Changes in suitable habitat for the critically endangered Northern white-cheeked gibbon (Nomascus leucogenys) in the Western Nghe An Biosphere Reserve, Vietnam: Implication for conservation

Abstract

Several recent studies have highlighted that change in land use and land cover (LULC) is the main threat causing the decline and extinction of certain species. Gibbons (Hylobatidae) could be excellent examples, on account of their potential for extinction in the near future under the effects of LULC changes due to their particular ecological traits. This study aims to model the current suitable habitat of the Northern white-cheeked gibbon ( Nomascus leucogenys (Ogilby, 1840)) in the Western Nghe An Biosphere Reserve (BR), Vietnam and assess the changes in its suitable habitat following the changes in LULC from 1990 to 2020. The maximum entropy approach (MaxEnt) was used to predict the suitable habitat of the gibbon using its occurrence localities and environmental predictors. The model analysis showed that the “Distance to Agriculture” variable had the strongest impact on the gibbons’ suitable habitat. Our results predicted the present suitable habitat of the gibbon species at approximately 4,022.42 km ² (30.95% of the overall BR area) in three spatially separated areas inside the Western Nghe An BR. Furthermore, the suitable habitat areas of the gibbon in 1990, 2000, and 2010 were projected at roughly 4,347.68 km ² , 4,324.97 km ² , and 2,750.21 km ² , respectively, following a decreasing trend from 1990 to 2010, but a gradual increase between 2010 and 2020. The suitable habitat of the gibbon inside three core protected areas (Pu Mat National Park, Pu Huong, and Pu Hoat Nature Reserves) showed a continually increasing trend from 1990 to 2020. Our results highlighted the influence of LULC changes and the role of the protected area network in gibbon conservation. The information from the study provides a quantitative baseline for the future conservation of the critically endangered gibbon in the Western Nghe An BR.

Full text

Changes in suitable habitat for the critically
endangered Northern white-cheeked gibbon
(Nomascus leucogenys) in the Western Nghe An
Biosphere Reserve, Vietnam:
Implication for conservation
Khoa Van Phung1, Dung Van Tran2, Hai anh Dong2,
Vinh Quang Luu2, Van Bac Bui3, inh Tien Vu2
1Department of Environmental Management, Faculty of Forest Resources and Environmental Management,
Vietnam National University of Forestry, Xuan Mai, Chuong My, Ha Noi, Vietnam 2Department of Wildlife,
Faculty of Forest Resources and Environmental Management, Vietnam National University of Forestry, Xuan
Mai, Chuong My, Ha Noi, Vietnam 3Department of Plant Protection, Faculty of Forest Resources and Envi-
ronmental Management, Vietnam National University of Forestry, Xuan Mai, Chuong My, Ha Noi, Vietnam
Corresponding author: Dung Van Tran (trandungfuv@gmail.com)
Academic editor: C. Knogge|Received 14 July 2022|Accepted 31 January 2023|Published 21 February 2023
https://zoobank.org/4778ABCC-2A21-45E0-8D8C-DAD53FB86D3A
Citation: Van Phung K, Van Tran D, anh Dong H, Quang Luu V, Bac Bui V, Tien Vu T (2023) Changes in suitable
habitat for the critically endangered Northern white-cheeked gibbon (Nomascus leucogenys) in the Western Nghe An
Biosphere Reserve, Vietnam: Implication for conservation. Nature Conservation 51: 167–188. https://doi.org/10.3897/
natureconservation.51.90373
Abstract
Several recent studies have highlighted that change in land use and land cover (LULC) is the main threat
causing the decline and extinction of certain species. Gibbons (Hylobatidae) could be excellent examples,
on account of their potential for extinction in the near future under the eects of LULC changes due to
their particular ecological traits. is study aims to model the current suitable habitat of the Northern
white-cheeked gibbon (Nomascus leucogenys (Ogilby, 1840)) in the Western Nghe An Biosphere Reserve
(BR), Vietnam and assess the changes in its suitable habitat following the changes in LULC from 1990
to 2020. e maximum entropy approach (MaxEnt) was used to predict the suitable habitat of the gib-
bon using its occurrence localities and environmental predictors. e model analysis showed that the
“Distance to Agriculture” variable had the strongest impact on the gibbons’ suitable habitat. Our results
predicted the present suitable habitat of the gibbon species at approximately 4,022.42 km2 (30.95% of
the overall BR area) in three spatially separated areas inside the Western Nghe An BR. Furthermore, the
suitable habitat areas of the gibbon in 1990, 2000, and 2010 were projected at roughly 4,347.68 km2,
Nature Conservation 51: 167–188 (2023)
doi: 10.3897/natureconservation.51.90373
https://natureconservation.pensoft.net
Copyright Khoa Van Phung et al. This is an open access article distributed under the terms of the Creative Commons Attribution License (CC
BY 4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
RESEARCH ARTICLE
Launched to accelerate biodiversity conservation
A peer-reviewed open-access journal

Khoa Van Phung et al. / Nature Conservation 51: 167–188 (2023)
168
4,324.97 km2, and 2,750.21 km2, respectively, following a decreasing trend from 1990 to 2010, but a
gradual increase between 2010 and 2020. e suitable habitat of the gibbon inside three core protected
areas (Pu Mat National Park, Pu Huong, and Pu Hoat Nature Reserves) showed a continually increasing
trend from 1990 to 2020. Our results highlighted the inuence of LULC changes and the role of the
protected area network in gibbon conservation. e information from the study provides a quantitative
baseline for the future conservation of the critically endangered gibbon in the Western Nghe An BR.
Keywords
Gibbon, land use & land cover, MaxEnt, Nomascus, species distribution modeling, suitable distribution
Introduction
Southeast Asia contains approximately 15% of the world’s tropical forests and is home
to at least four globally important biodiversity hotspots (Myers et al. 2000). However,
the region is also assessed as a deforestation hotspot. From 2005 to 2015, around 80
million ha (28%) of natural forests were lost in Southeast Asia (Estoque et al. 2019).
Researchers also predicted that the forest in Southeast Asia would experience a signi-
cant change, either shrinking by 5.2 million ha or growing by 19.6 million ha, depend-
ing on the choice of the pathway (Estoque et al. 2019). Several recent studies have high-
lighted that the change in land use and land cover (LULC) is the main threat driving
the decline and extinction of species, with loss in the habitat sustainability of wildlife
species (Jetz et al. 2007; Díaz et al. 2019; Su et al. 2021), especially for large-bodied
species. Gibbons could be excellent examples revealing the impacts of the extreme ef-
fects of LULC changes due to their particular ecological traits. e species are strictly
arboreal and highly frugivorous, preferring closed-canopy broadleaved evergreen forests
(Geissmann et al. 2000; Ruppell 2013). Territoriality and monogamy are observed in
almost all gibbon species (Mitani 1987; Brockelman et al. 2014). Gibbons also respond
sensitively to habitat degradation and fragmentation, as well as anthropogenic activities
(Geissmann et al. 2000; Rawson et al. 2011; Tran and Vu 2020; Sarma et al. 2021).
e northern white-cheeked gibbon (NWCG) (Nomascus leucogenys) is classied as
a Critically Endangered species on the International Union for Conservation of Nature
(IUCN) Red List (Rawson et al. 2020). e species is native to northwestern Viet-
nam, northern Lao PDR, and southwestern China (Rawson et al. 2011; Rawson et al.
2020). e gibbon is believed to be extinct or functionally extinct in China (Fan et al.
2014), while the largest remaining population probably persists in Lao PDR, although
its current status is still unclear (Rawson et al. 2011; Rawson et al. 2020). In Vietnam,
the distribution of NWCG is restricted in the north by the Black River (Geissmann et
al. 2000; Rawson et al. 2011) and limited in the south by the Rao Nay river (inh et
al. 2010). e species’ total population was estimated to be only around 300 groups
remaining in Vietnam, most of them persisting in a few isolated forest blocks close to
the Lao PDR border (Rawson et al. 2011).
e largest population of NWCG was found in Pu Mat National Park (NP),
with 22 conrmed groups; the site was also considered as the highest priority area for

Changes in suitable habitat for the critically endangered gibbon 169
conserving the species (Rawson et al. 2011). e occurrence of the species was also
conrmed in several other protected areas, although their population is likely low or
very low, for example, in Muong Nhe Nature Reserve (NR), Sop Cop NR, Xuan Lien
NR, Vu Quang NP, and Pu Huong NR (Rawson et al. 2011). It should also be noted
that the species seems to have been extirpated from many other protected areas, includ-
ing Hang Kia – Pa Co NR, Ngoc Son – Ngo Luong NR, and Pu Luong NR (Rawson
et al. 2011). Hunting and habitat loss are currently considered as the major threats to
NWCG, especially with the high rate of forest fragmentation and degradation pushing
the species into a few isolated forest areas (Rawson et al. 2011; Rawson et al. 2020).
To date, studies on gibbons in Vietnam have mainly focused on population size assess-
ments (Luu and Rawson 2009, 2010, 2011; Hoang et al. 2010; Ha et al. 2011; Tran
and Vu 2020). Gibbons require high-quality habitat with high food abundance and
a dense canopy (Geissmann et al. 2000; Sarma et al. 2021). However, limited eorts
have been made to monitor gibbon habitat, and no current studies report the changes
in suitable habitat for gibbons in Indochina.
In this study, we aimed to predict the suitable habitat of NWCG in the Western
Nghe An BR by using species distribution modeling (MaxEnt) based on identied
presence localities of the species and environmental predictors. We also attempted to
project suitable habitat uctuation in relation to the changes in LULC from 1990 to
2020. Based on the results, we discussed the impact of LULC changes and the role
of the protected area network in gibbon conservation in Vietnam. Our results are
a baseline for researchers, conservationists, and wildlife and habitat managers to aid
decision-making and plan future conservation strategies for NWCG in the Western
Nghe An BR, Vietnam.
Methods
Study area
e Western Nghe An Biosphere Reserve (BR) was recognized by UNESCO as the
6th Biosphere Reserve of Vietnam in 2007. e Western Nghe An BR (18°34'43"–
19°59'44"N, 103°52'28"–105°30'07"E) is located in the western part of Nghe An
province, central Vietnam, covering an area of 12,997.95 km2 with three functional
zones: the core zone (1,683.01 km2), buer zone (6,085.47 km2), and transition zone
(5,229.47 km2). e Biosphere Reserve is expected to create a green corridor between
the three protected areas: Pu Mat NP, Pu Huong NR, and Pu Hoat NR. e forest
cover of the Western Nghe An BR is approximately 66.4% of the total area. e area is
home to more than 3,000 vascular plants and more than 940 vertebrate animals, with
several species listed as threatened, rare, and endemic. More than 900,000 people in
ve indigenous minority groups currently reside within Western Nghe An BR (West-
ern Nghe An Biosphere Reserve 2017). Most wild groups of NWCG in the Western
Nghe An BR were detected in the three protected areas, including Pu Mat NP, Pu
Huong and Pu Hoat NR. Pu Mat NP is one of the largest remaining natural forests in

Khoa Van Phung et al. / Nature Conservation 51: 167–188 (2023)
170
Vietnam and is believed to be home to the largest population of the NWCG with an
estimated 455 gibbons in 130 dierent groups (Luu and Rawson 2011). However, the
forests in Pu Huong and Pu Hoat NR are isolated and probably not eectively linked
ecologically to the Pu Mat NP. More than a decade ago, researchers conrmed at least
seven groups of the gibbon remaining in the Pu Hoat NR (Luu and Rawson 2009).
However, the last eld survey in Pu Hoat NR and adjunction areas detected at least 40
gibbon groups (Pu Hoat NR 2021).
Occurrence data
To predict the suitable habitat of NWCG in the Western Nghe An BR, we collected
the occurrence of the endangered species through our eld surveys in Pu Mat NP, Pu
Huong NR, and Pu Hoat NR from March–May 2021 (Phung and Dong 2021), and
another published document (Pu Hoat NR 2020). e localities of gibbons in Pu Mat
NP, Pu Huong, and Pu Hoat NR were used as presence data for the MaxEnt model
in 2020. For presence data on the species for the 2010 model, we gathered data from
published documents (Luu and Rawson 2009, 2010, 2011). Initially, we collected 36
presence localities for the 2010 model, and 98 localities for the 2020 model. To avoid
spatial autocorrelation among localities that could cause an overestimation, we used the
“spin” package (Aiello-Lammens et al. 2015) in R version 4.1.2 to thin out the loca-
tions of gibbons within one km radius by randomly selecting one location and remov-
ing the others, similar to the models for the Southern white-cheeked gibbon (N. siki) in
Tran et al (2023). One km was used as a criterion for thinning because gibbons in the
genus Nomascus have a relatively small home range, around 0.45 km2 for the Southern
yellow-cheeked gibbon (N. gabriellae; Hai et al. 2020). Consequently, we used 25 and
40 occurrence localities of NWCG for the 2010, and 2020 models, respectively (Fig. 1).
Environmental predictors
To predict the suitable habitat of NWCG in the Western Nghe An BR, we obtained
variables on climate, topography, and LULC. e variables were selected based on our
knowledge of the habitat of the gibbon and consulting the available research sources.
For the climatic variables, we downloaded 19 predictors with the highest available res-
olution (30 arcseconds) from the World Clim database (http://www.worldclim.com/;
Fick and Hijmans 2017). e variables include 11 and eight layers of temperature
and precipitation, respectively, which were derived from the monthly temperature and
rainfall values in order to generate more biologically meaningful variables. According
to Tran and Vu (2020), elevation might also highly contribute to the species distribu-
tion modeling of gibbons. erefore, we used the Digital Elevation Model with a
resolution of 30 × 30 m (available from https://earthexplorer.usgs.gov) as elevation
variables. en, we calculated the Slope and Aspect variables based on the Digital El-
evation Model in ArcMap version 10.2 (ESRI). To assess the change of habitat suitabil-
ity of the threatened gibbon, we extracted the LULC of the Western Nghe An BR from
the LULC of Vietnam in 1990, 2000, 2010, and 2020 at 30 × 30 m resolution (Fig.2)

Changes in suitable habitat for the critically endangered gibbon 171
Figure 1. e occurrence localities in 2020 (red) and 2010 (green) of NWCG in the Western Nghe An
BR were used for predicting the suitable habitat by the MaxEnt model.
available at https://www.eorc.jaxa.jp/ALOS/en/dataset/lulc/lulc_vnm_v2109_e.htm
(Phan et al. 2021). e distance to the settlement plays an important role in modeling
suitable habitat for gibbons (Tran and Vu 2020). Here, we extracted the settlement
area from the LULC layer (Phan et al. 2021). e agricultural land was also selected
from the LULC layer (Phan et al. 2021), including rice paddies, woody crops, other
croplands, and in-house crops’ categories. To assess the eect of settlement and ag-
ricultural land on the suitable habitat of the gibbon, we calculated the distance to
settlement and agricultural land by the Euclidean Distance tool in ArcMap version
10.2 to create two variables: Distance to Settlement, and Distance to Agriculture in
1990, 2000, 2010 and 2020. Due to the dierence in the resolution of variables, we
resampled all variables to a resolution of 30 × 30 m using the Resample tool in ArcMap
version 10.2. e collinearity among environmental predictors might cause the overes-
timation of species distribution modeling. us, we calculated Pearson’s correlation in-
dex (r) using ENMTools version 1.4.4 (Warren et al. 2010). With the highly correlated
predictor pair (|r| > 0.80), we eliminated one variable and kept the remaining variables
for nal analysis (Nazeri et al. 2012). To select or remove variables, we considered our
understanding of the ecology of gibbons, and other publications on predicting habitat

Khoa Van Phung et al. / Nature Conservation 51: 167–188 (2023)
172
Figure 2. e LULC in the Western Nghe An BR in A 1990 B 2000 C 2010, and D 2020 extracted
from Phan et al. (2021).
suitability, including Sarma et al. (2015, 2021); Tran and Vu (2020). Additionally, we
also ran a model with all the variables and then considered the contribution of each
variable. Finally, we used 13 predictors for the nal models (Table 1).

Changes in suitable habitat for the critically endangered gibbon 173
Ecological niche modeling processing
To generate the suitable habitat model for NWCG in the Western Nghe An BR, we ap-
plied the Maximum Entropy approach (MaxEnt) version 3.4.4 available from https://
biodiversityinformatics.amnh.org/open_source/maxent/. e MaxEnt model predicts
the habitat suitability of species based on the probable occurrence of species in dis-
tinct localities nding the maximum entropy distribution of environmental predictors
(Phillips et al. 2006). e model was assessed as one of the most accurate species distri-
bution models and has been broadly applied to predict suitable habitats for species, es-
pecially for the small sample size of presence data (Elith et al. 2006; Phillips et al. 2006;
Raxworthy et al. 2007). Here, we applied the default setting for the MaxEnt models
as suggested by the model developer (Phillips et al. 2006). e area under the receiv-
ing operator curve (AUC) has been widely used to evaluate the accuracy of the model
performance (Elith 2000; Phillips et al. 2006; Nazeri et al. 2012). e higher AUC
values indicate better model performance: inability to predict model (AUC < 0.5),
poor performance (AUC = 0.5–0.7), moderate performance (AUC = 0.7–0.9), excel-
lent performance (AUC = 0.9–1) (Peterson et al. 2011).
To assess the changes in the suitable habitat of NWCG from 1990 to 2020, we
ran separately two models for 2010 and 2020 based on the presence localities and
environmental data set of each year. For the model of 2020, we ran the model using
10-folds cross-validation to evaluate the model, while we applied the jackknife method
due to the small sample size for the model 2010 (Pearson et al. 2007). Due to the lack
of occurrence data of the gibbons before 2000, we applied the projection function of
the MaxEnt model to project the suitable habitat into the past (2000 and 1990) based
on the presence localities in 2010, 2020 and the available environmental data set in
2000 and 1990. Because 2010 was closer to 2000 and 1990 than 2020, we assumed
that the suitable habitat models in 2000, and 1990 that were projected using presence
localities in 2010, would be more accurate than the models projected using localities
Table 1. e predictor variables used for generating the habitat suitability of NWCG in the Western
Nghe An BR.
No. Name Sources Description
1 Bio01 WorldClim Annual Mean Temperature
2 Bio03 Isothermality (BIO2/BIO7) (×100)
3 Bio07 Temperature Annual Range (BIO5-BIO6)
4 Bio12 Annual Precipitation
5 Bio13 Precipitation of Wettest Month
6 Bio15 Precipitation Seasonality (Coecient of Variation)
7 Bio16 Precipitation of Wettest Quarter
8 Elevation Earthexplorer.usgs.gov Height above sea level
9 Slope Degree of rise/run
10 Aspect Direction a slope face
11 LULC ALOS Research and Application
Project
Land Use and Land Cover
12 Distance to settlement Distance to residential land
13 Distance to Agriculture Distance to agricultural land

Khoa Van Phung et al. / Nature Conservation 51: 167–188 (2023)
174
in 2020. erefore, we used the models projected from localities in 2010 to analyze
the gibbon’s suitable habitat in 2000 and 1990 instead of the projected models from
localities in 2020. However, the results of models in 1990, and 2000 projected from
localities in 2020, were also presented in the Suppl. material 1: SI–III, and IV. For the
environmental variables, we kept topographic variables as unchanged while LULC,
Distance to Settlement, and Distance to Agriculture variables were changed following
the time. We also assumed that the climate did not change much during this time by
keeping the climatic variables as unchanged due to the lack of climate data in the study
area for each study period.
e relationship between the predicted suitable habitat of species and environmen-
tal variables in the MaxEnt model was shown by response curves. We presented the
response curves of the most contributing variables to identify the main impact of the
variables on the change of suitable habitat of NWCG in the Western Nghe An BR.
e result of the MaxEnt model is presented in a logistic format ranging from 0 to
1, in which the higher values mean higher suitability. e threshold selection should
be determined according to the objectives of the study (Merow et al. 2013; Vale et
al. 2014). To maximize the area for conservation purposes, we applied the threshold
“Minimum training presence logistic threshold” to determine the suitable/unsuitable
categories. To clearly show the changes in suitable habitat, we also categorized suitable
areas into three sub-levels by dividing equally suitability values: High, Moderate, and
Low suitability (Tran et al. 2020; Sarma et al. 2021).
Results
Performance of models and importance of variables
e MaxEnt model predicted the habitat suitability of NWCG based on the avail-
able presence localities and predictor data set of LULC, climate, and topography with
mean test AUC at 0.896 ± 0.043 and 0.936 ± 0.084 for the model of 2020 and 2010,
respectively (Fig. 3). e AUC values showed the high discrimination capacity of the
model to predict the suitable habitat of NWCG in the Western Nghe An BR.
For the contribution of each environmental variable in the model of suitable habi-
tat in 2020, the top three variables were “Distance to Agriculture” (39.2%), “Tempera-
ture Annual Range – Bio7” (25.3%), and “Precipitation of Wettest Quarter – Bio16”
(23.3%). For the model of 2010, the highest contribution variables were “Distance to
Agriculture” with 55.4%, followed by “Distance to Settlement” and “Elevation” with
29.3% and 6.6%, respectively (Table SI-II). e response curves in the MaxEnt model
show the changes in habitat suitability in response to the changes in the predictors
used in the model. e response curve of the highest contribution variables of both
models (2020 and 2010) was mainly identical when the habitat suitability of NWCG
increased signicantly for the distance to agricultural land below around one km, and
rose gradually for the distance above about one km (Fig. 4).

Changes in suitable habitat for the critically endangered gibbon 175
Figure 3. e results of AUC curves in developing the suitable habitat model of NWCG in A 2020,
and B 2010. e red line showed the mean of AUC, and the blue area showed the standard deviation.
Figure 4. e response curve of Distance-to-Agriculture, the most contribution variables to predicting
the suitable habitat of NWCG by the MaxEnt model in A 2020, and B 2010.

Khoa Van Phung et al. / Nature Conservation 51: 167–188 (2023)
176
Figure 5. e suitable habitat of NWCG under present landcover condition (2020) predicted by the
MaxEnt model.
The suitable habitat of NWCG in the Western Nghe An Biosphere Reserve
in 2020
e total suitable habitat of NWCG in Nghe An BR under the current land cover con-
dition was estimated at around 4,022.42 km2 (30.95% of the overall BR area; Table2)
and mainly concentrated in three areas including the (i) the northern part of Que Phong
and Quy Chau district – close to anh Hoa province, (ii) Pu Huong NR, and (iii) in
the southern Tuong Duong, Con Cuong district – within Pu Mat NP. However, the high
habitat suitability with around 380.64 km2 only occurred in the northern part (adjacent to
anh Hoa province) and the southern part of Nghe An BR (within Pu Mat NP; Fig. 5).
Table 2. Area of suitable habitat of NWCG projected by the MaxEnt model in 1990, 2000, 2010, 2020
(unit: km2).
Habitat suitability Year
1990 2000 2010 2020
Low 3,091.51 3,221.81 1,719.21 2,488.18
Moderate 871.28 793.30 708.17 1,173.40
High 384.89 309.86 322.83 360.84
Total 4,347.68 4,324.97 2,750.21 4,022.42

Changes in suitable habitat for the critically endangered gibbon 177
Figure 6. e suitable habitat of NWCG in the Western Nghe An BR in A 1990 B 2000 C 2010, and
D2020 (Pu Mat NR was established in 1995, and upgraded to be national park in 2001; Pu Huong, and
Pu Hoat NRs were established 2003, and 2013, respectively).
Habitat suitability changes of NWCG in the Western Nghe An Biosphere
Reserve from 1990 to 2020
Using the MaxEnt model, we projected the suitable habitat of NWCG in 1990
and 2010 (Fig. 6A–C). e predicted suitable habitat of NWCG in the Western
Nghe An BR under the changes in LUCL showed a uctuating trend from 1990 to

Khoa Van Phung et al. / Nature Conservation 51: 167–188 (2023)
178
2020. In particular, the total suitable habitat gradually declined by around 22.72
km2 from 1990 to 2000. e moderately and highly suitable habitat shrank by
around 77.98km2 (8.95%) and 75.04 km2 (19.49%), respectively. Between 2000
and 2010, there was around 1,574.76 km2 contraction of suitable habitat for the
gibbons in the Western Nghe An BR. e majority of habitat lost lay within the
low to moderate suitable habitat types. e predicted suitable habitat of NWCG
increased signicantly from 2750.21 km2 to 4,022.42 km2 between 2010 and 2020.
All suitable categories of habitat showed an expanding trend in the period. e
high, moderate, and low habitat suitability categories were estimated to increase by
around 38.01 km2 (11.77%), 456.23km2 (65.69%), and 768.97 km2 (44.73%),
respectively (Table 2).
e extracted LULC in the Western Nghe An BR from 1990 to 2020 consti-
tuted seven main types, including Settlement, Agriculture, Barren land, Broadleaf
forest, Plantation, Water, and Bamboo forest (Table 3). e Broadleaf Forest type
occupied the largest area and showed dierent trends between 1990 and 2020:
6,366.36km2 (46.01%), 7,145.05 km2 (51.64%), 6902.46 km2 (49.89%), and
7695.72 km2 (55.62%) in 1990, 2000, 2010, and 2020, respectively (Table 3).
On the other hand, the Settlement area was the smallest category but presented
an increase from around 0.56 km2 (0.004%) in 1990 to 34.75 km2 (0.25%) in
2020(Table 3).
Table 4. Area of suitable habitat inside and outside borders of protected areas (Pu Mat NP, Pu Huong
NR, and Pu Hoat NR*) of NWCG projected by the MaxEnt model in 1990, 2000, 2010, 2020.
Suitability 1990 2000 2010 2020
Inside Outside Inside Outside Inside Outside Inside Outside
Low 0 3,091.51 382.32 2,839.49 437.10 1,282.11 723.43 1,764.75
Moderate 0 871.28 348.16 445.15 444.19 263.99 613.29 560.11
High 0 384.89 137.57 172.28 228.84 93.99 236.99 123.85
Total 0 4,347.68 868.05 3456.91 1,110.13 1,640.08 1,573.71 2,448.70
*: Pu Mat NR was established in 1995, and upgraded to national park in 2001; Pu Huong, and Pu Hoat NR were
established 2003, and 2013 respectively.
Table 3. Area and change of the LULC categories in the Western Nghe An BR from 1990 to 2020 ex-
tracted from Phan et al. (2021) (unit=km2).
No LULC
categories
Year Change
1990 2000 2010 2020 1990–2000 2000–2010 2010–2020
1 Settlement 0.56 1.86 13.79 34.75 1.30 11.93 20.97
2 Agriculture 1,789.05 2,008.77 1,835.63 1,707.34 219.72 -173.14 -128.29
3 Barren land 791.70 869.97 899.11 549.95 78.27 29.13 -349.16
4 Broadleaf Forest 6,366.36 7,145.05 6,902.46 7,695.72 778.69 -242.59 793.26
5 Plantation 3,930.07 2,681.34 2,733.05 2,275.61 -1,248.73 51.71 -457.43
6 Water 68.27 83.86 69.91 125.89 15.59 -13.94 55.96
7 Bamboo forest 889.56 1,044.72 1,381.62 1,446.31 155.16 336.90 64.68

Changes in suitable habitat for the critically endangered gibbon 179
Habitat suitability changes of NWCG within existing protected areas inside
the Western Nghe An BR.
e overlay between the existing protected areas within the Western Nghe An BR, in-
cluding Pu Mat NP, Pu Huong, and Pu Hoat NRs, and the suitable habitat of NWCG
predicted by the MaxEnt model was only approximately 868.05 km2 (25.11%) of suit-
able habitat lies within protected areas (PAs) in 2000. In 1990, three protected areas in
the core zone of the BR were not yet established. In 2010, the areas of suitable habitat
inside PAs increased to 1,110.13 km2. en, the predicted suitable habitat rose to ap-
proximately 1,573.71 km2 in 2020 (Table 4).
Discussion
e AUC value of both models in 2010 and 2020 shows a high capacity to predict
the suitable habitat of NWCG with a good performance. e MaxEnt model has been
broadly employed to predict the potential distribution of primates (Sarma et al. 2015,
2021; Tran and Vu 2020; Widyastuti et al. 2020; Yang et al. 2021), and also to assess
the changes in the suitable habitats of several species under changes in land cover or
climate (Tran et al. 2020; Vu et al. 2020; Sarma et al. 2021; Trinh et al. 2021; Blair
et al. 2022; Ngo et al. 2022). e predicted suitable habitat of the 2020 model of
NWCG in the Western Nghe An BR covered all of the known presence points within
Pu Hoat, Pu Huong NR, and Pu Mat NP and adjacent areas (Fig. 5). e predicted
suitable habitat of the 2010 model also tted with the known occurrence of the species
in Luu and Rawson (2009, 2010, 2011), and Rawson et al. (2011). Our result indi-
cates that the model could reliably predict the change of suitable habitat over the years
1990–2020 of NWCG in the Western Nghe An BR.
In the present study, we tried to predict the suitable habitat changes of NWCG
in the Western Nghe An BR. Modeling a partial distribution could not provide a
complete species range because a part of the environmental conditions and presence
records of species outside of the target area of the model are not included (Carretero
and Sillero 2016; Sillero et al. 2021). Models for the entire distribution of a species can
provide a better result than models of partial distribution (Barbet‐Massin et al. 2010;
Carretero and Sillero 2016). However, it also should be noted that a model on a partial
range of a species can forecast other distribution constraints on this species, which may
uctuate in distinctive parts in the whole range (Martínez‐Freiría et al. 2008; Vale et
al. 2015; Sillero et al. 2021). Our study focused on a small part of the species’ overall
distribution range that spans Vietnam, northern Lao and southern China (Rawson et
al. 2011). us, the results could reveal the distinctiveness of environmental variables
in the Western Nghe An BR, resulting in changes in suitable habitat of the Western
Nghe An BR under changes in LULC between 1990 and 2020. e results also pro-
vided valuable insights that could assist in predicting changes in habitat suitability in
the entire distribution range of the species.

Khoa Van Phung et al. / Nature Conservation 51: 167–188 (2023)
180
Our models indicated that the “Distance to Agriculture” was the most signicant
variable for both predicted models 2020, and 2010 of NWCG, with the highest contri-
butions to the models. e habitat suitability was likely to increase when the distance
to agricultural land value was high (Fig. 3), which was also shown in a model of the
Northern yellow-cheeked gibbon by Tran and Vu (2020). Gibbons are very sensitive
to human disturbances (Geissmann et al. 2000) and habitats near agricultural land can
be more easily accessed. erefore, these areas are considerably impacted by anthro-
pogenic activities such as illegal logging or hunting. Previous studies also emphasized
that habitat loss is an important issue and has caused a dramatic reduction of gibbon
populations throughout their distribution (Rawson et al. 2011; Rawson et al. 2020).
e suitable habitat generated by species distribution modeling could be very use-
ful in suggesting areas to nd new populations, especially for rare and threatened spe-
cies (Pearson et al. 2007; orn et al. 2009; van Schingen et al. 2016; Ngo et al. 2019,
2022). Our model for the present suitable habitat of NWCG in the Western Nghe An
BR showed a concentration in three distinctive areas, including the northern part of
Que Phong and Quy Chau districts, Pu Huong NR, and the southern Tuong Duong,
Con Cuong districts – within Pu Mat NP (Fig. 4). Our result also revealed that the pre-
dicted suitable habitat area lies outside the existing protected areas at around 2,448.70
km2. is area is primarily located in the northern region of Pu Hoat NR, near anh
Hoa province, and in the northwest and southeast areas of Pu Mat NP, Nghe An prov-
ince (Fig. 5). Rawson et al. (2011) indicated that the majority of NWCG populations
in Vietnam were detected within existing protected areas, while a large forest area out-
side the established protected area system has not been surveyed. Recently, researchers
also found at least 40 groups of NWCG, mainly inhabiting an area to the north of Pu
Hoat NR, outside the current protected systems (Pu Hoat NR 2021). us, we believe
that a large population of NWCG outside protected areas probably have not been de-
tected yet. We recommend that more survey eorts on gibbons should be spent on the
higher suitable habitat areas outside protected areas that were predicted by our model,
besides the priority for gibbon surveys within Pu Hoat NR, Pu Huong NR, and Pu
Mat NP. Our suitable habitat results given by the MaxEnt model can allow managers,
conservationists, and researchers to easily plan eld surveys for exploring the unre-
corded population of NWCG in the Western Nghe An BR with greater condence.
Our current study indicates that LULC changes may signicantly impact the distri-
bution of gibbons in the Western Nghe An BR, central Vietnam, as projected from 1990
to 2020. Knowledge of habitat transformation plays a crucial role in making decisions
regarding biodiversity conservation (Fletcher et al. 2018; Su et al. 2021). Additionally,
monitoring gibbons and their habitat is extremely important for conservation due to
their shrinking populations globally (Sarma et al. 2021). e Western Nghe An BR was
assessed as the most important site for the conservation of NWCG (Rawson et al. 2011),
while Pu Mat NP might hold the largest population of the species, with an estimated
population of 130 groups through gibbon surveys in 2010 (Luu and Rawson 2011).
Our models showed that the habitat suitability for NWCG slightly declined from 1990
to 2000, and greatly decreased between 2000 and 2010. On the other hand, from 2010
to 2020, the predicted suitable habitat increased by around 1272.20 km2. It showed that

Changes in suitable habitat for the critically endangered gibbon 181
the suitable habitat area of the NWCG strongly depended on the change of LULC. e
main LULC types of the Western Nghe An BR constituted Agriculture land, Broadleaf
forest, and Plantation, while changes in LULC in the area showed complex trends con-
sistent with a case study in the highland area of Nghe An Province (Leisz 2009).
e Distance to Settlement variable was the second most important variable for the
habitat suitability model of NWCG, although the Settlement was the smallest area among
seven main LULC types in the Western Nghe An BR, probably indicating the high impact
of human residence settlement areas to habitat suitability of NWCG. e settlement area
in the transition zone of the Western Nghe An BR continually increased from 1990 to
2020, probably due to urbanization and population expansion. Between 1990 and 2000,
the agricultural land area increased, leading to a slight decrease in the suitable habitat of
NWCG. e broadleaf forest areas signicantly decreased between 2000 and 2010, re-
ecting the ineective activities for habitat conservation in the area. In contrast, there was
an increase in the broadleaf forest from 2010 to 2020 that was probably facilitated by
government policies, notably the closing of natural forest policy in the late 2010s, and the
National Action Programme on REDD+ in the period 2011–2020 (Government of Viet-
nam 2012). Gibbons are known to be sensitive to human disturbance and to prefer pristine
broadleaf forests. Based on our model results on habitat suitability, we strongly recommend
protecting the suitable habitat for NWCG and increasing the broadleaf forest areas.
Although the Western Nghe An BR has been established since 2007 with only
three core protected areas, including Pu Mat NP, Pu Huong, and Pu Hoat NR, most
areas of the Biosphere Reserve have remained currently intact even with the increas-
ing settlement, agricultural land, and the uctuation of broadleaf forest areas. As a
positive human policy, establishing protected areas for protecting all environmental
components, including populations, habitats of wild species and natural ecosystems,
is crucial because of its mitigations of negative impacts on biodiversity at a local scale
(Margules and Pressey 2000; Estrada and Real 2018). rough the eorts of the Viet-
namese government, Pu Mat NR was established in 1995 and upgraded to the national
park level in 2001; Pu Huong and Pu Hoat NR were established in 2003 and 2013,
respectively. e establishment is probably the major reason for the increase of suitable
habitats for NWCG inside protected areas. In the general planning for biodiversity
conservation of the country to 2020, orientation to 2030 in Degree 45/QĐ-TTg/2014
(Government of Vietnam 2014), the government also proposed three biodiversity cor-
ridors inside the Western Nghe An BR to support the movement of wildlife species
under the impact of climate change. However, connectivity habitat for gibbons among
the three core protected areas (Pu Mat NP, Pu Luong, and Pu Hoat NR) is very poor
when the suitable habitat within these protected areas seems to be isolated. We suggest
that areas for connecting three core protected areas inside the Western Nghe An BR
also should receive high priority in habitat monitoring of the gibbon.
In this study, land-use change has a strong relationship with the suitable habitat of
NWCG, which resulted in the contraction of the suitable habitat of the endangered
gibbon from 2000 to 2010. Interestingly, an expansion of suitable habitat of NWCG
was revealed in the period 2010 – 2020, which provides a positive sign for habitat con-
servation in the study areas. Additionally, a limitation of the study is the lack of infor-

Khoa Van Phung et al. / Nature Conservation 51: 167–188 (2023)
182
mation on specic habitat parameters that might be critical for dening the selection
habitat of gibbons, such as the height of tree canopy, food preference, and availability.
We recommended that more intensive surveys should be focused on the suitable habi-
tat areas suggested by our model to reveal the forest quality of these areas.
Vietnam is located in the Mekong region, which was assessed as one of the 11 areas
experiencing the largest forest loss. e average rate of deforestation in the Mekong
region was approximately 0.4% per year (Leinenkugel et al. 2015). Researchers also pre-
dicted that the Mekong area will have lost about 15–30 million hectares of natural forest
by 2030 if there are no eective urgent measures from governments (WWF 2018). In
the best scenario, countries have strong reforms toward green and clean development,
and the natural forest of countries in Southeast Asian countries can increase by around
19.6 million hectares (Estoque et al. 2019). However, we have not yet had any ne-scale
forecasts on forest status scenarios for the region in the future. With an increasing trend
in forest areas, and the eorts and commitments of the Vietnamese government in recent
times (De Jong et al. 2006; Government of Vietnam 2017), we predict that the natural
forest area in the Western Nghe An BR will continually show a slight increasing trend in
the next few decades, leading to an increase in the suitable habitat of gibbons in the study
area. But it also should be noted that the increased forest area comes mainly in the form
of young forests, which cannot meet the habitat needs of gibbons. Gibbons are special-
ized in arboreal life and require good-quality forests with a high canopy closure and large
trees for their survival (Geissmann et al. 2000; Nadler and Brockman 2014). erefore,
we strongly recommend that activities to enrich young forests are also necessary for the
conservation activities of NWCG and its habitats in the Western Nghe An BR.
Finally, we highly recommend the implementation of the following actions for pro-
tecting existing populations of NWCG and its habitat in the Western Nghe An BR: 1)
implement population surveys, monitoring and conservation awareness programmes not
only in three core protected areas (Pu Mat NP, Pu Hoat, and Pu Huong NR) but also in
other highly suitable habitats that are predicted by our model, especially for the northern
Pu Hoat NR, northern and southern Pu Mat NP; 2) upgrade the Pu Hoat watershed
protection forest, the highly suitable habitat in the north of Pu Hoat NR, to be a Nature
Reserve (Fig. 4). 3) improve green corridors connecting three core protected areas inside
the Western Nghe An BR and other highly suitable habitats; 4) predict the suitable habi-
tat of the NWCG in its whole distribution range, including northern Vietnam, Laos, and
southern China, then, propose transboundary conservation programs, especially between
Vietnam and Laos; 5) conduct surveys to obtain specic habitat parameters at the suit-
able habitat area and conrmed the presence of gibbon, then implement the enrichment
activities in young forest areas that do not meet the requirements of gibbons for living.
Conclusion
ere is currently a lack of research on the changes in suitable habitats for gibbons,
leading to limited eorts to protect and conserve their habitats. Here, we applied the
species distribution modeling to predict the suitable habitat of NWCG in the Western

Changes in suitable habitat for the critically endangered gibbon 183
Nghe An BR, and assessed the impact of changes in land use and land cover from
1990 to 2020 on the suitable habitat of this species. e current suitable habitat for
the gibbon in the Western Nghe An BR was estimated at approximately 4,022.42 km2,
mainly concentrated in three distinctive areas (Pu Mat NP, Pu Huong, and Pu Hoat
NR), while the predicted suitable habitat area lies outside the existing protected areas
at around 2,448.70 km2. Our result also indicated that due to the changes in LULC,
the predicted suitable habitat of the gibbon decreased from 1990 to 2010 but gradually
increased from 2010 to 2020. Based on our ndings, we suggested that more survey
eorts should be focused on areas of predicted suitable habitat regions outside exist-
ing protected areas, in addition to continuing eorts for suitable habitat areas within
protected areas. We also highly recommend enriching young forests and protecting
the preferred habitat for the gibbon, specically broadleaf forests. Finally, we proposed
several actions to safeguard NWCG and its habitat in the Western Nghe An BR.
Acknowledgements
We would like to thank the “Mainstreaming Natural resource management and Biodi-
versity conservation objectives into socio-economic development planning and man-
agement of biosphere reserve in Viet Nam” project (BR project) for their supports.
We are grateful to Dr. Greg Nagle for proofreading this manuscript. We wish to thank
Management Broad of Western Nghe An BR, Pu Mat NP, Pu Hoat, and Pu Huong
NR, and our eld assistants for their support during our eld surveys. Many thanks to
anonymous reviewers for improving a previous version of the manuscript.
References
Aiello-Lammens ME, Boria RA, Radosavljevic A, Vilela B, Anderson RP (2015) spin: An R
package for spatial thinning of species occurrence records for use in ecological niche mod-
els. Ecography 38(5): 541–545. https://doi.org/10.1111/ecog.01132
Barbet‐Massin M, uiller W, Jiguet F (2010) How much do we overestimate future local ex-
tinction rates when restricting the range of occurrence data in climate suitability models?
Ecography 33(5): 878–886. https://doi.org/10.1111/j.1600-0587.2010.06181.x
Blair ME, Nguyen TA, Le MD, Liu Z, Meng T, Horning N, Sterling EJ, ach HM, Xu M,
Galante PJ (2022) Karst as an abiotic driver of François’ langur distribution, with predic-
tions for biological communities on karst under climate change. Frontiers of Biogeography
0(0): 0–13. https://doi.org/10.21425/F5FBG51838
Brockelman WY, Nathalang A, Greenberg DB, Suwanvecho U (2014) Evolution of Small-
Group Territoriality in Gibbons. Primates and Cetaceans. Springer, Tokyo, 213–230.
https://doi.org/10.1007/978-4-431-54523-1_11
Carretero MA, Sillero N (2016) Evaluating how species niche modelling is aected by par-
tial distributions with an empirical case. Acta Oecologica 77: 207–216. https://doi.
org/10.1016/j.actao.2016.08.014

Khoa Van Phung et al. / Nature Conservation 51: 167–188 (2023)
184
De Jong W, Do DS, Hung TV (2006) Forest Rehabilitation in Vietnam: Histories, Realities,
and Future: Histories, Realities, and Future. Center for International Forestry Research
(CIFOR), Bogor, Indonesia, 1–90.
Díaz S, Settele J, Brondízio ES, Ngo HT, Agard J, Arneth A, Balvanera P, Brauman KA, Butchart
SHM, Chan KMA, Garibaldi LA, Ichii K, Liu J, Subramanian SM, Midgley GF, Miloslav-
ich P, Molnár Z, Obura D, Pfa A, Polasky S, Purvis A, Razzaque J, Reyers B, Chowdhury
RR, Shin Y-J, Visseren-Hamakers I, Willis KJ, Zayas CN (2019) Pervasive human-driven
decline of life on Earth points to the need for transformative change. Science 366(6471):
eaax3100. https://doi.org/10.1126/science.aax3100
Elith J (2000) Quantitative Methods for Modeling Species Habitat: Comparative Performance
and an Application to Australian Plants. Quantitative methods for conservation biology.
Springer, 39–58. https://doi.org/10.1007/0-387-22648-6_4
Elith JH, Graham CP, Anderson R, Dudík M, Ferrier S, Guisan AJ, Hijmans R, Huettmann
FR, Leathwick J, Lehmann A (2006) Novel methods improve prediction of species’ dis-
tributions from occurrence data. Ecography 29(2): 129–151. https://doi.org/10.1111/
j.2006.0906-7590.04596.x
Estoque RC, Ooba M, Avitabile V, Hijioka Y, DasGupta R, Togawa T, Murayama Y (2019)
e future of Southeast Asia’s forests. Nature Communications 10(1): 1–12. https://doi.
org/10.1038/s41467-019-09646-4
Estrada A, Real R (2018) Assessment of the National Park network of mainland Spain by
the Insecurity Index of vertebrate species. PLoS ONE 13(5): e0197496. https://doi.
org/10.1371/journal.pone.0197496
Fan PF, Fei HL, Luo AD (2014) Ecological extinction of the critically endangered northern
white-cheeked gibbon Nomascus leucogenys in China. Oryx 48(1): 52–55. https://doi.
org/10.1017/S0030605312001305
Fick SE, Hijmans RJ (2017) WorldClim 2: New 1‐km spatial resolution climate surfaces for
global land areas. International Journal of Climatology 37(12): 4302–4315. https://doi.
org/10.1002/joc.5086
Fletcher Jr RJ, Didham RK, Banks-Leite C, Barlow J, Ewers RM, Rosindell J, Holt RD,
Gonzalez A, Pardini R, Damschen EI, Melo FPL, Ries L, Prevedello JA, Tscharntke T,
Laurance WF, Lovejoy T, Haddad NM (2018) Is habitat fragmentation good for biodiver-
sity? Biological Conservation 226: 9–15. https://doi.org/10.1016/j.biocon.2018.07.022
Geissmann T, Nguyen XD, Lormée N, Momberg F (2000) Vietnam Primate Conservation Sta-
tus Review 2000, Part 1: Gibbons. Fauna and Flora International, Indochina Programme,
Hanoi, 2000.
Government of Vietnam (2012) Decision 799/QD-TTg dated June 27, 2012 on approval of
the “National Action Programme on REDD+ in period 2011-2020”.
Government of Vietnam (2014) Decision 45/QD-TTg dated January 8, 2014 on approval
the general planning for biodiversity conservation of the country to 2020, orientation
to 2030.
Government of Vietnam (2017) Decision 419/QD-TTg dated May 4, 2017 on Approval of
the National Action Programme on the Reduction of Greenhouse Gas Emissions through
the reduction of Deforestation and Forest Degradation, Sustainable Management of Forest
Resources, and Conservation and Enhancement of Forest Carbon.

Changes in suitable habitat for the critically endangered gibbon 185
Ha TL, Tam NA, Minh HT, Tinh NT, Tuan BV (2011) Survey of the Northern Bu-Cheeked
Crested Gibbon (Nomascus annamensis) in Kon Ka Kinh National Park, Gia Lai Province,
Vietnam. Fauna and Flora International/Conservation International, Hanoi, Vietnam.
Hai B, Chen J, Tiwari A, Duc H (2020) Ranging behavior of the southern yellow-cheeked
gibbon (Nomascus gabriellae) in response to food resources and environmental variables.
Primatology 3(2): 1–22.
Hoang MD, Tran VB, Vu L (2010) Population Status of the Yellow-Cheeked Crested Gibbon
(Nomascus gabriellae) in Ta Dung Nature Reserve, Dak Nong Province, Vietnam. Fauna &
Flora International and Conservation International, Hanoi, Vietnam.
Jetz W, Wilcove DS, Dobson AP (2007) Projected impacts of climate and land-use change on
the global diversity of birds. PLoS Biology 5(6): e157. https://doi.org/10.1371/journal.
pbio.0050157
Leinenkugel P, Wolters ML, Oppelt N, Kuenzer C (2015) Tree cover and forest cover dynamics
in the Mekong Basin from 2001 to 2011. Remote Sensing of Environment 158: 376–392.
https://doi.org/10.1016/j.rse.2014.10.021
Leisz SJ (2009) Dynamics of land cover and land use changes in the upper Ca river basin of
Nghe An, Vietnam. Japanese Journal of Southeast Asian Studies 47(3): 287–308.
Luu TB, Rawson B (2009) An Assessment of Northern White-Cheeked Crested Gibbon (No-
mascus leucogenys) Population Status in Pu Huong Nature Reserve, Nghe An Province,
Vietnam. Fauna & Flora International and Conservation International, Hanoi, Vietnam.
Luu TB, Rawson B (2010) An Assessment of the Status of the Northern White-Cheeked
Crested Gibbon (Nomascus leucogenys) in Pu Hoat Proposed Nature Reserve, Que Phong
District, Nghe An Province, Vietnam. Fauna & Flora International and Conservation In-
ternational, Hanoi, Vietnam.
Luu TB, Rawson B (2011) Population Assessment of the Northern White-Cheeked Crested
Gibbon (Nomascus leucogenys) in Pu Mat National Park, Nghe An Province. Fauna & Flora
International and Conservation International, Hanoi, Vietnam.
Margules CR, Pressey RL (2000) Systematic conservation planning. Nature 405(6783):
243–253. https://doi.org/10.1038/35012251
Martínez‐Freiría F, Sillero N, Lizana M, Brito JC (2008) GIS‐based niche models identify environ-
mental correlates sustaining a contact zone between three species of European vipers. Diver-
sity & Distributions 14(3): 452–461. https://doi.org/10.1111/j.1472-4642.2007.00446.x
Merow C, Smith MJ, Silander Jr JA (2013) A practical guide to MaxEnt for modeling spe-
cies’ distributions: What it does, and why inputs and settings matter. Ecography 36(10):
1058–1069. https://doi.org/10.1111/j.1600-0587.2013.07872.x
Mitani JC (1987) Territoriality and monogamy among agile gibbons (Hylobates agilis). Be-
havioral Ecology and Sociobiology 20(4): 265–269. https://doi.org/10.1007/BF00292179
Myers N, Mittermeier RA, Mittermeier CG, Da Fonseca GAB, Kent J (2000) Biodiversity hotspots
for conservation priorities. Nature 403(6772): 853–858. https://doi.org/10.1038/35002501
Nadler T, Brockman D (2014) Primates of Vietnam. Endangered Primate Rescue Center, Cuc
Phuong, Ninh Binh, Vietnam, 376 pp.
Nazeri M, Juso K, Madani N, Mahmud AR, Bahman AR, Kumar L (2012) Predictive mode-
ling and mapping of Malayan sun bear (Helarctos malayanus) distribution using maximum
entropy. PLoS ONE 7(10): e48104. https://doi.org/10.1371/journal.pone.0048104

Khoa Van Phung et al. / Nature Conservation 51: 167–188 (2023)
186
Western Nghe An Biosphere Reserve (2017) Western Nghe An Biosphere Reserve. http://sin-
hquyennghean.vn/khu-du-tru-sinh-quyen-1-8-2707.html
Ngo HN, Le QT, Pham LM, Le DM, van Schingen M, Ziegler T (2019) First record of the Cat
Ba tiger gecko, Goniurosaurus catbaensis, from Ha Long Bay, Quang Ninh Province, Viet-
nam: Microhabitat selection, potential distribution, and threat evaluation. Amphibian &
Reptile Conservation 13(2): 1–13. https://doi.org/10.3897/natureconservation.32.33590
Ngo HN, Nguyen HQ, Phan TQ, Nguyen TQ, Gewiss LR, Rödder D, Ziegler T (2022)
Modeling the environmental refugia of the endangered Lichtenfelder’s Tiger Gecko
(Goniurosaurus lichtenfelderi) towards implementation of transboundary conservation.
Frontiers of Biogeography 14(1): e51167. [1–14] https://doi.org/10.21425/F5FBG51167
Pearson RG, Raxworthy CJ, Nakamura M, Peterson AT (2007) Predicting species distributions
from small numbers of occurrence records: A test case using cryptic geckos in Madagascar. Jour-
nal of Biogeography 34(1): 102–117. https://doi.org/10.1111/j.1365-2699.2006.01594.x
Peterson AT, Soberón J, Pearson RG, Anderson RP, Martínez-Meyer E, Nakamura M, Araújo
MB (2011) Ecological niches and geographic distributions (MPB-49). Ecological Nich-
es and Geographic Distributions. Princeton University Press. https://doi.org/10.23943/
princeton/9780691136868.003.0003
Phan DC, Trung TH, Truong VT, Sasagawa T, Vu TPT, Bui DT, Hayashi M, Tadono T, Nasa-
hara KN (2021) First comprehensive quantication of annual land use/cover from 1990 to
2020 across mainland Vietnam. Scientic Reports 11(1): 1–20. https://doi.org/10.1038/
s41598-021-89034-5
Phillips SJ, Anderson RP, Schapire RE (2006) Maximum entropy modeling of species geograph-
ic distributions. Ecological Modelling 190(3–4): 231–259. https://doi.org/10.1016/j.ecol-
model.2005.03.026
Phung VK, Dong TH (2021) Monitoring the Northern white-cheeked gibbon in the Western
Nghe An Biosphere Reserve. Technical report in the project of Mainstreaming Natural
resource management and Biodiversity conservation objectives into socio-economic devel-
opment planning and management of biosphere reserve in Viet Nam” project (BR project),
Vietnam. [In Vietnamese]
Pu Hoat NR (2020) Study on the mammal fauna in Pu Hoat Nature Reserve. e technical
report in Pu Hoat NR, Nghe An, Vietnam. [in Vietnamese]
Pu Hoat NR (2021) Công tác bảo tồn loài Vượn đen má trắng tại Khu BTTN Pù Hoạt, tỉnh
Nghệ An/Conservation activities the Northern white-cheeked gibbon in Pu Hoat NR.
http://puhoat.vn/news/27086p24c32/cong-tac-bao-ton-loai-vuon-den-ma-trang-tai-khu-
bttn-pu-hoat-tinh-nghe-an.htm
Rawson BM, Insua-Cao P, Nguyen MH, Van NT, Hoang MD, Mahood S, Geissmann T, Roos
C (2011) e conservation status of gibbons in Vietnam. Fauna & Flora International and
Conservation International.
Rawson BM, Nguyen MH, Coudrat CNZ, Roos C, Jiang X, Duckworth JW (2020) Nomas-
cus leucogenys (errata version published in 2020). e IUCN Red List of reatened Spe-
cies 2020: E.T39895A180816530. https://doi.org/10.2305/IUCN.UK.2020-2.RLTS.
T39895A180816530.en
Raxworthy CJ, Ingram CM, Rabibisoa N, Pearson RG (2007) Applications of ecological
niche modeling for species delimitation: A review and empirical evaluation using day

Changes in suitable habitat for the critically endangered gibbon 187
geckos (Phelsuma) from Madagascar. Systematic Biology 56(6): 907–923. https://doi.
org/10.1080/10635150701775111
Ruppell JC (2013) Ecology of White-Cheeked Crested Gibbons in Laos. PhD esis, Portland
State University, USA.
Sarma K, Kumar A, Krishna M, Medhi M, Tripathi OP (2015) Predicting Suitable Habitats for
the Vulnerable Eastern Hoolock Gibbon, Hoolock leuconedys, in India Using the MaxEnt
Model. Folia Primatologica 86(4): 387–397. https://doi.org/10.1159/000381952
Sarma K, Saikia MK, Sarania B, Basumatary H, Baruah SS, Saikia BP, Kumar A, Saikia PK
(2021) Habitat monitoring and conservation prioritization of Western Hoolock Gibbon
in upper Brahmaputra Valley, Assam, India. Scientic Reports 11(1): 1–11. https://doi.
org/10.1038/s41598-021-94844-8
Sillero N, Arenas-Castro S, Enriquez‐Urzelai U, Vale CG, Sousa-Guedes D, Martínez-Freiría
F, Real R, Barbosa AM (2021) Want to model a species niche? A step-by-step guideline on
correlative ecological niche modelling. Ecological Modelling 456: e109671. https://doi.
org/10.1016/j.ecolmodel.2021.109671
Su X, Shen Y, Zhou W, Liu Y, Cheng H, Yang M, Zhou S, Zhao J, Wan L, Liu G (2021) Land-use
changes conservation network of an endangered primate (Rhinopithecus bieti) in the past
30 years in China. Diversity & Distributions 00: 1–14. https://doi.org/10.1111/ddi.13446
inh VN, Mootnick AR, Geissmann T, Li M, Ziegler T, Agil M, Moisson P, Nadler T, Walter
L, Roos C (2010) Mitochondrial evidence for multiple radiations in the evolutionary his-
tory of small apes. BMC Evolutionary Biology 10(1): 1–13. https://doi.org/10.1186/1471-
2148-10-74
orn JS, Nijman V, Smith D, Nekaris KAI (2009) Ecological niche modelling as a technique
for assessing threats and setting conservation priorities for Asian slow lorises (Primates:
Nycticebus). Diversity & Distributions 15(2): 289–298. https://doi.org/10.1111/j.1472-
4642.2008.00535.x
Tran DV, Vu TT (2020) Combining species distribution modeling and distance sampling
to assess wildlife population size: A case study with the northern yellow‐cheeked gibbon
(Nomascus annamensis). American Journal of Primatology 82(9): e23169. https://doi.
org/10.1002/ajp.23169
Tran DV, Vu TT, Tran BQ, Nguyen MD, Vu PT, Tran TH, Nguyen HT, Pham TV, Nguyen
TC (2020) Modelling the change in the distribution of the black-shanked douc, Pygathrix
nigripes (Milne-edwards) in the context of climate change: Implications for conservation.
e Raes Bulletin of Zoology 68(October): 769–778. https://doi.org/10.26107/RBZ-
2020-0088
Tran DV, Le TT, Vu TT, Nguyen DD, Khotpathoom T, Phung VK, Tran DH, Le VN, Le CT
(2023) A review on the status and modeling of suitable habitats of the southern white-
cheeked gibbon. Primates: 1–11. https://doi.org/10.1007/s10329-022-01047-4
Trinh DH, Nguyen AT, Le MD, Li X, Cao NTH, Blair ME (2021) Assessment of climate
change impacts on one of the rarest apes on Earth, the Cao Vit Gibbon Nomascus nasutus.
Frontiers of Biogeography 0(0). https://doi.org/10.21425/F5FBG53320
Vale CG, Tarroso P, Brito JC (2014) Predicting species distribution at range margins: Testing
the eects of study area extent, resolution and threshold selection in the Sahara–Sahel tran-
sition zone. Diversity & Distributions 20(1): 20–33. https://doi.org/10.1111/ddi.12115

Khoa Van Phung et al. / Nature Conservation 51: 167–188 (2023)
188
Vale CG, da Silva MJF, Campos JC, Torres J, Brito JC (2015) Applying species distribution
modelling to the conservation of an ecologically plastic species (Papio papio) across biogeo-
graphic regions in West Africa. Journal for Nature Conservation 27: 26–36. https://doi.
org/10.1016/j.jnc.2015.06.004
van Schingen M, Le MD, Ngo HT, Pham CT, Ha QQ, Nguyen TQ, Ziegler T (2016) Is there
more than one Crocodile Lizard? An integrative taxonomic approach reveals Vietnamese
and Chinese Shinisaurus crocodilurus represent separate conservation and taxonomic units.
Der Zoologische Garten 85(5): 240–260. https://doi.org/10.1016/j.zoolgart.2016.06.001
Vu TT, Tran DV, Tran HTP, Nguyen MD, Do TA, Ta NT, Cao HT, Pham NT, Phan DV
(2020) An assessment of the impact of climate change on the distribution of the grey-
shanked douc Pygathrix cinerea using an ecological niche model. Primates 61(2): 267–275.
https://doi.org/10.1007/s10329-019-00763-8
Warren DL, Glor RE, Turelli M (2010) ENMTools: A toolbox for comparative studies of
environmental niche models. Ecography 33(3): 607–611. https://doi.org/10.1111/j.1600-
0587.2009.06142.x
Widyastuti S, Perwitasari-Farajallah D, Prasetyo LB, Iskandar E, Setiawan A (2020) Maxent
modelling of habitat suitability for the endangered javan gibbon (Hylobates moloch) in
less-protected Dieng Mountains, Central Java. IOP Conference Series. Earth and Envi-
ronmental Science 457(1): e012014. https://doi.org/10.1088/1755-1315/457/1/012014
WWF (2018) With less than 30% of Greater Mekong’s forests remaining, new WWF report
highlights eorts by people and communities trying to stop the devastation. https://asiapa-
cic.panda.org/our_work/greater_mekong/
Yang L, Shi KC, Ma C, Ren GP, Fan PF (2021) Mechanisms underlying altitudinal and hori-
zontal range contraction: e western black crested gibbon. Journal of Biogeography
48(2): 321–331. https://doi.org/10.1111/jbi.13998
Supplementary material 1
Supporting information
Authors: Khoa Van Phung, Dung Van Tran, Hai anh Dong, Vinh Quang Luu, Van
Bac Bui, inh Tien Vu
Data type: gures and tables
Explanation note: SI–I. Map of each variable that was used for the nal model. SI–II.
e percent contribution of environmental variables. SI–III. e suitable habitat of
N.leucogenys were predicted from occurrence localities in 2020. SI–IV. Area of suitable
habitat of N. leucogenys projected by MaxEnt model in 1990, 2000, 2010, 2020 from
the model of 2020 (unit: km2).
Copyright notice: is dataset is made available under the Open Database License
(http://opendatacommons.org/licenses/odbl/1.0/). e Open Database License
(ODbL) is a license agreement intended to allow users to freely share, modify, and
use this Dataset while maintaining this same freedom for others, provided that the
original source and author(s) are credited.
Link: https://doi.org/10.3897/natureconservation.51.90373.suppl1