Species distribution models and a 60‐year‐old transplant experiment reveal inhibited forest plant range shifts under climate change

Abstract

Aim: Climate change causes species to shift their distributions. Individual species, however, greatly vary in their capacity to track the macroclimatic temperature increase due to differences in demography and dispersal. To better predict range shifts to climate change we need a complementary integration of long-term empirical data and predictive modelling.Location: Belgium and North-West Europe.Taxon: Hyacinthoides non-scripta¸ forest understorey plants. Methods:Complementing species distribution models with demographic data from an exceptional 60-year-old over-the-range-edge transplant experiment measured not less than 45 and 60 years after installation, we evaluated the long-term consequences of climate change on one of the most emblematic but also among the slowest colonizing plant species of European forests, bluebell Hyacinthoides non-scripta. Results:We found bluebell able to establish viable populations beyond its natural range. These results were confirmed by the SDM, showing that bluebell's potential range is considerably larger than its current range. Colonization rates of only 2 m century −1 were observed in the transplanted populations. Beyond bluebell's current range, we observed decreasing trends in population growth rates over the past 15 years. By the end of the 21st century, substantial decreases in the southern parts of bluebell's range were predicted. Main conclusions:Based on empirical and modelling results, we expect serious population declines in large parts of its current natural distribution of bluebell. Although the species is able to establish viable populations beyond the natural range edge, slow demography and local colonization rates four orders of magnitude lower than the velocity of climate change make fast enough range shifts virtually impossible in this species.

Competition mediates understorey species range shifts under climate change

“…For all focal species, the SDMs predicted substantial amounts of suitable habitat beyond their current ranges, indicating that species' potential ranges are considerably larger than their realized ranges. These results corroborate previous modelling and experimental work, suggesting that also non‐climatic factors are important in setting range limits (Louthan et al, 2015; Sanczuk et al, 2022; Shepard et al, 2021). For example, Svenning et al (2008) showed that the ranges of many plants native to European forests are still subject to post‐glacial dispersal limitation (Svenning et al, 2008).…”

“…Although rarely done, experimentally validating model predictions with field experiments can be particularly informative, as field experiments provide an integrated test of plant and population responses to the macroclimatic, microclimatic and biotic context of the environment (e.g. Greiser et al, 2020; Lee‐Yaw et al, 2016; Sanczuk et al, 2022). In our study, the results observed in the transplant experiment indeed validated the results from the SDMs.…”

“…Four long‐lived understorey plant species were selected that represent a geographical gradient in the distance between the transplant sites and the nearest range edge. All study species are forest specialists that typically have slow colonization rates and a poor competitive ability (De Pauw et al, 2022; Sanczuk et al, 2022; Svenning et al, 2008). See Figure 3 and Table S1 for the realized distribution of each species.…”

“…Although one would expect structured patterns in range shifts that closely match the unidirectional upslope and poleward moving isotherms, documented range responses to climate change are more complex, with slower, absent or even disparate trends frequently observed (Chen et al, 2011; Lenoir et al, 2020). This suggests that factors other than macroclimate change are also mediating species redistributions (De Frenne et al, 2013; Nathan et al, 2011; Sanczuk et al, 2022; Shepard et al, 2021; Zellweger et al, 2020). Based on the idea that mainly abiotic factors structure range limits at the cold ends of a species' distribution (Darwin, 1859; Louthan et al, 2015), macroclimate has been traditionally identified as the key driver of the distribution of species at large spatial scales (>10,000–200 km), whereas biotic interaction effects were considered more important at smaller scales (<10 km–0.1 m) (Wisz et al, 2013).…”

“…We focus on forest specialist plants of temperate Europe. These species are typically perennial, with long life cycles and limited colonization rates, making range shifts that are fast enough to track climate change often impossible (Sanczuk et al, 2022; Svenning et al, 2008). In addition, these species also often have a poor competitive ability (De Pauw et al, 2022), and are therefore potentially constrained in their range shifts due to competitive interactions.…”