Intermediate disturbance hypothesis is not supported for tree species diversity in boreal forest

Background/Question/Methods

Empirical evidence often contradicts the theoretical prediction of the intermediate disturbance hypothesis (IDH). Most IDH analyses focus solely on species richness, such that little is known about evenness and Shannon's index. Some studies have shown that disturbance and climate regulate species diversity, but they rarely account for local site conditions. Here, we test the applicability of the IDH in a large landscape of disturbance-driven boreal forest by evaluating the relationships between species richness, Shannon's index, evenness, and time since stand-replacing fire (TSF). We tested whether each measure of diversity was quadratically related to TSF using boosted regression trees (BRT).

Results/Conclusions

BRT model for richness, Shannon's index and evenness, explained 24, 34 and 31% variation of tree species diversity, respectively. We found negative linear relationships between species richness and TSF, as well as Shannon's index and TSF, but species evenness increased with TSF. TSF and local site conditions were the strongest predictors in each BRT model, with relative influences ranging from about 27 to 48%. Our results suggest that diversity indices can explain the many debates about the IDH.