Response of tropical forest nitrogen isotope ratios to 13 years of nitrogen and phosphorus fertilization

Background/Question/Methods

Generally, phosphorus (P) is considered less abundant than nitrogen (N) relative to plant demand.  However, N:P co-limitation is considered widespread and N additions have been shown to increase reproductive litter fall and to ameliorate declining relative growth rates when combined with potassium in lowland rainforest of Panama.  In addition, N-fixation appears to occur throughout the maturation of tropical forests suggesting episodic localization of N-limitation may be common.  Our objective in this study was to understand how the response of four tropical forest tree species to factorial N and P fertilization was reflected in ratios of heavy-to-light stable N isotopes (δ¹⁵N) – presumed integrators of multiple N-cycling processes. Through measurement of δ¹⁵N values from canopy leaves, senesced litter, and labile soil N pools (NO₃^-, NH₄⁺, dissolved organic N) we sought to detect how flexible N and P uptake and retention was among co-occurring tropical tree species and, through ¹⁵N tracing of soil N sources, we also analyzed changes in preferred forms of soil N.  Soil N and P availability was measured from soils extracted the same day as sampling or from in situ resin bag incubations in 16 lowland rainforest plots.  Soil N pool δ¹⁵N values were measured using persulfate oxidation coupled to the denitrifier method.

Results/Conclusions

Soil δ¹⁵N values were relatively unaffected across treatments with the exception of an enriched residual pool of NO₃^- in the N and N+P treatments following high denitrification losses.  However, only one species became significantly ¹⁵N-enriched in the N treatment relative to the control, a pattern corresponding to greater leaf N content.  Litter fall C:N content was relatively unaffected by fertilizer additions whereas litter fall C:P was greatly elevated by P addition.  Foliar δ¹⁵N values were positively correlated with litter fall δ¹⁵N but generally un-modified by changes in N resorption dynamics.  After accounting for ¹⁵N fractionation by arbuscular mycorrhizae, foliar δ¹⁵N values could be explained by altered fertility influencing species-specific changes in N source uptake.  Mass balance mixing models indicated that the urea fertilizer was in some instances (34%) rapidly taken up following fertilization followed by a period where the tree species switched to more abundant and mobile NO₃^- relative to control and P fertilized treatments.  Combined with precipitation and fertilization induced spikes in nutrient availability, a nuanced picture emerges where tropical trees appear to switch N forms throughout the season.  In conclusion, tropical N cycles remain poorly understood and integrative δ¹⁵N measurements offer unique insight.