Seasonal patterns of climate controls over nitrogen fixation by Alnus viridis subsp. fruticosa in a secondary successional chronosequence in interior Alaska1

Patterns of and controls over N2 fixation by green alder were studied in post-fire, mid-succession, and white spruce upland forests in interior Alaska, focusing on the hypothesis that ecosystem-level nitrogen (N) inputs decrease with successional development. N2-fixation rates tracked plant phenology during the 1997 (drought) and 1998 (normal precipitation) growing seasons. The best model for predicting acetylene reductase activity (ARA, μmol C2H4·g noduleDWT·h) across all stands and sampling periods included a linear response to soil temperature and a quadratic response to Julian day (r2 = 0.23, P < 0.0001). There were few significant relationships between seasonal maximum values for ARA and measured leaf traits; however, we did detect an inverse correlation between foliar N to P ratio and seasonal maximum ARA. During 1998, stands with higher maximum ARA had higher soil %N, and maximum ARA was positively correlated with subcanopy %P in O and A soil horizons. During 1997, leaf %P and N resorption were lower and leaves were thinner compared to 1998. Drought effects were most pronounced in mid-succession, where alder exhibited reduced ARA, leaf %P, leaf thickness, and lower leaf resorption of P and N. Although ARA and nodule biomass did not differ among stand types, increases in alder densities with successional time translated to increasing ecosystem-level N inputs across the chronosequence. These results contradict established theory predicting a decline in N2-fixation rates and N2-fixer abundance during successional stand development.