Much of the high species diversity and abundance found along riparian corridors is maintained by in-stream secondary production. Although evidence is tangential and is based on visual observations, many riparian species (amphibians, spiders, birds and bats) obtain a large portion of their diet from emerging aquatic insects. Direct evidence, however, for the significance of in-stream production is lacking because prey items are often partially or fully digested, and so gut content analysis is inadequate.

The objective of this study is to quantify the proportion of nitrogen incorporated into obligate riparian insectivores from emerging aquatic insects using a stable isotopic nitrogen tracer (15N). As part of the larger 15N enrichment study, I propose to sample two forested streams (Walker Branch in Tennessee and Hubbard Brook in New Hampshire) and two desert streams (Sycamore Creek in Arizona and Gallina Creek in New Mexico) in summer and fall of 1997. Emergence production will be sampled quantitatively using standard emergence traps and light traps (Smock 1996) at sites upstream and downstream from the 15N addition site prior to, during and immediately after the release. Riparian habitat and terrestrial predators (spiders and salamanders) will be assessed by lateral transect surveys. Obligate terrestrial predators that will be analyzed for 15N incorporation are spiders from the genera Araneus and Meta (Araneidae), Pirata and Pardosa. (Lycosidae), Tetragnatha (Tetragnathidae), Dolomedes and Pelopatis (Pisauridae), and salamanders from the genera Desmognathus and Eurycea (Plethodontidae).

Preliminary data indicate that background 15N levels are similar among prey species (del 15N values from 1-3); hence terrestrial taxa should have an isotopic signal considerably lower than enriched emerging stream insects. After calculating weighted averages of enriched del 15N values from emerging insects, proportion of dietary consumption which comes from terrestrial prey (unlabeled) and that which comes from emergence production (labeled) will be determined. I predict that varying levels of in-stream production and extent and complexity of riparian vegetation (e.g. availability of emergence platforms, attachment sites and habitat refuges) will influence the strength of the riparian/ aquatic trophic link in each of these watersheds.