Aim: Species ranges in mountain areas may shift both horizontally and altitudinally, resulting from climate change and anthropogenic impact. Two hypotheses (the abundant center hypothesis and the contagion hypothesis) have been proposed to account for patterns of horizontal range contraction. However, undulating topograph causes a mosaic of unsuitable habitats, which may complicate the spatial pattern of range contraction. We develop a framework incorporating horizontal and altitudinal range contraction patterns of a species living in mountain areas, to better understand the underlying mechanisms of species range contraction.

Location: China, North Laos, and North Vietnam

Taxon: Western black crested gibbon, Nomascus concolor

Methods: We collected occurrence data of the gibbons from various sources and modelled their potential distribution range in the 1950s, 1980s, and 2010s, using ecological niche modelling. We compared distances from the center point of the potential range in 1950s to center points of the largest 100 patches in the 1950s and the 2010s to understand the patterns of horizontal range contraction. We also calculated potential distribution within different altitudinal range for six populations in each period to understand the patterns of altitudinal range contraction.

Results: Potential horizontal distribution of the gibbons decreased by 69% from the 1950s to the 2010s. We found the 100 largest patches in 2010s were further apart from the center point of the potential range in 1950s compared to the 100 largest patches in 1950s, supporting the contagion hypothesis. No populations extended their range to higher altitude, suggesting climate change did not have a profound effect on gibbon range upward shift. All populations lost a substantial proportion of their ranges in lower altitude (500 – 1,500 m) but to different degrees, suggesting that populations experienced different anthropogenic pressures.

Main conclusions: Anthropogenic threats probably including human population increase, agricultural expansion and hunting, were more likely than climate change to have caused range contraction in western black-crested gibbons. This study highlights the importance of studying horizontal and altitudinal range contraction simultaneously.

1. Climate variable

We downloaded climate data from CRU TS v. 4.03 (0.5° * 0.5° grid) (Harris et al., 2014; Harris, Osborn, Jones, & Lister, 2020). We calculated monthly climate variables at 1km² resolution by interpolation of complex multivariate data from climate data (CRU TS v. 4.03) using thin plate smoothing splines for each period (1950s, 1980s, and 2010s) following the downscaling method provided in CCAFS (The CGIAR Research Program on Climate Change, Agriculture and Food Security) (Navarro-Racines, Tarapues, Thornton, Jarvis, & Ramirez-Villegas, 2020). We then derived the bioclimate variables from the monthly climate variables (1km²) using “dismo” package in R. We also calculated an annual heat-moisture index (AHM) by Bio1 (annual mean temperature) and Bio12 (annual precipitation) using “raster” package in R (Formula: AHM = [Bio1*0.1 + 10]/[Bio12*0.001]) (Wang, Hamann, Spittlehouse, & Murdock, 2012).

2. Landcover

We downloaded landcover maps in 2010 (current) from the MCD12Q1 V6 product (Friedl & Sulla-Menashe, 2015). This map provides global land cover at 500 meter spatial resolution at annual time. Categorical broad-leaf forest and anthropogenic influence (including crop and urban area) variables were derived from landcover data. We therefore converted all variables to continuous variables by calculating Path Distance and resampled as a raster layer in 1km² resolution using ArcMap 10.2.2 software. We also downloaded topographic data from the SRTM 90m Digital Altitude Database v4.1 (1km²) (Reuter, Nelson, & Jarvis, 2007).

Pleace check the details in the paper.