Great Basin shrublands in the United States are rapidly converting to annual grass- dominated ecosystems, driven primarily by increased wildfire activity. Post-fire vegetation recovery trajectories vary spatially and temporally and are influenced by the effects of topography, climate, soils, and pre-fire vegetation. Our study leverages spatially continuous Landsat data alongside spatial environmental datasets to evaluate drivers of post-fire vegetation recovery. We first evaluated the spectral diversity hypothesis, which suggests that variation in remotely sensed spectral values relates to plant species diversity. In turn, plant species diversity is theorized to be an important predictor of ecological resilience to disturbance and resistance to invasive species. Weak relationships between spectral diversity measures and plant species diversity led us to explicitly model plant species richness with Landsat spectral information and environmental variables. We applied the model of plant species richness to produce annual maps of predicted species richness from 1994–2017.

We assessed post-fire recovery in terms of the impacts of frequent fire activity on post- fire communities, whether post-fire seeding improves recovery outcomes, and by explicitly modeling post-fire plant communities. We found that repeated fires had a cumulative effect leading to increased annual herbaceous invasion and diminished perennial plant components. Meanwhile, on average, post-fire seeding treatments had negligible influence upon post-fire perennial plant recovery. Importantly, post-fire recovery trajectories varied significantly across the region, underscoring the importance of spatial evaluations of recovery patterns. The model of post-fire recovery produced strong validation statistics when averaged across all fires and more tempered results when applied to new fires not included in model development. Notably, plant species richness was not a strong enough predictor variable to be included in the final model.

Spatially continuous analyses are important as they can account for variability in post-fire recovery of Great Basin shrublands. While such analyses have previously been hampered by data and computing limitations, our results suggest that these approaches are increasingly tractable. Most importantly, spatially explicit approaches such as this provide valuable maps to land managers that can inform data-driven post-fire management.