A Multilevel Study of Effects of Low pH on Amphibians of Temporary Ponds

Abstract

Most researchers who have studied effects of low pH on amphibians that breed in temporary ponds measured direct lethal effects on single species. We summarize here results of a study conducted over 4 years in central Pennsylvania that illustrate the potential ecological importance of direct and indirect effects of low pH at lethal and sublethal levels. We argue for use of a multilevel approach (field sampling and experimentation in the laboratory, simulated ponds, and natural ponds) to study such effects on embryonic and larval amphibians of temporary ponds. A greater proportion of our study ponds was lower in pH than any reported previously in North America. Ponds were soft and often had high (Al,t,,,). Median seasonal pHs of 9 ponds were negatively associated with total amounts of rainfall over 4 years, but were not associated with H+ deposition during that same period. The pHs of many ponds declined as total rainfall increased. Embryonic mortality of the salamander Ambystoma jeffersonianum was high in ponds below pH 4.5. Less pH-associated embryonic mortality was observed among the more tolerant salamander, Ambystoma maculatum, and frog, Rana sylvatica. Laboratory-determined LC,, values for hatching of R. sylvatica, A. maculatum, and A. jeffersonianum were 4.10, 4.31, and 4.51, respectively, and predicted embryonic mortalities in ponds with good accuracy. More larval R. sylvatica survived at pH 4.1 in the laboratory when initially contained with A. jeffersonianum than at pH >6. This was due to reduced survival of, and subsequent reduced predation by, A. jeffersonianum at pH 4.1. In the same experiment, survival of A. maculatum was not different between pHs as they were eaten by A. jeffersonianum at pH >6 and suffered pH-induced mortality at pH 4.1. More larval A. jeffersonianum survived at pH 4.1 over 7 days when larval R. sylvatica and A. maculatum were available as prey, than when no prey were available. Larval A. jeffersonianum did not survive to metamorphosis in pH 4.2, and survived in low numbers in pH >6, in simulated ponds. Ambystoma maculatum metamorphosed less often, took longer to metamor- phose, and weighed less at metamorphosis in simulated ponds of pH 4.2 versus those of pH >6. Repro- duction of the newt Notophthalmus viridescens in simulated ponds was significantly lower, and adults were trapped more often exiting ponds, in pH 4.2 than in pH >6. Presence of adult N. viridescens in simulated ponds resulted in significantly lower survival of R. sylvatica at metamorphosis; survival was not significantly affected by pH 4.2. Our results in simulated ponds were likely due to a combination of lethal and sublethal effects that impinged either directly on individual larvae, or indirectly via the food web. These results suggest that studies focused only on direct lethal effects of low pH could substantially underestimate total impacts of low pH on amphibians of temporary ponds. Thus, in addition to the pronounced direct lethal effects we observed on A. jeffersonianum, the chronic and intermittent fluctu- ations in pHs in our study ponds that were associated with total rainfall could conceivably have induced less obvious effects throughout the food webs.