Eutrophic lichens respond to multiple forms of N: implications for critical levels and critical loads research

Abstract

Epiphytic lichen communities are highly sensitive to excess nitrogen (N), which causes the replacement of native floras by N-tolerant, "weedy" eutrophic species. This shift is commonly used as the indicator of ecosystem "harm" in studies developing empirical critical levels (CLE) for ammonia (NH3) and critical loads (CLO) for N. To be most effective, empirical CLE and/or CLO must firmly link lichen response to causal pollutant(s), which is difficult to accomplish in field studies in part because the high cost of N measurements limits their use. For this case study we synthesized an unprecedented array of atmospheric N measurements across 22 long-term monitoring sites in the Los Angeles Basin, California, USA: gas concentrations of NH3, nitric acid (HNO3), nitrogen dioxide, and ozone (n = 10 sites); N deposition in throughfall (n = 8 sites); modeled estimates of eight different forms of N (n = 22 sites); and nitrate deposition accumulated on oak twigs (n = 22 sites). We sampled lichens on black oak (Quercus kelloggii Newb.), and scored plots using two indices of eutroph (N tolerant species) abundance to characterize the community-level response to N. Our results contradict two common assertions about the lichen-N response: (1) that eutrophs respond specifically to NH3 and (2) that the response necessarily depends upon the increased pH of lichen substrates. Eutroph abundance related significantly but weakly to NH3 (r2 = 0.48). Total N deposition as measured in canopy throughfall was by far the best predictor of eutroph abundance (r2 = 0.94), indicating that eutrophs respond to multiple forms of N. Most N variables had significant correlations to eutroph abundance (r2 = 0.36-0.62) as well as to each other (r2 = 0.61-0.98), demonstrating the risk of mistaken causality in CLE/CLO field studies that lack sufficient calibration data. Our data furthermore suggest that eutroph abundance is primarily driven by N inputs, not substrate pH, at least at the high-pH values found in the basin (4.8-6.1). Eutroph abundance correlated negatively with trunk bark pH (r2 = 0.43), exactly the opposite of virtually all previous studies of eutroph behavior. This correlation probably results because HNO3 dominates N deposition in our study region.