Relationships among phosphorus, molybdenum and free- living nitrogen fixation in tropical rain forests: results from observational and experimental analyses

Biological nitrogen (N) fixation is the primary source of “new” N to unmanaged ecosys­ tems, and recent analyses suggest that terrestrial N inputs via free-living N fixation may be more impor­ tant than previously assumed. This may be particularly true in some tropical rain forests, where free-living fixation could outpace symbiotic N fixation to repre­ sent the dominant source of new N inputs. However, our understanding of the controls over free-living N fixation in tropical rain forests remains poor, which directly constrains our ability to predict how N cycling will respond to changing environmental conditions. Although both phosphorus (P) and molybdenum (Mo) availability have been shown to limit free-living N fixation rates in the tropics, few studies have simul­ taneously explored P versus Mo limitation or the S. C. Reed ( ^ ) • A. R. Townsend Department of Ecology and Evolutionary Biology, INSTAAR, University of Colorado, Boulder, CO 80309, USA e-mail: [email protected] C. C. Cleveland Department of Ecosystem and Conservation Sciences, University of Montana, Missoula, MT 59812, USA potential importance of P x Mo interactions. Here, an archived set of foliar, litter, and soil samples from a Costa Rican tropical rain forest provided an opportu­ nity to simultaneously assess the relative strength of P versus Mo relationships with free-living N fixation rates. We also conducted a short-term, full-factorial (P X Mo) litter incubation experiment to directly assess nutrient limitation, allowing us to explore P and Mo controls over free-living N fixation rates using both observational and experimental approaches. We previously showed that N fixation rates were posi­ tively correlated with P concentrations in all substrates and, using the archived samples, we now show that Mo concentrations correlated with N fixation only in canopy leaves (where total Mo concentrations were extremely low). Likewise, fertilization with P alone (and not Mo) stimulated leaf litter N fixation rates. Thus, our results suggest that P availability domi­ nantly controls free-living N fixation at this site, and when taken with data from other studies, our results suggest that attempts to identify “the nutrient” that limits N fixation in “the tropics” may be misguided. Rather, nutrient controls over free-living N fixation appear to be more nuanced—and the true nature of nutrient limitation to N fixation likely varies over a variety of scales across the vast tropical rain forest biome.