Disturbance ecology meets macroecology: A new method for cross-system comparisons of ecosystems in transition

Background/Question/Methods

The ubiquity of disturbance in structuring ecological communities continues to motivate a search for generality in disturbance ecology. A better understanding of ecological perturbations and quantitative comparisons of their effects over multiple scales is required for both species-level and landscape-scale conservation efforts, however, few quantitative syntheses of cross-system comparisons of disturbance effects exist. Here we extend the Maximum Entropy Theory of Ecology (METE), an information entropy-based theory of macroecology, to disrupted and disturbed ecosystems. Despite explicitly not incorporating ecological interactions, METE has proven reliable for estimating the species-area relationship, endemics- area relationship, and species-abundance distribution in minimally disturbed, “steady state” ecosystems. As a first test of METE’s predictions for ecosystems with high levels of disturbance, we census plants in several ecosystems with known histories of ecological disruption. We quantitatively compare the macroecological response of biological communities in transition (including a primary succession landslide system, a fire-evolved conifer system, and a novel grazing regime in forb- dominated meadows).

Results/Conclusions

We find that because it is based on information entropy mathematics, the Maximum Entropy Theory of Ecology may accurately describe certain macroecological metrics, including predicted number of rare species, in disturbed and relatively undisturbed sites. Where METE fails to make accurate predictions, a “signature of disturbance” may emerge where data deviates from theory in predictable ways.  This signature is detectable in replicate plot measurements (though specific outcomes vary by plot).  Candidates for a disturbance signature include: steeper SAR curves than predicted, log-normality of the SAD, deviation from species-level aggregation predictions, bias in the predicted number of singleton species, or specific combinations of failures of theory to predict these metrics.