TROPHIC BASIS OF PRODUCTION AMONG RIVERINE CADDISFLIES - IMPLICATIONSFOR FOOD-WEB ANALYSIS

We determined the trophic basis of production and quantified the food web of caddisfly larvae from the submerged woody (snag) habitat of a C oastal Plain blackwater river (Ogeechee River). Production was dominat ed by three net-spinning taxa (Cheumatopsyche spp., Hydropsyche rossi, and Chimarra moselyi), comprising 97-98% of the total among at least 14 trichopteran species. Annual production (as dry mass) was among the highest estimates reported for caddisflies, ranging from 43.5 to 63.9 g/m(2) of snag surface (Or from 14.2 to 25.5 g/m(2) of channel bottom ) in two consecutive years. Although all taxa were present throughout the year, C. moselyi had much higher production in summer, H. rossi wa s highest from late summer to winter, and Cheumatopsyche spp. showed n o pattern. Ingestion was determined from production estimates, bioener getic efficiencies, and quantitative gut analyses. The two macrofilter ing taxa (Cheumatopsyche spp. and H. rossi) were omnivorous, with 50.6 and 64.4% of their production, respectively, due to eating animals. S omewhat lower amounts of their production (40.7 and 23.5%) were due to ingestion of amorphous detritus. Production of the microfiltering spe cies Chimarra moselyi was primarily due to eating amorphous detritus ( 91%). Removal of amorphous detritus from the system by the major taxa was highest in summer, and diatom removal was highest in fall, whereas animals were eaten consistently throughout the year. A quantitative f ood web showed that while the linkages among nine taxa were complex, t he ingestion pathways were dominated by amorphous detritus (total cons umption = 62.5 g.m(-2).yr(-1)) and animal prey (22.3 g.m(-2).yr(-1)), with >99% of food resources being consumed by the three dominant filte r-feeding taxa. A connectivity food web was misleading because it impl ied equivalence of all food resources and consumers when great variati on in strength of linkages actually existed. These comparisons provide a strong argument for considering energy flow as a measure of linkage strength when evaluating food webs, and they cast doubt on the useful ness of oversimplified connectivity webs as a basis for food web theor y.