For widely distributed species, understanding the scale over which genetic variation correlates to landscape structure and composition is critical. Particularly within the context of restoration, the evolution of genetic differences may impact success if seeds are maladapted to the restoration environment. In this study, we used Geum triflorum to quantify the scale over which genetic differences for quantitative traits important to adaptation have evolved, comparing the proportion of variance attributed to broad regional- and local population-level effects. Geum triflorum is a widely distributed species spanning a range of environments, including alvar and prairie habitats, which have extreme regional differences in soil-moisture availability. Alvar habitats are regions of thin soil over limestone that experience substantial seasonal variation in water availability, from flooding to desiccation annually. This contrasts with prairie habitats, whose deeper soils mitigate irregular flood–desiccation cycles. Using a common garden experiment, we evaluated 15 traits broadly grouped into three trait classes: resource allocation, stomatal characteristics, and leaf morphological traits for individuals sourced from prairie and alvar environments. We quantified the proportion of trait variance explained by regional- and population-scale effects and compared the proportion of regional- and population-trait variances explained across trait classes. Significant regional differentiation was observed for the majority of quantitative traits; however, population-scale effects were equal or greater than regional effects, suggesting that important genetic differences may have evolved across the finer population scale. Stomatal and resource allocation trait classes exhibited substantial regional differentiation relative to morphological traits, which may indicate increased strength of selection for stomatal and resource allocation traits relative to morphological traits. These patterns point towards the value in considering the scale over which genetic differences may have evolved for widely distributed species and identify different functional trait classes that may be valuable in establishing seed transfer guidelines.

The ability of plants to track shifting fitness optima is crucial within the context of global change, where increasing environmental extremes may have dramatic consequences to life history, fitness, and ultimately species persistence. However, to track changing conditions relies upon the complex relationship between genetic and environmental variance, where selection may favor plasticity, the evolution of genetic differences, or both depending on the spatial and temporal scale of environmental heterogeneity. Over three years, we compared the genetic and environmental components of phenological and life-history variation in a common environment for the spring perennial Geum triflorum. Populations were sourced from alvar habitats that exhibit extreme, but predictable annual flood-desiccation cycles and prairie habitats that exhibit similar, but less predictable variation in water availability. Narrow-sense heritabilities were generally higher for early life history (emergence probability) relative to later life history traits (total seed mass), indicating that traits associated with establishment within an environment are under stronger genetic control relative to later life-history fitness expressions, where plasticity may play a larger role. This pattern was particularly notable in seeds sourced from environmentally extreme, but predictable alvar habitats relative to less predictable prairie seed sources. Fitness landscapes based on seed source origin, largely characterized by varying water availability and flower production, described selection as the degree of maladaptation to the prairie common garden environment relative to seed source environment. Plants from alvar populations were consistently closer to the fitness optimum across all years. Annually, the breadth of the fitness optimum expanded primarily along a moisture gradient, with inclusion of more populations onto the expanding optimum. These results highlight the importance of temporally and spatially varying selection for the evolution of life history, indicating plasticity within perennial systems may over time become the primary mechanism to track fitness for later life history events.

Background and Aims The ability of plants to track shifting fitness optima is crucial within the context of global change, where increasing environmental extremes may have dramatic consequences to life history, fitness, and ultimately population persistence. However, tracking changing conditions relies on the relationship between genetic and environmental variance, where selection may favor plasticity, the evolution of genetic differences, or both depending on the spatial and temporal scale of environmental heterogeneity. Methods Over three years, we compared the genetic and environmental components of phenological and life-history variation in a common environment for the spring perennial Geum triflorum. Populations were sourced from alvar habitats that exhibit extreme, but predictable annual flood-desiccation cycles and prairie habitats that exhibit similar, but less predictable variation in water availability. Key Results Heritability was generally higher for early life history (emergence probability) relative to later life history traits (total seed mass), indicating that traits associated with establishment are under stronger genetic control relative to later life-history fitness expressions, where plasticity may play a larger role. This pattern was particularly notable in seeds sourced from environmentally extreme, but predictable alvar habitats relative to less predictable prairie environments. Fitness landscapes based on seed source origin, largely characterized by varying water availability and flower production, described selection as the degree of maladaptation of seed source environment relative to the prairie common garden environment. Plants from alvar populations were consistently closer to the fitness optimum across all years. Annually, the breadth of the fitness optimum expanded primarily along a moisture gradient, with inclusion of more populations onto the expanding optimum. Conclusions These results highlight the importance of temporally and spatially varying selection to life history evolution, indicating plasticity may become a primary mechanism needed to track fitness for later life history events within perennial systems.