High-severity fire drives persistent floristic homogenization in human-altered forests

Ecological disturbance regimes across the globe are being altered via direct and indirect human influences. Biodiversity loss at multiple scales can be a direct outcome of these shifts. Fire, especially in dry forests, is an ecological disturbance that is experiencing dramatic changes due to climate change, fire suppression, increased human population in fire-prone areas, and alterations to vegetation composition and structure. Dry western conifer forests that historically experienced frequent, low-severity fires are now increasingly burning at high severity. Relatively little work has been done looking at the effects of this novel disturbance type on affected plant communities, and little is known about how these impacts change over time. To fill in these knowledge gaps, we examined a fire that burned in a yellow pine and mixed conifer forest in the central Sierra Nevada in California, USA. We sampled at five time steps across the nine years following the fire (1, 3, 5, 8, and 9 years postfire). We found a generally unimodal relationship between fire severity and plant alpha and gamma diversity, but found that areas that burned at high severity supported progressively lower plant diversity as time since fire increased. Similarly, beta diversity decreased drastically through time for the high-severity areas, while remaining more static in the other severity classes. The combination of these findings indicates that significant floristic homogenization can result from high-severity fire in this ecosystem type. We also saw consistently lower diversity in unburned areas in comparison to area burned at low and moderate severity, underlining that both lack of fire and high-severity fire can have negative impacts on postfire plant diversity. Unburned areas that experienced forest thinning after the first sample year saw an increase in plant diversity over time, suggesting that some—but not all—of the effects of fire on plant diversity can be approximated through forest management.