The enemy release hypothesis as a hierarchy of hypotheses

In ecology, a hypothesis is usually not discarded if a few studies reject it, as long as there are other studies supporting it. How to assess the usefulness of ecological hypotheses is therefore not straightforward. Using the enemy release hypothesis as an example, we show how creating a hierarchy of hypotheses (HoH) can help reviewing and evaluating evidence for and against an ecological hypothesis. In a HoH, a broad, overarching hypothesis branches into more specific and better testable sub-hypotheses. The enemy release hypothesis is a major hypothesis in invasion ecology and posits that the absence of enemies in the exotic range of an alien species is a cause of its invasion success. Based on a systematic review of empirical tests of this hypothesis, we divided it into sub-hypotheses, differentiating among 1) indicators for enemy release, 2) types of comparisons, and 3) types of enemies. We identified 176 empirical tests and weighted each test according to the number of alien species studied and the research method (experimental vs observational, field vs enclosure vs laboratory). For the broadly formulated enemy release hypothesis, we found nearly as much supporting (36%) as questioning evidence (43%). At the sub-hypotheses level, however, we found that some sub-hypotheses are strongly supported by empirical evidence, whereas others receive hardly any support. These differences are further emphasized for some types of habitat and focal taxonomic groups. Our findings suggest that several specific formulations (i.e. sub-hypotheses) of the broad enemy release hypothesis are useful, whereas other formulations should be viewed more critically. In general, the approach outlined here can help evaluate major ecological hypotheses and their specific sub-hypotheses. Our study also highlights the need for a scientific debate on how much supporting evidence is sufficient to consider an ecological hypothesis to be useful.