ONEIL and WEINSTEIN: FEEDING HABITATS OF SPOT 



10_ 

 _ 20_ 



> 30_ 



t- 



§40: 



i 50_ 

 60_ 



1/273 



GU 



-DET 

 MISC 



CS 



Lp 



HA. 



NEM 



OST 



HAi 



NER 



PLA 

 SPI 



MISC 



4/380 



GD 



HAo 



Me 



PLA 



Lp 



La 



CS 



HAf 



ZSsZ 



MISC 



J-Et 

 GA 



NEM 



NER 



HA, 



CHL 



OLI 



MAL 



usr 



MISC 



-PLA 

 -Eh 



2/228 

 GS 



2/284 



BU 



NEM 



MAL 



HA, 



CS 



CHL 



OLI 



PLA 



DET 

 "PST' 



MISC 



-LF 

 ■Eh 

 -NER 

 -HA, 



5/277 

 BD 



MAL 



OLI 



NEM 



HA, 



CS 



Eh 



SPI 



"HTTT 



FOR 



MISC 



-ORI 

 ■LF 



17/311 

 BS 



Figure 5. — Cluster analysis of prey similarity among habitats for spot from a polyhaline (BD = Blevins 

 downstream, BU = creek, VA, 1982. Prey abbreviations cire listed in Blevins upstream, BS = Blevins 

 shoal) and a meso-oligohaline (GU = Goalders upstream, etc.) tidal Table 1. Ratios at the bottoms of each 

 column represent number of empty stomachs/total sample. 



tern, however, there were few clear differences 

 caused by presence or absence of particular prey 

 types. Instead, it was more a question of which 

 food item was dominant. Spot appeared to eat 

 more nematodes within the creek and more mal- 

 danid polychaetes on the shoal. 



To confirm the results of the classification anal- 

 ysis, reciprocal averaging was used on the same 

 food-habitat matrix (Fig. 6). Prey items located 

 near a given station "lollipop" are the dominant 

 food items utilized by spot at that station. Axis 1, 

 accounting for 45% of the data variance, clearly 

 separated the low and high salinity creek sys- 

 tems. Axis 2 (28% of the variance) isolated the 

 shoal stations relative to the intracreek sites. Ne- 

 matodes and maldanid polychaetes were again 

 closely associated with the Blevins Creek sites. 

 Nereids, Leptocheirus, and Monoculodes were 

 dominant at the Goalders Creek habitats. 



To compare seasonal patterns in food utiliza- 

 tion between habitats, classification dendro- 

 grams were also constructed using monthly data 

 for each creek (Figs. 7, 8). At Goalders Creek 



there was little overlap of prey utilized in April 

 compared to all other months (Fig. 7). The main 

 reason for this appears to be the large proportion 

 of calanoid copepods consumed in April. August 

 and October were grouped together because of the 

 amount of nereids eaten, and the remaining 

 months were added to this cluster individually, 

 depending on their overall dissimilarity. 



April was also an outlier at Blevins Creek, be- 

 cause of the dominance of calanoids in the diet 

 of young spot (Fig. 8). May and June were 

 clustered together because of the similarity in 

 the consumption of maldanid polychaetes and 

 nematodes. August and September were similar 

 in the proportions of four prey items utilized: mal- 

 danids, nematodes, nereids, and harpacticoid 

 copepods. Although these food items were proba- 

 bly incidental in their diet, July was a separate 

 group because of the large amount of Chloro- 

 phyta present in the stomachs examined from 

 that month; October was isolated because 

 Foraminifera became an important addition to 

 the diet. 



791 



