Patterns and controls of nitrogen fixation in a lowland tropical forest, Panama

This dissertation examines the landscape-level patterns and nutrient controls of nitrogen (N) fixation in lowland tropical forests. Nitrogen fixation is the primary source of N to many ecosystems and can alter community structure, nutrient balance, and response to disturbances such as forest clearance and climate change. Yet, it remains poorly studied in tropical forests, one of the systems where putative N fixers are most abundant. First, I provide an introduction to N fixation with a focus on key issues for tropical forests---describing how research in temperate ecosystems has generated a bias towards an obligate model of N fixation while physiological research suggests fixation is facultative. Second, I use a focal tree approach to examine variations in fixation by adult Inga trees across several forest types in central Panama. Canopy trees actively fix N in disturbed shoreline and secondary forest but form few nodules in mature forests. This pattern, combined with the up-regulation of fixation in disturbed mature forest (gaps), supports a facultative (rather than obligate) model of N fixation. Third, I compare direct measures of nodulation and fixation to the alternative 15N natural abundance method. The results indicate that foliar N isotopes failed to capture significant variations in fixation across sites, suggesting that the 15N natural abundance method uses an overly simplistic model of N uptake and allocation. Fourth, I examined the relationship between fixation by canopy trees and soil nutrients. Nodulation was inversely correlated with soil nitrate supply, suggesting that trees down-regulate fixation in sites where N is abundant relative to both plant and microbial demand. Fifth, I examined the response of heterotrophic N fixing bacteria in the soil to long term nutrient fertilization. N fixation was suppressed by N addition and responded to both phosphorus (P) and micronutrient fertilizer. Direct additions and measurement of the trace element molybdenum (Mo) confirmed that Mo limits fixation rates and that the apparent effect of P can be explained by Mo present in P fertilizer. Overall, my findings provide the first clear picture of tropical N fixation as a facultative process which responds rapidly to changes in soil nutrient supply.