Variable germination phenology leads to divergence in environmental tracking for Phacelia fimbriata and P. purshii populations of the American southeast

Phenological studies often assume different populations across a species' range undergo development at different times to track similar environmental conditions, and the onset of development differs because the onset of optimal conditions differs across sites (‘environmental convergence’). In contrast, phenology may differ across populations because populations are locally adapted to track different environmental optima (‘environmental divergence’). This study tested if geographic variation in germination phenology results in convergence or divergence of the environment experienced by seedlings for Phacelia fimbriata and P. purshii populations. Seeds were dispersed at three common garden field sites along an elevational gradient in the southern Appalachian mountains and piedmont. Germination phenology, the temperature of seedling establishment and the fitness components of survival and flower number were measured. Additionally, germination and seedling traits were measured in controlled laboratory assays to test if germination and seedling traits were positively associated, a signature expected to emerge with adaptive habitat tracking. Populations were found to differ significantly in the timing of germination and, consequently, the temperature of seedling establishment. Divergence in the seedling environment was further supported by the laboratory assays documenting variable and positively associated temperature‐dependent germination and seedling performance across populations. Additionally, field data were used to test if populations used environmental tracking to canalize a consistent seedling environment across field sites; populations did not, however, the observed shifts in environmental tracking (i.e. temperature) were putatively adaptive. Synthesis. Geographic variation in germination phenology of Phacelia populations leads to divergence, not convergence, of the temperature experienced by seedlings during establishment, reflecting that populations are likely specialized to different seedling environments. This work demonstrates that measuring phenology in an ‘environmental currency’ rather than an ordinal ‘date of event’ enables the testing of specific hypotheses about the nature of environmental tracking through phenology: namely, if populations differ in environmental tracking, if environmental tracking is adaptive and if consistent tracking can persist with environmental change. When phenological plasticity is not sufficient to maintain environmental tracking in new locations, differences in physiological tolerances, increased specialization and/or shifts in environmental cueing may need to jointly evolve with phenology for populations to persist.