Shapiro, J. Morel 1, and Y.Sadovy for helpful discussion of ideas and criticism of the manuscript. Work was supported by NIH grant S06RR08103. BIBLIOGRAPHY Barlow, G.W. 1981. Patterns of parental investment, dispersal and size among coral-reef fishes. Env. Biol. Fish. 6: 65-85. Brothers, E.B., & W.N. McFarland. 1981. Correlations between otolith microstructure growth, and life history transitions in newly recruited French grunts (Haemulon flavolineatum (Desmarest), Haemulidae). Rapp. P.-V. Reun. Cons. Int. Explor. Mer. 178: 369-374. 43 Clavijo, I. 1982. Aspects of the reproductive biology of the redband parrotfish Span' soma aurofrenatum. Ph.D thesis, University of Puerto Rico, Mayaguez. Fishelson, L. 1975. Ethology and reproduction of Pteroid fishes found in the Gulf of Aqaba (Red Sea), especially Dendrochirus brachypterus (Cuvier), (Pteriodae, Teleostei). Pubbl . Staz. Zool . Napoli 39 suppl . : 635-656. Hemple, G. 1979. Early life history of marine fish: the egg stage. Washington Sea Grant Publication, University of Washington Press, Seattle/London. Hewitt, R. 1981. The value of pattern in the distribution of young fish. Rapp. P.-V. Reun. Cons. Int. Explor. Mer. 178: 229-236. Hunter, J.R. 1980. The feeding behavior and ecology of marine fish larvae, p. 287- 330. I_n: J.E. Bardach, J.J. Magnuson, R.C. May, & J.M. Reinhart (eds.), Fish Behavior and its Use in the Capture and Culture of Fishes. ICLARM Conf. Proc. 5, ICLARM, Manila. Hunter, J.R., & C. Kimbrell . 1981. Some aspects of the life history of laboratory- reared Pacific mackerel larvae (Scomber japonicus). Rapp. P.-V. Reun. Cons. Int. Explor. Mer 178: 344. Iwai , T. 1980. Sensory anatomy and feeding of fish larvae, p. 124-145. jji: J.E. Bardach, J.J. Magnuson, R.C. May, & J.M. Reinhart (eds.), Fish Behavior and its Use in the Capture and Culture of Fishes. ICLARM Conf. Proc. 5, ICLARM, Manila. Johannes, R.E. 1978. Reproductive strategies of coastal marine fishes in the tropics. Env. Biol. Fish. 3: 741-760. Jones, R., "& W.B. Hall. 1975. Some observations on the population dynamics of the larval stage in the common gadoids, p. 87-102. J_n: J.S. Blaxter (ed.j, The Early Life History of Fish. Springer-Verlag, Berlin/New York. Lasker, R. 1981. Factors contributing to variable recruitment of the northern anchovy (Engraul is mordax) in the California current: contrasting years, 1975 through 1978. Rapp. P.-V. Reun. Cons. Int. Explor. Mer 178: 375-388. Luckhurst, B.E., & K. Luckhurst. 1977. Recruitment patterns of coral reef fishes on the fringing reef of Curacao, Netherlands Antilles. Can. J. Zool. 55: 681- 689. Milinski, M. 1977. Experiments on the selection by predators against spatial oddity of their prey. Z. Tierpsychol . 43: 311-325. Okubo, A. 1971. Oceanic diffusion diagrams. Deep Sea Res. 18: 789-902. Okubo, A. 1974. Some speculations on oceanic diffusion diagrams. Rapp. P.-V. Reun. Cons. Int. Explor. Mer 167: 77-85. Owen, R.W. 1981. Microscale plankton patchiness in the larval anchovy environment. Rapp. P.-V. Reun. Cons. Int. Explor. Mer 178: 364-368. Robertson, D.A. 1981. Possible functions of surface structure and size in some planktonic eggs of the marine fishes. N.Z. J. Mar. Freshw. Res. 15: 147-153. Rubenstein, D.I. 1978. On predation, competition, and the advantages of group- living, p. 205-231. j_n: P.P.G. Bateson & P.H. Klopfer (eds.), Perspectives in Ethology, Vol. 3, Social behavior. Plenum, New York/London. Sherman, K. , R. Maurer, R. Byron, & J. Green. 1981. Relationship between larval fish communities and zooplankton prey species in an offshore spawning ground. Rapp. P.-V. Reun. Cons. Int. Explor. Mer 178: 289-294. Talbot, F.H., B.C. Russell, & G.R.V. Anderson. 1978. Coral reef fish communities: unstable, high-diversity systems? Ecol . Monogr. 48: 425-440. Thresher, R.E. 1977. Ecological determinants of social organization of reef fishes. Proc. 3rd Int. Coral Reef Symp., Miami, 1: 551-557, Atmospheric Sciences, University of Miami, Miami, p. 551-557. Thresher, R.E. 1980. Reef fish: behavior and ecology on the reef and in the aquarium. Palmetto Publishing Corp., St. Petersburg. Treherne, J.E., & W.A. Foster. 1980. The effects of group size on predator avoidance in a marine insect. Anim. Behav. 28: 1119-1122. 44 Victor, B.C. 1983. Recruitment and population dynamics of a coral reef fish. Science 219: 419-420. Walsh, J.J., CD. Wirick, D.A. Dieterle, & A.G. Tingle. 1981. Environmental constraints on larval fish survival in the Bering Sea. Rapp. P.-V. Reun. Cons. Int. Explor. Mer 178: 24-27. Watson, W., & J.M. Leis. 1974. Ichthyoplankton of Kaneohe Bay, Hawaii. Sea Grant Tech. Rep. UNIHI-SEAGRANT-TR-75-01 . 45 SETTLEMENT AND LARVAL METAMORPHOSIS PRODUCE DISTINCT MARKS ON THE OTOLITHS OF THE SLIPPERY DICK, HALICHOERES BIVITTATUS Benjamin C. Victor Department of Biological Sciences and Marine Science Institute University of California, Santa Barbara, California 93106 ABSTRACT Early in the sequence of daily increments on wrasse otoliths there is a transition from clear increments to a band of very faint, often indecipherable, wide increments. After about five of these wide increments clear increments resume. Planktonic larvae do not have the transition on the edge of their otoliths, while newly-metamorphosed juveniles have both the transition and the subsequent band on theirs. Experiments showed that metamorphosis takes about five days, during which the larva remains buried in sand. The transition therefore corresponds to settlement of the planktonic larva onto the reef, and the band is formed during the process of metamorphosis. These findings demonstrate that 1) it cannot be assumed that newly-appeared fish on the reef settled that day and 2) experiments are necessary to ascertain the meaning of marks on the otolith. INTRODUCTION Reef fish ecologists have only recently begun to focus their attention on the early life history of reef fishes. These fishes are unusual in that virtually every one of the many thousands of species that live on coral reefs has a planktonic larval stage. The interest in the ecology of larvae and the process of recruitment is indeed timely, for it is becoming increasingly apparent that many of the patterns of diversity and abundance of coral reef fishes are being determined by processes occurring in the plankton. Several studies have demonstrated that reef fish populations are limited by the supply of larval recruits, rather than by some resource on the reef (Wi 11 iams , 1980; Doherty , 1983 ; Victor , 1983) . Furthermore, some of my recent findings (in prep) indicate 1) that species-specific larval behaviors determine when successful recruitment occurs in some Caribbean wrasses and thus promote the coexistence of species on the reef, and 2) that the duration of the planktonic larval stage affects the distribution of wrasse species in the Indo-Pacific and may well account for the differing degree of speciation that has occurred within wrasse genera. Much of the progress that has been made on this subject is a result of the power of the daily otolith increment aging technique (Brothers, 1981; Brothers & McFar land , 1981 ) . Because there are both daily increments and a mark corresponding to the settlement of the planktonic larva on wrasse otoliths (Victor , 1982) , the date of settlement for any individual can be calculated by subtracting the 47 number of daily increments between the settlement mark and the edge of the otolith from the date of capture. The duration of the planktonic larval period can also be estimated by counting the number of increments between the center of the otolith and the settlement mark. The settlement mark appears under the microscope as a transition where the previously prominent dark lines that delineate each increment abruptly disappear. Regular increments only reappear after a band without clear increments is formed (see Fig. 1C) . It is, however, often possible to discern several (usually five) faint increments making up this band in some areas of the otolith. The aim of this study is to clarify which events during the process of settlement account for the transition and the subsequent settlement band on the otoliths of the slippery dick, Halichoeres bivi ttatus , one of the most abundant wrasses in the Caribbean. METHODS Planktonic larvae of the slippery dick were captured at a nightlight with an aquarium dipnet on Ukubtupo reef in the San Bias Islands of Panama. They were identified both by fin ray counts and by raising in an aquarium. Only larvae that had their full complement of fin rays and were of a size ready to settle were present at the nightlight. Some of those captured were preserved in ethanol immediately, while others were transferred to a ten-gallon aquarium containing sand and sea-water. Collections of juvenile slippery dicks also were made on Ukubtupo reef about the same time. The otoliths of all individuals captured were obtained by first removing the top of the cranium and then extracting the larger pair, the sagittae, from the base of the skull and the smaller pair, the lapilli, from the sides of the skull with fine forceps. The otoliths were then cleaned and dried and placed in a drop of immersion oil on a glass slide. They were subsequently examined under a compound microscope with transmitted light and a polarizing filter at magnifications of 400X to 1000X. RESULTS All twenty of the slippery dick larvae captured at the nightlight and preserved immediately had no settlement transition on their otoliths. In these fish the alternating light and dark lines that make up daily increments continued all the way out to the edge of the otolith (Fig. 1A) . Those larvae that were put into an aquarium had all disappeared into the sand at the bottom of the tank by the next morning. After a number of days these fish emerged from the sand and after some exploring took up residence in a corner or near some rubble. During those days in the sand they had lost the transparency and the melanophore pattern of larvae and developed the stripes, spots, and colors typical of juveniles of this species. The juvenile slippery dicks usually emerged on the fifth day (mean of 5.5, n=17, 48 ijSSJV v TRANSITION Fig. 1: Otoliths of slippery dicks. A: 9.9 mm SLAbar = 100 microns. B: 10.1 mm SL, bar=20 microns. C: 21.0 mm SL, bar=30 microns . 2 CO s: CO CD J E- E- 2 E- CO CO s: en CO cc ex U CO 2 H mCl, < -J O.E- 2 CnW OS CO CO E-1 cqE- Su rscn 2 CO o Cb CO CO -30 -25 -20 -15 -10 o o oo oo <£ JUVENILES 6 1 i t^---|... *s I 20 40 60 80 DEPTH (FT) 100 120 Figure 2. depth. T standoffs lumber of interactions per m2 (+ or - S.D. lines) versus total number of interactions, 0 = overgrowths, S = 57 Th is gi aggre overg (Fig. overg taxon i nter {% of subor for e 0/0; 6/0, Wi sors the g was n depth aggre depth 10% o 1 i s t e thems Hemec (23%) e rel ven i ssors rowth 3A). rown omic actio all d i n a t ach t zoant 0/0, th in chang orgon ever . At ssor , . Si r gre d bel el ves tyon Ver a t i v e c n Figs. , monot s and t Coral species groups ns . Th interac es) for a x o n o m i hids - 1/3. c r e a s i n ed subs acean E_ recorde 30' th i n i t i a mi 1 arl y ater of o w with i n i t i a (23%), o n g i a ( ontribution of the three most frequent interactors 3A & 3B. Clearly sponges were the most common onically increasing both the frequency of their heir relative standing as aggressors with depth s on the other hand were by far the most frequently , especially at depth (Fig. 3B). Several other were also frequently involved in these overgrowth eir relative contributions are given below as the tions as aggressors )/ (% of all interactions as depths of 10', 30', 60', 90' and 120' respectively c group: millepore corals - 12/34, 16/5, 0/0, 0/0, 12/0, 4/0, 0/0, 0/0, 2/0; ascidians - 1/1, 10/1, g depth the relative importance of various aggres- tantially. At the shallowest depth studied, 10', rythropodi urn was the most dominant aggressor, but d as a e demo ting 2 , at 6 all o the p ted at 1 rcini 11%); n aggressor in any observations below this sponge Age! us was the most influential 5% of aTT aggressive interactions at that 0', 90' and 120', those species that initiated vergrowths were demosponges; these genera are ercentage of the total overgrowth that they each depth: 60' = Chondrilla (34%), a (11%); 90' = Chondrilla (40%), Hemectyon 120' = Verongia (34%), Chondrilla (26%). 100- 80- 60- 2 40" O 20- (/) ® AGGRESSORS O O ,oo <. ** w 60- 2 40- 20- 0 *GORGONIANS SPONGES' CORALS -.is^t • ^f III II I SUBORDINATES "V-f**-^.. *GORGONIANS CORALS* SPONGES* DEPTH (FT) Fi gure 3. Percent of all interactions versus depth. Each taxonomic group was scored as either A) an aggressor (was overgrowing another species) or B) a subordinate (was being overgrown). 58 DISCUSSION W i n o spec for from quan from demo maxi note at 1 over both over rema M spac chem & Gr pred spon 1973 that W atta 1 eas over espe (Ran 1 i m i rese e pre pen-r i es r space open ti tat thes spong mum i that 20' . whelm a hi growt in ob echan e are i c a 1 s een , atory ges f ; Jac i n h i hethe eking t i ne growt c i a 1 1 dall t the arch sent e v i eef comm i c h n e s s , that oc -reef ha i v e data e commun es appea n our st coral c Neverth s that o g h diver hs on co scure at isms by likely Some 1974), a in n a t u or aggre kson & B bits cor r corals sponges ff ecti ve hs by sp y by the & Hartma i n f 1 u e n is neede dene unit but curs bi ta hav i t i e rs t udy over el es f co sity ral s thi whi c to i spon ltho re ( ss i v uss , al g the i s ) be onge ang n, 1 ce o d to e he i es als in ts ( e be s . o be at 9 inc s, t ral s of , wh s ti h sp nvol ges ugh Baku e gr 197 rowt msel unce caus s at elf i 968; f ag tes re that of the o with these e R e i s w i g en a v a i Al thoug correl 0' and reases he tota at 90' sponges i 1 e all me . onges c ve rapi are kno these h s, 1981 owth ov 5) , but h been ves use r t a i n , e of th depth, shes Po H o u r i g g r e s s i v t this demosponges play an important role Caribbean, not only in terms of respect to the intense competition nvironments. Demosponges are known , 1973), but until now virtually no Table on sponge/coral interactions h the relative aggressiveness of ated with depth, reaching a 120', it is interesting to from a mean of 18.3% at 60' to 39.1% 1 species richness of sponges clearly and 120'. The processes that permit and a high frequency of aggressive owing such a high % coral cover, an acq d grow wn to ave ty ). Th er cor only i d e n t i al lei but th e high Howe macant ui re th r poss pi ca e us als rece fied oche i s s f re ver , hus , ma ates ess ny e of has ntly (Su mica eems quen pre arcu inta and vi ru been all been has 11 iv Is t unl cy o dati atus an , e_t e spon hypoth al_. ges esi s , in on c pre oral in and expand their /or al 1 el o- 1 ent toxi ns ( Bakus considered anti- el ochemical s by suspected (Bryan, a sponge toxin an, e_fc aj., 1983). o ward off i k e 1 y (or at f aggressi ve on on sponges , and P_. paru ) may help to but further P- s , ACKNOWLEDGMENTS Special thanks go to M. W. Colgan and D. L. Green for assistance in collecting field data at Salt River Canyon and to the personnel of NULS-1 Hydrolab for their excellent operations program and unselfish help during both missions. We also thank J. C. Ogden, D. K. Hubbard and J. Bayes for help with various logistic and/or computer problems. This study was funded by NOAA grants NA81AAA02690 , NA82AAA00470 , NA82AAA00872 and NA82AAA01344 to T.H.S. and by funds from the West Indies Laboratory. LITERATURE CITED Bakus, G. J. 1981. Chemical Reef, Australia. Science Bakus, G. J. , & G. Green. a geographic pattern. Benayahu , Y . , & Y . Loya . reef sessile organisms 514-522. defense 211:497- mechan i sms 499. on the Great Barrier 1974. Toxicity in sponges and hoi othurians ; Science 185:951-953. 1981. Competition for space among coral- at Eilat, Red Sea. Bull. Mar. Sci. 31: 59 Birkeland, C, L. Cheng & R. A. Lewin. 1981. Motility ascidian colonies. Bull. Mar. Sci. 31:170-173. Bryan, P. G. 1973. Growth rate, toxicity and distribut encrusting sponge T e r p i o s sp. ( Hadromeri da ; Suberiti Marianas Islands. Micronesica 9:237-242. Bunt, J. S., W. T. Williams & B. E. Chalker. 1982. Cor tions at depths of 45 to 125 feet in the Bahamian re Fourth Int. Coral Reef Symp., Manila, 1981, 1:707-71 Buss, L. W., & J. B. C. Jackson. 1979. Competitive ne transitive competitive relationships in cryptic cora environments. Amer. Nat. 113:223-234. Dayton, P. K. 1971. Competition, disturbance and commu organization: the provision and subsequent utilizati a rocky intertidal community. Ecol. Monogr. 41:351- Jackson, J. B. C.,& L. Buss. 1975. Allelopathy and spa tion among coral reef invertebrates. Proc. Nat. Aca 72:5160-5163. Karlson, R. H. 1980. Alternative competitive strategie periodically disturbed habitat. Bull. Mar. Sci. 30: Karl son , R H by 1983. Disturbance and monopolization of Zoanthus s o c i a t u s ( Coel enterata , Anthozo aggression by to the swift. s c 1 e r a c t i Bull. Ma resource Sci. 33:118-131. Lang, J. C. 1973. Interspecific II. Why the race is not only 21:952-959. Osman, R. W. 1975. The establishment and development o epifaunal community. Ecol. Monogr. 47:37-63. Randall, J. E., & W. D. Hartman. 1968. Sponge-feeding f West Indies. Mar. Biol. 1:216-225. Re i swig, H. M. 1973. Population dynamics of three Jama Demospongiae. Bull. Mar. Sci. 23:191-226. Rutzler, K. 1970. Spatial competition among Porifera: epizoism. Oecologia 5:85-95. Sammarco, P. W., J. C. Coll, S. LaBarre & B. Willis. 19 tive strategies of soft corals ( Coel enterata : Octoco allelopathic effects on selected sci eractin i an coral Reefs 1:173-178. Intertidal distribution of zoanthi of Panama: effects of predation and Sebens , K. P Cari bbean Bull. Mar Sheppard, C . 1982 coast Sci . R. C. corals with reference Ser. 1:237-247. Sheppard, C. R. C. 1982. Coral major controls. Mar. Ecol. Suchanek, T. H. 1983. Control distribution by Callianassa bioturbation. J. Mar. Res. Suchanek, T. H., & D. J. Green. 32:316-335. 1979. Interspecific aggression betw to their distribution. Mar. E of didemnid ion of the dae ) in Guam, al associa- gion . Proc : 4. tworks: non- 1 reef nity on of space in 389. tial competi- d. Sci. U.S.A. s in a 894-900. a spatial a) . Bull . Mar, n ian coral s r . Sci. f a mari ne ishes of the i can solution by 83. Competi- r a 1 1 i a ) : s. Coral ds on the desiccation. een reef col . Prog. populations on reef slopes and their Prog. Ser. 7:83-115. of seagrass communities (Crustacea, Thalassinide. 41:281-298. and sediment ■a) 1982. Interspecific co between Paly thoa cari baeor urn and other sessile inver St. Croix reefs, U.S. Virgin Islands. Proc. Fourth Reef Symp., Manila, 1981, 2:679-684. Sullivan, B., D. J. Faulkner & L. Webb. 1983. Siphonod metabolite of the burrowing sponge S i phonodi ctyon sp inhabits coral growth. Science 221:1175-1176. m p e t i t i o n tebrates on Int. Coral i c t i d i n e , a . that 60 DISTRIBUTION OF SWEEPER TENTACLES ON MONTASTRAEA CAVERNOSA Elizabeth A. Chornesky Division of Biological Sciences, The Univ. of Texas, Austin 78712 Susan L . Willi ams Mar. Sci. Res. Ctr., State Univ. of New York, Stony Brook, NY 11794 ABSTRACT I spac cni d desc but the swee whic enco cong both of s enco n d i e, s ae, ri be deve Cari per h do unte ener the weep unte rect ome s 0 r sw d thu 1 op a bbean tenta not rs . i c sp numb er te r . com pet i p e c i e s eeper t s f ar , s compe reef c c 1 e s d i necessa Neverth eci es M er of p ntacl es ase INTRODUCTION Reef corals are known to use a variety of mechanisms to compete for limited substrate space in crowded reef environments. Two of the best described are the use of mesenterial filaments (Lang, 1971, 1973; Sheppard, 1979) or of "sweeper tentacles" (Richardson, et al., 1979; Wellington, 1980; Bak , _et _al_., 1982; Chornesky, 1983~J by some corals to damage the tissues of neighboring corals. Mesenterial filaments are normally present in all polyps of every coral. When corals of different species are placed into direct contact, these digestive filaments are deployed rapidly and extracoel enteri c digestion of opponent tissues may take place within hours (Lang, 1971, 1973; Sheppard, 1979). The immediate "winner" (i.e., the animal remaining undamaged) in such interactions is generally predictable among various species pairs. Unlike mesenterial filaments, sweeper tentacles (elongate tentacles with specialized cnidae) are found only on certain species of coral (see Lewis and Price, 1975; Bak and Elgershuizen, 1976). Moreover, within these species, sweepers may not be present on all colonies, and, when present, may be erratically distributed over the colony surface. On some corals, sweeper tentacles develop specifically after damage by mesenterial filaments (Wellington, 1980; Bak, e_t al., 1982; Chornesky, 1983) or after contact with recognition XChornesky, 1983) of other corals. In natural interactions, this delayed development, and thus the ability of the coral to utilize sweepers against a neighbor, occurs some time after the interaction has begun (on the order of a month--Wel 1 i ngton , 1980; Bak, ejt al . , 1982; Chornesky, 1983). However, on at least one species of 61 coral "read coral 1 ocat Sw coral the 1 i n c 1 u 1977; water predi Lang, prese appen cl ose and " (Lang and d compe (Ell i , sw y" t s wh i on eepe Mon eepe o pa i ch on t r te tast r te rt i c g row he c ntac raea ntacl i pate too ol ony 1 es c cave es a i n cl OS rel ommo rnos ocat de: Ri c cur ctab unp nt , dage gro poly , 19 eve! titi s an i on con ha rd rent 1 e p ub . swee s (L wth f unc 79). opme ve i d So of s cent son , s (P atte data pers ewi s of a ti on He nt o nter 1 and weepe ratio et a^ rice, rn of ). P on M and dj ace al" s re we f swe actio er) . r te n a r 1-. 197 di s erha . ca re r comp e (d at i v nly a (L ntac ound 1979 3, 1 trib PS i vern out i etit epen e to a re i nna 1 es col ) ; m n de ut i o n pa osa Pric nt c true pre eper ns w e, 1 ompe ture sent ten i th 975) ti to s wh pre tacl the nely i ve i ding, that seen eus ) . on M. ony p ax ima n Har n ove rt be have , def rs (R i ch m 1 imi n es on conge pres nter of of on c De cav ent an action course nea rby ol on i e scri pt ernosa enm 1 ex tog, r co caus been ens i i cha ight ary M. neri eters pansi o 1977 ) 1 ony s e they desc r ve str rdson , serve data o cavern d a r s wi , on com s of i ve var ( den n i n ; or urfa are i bed uctu et bot n th osa c speci es e therefore th neighboring their pet i tors ) . the Caribbean patterns of i ousl y Hartog , response to a less ces ( J .C . usually as f eedi ng res to deter al., 1979), h functions e behavior in artificial M . annularis MATERIALS AND METHODS Th River dive Col on c i n d e for m rem a i when Ou other expan col on on th night annul ese e Cany in t h i es o r bio ethod ned b their r i ni cora si on i es o e M. afte a ri s conta M. ca resul i m p o s du ri n Ne the 1 M. £a^ al rea obser each col on both caver ct wi verno ting s i b 1 e g the vert h onger ve rno dy we ved o obser i es w coral nosa to the M. x pen m on , St e NOAA f M. a_ ck s us s) . C etween polyp t i a 1 i Is cl o patter f M. a_ cavern r cora digest th the sa rem wound s to ob rem a i el ess , -term sa on 1 1 doc n six v a t i o n ere ca s were wa s d i a nnul a ents . Cro NULS nnul a were c i x . T II Un ri s an l ng u oral s pair s and ntent se to ns of nnul a nderwa were ed col t i s s u was t col on its s ri s , 1 osa c 1 s we ed ne i r t i a i ned . Th serve nder we w conse which ument night , the ref u 1 reco v i d e d ri s . 0 1 o n i e re cem a rby e ssues. cont r is unp the b of the ere pr quence the 1 ed. T s over posi t 1 y map rded . into Bound ondu hey derw d M. ter arra on i e es w o ex i es weep ocat s we ente xpa n Su acte redi ehav sat ov i d s of ocat hese the i ons ped For regi ari e cted we re ater cave at a de i n i t i a t Habitat rnosa ( epoxy nged s of e re c amine of M. ers. i ons re ca d i n t ded p bsequ d i n cted i or o urat i ed wi such i on o i nte foil of s and t data ons a s of -putty so that M . annu pth ed d i n N=8) ( see a g 1 a ri ont ract whet he cavern ed d r i n osa Thus , of e x i s ref ul 1 y o place olyps o ently , the are b e h a v i o f their on dive th the i ntera f sweep ract i on owing t weepers he beha analys d j a c e n t these " pri o ting map , al f M. poly a su ral swe of 60 u r i ng March were Chorn ap of sand uri ng troduc woul d r to i sweep ped. 1 col o caver ps of rround respon eper t feet a sat of 19 cemen esky, about M . c_a the d ti on af fee nt rod er te The f ni es nosa the d i ng t se ma en tac i n Sal t urati on 82. ted to 1983, 1 cm vernosa ay. of t the uc i ng ntacl es i rst Of M. i n i g e s t e d he de it 1 es opportunity to follow ctions for colonies of er tentacles wa s s subsequentl y we re wo months. During on the M. cavernosa vior and condition of is, each col ony of M. to and not adjacent adjacent" and "non- 62 adjacent" regions were designated arbitrarily on maps resulting from the first set of observations, and then held constant for all subsequent observations. Data were analyzed by contrasting changes in the relative proportion of polyps with sweeper tentacles on adjacent and non-adjacent regions of colonies. Adjacent regions were consistently smaller than non-adjacent regions on the same colonies (approximately a third of the size). Therefore, comparison of the absolute number of polyps with sweepers between adjacent and non-adjacent regions yields a conservative estimate of their density on adjacent regions (i.e., when adjacent and non- adjacent regions have equal numbers of polyps with sweeper tentacles, adjacent regions actually would have greater densities of sweeper tentacles than non-adjacent regions). RESULTS Figure 1 compares the number of polyps with sweeper tentacles on: A) tissues adjacent to the M. annularis, and B) tissues not adjacent to the M . annularis . The median and a quarter of the range is plotted on this graph since these data were clearly non- normal and sample sizes were small (7-8). There is a significant correlation between the number of polyps with sweeper tentacles on adjacent tissues and time after initiation of the experiment 8- 4-> E o 2 1- 0 1- B — 20 27 44 days after start expt ■2. 60 FIGURE 3. Cumulative changes in the number of polyps with sweepers. on adjacent (A) and non-adjacent (B) regions. Each value incorporates summed changes from all previous observa t ions. 64 (Spearman rank correlation rs = .333, P < .05) for non-adjacent tissues is not statistically - .133, P > .05). The ratio of A:B is plotted demonstrates that the relative proportion of p tentacles on adjacent regions increased during period. Figure 3 shows the cumulative changes number of polyps with sweepers on adjacent (A) (B) regions at various times after initiation These data suggest a cumulative decrease in th on tissues not adjacent to colonies of _M. a nnu clearly true for at least two colonies of M. c sweeper tentacles on non-adjacent tissues disa experiments were begun (4 and 6 weeks, respect Necrotic wounds appeared on most colonies o close to the colonies of M. cavernosa . Most ( formed in intervals between observations, duri number of M. cavernosa polyps with sweeper ten the M. annularis also increased. Sweeper tent cavernosa often were observed touching live M. close to necrotic regions. . The s i g n i f i in fig ol yps w the ob (media and no of the e numbe 1 a r i s . avernos correl ati on cant ( rs = ure 2 and i t h sweeper servat i on n ) in the n-adjacent experiment . r of sweepers This was a on which after the ppeared ively) . f M. annularis 5 of 67 n g w h i c tacl es acl es o annul a wounds h time the adjacent to f the M. ris tissues DISCUSSION M swee with numb us ( per tent M with port poss the i ncu tent A woun Chor cave cavernosa pert ot he er of SLW) polyp acl es oreov swee i ons ibly inter rred acl es 1 thou ds in nesky rnosa thei si mi M. a dama For grad swee at a cave r rea 1 ar t nnu 1 a ge to examp ual 1 y pert grea rnosa by M et a^ and a nn entacl r cora polyp sugges and t may i er , ou per te of col may de action by the close gh cau coral , 1983 we re ch. W o thes ri s M. a n le, if i nto entacl t e r d i sweep u 1 a ri s app es c Is. s ha t th he s ncre r da ntac onie crea i nc pro to ti on -cor ). i capa heth e ex deve nul a aren 1 ose Alt ving at , i ze ase ta s 1 es s no se a reas duct the sho al i tap ble er t peri 1 opm ri s 1-. 1 exper 979; s imenta col cont es p stan er t and ee C 1 in on i e act , ri or ce t enta /or horn tera tly to houg swe i n a of t cl OS ugge ore t i n s th es . i on si te uld nter pear of d he c ment ent by M s of M. to han cl es dige esky cti o can the h ou eper ddit he a e to st t xpan vol v e nu If and/ of be e acti s th amag ours s (i of s . ca i ncr si te r da ten ion, cros the hat si on ed i mbe r so , ore comp xerc ons at t ing e of • e. , weep vern _M . annu ca vernos ease the of comp ta speci tacl es , the num pheres o site of ei ther t of swee n compet of expa this m i g x p a n s i o n eti ti ve i sed in (see Bak he sweep tissues natural di gest i er tenta osa swee 1 a ri s an d i ge the mig sti o , 19 ns s sti o 1 cm ht e n by 83, omet a may be n by M. gap in f f i ci ent M. a nnu for disc i m e s may number e t i t i v e f i cal ly observat ber of s n e x i s t i such en he numbe per tent i t i v e in nded swe ht ref 1 e of a d d i encounte a s c r i b i n , _et al . er tenta of M. a_n i nterac on of M. cl es by p e r s ) is d _M. ca v able to annul a ri these ex ly deter 1 a ri s [s^ u s s i o n o differ) of polyps with encounters reported the ions by one of weeper tentacles ng sweeper counters . r of polyps acl es on teracti ons epers close to ct a "cost" tional sweeper rs . g causes for , 1982; cl es of M. nularis within t i ons is cavernosa by M . cavernosa less clear. ernosa g row devel op s . Similarly, pe ri ment s , ^. close g rowth ensu Ri chardson , f how natural 65 On duri n gener may 1 col on d i s t r it is seem i n c r e on ti tenta after exten over of pa other g dire al ly p ater r i es of i buted parti to aff asi ng ssues cl es o an in ded pe a col o st com cor ct c rese egre M. i n cul a ect i n d away n M. tera ri od ny m peti al s f ompet nt , b ss (W caver or whi c i t i on h ut appe el 1 i ngt nosa no seemi rl y i the d ensi t from cave ngl y di nterest i stri bu y on ad the zo rnosa a ct i on of t ay re ti ve cea ses ime , th f 1 ect a encoun t h th as b ar i on , rmal sord i ng t i on jace ne o ppar and e di t le ers . e dev een d n res 1980; ly po ered that of s nt ti f the ently may stri b ast a el opme ocumen ponse Chorn ssess patter i ntera weeper ssues i nter do no remai n u t i o n short nt of sweeper tentacles ted, sweepers are not to the encounter and esky, 1983). Since sweeper tentacles ns over colony surfaces, ctions with other corals s within colonies-- and perhaps decreasing action. Since sweeper t necessarily regress on a colony for an of sweeper tentacles -term historical record ACKNOWLEDGMENTS We thank the staffs of the NULS II Hydrolab and the West Indies Laboratory for essential surface support before, during, and after our saturation mission. We also thank C. Brunet, B. Nyden, and CM. Wahle for their enthusiasm and moral support during the mission;. and J.C Lang and CM. Wahle for many helpful discussions and useful comments on earlier drafts of this manuscript. We are especially grateful to Bruce Nyden to extraordinary and generous assistance with all aspects of the field work. This reseach was funded by a NOAA Hydrolab grant to E.A. Chornesky and CM. Wahle. LITERATURE CITED Bak , R. P sed ime Bak , R. P coral and ep Chornesky by the compet den Ha rto sancti . M. , a nt reje . M. , R interac i fauna . , E. A. reef c i t i o n . g , J . C thomae nd cti . M ti o M 1 ora Bi B. W on i . Te ns : ar. 983. 1 Ag . El n co rmaa inf Biol In a ri c (Co Montastrea caver ol . 1977 rail nosa funct i Int. C Gables Lang , J . coral s Edwa rd Lang , J . coral s Mar. S Lang , J . cavern on , pp. oral Re , F 1 o r i C 197 . I. s and H C 197 . I I . 23 197 ci C. osa pol Carib., 15th 46 ef da . 1. The a im 3. Wh : 2 9. yf u mee 3-46 Symp Bull . T imor (Sc 9. ., M gersh ral s . t , an 1 uenc . 69: duced i a ag . (in he ma pha ri 1 erac J_n: i a m i , uize Ma d R. e of 215 dev a ri c pre rgin a ) a t i n i D. L Uni n. 19 r . Bio Dekke time, -222. el opme i tes : ss ) . al ten nd the a); th . Tayl v . of 76. Pattern 1 . 37: 105-1 r. 1982. C 1 ocati on of nt of sweepe a response s of oi 1 - 13. ompl ex i ty of interaction r tentacl es to di rect tacl es of R_hodacti s sweeper ten ei r en i dom a or ( ed . ) , Pr Miami Press, tacl es of nd possi bl e oc . Thi rd Coral Interspecific aggressi rediscovery of S col ymi e) . Bui 1 . Mar. Sci . 21 Interspecific aggressi y t he race i s not al way 60-279. Are the sweeper tentacles of Montastrea nctional organs? Assn ting: 7 (abstract) . on by sclera a c u b e n s i s ( 952-959. on by sclera rs to the swi cti ni an Milne ct i n i an ft. Bull Island Mar. Lab 66 Lang. 1979. of Montastrea Maintenance cavernosa . Richardson, D. A., P. Dustan, and J. C. of living space by sweeper tentacles Mar. Biol. 55: 181-186. Sheppard, C. R. C. 1979. Interspecific aggression between reef corals with reference to their distribution. Mar. Ecol. Prog. Ser. 1: 237-247. Wellington, G. M. 1980. Reversal of digestive interactions between Pacific reef corals: mediation by sweeper tentacles. Oecol ogia 47: 340-343. 67 RELATIONSHIPS BETWEEN FISHES AND MOBILE BENTHIC INVERTEBRATES ON CORAL REEFS Nancy G. Wolf Section of Ecology and Systematics Cornell University Ithaca, NY 14853 Eldredge B. Bermingham Department of Molecular and Popular Genetics University of Georgia Athens, GA 30602 Marjorie L. Reaka Department of Zoology University of Maryland College Park, MD 20742 ABSTRACT Observations of 3 types of artificial reefs at 20 m depths show that fish preda- tion alters the pattern of colonization of stomatopods, the largest and most mobile members of the cryptic reef fauna. Recruitment by polychaetes probably is adverse- ly affected by the presence of fish predation also. Possibly because of their secretive habits, the densities of the remaining taxa of cryptic invertebrates were unaffected by fish predators. The data also suggest that the presence of an in- vertebrate biota influences the colonization and abundance of invertebrate-eating fishes. INTRODUCTION The importance of predator-prey interactions in governing community structure has been demonstrated for some marine systems (e.g., Paine, 1966; Dayton, 1975; Menge and Sutherland, 1976; and many others), but their role in coral reef communi- ties remains poorly understood. Coral-eating fishes can influence the structure of coral reefs (Kaufman, 1977; Neudecker, 1979; Wellington, 1982), and herbivores exert both direct and indirect effects upon coral reef communities (Ogden and Lobel, 1978; Hay, 1981; Hixon and Brostoff, 1983). In laboratory microcosm experiments, Brock (1979) showed that parrotfish grazers influence the abundance and diversity of the benthic flora and fauna, and that the presence of refuges (3-dimensional surfaces) is an even more important determinant of benthic community structure than the densities of consumers. Additionally, it has been suggested that a variety of structural, behavioral, and chemical defense mechanisms found in benthic reef organisms represent adaptations to strong predation pressures in reef environments (e.g., Bakus, 1966, 1981; Vermeij, 1978; Reaka, 1980a, 1980b; Reaka and Manning, 1981). However, astonishingly little experimental and quantitative information is available regarding the relationships between fish predators and the abundant in- vertebrate fauna that inhabits the reef substrate in the field. Most of this benthic biota lives in cryptic refuges under and within the coral substrate. Jackson and Buss (1975) have suggested that the cryptic sessile fauna currently does not experience strong predation. Many of these encrusting organisms grow in inaccessible sites, and particularly the colonial organisms exhibit chemical defenses that are used in competitive interactions (see also Buss and Jackson, 1 Present address: Biology Department, Woods Hole Oceanographic Institution, Woods Hole, MA 02543. 69 1979). However, virtually no field studies have qualitatively investigated the effects of fish predators upon the teeming mobile cryptic fauna that inhabits this environment. Also, while these invertebrates are the primary food source for many reef fishes (Randall, 1967), no studies have experimentally examined the importance of this mobile invertebrate fauna for the colonization and maintenance of popula- tions of reef fishes. In July 1980 we initiated a 7h year of study that experimen- tally investigated reciprocal interactions between fishes and the mobile cryptic invertebrate fauna in Salt Canyon, St. Croix. A preliminary analysis of the first 6 months of the data is presented here. A more thorough analysis of data from the entire 2% year period is in progress. MATERIAL AND METHODS Using the Hydrolab, an underwater habitat operated by NOAA on St. Croix, U.S. Virgin Islands, 15 artificial reefs were established at a depth of 20 m in Salt River Canyon in July 1980. Salt River Canyon is a submarine canyon 70-100 m wide with a sand floor flanked by a vertical and a sloping coral reef wall. Five reefs provided habitat for fishes and invertebrates (A reefs); each consisted of 11 cinderblocks arranged in a pyramid 2 blocks wide and 3 blocks high, with 6 pieces of sun-dried dead coral rubble placed around the pyramid base. The components were tied together with nylon rope and anchored in place with iron reinforcement bars. Five reefs consisted of a cinderblock pyramid without rubble, providing habitat for fishes but little shelter for invertebrates (B reefs). Five reefs, each composed of rubble arranged in the same pattern as in the A reefs but lacking a cinderblock pyramid, provided habitat more suitable for cryptic invertebrates (C reefs). The 15 reefs were arranged serially (ABCABC...) 10 m apart and 10 m out from the sloping east wall of the canyon. Fishes on the reefs were censused visually one week after establishment, and at approximately 30-day intervals thereafter. Using information from the literature (Randall, 1967; Clavijo, et al., 1980) and personal observations, we assigned each individual to a feeding guild- (planktivore, herbiv- ore, piscivore, piscivore-invertivore, invertivore) based on its size class (post- larval, juvenile, adult) and species (Appendix Table 1). We have included fishes that eat only invertebrates and those that eat both invertebrates and smaller fishes here as "invertebrate-eaters", since both prey on invertebrates. The in- vertebrate-eaters and planktivores (which prey upon invertebrate larvae as well as holoplankton) are the fishes most likely to influence benthic prey populations; hence, these taxa are emphasized in this report. After 6 months, cryptic in- vertebrates in half of the rubble from each of the A and C reefs were sampled quantitatively by sealing the rubble in plastic bags in situ. On shore, the rubble was chiselled into small pieces and sieved (0.7 mm mesh), retaining all of the resident cryptic biota. These samples were preserved in formalin and sorted, counted, and measured microscopically (see Reaka, 1981, 1983, for more details). RESULTS All fishes colonized both types of cinderblock reefs (A,B) quickly, while >2 months passed before fishes on the rubble reefs (C) reached peak abundances (lower graph, Fig. 1). On all 3 types of reefs, total numbers of individuals subsequently declined (and remained low throughout the winter; numbers increased again following spring recruitment; Reaka, 1981, 1983). Planktivores (mostly juvenile grunts, Haemulidae) were by far the most abundant guild, so their pattern of colonization followed that described above for total numbers of fishes (Fig. 1). Invertebrate- eating fishes were less abundant than planktivores. The number of invertebrate- eating fishes on reefs with rubble-dwelling invertebrates (A, C) did not peak until 2-5 months after establishment of the reefs. This period coincided with the time 70 FIGURE 1. Mean numbers of fishes observed on different types of reefs. All 5 guilds are included in total fishes. Species included in these guilds are listed in Appendix Table 1. The A reefs (open squares) are comprised of cinderblocks and rubble, the B reefs (closed squares) are built of cinderblocks only, and the C reefs (stars) are made of rubble only. For purposes of illustration, error bars are omitted here, but they are included in Reaka, 1981, 1983. INVERTEBRATE- EATERS July Aug Sep Oct Nov Dec Jan '80 CENSUS DATE '81 required for full colonization of the new rubble habitat by invertebrates (Reaka, 1981, 1983). In all guilds, fishes on the C reefs generally were smaller than those on the cinderblock reefs (Wolf, Bermingham, et a]_. , unpub. data) . The data in Figure 1 sug- gest that reefs with habitat for invertebrates (A) generally were characterized by higher numbers of fishes than reefs without habitat for inverte- brates (B). Specifically, the number of invertebrate-eaters per reef was significantly higher on A reefs than on B reefs in August and October, and the data showed a strong tendency in that direction in November (Mann Whitney U tests; p< 0.02, p<0.01 , and p<0.058, respectively). In addition, the total numbers of invertebrate- eaters per census on the 5 type A reefs were consistently higher than those recorded on the 5 type B reefs for the 6 month interval (Wilcoxon matched-pairs signed-ranks test, p <0.01). Population levels of inver- tebrates may be affected by invertebrate-eaters that prey upon reef residents, and by planktivorous fishes that eat invertebrate larvae swimming near or settling on the reef. The total numbers of inverte- brate-eating fishes per census on the 5 type A reefs were consistently higher than those on the 5 type C reefs through- out the study period (Wilcoxon matched-pairs signed-ranks test, p < 0.01). However, there was no significant difference in the numbers of planktivorous fishes per cen- sus on A vs. C reefs over the 6 month interval (Wilcoxon matched-pairs signed-ranks test, n.s.). The number of 71 invertebrate-eaters per reef was lower on the C than on the A reefs in the first two censuses (July 1980) (Mann Whitney U tests; p<0.02, p< 0.058, respectively). Although fewer planktivores were recorded on A than on C reefs in July and November, individual C reefs were populated by more small planktivores than were the A reefs in October 1980 (Mann Whitney U tests; p< 0.0001, p<0.03, p<0.05, respectively) . TABLE 1. Numbers of individuals of major invertebrate taxa collected per piece of rubble in control (natural reef) and experimental reefs (A and C reefs) in January 1981 (6 months after establishment of the experimental reefs). Results of a one-way classification analysis of variance (df = 2,37) are given under F value. Means and standard errors (in parentheses) are based on raw data. All data were tested for homogeneity of variance (Bartlett's test), and, if necessary, trans- formed by In (x+1). A superscript t indicates that test results are based on transformed data. Superscripts a and b indicate means that are significantly different by a Student Neuman Keuls test; means with the same subscript are not significantly different. Taxon Control A Reefs C Reefs F value Bartlett's Reef Wall (cinderblock (rubble only) (df=2,37) Test Value + rubble) Stomatopods 0.5(+.2)b 1.0(+.2) 1.8(+.3)a 3.31 p<0.05 4.29 N.S. sSps9 7.K+1.3) 6.3(+l.l) 8.3(+.9) 0.90 N.S. 1.26 N.S. (nSn-alpheidsS) 10.3(+1.5) 10.7(+2.0) 10.3(+1.6) 0.02 N.S. 2.30 N.S. Crabs 17.9(+2.0) 12.5(+2.4) 13.7(+2.0) 1.90 N.S. 0.69 N.S. Peracarids 73.3(+17.5) 62.7(+13.6) 87.3(+15.2) 0.74 N.S. 0.16 N.S. Sipunculans 18.5(+5.8) 17.8(+4.6) 11.7(+2.5) 0.80 N.S. 5.47 N.S. Polychaetes 180.0(+60.7)a 55.6(+12.2)b 76.9(+16.8)ab 4.54t pO.05 6.32tN.S. Ophiuroids 4.9(+0.7)a 2.3(+0.6)b 2.1(+0.4)b 9.53 p<0.05 2.89 N.S. Gastropods 6.2(+0.5) 6.5(+0.9) 6.5(+0.8) 1.28 N.S. 4.87 N.S. Bivalves 10.5(+2.1) 9.4(+1.9) 6.8(+0.9) 0.84t N.S. 1.30tN.S. Chitons 0.2(+0.1) 0.1(+0.1) 0.K+0.1) 0.10 N.S. 1.02 N.S. Examination of the invertebrate fauna in the rubble from A reefs (with fish predators) and C reefs (with fewer fish predators), and in naturally occurring rubble from the east canyon wall adjacent to the experimental reefs (with fish predators) revealed several patterns. Although the abundances of 8 of the 11 major revealed taxa did not differ in the 3 sets of rubble, stomatopods showed a significant increase in numbers on C reefs compared to either A reefs or rubble from the canyon wall (Table 1). The naturally occurring rubble from the wall con- tained species of stomatopods characteristic of shallow to moderate reef habitats (Gonodactylus oerstedii, G. spinulosus, Meiosquilla schmitti), while the experi- mental rubble (A and C reefs) was inhabited by different species (Gonodactylus sp, nov., Meiosquilla sp. nov. , M. tricarinata, Pseudosquil la ciliata). Several of the latter species are generalists, occurring in grassbeds as well as rubble 72 (M. tricarinata, P_. ciliata) , and others (particularly Gonodactylus sp. nov.) are characteristic inhabitants of our deeper (35-50 m) control and experimental reef sites. These opportunists reached higher densities in the new rubble habitat (particularly in the absence of predation) than the populations that normally in- habit rubble on the reef slope at this depth. There were no differences in the species of stomatopods found in the A vs. C experimental reefs. Numbers of poly- chaetes were significantly lower on the experimental reefs exposed to predation (A) than in the control rubble from the reef wall, and intermediate numbers of polychaetes were found in rubble from the C reefs. Ophiuroids also showed relatively low recruitment to the new habitat on the experimental reefs, but were equally abundant on A and C reefs (Table 1). DISCUSSION Does the presence of cryptic invertebrates in rubble influence colonization by fishes? Coral rubble harbors hundreds of invertebrates (Table 1), providing an abundant food source for some fishes. Many authors have argued that living space is more important than food in limiting (or structuring) populations of reef fishes (e.g., Sale, 1978; Smith, 1978). Although evidence of the importance of space comes from several sources (discussed in Sale, 1980), other studies show that space is not always limiting (e.g., Talbot, et^ al_. , 1978; Robertson, et al., 1981). Evidence that food directly influences numbers of fishes is limited TTsuda and Bryan, 1973). In the present study, however, the timing of colonization suggests that fish re- cruitment to reefs is related to the availability of benthic food. On those reefs with rubble-dwelling invertebrate fauna (A, C), the number of invertebrate-eaters peaked after 2-5 months, which coincides with the colonization rate of inverte- brates in coral rubble at this depth (Reaka, 1981, 1983, and Reaka, et aj_. , in prep.). In addition, cinderblock reefs with rubble had more invertebrate-eating fishes than cinderblock reefs without rubble. Rubble around the base of the A reefs gave the latter a slightly more complex structure than the B reefs, but whether or not this contributed to the observed differences in fish populations is unclear. Other variations in the structure of small artificial reefs (differences in the sizes of available holes) have not been related to number of fishes present or species composition (Molles, 1978; Talbot, et a]_. , 1978). Separation of the effects of food vs. habitat complexity in the present study would require an experiment comparing colonization of reefs with rubble initially containing a natural complement of invertebrates to that of reefs with sun-dried (defaunated) rubble. Does the presence of fish predators influence invertebrate colonization? Stomatopods appear to be strongly influenced by the presence of fish predators in this habitat (Table 1). After 6 months, these mantis shrimps were more abundant on the experimental reefs with fewer invertebrate-eaters (C) than on those with more and larger fish predators (A). This could not have been due to differences in location or habitat, since the positions of A and C reefs were alternated regularly down the canyon at equal distances from the reef wall. Except for octopuses (which in this habitat are very rare compared to the rubble fauna re- ported here), stomatopods are the largest and most active of the mobile cryptic fauna. Due to periodic movements on the surface of their rubble (Reaka, 1980b; Dominguez and Reaka, in review), stomatopods may be more exposed to predators than are many of the smaller, more secretive taxa. Ophiuroids had slow rates of colonization, but appeared unaffected by the fishes. Although recruitment of polychaetes to the new rubble also was slow, the effects of fish predators upon polychaete population levels are enigmatic. Invertivorous and planktivorous fishes possibly are responsible for decreased survivorship of settling polychaete larvae. 73 Cage experiments in shallow water (3 m) confirm that polychaete recruitment occurs slowly (probably via larval settlement), and that numbers of recruits are strongly decreased by exposure to fish predation compared to controls (Reaka, unpub. data). The remaining groups of major invertebrate taxa appear to be unaffected by fish predators. Finer taxonomic resolution of these taxa (in progress) may yield additional effects. It is also possible that a greater difference in predation pressures between the A and C reefs would have shown a more pronounced effect. At the moment, however, we conclude that the secretive habits of many of these cryptic invertebrates protects them from pronounced effects of predators upon their popu- lations at this study site. ACKNOWLEDGMENTS This work was supported by grants from the NOAA National Undersea Research Program (NA80AAA03715, NA81AAA01500) and the National Science Foundation (0CE-78- 26605) to M.L. Reaka. In addition, we are grateful to Frank Pecora, Joseph Landsteiner, Cheryl Van Zant, Jane Dominguez, David Moran, Hugh Reichardt, Joseph Dineen, and Miriam Smyth for their assistance as either aquanauts or members of the surface research team during our Hydrolab missions. We particularly thank the Hydrolab staff for their support and cooperation throughout the study. This is Contribution No. Ill from the West Indies Laboratory. LITERATURE CITED Bakus, G.J. 1966. Some relationships of fishes to benthic organisms on coral reefs. Nature 210: 280-284. Bakus, G.J. 1981. Chemical defense mechanisms on the Great Barrier Reef, Australia. Science 211: 497-499. Brock, R.E. 1979. An experimental study of the effect of grazing by parrotfishes and role of the refuges in benthic community structure. Mar. Biol. 51: 381-388. Buss, L.W., and J. B.C. Jackson. 1979. Competitive networks: nontransitive competitive relationships in cryptic coral reef environments. Amer. Nat. 113: 223-234. Clavijo, I.E., J. A. Yntema, and J.C. Ogden. 1980. An annotated list of the fishes of St. Croix, U.S. Virgin Islands. West Indies Laboratory, Fairleigh Dickinson University. Dayton, P.K. 1975. Experimental evaluation of ecological dominance in a rocky intertidal algal community. Ecol . Monogr. 45: 137-159. Hay, M.E. 1981. Herbivory, algal distribution and the maintenance of between- habitat diversity on a tropical fringing reef. Amer. Nat. 118: 520-540. Hixon, M.A. and W.N. Brostoff. 1983. Damselfish as keystone species in reverse: intermediate disturbance and diversity of reef algae. Science 220: 511-513. Jackson, J. B.C., and L.W. Buss. 1975. Allelopathy and spatial competition among coral reef invertebrates. Proc. Nat. Acad. Sci. 72: 5160-5163. Kaufman, L. 1977. The threespot damselfish: Effects on benthic biota of Caribbean coral reefs. Proc. Third Int. Coral Reef Symp. 1: 559-569. Menge, B.A., and J. P. Sutherland. 1976. Species diversity gradients: synthesis of roles of predation, competition, and temporal heterogeneity. Amer. Nat. 110: 351-369. Molles, M.C., Jr. 1978. Fish species diversity and natural reef patches: experimental insular biogeography. Ecol. Monogr. 48: 289-305. Neudecker, S. 1979. Effects of grazing and browsing fishes on the zonation of corals in Guam. Ecology 60: 666-672. Ogden, J.C, and P.S. Lobel . 1978. The role of herbivorous fishes and urchins in coral reef communities. Env . Biol. Fish. 3: 49-63. 74 Paine, R.T. 1966. Food web complexity and species diversity. Amer. Nat. 100: 65-75. Randall, J.E. 1967. Food habits of reef fishes of the West Indies. Stud. Trop. Oceanogr. 5: 665-847. Reaka, M.L. 1980a. Geographic range, life history patterns, and body size in a guild of coral -dwelling mantis shrimps. Evolution 34: 1019-1030. Reaka, M.L. 1980b. On learning and living in holes by mantis shrimp. Anim. Behav. 28: 111-115. Reaka, M.L. 1981. An experimental analysis of ecological processes that structure fish and invertebrate reef communities. NULS-1 Final Scientific Report. Reaka, M.L. 1983. Community interactions on natural and experimental reefs along a depth gradient. Final Scientific Report to the N0AA National Undersea Research Program. Reaka, M.L., and R.B. Manning. 1981. The behavior of stomatopod Crustacea, and its relationship to rates of evolution. J. Crust. Biol. 1: 309-327. Robertson, D.R., S.G. Hoffman, and J.M. Sheldon. 1981. Availability of space for the territorial Caribbean damselfish Eupomacentrus planifrons. Ecology 62: 1162-1169. Robins, C.R., R.M. Bailey, C.E. Bond, J.R. Brooker, E.A. Lachner, R.N. Lea, and W.B. Scott. 1980. A list of common and scientific names of fishes from the United States and Canada, 4th ed. Amer. Fish. Soc. Spec. Pub. No. 12: 1-174. Sale, P.F. 1978. Coeixtence of coral reef fishes--a lottery for living space. Env. Biol. Fish. 3: 85-102. Sale, P.F. 1980. The ecology of fishes on coral reefs. Oceanogr. Mar. Biol. Ann. Rev. 18: 367-421. Smith, C.L. 1978. Coral reef fish communities: a compromise view. Env. Biol. Fish. 3: 109-128. Talbot, F.H., B.C. Russell, and G.R.V. Anderson. 1978. Coral reef fish communities: unstable high-diversity systems? Ecol . Monogr. 48: 425-440. Tsuda, R.T., and P.G. Bryan. 1973. Food preferences of juvenile Siganus rostratus and S. spinus in Guam. Copeia 1973: 604-606. Vermel j, G.J. 1978. Biogeography and Adaptation. Harvard Univ. Press, Cambridge, Massachusetts. 332 pp. Wellington, G.M. 1982. Depth zonation of corals in the Gulf of Panama: control and facilitation by resident reef fishes. Ecol. Monogr. 52: 223-241. 75 +-> ** E r- • "=}" ro 1 oo t 1 1 — +->| CO re CD| T3 C « co ro CO 4- cd re 01 •r— 01 re -O S- •<- O a; i — (O ro +-> .re •r- ro +-> CJ T3 •i — *r- 2 4- E T- O CD +-> <_> S- "D C ro cd rO CO ■o s- CD cn CO i — i U QJ QJ I— +-> E Cd en ro ro UJ 0) O re > .re ^ CO cj .c i— i •I— *r- CO 4- _rc •i — Q.U- 4- O cu s- ai +-> • S- co o re 4- +-> o O •i — CO +-> CO -l-> ro 01 C > •i- CU S- s- E CD o re CO Cn CDX3 cd -i- o +-> CO ro CO 1 — O i— oo " 1 1 1 " o _l re co CO O CTl > i — h- re ro X c_> i— t • Q S- i — 1 z: o> CL. •■- +-"1 Q- CC fO o cn "O CD +-> +J O CL co 4- CD 3 CO O CD ro C O CD cn CL > co co CD co ro O Q. co 01 CD S_ CD -id CO ID cj co co •r- ro co "O S- ro CD 2 s- -o CD ro CL CD O.JZ: CO CD CO 4_ -£j ro -i- ro S- 2 CD 2 enje: > CO re CD C -C _re CD CO co •r- X ir- S- -I- =3 i — O CJ ro TO i — co i — CD -O -Q c— - J^. 3 O) ro 3 CD CD ro X3 CD ro co >> c O Q O h- c_) a' ro O i— i CD CO CO CO CO i/i CD QJ QJ QJ QJ I. S_ 5- S- S- CD QJ QJ QJ QJ o O O O O sz _c: -CT jr -£Z CJ CJ CJ CJ CJ ,— , — i — , — ro re ro re ro X X _C X IE <+- 4_ CD S_ r— CD •i- cn 4- cn -a Z re -t-> ro co x: •i— co 4- -i- S 4- O -^ o c 3 -O s- E -•-> o (j x: >>+J CD O £= O O E -C co S- CD CD co O a O- S- ro CO CD CD 3 cx ro 4- 5- QJ Cn 3 cn+-> ■i- CD S- > -l-J CO « CD LU +-> cs: co ^~ ro OO CQ •1 n (0 ro $~ +-> i— CU ro C +-> CO S- o aj S- +J cn 3 CD co >i CT 4- O •r— 4- S_ o 5- 3 Cl, 4J CL S- >, >> CQ CJ =C Di I— OO o a o 4-J CJ re aj Lj CO CL>— i QJ S- S- ' •i— co 3 3 CT O CO CO 3 « S_ UJ O o CO •r— 4- -C CD CO S_ •i- S- QJ CT S- CO S- •^ CD 3 r= cr-i- co CL co 4- cn cd c: 0) o s- <— co CD rc QJ > 3 ■l~3 CO ja CO co ro •r- JD QJ cn a: >, QJ cj -a o <: o c: +-> O •!- CO o 3 CD "O ro CT-C C JD CD CO ro O i — -i— co -(-> S- 4- ro CD CO 3 CD t- CO S- CO ro ro CJ _Q ro ro -Q CD ro co +-> n to LU 3 <: c Q ro i—i S- ■z: s- CO UJ 3 J-i-d a> CD >> co •» ro co S- 3 t_ QJ ro S- o re 3 S- cn QJ -re +-> +-> cj re ro re 3 +-> QJ S_ E s- cn o u_ +-> QJ «. E E 3 3 •)-> +J ro ro QJ OJ re • C •F— i — | o ■ o > s_ ro 3 i r ro 4- 4- I QJ 4- +-> -i- ro re -n _^ i _^ -re cj cn ro UJ CO ro ro i — co re O E •■- CD 3 E CJ S- 3 re •o O ro ro ■— UJ CO CO =C 3 Q -(-> I — I QJ ■Z. 3 UJ O" ct UJ CJ OO 76 -C to 0J -SZ tO -£Z cu ■i— to •r— to r— CZ -r- 4- •1 — •!— QJ 4- S- i- 4- c > s_ QJ 0J to i — QJ 3 O >, •r- CL •<— cu > •r-5 M rz T3 Q- E to 3 ro rz i — ra o E ••■-: CD s- cu > O -CI rz 5. n3 4- r— JZl 00 to to to _rz TZ> _rz cu •i- C JD o 00 •1- 1/1 (_) to t^ 3 cu cr 0J s_ 4- ra i — to S_ •»— i/i en cu rz jz: CO fO 1 JZl CJJ •1 — *l — C o 4- ro £Z S- -i- D- to LU JZ) ai >> ro E s 1 — cr S- •i- en 4- rz •r— or. .*: i — j*: a; +-> O o o o 3 T 4- o u O ■— cr> 3 S- S- <_> -SZ "O ^ 'i— s_ > 00 03 qj ro •i — O JZ: JZ> • 1— 0) 3 tO ro QJ QJ I — r QJ i— S- JC ja i— cj JD -SZ r— cl rz to > \— JZ jD <_> o 1 — l/l ^ to O QJ o .,— ^. to «* a> to cu cu to n ■i — ru " -o *> -£Z rz Z • (— ^ *> #i 00 >> qj to jc 3 Z3 to c s_ to rz • to QJ Z5 ro u 3 cu CL-r- #. > 1 i — 3 QJ 3 4-> » to -TZ "I— JZl ro +J Q. 4-> , — to rz QJ 3 CL QJ -Q •1 — •i — 00 ro •i — rz 1 i. c_> qj u CT >) CD to S- qj to i — 4- a> to ■i — S- n ro ■a ■a QJ QJ •i- ro to S- 3 CD i- ^ 3 CL +-> ) rO +-> Z3 £Z en ro £Z -(-> to <+- ro 4-> to E O. XT to 00 _Q O -Q to CL 5- C C L ID CT3 UJ i — to cu 3 CL to O XJ 0) to n •f— ra to +-> =3 ro jz to 3 CD ro >> =5 Z5 C.0 t- aj C ra QJ in p^ - +-> cu rO * JZ rz o S_ «. * (_) E s_ to iyi s- o rz jo ■1 — ZJ ro UJ to to •I— to LU +-> CO CU en CU LU CU LU +J ra l/l 3 QJ s. •■- E s- -Q uo > -i- (Zj -r- Q -i— ■I — QJ 3 rz ■1— SZ -M U JC E CTl o ro i—i Q. LU c ro i — i U LU S- c lu fz i— i rz i— i e E U ro +J (J CL +-> CU QC -r- < ra S- zc ro i > a > r— > h- s- S- E LU -a OJ , — E LU ro JZl LU o -1-J ^ >! 1 — 1 ro QJ z: s_ i — i U zj i— i z3 z: 3 s: x: _£Z O =£ C_J> < ZD C_J <_> I — I i — i i — i Cl O CtL <=t r"3 s: cr -r- to QJ ' * > o rz 4- cl rz C\J i — i o ••- o s_ IZL-i- • C_> S- JZ o QJ ro 4- Q. CO u 4-> rz -i<: 1— H T3 ra ro to o r< Cl 00 >, QJ JZl QJ ro ■o 1 •r— cu s. o s_ rz i— aj CO rz rz +-> en O JZl 3 LU o o « +-> rz or: •i — o to rs « -M - •f— O to 3 to on to 3 ro +j > C •> +-> 3 QJ 3 E ■— rz i—i to QJ ro ro r— JZ OJ o I— QJ o > +-> s_ t/> C_J r. QJ u on 00 to i — ■*-> ra •r— ro to to rz ^^^ LU to "O 3 3 u 4- rz s- • f— ro S> ro ro 4- CO ro Q. o QJ i u • — ^ JZl -Q ro Q_ CL ro (J i — i 00 to to ro O ro CL rz 3 LU ro 3 3 3 +-) on 00 ro rO JZl 1 QJ i — ^~ . — rz CQ to OJ QJ QJ to ■N to to to oo 1 — 1 i — c CL CL c±. QJ CO >-, l-H 3 3 Q ■Z. LU LU lu1 co i—i q; zz 1 1 ^ > s_ ro CD CD -Q en c -a > OO on c ■r- >1 3 3 OO -o re ro >> o LU s- on C r— CC "O en on on ro 4- o c ro O cn > ro 2: C O -^ ►— 1 oo E . ra -^ -C on u c_> on ^ *> Q. U ID S- O oo CD - -l-> 00 E ro E CD <-> 1 — 1 -C on ro =5 o •<-> "O ro D. on 3 4-> 4-> O- »c ai ■r~ 'r~ on sz ro S- •r— 3 a. • 4- "O • i — "O ro e o Q.-0 CD on 3 S_ C JO, on o Q- S- E on s_ -t-> ro S- cr 4- oo on on ro s- ro CJ on *» ai o 3 >> N CD JZ1 on s- Q..C CD +-> rC X ■r- -(-> on on J*. CD Q. 03 >, ro s_ ro i — c ro =3 3 U J3 ro E CD C o ^ •1 — CD •— i — ■ ro 3 £Z 4-> 3 E O " on CD QJ ■rn S- 00 00 4- •— -£z lu on -E _£Z 3 CD »• 4-> i—< 2: >, l—i Q. Q- i — i ro t — i 3 2z 1 q ca LU Q - LU < Z3 o s: 00 oo O — 1 ca 4- i — on >> CD ,4- E -i- E S- •— CD 4- ro O CD i — on "O cn oo -O CD -a CD E -C CD >•> S- ro Q. on -t-> -^ cn-a ■r- •,- 4-> on 3 ■— 4- O 3 ro ro tj cu a " — s "O CD O on • ^- (_> ro O CD > CD •^ (J ■i — C -(— 3 1— 1 £L c •i — i — £Z +-> on - C ca on 3 4-> •i— CD c CD on •i— CY. •i — Q_ 00 JD r— ro -C 3 4-> LU 4- O o ro J3 j — on i — C zc i — on o -i— +J •r~ ^-™ o CD S- 3 ^ r~ ro 4- 3 o on o ai ro 4-) S- QJ ~^ E -a • — > o 00 CD ro o 3 CD on 1 — X) on on on +-> »r— CTl CD 3 3 3 -Q SZ •i- c LU r. i- S- S- E cz S_ •!- 1 LU 4-> +-> +J O aj +-> -C ca X \— E E E ^ CJ> i—i OO LU Z2 i — + .— CO CO 0J- S- cn en ro CL >) o Q. -a >> QJ i- co o ■ ro > CD -r- - S- X3 U S- d) (1) ■o x: in CO t—t CO C Li_ LxJ •i — rr X" I— O u zr > i- LlJ >— t -5 i — i co CO a_- TJ z: UJ C =r I ro D£ "O CD 01 4-> >> +-> +-> CU -Q CD ro o c_ l_ O "O fO T3 OJ ■r— t_ QJ X: T3 ■D O 2 C e CU ro •^ OJ 1 CO CO u L- c CU O -C OJ r— "O • p- CO l_ ro c +-> • r- T3 <-. ro CJ M- O ro fimm C o L. a; CU + CU CO -M • = +-> E If- e 1 — C ro o O ro • « ■r- "O ■i— O c CO +J ■r— o -^ CU CU c >>-!-> ai CU ro ro 4-> O 3 X ■r— , — = ro 3 u a. L. >i o X ro OJ r— CO OJ +-> CU • CO c C > , — ro U o ■r- •1 — <0 QJ 4-> ■i — s_ -C CO t- ro CO • f— 4-> o o r— L_ CU i_ x: CU cu x: 3 +-> c 1_ > -(-> 4- C o o CU 't— 0) t_ T3 t_ -O +-> +-> +-> C O o ro c ro 0) CU ^^ ro x: sz CO -o a> CO +-> CU e +-J CU ■f— c ro U in 14- CU •^ CU +-> r— T3 L. c ro CU CU m CU o T3 +-> CO CU ■1 — CU C *. CU s. "O J= CU L. ■M +-> C CO E ro ro CU ■i — ro 3 , — •» — -O a o 3 <_) CO o u O CO 3 -o • r— CO c O • ■U CO o t_ >> >> (_ IX} •1— c_ r— r^ ro -M «r CU t- CX l_ o > o ■ r— ro ■ f— o C CU t_ • +J o_ »— < x: cu ■m <_) +-> +-> c O) CU c ro Q. i_ • T3 •f— -o CO ro CU C c CU CO ro en 3 L. •M c c -O ro CU OJ •(— <0 •» -C > ro M CO •l-J QJ cn 'i — CU +-> l_ !_ l_ CJ CO ro ro 4- ( — o CO 3 CU QJ +-> r— o CO X: CJ +-> +-> CO CU ro u CO -!-> • t— > t- ro 3 (_ >*- •i — OJ L L. ro 4-> ■<-> CU o sz "ai ro c +-> o CO cn r— c <-> 5 E CU 1 — cr o ro c: CU QJ *-> u_ "O ro CO CO c O QJ ■ 1— ro rD .C CO • i — ■o *-> QJ r — < t_ CU c c_ O > 1- <+_ 1 ai A-> i — O L_ L_ i — +-> CO aj =J t_ i— CU QJ rsl cn L. CO c E fO rm CU o • r- O [_ Lc t-> CLr CO cn 83 with grazing intensity on the Great Barrier Reef, being lowest within cages, intermediate within damselfish territories, and greatest outside territories. However, both Lobel (1980) and Hixon and Brostoff (in prep.), working in Hawaii, found considerably more blue-green algae inside than outside territories. These discrepancies suggest possible regional differences in local distributions of blue-green algae. In any event, herbivorous fishes, especially territorial damsel fishes, extensively affect reef algae in a variety of ways. SYNTHESIS Attempting to synthesize the above studies into a single conceptual framework can be done only at the realized risk of over-generalization and over- simplification. So be it. In general, fishes appear to strongly influence the community structure of reef algae, much more so than that of corals. This difference may be due to coral polyps and their surrounding calcareous skeletons being less available, palatable, and productive than many algae. Indeed, Randall (1974) has indicated that truly coral livorous fishes are among the most highly evolved of fishes, suggesting that this form of predation has appeared only recently in evolutionary time. However, many algae are inferior sources of nutrition (e.g., Montgomery and Gerking 1980), and chemical defenses in algae are being discovered at an increasingly rapid rate (e.g., Norris and Fenical 1982, Paul and Fenical 1983). In any event, while transient grazing fishes certainly control the distribution and abundance of many algae and some corals, the direct and indirect effects of territorial damsel fishes appear to strongly alter a variety of components of reef benthos. These fishes truly can be considered "keystone species" (sensu Paine 1969) where they are abundant (Williams 1980, Hixon and Brostoff 1983). I thus submit Figure 1 as a flowchart summarizing the general scheme of fish- benthos interactions on a "typical" coral reef where damsel fishes are common. Some of these interactions are well documented; others are not. This "synthesis" should therefore be considered a set of working hypotheses rather than a list of facts. All that can be stated unequivocally is that, first, fishes do indeed affect benthic community structure on tropical reefs, and second, more data on this important topic clearly are needed. ACKNOWLEDGMENTS Many thanks to F. Lynn Carpenter for her eleventh-hour review and Pam McDonald for her twelfth-hour typing of the manuscript. LITERATURE CITED Bak , R.P.M., & M.S. Engel . 1979. Distribution, abundance and survival of juvenile hermatypic corals (Scleractinia) and the importance of life history strategies in the parent coral community. Mar. Biol. 54:341-352. Birkeland, C. 1977. The importance of rate of biomass accumulation in early successional stages of benthic communities Proc. 3rd Int. Coral Reef Symp. 1:15-21. Birkeland, C. 1982. Terrestrial runoff as a planci (Echinodermata: Asteroidea). Mar. Borowitzka, M.A. 1981. Algae and grazing in N.S. 5:99-106. Brawley, S.H., & W.H. Adey. 1977. Territorial behavior of threespot damselfish (Eupomacentrus planifrons) increases reef algal biomass and productivity. Env. Biol. Fish. 2:45-51. 84 to the < survival of coral recruits. cause of outbreaks of Acanthaster Biol . 69 : 175-185. coral ree f ecosysti ?ms. Endeavour, Brock, R.E. 1979. An experimental study on the effects of grazing by parrotfishes and role of refuges in benthic community structure. Mar. Biol. 51:381-388. Connell, J.H. 1978. Diversity in tropical rain forests and coral reefs. Science 199:1302-1310. Day, R.W. 1977. Two contrasting effects of predation on species richness in coral reef habitats. Mar. Biol. 44:1-5. Day, R.W. 1983. Effects of benthic algae on sessile animals: observational evidence from coral reef habitats. Bull. Mar. Sci. 33:597-605. Harmel in-Vivien , M.L., & Y. Bouchon-Navaro. 1981. Trophic relationships among chaetodontid fishes in the Gulf of Aqaba (Red Sea). Proc. 4th Int. Coral Reef Symp. 2:537-544. Hatcher, B.G. 1981. The interaction between grazing organisms and the epilithic algal community of a coral reef: a quantitative assesssment. Proc. 4th Int. Coral Reef Symp. 2:515-524. Hatcher, B.G., & A.W.D. Larkum. 1983. An experimental analysis of factors controlling the standing crop of the epilithic algal community on a coral reef. J. Exp. Mar. Biol. Ecol . 69:61-84. Hay, M.E. 1981a. Spatial patterns of grazing intensity on a Caribbean barrier reef: herbivory and algal distribution. Aquat. Bot. 11:97-109. Hay, M.E. 1981b. Herbivory, algal distribution, and the maintenance of between- habitat diversity on a tropical fringing reef. Arner. Nat. 118:520-540. Hay, M.E., T. Colburn, & D. Downing. 1983. Spatial and temporal patterns in herbivory on a Caribbean fringing reef: the effects on plant distribution. Oecologia 58:299-308. Hixon, M.A., & W.N. Brostoff. 1981. Fish grazing and community structure of Hawaiian reef algae. Proc. 4th Int. Coral Reef Symp. 2:507-514. Hixon, M.A., & W.N. Brostoff. 1982. Differential fish grazing and benthic community structure on Hawaiian reefs, pp. 249-257. Jjk G.M. Cailliet & C.A. Simenstad (eds.), Fish Food Habit Studies. Univ. Wash. Sea Grant Progr. Hixon, M.A.., and W.N. Brostoff. 1983. Damselfish as keystone species in reverse: intermediate disturbance and diversity of reef algae. Science 220:511-513. Irvine, G.V. 1982. The importance of behavior in plant-herbivore interactions: a case study, pp. 240-248. In: G.M. Cailliet & C.A. Simenstad (eds.), Fish Food Habit Studies. Univ. Wash. Sea Grant Progr. Kaufman, L. 1977. The threespot damselfish: effects on benthic biota of Carib- bean coral reefs. Proc. 3rd Int. Coral Reef Symp. 1:559-564. Lassuy, D.R. 1980. Effects of "farming" behavior by Eupomacentrus lividus and Hemiglyphidodon plagiometopon on algal community structure. Bull. Mar. Sci. 30:304-312. Lawrence, J.M., & P.W. Sammarco. 1982. Effect of feeding on the environment: Echinoidea, pp. 499-519. J_n: M. Jangoux & J.M. Lawrence (eds.), Echinoderm Nutrition. Balkema Press. Lobel , P.S. 1980. Herbivory by damselfishes and their role in coral reef community ecology. Bull. Mar. Sci. 30:273-289. Lubchenco, J., & S.D. Gaines. 1981. A unified approach to marine plant-herbivore interactions. I. Populations and communities. Ann. Rev. Ecol. Syst. 12:405-437. Miller, A.C. 1982. Effects of differential fish grazing on the community structure of an intertidal reef flat at Enewetak Atoll, Marshall Islands. Pac. Sci. 36:467-482. Montgomery, W.L. 1980a. Comparative feeding ecology of two herbivorous damsel- fishes (Pomacentridae: Teleostei) from the Gulf of California, Mexico. J. Exp. Mar. Biol. Ecol. 47:9-24. 85 Montgomery, W.L.. 1980b. The impact of non-selective grazing by the giant blue damselfish, Microspathodon dorsalis, on algal communities in the Gulf of California, Mexico. Bull. Mar. Sci. 30:290-303. Montgomery, W.L., & S.D. Gerking. 1980. Marine macroalgae as foods for fishes: an evaluation of potential food quality. Env. Biol. Fish. 5:143-153. Nelson, S.G., & R.N. Tsutsui. 1981. Browsing by herbivorous reef fishes on the agarophyte Gracilaria edulis (Rhodophyta) at Guam, Mariana Islands. Proc. 4th Int. Coral Reef Symp. 2:503-506. Neudecker, S. 1979. Effects of grazing and browsing fishes on the zonation of corals in Guam. Ecology 60:666-672. Norn's, J.N., & W. Fenical. 1982. Chemical defense in tropical marine algae. Smith. Contr. Mar. Sci. 12:417-431. Ogden, J.C., & P.S. Lobel . 1978. The role of herbivorous fishes and urchins in coral reef communities. Env. Biol. Fish. 3:49-63. Paine, R.T. 1969. A note on trophic complexity and community stability. Amer. Nat. 103:91-93. Patton, W.K. 1976. Animal associates of living coral reefs, pp. 1-36. J_n: O.A. Jones & R. Endean (eds.), Biology and Geology of Coral Reefs, III, Biology. Academic Press. Paul, V.J., & W. Fenical. 1983. Isolation of hal imedatrial : chemical defense adaptation in the calcareous reef-building alga Halimeda. Science 221:747-749, Potts, D.C. 1977. Suppression of coral populations by filamentous algae within damselfish territories. J. Exp. Mar. Biol. Ecol . 38:207-216. Randall, J.E. 1961. Overgrazing of algae by herbivorous marine fishes. Ecology 42:812. Randall, J.E. 1974. The effect of fishes on coral reefs. Proc. 2nd Int. Coral Reef Symp. 1:159-166. Reese, E.S. 1977. Coevolution of corals and coral feeding fishes of the Family Chaetodontidae. Proc. 3rd Int. Coral Reef Symp. 1:267-274. Risk, M.J., & P.W. Sammarco. 1982. Bioerosion of corals and the influence of damselfish territoriality: a preliminary study. Oecologica 52:376-380. Sammarco, P.W. 1983. Effects of fish grazing and damselfish territoriality on coral reef algae. I. Algal community structure. Mar. Ecol. Prog. Ser. 13:1-14. Sammarco, P.W., & J.H. Carleton. 1981. Damselfish territoriality and coral community structure: reduced grazing, coral recruitment, and effects on coral spat. Proc. 4th Int. Coral Reef Symp. 2:525-535. Sammarco, P.W., & A.H. Williams. 1982. Damselfish territoriality: influence on Diadema distribution and implications for coral community structure. Mar. Ecol. Prog. Ser. 8:53-59. Sheppard, C.R.C. 1982. Coral populations on reef slopes and their major controls. Mar. Ecol. Prog. Ser. 7:83-115. Stephenson, W., & R.B. Searles. 1960. Experimental studies on the ecology of intertidal environments of Heron Island. I. Exclusion of fish from beach rock. Aust. J. Mar. Freshw. Res. 2:241-267. Van den Hoek , C, A.M. Breeman, R.P.M. Bak , & G. Van Buurt. 1978. The distri- bution of algae, corals and gorgonians in relation to depth, light attenua- tion, water movement and grazing pressure in the fringing coral reef of Curacao, Netherlands Antilles. Aquat. Bot. 5:1-46. Vine, P.J. 1974. Effects of algal grazing and aggressive behaviour of the f i shes Pomacentrus lividus and Acanthurus sohal on coral-reef, ecol ogy. Mar. Biol. 24:131-136. Wanders, J.B.W. 1977. The role of benthic algae in the shallow reef of Curacao (Netherlands Antilles). III. The significance of grazing. Aquat. Bot. 3:357-390. 86 Wellington, G.M. 1982. Depth zonation of corals in the Gulf of Panama: control and facilitation by resident reef fishes. Ecol . Monogr. 52:223- 241. Wilkinson, C.R., & P.W. Sammarco. 1983. Effects of fish grazing and damsel- fish territoriality on coral reef algae. II. Nitrogen fixation. Mar. Ecol . Prog. Ser. 13:15-19. Williams, A.H. 1980. The threespot damselfish: a noncarnivorous keystone species. Amer. Nat. 116:138-142. Williams, A.H. 1981. An analysis of competitive interactions in a patchy back-reef environment. Ecology 62:1107-1120. 87 CORAL RECRUITMENT AT MODERATE DEPTHS: THE INFLUENCE OF GRAZING H. Carl Fitz Institute of Ecology, University of Georgia, Athens, GA 30602 Marjorie L. Reaka Dept. of Zoology, University of Maryland, College Park, MD 20742 El dredge Bermingham Dept. of Genetics, University of Georgia, Athens, GA 30602 Nancy G. Wolf Dept. of Ecology and Systematics, Cornell University, Ithaca, NY 14853 ABSTRACT The effects of grazers on the biomass of algae and coral recruitment have been investigated extensively in shallow water, yet the dynamics of this interaction in deeper water have received, by comparison, relatively little attention. Fifteen cement artificial reefs were established at a depth of 20m in Salt River Canyon on the north coast of St. Croix. One third of the reefs were exposed to all grazing organisms, one third were protected from macrograzers by exclusion cages, and one third had partial cages (controls). After a year-long immersion, 267 corals of two genera were found, and comparisons were made between the three reef treatments. Numbers of newest coral recruits (3mm diameter and less) were similar on all types of reefs, indicating similar rates of settlement. However, caged reefs, with visi- bly greater algal biomass, had significantly fewer of the larger juveniles than those that were uncaged and exposed to (primarily fish) grazers. Though relatively few herbivores occur at these depths, herbivory nevertheless appears to indirectly control the survival of coral recruits, and hence determines the structure of coral communities on moderately deep reefs. INTRODUCTION Coral reefs and their primary structural component, scleractinian corals, have been the focus of intensive study, with numerous investigators analyzing some aspect of the ecology of the corals. Yet an understanding of the factors influencing coral distribution is far from complete. The present study attempts to elucidate the ef- fects of fish (and echinoid) grazing on the settlement and survival of scleractinian corals at intermediate (20m) depths. The effects of competition for space between algae and coral planulae or settled corals has received some attention (e.g.,0art 1972, Vine 1974, Kaufman 1977, Potts 1977, and others cited below), particularly in shallow water. Sammarco (1980) man- ipulated densities of the echinoid Diadema antillarum in shallow water, finding that ungrazed areas of high algal biomass (free from predation or disturbance by Diadema) allowed the highest rates of coral settlement. Subsequent survival, however, was highest in areas subject to moderate grazing pressures: competition from algae in ungrazed areas — and predation/disturbance in heavily grazed areas — reduced coral survivorship. Certain fishes also crop the algae and/or prey on small corals. Brock's (1979) microcosm study revealed low rates of coral recruitment in areas exposed to grazing scarids. Given adequate spatial refuges for the corals, however, Present address: Biology Department, Woods Hole Oceanographic Institution, Woods Hole, MA 02543. 89 the highest recruitment was then found in the presence of the fishes, which cropped the algal competitor without harming the corals. Also, damself ishes , maintaining territories that form algal mats, can have either positive or negative effects on corals, depending upon the location of the coral relative to the territory (see Vine 1974, Kaufman 1977, Wellington 1982). Observing the browsing activities of Caribbean acanthurid and scarid fishes, Birkeland (1977) found evidence that fishes avoided corals as small as 3mm; yet Bak and Engel (1979) saw no such avoidance. Moreover, Neudecker (1977, 1979) observed fishes selectively preying on corals transplanted to different areas on Pacific reefs. The extent to which urchins and fishes prey upon corals may depend to a great extent on the availability of the preferred food resource. Diadema, for instance, shift to feeding on corals as the preferred algal food is diminished through grazing (Carpenter 1981). The differing observations reported above may be a reflection of different grazer densities and/or algal biomass in the various study sites. Manipulative studies such as Sammarco's (1980) and Brock's (1979) indicate the existence of optimal levels of grazing for coral survival, with these grazer densities apparently being closer to those commonly found on the reefs. We sought to determine the influence that grazers naturally present at interme- diate depths have on populations of juvenile corals. As a large number of cement artificial reefs were constructed and censused, extensive underwater time was re- quired and the work would not have been feasible without the saturation diving facilities of NOAA's underwater habitat, NULS 1 (Hydrolab). The study area is off the north coast of St. Croix, US Virgin Islands, in Salt River Submarine Canyon. The canyon extends seaward from a small estuary, with the canyon walls being variously covered by hard corals and coral rubble, sponges, gor- gonians, etc. (see Rogers et aj. 1983). The artificial reefs were established on the sandy floor of the canyon, well removed from the reef complex of the walls. This is an area of shifting sand uncolonized by benthic macroflora; during periods of heavy swells on this exposed coastline, sand scouring around the reefs is evi- dent. METHODS Artificial reefs were established during NOAA Hydrolab Mission #81-1 in January 1981. Each reef structure was identical, consisting of 11 cement cinder blocks stacked and lashed together in three layers of 6, 4, and 1 block-per layer. Total censused surface area of each reef was 0.93m vertical and 0.46m horizontal sub- strate. Corals inside the holes of the cinder blocks were not censused. Cage enclosures were composed of 2.5 X 5cm wire mesh over a 1.5 X 3 X 1.5m frame. A total of fifteen reefs were constructed at a depth of 20m on the canyon floor along a line a constant 15m distance from the junction of the sandy floor and the east coral wall. The reefs were divided equally into three treatment types, with each reef along the line being an alternate treatment. Uncaged reefs were exposed completely to fishes and invertebrates; caged reefs were totally enclosed by cages; and cage controls had cages with two sides and the top partly-open, allowing access by macrofauna. In January 1982 during Hydrolab Mission #82-1, all of the reefs were censused j_n situ for juvenile scleractinian corals. Maps of each reef were employed to record the location of the corals, its identification to genus and sometimes species, and its length and width (± 0.5-lmm). Deterioration of all the cages during storms in February 1982 did not allow repeat censuses of the corals on the caged reefs. After initial establishment of the reefs, periodic fish censuses were performed to quantify the number of resident fishes on the reefs. Transient fishes (including herbivorous fishes around the reefs) were not quantified. Numbers of Diadema antil- larum, being low, were recorded during only one census in February 1982. 90 Analyses of variance (ANOVA), both completely randomized and randomized block design, and Student's t-tests were employed for statistical analysis as was ap- propriate. Where the ANOVA indicated significant differences (p<0.05), Tukey's Stu- dentized Range Test was used to determine which cells differed significantly (p<0.05). RESULTS Cage effects: corals After the year- long reefs, all being on verti ites and Agaricia formed Difficulty with j_n situ i meter prevented assigning Agaricia sp. , approximate were P. astreoides. A Student's t-test ind vs. Agaricia on any reef Substantially more corals and 90 respectively) tha showed that the mean numb different, nor was the me immersion, a total of 267 corals were found on the fifteen cal edges. This included only two genera of corals: Por- 19% and 8!L% of the total number of corals, respectively, dentif ication of corals less than approximately 6mm in dia- species to the smallest corals. However, of the larger ly 2/3 were A. agaricites and 1/3 A. lamarki . All Porites icated no significant differences in the numbers of Porites treatment; thus all further analyses pooled both genera. settled on uncaged and cage control reefs (totals of 123 n on caged reefs (54 total). However, a one-way ANOVA er of corals on the three reef types was not significantly an diameter (Table 1). Table 1. Number of corals on each reef; total number and mean for each treatment type; mean diameter of corals for each treatment type. Numbers in parentheses are standard deviations; n.s. = not significant (p<0.05). Caged Uncaged Cage Control Rep.#l Rep. #2 Rep. #3 Rep. #4 Rep. #5 Total Mean number 10 19 9 1 15 12 26 41 34 10 54 10.8 (±6.8) n. s. 123 24.6 (±13.5) n. s. 14 28 20 10 18 90 18.0 (±6.8) Mean diameter (mm) 5.3 (±3.0) n. s. 6.8 (±3.0) n. s. 5.4 (±2.6) 60 -i 50 - CO 40 s- o u 30 ■4- O i- ^0 cu E 10 - \ \ 1 \ 1 1 \ - ' 1 , V 1 i \l caged ) control 22 uncaged // <3 1 > 3-6 1 > 6-9 1 >9-l 2 1 >12 1 Figure 1 Diameter (mm) Total number of corals in each size class for each treatment type. 91 Figure 2. A. Open artificial reefs built of cinder blocks as described in the text. B. Cage control. C. Caged artificial reef. D. Closer view of caged artificial reef, showing extensive algal growth about 3 months after initiation of the experiment. E. Close-up view of fleshy algal growth on the edge of a cinder block in a caged reef. 92 93 Though no overall differences in numbers of corals existed between types of reefs, the size (age) structure of corals on the reef types was dissimilar. Corals on each type of reef were divided into five size classes at 3mm intervals. The total number of corals in each size class for the combined five reefs of each type are depicted in Figure 1; the general trend of similar numbers of newest (^3mm) recruits and differing numbers of older corals between types of reefs is evident. A two-way ANOVA indicated a significant effect of type of reef treatment (blocked over size classes) on the mean number of corals per reef. With this knowlege, a one-way ANOVA on each size class was used to determine which size class(es) had different mean numbers of corals between the reef types. The only significant dif- ference was found in the >6-9mm class, with a Tukey's Studentized Range Test in- dicating that the caged reefs had significantly fewer corals than the uncaged reefs (Table 2). Shading by cages was insignificant, as the cage control did not differ from the uncaged reefs. Table 2. Mean number of corals in each size class for each treatment type. Five replicates in each cell, numbers in parentheses are standard deviations, asterisk indicates significant difference (p<0.05). Caged Uncaged Cage control ^3mm 3.8(±2.6) 2.6(±2.0) 4.4(±4.0) >3-6mm 3.8(±2.6) 10.6(±7.2) 8.6(±4.6) >6-9mm 1.8(±1.9) * 7.0(±3.7) 3.2(±2.5) >9-12mm 1.2(±1.3) 3.4(±4.0) 1.6(±1.1) >12mm 0.0(±0.5) 1.0(±1.2) 0.0(±0.5) Grazing fishes and invertebrates Diadema anti 1 larum was an uncommon echinoid on these reefs, well removed from the hard substratum of the East Wall of the canyon. One complete census showed an occasional Diadema on the ten uncaged and cage control reefs for a mean of 0.4 per reef. Total available surface area (including 12X12cm holes uncensused for corals) was 3.4m . The herbivorous fishes around the reefs were acanthurids, scarids and pomacen- trids. Bicolor damselfish (Eupomacentrus partitus) were common residents on the reefs, along with occasional juvenile blue tangs (Acanthurus coeruleus). Schools of roaming doctorfish (A. chirurgus) and ocean surgeons (A. banianus) were common but, due to their transient nature, difficult to quantify (see Wolf et aj. 1983 for the results of the fish censuses). Princess parrotfish (Scarus taeniopterus) were seen occasionally, again roaming about between different reefs. The cages effectively excluded these larger macrograzers , as evidenced by the very noticable difference in cover of filamentous algae. The algal 'turf on the exposed reefs was cropped extremely short, whereas thick, long tufts of filamentous algae and some fleshy algae were present on the completely caged reefs (see Figure 2). DISCUSSION Recruitment to these reefs, as evidenced by numbers of the smallest corals, was similar in both ungrazed and grazed situations. Though algal biomass was visibly greater on ungrazed reefs, space for coral settlement was apparently adequate in either type of habitat. Large differences in numbers of >6~9mm corals between caged and uncaged reefs indicate that coral survivorship had been greater in areas subject to some grazing pressure. Thus, any incidental predation on corals by the fishes (and echinoids) was more than offset by the reduction in competition with the algae. Though differ- ences between the types of reefs are suggestive in the >3-6mm class, the lack of 94 statistical significance is indicative of the time required for competitive effects to become evident and of the patchy nature of recruitment. The very small sample sizes of corals larger than 9mm prevents thorough statistical treatment of differ- ences in corals of this size. It seems unlikely that the trends toward greater survival in the corals of intermediate size on the exposed reefs would cease to be significant statistically as the community matures. For Agaricia spp. and Porites astreoides, Fitz (unpublished data) measured mean linear growth rates of 0.6 (±0.2)mm/mo for corals 5-10mm diameter, 0.9 (±0.3)mm/mo for corals 10-15mm diameter, and 1.3 (±0.3)mm/mo for corals 15-20mm diameter (n=45, 41 and 17, respectively). Few corals would be expected, therefore, to attain sizes greater than 9mm in the year-long immersion period. Some manipulations of the coral reef community are feasible and provide insights into the processes shaping that community. This caging experiment implies that the herbivores present at moderate depths play a crucial role in the survival of juve- nile corals. As has been shown in shallow water areas (e.g., Sammarco 1980), herbivory can reduce the deleterious effects of the algae on the corals by reducing the biomass of the algal competitor. Though not as abundant in deeper as in shallow waters, herbivores nevertheless appear to control the survival of coral recruits and determine the structure of this epibenthic sessile community. ACKNOWLEDGEMENTS Special appreciation goes to Jane Dominquez for many hours of cold saturation diving, and to the very competent surface support staff of the NOAA Hydrolab oper- ation. We also thank Frank Pecora, Cheryl Van Zant, Hugh Reichardt, Miriam Smyth, and Joseph Dinneen, who assisted in building the reefs either as aquanauts or as part of the scientific surface support team. This work was made possible by grants from NOAA (NA81AAA01500, NA82AAA01286) and NSF (0CE-78-26605) to Marjorie L. Reaka. This is Contribution No. 110 from the West Indies Laboratory. LITERATURE CITED Bak, R.P.M. , and M.S. Engel. 1979. Distribution, abundance and survival of ju- venile hermatypic corals (Scleractinia) and the importance of life history strategies in the parent coral community. Mar. Biol. 54:341-355. Birkeland, C. 1977. The importance of rate of biamass accumulation in early succes- sional stages of benthic communities to the survival of coral recruits. Proc. 3rd Int. Symp. Coral Reefs, Miami 1:15-21. Brock, R.E. 1979. An experimental study on the effects of grazing by parrotfishes and the role of refuges in benthic community structure. Mar. Biol. 51:381-388. Carpenter, R.C. 1981. Grazing by Diadema antillarum (Philippi) and its effects on the benthic algal community, j. Mar. Res. 39:749-765. Dart, J.K.G. 1972. Echiniods, algal lawn and coral recolonisation. Nature 239:50-51. Kaufman, L. 1977. The three spot damselfish: effects on benthic biota of Carib- bean coral reefs. Proc. 3rd Int. Symp. Coral Reefs, Miami 1:559-564. Neudecker, S. 1977. Transplant experiments to test the effect of fish grazing on coral distribution. Proc. 3rd Int. Symp. Coral Reefs, Miami 1:317-323. Neudecker, S. 1979. Effects of grazing and browsing fishes on the zonation of corals in Guam. Ecology 60: 666-672. Potts, D.C. 1977. Suppression of coral populations by filamentous algae within damselfish territories. J. Exp. Mar. Biol. Ecol. 28:207-216. Rogers, C.S., H.C. Fitz, M. Gilnack, J. Beets, and J. Hardin. 1983. Coral re- cruitment patterns at Salt River submarine canyon St. Croix USVI. NOAA Hydro- lab Final Scientific Report; Missions 81-3 & 82-8. 95 Sammarco, P.W. 1980. Diadema and its relationship to coral spat mortality: grazing, competition and biological disturbance. J. Exp. Mar. Biol. Ecol. 45:245-272. Vine, P.J. 1974. Effects of algal grazing and aggressive behavior of the fishes Pomacentrus lividus and Acanthurus sohal on coral reef ecology. Mar. Biol. 24:131-136. Wellington, G.M. 1982. Depth zonation of corals in the Gulf of Panama: control and facilitation by resident reef fishes. Ecol. Monogr. 52: 223-241. Wolf, N. G., E. B. Bermingham, and M. L. Reaka. 1983. Relationships between fishes and mobile benthic invertebrates on coral reefs, p. 69-78. In: M. L. Reaka (ed.), The Ecology of Deep and Shallow Coral Reefs. Symposia Series for Undersea Research, Vol. 1. Office of Undersea Research, NOAA, Rockville, Md. 96 BETWEEN-HABITAT DIFFERENCES IN HERBIVORE IMPACT ON CARIBBEAN CORAL REEFS Mark E. Hay University of North Carolina at Chapel Hill Institute of Marine Sciences, Morehead City, North Carolina 28557 Tim Goertemiller Marine Systems Laboratory, NHB W-310 Smithsonian Institution, Washington, DC 20560 ABSTRACT Transplanted sections of the seagrass Thalassia testudinum were used as a bioassay to assess between-habitat differences in herbivory on three Caribbean reefs. Consumption of Thalassia by herbivorous fishes on shallow (1-10 m) reef slopes was significantly higher than on deep (30-40 m) reef slopes or on shallow reef flats. Seaweeds typical of reef flat habitats were rapidly consumed when placed on shallow reef slopes. Seaweeds typical of either deep or shallow reef slopes were relatively resistant to herbivory and a high proportion of these species are known to contain secondary chemical compounds that appear to deter herbivorous fishes. Shallow reef flats provide seaweeds with a predictable spatial escape from major reef herbivores; algae characteristic of these habitats have evolved few, if any, characteristics that significantly reduce losses to herbivory. INTRODUCTION Although herbivory plays a major role in determining the distribution and abundance of seaweeds on coral reefs (Stephenson and Searles 1960; Randall 1961, 1965; Ogden et al_. 1973; Hay 1981a, b; Luchenco and Gaines 1981; Hay et al . 1983), few studies have addressed spatial variations in herbivory or the relative susceptibility to herbivory of seaweeds from different habitats. Recent investigations have focused on between- and wi thin-habitat variations in herbivory on individual reefs (Hay 1981a, b; Hay et a]_. 1983) and on changes in herbivory that occur over depth gradients on several reefs scattered throughout the Caribbean (Hay 1984). In this study we use transplanted sections of the seagrass Thalassia testudinum as a bioassay for herbivore activity in different habitats [reef flats, shallow (1-10 m) reef slopes, and deep (30-40 m) reef slopes] on 3 Caribbean reefs. We also transplant seaweeds from each of these habitats into areas with high herbivore activity in order to assess their relative susceptibility to herbivory. METHODS Thalassia was chosen as the bioassay organism because it is eaten in the field by both herbivorous fishes and urchins (Randall 1965, Ogden et _aJL 1973, Ogden 1976) and it is readily available on most reefs. Freshly collected sections of Thalassia were fastened in wooden clothespins; the latter were attached to small coral fragments and distributed haphazardly within the habitats where herbivory was to be measured. All Thalassia sections were 5 cm long and separated by a distance of 1-3 m when placed in the field. At the end of each 97 bioassay, removal of Thalassia was quantified by measuring the remaining length of each blade to the nearest .5 cm. During all tests, the clothespin and coral apparatus was positioned so that it would be equally approachable by both fishes and urchins. However, on most sections of these reefs, urchins are relatively uncommon and almost all Thalassia removal was due to grazing by fishes (Hay 1984). Removal of Thalassia on reef slopes varies with depth; portions of this pattern have been extensively analyzed elsewhere (Hay et a]_. 1983, Hay 1984). In this paper we compare herbivory on (1) reef flats that are exposed at lowest tides, (2) shallow (1-10 m) portions of reef slopes, and (3) deeper (30-40 m) portions of reef slopes. Seaweeds from each of these habitats were exposed to herbivorous reef fishes by placing small (3-4 cm long) pieces of each seaweed in a 3-stranded rope that was fastened to the reef slope at a depth of 1-5 m. Ten to 37 individuals of each test species were used at each location (for a description of each site, see Hay 1984). Seaweeds within a length of rope were separated from one another by a distance of about 7 cm. Thus, when an herbivore encountered a rope, all species of seaweed should have been equally apparent and available. At the end of an experiment, each species on each rope was recorded as either still present or totally eaten. Ropes were only placed in the field under completely calm conditions and were shaken to be sure that all individuals were securely attached. On the reef at Becerro, Honduras, where the feeding trial was of short duration (1.75 h), we were able to directly observe the ropes for most of the test period; no individuals were lost to any source other than herbivory. For the feeding trials of longer duration (19-24 h), we cannot absolutely rule out the possibility that some individuals were lost to breakage. However, the magnitude of such loss would have to be very small given the calm conditions and our inability to observe breakage during any of the observation periods. Assignment of seaweeds as characteristic of reef flat, shallow reef slope, or deep reef slope habitats was based on qualitative observations at each study site. For example, species that were common between 30 and 40 m deep and present but rare at 10 m deep were listed as characteristic of the deep reef slope. RESULTS Consumption rates for sections of Thalassia placed on shallow reef slopes were significantly higher than consumption rates on either reef flats or deeper sections of reef slopes (p<.05, AN0VA and Student Newman-Keuls Test) (fig. 1). A deep reef slope comparison could not be done at Becerro, Honduras, since the reef slope extended to a depth of only 9m. On all 3 reefs, all daytime removal of Thalassia was attributable to fishes, as evidenced by their crescent shaped feeding scars (see Hay et a]_. 1983, Hay 1984). On the two reefs where comparisons were made, removal of Thalassia on the deep reef slope was significantly higher than removal on the reef flat (fig. 1). However, the magnitude of this difference was small when compared to differences between the shallow reef slope and either of the other habitats. When seaweeds were transplanted onto shallow reef slopes, those from reef flats were consumed rapidly while those from either deep or shallow reef slopes were consumed slowly if at all (fig. 2). The one exception to this pattern was Padina sp. at Becerro, Honduras (fig. 2C). This reef flat species was not eaten when placed on the shallow reef slope, although Padina sanctae-crucis was rapidly consumed at Carrie Bow and Lighthouse reefs (figs. 2A, 2B). 98 (29) Reef Flot Reef Slope! I -IOm) Reef Slope (30-40m) (39) (30) CARRIE BOW BELIZE LIGHTHOUSE BELIZE BECERRO HONDURAS ® CARRIE BOW, BELIZE (24h) i pjpiNa sanctae-cbuos 2 ACANTHOPHOBA SPEClFERA 3 LAURENCIA INTRICATA 4 HALIMEDA TUNA 5 LIAGORA Sp. 6 SABGASSUM POLfCEBATIUM 7 5TYP0P0DIUM ZONALE 8 CRYPTONEMIA CBENULATA 9 HALIMEDA OISCOIDEA 10 HALIMEDA GOBEAUII 11 BHIPOCEPHALUS PHOENIX 12 L080PH0BA VABIEGATA D LIGHTHOUSE, BELIZE (I9h) 1 LAUBENCIA PAPILLOSA 2 PADINA SANCTAE-CBUCIS 3 SABGASSUM POLrCEBATIUM 4 DIGENIA SIMPLEX 5 TUBBINABIA TBICOSTATA 6 L080PH0BA VABIEGATA 7 HALIMEDA COPIOSA 8 GELIDIELLA ACEBOSA 9 HALIMEDA TUNA 10 HALIMEDA GOBEAUII 11 HALIMEDA OPUNTIA 12 BHIPOCEPHALUS PHOENIX INDIVIDUALS EATEN (%) ) 20 40 60 80 100 SIGNIFICANTLY DIFFERENT FROM 1125) (191 1(10) ^2(ii) E3(iO) B3(37) 8(20) BliOl 0i2O) 3(20) H(i5) 3(19) 20 40 60 80 100 © BECERRO, HONDURAS (l.75h) 0 20 40 60 BO I ACANTHOPHOBA SPICIFEBA 100 LAUBENCIA PAPILLOSA TUBBINABIA TRICOSTATA HALIMEOA OISCOIDEA HALIMEDA TUNA PADINA Sp BHIPOCEPHALUS OBLONGUS STTPOPODIUM 20NALE 4-12 4-12 4-12 -3 -3 -3 -3 -3 -3 -3 -3 -3 6- 6 6 8 8 1-3 1-3 1-5 1-5 I 12 12 12 12 12 11,12 11,12 11,12 12 5 8 ■9 REEF FLAT REEF SLOPE (I- IOm) , 2 . 2 - 3 - 3 -3 1REEF SLOPE (30-40m) Fig. 1. The mean % of Thalassia eaten in different habitats on the 3 study reefs. Vertical lines represent + 2 standard errors. Numbers in parenthe- ses = N. At each reef, all differences are significant (p < .05, ANOVA and Newman- Keuls Test). Fig. 2. Percentage of individuals eaten when placed on the shallow reef slope. For Carrie Bow, numbers in parentheses = N. Significant differences (p < .05) were evaluated using Contingency Table Analysis or Fishers Exact Test (if cell sizes were small ) . Susceptibility to herbivory of seaweeds from the shallow versus the deep reef slopes showed no consistent pattern. At Carrie Bow (fig. 2A), there were no significant differences between seaweeds from deep and shallow areas of the reef slope (p < .05, Fishers Exact Test). At Lighthouse (fig. 2B), there were some differences between species but these revealed no consistent between- habitat patterns. For a given algal species, susceptibility to herbivory showed similar patterns on different reefs (fig. 2). There were, however, a few interesting exceptions. Acanthophora specifera, Laurencia papillosa, and Padina sanctae- crucis were consumed rapidly on each reef where they were tested. Species of Halimeda and Rhipocephalus were consistently resistant to herbivory, as was Stypopodium zonale. Turbinaria tricostata was of intermediate preference. Sargassum polyceratium from the reef flat on Lighthouse was very susceptible to herbivory (fig. 2B); S^ polyceratium from the deep reef slope on Carrie Bow was very resistant (fig. 2A). 99 DISCUSSION On the scale used in this study, between-habitat differences in herbivory are shown to be consistent on 3 different reefs (fig. 1). Rates of macrophyte removal on reef flats or on deep reef slopes are significantly reduced relative to removal rates on shallow reef slopes. In addition to the reefs studied here, shallow reef flats also have been shown to function as spatial escapes from herbivory in the Virgin Islands (Adey and Vassar 1975, Steneck and Adey 1976), the Netherlands Antilles (van den Hoek et aj_. 1978), and Panama (Hay 1981c, Hay et a]_. 1983). Decreased herbivory on deep reef slopes has been hypothesize'd- to provide an explanation for the increased algal abundance that occurs at depth on some reefs (van den Hoek et aj_. 1978), and the rate of macrophyte removal recently has been shown to decrease with depth on a wide variety of undisturbed Caribbean reefs (Hay 1984). Reef flats and deep reef slopes usually are characterized by reduced topographic complexity; in areas where predatory fishes are abundant, these more simplified habitats may be avoided by herviborous fishes because they offer few places to hide when attacked by predators. On shallow reef slopes, herbivorous fishes concentrate their grazing in areas of greater topographic complexity (Hay £t jajL 1983) and, on heavily fished reefs where predatory fishes are relatively rare, herbivorous fishes make increased use of deeper reef slopes (Hay 1984). All of these patterns suggest that the probability of being preyed upon may play a significant role in determining the spatial pattern of foraging by herbivorous fishes. Previous studies have suggested that the evolution of herbivore resistance in seaweeds involves costs that result in decreased growth rates and decreased competitive ability in the absence of herbivores (Lubchenco 1980; Lubchenco and Gaines 1981; Hay 1981a, c; Hay et jH . 1983). The data presented in figure 2 provide a partial test of this hypothesis; if characteristics that promote herbivore resistance mandate costly tradeoffs, then herbivore resistance should not evolve in species that occur primarily in habitats subject to low rates of herbivory. Patterns exhibited by reef flat seaweeds support the hypothesis; they are subject to low rates of herbivory (fig. 1) and exhibit little resistance when exposure to herbivores is increased (fig. 2). Rates of Thalassia removal on deep (30-40 m) reef slopes were significantly higher than on reef flats, but the magnitude of difference was not large--17.5% versus 9.3% at Carrie Bow, and 12.8% versus 1% on Ligvhthouse (fig. 1). However, differences in herbivore resistance of species from these habitats were striking (fig. 2). Despite the low rate of removal of Thalassia that was documented on deeper sections of the reef slope, seaweeds from these deeper areas were very resistant to herbivory. Even though the Thalassia bioassay shows herbivory to be relatively low in both reef-flat and deep reef-slope habitats, herbivore resistance appears to have been selected for on the deep reef slope and selected against on the shallow reef flat. This apparent paradox can be explained if one considers rate of biomass removal by herbivores (i.e., the Thalassia bioassay) relative to rate of production through photosynthesis. Seaweeds in shallow waters may grow many times faster than seaweeds in deeper waters (Hay 1981a, b). The low rate of biomass removal that occurs on reef flats can rapidly be replaced by photosynthesis since light is plentiful and turbulence prohibits the formation of large diffusion gradients that would slow nutrient acquisition (Doty 1971). Since production of seaweed biomass is very slow on deeper reef areas, even small losses to herbivores may exceed gains and thus select for increased herbivore resistance. As an example, if herbivores removed equivalent amounts of plant material from deep and shallow sites, selective 100 pressure for the evolution of grazer deterrents would be much greater in the deeper habitats since losses would be a larger proportion of net growth and take longer to replace. Future studies on spatial patterns in herbivory should attempt to quantify herbivore impact as a proportion of plant production within each habitat studied. Most of the reef slope seaweeds are known to have naturally occurring chemical substances that appear to serve as defenses against herbivory (Fenical 1975, Norris and Fenical 1982). Many species in the genus Halimeda produce diterpenoid trialdehydes (Paul and Fenical 1983), Rhipocephalus contains similar compounds (Norris and Fenical 1982), Stypopodium zonale contains several related C27 compounds derived from a mixed biosynthesis of diterpenoid and acetate precursors (Gerwick and Fenical 1981), and Liagora produces an unusual acetylene containing lipid (Norris and Fenical 1982). These compounds are toxic to or deter feeding in reef fishes, and some even stop cell division in fertilized sea urchin eggs or motility in sea urchin sperm (Norris and Fenical 1982, Paul and Fenical 1983). The polyphenol ic compounds produced by Turginaria and Sargassum (Norris and Fenical 1982) do not appear to be especially effective, as evidenced by the feeding data in figure 2. The difference in susceptibility of Sargassum polyceratium from the deep reef slope at Carrie Bow and the reef flat at Lighthouse could result from population differences in defensive compounds or from between-reef differences in herbivorous fishes. In general, it appears that herbivores consume a significant proportion of reef slope production and that this has resulted in strong selection for herbivore deterrents in seaweeds from this habitat. Reef flats provide predictable escapes from herbivory, and seaweeds from these habitats are characterized by very little resistance to herbivory. ACKNOWLEDGMENTS This study was supported by a post-doctoral fellowship from the Smithsonian Institution and by the Marine Systems Lab of the Smithsonian Institution. LITERATURE CITED Adey, W. H., & J. M. Vassar. 1975. Colonization, succession and growth rates of tropical crustose coralline algae (Rhodophyta, Cryptonemiates) . Phycologia 14: 55-69. Doty, M. S. 1971. Physical factors in the production of tropical benthic marine algae, p. 1-25. I_n: J. D. Costlow (ed.), Fertility of the Sea. Gordon & Breach Science Publishers, New York, N. Y. Fenical, W. 1975. Halogenation in the Rhodophyta: a reveiw. J. Phycol . 11: 245-259. Gerwick, W. H., & W. Fenical. 1981. Ichthyotoxic and cytotoxic metabolites of the brown alga, Stypopodium zonale. J. Org. Chem. 46: 22-27. Hay, M. E. 1981a. Herbivory, algal distribution and the maintenance of between- habitat diversity on a tropical fringing reef. Amer. Nat. 118: 520-540. Hay, M. E. 1981b. Spatial patterns of grazing intensity on a Caribbean barrier reef: herbivory and algal distribution. Aquat. Bot. 11: 97-109. Hay, M. E. 1981c. The functional morphology of turf-forming seaweeds: persistence in stressful marine habitats. Ecology 62: 739-750. Hay, M. E. 1984. Patterns of fish and urchin grazing on Caribbean coral reefs: are previous results typical? Ecology 65(2): in press. Hay, M. E., T. Colburn, & D. Downing. 1983. Spatial and temporal patterns in herbivory on a Caribbean fringing reef: the effects on plant distribution. Oecologia 58: 299-308. 101 Lubchenco, J. 1980. Algal zonation in the New England rocky inter-tidal community: an experimental analysis. Ecology 61: 333-344. Lubchenco, J., & S. D. Gaines. 1981. A unified approach to marine plant- herbivore interactions. I. Populations and communities. Ann. Rev. Ecol. Syst. 12: 405-437. Norris, J. N., & W. Fenical. 1982. Chemical defense in tropical marine algae, p. 417-431. Jn_: K. Rutzler & I. G. Macintyre (eds.), The Atlantic Barrier Reef Ecosystem at Carrie Bow Cay, Belize. I. Structure and communities. Smithsonian Institution Press, Washington, DC. Ogden, J. C. 1976. Some aspects of herbivore-plant relationships on Caribbean reefs and seagrass beds. Aquat. Bot. 2: 103-116. Ogden, J. C, R. A. Brown, & N. Salesky. 1973. Grazing by the echinoid Diadema antillarum Philippi: Formation of halos around West Indian patch reefs. Science 182: 715-717. Paul, V. J., & W. Fenical. 1983. Isolation of hal imedatrial : chemical defense adaptation in the calcereous reef-building alga Halimeda. Science 221: 747- 749. Randall, J. E. 1961. Overgrazing of algae by herbivorous marine fishes. Ecology 42: 812. Randall, J. E. 1965. Grazing effects on seagrasses by herbivorous reef fishes in the West Indies. Ecology 46: 255-260. Steneck, R. S. , & W. H. Adey. 1976. The role of environment in control of morphology in Lithophyllum congestum, a Caribbean algal ridge builder. Bot. Mar. 19: 197-215. Stephenson, W., & R. B. Searles. 1960. Experimental studies on the ecology of intertidal environments at Heron Island, Australia. Aust. J. Mar. Freshw. Res. 11: 241-267. van den Hoek , C, A. M. Breeman, R. P. M. Bak , & G. van Buurt. 1978. The distribution of algae, corals, and gorgonians in relation to depth, light attenuation, water movement, and grazing pressure in the fringing coral reef of Curacao, Netherlands Antilles. Aquat. Bot. 5: 1-46. 102 QUANTIFYING HERBIVORY ON CORAL REEFS: JUST SCRATCHING THE SURFACE AND STILL BITING OFF MORE THAN WE CAN CHEW Robert S. Steneck Department of Zoology, University of Maine at Orono Darling Center Marine Laboratory, Walpole, Me 04573 ABSTRACT Herbivory was quantified using six different techniques simultaneously at nine discrete sites along a coral reef system on the Caribbean island of St. Croix. In order to study diverse assemblages of herbivores and algae, functional groups were used for both. The groups are based on shapes and structural properties of the algae and feeding capabilities of the herbivores. Herbivory was most frequent and intense in the shallow forereef sites where an average of over 5,000 herbivorous fish bites per meter square per hour was recorded. Although most of these were from herbivorous fishes that do not denude primary subtratum (i.e., small damself ishes), this site also had the highest frequency of grazing from herbivores capable of denuding (i.e., yellowtail damselfish and tangs) and excavating the calcareous substratum (parrotf ishes, urchins and limpets). Herbivory from all sources decreased in backreef, shallow algal ridge, and deep wall-reef habitats. The latter sites had the lowest levels of grazing. Herbivory on macrophytes was assessed using a Thalassia bioassay technique but the results at the forereef sites contradicted those of all other techniques. Caution is suggested in applying this technique as a single measure of herbivory. INTRODUCTION The process of herbivory is generally thought to be of primary importance to the distribution and abundance of benthic algae on coral reefs (reviewed by Lubchenco and Gaines 1981). While the units of measure and methods for determining the abundance of algal prey are well established, no such convention exists for determining the impact of their herbivorous predators. Measurements of percent cover, biomass, or number of individuals when applied to assemblages of reef dwelling herbivores are of dubious meaning for quantifying herbivory. For instance, how many foraging urchins equal the impact of a 20 kg parrotfish? The "apples and oranges" involved here result from trying to force units and techniques which are designed to determine patterns in herbivore abundance on the process of herbivory. Ecological processes are factors that result in observed patterns. It is generally assumed that the abundance of herbivores corresponds with their impact on algae. This assumption has never been tested. In this paper, I will report on several techniques used simultaneously on a single reef system in order to measure the impact of a diverse assemblage of herbivores on an assemblage of reef-dwelling algae. I will also provide an argument for considering this topic at a "functional group" level so that herbivores with similar effects and algae with similar ecological properties are treated together. ORGANISMS, STUDY SITES, AND METHODS Functional Groups Since herbivory involves the interaction of two diverse groups .of organisms, "functional group" subdivisions will be used. Algae have been subdivided into such groups based on shared anatomical and morphological characteristics (see Littler and Littler 1980, Steneck and Watling 1982). For the purposes of this paper, I will simplify these subdivisions to three groups: 1) ALGAL TURFS (diverse, microscopic 103 filaments; e.g., Pol vsiphoniat Space1aria» and Taenioma) , 2) MACROPHYTES (larger, more ridged forms; e.g.» Laurencia, Dictvota, Jania, and Asperogopsis) , and 3) ENCRUSTING CORALLINES (calcareous algal crusts; Porolithon, Neogoniol ithon and Paragoniol ithon) . Note that the "turfs" referred to here follow Neushul 1967, Randall 1967, Dahl 1972, John and Pople 1973, Van den Hoek et al_. 1975, Adey et_ al_. 1977, Benayahu and Loya 1977, Pichon and Morrisey 1981, but not Hay 1981a). Among the ecological properties correlated with algal functional groups (and this simplified scheme) is toughness of the thallus. From an herbivore's perspective turfs are easiest to consume, rnacrophytes intermediate and coralline crusts most difficult (Littler and Littler 1980, Steneck and Watling 1982, Littler ejt_ aj_. 1983). Herbivores fall into three categories with respect to grazing (Steneck 1983, see Table 1 for species): 1) NON-DENUDING (those incapable of, or unlikely to, denude the substratum of algae; e.g., some damselfishes [Eupornacentrus], anphipods and polychaetes; Brawley and Adey 1977, Kaufman 1977), 2) DENUDING (those that denude the substratum of turfs and smaller macroalgae but are incapable of excavating corallines and large leathery rnacrophytes; e.g., yellowtail damselfish [Microspathodon], tangs [Acanthurus] several non-limpet archaeogastropods and mesogastropods; Randall 1967, Jones 1968), and 3) EXCAVATING (those capable of consuming even the toughest forms of algae, such as encrusting corallines e.g., limpets CAcmaeaJ, chitons, some regular echinoids [Diadema ] and parrotfishes CScarus and Sparisoma] : Randall 1967). For a more complete discussion of these categories see Steneck 1983 . Study Sites This study was conducted at nine sites in three locations along the north shore of St. Croix (Fig. 1). The locations were selected as representative of three common reef types in the Caribbean 1) Alcal Ridge (Boiler Bay), 2) Bank barrier-reef (Tague Bay), and 3) Deep wall-reef (Salt River Canyon). All three are geographically close to one another, thus giving them about the same exposure to light, destructive storms and recruitment events. The predominant algal growth form at the bank-barrier and deep wall-reef sites are turfs with an average canopy height of 2mr.i. Turfs are highly diverse with about 30 to 50 species found in an average four square centimeter area (Adey and Steneck in ms.). Species composition is temporally variable with up to an 80/3 change in community dominance every 3-4 months (Steneck in prep.). Patches of encrusting coralline algae are also scattered throughout. Macroalgae are only abundant in places at the algal ridge site (see Connor and Adey 1977). At each site, circular slabs of coral (1 cm thick, and 10 cm diameter) were placed on racks and affixed to the reef (locations and depths in Fig. 1). The plates were placed in December of 1979, and most of the experiments were conductea more than three years later from March to May 1982. At the time of the experiments, the coral plates had the same algal community structure, canopy height, and biomass as the surrounding substrata. All the plates at the nine stations were covered with the same (algal turf) functional group. Details of specific experiments are described below where appropriate. RESULTS AND DISCUSSION To reduce some of the variables related to herbivory, the study was conducted at one spot and at one time at each of the nine sites. Several of the experiments ran simultaneously, focusing on the same six (100 to 200 crn sq. ) planar coral plate surfaces or their surrounding areas. 104 TABLE 1 HERBIVORES OBSERVED IN THIS STUDY NON-DENUDING HERBIVORES FISHES: Eupornacentrus dorsopunicans (dusky damselfish) Eupomacentrus planifrons (threespot damselfish) Eupornacentrus variablis (cocoa damselfish) Eupomacentrus leucostictus (beaugreyory) GASTROPODS: Fissurella angusta (keyhole limpet) REFERENCES Randall 1967, Pers. Obs. 3rawley and Adey 1977, Kaufman 1977 Randall 1967, Pers. Obs. Randall 1967, Pers. Obs. Steneck and Adey 1982 DENUDING HERBIVORES FISHES: Microspathodon chrvsurus (yellowtail damselfish) Ophioblennius atlanticus (redlip blenny) Acanthurus bahianus (ocean surgeon) Acanthurus coeruleus (blue tang) REFERENCES Randall 1967, Steneck 1983 Randall 1967, Steneck 1983 Randall 1967, Steneck 1983 Randall 1967, Steneck 1983 EXCAVATING HERBIVORES FISHES: Scarus croicensis (striped parrotfish) Sparisorna chrysopterum (redtail parrotfish) Sparisoma viride (stoplight parrotfish) ECHINOIDS: Diadema anti 1 la rum (long-spined sea urchin) GASTROPODS : Acmaea pustulata (limpet) REFERENCES Randall 1967, Randall 1967, Randall 1967, Steneck 1983 Steneck 1983 Steneck 1983 Steneck 1983 Steneck and Watling 1982, Steneck 1983 105 FIGURE 1. Diagrammatic composite representation of reef transects along the north of St. Croix. Numbers correspond to study sites in Table 2. KM 106 Observing Herbi vorv Quantifying fish grazing employed two methods. The first involved watching each rack of six coral plates at each station for five-rninute intervals several times a day and recording which species fed on them. The second method is identical to the first except that an underwater time-lapse movie camera takes the place of a diver. Each movie lapses between 10 and 12 hours and generates 3600 observations (frames). During the movies, divers avoided the area except to simultaneously count fish bites over a few five minute intervals. Since the plate areas are known, the number of bites per square meter per hour can be determined for both techniques. Fish grazing was greatest in shallow forereef habitats and least at depth (Table 2). Mean grazing frequencies of over 5,000 bites per meter square per hour were recorded using visual and time-lapse techniques in 1982 and visually in 1981. At the algal ridge and bank barrier-reef sites grazing was predominantly from non-denuding herbivores (i.e., Eupomacentrus) . This group of damselfishes Cfour species) accounted for 72£ (+ 30) and 50% (+ 22) of the bites observed at this site visually and on film, respectively. Pomacentrid grazing on the deep wall-reef was relatively low (6 7i + 12). Denuding herbivores (particularly Microspathodon) were most abundant at the rnid-depth forereef site. Excavating herbivores (Scarus and Spar i soma) were most abundant at the shallow forereef site. Juveniles of this group were the primary herbivores at the deep wall-reef sites. It is difficult to determine why a better correspondence between visual and tin.e-lapse movie techniques does not exist, since both were conducted at the same time (Table 2). Time-lapse, of course, gives high resolution data over an entire day whereas visual techniques give high resolution over only short intervals spread over several days. The latter technique also requires the presence of a human observer which may suppress normal grazing frequencies during the observation period. Invertebrate herbivores are easier to count, but more troublesome to actually observe feeding since their mouths are under their bodies. Since they have reduced mobility, their range of grazing influence is relatively restricted. Thus, some indication of their impact can be assessed by determining their population density in a given area. Excavating invertebrate herbivores (i.e., echinoid Diadema and limpet Acmaea), are most abundant in shallow forereef environments (Table 2). In fact, the two shallowest forereef stations have more echinoids per area than all seven other stations combined. Limpets were only found at the shallowest station. Impact of Herbivorv: The impact of herbivory requires studying the plants being eaten. Because different functional groups of algae have different structural and morphological properties that contribute to the rate at which they are consumed (discussed above), I will treat three functional groups of algae differently. In terms of areal coverage, minute turfs are the most abundant algal form on coral reefs. They are impossible to handle without damage so all turf experiments and observations were confined to the coral plates. To determine herbivore impact on turfs, the rate of biomass loss to herbivores was studied. For this, a set of six plates at each depth (at the deep wall-reef site) was suspended in the water column at the depth they had been growing in for the past three years. An identical set of plates remained on the benthos. No herbivores were observed grazing the suspended plates (nor was there any evidence of grazing using other techniques (i.e., Thalassia bloassay technique, described below). Table 2 shows that there 1s a steady (nearly linear) decline 1n the rate of turf algae biomass loss to herbivores with depth. Assuming that the productivity of suspended plates equals that of the benthic plates, the difference in algal biomass after a period of 6 days should indicate the amount of algal biomass that is eaten during that period. Negative numbers indicate that algal growth rate exceeds the 107 STATION NUMBERS DEPTHS (FEET) TABLE 2 COMPARISON OF RESULTS IN MEASURING HER3IV0RY 1234567 89 0.5 3.5 5 15 30 30 60 90 120 HERBIVORY (BITES/M2/H) FROM Visual Obs. 1981 Visual Obs. 1982 Time-lapse 1982 MEAN Non-Denuding Fishes 0 0 3,269 170 28 448 273 518 147 230 0 7,061 1,567 3,325 1,630 1,151 2,503 613 2,371 649 470 0 94 94 Visual Obs. 1981 Visual Obs. 1982 Time-lapse 1982 MEAN Denuding Fishes 3 0 2,774 170 28 0 0 532 193 48 39 936 2,687 0 4,326 1,151 313 896 1,102 1,563 409 39 12 12 Visual Obs. 1981 Visual Obs. 1982 Time-lapse 1982 MEAN Excavating Fishes 0 0 2,476 * 0 0 0 0 27 0 0 79 0 0 2,660 0 0 0 282 0 1,721 0 0 70 282 235 235 Benthic Plates Suspended Plates Diadema Acmaea Total Herbivory From Fishes 2,816 1,509 5,194 2,212 879 116 338 235 0 0 0 0 0 Excavating Invertebrates (No/rrr ) 0.8 3.8 14.3 17.5 3.9 0.2 0.3 0.5 0 002 0.3 000 00 Rate of Turf Biomass Loss GRAZING IMPACT ON: Algal Turfs (g dry/mVday) 2.8 1.35 0.933 -0.79 On benthos On Plates Suspended Plates % Grazed by Fish Macrophytes (% Thalassia eaten/day) 2.04 14.9 1.45 0.74 49.8 25.2 4.8 0.004 1.6 0.79 37.4 16.0 0 100 95 50 50 96 98 10.1 3.0 0.69 2.2 3.4 1.5 0 0 0 100 100 90 % Coralline Grazed 7<> Grazed by Urchins Coral 1 ines 0.6 54.2 82.7 48.0 44.6 0 26.6 57.9 42.0 28.0 NOTE: Blanks indicate no data 108 herbivore grazing rate. Macrophytes are not usually abundant on reefs. In the study areas, macrophytes were most abundant in the intertidal zone of the algal ridges at Boiler Bay (e.g., Gracilaria, Laurencia, Gelidiella, Dictvota and Acanthophora) and to a lesser extent on the Acropora cervicornis in the sand plain in front of the forereef sites (e.g., Asparaaopsis, Dictvota and Laurencia). A scattering of macroalgae can also be found at the backreef site (e.g., Acanthophora, Laurencia and Dictvota). To determine grazing pressure on macrophytes, a Thalassia bioassay technique of Hay (1981b, and Hay et jj. 1983) was used. For this, five centimeter long blades of Thalassia were placed in clothes pins. The blades were checked every several hours to determine the amount grazed by herbivores. Bite marks were interpreted as to whether they were from parrotfishes (semicircular bites) or urchins (a shredded appearance) (See Fig. 2 in Hay et _aj. , 1983). At each station, eight blades of Thalassia were affixed to the rack of coral plates and an additional 20 were scattered nearby on the bottom. The rate of Thalassia loss was greatest at the deepest forereef station (for both those scattered on the bottom and attached to the rack; Table 2). The next highest rate of loss at the Tague 3ay stations was in the backreef. In this case however, only those scattered on the bottom were heavily grazed. The deep wall-reef sites showed a consistent decrease in Thalassia grazing with depth. None of the suspended Thalassia blades were grazed. Nearly all grazing marks were attributable to parrotfishes. No urchin marks were identified on Thalassia blades attached to the racks of coral plates. At benthic stations, however, a few urchin marks were observed. The nighest proportion of urchin marks occurred in the shallow forereef stations of the Tague Bay locations (where urchin densities were highest). The 10X urchin bites recorded for the 120' station represents only one urchin-looking bite out of 10 recognizable bites. It is probably an error since no urchins were found below 90'. Other shredding herbivores such as crabs could have been responsible for the marks. The pattern of Thalassia loss across the Tague Bay reef is opposite that of all other measurements of herbivory. There is no indication from other measures (Table 2) that the deep forereef, and to a lesser extent the backreef, receive as high a rate of predation as the Thalassia bioassay suggests. Conversely, all other techniques indicate that the two snallow forereef stations are most heavily grazed by all herbivore groups rather than mininally grazed as the Thalassia bioassay indicates. There is no question, however, that wnatever the Thalassia bioassay measures, it does so consistently and repeatably. It is possible that fishes with a search image for Thalassia, or other conspicuous macrophytes of sand flats or lagoons, are attracted to the Thalassia of this experiment and consume it at a rate unrepresentative of overall grazing rates on the reef. Relatively few herbivores are capable of excavating crustose coralline crusts (discussed in detail in Steneck 1983). Since crusts can "erase" most graze marks as they grow undisturbed for 10 to 30 days (depending on the depth of the injury and the growth rate of the coralline), marks on corallines are a rough indication of the rate of grazing by excavating herbivores. In addition, the bite marks of the major groups of excavating herbivores (i.e., parrotfishes, urchins, limpets and chitons) are readily identifiable. The pattern of graze marks among excavating herbivores indicates that the greatest grazing pressure occurs in the shallow forereef. Data were not collected for the deep wall -reef site. Overall urchin grazing was most conspicuous on the crusts, with the proportion of graze marks generally corresponding with the abundance of urchins in the area (Table 2). 109 Patterns of Herbivorv on Reefs Herbivory is greatest in shallow forereef habitats and reduced in shallow algal ridge and backreef habitats. The reduced herbivory around massive algal ridges is likely due to the absence of habitat and refuge space in those environments coupled with the characteristic turbulence (discussed in Steneck and Adey 1976). Backreef habitats are an enigma. They consistently indicate lower levels of grazing (both intensity and frequency) but no simple explanation can be offered. The decrease in herbivory with depth is not surprising since the trophic carrying capacity for herbivores probably diminishes with depth as a function of reduced benthic productivity (Steneck in prep.). Ecologists are far from agreeing on a method (or methods) for measuring herbivory on reefs. While this does not seem to inhibit publications on the subject, comparisons between reefs are impossible. The simplest and most easily replicated technique, the Thai ass i a bioassay, is of dubious meaning and must be examined critically in other comparative studies before its general application can be accepted (see Hay 1983 for caveat). The only clear message revealed in this study is that quantifying a process is infinitely more difficult than quantifying patterns in the abundance of herbivores. So far we have only scratched the surface and still we are biting off more than we can chew. ACKNOWLEDGEMENTS: This project was funded by a grant from NOAA NA82AAA01460 for Hydrolab mission 82-6. D. Blanchard, S. Brusila, R. Carpenter, D. Cancilla, D. Estler, E. Fleishmann, M. Hay, L. Mayer, B. Milliken, D. Morrison, J. W. Porter, and L. Schick assisted in various phases of the study. LITERATURE CITED: Adey, W. H., P. Adey, R. B. Burke, and L. Kaufman. 1977. The Holocene reef systems of eastern Martinique. Atoll Res. Bull. 218:1-40. Benayahu, Y., and Y. Loya. 1977. Seasonal occurence of benthic algal communities and grazing regulation by sea urchins. Proc. 3rd Int. Coral Reef Symp. 383-389. Brawley, S. H., and W. H. Adey. 1981. The effect of mirograzers on algal community structure in a coral reef microcosm. Mar. Biol. 61:167-177. Connor, J. L., and W. H. Adey. 1977. The benthic algal composition, standing crop and productivity of a Caribbean algal ridge. Atoll Res. Bull. 211: 1-15. Dahl, A. 1972. Ecology and community structure of some tropical reef algae in Samoa. Proc. Int. Seaweed Symp. i7:36-39. Hay, M. E. 1981a. The functional morphology of turf-forming seaweeds: persistence in stressful marine habitats. Ecology 62: 739-750. Hay, M. E. 1981b. Spatial patterns of grazing intensity on a Caribbean barrier reef: herbivory and algal distribution. Aquat. Bot. 11:97-109. Hay, M. E. 1983. Patterns of fish and urchin grazing on Caribbean coral reefs: Are previous results typical? Ecology (in press). Hay, M. E., T. Colburn, and D. Downing. 1983. Spatial and temporal patterns in herbivory on a Caribbean fringing reef: the effects on plant distribution. Oecologia 58:299-308. John, D., and W. Pople. 1973. The fish grazing of rocky shore algae in the Gulf of Guinea. J. Exp. Mar. Biol. Ecol . 11:81-90. Jones, R. S. 1968. Ecological relationships in Hawaiian and Johnston Island Acanthuridae (Surgeonf ishes). Micronesica 4:309-3bl. Kaufman, L. 1977. The three spot damselfish: effects on benthic biota of Caribbean coral reefs. Proc. 3rd Int. Coral Reef Symp. 1:559-564. Littler, M. m. , and D. S. Littler. 1980. The evolution of thallus form and survival 110 strategies in benthic marine macroalgae: field and laboratory tests of a functional form model. Amer. Mat. 116: 25-44. Littler, M. M., and D. S. Littler, and P. R. Taylor. 1983. Evolutionary strategies in a tropical barrier reef system: functional-form groups of marine macroalgae. J. Phycol. 19: 229-237. Lubchenco, J., and S. D. Gaines. 1981. A unified approach to marine plant-herbivore interactions. I. Populations and communities. Ann. Rev. Ecol. Syst. 12: 405-437. Neushul, M. 1967. Studies of subtidal marine vegetation. Ecology 48: 83-94. Pichon, M. and J. Morrissey. 1981. Benthic zonation and community structure of South Island Reef, Lizard Island (Great Barrier Reef) 31:581-593. Randall, J. E. 1967. Food habits of reef fishes of the West Indies. Stud. Trop. Oceanogr., Miami 5:665-847. Steneck, R. S. 1983. Escalating herbivory and resulting adaptive trends in calcareous algal crusts. Paleobiology 9: 44-61. Steneck, R. S., and W. H. Adey. 1976. The role of environment in control of morphology in Lithophvl lum congestum, a Caribbean algal ridge builder. Bot. Mar. 19:197-215. Steneck, R. S., and L. Watling. 1982. Feeding capabilities and limitations of herbivorous molluscs: a functional group approach. Mar. Biol. 68: 299-319. Van den Hoek, C. , A. Cortel-Breeman, and J. Wanders. 1975. Algal zonation in the fringing coral reef of Curacao, Netherlands Antilles, in relation to zonation of corals In gorgonians. Aquat. Bot. 1: 269-308. Ill DIFFERENTIAL EFFECTS OF CORAL REEF HERBIVORES ON ALGAL COMMUNITY STRUCTURE AND FUNCTION Robert C. Carpenter Insti tute of Ecol ogy University of Georgia Athens, Georgia 30602 ABSTRACT Fu i n t e n remov algal revea herbi speci forms of pr those algal to su n c t i o si ty al , s comm led d vores es co domi imary i n n comm bsequ nal and peci unit iffe ran mpos na ti pro on-u unit ent herbivore groups were categorized according to the frequency of disturbance resulting from feeding. Biomass es composition and primary productivity of experimental ies subjected to herbivory by each functional group rences among groups. Combined removal of biomass by all ged from 32 to 84% of daily algal production. Algal it ion varied according to grazing regime, with crustose ng the most intensely grazed treatments. Specific rates duction of urchin grazed treatments were 2 to 4 times renin grazed treatments. Species composition of reef ies is affected differentially by herbivores, which leads changes in community metabolism. INTRODUCTION C prod orga free nota alga herb reef Carp comm and such the thei whi c diff T the reef desc orSl re u c t i o n n i s m s . -living bly i nc 1 commu i v o r e s algae enter, unities support i n t e n s s tructu r m a g n i h such erent g his pap effects algal r i p t i o n ef communitie that supports The majority algal compon onspicuous wh nities such a has been repe (Stephenson a 1981 ) . The pr invites spec ed by the app e disturbance re of the al g tude and mode removal takes roups of herb er reports so of different community str of methods a s are ch an abun of this ent (Mar en compa s kelp f atedly s nd Sear! edomi nan u 1 a t i o n arently ( herbi v al commu of b i o m place, i v o r e s . me p r e 1 i f uncti o ucture a nd resul arac dant prod sh , red ores hown es , ce o on h spar ory) nity ass the t e r i z and u c t i o 1976) to ot ts. I to 1 1960; f sue ow su se al may . Bee remov ef f ec ed by diver n i s ; how her h ntens i m i t Samm h her ch bi gal c have ause al an ts ma a hig se ass accomp ever, t i g h 1 y e graz the st arco, e b i v o r e omass ommuni s i g n i f h e r b i v d the y not h 1 evel embl age 1 i s h e d his com product ing pre a n d i n g t aj_. , s in re can be ty. In i c a n t e ores d i f requen be the of primary of by the p o n e n t is i v e marine ssure by crop of 1974; ef achieved addition, ffects on f f e r in cy with same for minary results of a study to quantify nal groups of herbivores on shallow nd function. A more detailed ts will be presented elsewhere. MATERIALS AND METHODS Functional herbivore groups were categorized using two criteria: the size of patch (disturbance) created per feeding bout, and the probability (frequency) of regrazing the same patch on a subsequent feeding bout. The former is mainly dependent on the size of the feeding apparatus and foraging behaviors such as the number of bites in a particular feeding bout. The second criterion depends on the mobility of the organism as well as on behaviors which determine the size of the foraging range (i.e., homing). These groups cross 113 phylog only f member even o d i s t r i s i s t s i d a e a catego the ec am phi p e n e t i c or ill s of e n a mi b u t i o n of sev nd her ry. Th h i n o i d ods , t boundaries (Figure 1) and are depicted as non-overlapping ustration. The organisms listed are potential Caribbean ach group; the actual composition at a given locality, cro-habitat level, will depend on patterns of species and abundance. For the present study site, Group 1 con- eral fish species in the families Scaridae and Acanthur- eafter will be referred to as the herbivorous fish e most important herbivore in Group 2 at this locality is Pi ad em a a n t i 1 1 a r u m . Members of Group 3 include several ana ids, syllid polychaetes and small gastropods. 3 high 3EQ. OF REGRAZING SAME PATCH Microher bivores amphtpods syllid polychaetes tanaids sm gastropods Diadema Mithrax sp. Blenmdae Acanthunda Kyphosldae Scaridae PATCH SIZE CHEATED PER UNIT TIME large Figure 1 . Diagramatic representation of herbivore functional groups. Group members are potential Caribbean herbivores. Algal communities were allowed to develop under ambient backreef conditions (depth, 2 m) in St. Croix, U.S. Virgin Islands on exper- imental substrata consisting of 8 cm x 8 cm x 1 cm cross-sections of the coral Acropora palmata. These experimental plates were randomly placed in one of the following treatments: grazing by groups 1 and 3, grazing by groups 2 and 3, grazing by group 3 only, no grazing and control. The accessibility of plates to functional groups was con- trolled using a combination of cages and suspended platforms at the same depth. Each treatment was replicated once. A detailed descrip- tion of the experimental design and methods is forthcoming (Carpenter, ms. in prep.). Algal community primary productivity was estimated monthly from changes in dissolved oxygen in well-stirred light and dark plexiglass chambers on randomly chosen plates from each treat- ment. From the same plates samples were taken to estimate algal bio- mass (decalcified dry wt.), species composition and chlorophyll a_ content . The amount of algal biomass removed by each functional group was estimated in two ways. The difference in algal standing crop between plates not grazed and those grazed by a given group gave one estimate. The second, more accurate method converted the areal production rate (corrected for a constant percent of dissolved organic matter excret- ion (Fogg, 1976)) for a treatment to carbon equivalents, and assuming a constant carbon content, this was translated into the expected dry, decalcified biomass accumulation over a time period. The difference between this expected amount and the actual algal standing crop was assumed to be removed by herbivores. All data reported here were obtained at least one year after the initiation of the described treatments. 114 RESULTS Removal of Algal Biomass Herbivores removed between 0.88 and 2.10 g dry wt. m day , which represents between 32 and 84% of the daily production of algal biomass (Table 1). The temporal variation in the total biomass removed was mainly the result of variation in the amount removed by the herbivorous fish component. The percent of the total biomass re- moved by all herbivores that was taken by fishes varied between 3 and 39%. The percent removed by Pi adema only varied from 41 to 68%, and micro-herbivores removed from 11 to 28% of the biomass. Table 1. Mean decalcified dry wt. removed (g m'^d"^) + S.D. by each functional herbivore group; the % of daily algal produc- tion that this represents; and, in parentheses, the % of the total amount removed (N=4). Note that the biomass removed by all herbivore groups is an independent estimate, not the sum of the removals by the three separate groups. GROUP JUNE JULY AUGUST wt % 19(38) 26(51) 6(11) 51 wt % wt % 1 2 3 ALL 0.4 5+0. 12 0.65+0.05 0.09+0.07 1.27+0.28 0.78+0. 18 33(39) 0.85+0.14 35(41) 0.31+0.20 16(20) 2.10+0.61 84 0.03+0.05 1(3) 0.59+0.11 22(68) 0.25+0.17 9(28) 0.88+0.26 32 Based on herbivore densities in the study area sible for approximately 50% of the herb ivory, fishes herbivores the remaining 20-30% of the biomass removed D i a d e m a is respon- 20-30%. and micro' Algal Species Composition Alga were pr p 1 e t e, s presenc reg ime . Pi adema domi nat with mi Ch 1 orop wel 1 re of thes by a si u 1 o i d e s that th and oth I n c a commu of the Amp hi ro b e n t h i c 1 spe esent o per e/abs Plat show ed by nute hyta, prese e pi a ncl e ) tha is sp er f i ontra nity f a m i 1 a f ra ci es c , and cent r ence a es gra s i m i 1 crust (10-50 Phaeo nted . tes. P specie t form eci es 1 ament st , pi domi na y Gel i g i 1 i s s ompo thei elat nd d zed ar p ose urn phyt No 1 late s of ed p i s a s we ates ted di ac ima ) foraminifera. si ti r re i ve omi n by a atte form di am a an arge s f r fil atch ct i v re a gra by 1 eae, to This on d lati abun ance 11 h rns s (U eter d Rh cor om t amen es o ely 1 so zed arge the one shi iffe ve a danc dat erbi of a lvel red bo bundan e can a s how vore g Igal c la Ten Tfi odop ti ca he f tous f va avoi of p only r al upr com ft i 1 ament hyta . ted al i s h g r brown rying ded by ri mary by mi gae (C i g h t c prised n comm th in ce . An not be a pat roups ommuni s_ and ous re Cyano- gal sp a z i n g algae size. herbi impor cro-he oel oth term alys gi v tern and ty s crus pres bact eci e trea (S_£ Obse voro tanc rbi v r i x s of what species es are not yet corn- en ; however, related to grazing those grazed by tructure. Each is tose corallines) entatives of the eria were also s were found on any tment were dominated h a c e 1 a r i a t r i b - rv at ions suggest us fishes. Crusts e on these pi ates . ores fluctuated from i rregul ari s , species oral 1 i prima uni ty nes ri ly stru J a n i a adherens and of crust c t u r e c o i s and n c i d e d with 115 the re This w mass r occur r The and se or m i n specie presen At algal Algal c r u i t as al emove ed on fila veral i m a 1 . s (Ch t onl this s p e c i Prima ment so a d by the ment blu Pen ampi of th ccompa micro se pi a ous sp e-gree nate d a parv e g n i e -he tes eci ns i at ul a Pro treatm per ch betwee d u c t i ents 1 orop n gra y on time es o ry P these there c c u r r i roduct Pi ap ng i on astropod C e r i t h i u m 1 i tt d by an increase in the rbivores (Table 1). Sma es Herposiphonia secund dominated plates where oms were also abundant. , C e r a m i u m n i t e n s and L a tes . pears under eratum onto the plates amount of algal bio- 11 filaments also a_, G i f f o r d i a r a 1 1 s i a e herbivory was absent Several 1 arger algal a u r e n c i a obtusa ) were to be no clear pa each type of graz on p (Fig hyll zing er utii ure 2 ) a wei r e g i m t a . W ght es . rea was not significant hen expressed per unit , productivity rates va A consistent pattern o ttern in the number of i n g regime. ly different between algal weight or r i e d widely f higher specific 0.8i .C I O Month Month Month Figure 2. Mean net community primary production + 95% confidence limits; A) per area, B) per unit algal weight, and C) per unit chl. a weight. (X)- all herbivore treatment. (□)- group 1 and 3 treat- ment, (a)- group 2 and 3 treatment. (■)- group 3 only treatment. (•)- no grazing treatment. J, J, and A on the x axis indicate June, July, and August. Plates were submerged initially in December, 1981, and these data were collected during the summer of 1983. production rates in urchin grazed treatments (all herbivore treatment and urchin only treatment) than in fish grazed, micro-herbivore grazed and non-grazed treatments, was mai ntai ned. Net specific com- munity primary production rates (ug 02 ug chl. a'1 hr."1) differed by a factor of two to four among the treatments, The micro-herbivore grazed and ungrazed treatments had the lowest specific rates of algal production. Specific rates were inversely related to algal standing crop . DISCUSSION These data provide evidence that different herbivores have variable effects on the benthic algal community. Data on algal species compos- ition and specific primary productivity between treatments demonstrate that functionally different herbivores have qualitatively and quanti- tatively different effects on these parameters. Algal communities grazed by Diadema apparently consist of algal species with much higher specific rates of production. Such low biomass communities are main- 116 t a i n e d in grazing , i m a t e 1 y t dominant, fish g r a z algal s pe specific sity betw the abund to m a i n t a 1 o c a 1 i t i e maintain species, if truly The ra those fou and, thoug daily alg removed c taken by a t i o n of i n the pe variation amount of taken by that al 1 b i o m a s s a However, than in m contrast is that H for all o by fishes bably ref site. For al different complex r areas , wh structure dance bet ability i communi ti factors ( but also ses which a hi s p e c i wo, an This ing, ci es produ een t ance i n a s, th a s i m bi oma gener tes o nd f o h tern al pr an re h e r b i both rcent s i n bi om h e r b i the n ccumu a 1 ar ost e betwe atche f the was 1 ects gh tu fie p d it diff al 1 ow (Ogde c t i v i reatm of Di very e abu i 1 ar ss wo al i ze f alg r her poral oduct suit vores these bi om algal ass r vores et pr 1 a t i o ger p cosys en th r (19 alga 1 ess the move roduc appea erenc ing d n and ty ar ents adema produ ndanc algal uld b d gra a 1 ' b i b i v o r ly va ion . from , var fact ass r prod emove may i mary n tha ropor terns ese r 81 ) f 1 con than rel at r state t i o n ra rs that e may b i s c r i m i Lobel , e in pa ( amount (and g c t i v e , e of he commun e h i g h e zing o c omass r es on t r i a b 1 e , V a r i a t i changes i a t i o n s ors . Fo emoved u c t i o n d. This be an u produc t is a v t i o n of ( Wi eger esul ts ound th s u m p t i o 50%, wit i v e abu by tes di f e th nati 197 rt d of razi 1 ow rbi v ity. r ( a curs emov he G rep ons i n i n r th did and est nder ti on ai 1 a mat t an and at h n. I h ur ndan intense herb ivory. Without urchin decrease by a factor of approx- ferent algal species become e result of the visual nature of on and preferences for certain 8). Differential effects on algal ue to differences in grazing inten- biomass removed). In this locality, ng intensity) is sufficiently great bi omass algal community. In other orous fishes may be high enough to However, if fishes avoid some algal nd specific production lower) than al re reat resen in th the a rates e dat not c repre i m a t e e s t i m (les ble f e r i a 1 d Owe those e r b i v n thi chins ces o port Barr t be e pe bsol of a pr ons i sent of at io s ex or c goe n, 1 obt orou s s t con f th ed h i er twee rcen u te prod esen s ten cha the n du cret onsu s to 971 ) a i ne s fi udy , sumi ese ere Reef n 50 t of amou ucti ted tly nges perc e to ion ) mpti her . An d i n s hes the ng 4 herb are ver (Hatch and 60 daily nt of b on, or a here , v correl a in the ent pro the as is g o i on by h b i v o r e s i n t e r e Austra were r amount 0-70%. i v o r e s y cl er , % of prod i oma com aria te w abs duct sump ng i erb i i n s ti n 1 i an espo rem This ate ose to 1981 ) net u c t i o n ss b i n a - t i o n s i th ol ute i on ti on nto vores . this 9 reefs n s i b 1 e oved pro- ach gal comm l a 1 e f f e eefs, how ich may and fun ween and n these es are n water mo by biolo in turn unities a cts of va ever, prec lead to m c t i o n . Pa within r parameter ot only s vement, n gi cal ly , affect f cces r i ou 1 ude osai tter eef s. T true utri poss unct s i b 1 e s her acce cs of ns of zones his s tured ent a ibly i o n a 1 to all herbivore groups, the bi vores may be masked. Spatially ss by all groups, at least in some lgal communities with different and abun- 117 LITERATURE CITED Carp ef Fogg Au Hate ep me Mars Mi Ogde ur Samm CO (E 32 Step ec Fr Wieg so ec enter , R . fects on , G.E. 19 str. J. P her, B.G. i 1 i t h i c a n t . Proc . h, J. A. 1 crones ica n , J . C ., a chins in arco , P . W ntrol of chi no derm : 47-53 . henson , W ology of e s h w . Res ert, R . F. urces and osystems . C. 1981 the ben 76. Rel lant Ph 1981 . Igal co 4th In 976. En 12:13- nd P.S coral r . , J.S. commun i ata : Ec . G r a z i n t h i c a 1 g ease of y s i o 1 . 3 The i n t e mmuni ty t . Coral e r g e t i c 21. Lobel. 1 eef comm L e v i n t o ty struc h i n o i d e a g by Pi ad em a anti 1 1 arum Philippi and its al community. J. Mar. Res. 39:749-765. glycollate from tropical marine plants. : 57-61 . raction between grazing organisms and the of a coral reef: a quantitative assess- Reef Symp. 2:515-524. role of algae in reef ecosystems. 978. u n i t i n, and ture ) : a prel imi nary The role of herbivorous fish and es. Env. Biol. Fish. 3:49-63. J.C. Ogden. 1974. Grazing and by D i a d e m a a n t i 1 1 a r u m Philippi study. J . Mar. Res . ., and R . B . Sear! es . intertidal environme . 2:241-267. and D.F. Owen. 1971 population density J. Theor. Biol. 69: 1960. Experimental studies on the nts at Heron Island. Austr. J. Mar. . Trophic structure, available re- in terrestrial vs. aquatic 69-91 . 118 NEARSHORE AND SHELF-EDGE OCULINA CORAL REEFS: THE EFFECTS OF UPWELLING ON CORAL GROWTH AND ON THE ASSOCIATED FAUNAL COMMUNITIES John K. Reed Harbor Branch Foundation, Inc Ft. Pierce, Florida Rt. 1, Box 196 33450 ABSTRACT Colonies of Ocul i na vari cosa were collected from nearshore, mid- shelf, and shelf-edge reefs off central eastern Florida. The shelf- edge reefs are inundated episodically throughout the year by upwell- ing of cool, nutrient rich water. On the inner shelf, cyclic seasonal factors predominate and upwelling intrudes only for a few weeks during the summer. Growth rates of the coral are significantly greater on the shelf-edge reefs than nearshore, even though at the shelf-edge temperatures are cooler and the coral lacks zooxanthel 1 ae. Species diversities of assemblages of decapods and mollusks associ- ated with the coral are greater at the shelf-edge reef than at the mid- and inner-shelf reefs. It is suggested that the upwelling of nutrient rich water onto the shelf-edge Ocul i n a reefs enhances the growth rate of the coral and facilitates the greater species diversity of the associated faunal communities compared to the inner-shelf reefs . INTRODUCTION Fl ori vari c High da a osa on th water epi so year, few w upwel Novem event prima fish and T targe ture prima Genti Ocul i menta epi so e sh ree di c whe eeks 1 ing ber s of ry P popu arge ted on k ry s en , This na T- pa die lat re c Lesu elf, fs ( upwe reas dur occ to A f ce rodu lati tt, as t el p ourc 1982 pap reef rame upwe i tud ompr eu r ran Reed llin the ing urre pri 1 ntra cti o ons 1983 he p beds e of )• er c s on ters llin e cora i sed o 1820. ging f , 1980 g of c nears the su d al on , and 1 Flor n with (Atkin ). Up rimary off S n u t r i 1 re f a Dis rom )• old hore mmer 9 th Smi t i da . a c son , wel 1 det . Af ents ef s sing June near The and ree . Y e so h (1 Up onco et ing ermi ri ca for i n t 1 e s t po shor shel nutr f s a oder uthe 981) wel 1 mmi t al., and nant (Wu the he we pecie pul at e sha f-edg i ent re af , et ast U docu ing i ant i 1978 downw of t Iff a Grea stern s of b ions o 1 1 ow r e reef rich w f ected aH. (1 . S . ov mented s know ncreas ; Paff e 1 1 i n g he ben nd Fie t Barr Atl ant r a n c h i f the eef s t s are ater t by up 983) r er 50% summe n to c e i n z enhbf e event thi c c Id, 19 i er Re 1 c o ng c cora o sh subj hrou wel 1 epor of r up ause oopl r, 1 s ha ommu 83) ef ( f f cen oral , 1 are elf-ed ect to ghout ing on ted th the ti w e 1 1 i n i ncre ankton 980; A ve bee n i ty s and as Andrew tral Ocul i n a present ge deep- the ly a at me from g ased and tki nson n truc- a s and ompares the growth and community structure of the inner shelf where cyclic and seasonal environ- predominate, to reefs on the outer shelf where g occurs year round. METHODS AND MATERIALS Colonies of Ocul i na reef (27°30'N, 80°17'W) van cosa and with were sampled with SCUBA at the 6 m the Johnson-Sea-Link submersibles at 119 the offsh at each s traps , th the 6 and Vari o total of height we determine bif urcati branch di cal i cal d coral 1 i te c o r a 1 1 i t e branch si rel ati ve Growt (1981). plastic b periodica The f colonies; s i tes . A by placin The decap ore r i te , ermog 80 m us me 37 co re me d: ti on) , amete iamet ) , in and ze wa tip 1 h rat Li nea ands I ly w aunal 2-4 II th g a 0 od co eefs at but noti raphs , c sites f asuremen 1 o n i e s . asured a p 1 ength tip di am r (measu er (maxi tercal i c the rim s the ra ength wa es of 44 r growth and Al iz e i g h i n g communi colonies e moti le .5 mm me mmuni ty 27, 42 ceably u rrent or one ts of c For ea nd an a ( di sta eter ( d red at mum wid al di st of the ti o of s the r c o 1 o n i rates ari n dy the cor ty was were c , epili sh Nyte i s desc and dama mete year oral ch c vera nee iame 1-2 th f ance near mean ati o es w of b e . al 1 samp olle thic x ba ri be 80 m. C ged colo rs , and ol oni es were nies were re 1 i ght meters mor ol on ge f from ter cm i rom (me est bra of ere ranc Call n s i led cted , an g ov d by phol y (c ore the of a ncre rim asur neig nch mean comp h ti f i ca tu . by a eve d en ere Ree ogy w oral 1 ach o apex bran ments to ri ed be hbori di a me ti p ared ps we ti on ere det urn) , ma f the f of a b c h tip throug m of a tween t ng cora ter to 1 ength and des re meas rates w col 1 ecti on ry 2-3 month dolithic fau ach colony d d, et al . (1 randomly selected jected. Sediment were deployed at ermined from a ximum width and ollowing was ranch to the first at i ts mi dpoi nt ) , hout a col ony ) , polyp ' s cup , i.e., he cup rim of a 1 1 i te) . Rel ati ve coral 1 urn width, and to tip diameter, cribed by Reed ured by use of ere determined by of a total of 42 s at the study na were captured uring collection. 982) . RESULTS Habitat and hydrology T consi paral of al rock occur (80 m 5-25m The c and d T sedim edge (0-31 ( i ncl i nund T nears upwel appar Betwe major edge cause Gibso area he inne st of c lei the gae, sp pavemen s in d e ) , disc rel i ef oral gr ead rub he near e n t a t i o reef (R %) , dep udi ng z ated wi emperat hore (x ling wh e n 1 1 y i en 1977 i ntrus reef in s tempe n , pers shows t r-sh oqui coa onge t i n nse onti and ows ble shor n ra eed , endi ooxa th a ures = 1 i ch ntru and i on di ca ratu . co hat eTfT noi d st. and hibit popul nuous are to 1. areas e ree tes t 1982 ng on nthel turb are 6.2 a produ des o 1979 a yea ted e res t mm. ) 1 eve! 6m) and mid-shelf (27m) reef limestone pavement and ledge Colonies of Ocul i na are i n t e octocoral . At the 42 m site s algal and sponge attachmen ations on isolated patch ree pinnacles form a ridge (Ave covered with colonies of apo 5 m in height and forms exte f is o be )• sur lae) id, sign nd 2 ces nto , th r, 1 pi so o dr on u s of i nf 1 ue 10 tim Transmi ge. Li growth bottom i f i c a n t 4.6°C, a 3-7°C the nea ermogra asti ng die i n t op belo pwel 1 i n nutri e need by es great ttance o ght leve at the n e p h e 1 o i ly col de respecti drop in rshore r phs at t 2-3 week rusi ons w 10°C. g on the nts and strong wa er than t f 1 i ght i Is are to 80 m site d 1 ayer . r at the vely; Ree temperat eefs only he nearsh s. Tempe throughou Unpubl i s outer sh chl orophy s are disjunct and s (1-5 m rel ief ) whi ch rspersed among a cover , a sand veneer over a t ; however, Ocul i na fs. At the shelf edge nt , e_t aj_. , 1977) with symbiotic Ocul i n a . nsive living thickets ve surge which causes hose at the shelf- s variable nearshore o low to support algal which often is shelf-edge reef than d, 1981). Cold-water ure (Smith, 1981) during the summer, ore reef recorded one ratures at the shelf- t the year which hed research (R. elf near this study 11 increased nearly an 120 order of magnitude during these intrusions of cold bottom water (nitrates- <2 yM during non-upwel 1 i ng to 9-18 during upwelling, phosphate- from < 0.25 to 0.5-2 yM, chlorophyll a- from <1 to 1-9 mg/m ). Table 1. Comparison of temperatures (mean and range) and various measurements of Ocul i na v a r i c o s a colonies collected from inner- (6m), mid- (27m), outer-shel f (42m) , and shelf-edge (80m) reefs off central eastern Florida. RTL= relative tip length, CD= intercalical distance, and RBS= relative branch size. Site 6m 27m 42m 80m Temperature (°C) 24.6(13.7-31.0) 18 16 4(8. 0' 2(7.4- 27 26 8) 7) x Colony Size (g) 175 115 1049 3173 RTL 2.29 3.23 4.04 4.84 CD (mm) 1.48 1.67 2.24 2.66 RBS 092 070 039 018 x Growth ? cm/yr g/cm /yr 1.13 1.61 508 651 Corallum morphology Colony size, re general ly showed branch diameter di ameter "f rom th pronounced as de the nearshore po the inner-shelf significantly sh hei ght and has a shelf edge have 1.5 m. Cal i cal significant morphol ogi c the depth r Coral growt rel ati a prog decreas e top t p t h i n c pul atio col onie orter a wide e si gni f i Th shelf- growth but ne shelf- the va be si i Faunal e mea edge was gati v edge r i anc ghtly asso di amet change al char ange of h n growt reef th posi ti v ely cor reef, t e of gr greate ci ation er wa betw acter the h rat an ne ely c rel at emper owth . r at s wit s the only conservative feature and showed no een depths (t-test, p <.05). The other s generally showed a c 1 i na 1 progression over 4 disjunct populations (Table 1). Fo number (N=230 simi 1 a areal Both t N) by and mo 10-11% Bo the co are fa r the s of 0, S = r f or rel at he mo a few 1 1 use of t th as ral , cul ta 42 cor i ndi vid 50; Ree these i onshi p Husks abunda an taxa he tota sembl ag r e q u i r i t i v e as al s~a ual s d, et two a s wi t and d nt sp were 1 ind es ha ng 1 i soci a e of arsho orrel ed wi ature Ini the 8 hOcu mpl es (N = 51 al . , ssemb h the ecapo ecies rare i v i d u ve f e ve co tes t Ocul re { ated th s , cu tial 0 m 1 i na 7Th" 32) 198 lage cor ds w (5- (N als w sp ral hat i na was signi f icantly Reed, 1981). At the with water temperatu edimentation and curr rrent and 1 i ght accou studies showed cal ci site than at 6 m (Tab e mo and 2). s i n al, ere 12% < 10 coll eci e ti ss are 1 1 usks taxa ( Commu terms and be numeri of S). ) ; the ected . s that ue and more d were consi S=230) than nity struct of dominan tween-stati cally domin Over 50% se species are o b 1 i g a mucus. Mo ependent on greater at the nearshore si te , re and i nsol ation ent. At the nted for 84% of fication rates to le 1). derably richer in the decapods ure, however, was ce-di versi ty , on similarity, ated (65-70% of of the decapod comprised only te symbionts with st of the taxa the size of the 121 dead th 49% of amount had a s due to vol ume the cen The di f f ere nearsho edge si f aunal decapod R e c i p r o 1982). decapod the por to the contras feeders on the sites, similar types s The den amount abundan live co an t the of d tron dead of 1 tral ass nt a re r te h si mi s an cal The s, t cell hi gh t, t sue size For at uch s i t i of d ce o ral . he live vari ance ead cora g relati coral s i v i n g s p portion embl ages t the i n eef and ad great larity b d 3.2% f Averagi n se two r he nears ani d era t u r b i d i he shelf h as the of the the mol 6 and 80 as coral es of th ead cora f food i port of 1 pr onsh i ze . ace s di of ner- shel er s etwe or t g al eef s hore b Me ty a -edg pag dead lusk m, 1 i vo e ca l; t terns i ons N and esent ip, wi The fort e . both shelf f-edg peci e en th he mo so sh show reef gal ob of the coral . S, respect iv (Reed, et. aj_ th 52 and 79% live coral , he associated the m and e ree s di v e 6 a Husk owed ed di was rachi nd wa e ree uri d cora s , pe but t res , rni vo hese that ve su f s we crabs 1 whi rcent he 80 paras res w possi were ollu shel f s w ersi nd 8 s . stro sti n domi urn s^ rge re d . T ch w ages m s i tes ere bly ava sks f-ed ere ty ( 0 m The ng s ct t nate oria and dec ge reef both sp Bri 1 1 ou sites w ordi nat ite sep r o p h i c d by fi turn; th whi c omi n hese ere of ite and sign were ilab h keep ated by specie used fo filter had mor carni v i f i c a n t respon 1 e with apod s. A eci e i n ' s as o i on arat part Iter ese food det s we r de feed e sp ores ly c di ng dea s we ltho s ri H). nly tech i on i ti o fee may i n ri tu re m trit ing eci a tha orre to d co re st ugh t ch, t The 5.7% ni que (Reed ni ng . ders be re suspe s and ore d al co taxa 1 i z e d n the lated the g ral r r i k i he he s per for of , et Fo such spon nsio dep epen llec were fee 6 m wi t reat athe ngly helf- cent the al . , r the as di ng n . In osi t dent tion di ng site, h the er r than DISCUSSION The Fl ori da the yea only oc peri ods respons upwel 1 i cycl es reefs w negated The tip 1 en to diff upwel 1 i in calm cal i cal areas o cal i ces excreti the str deep-wa L o p h e 1 i confine (Tei che epi Str r, b curs i n es , ng e of t here by cli gth , eren ng . (de rel f hi app on o uctu ter a an ntal rt, sodi ai ts ut a dur this if a vent empe as o epi s nal and ces For ep o ief gh t ears f me re o Loph c upw inun ppare i ng t repo ny, o s. H ratur n the odi c chang i nte i n wa exam r she i s mo urbi d to b tabol f the el i a d Oc sTTe 1958 u 1 i n a 1 ves ; Ree ell in dates ntly he su rt we f the oweve e, li shel upwel e i n real i ve en pie, 1 tere re ef ity ( e rel i tes shel cora 1 occu where d, 19 g of the furt mmer re t cor r, g ght f-ed 1 i ng cora cal ergy many d) e f i ci Hubb ated (Yon f-ed ban rat wat 80). cool , nutr shel f-edge her onshore (Atkinson, oo long to al or fauna eneral tren and turbidi ge reefs se 1 1 urn dist and cor nvi r ent ard to ge, ge 0 ks , ere er w morpho ance) o 1 ight al s bee onments in remo and Poc e f f i c i e 1973; W c u 1 i n a may be sts of ells up ient rich water from the region frequently throughout movement of upwelled water et ajk , 1978). The sampling distinguish any immediate 1 communities to individual ds may be inferred. Seasonal ty predominate at the nearshore asonal cycles are partially logy (i.e., branch diameter, f 0 c u 1 i n a most likely is due levels but is not a result of ome more spindly and fragile (Stoddard, 1969), high ving sediment in nearshore ock, 1972), and density of ncy of light gathering and ijsman-Best, 1974). However, reefs, which are similar to a result of upwelling; both escarpments or at edges of and around promontories 122 Th where theore popul a water subseq aj_. , 1 shelf- parti c growth pul sed has st other reefs contro questi Di nearsh the mi The hi stable surrou from a Ocul i n of san fact , that t resul t regi on upwel 1 e ap temp ti ca tion onto uent 978; edge ul at at , oc ored fact al so lied on . sti n ore d-sh gh s the ndi n lgal a^ re d, an a 13 he h of . T ed w osymb eratu lly s whi c the i ncr Paff Ocul e org 80 m curri up t ors s coul envi ct as and s elf r peci e rmal g hab -spon ef si d tern .4°C igh s the u he sp ater i o t i c res a houl d h pos shelf eases enhof i na . am c shows ng du he nu uch a d be ronme pop vera hav sess edg of er , The mate a p ri ng trit s gr impo ntal ulations of Ocul i na at the shelf-edge reefs, ge 8.4°C cooler than the nearshore reefs, e lower growth rates than the shallow es zooxanthel 1 ae . Upwelling of nutrient rich e of the study area (Gibson, pers. comm.) and phytopl ankton and zoopl ankton (Atkinson, ejt 1980) may enhance growth rates of the shelf-edge coral must rely on dissolved and rial and plankton for nutrition. Even though ositive response to temperature, growth may be warmer non-upwel 1 i ng periods after the coral ion gained during the upwelling. Of course eater sedimentation rates on the nearshore rtant. Short term growth studies under conditions are necessary to resolve this semb helf eef s s ri envi i tat ge, te, pera vari peci pwel eci f to t 1 ages -edge and chnes ronme hete rock , howev ture ati on es di 1 ing i c pa he co of bot reefs . d i v e r s i s at th nt. Al rogenei and sa er , is f 1 uctua was re versi ty which i thways ral and h de Th ty i e ne so t ty; bell an i tion cord on nduc of e f au capods a ese two s greate arshore he nears thus, th a r i i d wo sol ated s from u ed wi thi the shel es incre nergy tr nal comm nd mo reefs st at reef hore e spe rm re pinna pwel 1 n a 2 f-edg ased ansf e u n i t i 1 1 usks occur on the have more species than the shel f-edge reef, may be due to a more reef has greater cies pool may be drawn ef biotopes. The 80 m cle surrounded by miles ing are extreme; in 4 h period. I suggest e Ocul i na reefs is a producti vi ty of that r from the nutrients of es remain unknown. ACKNOWLEDGEMENTS R.M. Avent is gratefully acknowledged for initiating these studies on Ocul i na . H.L. Edmiston did much of the research on the morphology of the Ocul ina corallum. P.M. Mikkelsen assisted with the molluscan research. CM. Hoskin, R.W. Virnstein, and K.J. Eckelbarger are thanked for their critical reviews of the manuscript. This is Contribution No. 355 from Harbor Branch Foundation, Inc. LITERATURE CITED Andrews, J.C., & P. Gentien. 1982. Upwelling as a source of nutrients for the Great Barrier Reef ecosystem: a solution to Darwin's question? Mar. Ecol . Prog. Ser. 8:257-269. Atkinson, L.P., G.A. Paf f enhbf er , & W.M. Dustan. 1978. The chemical and biological effect of a Gulf Stream intrusion off St. Augustine, Florida. Bull. Mar. Sci. 28:667-679. Atkinson, L.P., & T.E. Targett. 1983. Upwelling along the 60-m isobath from Cape Canaveral to Cape Hatteras and its relationship to fish distribution. Deep-Sea Res. 30:221-226. Avent, R.M., M.E. King, & R.H. Gore. 1977. Topographic and faunal studies of shelf-edge prominences off the central eastern Florida coast. Int. Revue ges. Hydrobiol. 62:185-208. 123 Hubbar sc ti Paffen su Bu Reed, va 30 Reed, 0c wi 2: Reed, CO de CO Smith, At Hy Stodda Bi Tei che As Wi j sma (F Co Wulff , pa CO Yoder , 19 ou Re Yonge, Bu d, J. I erac on . hbfer mmer II . M J.K. ri cos : 667- J.K. ul i na thout 201-2 J . K . , mposi capod ral r N.P. 1 anti drody rt, D ol . R rt,' C soc . n-Bes avi id ral R F.V. thway ndi ti J. A. 83. ter s s. 1: CM. 11. M A. , & Y. t i n i a n c Geol . Ru , 6. A. hydrogra ar . Sci . 1980. a_ coral 677. 1981. vari cos on 80-m 06. R.H. Go tion , st s associ eefs. Bu 1981. c coast nami cs , .R. 196 ev. 44:4 . 1958. Petrol . t, M. 1 ae) and eef Symp , & J.G. s i n a n ons . Mar , L.P. A Effect o outheast 385-404. 1973. ar . Sci . P. Pococ oral s : a nd. 61:5 1980. Z phi c con 30:819- D i s t r i b u reefs of In situ a occurr banks . re , L . E. ructure , ated wit 11 . Mar. An i n v e of Fl ori p. 79-98 9. Ecol 33-498. Cold- Geol . 42 974. Ha its cons . 2:217- Field. earshore . Ecol . tki nson , f upwel 1 ern Unit k. 1972. Sediment rejection by recent key to pal aeo-envi ronmental reconstruc- 98-626. ooplankton distribution as related to ditions in Onslow Bay, North Carolina. 832. tion and structure of deep-water Ocul i na f central eastern Florida. Bull. Mar. Sci. growth rates of the sci eracti ni an coral ing with zooxanthel 1 ae on 6-m reefs and Proc. Fourth Int. Coral Reef Symp. Scotto, & K.A. Wilson. 1982. Community areal and trophic relationships of h shallow- and deep-water 0 c u 1 i n a vari cosa Sci. 32:761-786. stigation of seasonal upwelling along the da. Proc. Twelfth Int. Liege Coll. Ocean ogy and morphology of recent coral reefs. and deep-water coral banks. Bull. Amer. : 1064-1082 . bitat induced modification of reef corals equences for taxonomy. Proc. Second Int. 228. 1983. Importance of different trophic community under upwelling and downwelling Prog. Ser. 12:217-228. S.S. Bishop, E.E. Hofmann, & T.N. Lee. ing on phy topi ankton productivity of the ed States continental shelf. Cont. Shelf The nature of reef-building (hermatypic) corals 23:1-15. 124 CHAPTER V: THE ORGANIZATION OF CORAL REEF ECOSYSTEMS 1 • L • * -A II. %aAw vlfo.-: 1 \ •^ ^^ ^*t®kr z&af**' V € i * ^ ^1 .'■^ fc - m ^r-' %3 a»jfr^dffbS^JM •-' :---Jfl£> ^ —JMk 125 NET PRODUCTION OF CORAL REEF ECOSYSTEMS S. V. Smith Hawaii Institute of Marine Biology P.O. Box 1346, Kaneohe, Hawaii 96744 INTRODUCTION How do coral reefs survive as rich and diverse ecosystems generally found in nutrient-deficient water? This fundamental question has directed much research on the community metabolism of coral reefs towards two ancillary questions: 1. Does the high biomass of coral reef communities indicate high gross organic carbon production? 2. Are coral reef communities net importers or exporters of organic material? If coral reefs were found to produce organic carbon very slowly, then their survival in nutrient-deficient waters would be more easily understood; their nutritional requirements would be minimal. However, early work (Sargent and Austin, 1949, 1954; Odum and Odum, 1955) through to recent research (see references in Kinsey, 1979; Smith, 1981) has consistently demonstrated that the gross production rate of these reef communities is relatively high. The best estimate for the average gross production rate of these communities is 7 gC m_2 d"l , not an entry to the Guinness Book of Records, but impressive nevertheless. The organic and inorganic nutritional requirements of any biological system are determined by the difference between the amount of organic material produced and that which it consumes. We therefore turn to the second question. Can we learn something about the survival mechanisms and ecosystem requirements of coral reefs by considering their net trophic status? The flow respirometry model advanced by Sargent and Austin (1949) and popularized by Odum and Odum (1955) to study the community metabolism of coral reef flats was an elegant approach to community metabolism. Much has been done to refine the model and to adapt it to communities besides reef flats. Probably to the detriment of the original question about the survival of reefs in nutrient- deficient waters, the flow respirometry model and its extensions have focused attention on the metabolic performance of coral reef communities, not coral reef ecosystems. By now we know that individual coral reef communities can be net producers or consumers of organic material, by substantial margins in either direction (Smith and Marsh, 1973; Kinsey, 1979). However, net community production is not directly relevant to the question of net ecosystem performance, because we know that the transfer of particulate and dissolved materials across boundaries between adjacent autotrophic and heterotrophic communities tends towards an intercommunity trophic balance. Such a tendency towards trophic balance has been demonstrated by examining groups of distinct communities (or zones) on coral reefs (Smith and Marsh, 1973; Kinsey, 1979) and is discussed in some detail for a non-reef ecosystem that is metabol ical ly dominated by seagrass and soft bottom communities (Smith and Atkinson, 1983). This characteristic also has been noted as a general characteristic of adjacent communities in flowing water (Odum, 1956). But do we really have a firm notion about the metabolic balance among the entire consortium of biotic communities which form a coral-reef ecosystem? I 127 believe that the answer to this question in any explicit fashion must be "No!" We simply cannot expect to weight and sum the component estimates for individual communities in most ecosystems to a quantitatively satisfying estimate of net ecosystem production. Neither the sizes of communities within the ecosystems nor the community production rates can be measured with sufficient accuracy or precision to accomplish such a summation to within a few percent. At least such precision is required. RESULTS I have recently revisited a subset of my own research papers (and have literally revisited one site and added another) in an attempt to derive an explicit answer about the net production rate of entire coral reef ecosystems. This subset deals with coral atoll lagoons. The lagoons in question are a special case of atoll lagoons: I have examined lagoons virtually surrounded by land (rather than by the more usual oceanic reef rim). Within the lagoons are reef flats, slopes, and inter-reef communities--al 1 of the essential units of more classical "open" coral reefs. The advantage of these confined lagoonal reef ecosystems over more open reef ecosystems is the clear definition of ecosystem boundaries and fluxes. These lagoonal reef systems apparently function as self-contained entities with little communication (with respect to carbon flow, at least) with the oceanic reefs beyond the enclosing islands. The sites being considered are Fanning (Smith and Pesret, 1974, plus new data), Canton (Smith and Jokiel, 1978), and Christmas (Smith, et aK , 1983), all in the central Pacific Ocean. I will also include some relevant discussion with respect to a more open atoll lagoonal system in the Indian Ocean (the Abrolhos Islands; Crossland, et al. , 1983) and a seagrass-dominated coastal lagoon in Western Australia (Shark Bay; Smith and Atkinson, 1983). These last two systems provide additional insight into material processing at the scale of enti re ecosystems. I have used water, salt, and CO2 budgets to examine the net metabolism of these confined lagoonal systems for clues about the total system metabolism of coral reef ecosystems. Shark Bay provides the clue that my suggestions about the metabolism of coral-reef ecosystems may be generalized to other well-defined ecosystems with little input from the surroundings, and the Abrolhos lagoon is an anomaly which strengthens our understanding of the mechanisms controlling these systems. The initial work at Fanning and Canton resulted in values for CaC03 production and total CO2 flux, but failed to separate organic metabolism from gas flux across the air-water interface. In that early work, we were initially very conservative about the gas transfer coefficient. Expressing this coefficient in terms of "piston coefficients," we assumed a possible range between 0 and 20 m d"l. More recent assessment of available data suggests that at wind speeds below 7 m s~* (i.e., about the median wind speed on most coral reefs), the piston coefficient lies in the much narrower range of 0.3 to 3 m d~l (Smith and Atkinson, 1983). If the average piston coefficient is applied to the data set previously mentioned, we derive the following organic carbon production rates (Table 1): 128 Table 1. Organic carbon production (+) or consumption (-) rates for atoll lagoon reef ecosystems. Note that the calculated rates reflect the budgets adjusted to the presently estimated piston coefficient based on other budgetary data presented in the original reference citations. Location Rate Reference mgC m"2 d~l Canton 1973 170 Smith and Jokiel (1978) Christmas 1983 70 Smith, et al_. (1983) Fanning 1983 0 Smith (unpublished data) Fanning 1972 -10 Smith and Pesret (1974) It should be noted that Canton, by its proximity to equatorial upwelling and because of its ocean/lagoon exchange characteristics, has the highest nutrient concentration and loading of these sites. Moreover, it is clearly demonstrated in Table 1 'that this site exhibits the highest net ecosystem production observed according to this model. Shark Bay, although not a coral reef, is another confined ecosystem showing a very low net production rate (15 mgC m_2 d_l ; Smith and Atkinson, 1983). The Abrolhos lagoon is an open atoll lagoon surrounded by oceanic reefs. The slopes of these oceanic reefs have a very large standing crop of typically temperate- water macroalgae (kelp), much of which gets swept into the lagoon during storms (Crossland, ejt a]_. , 1983). Based on dissolved nutrient export from that lagoonal reef system, we concluded that that system is marginally heterotrophic. We could not quantify the degree of heterotrophy there, although research on that question continues (B. Hatcher, personal communication). DISCUSSION Based on this very limited set of data from a particular subset of coral reefs, I suggest that the net organic carbon metabolism of coral reefs and related ecosystems in low-nutrient waters is very low, probably generally averaging well less than 100 mgC m"2 d-l . The available data show a slight bias towards net autotrophy, but it seems likely that individual systems can be influenced to swing one way or the other. This swing will reflect variations in availability (and lability) of inorganic and organic materials used for metabol ism. The important conclusion to be drawn from these calculations does not ultimately concern the trophic status of coral reefs and related systems. Rather, note the very low absolute values of net organic carbon flux in these ecosystems. Let us compare, for example, the net nutritional requirements of coral reefs with the requirements of the surrounding ocean. Eppley and Peterson (1979) defined "new production" of plankton to be that production in the surface waters supported by external nutrient inputs, mostly from upward nutrient flux out of the aphotic zone. Those authors estimated an ocean-averaged new production rate of about 30 mgC m-2 d_l , within the range for net coral-reef production (Table 1). 129 Coral reefs and other ecosystems dominated by benthic plant production apparently have a nutritional advantage over plankton. Rather than approximating the Redfield C:N:P ratio of 106:16:1, benthic plants have a C:N:P ratio of about 550:30:1 (Atkinson and Smith, 1983). The nutritional implication of this observation is that reefs and other benthos-dominated ecosystems can produce more net C per unit of P and N availability than can plankton systems. Moreover, it is well established (e.g., Webb, et aU , 1975; Wiebe, et al . , 1975) that coral reef communities can fix large amounts of N. Recent work suggests that entire reef ecosystems, as well as other confined aquatic ecosystems, can provide most of their N requirements via N fixation (Smith, et a!., 1983; Smith and Atkinson, submitted). Thus, in any net sense, the nutritional requirements of coral reefs do not differ greatly from the requirements of the plankton in the surrounding ocean. We need not look for upwelling or other exogenous nutrient sources to explain the survival of coral reefs. This conclusion does not imply that coral reef metabolism will not respond to increased subsidy of inorganic or organic nutrients. To the contrary, Canton and the Abrolhos appear metabol ical ly responsive to inorganic and organic subsidies, respectively. Direct evidence of reef metabolism response to increased nutrient loading has been documented (Kinsey and Domm, 1974), and both community metabolism and community structure of a reef system have been demonstrably altered by long-term elevation of the supply of inorganic and organic nutrients (Smith, et a]_. , 1981; Brock and Smith, 1983). We can therefore return to the original question about reef survival in nutrient-poor waters. Despite the rapid metabolic activity of components within coral reefs, the net nutritional demands of entire coral-reef ecosystems are low. The key to the success of coral reefs as biologically rich and diverse entities would appear to be the accumulation of a large biomass into a network of biotic communities which are effective at preventing leaks to the surrounding ocean. LITERATURE CITED Atkinson, M. J., and S. V. Smith. 1983. C:N:P ratios of benthic marine plants. Limnol . Oceanogr. 28: 568-574. Brock, R. E., and S. V. Smith. 1983. Response of coral reef cryptofaunal communities to food and space. Coral Reefs 1: 179-183. Crossland, C. J., B. G. Hatcher, M. J. Atkinson, and S. V. Smith. 1983. Dissolved nutrients of a high latitude coral reef, Houtman Abrolhos Islands, Western Australia. Mar. Ecol . Prog. Ser. (in press). Eppley, R. W., and B. J. Peterson. 1979. Particulate organic matter flux and planktonic new production in the deep ocean. Nature 282: 677-680. Kinsey, D. W. 1979. Carbon turnover and accumulation by coral reefs. Ph. D. thesis, Univ. Hawaii. 248 p. Kinsey, D. W., and A. Domm. 1974. Effects of fertilization on a coral reef environment—primary production studies. Proc. 2nd Coral Reef Symp. (Brisbane) 1: 49-66. Odum, H. T. 1956. Primary production in flowing waters. Limnol. Oceanogr. 1: 102-117. Odum, H. T., and E. P. Odum. 1955. Trophic structure and productivity of a windward coral reef community on Eniwetok Atoll. Ecol. Monogr. 25: 291-320. Sargent, M. C, and T. S. Austin. 1949. Organic productivity of an atoll. Trans. Amer. Geophys. Union 30: 245-249. 130 Sargent, M. C, and T. S. Austin. 1954. Biological economy of coral reefs, Bikini and nearby atolls, Marshall Islands. U. S. Geol . Surv. Prof. Paper 260-E: 293-300. Smith, S. V. 1981. The Houtman Abrolhos Islands: carbon metabolism of coral reefs at high latitude. Limnol. Oceanogr. 18: 106-120. Smith, S. V., and M. J. Atkinson. 1983. Mass balance of carbon and phosphorus in Shark Bay, Western Australia. Limnol. Oceanogr. 28: 625-639. Smith, S. V., and M. J. Atkinson, (submitted). Phosphorus limitation of net production in a confined aquatic ecosystem. Nature. Smith, S. V., S. Chandra, L. Kwitko, R. C. Schneider, J. Schoonmaker, J. Seeto, T. Tebano, and G. W. Tribble. 1983. Chemical stoichiometry of lagoonal metabolism. Univ. Hawaii/Univ. South Pacific Internat. Sea Grant Programme Tech. Rept. (in press). Smith, S. V., and P. L. Jokiel. 1978. Water composition and biogeochemical gradients in the Canton Atoll lagoon. Atoll Res. Bull. 221: 15-53. Smith, S. V., W. J. Kimmerer, E. A. Laws, R. E. Brock, and T. W. Walsh. 1981. Kaneohe Bay sewage diversion experiment: perspectives on ecosystem responses to nutritional perturbation. Pac. Sci. 35: 279-395. Smith, S. V., and J. A. Marsh, Jr. 1973. Organic carbon production on the windward reef flat of Eniwetok Atoll. Limnol. Oceanogr. 18: 953-961. Smith, S. V., and F. Pesret. 1974. Processes of carbon dioxide flux in the Fanning Island lagoon. Pac. Sci. 28: 225-245. Webb, K. L., W. D. DuPaul , W. Wiebe, W. Sotille, and R. E. Johannes. 1975. Enewetak (Eniwetok) Atoll: aspects of the nitrogen cycle on a coral reef. Limnol. Oceanogr. 20: 198-210. Wiebe, W. J., R. E. Johannes, and K. L. Webb. 1975. Nitrogen fixation in a coral reef community. Science 188: 257-259. 131 FUNCTIONAL ASPECTS OF NUTRIENT CYCLING ON CORAL REEFS Alina Szmant Froelich Rosenstiel School of Marine and Atmospheric Science University of Miami 4600 Rickenbacker Causeway Miami, Florida 33149 ABSTRACT Coral reef waters are generally very low in nutrients, yet benthic productivity is among the highest in the world. A general concern of reef ecologists has been to identify the sources of nutrients for reef productivity and the processes responsible for nutrient recycling on reefs. One potentially important process that has been poorly studied is the recycling of regenerated nutrients from sediments and feces trapped in reef holes and crevices. These nutrients, if available to the benthos in concentrated form, could have a major influence on reef productivity. INTRODUCTION Coral reef communities are extremely diverse and variable, and we should not expect that the relationships between reef organisms or the dynamics of the system to be simple to understand. On the other hand, by analyzing these complex systems in terms of functional groups, we may gain a better understanding of important processes, and identify research areas that need our attention. In this paper, I will try to identify processes and functional groups involved in nutrient dynamics on coral reefs. Understanding nutrient dynamics on coral reefs is important because, as in any other ecosystem, primary production is controlled by the availability of nutrients, and primary production is one of the main factors that determines community biomass and secondary production. Coral reefs occur in tropical areas, where oceanic waters are generally very low in nutrients, yet these complex ecosystems have some of the highest biomasses and productivity that have been reported (Lewis, 1977). This dilemma can been termed the 'paradox of the coral reef. There has been a great interest in identifying mechanisms that could explain this paradox. Two important ones that have been well studied are the recycling of nutrients between algal/invertebrate symbionts and the fixation of N2 by blue- green algae and bacteria. Another approach has been to identify additional sources of dissolved nutrients, such as from ground water seepage or upwelling of nutrient enriched deepwater. A final approach has been to examine the reef ecosystem for structural and functional properties that lead to enhanced conservation of nutrients. This is the approach that will be emphasized here. SOURCES OF NUTRIENTS There are two general sources of nutrients for primary production: New nutrients and regenerated (recycled) nutrients. If only regenerated nutrients were available, gross photosynthesis could not excede respiration (i.e., P/R ratios could not excede 1.0) unless there was a change in elemental ratios (e.g., an increase in the C/N and C/P ratios). Net production (growth) requires the input of new nutrients into the system, as does net export from the system. It is important to point out that nutrient recycling mechanisms, even when 100% efficient, cannot supply nutrients for a positive net production. Furthermore, if recycling mechanisms are inefficient it will take an input of new nutrients to maintain a steady state biomass. 133 New Nutrients: New nutrients can enter reef systems from both terrestrial and oceanic sources and, in the case of nitrogen, by in situ N2 fixation. Nutrient input to reefs from terrestrial sources remain poorly studied. However, we can make several general statements. Coral reefs located off the coasts of high volcanic islands or on continental shelves may receive a considerable supply of nutrients via terrestrial runoff (Marsh, 1977). However, coastal areas with high runoff are usually devoid of coral reefs unless estuarine systems, such as mangrove forests, trap the nutrients near shore. It is not clear at this time whether the negative effect of runoff on reef development is due to siltation stress, salinity stress or nutrient stress (in the latter case, high nutrient concentrations can promote high phytoplankton and macroalgal growth rates, which in turn can result in unfavorable environmental conditions for reef-building corals). Most likely, all three stresses contribute to the effect. In any case, the on-shore current patterns that generally dominate coral reef circulation should limit the amount of terrestrial runoff that reaches most offshore reefs. One interesting phenomenon that has been reported recently is the submarine seepage of NO.j enriched fresh water inshore of a fringing reef (D'Elia, et al., 1981; Johannes, 1980). It is unlikely that this kind of seepage will reach offshore reefs^ and there is little or no information on how frequently this phenomenon occurs. Reefs located offshore of low carbonate islands should receive little input of terrestrial nutrients because runoff from these terrains are usually low in nutrients unless fertilizer has been applied for agricultural purposes. Atoll reefs generally have no significant land masses nearby from which to receive terrestrial runoff. Oceanic sources of nutrients will be dependent on the concentration of nutrients in the source water, the rate of flow of water over the reef, and the ability of primary producers to take up the nutrients at the given concentration. Nutrient concentrations in tropical oceanic waters are generally near limits of detection, but there are reports that upwelling may occasionally result in higher concentrations (Thompson and Golding, 1981; Andrews and Gentien, 1982). The rates of water flow over the reef are high, and several species of reef coral have been found to be able to take up nutrients from these low concentrations (Franzisket, 1974; D'Elia, 1977; D'Elia and Webb, 1977; Webb and Wiebe, 1978). Atkinson (1981) has estimated that there was a sufficient uptake of dissolved phosphorus from water crossing the reef flat to account for the measured primary production (&C>2) if a Redfield ratio (C:P) of 490 was assumed (the oceanic C:P ratio is 106). He also found that marine algae from a variety of sources had C:N:P ratios much higher than the Redfield ratio (Atkinson and Smith, 1983), a possible indication of an evolutionary adaptation to low nutrient conditions. An alternative explanation is that macroalgae need a greater amount of C-rich structural material than phytoplankton, which results in higher C:N:P ratios for the former. It should be instructive to measure the C:N:P ratio of other reef organisms, especially the microcrustaceans, which have body structures similar to planktonic organisms. In any case, a quick calculation shows that for the following average conditions: dissolved inorganic nitrogen (DIN) concentrations in the range of 0.5 to 1.0 /jM, currrent velocities in the order of 0.06 to 0.50 cm/s, annual gross productivity of 3,220 gC • m -2 . y- 1, P/R ratio of 1.4 and net productivity of 950 gC • m-2 . y-1, the first m2 of reef substrate to make contact with oceanic water would need to strip out 0.08 to 1.3% of the DIN to support net production and maintain a C:N of 6.6. This assumes a mixed layer 1 m deep; twice that percentage would have to be taken up if the organisms can only strip nutrients from the bottom-most 0.5 m of water. Uptake of nutrients to support all of gross production (i.e., if no recycling occurred) would require stripping somewhere between 0.25 to 8.5% of DIN. This means that for many reef flats, where the water column is usually less than a meter deep, waters flowing over downstream areas may be significantly depleted in nutrients. In fore-reef zones, vertical mixing should prevent this type of depletion from being as important, but it is still apparent that at low concentrations and at slow current velocities it is unlikely that uptake of new oceanic nutrients could account for much more than net productivity. Otherwise, we should see a much greater drop in DIN and dissolved 134 inorganic P concentrations during upstream-downstream experiments (Pilson and Betzer, 1973). In fact, DIN and dissolved organic N (DON) concentrations generally increase as oceanic waters cross the reef, and a net export of N has been found for some reefs (Johannes, et al., 1972; Webb, et al., 1975) . This implies a source of fixed N from within the reef community, which can be attributed to N2 fixation by benthic blue-green algae and N2 -fixing bacteria (Mague and Holm-Hansen, 1975; Burris, 1976; Capone, et al., 1977; Wiebe, et al., 1975). Rates of N2 fixation on coral reefs have been found to exceed those of alfalfa fields, the terrestrial community with the highest reported rates of N2 fixation. These high rates have led many investigators to conclude that fixed N is plentiful on reefs, and that concentrations of P must be controlling reef productivity. However, measurements of N2 fixation generally have been restricted to shallow reef flats or back reef areas and denitrification rates have not been adequately measured in any reef environment. Until they are, any conclusions about N availability are premature. Regenerated Nutrients; It generally is believed that the main evolutionary adaptation to low nutrient conditions in reef environments has been the evolution of relationships that lead to efficient recycling of nutrients. The foremost example of this type of relationship is the endosymbiosis between algae and invertebrates. Present day coral reefs are physically dominated by a variety of orders and classes of coelenterates and virtually all of them have symbiotic dinoflagellates (zooxanthellae) in their tissues. It has been repeatedly demonstrated that these animals do not excrete waste products as do other nonsymbiotic animals (Kawaguti, 1953; Pomeroy and Kuenzler, 1969) and that there is even a measurable uptake of dissolved nutrients by them attributable to the presence of the algae (see earlier references). Other invertebrate groups, including sponges, molluscs and ascidians, also have some species with algal symbionts. This form of recycling is the most efficient possible (often 100%) as the nutrients are available to the algae in concentrated form. It should cost the algae much less energy to take up the nutrients they need from a concentrated source than to take them up once they had been excreted and diluted. As important as this type of relationship may be, there are still many groups of algae and many more of vertebrates and invertebrates living on the reef that are not involved in endosymbiotic relationships. These plants and animals need nutrients and excrete nutrients, respectively, and depending on their biomass these fluxes may be quite large. There are few estimates of the absolute and/or relative biomasses of reef organisms. Odum and Odum (1955) found that zooxanthellae make up roughly 15% of the biomass of primary production on a Enewetak reef flat and that coral polyps make up about 50% of the biomass of consumers. However, they had very poor estimates for the biomass of reef fishes and their samples were from the reef flat, which is topographically and biologically less diverse than most forereef areas. Measurements of biomass on a Caribbean fore reef (Szmant-Froelich, 1972) show a greater percentage of the algal biomass to be made up of zooxanthellae and a smaller percentage of the consumer biomass to be made up of symbiotic coelenterates. In order for the reef as a whole to be efficient in recycling nutrients, there must be mechanisms for recycling nutrients among these free-living plants and animals. The main problem arises when one considers that the same high water flow over the reef that assures a large source of low-nutrient oceanic water also assures that any nutrients excreted into the water by animals will be rapidly diluted and carried away. Therefore, what is needed is a mechanism to prevent dilution and loss. I would like to bring attention here to a little studied mechanism, that of particle entrapment and nutrient regeneration within the reef framework. REGENERATIVE SPACES Coral reefs are riddled with holes and tunnels of all sizes. From 50 to 75% of the reef volume can be made up of these voids (Ginsburg, 1983). These holes contain varying amounts of sediment which comes from a variety of sources, including carbonate sediments 135 generated by degradation of the reef structure by borers, fecal material from fishes and invertebrates that use these holes as shelters or encrust the walls of the holes, and non- reefal material (including terrestrial and pelagic) that is trapped inside the reef as seawater percolates through the porous structure. Organic materials in these sediments are metabolized by microorganisms, and in the process, nutrients are regenerated. Elevated concentrations of nutrients have been measured in waters from reef cavities (Di Salvo, 1969; Andrews and Muller, 1983; Szmant-Froelich, 1983). I suggest that 'burps' of nutrient enriched water exit these holes and provide benthic primary producers with short episodes of exposure to higher nutrient concentrations. During Hydrolab Mission 83-10, whose objective was to study the role of herbivorous fish in nutrient regeneration, we measured nutrient (Nrty, N03 and total dissolved N) concentrations of surface and bottom waters at two sites, one near shore ("Habitat" site), the other less protected and more offshore (East Slope site), as well as waters from about 400 m offshore of the reef, and from inside caves at the East Slope site (Table 1). The results of these measurements show that reef water is 3 to 4 times higher in NO3 and slightly higher in NH4 and organic N concentrations than oceanic waters. The most dramatic difference in nutrient concentration, however, can be seen between the offshore water and the cave water. Cave water concentrations are 13 times higher in NO3, 2 times higher in NH4 and 3 times higher in organic N than offshore waters. These enrichments in the caves represent a signficiant increase in nutrients for any primary producers that might have access to them. Dye injections into caves showed that there was rapid outwelling of cave waters onto the reef, and, importantly, that these cave waters flowed within 1 m of the bottom for 10-15 min or longer before mixing upwards. This indicates that there is a process that restricts vertical mixing and dilution of nutrient enriched outwelled cave waters, such that benthic primary producers would have sufficient time to strip the nutrients from these waters before they mix upwards. Corollary evidence that 'burps' of enriched water exist near the reef bottom were obtained from fish excretion experiments during which fishes were incubated inside PVC pipes. Concentrations of NO, were always constant during each incubation but varied considerably between incubations (Table 1). Since these were short incubations (30 min ) replicated repeatedly with new fish during 4 to 5 hour periods, and conducted at various locations over the reef, the differences in NO, concentration reflect true spatial and temporal differences in NO, concentrations of reel bottom waters. The rapid outwelling of the dye from the caves further indicates that rates of nutrient regeneration in cave sediments must be fairly high in order to maintain the observed enrichments in spite of the high water flow. Table 1. Mean nutrient concentrations at various sites in Salt River Canyon, St. Croix (Hydrolab Mission 83-10) during August 4-10, 1983 (mean + 1 std. dev.). Depth No. Sarpling uM Site (m) periods NH^ ND3 DIN DCN Offshore 0 2 0.40 + 0.04 0.07 + 0.03 7.7 + 0.2 7.2 15 2 0.28 + 0.06 0.07 + 0.04 8.2 + 0.4 7.8 East Slope 0 8 0.49 + 0.19 0.24 + 0.12 7.8 + 1.4 7.1 15 17 0.45 + 0.27 0.30 + 0.13 9.2 + 3.3 8.4 "Habitat" 0 9 0.57 + 0.25 0.27 + 0.05 9.0 + 1.8 8.2 15 16 0.53 + 0.24 0.31 + 0.11 8.6 + 1.8 7.8 Caves 15-30 15 0.67 + 0.26 0.93 + 0.42 24.6 + 11.0 23.0 Fish Incu- bation 15 49 - range 0.14 - 0. 40 *" 136 Nutrient regeneration also occurs in lagoon sediments (Entch, et al., 1983) where nutrients are utilized by extensive macroalgae and turtle grass beds. These beds are heavily grazed upon by reef fish. Many of these fish exhibit diel migration patterns, whereby they graze on the backreef and in the lagoons by day but shelter in the deeper forereef by night. Material that they transport in their guts at dusk is defecated over the forereef or in their nocturnal shelters. This form of transport may be an important means of upstream nutrient recycling on coral reefs. The reverse cycle of migration (nocturnal feeding, daytime sheltering) by juvenile grunts has been shown to have measureable impact on nutrient concentrations around coral heads (Meyer, et al., 1983). PATHWAYS OF NUTRIENT TRANSFER DIRECT TRANSFER PARTICULATE ' DISSOLVED DETRITUS » FEEDERS Figure 1. Diagramatic representation of nutrient pathways on a coral reef. FUNCTIONAL GROUPS The various processes and functional groups involved in coral reef nutrient dynamics are diagramatically represented in Figure 1. New nutrients enter the system in both dissolved and particulate form, or are generated in situ by N2 fixation. Dissolved nutrients and some particulates are taken up by the organisms included in the "fine-pore concentrators" group made up of algae and zooxanthellae. Other particulates are trapped by the reef framework and by filter-feeding organisms, which I call the "coarse-pore concentrators". Planktivorous fish also concentrate particulates and have been shown to excrete and defecate significant amounts of NH4 and organic material in their nocturnal shelters (Bray, 1982). Herbivores graze on the algae and corals, and carnivores, in turn, feed on the herbivores. The fecal material from both of these groups, many of which spend about one-half of their time sheltering in reef crevices, 137 are deposited either into reef crevices or released just above the reef surface, where it rains onto what can be viewed as a benthic wall-to-wall carpet of mouths. I have observed particles of cardinal fish feces being consumed by Agaricia agaricites and by other fish. Fish feces contain a lot of mucus, and I believe that fish fecal material is the source of much of the "organic-aggregate" material reported from reef waters. Fecal particles that are eaten form a direct link in nutrient recycling. They are an important nutritional source for corals and other sessile invertebrates that previously has not been taken into account when considering questions of food availability. Regenerated nutrients outwelling from reef interstices form an indirect link in nutrient recycling on the reef, possibly a critical one in maintaining the high rates of reef productivity. LITERATURE CITED Andrews, 3.C., and P. Gentien. 1982. Upwelling as a source of nutrients for the Great Barrier Reef ecosystems: a solution to Darwin's question? Mar. Ecol. Prog. Ser. 8: 257- 269. Andrews, 3.C., and H. Muller. 1983. Space-time variability of nutrients in a lagoonal patch reef. Limnol. Oceanogr. 28:215-227. Atkinson, M.J. 1981. Phosphate metabolism of coral reef flats. Ph.D. Dissertation, Univ. Hawaii. 90 pp. Atkinson, M.3., and S.V. Smith. 1983. C:N:P ratios of benthic marine plants. Limnol. Oceanogr. 28: 568-574. Bray, R. 1982. Role of planktivorous fishes as nutrient importers into a tropical reef community. Hydrolab Mission 82-3, Final Report. Burris, R.H. 1976. Nitrogen fixation by blue-green algae of the Lizard Island area of the Great Barrier Reef. Aust. 3. Plant Physiol. 3: 41-51. Capone, D.G., D.L. Taylor, and B.F. Taylor. 1977. Nitrogen fixation (acetylene reduction) associated with macroalgae in a coral-reef community in the Bahamas. Mar. Biol. 40: 29-32. D'Elia, C.F. 1977. The uptake and release of dissolved phosphorus by reef corals. Limnol. Oceanogr. 22: 301-315. D'Elia, C.F., and K.L. Webb. 1977. The dissolved nitrogen flux of reef corals. In: Proc. Third Int. Coral Reef Symp., Miami 1: 325-330. D'Elia, C.F., K.L. Webb, and 3.W. Porter. 1981. Nitrate-rich ground water inputs to Discovery Bay, 3amaica — a significant source of N to local coral reefs? Bull. Mar. Sci. 31: 903-910. Di Salvo, L. 1969. On the existence of a coral reef regenerative sediment. Pac. Sci. 23: 129-135. Entsch, B., K.G. Boto, R.G. Sims, and 3.T. Wellington. 1983. Phosphorus and nitrogen in coral reef sediments. Limnol. Oceanogr. 28: 465-476. Franzisket, L. 1974. Nitrate uptake by reef corals. Int. Revue ges. Hydrobiol. 59: 1-7. Ginsburg, R. 1983. Geology inside coral reefs. Assoc. Is. Mar. Lab. Carib., May, 1983, Miami: 40 (abstract). 3ohannes, R.E. 1980. The ecological significance of the submarine discharge of groundwater. Mar. Ecol. Prog. Ser. 3: 365-373. 3ohannes, R.E., et al. (22 others). 1972. The metabolism of some coral reef communities: a team study of nutrient and energy flux at Enewetak Atoll. Bioscience 22: 541-543. Kawaguti, S. 1953. Ammonium metabolism of the reef corals. Biol. 3. Okayama Univ. 1: 171-176. Lewis, 3.B. 1977. Processes of organic production on coral reefs. Biol. Rev. 52: 305-347. Mague, T.H., and O. Holm-Hansen. 1975. Nitrogen fixation on a coral reef. Phycologia 14: 87-92. Marsh, 3. A. 1977. Terrestrial inputs of nitrogen and phosphorus on fringing reefs of Guam. 138 Proc. Third Int. Coral Reef Symp., Miami 1: 331-336. Meyer, J.L., E.T. Schultz, and G.S. Helfman. 1983. Fish schools: an asset to corals. Science 220: 1047-1049. Odum, H.T., and E.P. Odum. 1955. Trophic structure and productivity of a windward coral reef community on Eniwetak Atoll. Ecol. Monogr. 25: 291-320. Pilson, M.E.Q., and S.B. Betzer. 1973. Phosphorus flux across a coral reef. Ecology 54: 581-588. Pomeroy, L.R., and E.J. Kuenzler. 1969. Phosphorus turnover by coral reef animals. Proc. Sec. Conf. Radioecol: 474-482. Szmant-Froelich, A. 1972. The zonation and ecology of Jobos Bay coral reefs. Ann. Report PRNC-162 (Puerto Rico Nuclear Center): 174-224. Szmant-Froelich, A. 1983. The role of herbivorous fish in the recycling of nitrogenous nutrients on coral reefs. Hydrolab Quick Look Report 83-10. Thompson, R.O.R.Y., and T.J. Golding. 1981. Tidally induced "upwelling" by the Great Barrier Reef. J. Geophys. Res. 86: 6517-6521. Webb, K.L., W.D. DuPaul, W. Wiebe, W. Sottile, and R.E. Johannes. 1975. Enewetak Atoll: Aspects of the nitrogen cycle on a coral reef. Limnol. Oceanogr. 20: 198-210. Webb, K.L., and W.I. Wiebe. 1978. The kinetics and possible significance of nitrate uptake by several algal-invertebrate symbioses. Mar. Biol. 47: 21-27. Wiebe, W.J., R.E. Johannes, and K.L. Webb. 1975. Nitrogen fixation in a coral reef community. Science 188: 257-259. 139 CONTRASTS IN BENTHIC ECOSYSTEM RESPONSE TO NUTRIENT SUBSIDY: COMMUNITY STRUCTURE AND FUNCTION AT SAND ISLAND, HAWAII S. J. Dollar Hawaii Institute of Marine Biology Department of Oceanography University of Hawaii INTRODUCTION Controlled technological events may provide a well-defined framework for experi- ments that demonstrate community metabolic response to large-scale nutrient sub- sidies (Smith, et al_. , 1981). Sand Island, Oahu is an Hawaiian site where a major change in sewage discharge technology has presented an opportunity to compare response patterns of two yery dissimilar benthic communities to essentially the same effluent. From 1955 to 1977 the Sand Island sewage outfall discharged approximately 2.3 x 10^ m^ raw effluent per day from a point source discharge at a depth of 10 meters. Prior to discharge the receiving environment was a coral reef typical of those off leeward areas of Hawaii. Two research programs were conducted at the shallow discharge site, one in 1975 while sewage was still being discharged (Grigg, 1975) and one in 1979, 1.5 years after sewage diversion (Dollar, 1979). In 1977, in response to federal mandate, the Sand Island discharge was shifted to a newly constructed multi-port deepwater outfall. The receiving environment, at a depth of approximately 70 m is a homogeneous calcium carbonate sandy substratum. For the first 5 years of operation effluent was discharged raw, while for the last two years effluent has undergone primary treatment. In 1981, an ongoing research program designed to examine the response of the benthic ecosystem to the Sand Island outfall was begun. The questions posed in this study were based on the functional metabolic approach developed by Smith, et a]_. (1981): can a total system response be ascertained from a description of nutrTent addition, community meta- bolism and resultant nutrient fluxes? The purpose of this paper is to present a summary of results and a short discussion of the contrasting effects to the benthic ecosystems resulting from the nutrient subsidies from the Sand Island sewage out- falls. Of particular interest in this regard are the contrasting approaches to defining ecosystem response by community structure and community function analyses. RESULTS Shallow Outfall The 1975 and 1979 benthic surveys at the shallow Sand Island Outfall consisted of series of line transect and photographic quadrats used to estimate quantita- tively the effects of sewage on macrobenthos at 28 stations ranging from 3000 m east to 11,000 m west of the outfall. A clear pattern of community alteration associated with the sewage impact was distinguished by two distinct zones; an acute impact zone included approximately 4 km2 of bottom and was characterized by a total lack of living reef coral. The epibenthos was dominated by large mounds of deposit feeding worms (Chaetopterus sp.) apparently adapted to the high organic particulate loading from the outfall. The zone of intermediate impact covered approximately 20 km2 in an asymmetric pattern reflecting prevailing current patterns which transported the sewage laden plume to the southwest. Coral mortality was high, but not total in this area. Macrobenthic species diversity was highest in the zone of intermediate impact due to the co-occurrence of species found in normal, unstressed communities and those species directly associated with the particulate loading. 141 Following sewage abatement, the pattern remained essentially the same, and was reflected in the physical condition of the benthic surface, as well as in species distribution (Dollar, 1979). In the acute impact zone the reef platform was re- duced to a pitted flat limestone pavement by biological/chemical erosional activity. The dense aggregations of polycheate worms were totally absent. Termination of the heavy fallout of particulate organic material emanating from the outfall resulted in a complete change in trophic community structure in the acute impact zone. Apparently qualitative alteration of the limestone surface affected the recon- ditioning process that is a precursor to hermatypic coral colonization. Deep Outfall Preliminary observations revealed that diversion of sewage to the deep water discharge resulted in none of the severe community structure alterations that were apparent at the shallow site. No changes occurred either by removal of organisms from adverse environmental conditions or additions of new species in response to increased nutritional loading. In addition, there were no apparent changes in water column productivity and nutrient characteristics in the vicinity of the out- fall (Laws and Terry, 1983). However, results of benthic nutrient flux experiments performed with the Hawaii Undersea Research Laboratory Submersible Makali ' i showed a very distinct pattern of metabolic community response to nutrient subsidy from sewage loading. Table 1 summarizes the results of the deep Sand Island outfall study and includes for purposes of comparison similar data from a Hawaiian estuary, Kaneohe Bay (Harrison, 1981) and a deep ocean site underlying an area of intense upwelling off the coast of California known as the Patton Escarpment (Smith, et a_l_. 1979). Results of oxygen uptake experiments indicate that within a horizontal distance of 25 meters from the diffuser, metabolic oxygen uptake varies from levels of the deep ocean (3800 m) to an estuary subjected to high levels of terrigenous input from stream runoff and recycled sewage-derived nutrients. NUTRIENT FLUX (uM M-2day-1) STATION DEPTH DISTANCE (M) FROM OUTFALL (M) O.FLUX (mM K_2day-1) NH+ 4 PO/. SEDIMENTATION RATE-ORGANIC C (grams M~2day_1) % ORGANIC C IN BOTTOM SEDIMENT X SEDIMENTED ORG. CARBON OXIDIZED SAND ISLAND 72 5 -16.9 s-7.4 n-13 2209.0 s-1186 n=13 196.5 6-82.9 n-13 222.4 s-139.2 n-13 1.20 S-.55 n-10 0.60 S-.06 n=6 12.9 SAND ISLAND 72 15 -13.2 8-6.1 n-8 328.2 s-111.7 n-8 347.3 s-237.6 n-8 -12.45 s-62.77 n-8 1.30 s-,26 n-8 0.57 s-,03 n=6 9.3 SAND ISLAND 72 25 -4.4 8-4.5 n-9 319.0 s-173.0 n-9 216.2 8=51.1 n-9 56.0 s-14.4 n-9 0.93 s-,52 n-9 0.76 S-.08 n=7 4.3 SAND ISLAND 72 5000 -1.3 s-4.0 n=6 90.2 s-59.2 n=6 50.4 s-8.3 n=6 -14.6 s-27.7 n=6 0.40 s=0.2 n=6 0.80 s-O.OS n=4 3.0 KANEOHE BAY (presewage diversion) 10 500 -18.3 8-4.6 n-17 1857 s-958 n-17 182 s-93 n-17 110 s-68 n-16 0.49* 8-. 22 3.2* 34.4 KANEOHE BAY (post sewage diversion) 10 - -11.9 s-6.6 n=16 959 s-556 n=15 218 s-161 n-15 82 8=6.0 n-15 PATTON ESCARPMENT 3815 ■- -2.4 6-. 7 n-3 283.2 640.3 -5.4 1.22 TABLE 1. Summary of results of Sand Island nutrient flux and sedimentation rate studies. Positive fluxes indicate efflux from sediments; negative fluxe6 indicate uptake by sediments. Kaneohe Bay data is from Harrison (1981) Patton Escarpment, North Pacific data is from Smith et. al. (1979). 142 Table 1 also shows the amount of organic carbon that was intercepted in sedi- ment trap deployments as a function of distance from the outfall. While oxygen uptake (carbon utilization) increased by a factor of 13 between control and the 5 m outfall station, organic sedimentation increased by only a factor of 3. Organic content of bottom sediment remained relatively constant throughout the stations. All fluxes of nitrogeneous dissolved substances, NH^"1", NO3" plus N02~, were posi- tive, indicating release to the water column from the sediments. Mean NH4+ fluxes were highest near the outfall, and decreased with distance from the source of nutrient particulates. Ammonium flux at the station nearest the diffuser (2.2 mM m^day-l) was very close to the ammonium flux measured in Kaneohe Bay (1.86 mM m2 day-'). Nitrate plus nitrite did not show the same distinct progressive decrease in magnitude with distance from the outfall. Phosphate fluxes showed even higher variability since even the direction of flux changed from station to station. Such high variability appears to indicate that net phosphorous exchange is not significantly influenced by the effluent discharge. Data from the City and County of Honolulu show that the ratio of particulate total N to P in the sewage effluent is about 5.5, while planktonic organic material has a N:P ratio of 16:1 (Ryther and Dunstan, 1971). The ratio of total N to P fluxed from the sediment at the Sand Island stations was 8.5. The low flux ratios of N:P may be due to metabolism of exogeneous sewage rather than natural sedi- menting organics. However, Nixon (1981) also has measured N:P ratios of 7:1 in benthic fluxes in Naragansett Bay, Rhode Island. He theorizes that the low nitrogen fluxes are due to sequential nitrification of ammonium produced in aerobic metabolism, diffusion of nitrate and nitrite into the lower anaerobic sediment column, and subsequent denitrification to di-nitrogen. The end product is lost from the dissolved fixed nitrogen pool, and may be the cause for nitrogen limitation in some shallow marine systems. Approximately 1()6 moles of particulate organic carbon are discharged each day from the outfall. Extrapolation of sediment trap fluxes to the total area of im- pact around the outfall structure indicate that only about 10^ moles of this par- ticulate carbon reaches the benthic surface. If these fluxes are of the correct order of magnitude, over 99.9% of the discharged material is dispersed in the water column under the normal regime of tide and current flow. While the propor- tions of material cycled within the benthic boundary layer are infinitesimally small compared with total effluent volume, it is significant that there is any increased- signal at all indicating that benthic nutrient metabolic process are very sensitive to nutrient inputs. DISCUSSION Both community structure and community function analyses are useful approaches to determining ecosystem response to nutrient subsidy. Structural analysis of benthic macrofaunal community assemblages, including measures of species composition, abundance, diversity, trophic position and biomass, are aimed at determining the ultimate effect of a large scale nutrient subsidy on macrobiota. With this approach, however, there are no provisions to link structural characteristics of the community with changes in nutrient parameters via examination of metabolic pathways that are changed by the perturbation. Unless the trophic response of the macrofaunal community is extremely visible, as was the case at the shallow Sand Island outfall, there is little or no understanding of the causative relationships between biotic structure and the altered environment. A community metabolism approach can, however, be used successfully to assess the effect of sewage stress or other nutrient subsidies on the marine ecosystem. Quantitative determination of rates of processes that cycle materials provide a 143 means to estimate the fate and effect of organic material within the framework of the whole system. While this type of functional approach entails a higher level of technology, it may actually be easier and quicker than the purely structural approach, since the latter requires establishment of a statistically valid sampling procedure and tedious examination of large numbers of samples. Thus, assessing and quantifying cycling of natural materials by mass balances derived from the measure of nutrient fluxes may be used to identify material sources and sinks- important information for improving the potential to predict ecosystem response to other similar situations. We have chosen to investigate the biogeochemical processes occurring within the benthic boundary layer because the sediments act as a storage reservoir of both particulate and dissolved nutrient material. Thus, it is at the sediment-water interface that one might expect to see the most subtle or the most intense alterations to ecosystem function. At this point in our research effort, we can already conclude that nutrient regeneration in the sediments and release to the overlying water has been responsive to nutrient loading from the Sand Island Out- fall. REFERENCES Dollar, S. J. 1979. Ecological response to relaxation of sewage stress off Sand Island, Oahu, Hawaii. Water Resources Research Center, Technical Report No. 124. . Grigg, R. W. 1975. The effects of sewage effluent on benthic marine ecosystems off Sand Island, Oahu. Proc. 13th Pac. Sci. Congress. Univ. B.C., Vancouver (abstract). Harrison, J. T. 1981. The influence of Alpheus mackayi on ecosystem dynamics in Kaneohe Bay. Ph.D. dissertation, Univ. Hawaii" 112pp. Laws, E. A., and K. L. Terry. 1983. The impact of sewage discharge at ocean out- falls on phytoplankton populations in waters surrounding the Hawaiian Islands. Mar. Env. Res. 8:101-117. Nixon, S. W. 1981. Remineralization and nutrient cycling in coastal marine eco- systems, p. 111-138. _In.: B- J- Nielson and L. E. Cronin (eds.), Estuaries and Nutrients. Humana Press, Clifton, N.J. Ryther, J. H., and W. W. Dunstan. 1971. Nitrogen, phosphorus and eutrophication in the coastal marine environment. Science 171:1008-1013. Smith, K. L., G. A. White, and M. R. Laver. 1979. Oxygen updake and nutrient exchange of sediments measured in-situ using a free vehicle grab respirometer. Deep Sea Res. 26A: 337-346. Smith, S. V., W. J. Kimmerer, R. E. Brock, E. A. Laws, and T. W. Walsh. 1981. Kaneohe Bay sewage diversion experiment: perspectives in ecosystem responses to nutritional perturbation. Pac. Sci. 35:279-402. 144 METABOLISM OF INTERREEF SEDIMENT COMMUNITIES John T. Harrison, III Mid-Pacific Research Laboratory P. 0. Box 1346, Kaneohe, Hawaii 96744 ABSTRACT Metabolic and nutrient fluxes of lagoon sediment communities at Enewetak, Marshall Islands, were measured using hemispheric incuba- tion chambers. Simultaneous determination of oxygen and carbon dioxide fluxes allowed empirical derivation of an RQ value of 1.5 and a PQ of 0.8. More carbon is respired by the sediment community than is produced, and both production and respiration are closely correl- lated with depth. Biotic and functional comparisons between Enewetak and Kaneohe Bay, Hawaii, suggest metabolic and structural similar- ities between these physiographically disparate coral reef systems. INTRODUCTION Ecosystem analysts increasingly are turning to coral reefs as foci for studies of community metabolism. However, most researchers have concentrated on the conspicuous and highly productive perimeters of reef systems; with a few exceptions (Sournia, 1976; Kinsey, 1979; Harrison, 1981), interreef areas have been largely ignored, despite the fact that these regions usually comprise the substantial areal majority of the coral reef ecosystem. Lagoon floor communities have long been regarded as sinks for the excess production of fringing reef crests. However, neither sediment community metabolic charac- teristics nor overall dimensions of the functional relationship between organic sources and sinks within reef systems have been ade- quately described. Enewetak Atoll in the northern Marshall Islands has been the site of numerous pioneering investigations into coral ecosystem dynamic processes. Perhaps more than any other major system, Enewetak pro- vides an opportunity to achieve an empirical synthesis of all the main components of coral reef ecosystems for an overall budgetary analysis of organic metabolism. There remain only two major system compartments needing detailed description: the lagoon water column, and the lagoon benthos. In this report, I intend to provide suffi- cient data to characterize the latter compartment, and I further hope to provide impetus to encourage pursuit of the final area of uncer- tainty, the lagoon planktonic community. METHODS AND RESULTS Clear acrylic hemispheres (domes) with a radius of 0.5 meters were used to isolate sample areas for metabolic and dissolved inorganic nutrient fluxes. Oxygen concentrations were measured polarographic- ally; carbon dioxide fluxes were calculated from temperature, pH , and alkalinity measurements (Smith and Kinsey, 1978). Preliminary alkalinity determinations confirmed Smith and Harrison's (1977) sug- gestion that net CaC03 production of the sand-rubble component of the Enewetak marine environment is effectively zero. Thus, the majority of C02 calculations were based on temperature and pH measurements alone. Dissolved inorganic nutrient concentrations were determined 145 using standard automated techniques. Incubations at depths down to 35 meters were performed by use of SCUBA. At depths in excess of 35 m, the submersible, Makali'i was used to emplace and sample the domes. Oxygen data were normalized to a half-sine wave model of diurnal production and integrated to esta- blish net production (Marsh and Smith, 1978). For lack of a more accurate estimate, nighttime respiration measurements were extrapol- ated throughout the day. A summary of the data is given in Tables 1 and 2. Due to equipment failures, reliable CO2 data were obtained only for two of the six sampled depth intervals, yet the close cor- respondence between these figures suggests a uniformity of RQ and PQ values throughout the sampled depth range. Table 1. Summary of Enewetak lagoon sediment metabolic data. Depth in meters; respiration and gross production means ± standard deviation. Units: mmoles O2 m' DEPTH n RESPIRATION PRODUCTION 4 10 60 ± 12.5 64 ± 16.7 7 26 58 ± 34.9 64 ± 36 10 12 54 ± 8.6 50 ± 7.8 16 32 36 ± 14.2 36 ± 13.2 27 9 30 ± 7.9 27 ± 9.8 55 5 12 ± 19.8 13 ± 21.5 day _ 1 RQ 1.5 ± 0.6 (n = 21) 1.5 ± 0.6 (n = 23) NET PQ 1.0 ± 0.7 (n = 22) 1.0 ± 0.5 (n = 24) TRUE PQ 0.8 0.8 Table 2 meters ; theses. Units : DEPTH 4 7 10 16 27 55 Enewetak lagoon sediment nutrient fluxes. Depth in mean fluxes ± standard deviation. Sample size in paren- Negative sign denotes uptake by the sediments, micromoles P04 m 2 day - 1 -47 ± 62.3 (7) 46 ± 352.7 (22) 7 ± 16.0 (7) -54 ± 190.8 (25) -2 ± 21.2 (5) -18 ± 88.9 (9) N02 + N03 -7 ± 24.5 (7) 38 ± 111.7 (23) -10 ± 48.4 (7) -7 ± 94.2 (24) -11 ± 61.6 (5) -40 + 53.9 (9) NH, 131 ± 174.9 (6) 206 ± 523.2 (23) 306 ± 563.8 (7) 132 ± 536.3 (25) 86 ± 1132.8 (5) -45 ± 113.2 (9) Oxygen data are depicted in the histogram (Figure 1) by depth. The evident depth trends of both respiration and gross production are not found in the sediment nutrient fluxes (Table 2). In general, nutrient fluxes, particularly phosphate and nitrite + nitrate, are barely above limits of detection. However, as noted in previously studied lagoon floors (Harrison, 1981), both oxygen and dissolved 146 inorganic nutrient fluxes are highly variable i CM CM O l/> 0) o E E | RESPIRATION f~1 GROSS PRODUCTION Figure and gro 3-5 5-8 8-14 14-20 DEPTH (m) 1. Histogram of pooled oxygen data ss production at increasing depths. 20-30 >30 24 hour respiration DISCUSSION Depth dependence of community metabolism. The evident balance between gross production and respiration throughout the depth range of Enewetak lagoon (Fig. 1) is misleading, as will be discussed subsequently. However, the figure clearly ill- ustrates that both parameters are depth dependent. Based on the pooled data, gross production, regressed exponentially against depth, yields a decay coefficient of 0.031 m~ * , with a correlation coeff- icient (r2) of 0.971. Similarly, the respiration decay coefficient is also 0.031 m~ 1 , with r2 equal to 0.985. The close adherence to an exponential model establishes the correlation of sediment community metabolism with incident light. Unfortunately, the lagoon extinction coefficient has not been determined, but in the open ocean immediately east of Enewetak, the extinction coefficient is 0.045 m~ ' (Colin, et al., in prep.). Since the lagoon water column contains more partic- ulate material than does the open ocean, the lagoon extinction coeff- icient cannot be lower than 0.045 m~ l . Thus, although production and respiration of lagoon floor communities are clearly light-driven, benthic community activity is increasingly efficient at greater depths. Lagoon floor heterotrophy . Assuming that the empirically determined values for RQ and PQ are uniform throughout the lagoon floor, production and respiration can be expressed as carbon equivalents. If measured oxygen fluxes are con- verted to carbon equivalents by use of RQ and PQ values, it turns out that more carbon is respired by the system than is produced, and the 147 sediments are distinctly heterotrophic. Although no direct assays have been conducted, the RQ value of 1.5 indicates a substantial com- ponent of anaerobic metabolism in lagoon sediment communities. Kinsey (1979) likewise derived high RQ values for sand-rubble systems, also invoking anaerobic metabolism as an explanation. Net carbon deficits for each depth interval are presented (Table 3) along with depth distribution data from Emery, et al (1954) to derive a mean carbon requirement for Enewetak sediment communities of 41 mgC m-2 day-1 . Table 3. Lago on sec DEPTH (m) AREA (km2) 3-5 22.8 5-8 22.8 8-14 25.4 14-20 50.8 20-30 40.6 >30 769.6 Lagoon sediment community carbon deficits CARBON DEFICIT (mgC m~2 day"1) AREAL DEFICIT (kg C) 121 2,759 74 1,687 216 5,486 110 5,588 138 5,603 22 agoon agoon deficit area 16,931 Total 1 Total 1 38,054 kg C 932 km2 Mean deficit 41 mgC m"2 day - 1 Preliminary calculations based on organic carbon production of wind- ward reefs (Smith and Marsh, 1973) indicate that excess production of these reefs is probably more than sufficient to provide the carbon requirement of the sediment community. However, in the absence of detailed information on metabolism in the lagoon water column, mechanisms and precise characteristics of the trophic relationship between organic sources and sinks within Enewetak Atoll must remain speculative . Macroinf aunal influences. Inspection of the summarized results (Tables 1, 2) reveals substan- tial variability in both metabolic and dissolved nutrient flux data. Previous sediment metabolic surveys have related high levels of para- meter variability to the presence of actively bioturbating infaunal communities in interreef lagoonal systems. In Kaneohe Bay, Hawaii, alpheid shrimps thoroughly mix the upper 30 cm of sediments, greatly enhancing rates of sediment community metabolism and nutrient cycling (Harrison, 1981). The lagoon floor at Enewetak contains an extensive population of callianassid shrimps actively turning over large vol- umes of sediment daily. Callianassid burrow systems penetrate in excess of 2 m below the sediment surface (Colin, et al . , in prep.), providing direct routes for transport of organic cletritus into the sediment column, and suggesting mechanisms for stimulation of the apparently high levels of anaerobic metabolism inherent in these com- munities. In this respect, and in others, there is a remarkable cor- respondence between both structural and functional characteristics of widely disparate reef ecosystems. 148 The substantial physiographic differences between a high island system such as Kaneohe Bay and an oceanic reef like Enewetak are ref- lected in the different levels of organic and inorganic supply to which their respective biota must adapt. High island reefs must with- stand heavy nutrient and organic loading from terrigenous runoff; by contrast, external supply to atolls is minimal. Despite such opposi- tion of material supply regimes, reef structure in the two environ- ments is surprisingly similar. To be sure, systematic and morpholo- gical differences in coral reef communities are prevalent, but both oceanic margins are metabolically and structurally alike, as are the lagoon patch reefs. It would appear that the major metabolic differ- ences between high island and oceanic reef systems are confined to interreef and possibly planktonic compartments. Reliable estimates of RQ and PQ values for Kaneohe Bay sediment communities are not available, so direct comparisons of Enewetak and Kaneohe Bay carbon metabolism cannot be made. However, the near-zero net balance of aerobic metabolism of Enewetak sediments (Fig. 1) contrasts with a daily oxygen deficit at the Kaneohe Bay floor of roughly 9 mmoles m-2. The increased heterotrophy of the Kaneohe sediment community reflects the abundance of systemic material supply. Despite metabolic and visual dissimilarities, the lagoon floors of Enewetak and Kaneohe Bay are functionally and biologically comparable. Such correspondence between major components of disparate reef systems supports the validity of general models of coral reef ecosystems. LITERATURE CITED Emery, K. 0., J. J. Tracey , Jr., and H. S. Ladd. 1954. Geology of Bikini and nearby atolls: part 1, geology. U. S. Geol. Survey Prof. Paper 260- A. Harrison, J. T. 1981. The influence of Alpheus mackayi on ecosystem dynamics in Kaneohe Bay. Ph.D. Thesis. Univ. Hawaii, Honolulu. Kinsey , D. W. 1979. Carbon turnover ana accumulation by coral reefs. Ph.D. Thesis. Univ. Hawaii, Honolulu. Marsh, J. A., and S. V. Smith. 1978. Productivity measurements of coral reefs in flowing water, p. 361-377. In: D. R. Stoddart and R. E. Johannes (eds.), Coral Reefs: Research Methods. Monogr. Oceanogr. Methodol. 5, UNESCO. Smith, S. V. , and J. T. Harrison. 1977. Calcium carbonate produc- tion of the Mare Incognitum, the upper windward reef slope, at Enewetak Atoll. Science 197: 556-559. Smith, S. V., and D. W. Kinsey. 1978. Calcification and organic carbon metabolism as indicated by carbon dioxide, p. 469-484. In: D. R. Stoddart and R. E. Johannes (eds.), Coral Reefs: Research- Methods. Monogr. Oceanogr. Methodol. 5, UNESCO. Smith, S. V. , and J. A. Marsh. 1973. Organic carbon production and consumption on the windward reef flat of Enewetak Atoll. Limnol. Oceanogr. 18: 953-961. Sournia, A. 1976. Primary production of sands in the lagoon of an atoll and the role of f oraminif eran symbionts. Mar. Biol. 37: 29-32. 149