Weak local adaptation to drought in seedlings of a widespread conifer

Abstract

There is an urgent need for better understanding how populations of trees will respond to predictable changes in climate and the intensification of extreme weather events such as droughts. The distribution of adaptive traits in seedlings is a crucial component of population adaptive potential and its characterization is important for development of management approaches mitigating the effects of climate change on forests. In this study, we used a large-scale common garden drought experiment to characterize the variation in drought tolerance, growth, and plastic responses to extreme summer drought in seedlings of 73 natural provenances of the two main varieties of Douglas-fir (Pseudotsuga menziesii var.menziesiiand var.glauca), sampled across most of their extensive natural ranges. We detected large differences between the two Douglas-fir varieties for all traits assessed, with var.glaucashowing higher tolerance to drought but slower height growth and less plasticity than var.menziesii. Surprisingly, signals of local adaptation to drought within varieties were weak within var.glaucaand nearly absent within var.menziesii. Temperature-related variables were identified as the main climatic drivers of clinal variation in drought tolerance and height growth species-wide, and in height growth within var.menziesii. Furthermore, our data indicate that higher plasticity under extreme droughts could be maladaptive in var.menziesii. Overall, our study suggests that genetic variation for drought tolerance in seedlings is maintained primarily within rather than among provenances within varieties and does not compromise growth at early stages of plant development. Given these results, assisted gene flow is unlikely to help facilitate adaptation to drought within Douglas-fir varieties, but selective breeding within provenances could accelerate adaptation.

Future carbon sequestration potential in a widespread transcontinental boreal tree species: Standing genetic variation matters!

“…Our model incorporates plasticity by integrating annual climate variation from several gardens, enabling the response of populations to vary with both temporal and spatial climate variation (Gárate‐Escamilla et al., 2019). However, a comprehensive examination of plasticity would require a more targeted and reciprocal experimental design, possibly involving cloned individuals or a series of treatments, which falls beyond the scope of the current study (e.g., Candido‐Ribeiro & Aitken, 2024; Cornille et al., 2022).…”

“…Tree age is known to have a strong influence on growth (e.g., Girardin et al., 2011) and, as such, it was included as a covariate. Similarly, climate of the provenances was included to account for local adaptation (Candido‐Ribeiro & Aitken, 2024; Gárate‐Escamilla et al., 2019). Finally, since it has been shown previously that local adaptation in black spruce populations varies within and among clusters (Prunier et al., 2012), the cluster composition of the populations was included to account for the phylogeographic structure of the species.…”

“…When the home site of every population is included as an experimental site, local adaptation (i.e., the fitness advantage of local populations over foreign ones at a given site) can be assessed (Kawecki & Ebert, 2004; Lortie & Hierro, 2022; Savolainen et al., 2007). A correlation between population responses and the climate at their respective origins, may also be interpreted as indicative of local adaptation (Candido‐Ribeiro & Aitken, 2024; Gárate‐Escamilla et al., 2019; Leites & Benito Garzón, 2023; Savolainen et al., 2013), as relationships between traits, functions and climate potentially indicates climate‐driven selection processes (De Villemereuil et al., 2016; Etterson et al., 2016; Schwinning et al., 2022). Therefore, common gardens allow the direct assessment of intraspecific variation in both growth response and potential adaptive response, and the identification of the main underlying environmental factors associated with any observed phenotypic divergence (Aitken & Bemmels, 2016).…”