Merging theory and data to understand broad patterns in ecosystem ecology

Background/Question/Methods

Many of the most interesting and important patterns in ecosystem ecology occur at very broad spatial and temporal scales.  Understanding these broad patterns presents a scientific challenge because experiments on the right spatial and temporal scales are logistically impossible.  Mathematical theory, large-scale observations, and targeted experiments each contribute pieces of the puzzle, and I believe that pursuing all three approaches offers the best path to understanding these broad patterns.  In this talk I will focus on two patterns that I find interesting and some theoretical and empirical work that my collaborators and I have done to understand them.  The first pattern is that N-fixing trees in North America are 10x more abundant south of 35° latitude than they are to the north.  The second pattern is that N-fixing trees at higher latitudes drop out in succession, whereas at lower latitudes they are relatively common throughout succession.  These patterns would be easy to understand if lower latitude forests were more N limited, but this does not seem to be the case.  Why, then, are N-fixing trees so much more abundant at lower latitudes, particularly late in succession?

Results/Conclusions

I will discuss a variety of hypotheses that can explain these patterns.  The first concerns a latitudinal transition in the degree to which N-fixing trees regulate how much N they fix based on how much they need.  “Obligate” N-fixing trees—those that do not adjust how much N they fix based on how much they need—succeed in early succession but are excluded fairly rapidly during succession.  At the landscape level, therefore, they are not very abundant.  On the other hand, “facultative” N-fixing trees—those that rapidly adjust N fixation based on their current nutritional requirements—survive throughout succession, and are thus fairly abundant at the landscape scale.  A transition across latitude from facultative to obligate N-fixing trees, therefore, can explain the patterns.  A second hypothesis concerns latitudinal changes in the costs and benefits of N fixation itself, assuming that facultative N fixation is ubiquitous.  For this hypothesis to work, the threshold degree of N limitation at which the benefits of N fixation exceed the costs must change with latitude.  I will discuss a variety of mechanisms by which this could occur.  Empirical support for each of these hypotheses exists, but is not yet conclusive.