North American tree migration paced by fecundity and recruitment through contrasting mechanisms east and west

Global forest diebacks are the beginnings of change that will be controlled by tree migration, which combines two uncertain processes, tree fecundity and recruitment. Knowledge of how, and where, tree migration can proceed is critical for adaptive management of forest resources and conservation efforts. The initial stage of seed production is erratic and poorly observed, with most studies limited to few trees, few species and few sites. At the next stage, tree recruitment is typically too sporadic to characterize at landscape scales. Neither seed production nor seedling recruitment have been quantified or linked to climate and habitat variables at scales needed to evaluate the changes happening now or to anticipate the diversity and structure of 21st century forests. As part of the masting inference and forecasting (MASTIF) project, we synthesized continental-scale data for tree fecundity gathered over the last half century and combined it with forest inventories to connect adult trees (basal area) to i) fecundity (seeds per basal area) and ii) recruitment (recruits per seed). A dynamic model fitted to >107 tree years of fecundity data provided estimates tree-by-year fecundity. A predictive distribution for the continent combines the fitted mode with 105 trees from Forest Inventory Analysis (FIA), Canadian National Forest Inventory(CNFI) and the National Ecological Observatory Network (NEON).

Results show continent-wide migration as a balance between regional shifts in fecundity that can diverge from conditions that favour establishment, with clear differences in eastern and western North America. In moist eastern states, the geographic centers for fecundity are most commonly displaced south of tree basal area for the same species. This relationship would be expected if optimal conditions for seed production lie to the south of optimal conditions for growth and survival, despite potential benefits of warming poleward. In the dry west and north-central, fecundity is for some species displaced northwest of tree basal area, as would be expected if the high-rainfall north-west is predisposed to lead migration as the continent warms. The east-west contrast diminishes at the transition from fecundity to recruits per seed, which tends to be shifted north in both regions. The net continent-wide migration by contrasting east-west controls highlight interactions, with fecundity primed to lead tree migration in the west, and fecundity slowing progress in the east. The possibility of fecundity limitation offers one explanation for migration lag in species expected to track climate warming by expanding poleward.