Vegetation changes along environmental gradients have long captured the attention of ecologists, yet predictive models with explanatory value have been elusive. Here, we evaluate the predictive power and generality of Shipley’s maximum entropy (maxent) model of community assembly in the context of 96 quadrats, distributed over broad environmental gradients, within a 120 km² area of ponderosa pine forest that has a large (79 species) pool of understory herbaceous plants. The maxent model is a statistical translation of the concept of trait-based environmental filtering. The model states that, in a given environment, species relative abundances are functions of their traits, because species’ trait values bias their probability of occurrence in that environment. If validated, this model would have both explanatory and predictive ability in natural vegetation occurring at scales typically studied by community ecologists.

Results/Conclusions

The maxent model accurately predicted species relative abundances when observed community-weighted mean trait values were used as model constraints, and this predictive ability largely surpassed that of NMDS or DCA ordinations. Although only 53% of the variation in observed relative abundances was associated with a combination of 12 environmental variables, the maxent model based only on the environmental variables provided highly significant predictive ability and accounted for 72% of the variation that was possible given these environmental variables. Using cross-validation with 1000 independent runs, the median correlation between observed and predicted relative abundances was 0.560. The qualitative predictions of the model were also noteworthy: dominant species were correctly identified in 53% of the quadrats, 83% of rare species were correctly predicted to have a relative abundance of less than 0.05, and the median predicted relative abundance of species actually absent from a quadrat was 5x10^-5. Our results highlight both the potential for general applications of the maxent model as well as its empirical limitations.