Climate change may alter the signal of plant facilitation in Mediterranean drylands

Facilitation is an ecological interaction that has allowed plant lineages to survive past climate aridification. This same interaction can be expected to buffer the effects of current climate change, which is tending to become more arid in the Mediterranean basin. However, facilitation may wane when stress conditions are extreme. Here we argue that the erosion of the facilitation signal between Quercus ilex and its nurses detected by García-Fayos et al. (2020) along 50 years in the eastern Iberian Peninsula may have been due to the reversion of facilitation to competition imposed by an increasingly arid climate. To support this speculation, we reconstructed the climatic niche of Q. ilex and its nurses as well as the local climate change occurring in the populations studied. We found that the decreasing trend in precipitation is pushing Q. ilex out of its climatic optimum in the stressful (semi-arid) but not in the mild (sub-humid) habitats. These results suggest that facilitation will be unable to mitigate the effects of climate change, especially those related to aridification. However, other scenarios linking climatic change with herbivory and rural abandonment should be considered to fully understand the past, present and future of facilitation interactions. Reconstructing past interactions can serve as an early warning signal about the future of populations in the face of climate change.

Methods
Niche modelling:

Studied species occurrences were extracted from the Global Biodiversity Information Facility (GBIF) database (GBIF 2023, www.gbif.org). For each occurrence we extracted the precipitation of that location from the CHELSA database (Karger et al. 2017, https://chelsa-climate.org). Final species datasets used for the analyses were cleaned and redundant data were droped.

Precipitation trend in our study area:

Studied plots precipitation were extracted from quality controlled and homogenized meteorological observations from high spatial resolution databases (González‐Hidalgo et al. 2011, 2022)