Forecasting range shifts of dioecious plants under climate change

Abstract

Global climate change has triggered an urgent need for predicting the reorganization of Earth's biodiversity. For dioecious species (those with separate sexes), it is unclear how commonly unique climate sensitivities of females and males could influence projections for species-level responses to climate change. We developed demographic models of range limitation, parameterized from geographically distributed common garden experiments, with females and males of a dioecious grass species (Poa arachnifera) throughout and beyond its range in the south-central U.S. We contrasted predictions of a standard female-dominant model with those of a two-sex model that accounts for feedbacks between sex ratio and vital rates. Both model versions predict that future climate change will induce a poleward shift of niche suitability beyond current northern limits. However, the magnitude of the poleward shift was underestimated by the female-dominant model because females have broader temperature tolerance than males but become mate-limited under female-biased sex ratios, which are forecasted to become more common under future climate. Our results illustrate how explicitly accounting for both sexes can enhance population viability forecasts and conservation planning for dioecious species in response to climate change.

Keywords: global warming; matrix projection model; population dynamics; sex ratio.

Fig. 1.

Geographic distribution and climatic variation of P. arachnifera in Texas, Oklahoma, and Kansas. (A) Colored points indicate common garden locations in Texas, Oklahoma, and Kansas. Gray diamonds represent GBIF occurrences, + show source populations. (B and C) Changes in growing and dormant season climate for each site. Arrows in B and C connect past (1901–1930) and present (1990–2019) climate normals, and present and future (2071–2100) climate normals under RCP4.5 and RCP8.5 forecasts. Future forecasts are from MIROC5 but other climate models show similar patterns.

Fig. 2.

Sex specific demographic response to climate across species range. (A–D) Probability of survival as a function of precipitation and temperature of the growing and dormant season. (E–H) Change in number of tillers as a function of precipitation and temperature of the growing and dormant season. (I–L) Probability of flowering as a function of precipitation and temperature of growing and dormant season. (M–P) Change in number of panicles as a function of precipitation and temperature of the growing and dormant season. Points show means by site for females (orange) and males (green). Points sizes are proportional to the sample sizes of the mean and are jittered. Lines show fitted statistical models for females (orange) and males (green) based on posterior mean parameter values. The fitted lines were estimated using only one climate covariate, while the other covariates and size were held constant (mean).

Fig. 3.

Niche suitability across seasonal climate space predicted by female-dominant and the two-sex models. Panels (A–D) show predicted probabilities of self- sustaining populations, Pr ($\lambda >1$) conditional on precipitation and temperature of the dormant and growing season. Panels (E) and (F) show the difference in niche estimation between the female dominant model and the two-sex model for each season. The pink color indicates the female dominant (F) while the violet represents the two-sex models (FM). The dash line represents the mean probability for each model.

Fig. 4.

Geographic projections of population viability predicted by the female-dominant and two-sex models. Maps show past, current, and future (RCP 4.5 and RCP 8.5) range shifts based on the predicted probabilities of self-sustaining populations Pr ($\lambda >1$). The last panel shows the difference in geographic projections of population viability between the female-dominant model and the two-sex model for each season. Future projections are based on the CMCC-CM climate model. Black dots on the panel showing current climatic conditions represent all known presence records from GBIF. These occurrences are located within areas of higher population fitness Pr ($\lambda >1$), supporting the validity of our approach in predicting range shifts.