STONER: FEEDING ECOLOGY OF LAGODON RHOMBOIDES 



with increasing food abundance did not hold in 

 certain instances because the abundance of 

 macrobenthic organisms at the study site was 

 closely related to standing crop of benthic macro- 

 phytes at the site (Stoner in press). Also, since 

 seagrasses and epiphytes serve as important di- 

 etary components of larger pinfish, prediction of 

 dietary diversity is further complicated. 



Data provided in this study further verify the 

 conclusion that L. rhomboides is a generalist 

 feeder. Schoener (1969) suggested that generalist 

 feeding strategy is favored when: 1) food density is 

 low and there is a premium on the ability of the 

 animal to take a range of prey, 2) the predator has 

 a relatively long period to gain energy, and 3 ) prey 

 densities fluctuate widely. Given the relatively 

 low abundance of prey species in Apalachee Bay, 

 the great diversity of potential prey items, and 

 their high degree of variability with time and 

 space, the generalist feeding strategy exhibited by 

 L. rhomboides would be predicted by Schoener's 

 model. 



The detailed analysis of the food habits of L. 

 rhomboides provided in this study accentuate 

 difficulties inherent in the description of a trophic 

 niche. Because of dramatic variation in food habits 

 and dietary breadth in coastal fishes, serious 

 methodological problems arise in description of 

 food habits. In most cases, length-frequency dis- 

 tributions offish are not constant with time; con- 

 sequently, when animals are not placed in size 

 classes or when placed in overly large size classes, 

 dietary variation may be due to either seasonal 

 changes in food habits or increasing fish size. The 

 diet of a group of fish will be a function of the 

 length-frequency distribution of the population if 

 variation with size occurs. When food habits are 

 examined by size and not by season, variation 

 within a given size class may appear greater than 

 is actually true at any particular time, and sea- 

 sonality of diet is completely obscured. Variability 

 in food habits as a function of space (a common 

 occurrence) adds still one more dimension to the 

 problem of describing an animal's food and feeding 

 habits, but spatial variation is usually ignored. 

 Keast (1970), in a study of the bioenergetic inter- 

 relationships of cohabiting freshwater fishes in 

 Ontario, provided insight into the complex in- 

 teractions of fish size and season in determining 

 food habits. One other study (Nakashima and 

 Leggett 1975), an investigation of responses of 

 yellow perch to different levels of phytoplankton 

 and benthic biomass in Lake Memphremagog, 



Quebec- Vermont, showed interactions of time and 

 fish size in diet determination. The dimension of 

 space was added by comparing the diets of fish 

 from the northern and southern basins of the lake. 

 Few studies have described more than one dimen- 

 sion of an animal's food habits. 



Peters (1977) stated that the "Trophodynamic 

 Concept" (Lindeman 1942) is based upon the 

 premise that organisms in an ecosystem are 

 categorized according to their distance along a 

 food chain from the sun. He pointed out, however, 

 that the real world is constructed of complex food 

 webs and organisms do not fall into neat 

 categories such as "primary consumer" or "sec- 

 ondary carnivores." This is not a new idea. In 

 1961, Darnell asserted that "trophic level" is an 

 inoperational term since: 1 ) animals of a given size 

 and belonging to a single species take food from 

 several sources, 2) alternate foods are frequently 

 utilized as a function of their availabilities, 3) an 

 ontogenetic progression of food habits is common 

 in animals, and 4) many animals are dependent 

 upon detrital material which is itself of a complex 

 origin and an undefined distance from primary 

 producers. Regier and Henderson (1973) and 

 Kercher and Shugart (1975) provided similar 

 reasoning for the inadequacies of the trophic level 

 concept. Darnell recommended a spectral ap- 

 proach to the food habit problem and Kercher and 

 Shugart defined an "effective trophic position," 

 actually a continuous index of trophic position 

 rather than the conventional discrete level. 

 Neither solution to the problem addressed all of 

 Darnell's objections and gave an accurate por- 

 trayal of the functional role of the organism in its 

 ecological context. Data provided in this paper 

 show that, in addition to ontogenetic pattern in 

 food habits, animals within given size classes take 

 foods from several sources with the possible excep- 

 tion of postlarval pinfish which are collected only 

 in the winter and early spring. Fish of many size 

 classes consumed significant quantities of cala- 

 noid copepods which are probably herbivores; har- 

 pacticoid copepods which may be detritovores, 

 herbivores, or carnivores; amphipods which show 

 wide variety in food habits; plus shrimp, inverte- 

 brate eggs, and many other invertebrate taxa. In 

 most cases, the prey species themselves cannot 

 reliably be placed in any one trophic level and, 

 since individuals of a given species were consumed 

 at different developmental stages, and prey 

 species may show trophic ontogeny, the problem of 

 assigning a trophic level to the predator is further 



349 



