Climate as an agent of selection
The strength of the correlations between traits and provenance climate varied across gardens (Table 3, Fig. 3), supporting our third hypothesis of strong associations between phenotype and climate for some traits in some environments. A single axis (PC1) explained 95.8% of the variation in provenance climate and was influenced primarily by temperature and growing season-related climate variables (Supplemental Table 2). Populations sourced from areas with higher temperatures, lower precipitation, lower elevation, and longer growing seasons had higher PC1 scores. Bud set exhibited the strongest relationship with provenance climate across the gardens (R2 = 0.67 to 0.77), while bud flush showed significant correlations in the two warmer gardens (R2 = 0.49 in Yuma and 0.66 in Agua Fria), but not in the cold garden (R2 = -0.01) (Table 3, Fig. 3a). In Canyonlands, population variation was constrained as all trees flushed at approximately the same time, late in the spring. Specific leaf area did not show significant trait-climate correlations in any garden, although we see overall SLA values increasing from the hot to the cold garden site (Fig. 3a).
Tree growth traits were more likely to show garden-dependent relationships between population origin and performance (Table 3, Fig. 3b). Tree height showed no climate relationships when planted in the hottest garden, however the correlation become stronger in the mid to cold gardens. When planted at the coldest garden, trees sourced from colder, wetter climates, including the three populations from the Colorado Plateau, were taller than populations from hotter, drier environments. Similarly, the DRC relationship in the hottest garden showed trees sourced from warm, hot environments had larger trunk diameters compared to trees from colder climates (Fig. 3b). Together, tree height and basal trunk diameter act as indicators that overall tree performance is consistent with local adaptation, with hot, southern populations growing larger in the hottest Arizona garden, and northern, cold adapted populations growing larger in the coldest Utah garden.
Phenotypic plasticity was significantly correlated with population source climate for all traits except SLA (Table 3, Fig. 3). Populations sourced from hot, dry climates exhibited increased plasticity in leaf-level phenology traits relative to the colder populations, as previously reported in Cooper et al. (2019). Tree-level growth traits showed the opposite pattern of increased plasticity in those populations sourced from the colder, high elevation environments. SLA showed a similar trend to phenology, with warmer populations exhibiting higher plasticity compared to populations sourced from cooler climates, but was not significant. Again, this may be due in part to the lower sample size of seven populations that did not include populations from the hottest, driest sites (Fig. 3a).