Figure 1.
Predicted probability distribution for Procapra przewalskii using Maxent at 1×1 km.
The lapis lazuli area indicates the Qinghai Lake.
Figure 2.
Projected impacts of climate change on the distribution of Procapra przewalskii.
Percentage of range loss, range gain, range change and range turnover as predicted using presence by pixel across the three general circulation models (CCCMA, CSIRO and HADCM3) and the two climate change scenarios: A2a (liberal) and B2a (conservative), for three time slices (2020, 2050 and 2080). The solid horizontal line represents the median, the square symbol represents the mean, edges of box are quartiles, whiskers are 5th and 95th percentiles and circles are outliers.
Figure 3.
Close-up of predicted distribution of Procapra przewalskii for three time slices: 2020, 2050 and 2080.
Models are obtained with an ensemble-forecast approach across the three general circulation models (CCCMA, CSIRO and HADCM3) and the two climate change scenarios (A2a and B2a). Suitable ranges are selected by the thresholds of 0.54, 0.80 and 0.95 (panels a–c; d–f; g–i, respectively) for current and future predictions. For all panels, red indicates the current suitable habitats predicted to be unsuitable in the future; yogo blue indicates the current nonsuitable habitats predicted to be suitable in the future and blue indicates current suitable habitats predicted to stay suitable in the future. The gray solid lines represent county boundaries and yellow dotted lines, the boundaries of protected areas. The lapis lazuli area indicates the Qinghai Lake.
Figure 4.
Close-up of model disagreement among time slices (current, 2020, 2050 and 2080).
The disagreement is estimated through maximal divergences (DIVERGmax) for the two climate change scenarios: A2a (panel a) and B2a (panel b). The gray solid lines represent county boundaries and yellow dotted lines, the boundaries of protected areas. The lapis lazuli area indicates the Qinghai Lake.
Figure 5.
Predicted probability of occurrence for known Procapra przewalskii localities.
The probability of occurrence is extracted for the four time slices considered (current, 2020, 2050 and 2080) based on the predicted distributions from an ensemble-forecast approach. The solid horizontal line represents the median, the square symbol represents the mean, edges of box are quartiles, whiskers are 5th and 95th percentiles and circles are outliers.
Table 1.
Results of both multivariate analysis of variance (MANOVA) and univariate ANOVAs on the effects of time slice, threshold, GCMs, and their interactions on the four measures for estimating potential impacts of climate change: the percentage of range loss, range gain, range change and range turnover predicted.