1 INTRODUCTION
Climate change is one of the most significant contemporary threats to
biodiversity worldwide (Isaac & Williams, 2007; Jing et al., 2022) and
is already affecting species and ecosystems on a global scale, and these
effects are projected to become more rapid and extensive (Field et al.,
2014). In particular, many species that are already vulnerable, such as
those with limited geographical range and/or restricted habitat
requirements, could find their available habitat area reduced or
eliminated (Hobbs et al., 2009). And these endangered species may be
threatened by a combination of disturbances associated with climate
change (such as temperature or precipitation) and ongoing anthropogenic
activities (such as habitat fragmentation, pollution, urbanization)
(Holling, 1973). For example, deforestation and forest degradation are
threatening the survival of about half of all bamboo species worldwide,
and climate change may present an additional significant threat,
especially, many bamboo species provide essential food and shelter for
wildlife species, including one of the most endangered species, the
giant panda (Tuanmu et al., 2013). Many species are responding to
changing environments by shifting their geographical ranges (Chen et
al., 2011; Parmesan, 2006).
Francois’ langur (Trachypithecus francoisi ) is an endangered
species endemic to the limestone forests of the tropical and subtropical
zones of northern Vietnam and southwestern China (Nadler et al., 2004),
with a current estimated global population of approximately 2,000
individuals (Wang et al., 2011; Zeng et al., 2013). In China, the
distribution of Francois’ langur was largely restricted to areas
characterized by karst topography at altitudes of 120-1,800 m in
Guangxi, Guizhou and Chongqing (Han et al., 2021). These areas have some
unique features that differentiate them from mountain forests, such as
steep cliffs, which cover about 10-20% of the area, lack of surface
water (Huang, 2002), and relatively poor, but more diverse vegetation
(Zhou et al., 2013). Due to forest loss, habitat destruction and
fragmentation, some Francois’ langur populations have been locally
extirpated from some parts of their historic range (Li et al., 2007;
Zeng et al., 2013). For instance, this species has become restricted to
only five isolated sites in Guizhou (Hu et al., 2011). Although the
Francois’ langur was listed as endangered on the Red List of the
International Union for Conservation of Nature (IUCN, 2020), and was
considered a first-grade protected wildlife species in China (Niu et
al., 2016; Wang et al., 2011; Wen et al., 2010), little was known about
its habitat use and factors affecting habitat selection or association,
except for some roost selection and behavioral ecology, moreover,
previous studies were limited to specific nature reserves or locations
(Hu et al., 2011; Li et al., 2020; Liu & Bhumpakphan, 2020; Zhou et
al., 2013, 2018).
Climate change effects on distribution of species, particularly rare and
endangered ones, have attracted increasing attention, and predicting how
these species will react to climate change is a major current challenge
in ecology. Some species, including plants (Mahmoodi et al., 2022; Pan
et al., 2020; Wu et al., 2016; Zhang et al., 2020), freshwater fishes
(Chu et al. 2005), mammals (Wilkening et al., 2019), coral reefs (Sill
& Dawson, 2021) and invertebrates (Martín & Abellán, 2022), have been
predicted distribution affected by the climate change with some model
methods, such as bioclimatic prediction system (BioCLIM), random forest
model (RFM) and maximum entropy models (MaxEnt) (Phillips et al., 2004,
2006). Among these models, MaxEnt is the most accurate and commonly used
model (Bellard et al., 2018; Tsoar et al., 2007). The MaxEnt is an
open-source software package for modeling species niches and
distributions through the application of a machine-learning technique.
It combines the current distribution data of the species with the
background environmental variables, associates the known occurrences
with a set of unique environmental conditions and then predicts the
probability of presence for unknown locations via the use of multiple
algorithms, and evaluates the importance of environmental variables
using jackknife (Yin et al., 2022). Thus, MaxEnt can be applied to well
predict the distribution of Francois’ langur as well as habitat
selection.
In this study, MaxEnt was applied to construct the potential
distributions of Francois’ langur. The primary objectives were as
follows: (1) to model current potential distribution of the Francois’
langur habitat across the entire range in China, (2)to explore the
relationships among the potential distribution patterns and environments
of the Francois’ langur and determine relevant influencing variables,
(3)to predict potential distribution in the future different periods,
and (4)to evaluate how the climate change in the future affect the
potential distribution of this species. Additionally, the study aims to
provide suggestions and new insights for the biodiversity conservation
of rare species.