Temperate forests are predicted to increase productivity in response to anthropogenic nitrogen (N) deposition, but this prediction is based on the assumption that N is growth-limiting and will remain so in the future. Mounting evidence suggests that productivity in many terrestrial systems is already stoichiometrically constrained by additional nutrients, especially phosphorus (P). Thus, under continued N elevation, forest carbon sequestration may increasingly depend on the size and accessibility of soil P pools. While it is known that N deposition tends to be higher than P relative to plant demand, and that experimental N additions increase plant effort towards P acquisition, evidence of long-term shifts in stoichiometric demand from unmanipulated systems has not yet been shown. In deciduous trees, leaf senescence is the largest annual loss of nutrients, but trees can control their relative losses of N and P though nutrient resorption. Here we examine autumn leaf litter concentrations collected annually from 1996-2013 at the Coweeta Hydrologic Laboratory, Otto, NC, USA. Freshly fallen leaves were collected from five forest stands of varying soil P status, and representing the major forest types of the southern Appalachian region. We hypothesized that N concentrations of senesced leaves would increase and P concentrations would decrease due to shifting nutrient limitation, leading to an overall increase in N:P ratios through time.
Results/Conclusions
Senesced leaf N:P ratios increased significantly (P ≤ 0.01) through time in four out of the five stands in our study. This trend was driven in large part by changes in N concentrations, which showed at least marginally significant increases (P ≤ 0.1) in all stands, while we did not detect significant declines in litterfall P concentrations as predicted. Net P lost from litterfall increased during this period, despite the relatively unchanging P concentrations due to an increase of litterfall mass flux. We observed an increase in acid extractable P in the top layers of soil in all stands during this time (P<0.05), suggesting a net movement of P from slower cycling pools. Collectively, our results suggest that chronic N enrichment has resulted in faster internal P cycling in these forests, and that this increase is being driven by plant demand.