Demographic importance of ecological interactions: how much do statistics tell us?

Many investigators demonstrate the existence of intra- and interspecific interactions through rigorous statistical hypothesis testing. Statistical significance of an interaction, however, does not necessarily imply that it controls a species population size or distribution. Other analytical methods must be used to make these important determinations. This paper demonstrates one method, with an example, to quantify the demographic importance of statistically significant interactions. In our example, some of the statistically significant interactions with the polychaete Clymenella torquata (Leidy) have large impacts on the population growth rate of the bivalve Gemma gemma (Totten), while others do not. This demonstrates the necessity of distinguishing between statistical significance and demographic importance. In some cases, the demographic importance of the same interaction changes through time because of changes in the bivalve's life history. Interactions affecting juvenile bivalves often have a greater demographic importance than those affecting adults; but this is not always the case.