The food web of a severely contaminated site in Pennsylvania following restoration

Background/Question/Methods

   We present results of a study of the food web of the Lehigh Gap Wildlife Refuge, which was heavily impacted by emissions of sulfur dioxide, zinc, cadmium, and lead from two large zinc smelting plants. In 2006 the site was reseeded by warm season grasses that use the C₄ pathway and therefore have more of the heavy isotope ¹³C than the surrounding vegetation and the few forbs on the site, which possess the C₃ pathway. We used stable isotopes to distinguish between those components of the food web that depended on the introduced grasses and those that depended on the surrounding vegetation.  Two questions motivated our study: Are higher trophic levels using carbon from the introduced grasses? Is the food web that is based on the introduced grasses connected to the surrounding landscape? Invertebrate and vertebrate samples were collected from a 1200 m² plot located close to the center of the recently revegetated area and analyzed for ¹³C and ¹⁵N. The δ¹³C values were used to calculate the percent of carbon that came from C₄ plants using a two source model. Trophic position of consumers was estimated from 1 + (δ ¹⁵N_sample- δ¹⁵N_base)/ 3.4‰.

Results/Conclusions

   The food web appears to have at least two channels for carbon flux, one based on the warm season grasses and the other on C₃-derived carbon. Carbon from the recently planted grasses was apparently passed through to the higher trophic levels, as shown by the abundance of predators (mainly spiders) with a high proportion of C₄-derived carbon.

   Mean aboveground biomass was 165.6 g/m^-2 . Species with the C₃ photosynthetic pathway were less abundant than those with the C₄ pathway, appearing in 31% of sampled quadrats whereas C₄ species appeared in all of the quadrats. Plants with the C₃ pathway contributed 1.16% of the total plant biomass on the site while warm season grasses with the C₄ pathway contributed the remainder. The proportion of tissue samples that acquired most of their carbon from C₄ sources did not reflect the composition of the vegetation. Forty-eight of the 135 samples (36%) had more than 75% of their carbon from C₄-derived sources while 24% of the samples had less than 25%. Even when wide-ranging animals such as birds and wasps were eliminated from the data set, much of the carbon in the food web still appeared to come from sources that are poorly represented on the site.