Modelling pin-point plant cover data along an environmental gradient

Devil in the details: how can we avoid potential pitfalls of CATS regression when our data do not follow a Poisson distribution?

BACKGROUND

Community assembly by trait selection (CATS) allows for the detection of environmental filtering and estimation of the relative role of local and regional (meta-community-level) effects on community composition from trait and abundance data without using environmental data. It has been shown that Poisson regression of abundances against trait data results in the same parameter estimates. Abundance data do not necessarily follow a Poisson distribution, and in these cases, other generalized linear models should be fitted to obtain unbiased parameter estimates.

AIMS

This paper discusses how the original algorithm for calculating the relative role of local and regional effects has to be modified if Poisson model is not appropriate.

RESULTS

It can be shown that the use of the logarithm of regional relative abundances as an offset is appropriate only if a log-link function is applied. Otherwise, the link function should be applied to the product of local total abundance and regional relative abundances. Since this product may be outside the domain of the link function, the use of log-link is recommended, even if it is not the canonical link. An algorithm is also suggested for calculating the offset when data are zero-inflated. The relative role of local and regional effects is measured by Kullback-Leibler R2. The formula for this measure presented by Shipley (2014) is valid only if the abundances follow a Poisson distribution. Otherwise, slightly different formulas have to be applied. Beyond theoretical considerations, the proposed refinements are illustrated by numerical examples. CATS regression could be a useful tool for community ecologists, but it has to be slightly modified when abundance data do not follow a Poisson distribution. This paper gives detailed instructions on the necessary refinement.

Evidence for acidification-driven ecosystem collapse of Danish Erica tetralix wet heathland

We report observations of disappearance of Erica tetralix in wet heathland, which is unlikely to be caused by competition, as E. tetralix is dying before its place is taken up by other species. To investigate the causes, we used both old and new data. Results showed that presence of Molinia caerulea and Calluna vulgaris were substantial in the former E. tetralix dominated areas. Measurements of the C/N ratio in the morlayer were between 21 and 26 under the E. tetralix stands. As the expected C/N ratio in a healthy nutrient poor ecosystem like the E. tetralix wet heathland is around 30, this indicates that the ratio is probably decreasing and, correspondingly, the probability of nitrogen leaching from the ecosystem is increasing. The morlayer pH was extremely low-between 3.03 and 3.78. This represents a pH decline since the 1960s, where pH values generally were above 4. This supports the hypothesis that the decrease in morlayer pH is the major factor explaining the disappearance of E. tetralix and that measures to increase pH should be considered as part of the recommendations for relevant future management.

Spatial structure of the abiotic environment and its association with sapling community structure and dynamics in a cloud forest

Analyzing the relationship between the spatial structures of environmental variables and of the associated seedling and sapling communities is crucial to understanding the regeneration processes in forest communities. The degree of spatial structuring (i.e., spatial autocorrelation) of environmental and sapling community variables in the cloud forest of Teipan, S Mexico, were analyzed at a 1-ha scale using geostatistical analysis; after fitting semivariogram models for each set of variables, the association between the two sets was examined through cross-variograms. Kriging maps of the sapling community variables (density, cover, species richness, and mortality and recruitment rates) were obtained through conditional simulation method. Canopy openness, total solar radiation, litter depth, soil temperature and soil moisture were spatially structured, as were sapling density, species richness and sapling mortality rate. Mean range in semivariograms for environmental and sapling community variables were 13.14 ± 3.67 and 12.68 ± 5.71 m (±SE), respectively. The spatial structure of litter depth was negatively associated with the spatial structures of sapling density, species richness, and sapling community cover; in turn, the spatial structure of soil moisture was positively associated with the spatial structure of recruitment rate. These associations of the spatial structures of abiotic and sapling community variables suggest that the regeneration processes in this cloud forest is driven by the existence of different microsites, largely characterized by litter depth variations, across which saplings of tree species encounter a range of opportunities for successful establishment and survival.