Extending the CWM approach to intraspecific trait variation: how to deal with overly optimistic standard tests?

Community weighted means (CWMs) are widely used to study the relationship between community-level functional traits and environment. For certain null hypotheses, CWM-environment relationships assessed by linear regression or ANOVA and tested by standard parametric tests are prone to inflated Type I error rates. Previous research has found that this problem can be solved by permutation tests (i.e., the max test). A recent extension of the CWM approach allows the inclusion of intraspecific trait variation (ITV) by the separate calculation of fixed, site-specific, and intraspecific CWMs. The question is whether the same Type I error rate inflation exists for the relationship between environment and site-specific or intraspecific CWM. Using simulated and real-world community datasets, we show that site-specific CWM-environment relationships have also inflated Type I error rate, and this rate is negatively related to the relative ITV magnitude. In contrast, for intraspecific CWM-environment relationships, standard parametric tests have the correct Type I error rate, although somewhat reduced statistical power. We introduce an ITV-extended version of the max test, which can solve the inflation problem for site-specific CWM-environment relationships and, without considering ITV, becomes equivalent to the "original" max test used for the CWM approach. We show that this new ITV-extended max test works well across the full possible magnitude of ITV on both simulated and real-world data. Most real datasets probably do not have intraspecific trait variation large enough to alleviate the problem of inflated Type I error rate, and published studies possibly report overly optimistic significance results.

Better environmental data may reverse conclusions about niche- and dispersal-based processes in community assembly.

Variation partitioning of species composition into components explained by environmental and spatial variables is often used to identify a signature of niche- and dispersal-based processes in community assembly. Such interpretation, however, strongly depends on the quality of the environmental data available. In recent studies conducted in forest dynamics plots, the environment was represented only by readily available topographical variables. Using data from a subtropical broad-leaved dynamics plot in Taiwan, we focus on the question of how would the conclusion about importance of niche- and dispersal-based processes change if soil variables are also included in the analysis. To gain further insight, we introduced multiscale decomposition of a pure spatial component [c] in variation partitioning. Our results indicate that, if only topography is included, dispersal-based processes prevail, while including soil variables reverses this conclusion in favor of niche-based processes. Multiscale decomposition of [c] shows that if only topography was included, broad-scaled spatial variation prevails in [c], indicating that other as yet unmeasured environmental variables can be important. However, after also including soil variables this pattern disappears, increasing importance of meso- and fine-scaled spatial patterns indicative of dispersal processes.

Measuring size and composition of species pools: a comparison of dark diversity estimates.

Ecological theory and biodiversity conservation have traditionally relied on the number of species recorded at a site, but it is agreed that site richness represents only a portion of the species that can inhabit particular ecological conditions, that is, the habitat-specific species pool. Knowledge of the species pool at different sites enables meaningful comparisons of biodiversity and provides insights into processes of biodiversity formation. Empirical studies, however, are limited due to conceptual and methodological difficulties in determining both the size and composition of the absent part of species pools, the so-called dark diversity. We used >50,000 vegetation plots from 18 types of habitats throughout the Czech Republic, most of which served as a training dataset and 1083 as a subset of test sites. These data were used to compare predicted results from three quantitative methods with those of previously published expert estimates based on species habitat preferences: (1) species co-occurrence based on Beals' smoothing approach; (2) species ecological requirements, with envelopes around community mean Ellenberg values; and (3) species distribution models, using species environmental niches modeled by Biomod software. Dark diversity estimates were compared at both plot and habitat levels, and each method was applied in different configurations. While there were some differences in the results obtained by different methods, particularly at the plot level, there was a clear convergence, especially at the habitat level. The better convergence at the habitat level reflects less variation in local environmental conditions, whereas variation at the plot level is an effect of each particular method. The co-occurrence agreed closest the expert estimate, followed by the method based on species ecological requirements. We conclude that several analytical methods can estimate species pools of given habitats. However, the strengths and weaknesses of different methods need attention, especially when dark diversity is estimated at the plot level.