FISHERY BULLETIN: VOL. 86, NO. 3 



67°irW). The study site and shrimp and fish pop- 

 ulations in the lagoon have been described by Stoner 

 (1986, 1988). Briefly, Laguna Joyuda is a polyhaline 

 basin with a total surface area of 1.21 km^ and an 

 average depth of approximately 1.5 m. One narrow 

 channel 0.5 km long represents the only connection 

 to the Puerto Rico shelf. Lagoon sediments are 

 mostly fine mud and sand with very high organic 

 content derived from red, white, and black man- 

 groves which surround approximately 75% of the 

 shoreline. Over the last 20 years the lagoon has had 

 a salinity range of 4 to 44''/oo, depending upon pre- 

 cipitation and degree of channel closure (A. W. 

 Stoner, unpubl. data). 



Four penaeids utilize the lagoon as a nursery area. 

 Penaeus notialis and P. subtilis are approximately 

 equal in abundance and together comprise over 92% 

 of the total penaeid assemblage (Stoner 1988). 

 Penaeus brasiliensis made up 7.0% of the total col- 

 lection in 1984 to 1985, and P. schmitti was rela- 

 tively uncommon, making up <0.3% of the total 

 (Table 1). For this study, we examined the diets of 

 the three most abundant shrimp species. 



Penaeids were collected with a 5 m otter trawl 

 with 2.5 cm wings and body, and 5 mm cod end liner. 

 All collections were made between nautical twilight 

 and midnight at three sites: a northern muddy bot- 

 tom arm of the lagoon (station 5), a sandy mud site 

 in the central basin (station 3), and a sandy mud site 

 near the channel (station 1). Collections were made 

 monthly, during the last quarter of the moon, from 

 July 1985 to June 1986. 



Shrimps were divided into size classes on the basis 

 of carapace length (CL) for gut content analyses. 

 Penaeid shrimps between 3 and 6 mm CL could not 

 be identified to species and were simply identified 

 as Penaeus juveniles (all were of the grooved vari- 

 ety and, therefore, did not include P. schmitti). 

 Shrimps larger than 6 mm CL could be identified 

 and were examined by individual species in 4 mm 

 size classes, up to 26 mm CL in the case of P. sub- 

 tilis. To yield sufficient numbers of individuals in 



Table 1 . — Composition of the penaeid shrimp assemblage at three 

 stations in Laguna Joyuda, Puerto Rico, during the 12-mo study 

 period. Values are total numbers collected in 72 trawl samples and 

 percentages of totals at the individual sites (parentheses). 



all size classes for each monthly collection, members 

 of individual species were pooled from all sampUng 

 areas. In the case of P. brasiliensis, collections from 

 2-mo intervals were pooled to analyze seasonal 

 variation in the diets of this less abundant species. 

 Food items taken from the proventriculus of up to 

 25 shrimp were pooled for each sampling date and 

 size class, and preserved with 70% isopropanol and 

 a dilute solution of rose bengal stain. 



We used the gravimetric sieve fractionation pro- 

 cedure developed by Carr and Adams (1972) to 

 analyze gut contents of the shrimp. This procedure 

 has been widely used for juvenile fishes (Sheridan 

 1979; Stoner 1980; Livingston 1984) and a variety 

 of decapod crustaceans, including Penaeus spp. 

 (Laughlin 1982; Leber 1983). Gut contents were 

 washed through a series of six sieves of decreasing 

 mesh size (2.0-0.075 mm mesh) and each sieve frac- 

 tion was examined with a dissecting microscope. 

 Because all of the items in a particular sieve frac- 

 tion were of approximately equal size, the relative 

 proportion of the gut contents made up of each food 

 type was measured directly by counting. After ex- 

 amination, each sieve fraction was dried overnight 

 at 80°C and the total contribution of each food type 

 to total dry weight was calculated. 



With few exceptions, each food particle was placed 

 in a mutually exclusive category (Table 2). In most 

 cases, food items or fragments could be identified 

 to major taxonomic group such as Amphipod or 

 Polychaete. The classification "Animal Remains" 

 was applied where fragments were unidentifiable 

 to taxon, but where the tissue was stained by rose 

 bengal. The major food categories were used for 

 statistical interpretation of diets; however, when- 

 ever an animal or plant could be identified to a lower 

 taxonomic level, this information was recorded. 



Similarities between and among the diets of 

 various shrimp species and size classes were mea- 

 sured with Czekanowski's coefficient (Bray and 

 Curtis 1957; Field and McFadane 1968). Dendo- 



Table 2. — List of the general food categories encountered in the 

 foreguts of Penaeus species and the codes employed in histograms 

 for shrimp diets. 



AM Amphipod OS 



CC Calanoid copepod PM 



CY Cyclopoid copepod PO 



CZ Crab zoea RU 



DE Detritus SA 



FO Foraminifera TA 



FR Fish remains TH 



HC Harpacticoid copepod 



GA Gastropod MS 



IE Invertebrate egg 



NE Nematode 



Ostracod 



Plant material (green) 



Polychaete 



Ruppia maritimia 



Sand 



Tanaidacean 



Thalassia testudinum 



Miscellaneous— used in 

 histograms for all food 

 items making up <4% of 

 the total dry weight. 



544 



