Adaptive variation, including local adaptation, requires decades to become evident in common gardens.

Population-level adaptation to spatial variation in factors such as climate and soils is critical for climate-vulnerability assessments, restoration seeding, and other ecological applications in species management, and the underlying information is typically based on common-garden studies that are short duration. Here, we show >20 yr were required for adaptive differences to emerge among 13 populations of a widespread shrub (sagebrush, Artemisia tridentata ssp wyomingensis) collected from around the western United States and planted into common gardens. Additionally, >10 yr were required for greater survival of local populations, that is, local adaptation, to become evident. Variation in survival was best explained by the combination of populations' home ecoregion combined with grouping of minimum temperature and aridity. Additional reductions in survival were explained by ungrouped (i.e., continuous) measures of garden-to-population-origin separation in geographic distance (5% decrease in survival per 100 km increase in separation; R2  = 0.22) and especially in minimum temperature in younger plants (-4% per + °C difference, R2  = 0.56 vs. 0.29 in the 14th vs. 27th post-planting years, respectively). Longer-term common garden studies are needed. While we await them, uncertainty in adaptive variation resulting from short-term observations could be quantitatively estimated and reported with seed-transfer guidelines to reduce risks of introducing maladapted provenances in restoration.