Alpine species in dynamic insular ecosystems through time: conservation genetics and niche shift estimates of the endemic and vulnerable Viola cheiranthifolia

Abstract

Background and aims: Alpine oceanic ecosystems are considered amongst the most ephemeral and restricted habitats, with a biota highly vulnerable to climate changes and disturbances. As an example of an alpine insular endemic, the past and future population genetic structure and diversity, and the future distribution of Viola cheiranthifolia (Violaceae), endemic to Tenerife (Canary Islands), were estimated. The main goals were to predict distribution changes of this alpine oceanic plant under climate change, and to assist in actions for its conservation.

Methods: To perform population genetic analysis, 14 specific microsatellite markers and algorithms which considered the polyploid condition of V. cheiranthifolia were employed. The niche modelling approach incorporated temperature gradients, topography and snow cover maps. Models were projected into climate change scenarios to assess the extent of the altitudinal shifts of environmental suitability. Finally, simulations were performed to predict whether the environmental suitability loss will affect the genetic diversity of populations.

Key results: Viola cheiranthifolia presents short dispersal capacity, moderate levels of genetic diversity and a clear population genetic structure divided into two main groups (Teide and Las Cañadas Wall), showing signs of recolonization dynamics after volcanic eruptions. Future estimates of the distribution of the study populations also showed that, despite being extremely vulnerable to climate change, the species will not lose all its potential area in the next decades. The simulations to estimate genetic diversity loss show that it is correlated to suitability loss, especially in Las Cañadas Wall.

Conclusions: The low dispersal capacity of V. cheiranthifolia, coupled with herbivory pressure, mainly from rabbits, will make its adaptation to future climate conditions in this fragile alpine ecosystem difficult. Conservation actions should be focused on herbivore control, population reinforcement and surveillance of niche shifts, especially in Guajara, which represents the oldest isolated population and a genetic reservoir for the species.

Keywords: Viola cheiranthifolia; Alpine; Canary Islands; climate change; conservation genetics; genetic diversity loss; microsatellites; niche modelling; oceanic; polyploid; short-distance dispersal; volcanism.

Fig. 1.

(A) Geographical situation of the Canarian archipelago. (B) Habitat and aspect of Viola cheiranthifolia. (C) Tenerife island with the limits of Teide National Park (orange line) and (D) the Viola cheiranthifolia distribution divided in the groups Teide and the Cañadas Wall.

Fig. 2.

(A) Bar plot of co-ancestry inferred from Bayesian cluster analysis implemented on STRUCTURE and CLUMPP, with the whole set of Viola cheiranthifolia sampled individuals and localities (K = 2). (B) Bar plot of the individuals within Teide. (C) Bar plot of the individuals within the Las Cañadas Wall. (D) Principal co-ordinate analysis (PCoA) with Bruvo genetic distances (Bruvo et al., 2004) between individuals. (E) Representation of the most probable demographic scenario (Scenario 2) and the second most probable scenario (Scenario 3) with the ABC method implemented in DIYABC (Cornuet et al., 2014). The populations studied were inferred from the STRUCTURE and pairwise Rho distance results. t1, t2, t2a: time scale of divergence times measured in generations, since the present (t = 0). N1, N2, N3 and Na refer to effective population sizes, respectively, of standing populations (Teide, Guajara, and Topo de la Grieta) and from a non-sampled ancestral population. See the tested demographic scenarios in Supplementary Data Fig. S1

Fig. 3.

Average kinship coefficients F_ij between pairs of individuals at geographical distance intervals every 20 m (log scale) in Teide (A) and Guajara (B). Grey lines show the 95 % confidence intervals and the dots that are out of these margins indicate a significant deviation of the random spatial distribution (P < 0.05).

Fig. 4.

(A) Current topoclimatic suitability of Viola cheiranthifolia. (B) Projections of topoclimatic suitability under climate change scenarios. Suitability is displayed in a continuous range for the present, and in presence/absence for climate change scenarios for a threshold of 650 according to the TSS score (see the Results).

Fig. 5.

Number of suitable cells of Viola cheirantholia (suitability >650) for each climate change scenario.

Fig. 6.

Simulated levels of heterozygosity across time for the combinations of RCP and GCM for both populations, Las Cañadas Wall and Teide.