The impacts of climate change on biota are often thought to be globally coherent, with large-scale shifts in species distributions towards higher elevations or away from equatorial latitudes in response to warming temperatures.  However, at smaller spatial scales, the effects of climate change may not adhere to broad scale patterns.  We used historic (ca. 1930) and modern (ca. 2000) vegetation surveys of nearly 10,000 plots each from the mountain ranges of California to assess if species optimum elevations had shifted in response to 20^th century climate change.  We used Gaussian response curves and a profile deviance approach to estimate optimum elevations and confidence intervals for more than 50 vascular plant species in each time period.  We conducted analyses across our entire study area and then separately within ecoregions to determine if observed patterns in climate change impacts were consistent across spatial extents and among ecoregions.

Results/Conclusions At the study area scale we found evidence of species’ optimum elevations shifting both uphill and downhill, with no definitive pattern as expected under climatic warming.  Evidence of both uphill and downhill shifts were also detected among individual ecoregions, suggesting that climate change impacts in our study area may be scale dependent.  Although our study area has experienced a general trend of warming during the 20^th century, it has also experienced substantial increases in precipitation that have led to widespread decreases in water deficit, which may explain our failure to detect large-scale uphill shifts in elevation.  At the ecoregion scale, our findings highlight the spatial variability in climate change impacts and may be a result of variability in warming trends at local scales.  These results suggest that climate change impacts are not globally coherent and may diverge from expected patterns, especially at smaller spatial scales.