OOS 25-5
One size does not fit all: Multi-scale heterogeneity in the lowland tropical N cycle

Tuesday, August 11, 2015: 2:50 PM
310, Baltimore Convention Center
Alan Townsend, Nicholas School of the Environment, Duke University
Cory C. Cleveland, Department of Ecosystem and Conservation Sciences, University of Montana, Missoula, MT
Gregory Asner, Department of Global Ecology, Carnegie Institution for Science, Stanford, CA
Stephen Porder, Ecology and Evolutionary Biology, Brown University, Providence, RI
Philip G. Taylor, Nicholas School of the Environment, Duke University, Durham, NC
Brooke B. Osborne, Institute at Brown for Environment & Society, Brown University, Providence, RI
Megan K. Nasto, Department of Ecosystem and Conservation Sciences, University of Montana, Missoula, MT
Will R. Wieder, TSS / CGD, National Center for Atmospheric Research, Boulder, CO
Benjamin Sullivan, University of Nevada, Reno
Background/Question/Methods

A longstanding paradigm in ecosystem ecology states that lowland tropical forests cycle nitrogen (N) in relative excess, such that soil inorganic pools of N, along with gaseous and hydrologic losses, are comparatively high. However, more recent work shows this construct does not hold across all lowland systems, perhaps most especially those on the wetter end of the lowland tropical biome.   In addition, evidence is emerging that variation in other key state factors, including both tree species and geomorphology, can cause substantial heterogeneity in N cycling and losses at the sub-regional to local scale.  Here, we present highlights from several years of studying the N cycle on the Osa Peninsula of Costa Rica, a region with highly diverse and productive primary tropical forests, complex terrain and annual rainfall values that range from ~ 3.5 to 7 meters per year.  The work we present integrates data from soils, canopies and streams, as well as remote sensing data from the Carnegie Airborne Observatory, and is aimed at understanding multi-scale controls over the N cycle in this and other tropical forests.

Results/Conclusions

Taken as a whole, N does not appear to cycle in relative excess in Osa forests. 15-N data show high rates of immobilization, and relatively low rates of nitrification.  Soil nitrate pools are low, and inorganic N declines even further between 20 and 50 cm soil depth, suggesting a low potential for hydrologic export.  Low concentrations of inorganic N in primary streams are consistent with these soil data, as are comparatively low values for nitrous oxide emissions.  Hydrologic losses of particulate N are the largest N export vector, suggesting a mechanism by which high rainfall and complex terrain help maintain the system in a more N-constrained state.   However, while the overall picture is not one of N excess, metrics of the N cycle also vary substantially at both local and regional scales.   Soil 15-N values decline by 3 parts per mil along regional rainfall gradients that range from ~3.5-7 m/yr.  Multiple metrics suggest a more open N cycle on flatter, upland ridgetops than on neighboring slopes, though these topographic patterns appear to vary with climate and geomorphic status.  And canopy N data show significant local variation including areas of local canopy “hotspots” associated with higher soil N status.