| In this paper, I present a method developed to detect 'extra', or novel, changes in diversity in a naturally fluctuating environment. When comparing samples in order to evaluate changes in community structure, the sampling procedure will inevitably induce randomness in observed species composition and abundance, so two samples may show considerable differences, even if they come from exactly the same community. Sometimes the effort put into the sampling varies as well, leading to the expectation of a further increase in difference. Finally, if there were a temporal distance between the samples, we would like to correct for variations in species abundance occurring naturally due to fluctuating environmental conditions. The test method presented here... |
