Enhancing the credibility of ecology: Can research be made more comparable and predictive?

A sounder theoretical basis for extrapolation and prediction, typologies of what is and what is not comparable, and representative natural areas for long-term field research are three geographically-based ingredients of comparative research in ecology. Two types of conceptual model allow assessment of the effects of phylogenesis and environment on ecosystem structure and functioning. The first is based on 'fixing' of genetic variables and study of biological performance along a gradient of environmental conditions — an approach of sequential analysis of environmental gradients. The second is based on 'fixing' of environmental variables and comparison of performance of different phylogenetic stocks in non-contiguous geographical areas — an approach entailing matching of disjunct ecological analogues. The two approaches are important complements in examining hypotheses of convergent and divergent evolution, in helping to define what really can be compared in comparative ecological research, and in contributing to ecology becoming a more predictive and credible science.