General College and
James Ford Bell Museum of Natural History University of Minnesota
Minneapolis, MN 55455

27 March 1998

MATERIALS AND METHODS

RESULTS AND DISCUSSION/div>

The 65 specimens examined in this study consumed 25 different categories of food, including eight insect taxa (the dipterans Ceratopogonidae and Ephydridae were consumed but not listed separately), six taxa of microcrustacea, four taxa of worms (including bryozoa), the larvae of at least one species of cyprinid fish, a variety of filamentous green algae and other vascular plant matter, and organic detritus (Table 1). One specimen (from the upstream site in Elk Creek) not recorded in Table 1 consumed 8 Corixidae nymphs. More Topeka shiners utilized larval chironomids than any other food item, but they never consumed them in large numbers and most of the larvae eaten were small (mean + 2SE = 3.1 + 0.6 mm, N = 28). Shiners included chironomid pupae and simuliid larvae in their diet but in much smaller proportions (Fig. 2A) and only on two dates (Table 1). Proportionately, corixids were the dominant insect food item (Fig. 2A), but they were utilized by only 14% of the fish and mostly in Elk Creek and the Rock River (Table 1). All of the corixids were small nymphs ranging from 0.8 to 1.5 mm long.

Topeka shiners utilized microcrustaceans sparingly, except in August when they consumed large numbers of Bosmina and moderate numbers of two other cladoceran genera (Fig. 2B). A single Chydorus was ingested by a fish from Mound Creek site 2. Carapace lengths for these animals were 0.3-0.4 mm for Bosmina, 0.5-0.8 mm for Ceriodaphnia, and 0.5-0.9 mm for Daphnia. Although ostracods made up almost 20% of the 1268 microcrustaceans consumed, they were eaten by a single Topeka shiner at the same site where most other shiners were ingesting corixid nymphs (Table 1).

We did not quantify all of the plant matter we found, but we believe it was an important physical component of the diet. Seeds of several flowering plant species were important numerically and in terms of frequency of ingestion (Table 1). We measured length and width of each seed type, and based on these measurements as compared to those of chironomid larvae, we estimate the volume of seed ingestion exceeded that of chironomid volume by at least a factor of two. Ingestion of filamentous green algae (Chlorophyta) was infrequent and the filaments few, suggesting that they may simply represent incidental consumption. The other plant tissue ingested included pieces of vascular stem and root. Although the size of these fragments varied, they were routinely as large as or larger than any other food item in a given gut, with the exception the guts that contained larval fish. Further, the soft inner tissues of the vascular plant matter (seeds, stems and roots) very clearly was being digested as it passed through to the distal gut.

More extensive studies are being undertaken to quantify ontogenetic, seasonal, and geographic differences in food use and to determine the relative nutritional value of food items. However, it is clear that Topeka shiners in the Rock River drainage at least utilize many food sources beyond chironomid larvae and other immature aquatic insects. They feed from at least three trophic levels when they include corixid nymphs or larval fish in their diets and must be considered omnivores (Vadas 1990). Further, the consistent occurrence of detritus and sand in the gut, along with dipteran larvae as well as pupae, and the ingestion of Bosmina and Daphnia (planktonic cladocerans) indicates that Topeka shiners function both as benthic and nektonic feeders (Gatz 1979). The inclusion of Ceriodaphnia and Chydorus (substrate-dwelling cladocerans), may also indicate that Topeka shiners feed from the surfaces of plants (Whiteside and Lindegaard 1980).

The overall feeding pattem seen here is similar to that of the sand shiner (N. stramineus), which is the sisters species of the Topeka shiner (Schmidt and Gold 1995). Starrett (1950) studied sand shiners in the Des Moines River watershed of Boone County, Iowa, and considered them to be the only omnivorous generalist among 13 minnow species. Most of the diet consisted of bottom ooze, a variety of aquatic immature insects, microcrustaceans and plant matter. He did not report consumption of fish larvae. Stegman (1960) reported bottom ooze in addition to dipteran larva and other aquatic insects as the major components of sand shiners in Kinkaid Creek, Illinois. The swallowtail shiner (N. procne), which is the sister species to the sand shiner-Topeka shiner clade, is also a generalized omnivore, although its specific diet reflects the streams east of the Appalachians where it is endemic (Vadas 1990). Starrett (1950) noted that generalized omnivores in the Des Moines River were capable of altering their feeding habits easily when faced with potential competition from more specialized minnows. Since Topeka shiners and sand shiners are syntopic at most sites in the Rock River drainage and both appear to be generalized omnivores, it will be important to examine their feeding relationships. Sand shiners have not undergone the same decline in prairie streams that has been noted for Topeka shiners.