Sweeney, B.W., and R.L. Vannote. 1986. Ecology 67:1396–1410.
The natural influx of leaf litter from a hardwood forest to sections of two spring seeps and a spring brook was excluded for two years and replaced with equivalent amounts of leaves from specific tree species. Larval growth, mortality, and productivity, as well as the timing of emergence and the size of adults, were assessed in the naturally occurring populations of Soyedina carolinesis, and the differences related to differences in available food and in temperature among study sites. During the first experimental year, larvae fed only on sugar maple or chestnut oak leaves exhibited the same rate and magnitude of growth as larvae fed the natural mix of leaves collected from the surrounding forest. Larval growth rates and adult size seemed lower on monospecific diets of hickory, American beech, and red oak leaves relative to the natural, mixed leaf diets. Adult emergence occurred on or about the same date for most diets. larval production ranged from 1382 to 5500 mg°m–2°yr–1; there was no correlation between larval growth rate and productivity on a given diet. During the second experimental year, most sites that had been supplied with single-species leaf diets during the first experimental year were provided with sugar maple leaves only. This was an attempt to evaluate site-to-site differences in habitat quality (other than diet) that might affect the growth performance of larvae. The amount of variation in larval growth rate and adult size among sites having a common diet was equal to or greater than among-site variation during the previous year, when each site differed in leaf diet. In the spring brook experiment, both diet and location within the brook exerted significant effects on larval growth rate and adult size. The results of the two-year experiment suggest that diet exerts some influence on the seasonal pattern and magnitude of larval growth. However, site-to-site differences in the seasonal pattern and magnitude of temperatures (especially during the December-February period) exhibited the best correlation with observed differences in larval growth and adult size.