Averaged 30 year climate change projections mask opportunities for species establishment

Survival of early life stages is key for population expansion into new locations and for persistence of current populations (Grubb 1977, Harper 1977). Relative to adults, these early life stages are very sensitive to climate fluctuations (Ropert-Coudert et al. 2015), which often drive episodic or ‘event-limited’ regeneration (e.g. pulses) in long-lived plant species (Jackson et al. 2009). Thus, it is difficult to mechanistically associate 30-yr climate norms to dynamic processes involved in species range shifts (e.g. seedling survival). What are the consequences of temporal aggregation for estimating areas of potential establishment? We modeled seedling survival for three widespread tree species in California, USA (Quercus douglasii, Q. kelloggii, Pinus sabiniana) by coupling a large-scale, multi-year common garden experiment to high-resolution downscaled grids of climatic water deficit and air temperature (Flint and Flint 2012, Supplementary material Appendix 1). We projected seedling survival for nine climate change projections in two mountain landscapes spanning wide elevation and moisture gradients. We compared areas with windows of opportunity for seedling survival – defined as three consecutive years of seedling survival in our species, a period selected based on studies of tree niche ontogeny (Supplementary material Appendix 1) – to areas of 30-yr averaged estimates of seedling survival. We found that temporal aggregation greatly underestimated the potential for species establishment (e.g. seedling survival) under climate change scenarios.