Observational and experimental studies have shown that an interaction class between two species (be it mutualistic, competitive, antagonistic, or neutral) may transform into a different class depending on the biotic and abiotic contexts within which species are observed. This complexity arising from the evidence of context-dependencies has underscored a difficulty in establishing a testable theory about the frequency at which different interaction classes are expected to transform in nature.
Results/Conclusions
We introduce a structural approach to derive null expectations about how often such transformations should happen as a function of different biotic and abiotic contexts. Transformations are modeled as a combination of selection and fixation processes matching biotic and abiotic constraints, respectively. By accounting for transformation variations across different abiotic conditions within biotic contexts, our model can explain the apparently contrasting results showing that mutualistic interactions are on average the most likely to change under predetermined abiotic conditions (as in experiments) but not under changing abiotic conditions (as in nature). Overall, our framework provides a structural and systematic platform to study the frequency of transformations of species interactions within a context-dependent framework.