Fine root respiration in northern hardwood forests in relation to temperature and nitrogen availability.

We examined fine-root (< 2.0 mm diameter) respiration throughout one growing season in four northern hardwood stands dominated by sugar maple (Acer saccharum Marsh.), located along soil temperature and nitrogen (N) availability gradients. In each stand, we fertilized three 50 x 50 m plots with 30 kg NO(3) (-)-N ha(-1) year(-1) and an additional three plots received no N and served as controls. We predicted that root respiration rates would increase with increasing soil temperature and N availability. We reasoned that respiration would be greater for trees using NO(3) (-) as an N source than for trees using NH(4) (+) as an N source because of the greater carbon (C) costs associated with NO(3) (-) versus NH(4) (+) uptake and assimilation. Within stands, seasonal patterns of fine-root respiration rates followed temporal changes in soil temperature, ranging from a low of 2.1 micro mol O(2) kg(-1) s(-1) at 6 degrees C to a high of 7.0 micro mol O(2) kg(-1) s(-1) at 18 degrees C. Differences in respiration rates among stands at a given soil temperature were related to variability in total net N mineralized (48-90 micro g N g(-1)) throughout the growing season and associated changes in mean root tissue N concentration (1.18-1.36 mol N kg(-1)). The hypothesized increases in respiration in response to NO(3) (-) fertilization were not observed. The best-fit model describing patterns within and among stands had root respiration rates increasing exponentially with soil temperature and increasing linearly with increasing tissue N concentration: R = 1.347Ne(0.072T) (r(2) = 0.63, P < 0.01), where R is root respiration rate ( micro mol O(2) kg(-1) s(-1)), N is root tissue N concentration (mol N kg(-1)), and T is soil temperature ( degrees C). We conclude that, in northern hardwood forests dominated by sugar maple, root respiration is responsive to changes in both soil temperature and N availability, and that both factors should be considered in models of forest C dynamics.