FISHERY BULLETIN: VOL. 82, NO. 3 



TABLE 3. — Two-way coincidence table comparing stations (Groups A-D) and species 

 (groups I-VI) associations at York River estuary sites. Clustering by unweighted-pair 

 group-average; similarity index = Camberra metric, all data log 10 (x + 1) transformed, 

 single station occurrences dropped. G = Goalders Creek, B = Blevins Creek, U = up- 

 stream, D = downstream, S = shoal stations, W = Wegener ring samples. 



salinity) examined. Nonetheless, within this 

 group were several species which displayed area- 

 specific distributions, in either relative numbers 

 or presence/absence in a given creek system. 

 Examples of the former include the hogchoker and 

 bay anchovy and of the latter, white perch, which 

 was far more prevalent upstream. The Goalders 

 Creek nekton community was also dominated by 

 members of groups I and II, whose members were 

 rare or absent at downstream localities. Remain- 

 ing species were generally not captured in suffi- 

 cient numbers to depict their role in defining 

 community structure in each area. 



Trophic Analysis 



Six species (Fig. 4) were sufficiently abundant in 

 time and space to allow a comparative trophic 

 analysis to be undertaken. Collectively, they ex- 

 ceeded 989r of the total number of individuals cap- 

 tured during this study. Prey taxa were defined on 

 the basis of 39 categories (Table 4). All but two — 

 miscellaneous (MISC) and unidentified (UID) — 

 were mutually exclusive. These two did not exceed 

 179c of the total diet composition of any one species 

 and were generally much lower than this amount. 

 The dietary relationships of these six species are 

 summarized across all sampling strata by the 

 dendrograms appearing in Figure 4. With the ex- 

 ception of summer flounder, Paralicthys dentatus, 



sufficient numbers of individuals were captured to 

 also allow partitioning by size classes. Such on- 

 togenetic summaries are shown in Figure 5. 



Although more than 2,600 specimens were 

 examined for diet composition, sample sizes were 

 not sufficient in the first year of the study to 

 examine details of seasonal nor spatial food utili- 

 zation in all species except spot (O'Neil 1983). In 

 addition, several species were only abundant in a 

 restricted area (Table 1) or attained peak abun- 

 dance in a relatively narrow time frame, preclud- 

 ing dietary comparison of dominant species 

 (Fig. 6). 



Diet overlap was greatest between white perch 

 and hogchoker (Fig. 4). Dietary items shared by 

 these species included clam siphons, Leoptochirus 

 plumulosus, and other gammarid amphipods. 

 Both predators were generally habitat-specific, 

 with young-of-year white perch more abundant in 

 the creeks and hogchoker prevalent on the river 

 shoals (Table 1). Largely because of fish (TEL) 

 included in the diet of larger individuals, white 

 catfish displayed somewhat less overlap in its diet 

 compared with the former species, but because of 

 its partial piscivorous habits shared this similar- 

 ity with the major fish predator captured in our 

 trawl samples — the summer flounder. Both small 

 summer flounder and white perch (<61 mm) also 

 consumed substantial quantities of the mysid 

 shrimp Neomysis americana. 



460 



