Adapting Mediterranean forests to climate change by testing alternative species: Insights from recent planting trials across an elevational gradient in southeastern France

Abstract

To adapt Mediterranean forests to climate change-induced disturbances—such as severe droughts and temperature extremes—introducing "alternative" tree species whose field performance remains poorly tested is a critical management option. This study evaluated the early survival and height growth of 12 coniferous and broadleaved species, comprising non-native taxa and underutilized native species planted outside their traditional geographic ranges. Over 14 000 seedlings were introduced across an elevational gradient (72–1575 m a.s.l.) encompassing 29 sites and 78 planting units in southeastern France. Survival and height growth were monitored 5.2 ± 0.04 years post-plantation. Species-specific responses were analyzed using generalized linear and linear mixed models against climatic and edaphic predictors derived at the site and planting unit scales. Early performance was highly species-specific and primarily driven by the summer and annual climatic water balance and thermal stress, followed by winter thermal constraints and localized site factors. The highest overall performance (co-occurring high survival and growth) was exhibited by two native broadleaves ( Fraxinus ornus and Sorbus domestica ). In contrast, non-native species displayed lower growth rates (except Alnus cordata ) and divergent survival trajectories: survival was high for Abies cephalonica , Cedrus atlantica and Calocedrus decurrens but low for Pseudotsuga menziesii , Pinus heldreichii and Alnus cordata . Despite inherent study limitations and the necessity for long-term monitoring, these results highlight the promising potential of select underutilized indigenous species. Conversely, the highly variable performance of non-native species underscores the need for rigorous multi-site testing before integrating them into climate-adaptive forest management and species selection frameworks.