A method for experimental warming of developing tree seeds with a common garden demonstration of seedling responses

Background: Forest dieback driven by rapid climate warming threatens ecosystems worldwide. The health of forested ecosystems depends on how tree species respond to warming during all life history stages. While it is known that seed development is temperature-sensitive, little is known about possible effects of climate warming on seed development and subsequent seedling performance. Exposure of seeds to high air temperatures may influence subsequent seedling performance negatively, though conversely, warming during seed development may aid acclimation of seedlings to subsequent thermal stress. Technical challenges associated with in-situ warming of developing tree seeds limit understanding of how tree species may respond to seed development in a warmer climate. Results: We developed and validated a simple method for passively warming seeds as they develop in tree canopies to enable controlled study of climate warming on seedling performance. We quantified thermal effects of the cone-warming method across individual pine trees and stands by measuring the air temperature surrounding seed cones using thermal loggers and the temperature of seed cone tissue using thermocouples. We then investigated seedling phenotypes in relation to the warming method through a common garden study. We assessed seedling morphology, physiology, and mycorrhizal nodulation in response to experimental cone-warming in 20 seed-source-tree canopies on the San Francisco Peaks in northern Arizona, USA. The warming method increased air temperature surrounding developing seed cones by 2.1 degrees C, a plausible increase in mean air temperature by 2050 under current climate projections. Notable effect sizes of cone-warming were detected for seedling root length, shoot length, and diameter at root collar using Cohen's Local f(2). Root length was affected most by cone-warming, but effect sizes of cone-warming on root length and diameter at root collar became negligible after the first year of growth. Cone-warming had small but significant effects on mycorrhizal fungal richness and seedling multispectral near-infrared indices indicative of plant health. Conclusions: The method was shown to reliably elevate the temperature surrounding seed cones and thereby facilitate experimental in-situ climate warming research on forest trees. The method was furthermore shown to influence plant traits that may affect seedling performance under climate warming.