Plant species climatic niche and its relationship with population responses to extreme drought

Resumen

Understanding how climate affects species’ distribution and performance is a central issue in ecology since its origins. In last decades, however, the interest in this question has been reactivated by the current context of climate change. Species Niche Modelling has been widely used to assess shifts in species distribution and to test the relationship between species’ climatic niche and species physiological and demographic performance, implicitly assuming that species occurrence portrays the environmental and biotic species’ suitable conditions. Nevertheless it is still largely undetermined whether these models can portray population and community responses, particularly in relation to extreme climatic episodes. In this thesis I aim at exploring the capacity of niche modelling to predict species decay under extreme climatic conditions, particularly droughts, addressing some constraints of this approach and proposing possible solutions. To achieve this goal, I counted with 3 vegetation decay datasets measured in the Spanish SE after the extreme drought year 2013-2014. Two of these datasets were based on defoliation sampling of individual plants belonging to more than 40 semiarid shrubland species (chapters 2, 4 and 5), while the other one was based on re- gional compiled data of Pinus halepensis L. affectation in plots of 1km2 (chapter 3). In second chapter I used different Species Distribution Model (SDMs) algorithms to estimate species’ climatic suitability before (1950-2000) and during the extreme drought, in order to test the possible correlation between suitability and decay, and whether the existence of this rela- tionship depended on the applied SDM algorithm. I consistently found a positive correlation between remaining green canopy and species’ climatic suitability before the event, suggesting that populations historically living closer to their species’ tolerance limits are more vulnerable to drought. Contrastingly, decreased climatic suitability during the drought period did not correlate with remaining green canopy, likely because of extremely low climatic suitability val- ues achieved during the exceptional climatic episode. In order to test whether this extremely low suitability values could derive as a consequence of only considering climatic averages when calibrating SDMs, in the third chapter I developed a method to include inter-annual climatic variability into niche characterization. I then compared the respective capacities of climatic suitabilities obtained from averaged-based and from inter-annual variability-based niches to explain demographic responses to extreme climatic events. I found that climatic suitability obtained from both niches quantifications significantly explained species demo- graphic responses. However, climatic suitability from inter-annual variability-based niches showed higher explanatory capacity, especially for populations that tend to be more geo- graphically marginal. In the fourth chapter I tried to overcome the inability of the SDMs to predict populations decay during extreme conditions, as observed in the second chap- ter, by using Euclidean distances to species’ niche in the environmental space. I compared the capacities of both population distances in the climatic environmental space and popu- lation climatic suitability derived from SDMs to explain population observed physiological and demographic responses to an extreme event. Additionally, I tested such relationship in populations located in three different bedrock sites, corresponding to a gradient of water availability. I found that SDMs-derived suitability failed to explain population decay while distances to the niche centroid and limit significantly explained population die-off, highlight- ing that population displaced farther from species’ niche during the extreme episode showed higher vulnerability to drought. The results also suggested a relevant role of some bedrocks buffering species decay responses to extreme drought events mainly according to soil water holding capacity. Finally, in the fifth chapter, I used species niche characterizations in the environmental space and demographic data to address the impact of extreme events at com- munity level. Particularly, I estimated the community climatic disequilibrium before and after a drought episode along a gradient of water availability in three bedrock types. Dise- quilibrium was computed as the difference between observed climate and community-inferred climate, which was calculated as the mean of species’ climatic optimum weighted by species abundance collected in field surveys. I found that extreme drought nested within a decadal trend of increasingly aridity led to a reduction in community climatic disequilibrium, partic- ularly when combined with low water-retention bedrocks. In addition, community climatic disequilibrium also varied before the extreme event across bedrock types, according to soils water-retention capacity. In conclusion, by developing different techniques, derived from species distribution, that characterize climatic accuracy at population and community level, this work reveals the capacity of species climatic niche to explain demographic responses under climate change-induced episodes of extreme drought.