Sasha C. Reed, Cory C. Cleveland, and Alan R. Townsend. University of Colorado, Boulder
Free-living nitrogen (N) fixation can be a significant source of N to terrestrial ecosystems. For plants capable of symbiotic associations, species-specific differences in N fixation rates are frequently documented, yet we have little understanding of how different plant species affect the free-living N fixation rates occurring on and beneath them. In particular, many species-rich tropical rain forests appear to support high rates of free-living N fixation, yet the degree to which tree species variation controls such rates is unknown. Here, we report results for surveys of canopy leaf, leaf litter, and soil N fixation rates on and beneath six species of trees from a lowland tropical rain forest in southwestern Costa Rica. The site contains 100-200 tree species per hectare, averages over 5 m of precipitation annually, and our six selected species represent six of the most common families in the region. Our data showed that rates of free-living N fixation varied by orders of magnitude along a vertical profile spanning the canopy to the soil. Moreover, free-living N fixation rates varied significantly by tree species within each profile component: showing sixty-, six-, and three-fold inter-specific rate differences for canopy leaves, leaf litter, and soils, respectively. Indices of phosphorus (P) availability such as foliar and litter P content, N:P ratios, and labile soil P pools also consistently varied by tree species, and were significantly correlated with rates of free-living N fixation. Thus, these data suggest that: 1) free-living N fixation in tropical forests varies considerably in both vertical and horizontal dimensions; 2) species-specific variation in N fixation rates may be predictable from standard ecosystem metrics such as foliar and soil P content and/or N:P ratios; and 3) variability in P availability and N inputs could drive inter-specific differences in other biogeochemical processes such as soil carbon storage.