Divergent selection has shaped population differences
We found the largest QST values for spring bud flush, followed by fall bud set, consistent with other studies showing high phenological divergence across latitudinal clines (Hurme 1999; Howeet al. 2003; Hall et al. 2007; Evans et al. 2016). Spring bud flush is highly differentiated among P. fremontiipopulations, with a difference of up to eight weeks observed in flush timing in three separate common garden studies (Grady et al.2015; Cooper et al. 2019; Davis et al. 2020). We also found large population differences in fall bud set timing of ~2-5 weeks across the common gardens.
Bud flush and bud set also showed evidence of divergent selection (QST > FST) in the congeneric species, P. angustifolia (Evans et al. 2016), as well as in P. tremula (Hall et al . 2007). In contrast to these studies, however, we found bud flush had a higher degree of population differentiation compared to bud set (Fig. 4). This result is intriguing, since spring bud flush is primarily governed by the accumulation of degree-days above a specific temperature, while fall bud set is often linked to precise day length periods (Howe et al . 2003). While day length is driven by latitude and is constant from year to year, temperature can vary each year. Our result of less differentiated bud set timing compared to other Populus examples could be because our study did not encompass as large a latitudinal gradient. Evans et al. (2016) used three gardens spanning over 10 degrees of latitude and Hall et al. (2007) employed two gardens approximately seven degrees apart, while this study used gardens just over five degrees of latitude in separation (Supplemental Table 1). This reduced latitudinal gradient means smaller day length differences experienced across the gardens, perhaps resulting in reduced differences in bud set expression.
Alternatively, our results of higher bud flush versus bud set differentiation could be because our source populations span the winter frost line. The southern populations in our study do not experience predictable freezing winters, which could mean that there is more room for temperature-related differentiation in bud flush earlier in a risk-free spring, and/or less selection for differentiation in bud set among populations south of the frost line late in the fall. Overall, these two phenology traits showed both the highest degree of population differentiation of all our traits and the strongest correlations with provenance climate (highest QST values; Fig. 5). This result is suggestive of climate as a strong driver of phenotypic differences among these Arizona populations.