Demographic effects of interacting species: exploring stable coexistence under increased climatic variability in a semiarid shrub community

Abstract

Population persistence is strongly determined by climatic variability. Changes in the patterns of climatic events linked to global warming may alter population dynamics, but their effects may be strongly modulated by biotic interactions. Plant populations interact with each other in such a way that responses to climate of a single population may impact the dynamics of the whole community. In this study, we assess how climate variability affects persistence and coexistence of two dominant plant species in a semiarid shrub community on gypsum soils. We use 9 years of demographic data to parameterize demographic models and to simulate population dynamics under different climatic and ecological scenarios. We observe that populations of both coexisting species may respond to common climatic fluctuations both similarly and in idiosyncratic ways, depending on the yearly combination of climatic factors. Biotic interactions (both within and among species) modulate some of their vital rates, but their effects on population dynamics highly depend on climatic fluctuations. Our results indicate that increased levels of climatic variability may alter interspecific relationships. These alterations might potentially affect species coexistence, disrupting competitive hierarchies and ultimately leading to abrupt changes in community composition.

Figure 1

Relationships between plant size and vital rates (survival, growth, probability of reproduction, fecundity) in H. squamatum and L. subulatum modelled with GLMMs. Black dots are observed values, mean values with standard errors in the cases or survival and probability of reproduction. Grey dots represent predicted values. Lines represent variations in adjusted values related to variations in two of the explanatory variables included in the most explanatory model selected by species and vital rate. Intra intraspecific interaction index, Inter interspecific interactions index, WB summer water balance, P spring rainfall.

Figure 2

Variation in population growth rate (lambda) with climatic variables (minimum winter temperature, spring rainfall and summer water balance) for Helianthemum squamatum and Lepidium subulatum. Lambdas were obtained from the Integral Projection Model built for each species separately, adjusting it with the observed values of climatic variables from 2001 to 2015. Models were adjusted for block A and intra- and interspecific covers of 30%; results for block B were similar (see Supplementary Fig. S1). Right hand panels represent population growth rate per year combining the three climatic variables. Note that the highest value of simulated lambda (black line and open symbols) corresponded to 2012 in H. squamatum and 2006 in L. subulatum, whereas the highest increase in observed population growth rate (grey lines and filled symbols) occurred in 2010 and was much higher than that adjusted by the model.

Figure 3

Variation in population growth rate (lambda) depending on intra- and interspecific covers for H. squamatum and L. subulatum. Lambdas were obtained by running the Integral Projection Models built for each species separately, using contrasting climatic conditions and adjusting intra and interspecific covers in each iteration to cover all possible combinations. White lines indicate the limit between negative (darker) and positive (lighter) population growth rates (i.e., lambda = 1). Models were adjusted for block A; results for block B were similar (see Supplementary Fig. S2). Note that covers of 80% of the two species are unrealistic.

Figure 4

Median and quartiles of stochastic population growth rates obtained from 100 simulations of 14 annual transitions in Helianthemum squamatum and Lepidium subulatum. Upper panels consider parameters specified in Supplementary Table S9 for simulations 1–4 (left) and 5–8 (right); mid panels consider parameters for simulations 9–12 (left) and 13–16 (right); lower panels correspond to simulations 17–20 (left) and 21–24 (right). Numbers in the upper-right corner indicate the year from which climatic conditions (specified in each panel) were more frequently used. Open symbols represent simulations where the interaction between H. squamatum and L. subulatum was null, whereas filled symbols represent simulations where the interaction was considered. Stable population growth rate (λ = 1) is marked with a dashed horizontal line in all plots. The letters A and B in the X axis indicate blocks. See Supplementary Table S10 for exact values of climatic variables.

Figure 5

Effect of varying initial densities (2 and 20 individuals m⁻²) of both species on the stochastic population growth rate under different climatic conditions and when recruitment is increased tenfold. Open symbols represent simulations where the interaction between H. squamatum and L. subulatum was null, whereas filled symbols represent simulations where the interaction was considered. Stable population growth rate (λ = 1) is marked with a dashed horizontal line in all plots. The letters A and B in the X axis indicate blocks. See Supplementary Table S10 for exact values of climatic variables.