Theoretical habitat templets, species traits, and species richness: a synthesis of long‐term ecological research on the Upper Rhône River in the context of concurrently developed ecological theory

Abstract

SUMMARY Based on information obtained from analysis of thirteen taxonomic groups of plants and animals occurring in the alluvial floodplain habitats of the Upper Rhône River, France, we synthesize results obtained on: (i) relationships among species traits; (ii) habitat utilization by species; (iii) the relationship between species traits and habitat utilization; (iv) trends in species traits in a framework of spatial and temporal variability; and (v) tests of trends predicted for species traits and species richness in the framework of spatial and temporal habitat variability in terms of the river habitat templet and patch dynamics concept. Species traits describing reproductive characteristics, food, and size had the closest relationships with each other in the various correspondence analyses performed. Faunal and floral separation by species traits produced groupings similar to those based on traditional taxonomy. Two major gradients appear in the utilization of the floodplain habitats: a vertical gradient from interstitial to superficial habitats; and a transverse gradient from the main channel to oxbow lakes, temporary waters, and terrestrialized habitats. For the majority of the groups examined, a statistically significant relationship was evident between the structure of the species trait and habitat utilization arrays. For these groups, the characteristics of the habitat act as a templet for species traits. Moreover, species trait modalities (i.e. categories defining traits) were significantly arranged along the axis of spatial and temporal variability for most groups, which indicates that such variability acts as a templet for species traits. Species traits did not conform to predictions of the river habitat templet because the observed modality sequences did not follow the trends predicted in a framework of spatial and temporal variability. Moreover, there was no clear pattern in the distribution of species traits along an axis of temporal variability for groups of organisms having different sizes, which is a correlate of longevity, nor did modalities of species traits that occur under conditions of low temporal variability also tend to occur under conditions of high spatial variability (or vice versa). Clearly, species traits occur as alternative suites of characteristics in various groups of organisms. The patch dynamics concept, which predicts that highest species richness occurs at intermediate levels of temporal variability and highest levels of spatial variability, was supported by observations in only two of the thirteen groups exaniined, and only partially (for spatial variability) when all 548 taxa were examined together. The predictions of the river habitat templet and patch dynamics concepts were not supported, perhaps because templet theories do not yet accommodate alternative suites of characteristics and trade‐offs between combinations of traits, or perhaps because the single scale of variability considered in the analyses, the inhomogeneity of the available biological information, and the aggregation of spedes traits that were used created methodological problems. Ecologically sound river management polides eventually may be based on two key points that emerged from this synthesis: that the habitat acts as a templet for spedes traits, and that composite taxonomic groupings represent relatively homogeneous assemblages of spedes trait modalities. The use of statistical approaches developed in this project to analyse other long‐term data sets may clarify questions about the applicability of habitat templet theories to river ecology, and hasten development of ecologically sound river management policies.