Seasonal abundance oC tbe demersal copepod Pseudodiaptomus cokeri ( Calanoidea : Pseudodiaptomidae ) in a Caribbean estuarine environment

Seasonal abundance variations of the demersal copepod Pseudodiaptomus cokeri were examinated at three locations (center, north shore and south shore) in Phosphorescent Bay, Puerto Rico, throughout an annual cycle. Nocturnal oblique tows (21 :00-23 :00 hr) were taken biweeldy (three replicates) at each location with a coDicaI net (mouth diamcter = 0.5 m; mesh size = 1 35 11m) and a standard calibrated flowmeter. Water temperature and salinity measurements were taken at the surface and near the bottom in each location before towing activities using a SCT meter. o Chlorophyll-a concentrations were fluorometricaIly determined. The adult and copepodite stages of this species accountcld for approximately 1 .6% of the annual mean total zooplankton abundance of the bayo Higher abundance of P. cokeri (mean:l:l SD = 4 1 9 1 :!: 1 444 iÍldividuals m-3) was associated with �ool water temperatures and dry conditions (cooVdry season) which prevailed betw'ee n December and March relative to the period between April and November (wann/wet seasón) with lower abundance. F1uctuations of this population followed progressive increments in chloro­ phyll-a concentrations at the three sampling statigns (One-way ANOVA, p<O.05). However, abundance was generalIy higher on the north shore of the bay (Tukey's test, p<O.05). Differences in abundance of P. cokeri between sampling sta:­ tions could be related to the different types of substratum found in the bayo Apparently, this species prefers areas with heterogeneous substrata. Its demersal behavior may contribute to the observed distribution inside the bayo

Substantial attention has been devoted to descriptions of the spatial and temporal distribu tion pattems of zooplankton in nearshore envi ronments of Puerto Rico (Coker and González 1960, Suárez-Caabro and Shearls 1972, Youngbluth 1976, Pesante 1979, Zayas 1979, Rojas and Suárez-Caabro 1979, Youngbluth 1980, Yoshioka et al. 1985, García and López 1989, García and Durbin 1993. These studies give information about the physical and biological factors that regulate the development and persistence of these communities. Most of the plankton research in Puerto Rico, however, has been focused at the community level; stud ies on the ecology of particular populations are few and limited to copepod species of the genus Acartia (González 1980, Olivieri-Vega 1987, Santiago 1988).
Among the characteristic zooplankters of the coastal waters of Puerto Rico is the calanoid copepod Pseudodiaptomus cokeri González and Bowman 1965. This species was first col lected in Phosphorescent Bay in 1957, and described later by González and Bowman (1965). It has been subsequently reported in collections from several bays along the south ern coast of the island as Jobos Bay (Youngbluth 1980) and the Guayanilla and Tallaboa Bays (Rojas and Suárez-Caabro 1979). This copepod is resident inside these bays, and it was not found in adjacent neritic waters. Pseudodiaptomus cokeri has been reported also from Jamaica, Puerto Rico, Antigua and St. Lucia (González and Bowman 1965), Trinidad (Bacon 1971), the Atlantic coast of the Canal Zone (Bowman 1978), Venezuela, Panama and Belize (Walter 1989), and Cuba and the Gulf of México (Campos and Suárez-Morales 1994).
The demersal species of the genus, includ ing P. cokeri, tend to remain near, on or in the bottom during the day, but rise towards the water column at night (González and Bowman 1965, Bowman 1978. Demersal zooplankters are usualIy undersampled by traditional plank ton tows performed during daytime. This may be the reason why most studies do not report and possibly underestimate the abundance of Pseudodiaptomus cokeri in Caribbean embay ments, including Puerto Rico. Recent plankton studies using nocturnal tows, emergence traps or diver-towed nets near the bottom have pro duced new information on the abundance and distribution of demersal zooplankton from shallow coastal environments (Ohlhorst 1982, Youngbluth 1982, Jacoby and Greenwood 1988, 1989, Saínt-Jean and Pagano 1990, Buskey 1993. In Puerto Rico, there is evidence of higher abundance of most taxa, including copepods, during night-time collections in sev eral coastal locations around the island (Youngbluth 1979(Youngbluth , 1980. Most of the fieId studies on species of Pseudodiaptomus in tropical and subtropical environments consist of descriptions of their vertical movements or substratum preference (Rart and Allanson 1976, Hart 1978, Hutchins 1985, Jacoby and Greenwood 1988, 1989, Saint-Jean and Pagano 1990. Other research focuses on the diel feeding rhythmicity (Hart 1977). Experimental research using cultures of the three species of Pseudodíaptomus found in the Caribbean, P. acutlls, P. cokeri and P. marshi, have provided important informatíon on their life history and mating behavior (Jacoby and Youngbluth 1983). No ecological work has been conducted on natural popula tions oí P. cokeri. This paper investigates the relationship between environmental factors and the temporal and spatial patterns of abundance of Pseudodiaptomlls cokeri in Phosphorescent Bay, Puerto Rico, throughout an annual cycle. Special emphasis was devoted to describe the variations in abundance of the copepodite and adult stages (e.g. male, female and ovigerous female) of Pseudodiaptomus cokeri.

MATERIALS AND METRODS
The southern coastline of Puerto Rico is sculptured with small bays, which may be broadly open or nearly enclosed, and is gener ally bordered by dense growths of mangroves. Orte of these bays is Phosphorescent Bay, which is located towards the southwest end of the island (17°58'30" N; 67°01'10" W). The bay is a protected estuarine environment with a surface area of ca. 19 hectares and a mean depth of 2.5 m (Coker and González 1960). Ir connects wíth offshore waters through a narrow opening about 158 m wide. A sandy substratum covers most of the shallow borders of the bay together with small patches of turtle grass, Thalassia testlldinllm, and a variety of macroal gal species. Turtle grass, sometimes mixed with manatee grass, Syringodillm filiforme, grows a1so on sections of the south border of the bay. The central area of the bay has fine muddy sediment. This area constitutes the deepest and most extensive portion of the bay, with depths ranging down to 4.5 m. Organic detritus mainly originating from the surround ing red mangroves constitute the majority of this loose muddy substratum. Most of the cen tral portion of the embayment has reduced light penetration with the deeper areas considerably darker. Secchi disk readings vary from 0.9 to 3.4 m (mean = 2.0 m) (González 1967). There is 110 macroalgal or seagrass growth in this area. High turbidity in Phosphorescent Bay has been associated with high phytoplankton con centrations, debris from the mangrove swamp, rainfall runoff and stirring of loose sediments  December. Peak values were associated with extended periods of two or more months of low rainfall in JuIy and January. Low salinity val ues were registered during and after high pre cipitation events in November and December ( Fig. 1). There was a slight difference in water temperature and salinity between sampling sta tions (Coefficient ofVariation of temperature = 1.27 %; C.v. salinity == 0.78 %). No stratifica tion was observed in the water column during the period of study. The greatest difference ..,
The seasénal v�ations in Pseudodiaptomus cokeri,.in PhosPnGtescent Bay appear to be,regulaté(t py interactions . of climatic. �d biolpgical factOI's. Thepattem of P . coketl.abundance coincided' with variations in· chlorophyU-a.:(;oncentrations, with lrigher numberaf-"-���g du� . � coolldry seasGnj�jwhen·chl orophyll�a v�ües were also highesti?mhís f�ct suggest�that¡�Ope1 pod.abundance . wa��at leas��n: part' i� eriktted  by food availability (e.g. phytoplankton). Phytoplankton increased after periods of higher cummulative monthly precipítation in the bay. Exogenous nutríent inputs from rainfall runoff may have a fertilizíng effect on phytoplankton either ímmediately or after a tÍme lag (Purcell 1980, García and López 1989, Jordan et al. 1991. Nutrient enríchment associated with the seasonal rainfall patterns is a key factor regulat ing temporal varíations in standing stocks of phytoplankton and zooplankton (mostly cope pods) in coastal environments of Puerto Ríco (García and López 1989) and the U.S. Virgin Islands (Purcell 1980). Nutríent remineraliza tíon after the rainy season was also considered an important factor of plankton abundance in Biscayne Bay, Florida (Woodmansee 1958, Reeve 1964) and in Laurel Reef, La Parguera, Puerto Rico (Glynn 1973). During the short perlod of lower rainfall between December and March, nutríent ínputs from the watershed prob ably declíned in Phosphorescent Bay. Phytoplankton standíng stocks, however, prob ably remained high duríng this period due to autochthonous remineralization processes. The abundance decline of Pseudodiaptomus cokeri duríng the warmJwet season may be explained by a decrease of phytoplankton standing stocks, as indicated by chlorophyll-a concentrations. Salínity appears to play a mínor role in influencíng temporal distribution pat terns of this species in Phosphorescent Bay. No significant relationship was found between saliníty and total P. cokeri abundance. The neg ative assocíation between temperature and total P. cokeri abundance reflects the in verse rela tíonshíp existíng between temperature and phytoplankton (chlorophyll-a concentration).
Results of this ínvestigaríon support the hypothesís that food limitatíon (e.g. phyto plankton standing stocks) regulate the abun dance of Pseudodiaptomus cokeri in Phosphorescent Bay. The concept of food limi tation has been supported by laboratory experi ments and field observations which have shown that growth and reproduction in copepods (e.g. Acartia tonsa) can be lirnited by food concen tratíon (Checkley 1980, Durbín et al. 1983, Ambler 1985, PaffenhOfer and Stearns 1988. Thus, since mean chlorophyll-a concentration recorded ín Phosphorescent Bay was sígnífi cantly lower duríng the warm/wet season rela tive to the cool/dry season (Student-T test, p < 0.01), low food availabílíty could be expected to have had a limiting effect on herbívorous zooplankton. Probably the fast turnover of phytoplankton was a key factor that permitted the high numbers of P. cokeri In Phosphorescent Bay.
Similar variations in abundance of copepod populations have been reported in other nearshore localities of Puerto Rico (Youngbluth 1979, García andLópez 1989) and complement those described for the total zooplankton abun dance in Phosphorescent Bay (Ríos-Jara, in prep.). Fluctuations in these localities were closely related to the annual rainfall pattern with associated nutrient inputs from land drainage and consequent increases in phyto plankton productivíty. In other embayments of Puerto Rico, P. cokeri is occasionally very abundant (Youngbluth 1979(Youngbluth , 1980. Ovigerous females and copepodites of P. cokeri were always present in the bay, indicating a continu ous reproductive activity. Other species of Pseudodiaptomus (e.g. P. marinus) are listed among the copepods recommended for mass cultivation (Iwasaki and Kamiya 1977). The data from this investigation may be useful in estimating the potential production of P. cokeri in Phosphorescent Bay as a food source for fish larvae. This ínformation may also be useful for determinations of secondary production in trop ical estuarine environments.
The seasonal changes of zooplankton abun dance may be influenced by the abundance of predators as well. Predator-prey interactions have been proved to be important in structuring the zooplankton community (e.g. Fulton 1985, Jenkins 1988, Bamstedt 1990), and a major factor determining the distribution of copepods (Ríppingale and Hodgkin 1974, Iwasa 1982, Fancett and Kimmerer 1985, García and Durbin 1993. Pseudodiaptomus hessei forms a major part in the diet of estuarine fish and rnysids (Wooldridge and Bailey 1982). Escape response through vertical migratíon and a ben thic daytime existence may reduce relative pre dation in demersal copepod populations (Jerling and Wooldridge 1991). Gelatinous zooplankton were sometimes present but inadequately sam pled in Phosphorescent Bay. Sorne ctenophore blooms were detected after heavy rain events in September and October. However, the lack of population estimates of abundance of theseand other possible predators prevent any further considerations in this study.
Dífferences in abundance of P. cokeri between sampling stations could be related to the different types of substratum found in Phosphorescent Bay. Both adult and copepodite stages of P. cokeri were more abundant al sta tions S-2 andJor S-3 (north and south shores, respectively), relative to S-1 (center of the bay) on 19 out of 24 sampling dates. Daytime col lections showed that P. cokeri was most fre quent over the vegetated bottoms found in the shallower coastline around the bay than in the unvegetated mud substratum of the central part of the bayo Apparently, this species prefers areas with heterogeneous substrata. Jacobs (1961) observed that P. coronatus sinks rapidly when not swimming and can c1ing tight1y to substra ta; P. cokeri has a similar behavior (Jacoby and Youngbluth 1983). This has been interpreted as an active mechanism that prevents individuals from being washed out of the system (Castel and Veiga 1990).
Pseudodiaptomus cokeri can be c1assified as endemic, in fue sen se that it inhabits only enc10sed bays, its demersal behavior may con tribute to its restricted distribution inside the bay (González and Bowman 1965). Transport studies in planktic animals have demonstrated that individual s remaining in surface waters are passively transported outside coastal systems by tidal currents to adjacent neritic waters (Fontier and Leggett 1982); whereas fuose remaining in near-bottom waters accumulate inside (Laprise andDodson 1990, 1993). Consequently, vertical migration may be an important mechanism allowing retention of drifting animal s inside coastal semi-enclosed ecosystems. Copepodite and adult stages showed strong vertical migratíon and probably maintained their vertical position in the water column of the bay by means of vertical move ments through the water column (Ríos-Jara, in prep.). This has been demonstrated for severa! coastal specíes (Trinast 1975, Jacobs 1968, Wooldridge and Erasmus 1980, Alcaraz 1983. Restriction to nearshore environments has been explained as a strategy to remain in high-food environments (Paffenhofer and Stearns 1988). Pseudodiaptomus is a typical estuarine-coastal genus, found only in shallow inshore waters, where the food supply is generally abundant (González and Bowman 1965).
There are temporally stable are as of high or low den sities of P. cokeri in Phosphorescent Bay, causal mechanism(s) of these temporal changes in abundance affect differently in the various areas of the bay throughout the year. The effect of these factors are les s apparent dur ing the warrn/ wet season. Thus, it is suggested that P. cokeri displays a small scale patchiness. This pattern is comparable to observations on the spatial distribution of zooplankton abun dance from other coastal (Youngbluth 1980, García andDurbin 1993) and oceanic (Yoshioka et al. 1985) environments of Puerto Rico. Most importantly, these results suggest that abiotic and biotic relationships directly influence abundance of plankton in Caribbean coastal environments and that zooplankton assemblages are not simply the result of species independently sorting in diverse envrronments according to physiological tolerances. A deeper understanding of the relative importance of the factors resposible for the variability observed in abundance requires further studies on the appropiate temporal and spatial scales, sínce the distribution of demersal copepods is heavily dependent on the coupling between the physical and biological processes.

ACKNOWLEDGMENTS
This study is part of a research on the zoo plankton community of Phosphorescent Bay which was submitted in partíal fulfil1ment of the requirements for the Doctor of Philosophy degree in Marine Sciences from the Department of Marine Sciences, University of Puerto Rico in Mayagüez (DMS-UPR). 1 thank the members of my dissertation committee, Juan G. González Lagoa, Jorge R. García Sais, ABen R. Lewis, Stuart Ramos Bíaggi and James S. Beaver, for invaluable help throughout the course of this research. This study was partially supported by the Universidad de Guadalajara, México and the DMS-UPR.