PARENT SESSION
Oral Session 110: Biogeochemistry VII: Soil I and Nutrient Cycling I.
Presiding: K Lajtha
Thursday, August 5, 1:30 PM to 5:00 PM, Meeting Room A 106.

Nutrient losses from undisturbed forests in New Zealand: C, N and P loss as a function of climate and soil properties.

Baisden, W.^*,1, McGroddy, Megan², Hedin, Lars², ¹ Landcare Research, Palmerston North, New Zealand² Princeton University, Princeton, NJ

ABSTRACT- To understand the possible drivers of organic N losses from southern hemisphere forests with no known anthropogenic N inputs, we studied first-order streams draining native forests in New Zealand. This research builds on previous findings that organic forms dominate N loss from forests in temperate South America. To understand patterns of N loss and possible interactions with C and P, we use high-resolution spatial data describing climate and soil properties derived from the Land Environments of New Zealand project. We analyzed a preliminary dataset of 68 sites sampled in early spring using stepwise multivariate regression to identify the main predictors of C, N and P losses, as well as C:N and N:P ratios. Regressions for all variables analyzed were highly significant, and climate variables were more significant predictors of organic and inorganic C, N and P concentrations and ratios of C:N and N:P than soil properties. In all cases, suites of 3 - 7 predictor variables explained 30 - 60% of the variance or log-variance in various C, N and P concentrations and C:N and N:P ratios. Mean annual temperature, and minimum temperatures are the most influential predictors of DON concentrations and DON:P_org. Winter and/or summer solar radiation and water deficits or evapotranspiration are strong predictors of DOC and P_org concentrations as well as DOC:DON. Among soil properties, acid soluble P, drainage, induration, and particle size showed some explanatory value. These results demonstrate that patterns of nutrient loss can be predicted empirically for ecosystems in steady state, and further understanding can be generated with additional study. When combined with patterns of nutrient input, these results will allow the size of C and nutrient pools to be estimated for undisturbed ecosystems -- results which can be used to initialize biogeochemistry models.

Key words: phosphorous, stoichiometry, organic nitrogen, carbon