SEARCH

SEARCH BY CITATION

Ecologists need not despair ot discovering the mechanisms that lead to large scale patterns. The search for process at higher scales has already led to enhanced confidence in the patterns and to improvements in their description. For example, species-area relationships turn out to form not one. but three patterns. Each is controlled by gain-loss dynamics at its own scale. At the macroscale, origination and global extinction reign. At the archipelagic scale, immigration and island extinction determine the results. At the local scale, metapopulation processes do. The three scales exhibit species-area curves with systematically different slopes in logarithmic space. We use the three scales of species-area to illuminate the relationship between local and regional diversity. Algebra shows that the latter pattern is an echo of species-area curves, and that those echoes ought to be nearly linear. So. we call the relationship of local and regional diversity, the Echo pattern. Ecology has long known that species-area curves within a region reflect the accumulation of habitat variety. Thus, their connection to Echo patterns argues against concluding that local diversity has little or nothing to do with population interactions. To obtain a pure Echo pattern, one should draw data from independent regions rather than separate islands. The independence allows natural selection to adjust the fundamental niches of species to diversity. Theory suggests that higher diversity should shrink niches, allowing the coexistence of more species locally. Hence, independence should tend to produce the straightest Echoes. However, archipelagic species-area curves predict that even when different islands are used as the regions, the Echoes should show only very gentle curvatures. Flouting theory, some archipelagic Echoes approach an asymptote as regional diversity increases. These must have logarithmic slopes that increase with regional pool size. We do not understand why.