THE ROLE OF TROPHIC BOTTLENECKS IN STUNTING - A FIELD-TEST OF AN ALLOCATION MODEL OF GROWTH AND REPRODUCTION IN YELLOW PERCH, PERCA-FLAVESCENS

Stunting in freshwater fish populations may be due to low availability of one or more prey components within their diet. If the limiting pre y constitute a trophic level (i.e. zooplankton, benthos, or fish), we define the phenomenon as a trophic bottleneck. Growth of a non-stunted population of yellow perch, Perca flavescens, was simulated using an allocation model based on reported ontogenetic shifts in diet (plankti vory to benthivory to piscivory). The model was then perturbed by limi ting the available ration individually for each of the three feeding t ypes. The resulting simulated growth curves all reflected lower growth rates than the unperturbed state and the shape of the curve differed between the limited food types. Only the reduced benthic ration produc ed a simulated growth that matched that observed for the stunted yello w perch of Lac Hertel (Quebec). To test the hypothesis that benthic ra tion was limiting growth in Lac Hertel, all the fish species in the la ke were sampled for diet composition and total length at age. We predi cted that species feeding upon benthic invertebrates should exhibit th e slow growth characteristic of stunting (based on the results of the allocation model). Of the seven fish species found in Lac Hertel, four were stunted: yellow perch, pumpkinseed, Lepomis gibbosus, rock bass, Ambloplites rupestris, and brown bullhead, Ictalurus nebulosus. Three species were non-stunted: northern pike, Esox lucius, golden shiner, Notemigonus crysoleucas, and white sucker, Catostomus commersoni. All stunted species fed on benthic invertebrates and all non-stunted speci es, except the white sucker, did not feed upon benthos. The prediction of the allocation model was thus supported.