Local adaptation to climate
Whereas leaf phenology and morphology traits (bud flush, bud set, and SLA) had the highest degree of differentiation in our study, likely due to climate-related divergent selection, tree growth traits (height and trunk basal diameter) were less differentiated. This result suggests that the neutral processes of gene flow and drift may override weak selection on growth traits (McKay & Latta 2002), or that growth is strongly constrained by the relative success dictated by the rest of a plant’s phenotype (Saint-Laurent et al. 2007; Leinonen et al . 2013). In this case, the latter explanation is very likely. Although phenology and growth traits showed significant regressions with climate of origin (Fig. 3), only the growth traits reversed the sign of that relationship across gardens. Specifically, phenology trends were mostly constant across gardens, with warmer source populations setting bud later and flushing earlier regardless of growing environment (except for bud flush in Canyonlands). However, height and trunk diameter declined as transfer distance increased (in terms of both hotter and colder climates) for populations relative to their home sites. This indicates local adaptation, where the highest productivity is observed in populations whose source climate best matches that of the garden climate. In the hottest garden (Yuma), there was a positive relationship between trunk diameter and warmer provenance climates. In the coldest garden (Canyonlands), the reverse was true, where trees from cooler provenances grew significantly taller than those from the warmer sites (Fig. 3b). Whether this higher performance of local populations is enabled by their leaf phenology and morphology traits measured here vs. additional plant functional traits is an important area for further study.