Genetic variation in temperature among sagebrush populations in a common garden revealed by unoccupied aerial system (UAS) thermal imagery

Peter J. Olsoy, Andrii Zaiats, Anna V. Roser, Jennifer S. Forbey and T. Trevor Caughlin, Department of Biological Sciences, Boise State University, Boise, ID, Matthew J. Germino, Forest and Rangeland Ecosystem Science Center, US Geological Survey, Boise, ID, Donna M. Delparte, Department of Geosciences, Idaho State University, Pocatello, ID, Bryce A. Richardson, Forest Sciences Laboratory, USDA Forest Service, Moscow, ID, Megan E. Cattau, Human-Environment Systems, Boise State University, Boise, ID

Background/Question/Methods
Studying interactions between environmental stress and genetic variation is crucial to better understand species’ adaptive capacity and resilience to climate change. The primary means for detecting genetic variation among plant populations is in a common garden experiment. Thermal snapshots of common gardens using unoccupied aerial systems (UAS) could be an innovative way to assess thermal variation, helping eliminate temporal variation in measurement conditions among plants. We used a UAS equipped with a thermal IR camera to assess temperature variation among populations of big sagebrush (Artemisia tridentata), a keystone plant species throughout much of the western USA that has a high level of genetic and phenotypic adaptations to hot, dry summers. Flights were completed on June 10, 2015, and on three dates across the 2019 growing season (June 5, July 11, and August 28) at a common garden in Idaho, USA, and were compared to a garden census (survival) and simultaneous leaf-level ecophysiology. The common garden experiment started in 2010 with seeds collected across the western USA representing three subspecies (Basin/tridentata, Wyoming/wyomingensis, and mountain/vaseyana) and two ploidy levels (2n and 4n). Our objectives were to (1) measure thermal differences in genetics (subspecies and ploidy level) and size; (2) predict survival; and (3) quantify differences between gas exchange measured at the leaf-level and energy balance occurring at the plant-level.
Results/Conclusions
UAS-derived temperatures were between 5.4 and 16.0 °C lower for foliar than surrounding soil surfaces, dependent on time of day and time of year. We also found that larger plants were cooler, but the strength of that relationship ranged from -0.97 °C/m2 in June 2019 (r2 = 0.14) to -3.68 °C/m2 in June 2015 (r2 = 0.43). The temperature difference was more stable later in the growing season with -2.40 °C/m2 in July (r2 = 0.29) and -1.98 °C/m2 in August (r2 = 0.21). While survival between 2015 and 2019 was best predicted by subspecies, a combined model of plant size and UAS-derived temperature performed similarly to ploidy level. UAS-derived temperature did not correlate with leaf-level ecophysiology measurements (r2 < 0.02), suggesting a mismatch in scale between gas exchange measured at the leaf-level and energy balance occurring at the plant-level. These results are a promising example of how fine-scale thermal mapping with UAS can help ecologists decouple the contribution of genetics and the environment on temperature dynamics in common garden experiments.