Biological nitrogen fixation in tropical dry forests of Costa Rica: patterns and controls

Abstract

In tropical forests, new nitrogen (N) inputs fuel a large proportion of global net primary productivity. However, global estimates of tropical N fixation are biased towards wet forests and other areas such as tropical dry forests are understudied. In the dry forests of Guanacaste, Costa Rica, N fixing legume trees are highly abundant throughout forest successional stages, thus I hypothesized that in tropical dry forests legume trees are critical regulators of ecosystem level N dynamics. I addressed this question from multiple approaches that included a shade house experiment and field surveys of N fixing legume trees in plantations or in diverse secondary forests using a common set of species: <italic>Acosmium panamense<italic>, <italic>Dalbergia retusa<italic>, <italic>Enterolobium cyclocarpum<italic>, <italic>Gliricidia sepium<italic>, and <italic>Lysiloma divaricatum<italic>. Individual legume species had measurable influences on a number of soil properties, but this effect is more pronounced than the influence of legumes as a functional group. I observed species-specific variation in belowground foraging strategies and in the timing and degree of nodulation. In the shade house experiment, species differed in their nodulation effort and in how they regulated N fixation with respect to available resources. These five legume species could be arrayed along a continuum defined by strategies of nutrient conservation and nutrient acquisition, which coincided with degrees of fine-tuning of N fixation. In the field study, I did not find evidence of down-regulation of fixation with soil N. I hypothesized that the adjustment of N fixation to soil nutrients occurs indirectly and is mediated by water availability and its effects on nutrient pulses. My stand-level estimates N fixation by legumes showed that legumes are responsible for the largest contribution of new N inputs to this ecosystem relative to other inputs such as free-living fixation or wet deposition, but which are modest relative to N recycling through leaf litter and fine root decomposition. Different legume strategies could represent different ways of dealing with the transient and seasonal water availability of this ecosystem. Collectively, my results suggest that the conceptual models of how N fixation works in tropical wet forests may not necessarily be the same in seasonally dry forests.