2.2 Ecological niche modeling
We used ecological niche modeling (ENM; Soberon & Peterson 2005) to characterize the spatial distribution of suitable conditions forC. chuniana and locate potential distributional areas in conjunction with historical biological inferences. We based the analysis on high-resolution paleoclimate data inferred for the Last Interglacial (LIG, 0.14 ~ 0.12 Ma), Last Glacial Maximum (LGM, ≈ 0.02 Ma), Middle Holocene (MH, ≈ 0.006 Ma), and current. Bioclimatic variables were downloaded from the WorldClim database (http://worldclim.org/download; Fick & Hijmans, 2017) for the four different stages with 2.5-minute spatial resolution. The LIG, LGM, and MH data were obtained from circulation model simulation of the Community Climate System Model (CCSM) (Collins et al., 2006), which provides downscaled high-resolution estimates of the climate parameters (Hijmans, Cameron, Parra, Jones, & Jarvis, 2005). We used the maximum entropy modeling method with Maxent v3.3.2 (Phillips, Anderson, & Schapire, 2006). Herbarium specimen records of C. chuniana from nine herbaria (A, IBEC, IBK, IBSC, KUN, LBG, NMNH, PE and SCFI) as well as our sample collection locations were used to determine the locations of populations considered to occur at present. The analysis pipelines and parameter settings, including the occurrence points, current/past bioclimatic variables as well as the convergence threshold and maximum number of iterations were all as in Dai et al. (2011) and Gong et al. (2016). Model accuracy was assessed by evaluating the area under the curve (AUC) of the receiver-operating characteristic (ROC) plot (Phillips et al. , 2006), where scores higher than 0.70 were considered to show good model performance (Fielding & Bell, 1997). This approach is thus conservative, identifying the minimum predicted area possible while maintaining zero omission error in the training dataset (Pearson, Raxworthy, Nakamura, & Peterson, 2007). The most influential climate factors were also compared, including precipitation and temperature in each month or on average.