Retreating or standing: responses of forest species and steppe species to climate change in arid Eastern Central Asia

Abstract

Background: The temperature in arid Eastern Central Asia is projected to increase in the future, accompanied by increased variability of precipitation. To investigate the impacts of climate change on plant species in this area, we selected two widespread species as candidates, Clematis sibirica and C. songorica, from montane coniferous forest and arid steppe habitats respectively.

Methodology/principal findings: We employed a combined approach of molecular phylogeography and species distribution modelling (SDM) to predict the future responses of these two species to climate change, utilizing evidence of responses from the past. Genetic data for C. sibirica shows a significant phylogeographical signal (N ST > F ST, P<0.05) and demographic contraction during the glacial-interglacial cycles in the Pleistocene. This forest species would likely experience range reduction, though without genetic loss, in the face of future climate change. In contrast, SDMs predict that C. songorica, a steppe species, should maintain a consistently stable potential distribution under the Last Glacial Maximum (LGM) and the future climatic conditions referring to its existing potential distribution. Molecular results indicate that the presence of significant phylogeographical signal in this steppe species is rejected and this species contains a high level of genetic differentiation among populations in cpDNA, likely benefiting from stable habitats over a lengthy time period.

Conclusions/significance: Evidence from the molecular phylogeography of these two species, the forest species is more sensitive to past climate changes than the steppe species. SDMs predict that the forest species will face the challenge of potential range contraction in the future more than the steppe species. This provides a perspective on ecological management in arid Eastern Central Asia, indicating that increased attention should be paid to montane forest species, due to their high sensitivity to disturbance.

Figure 1. Sampling locations and distributions of population groups in Clematis sibirica (A) and C. songorica (E) in arid Eastern Central Asia.

Population codes are consistent with the population names shown in Table S1. The color of each pie corresponds to the cpDNA population groups; the symbols with different colors correspond to the nrITS population groups. (B, F) Median-joining networks for recovered chlorotypes in these two species. The colors within each chlorotype are consistent with the cpDNA population groups. (C) Median-joining networks for recovered ribotypes in Clematis sibirica. The colors within each ribotype are consistent with the nrITS population groups. (D, G) Habitats of these two species.

Figure 2. Spatial genetic landscapes constructed from the cpDNA sequences (left) and nrITS (right) sequences across the total distributions of Clematis sibirica (A) and C. songorica (B).

Figure 3. Divergence time (ka BP.) of Clematis sibirica in cpDNA (A) and nrITS (B) datasets based on BEAST analysis.

Figure 4. Modelling distribution of Clematis sibirica.

Maps depicting potential distribution of Clematis sibirica in arid Eastern Central Asia during present-day and at LGM based on the MIROC model, and future (2080’s) based on the A1b and A2a scenarios.

Figure 5. Modelling distribution of Clematis songorica.

Maps depicting potential distribution of Clematis songorica in arid Eastern Central Asia during present-day and at LGM based on the MIROC model, and future (2080’s) based on the A1b and A2a scenarios.