pure.southwales.ac.uk€¦ · Web view1 Lee et al.: Habitat utilisation by sun bears *...

1 Lee et al.: Habitat utilisation by sun bears

Understanding landscape and plot-scale habitat utilisation by Malayan sun bear (Helarctos

malayanus) in degraded lowland forest

David C. Lee a, b, *, Victoria J. Powell c, Jeremy A. Lindsell b, d

a School of Applied Sciences, University of South Wales, Pontypridd CF37 4BD, UK.

b RSPB Centre for Conservation Science, Royal Society for the Protection of Birds, Sandy SG19

2DL, UK.

c Harapan Rainforest, Jambi, Sumatra, Indonesia.

d A Rocha International, David Attenborough Building, Pembroke Street, Cambridge CB2 3QZ,

UK.

* Corresponding author. E-mail address: [email protected] (D. Lee).

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24


ABSTRACT

Malayan sun bear (Helarctos malayanus) is a forest-dependent species globally threatened by

loss of suitable habitat and hunting. Understanding how sun bears utilise habitat in more

degraded landscapes is increasingly important for the effective conservation of the species. We

studied how landscape and plot attributes affect sun bear habitat use along a gradient of logging

disturbance in a lowland forest site of Sumatra. We conducted surveys of bear claw marks to

indicate sun bear habitat use at plot and landscape scales, and inform forest restoration strategies

that benefit the conservation management of the species. We recorded 12 habitat features and the

presence/absence of claw marks in 262 plots in four different habitat types. We reduced the

number of habitat variables using Principal Components Analysis (PCA), resulting in four

derived habitat factors. We used these factors in a Discriminant Analysis to refine habitat

classifications of plots, and modelled presence/absence of claw marks using the PCA factors in a

binary logistic regression. We inventoried tree species in a subset of randomly selected plots with

claw marks alongside paired control plots with no claw marks. We compared tree community

compositions in these plots using ANOSIM and SIMPER analyses. Based on claw mark signs,

sun bear habitat use appeared to be non-random and was significantly associated with gradients

of increasing habitat intactness, from non-forest habitat to least disturbed forest. Two PCA

factors explained the probability of bears utilising a given habitat, which increased with tree

biomass and decreased with understorey cover. At the plot level, tree family and species

compositions were significantly different between plots without and with claw marks. The

abundance and use of Olacaceae stems was significantly higher in plots with claw marks.

Incorporating forest restoration strategies that enhance or increase more intact forest and the

availability of key tree resources should benefit the conservation of sun bears and encourage

natural forest regeneration in these degraded landscapes. We also emphasise the conservation


25

26

27

28

29

30

31

32

33

34

35

36

37

38

39

40

41

42

43

44

45

46

47

48


value of degraded forest habitats for this species while ensuring bear movement and connectivity

within modified landscape matrices.

Keywords: Conservation ecology; forest structure; sun bear; tree community composition;

tropical forest restoration.

1. Introduction

Deforestation rates in Sumatra are among the highest in Southeast Asia (Miettinen et al.,

2011). Land conversion has resulted in the loss of nearly half of natural forest cover in the last

25-30 years: in southern Sumatra, 53-69% of forest cover has been lost since 1985 (Uryu et al.,

2010). While large-scale forest loss has generally been prevented within protected areas in

Sumatra (Gaveau et al., 2009), few areas of lowland forest remain outside of these, while large

areas of degraded forest within logging concessions are at risk of permanent conversion to

alternative land-use (Gaveau et al., 2012). Consequently, there is a need to understand what

conservation value these degraded areas may have for key components of forest biodiversity.

The Malayan sun bear (Helarctos malayanus) is a forest-dependent species threatened by

habitat degradation, fragmentation and loss, and poaching (Scotson et al., 2017). It is listed as

Vulnerable by the IUCN due to suspected population declines resulting from these threats

throughout its range in Southeast Asia (Scotson et al., 2017). While sun bears are reported in

secondary and logged forests (Wong, 2002; Wong et al., 2004; Fredriksson, 2005; Meijaard et al.,

2005; Linkie et al., 2007; Lindsell et al., 2013; Sethy and Chauhan, 2016), in the absence of

adjacent forest, it is uncertain how sustainable sun bear populations are in degraded habitats

(Augeri, 2005; Fredriksson, 2012).


49

50

51

52

53

54

55

56

57

58

59

60

61

62

63

64

65

66

67

68

69

70

71

72


Food availability and diversity influence sun bear ranging behaviour and habitat use,

particularly in relation to fruiting events (McConkey and Galetti, 1999; Wong et al., 2004;

Augeri, 2005; Fredriksson et al., 2006; Fredriksson, 2012). Although sun bears have a varied diet,

which enables them to utilise a range of forest habitats (Linkie et al., 2007), they feed primarily

on insects, honey and fruits (Wong et al., 2002; Augeri, 2005; Fredriksson et al., 2006). In terms

of natural forest regeneration and active restoration, they are considered important seed dispersers

(Leighton, 1990; McConkey and Galetti, 1999). Despite this, effective conservation management

is limited by a lack of ecological information on spatial distribution and habitat preferences

(Servheen, 1999; Wong and Linkie, 2013), particularly when increasing landscape degradation

and fragmentation drive a growing need to understand how sun bears use human-altered forests

and how these can be managed to enhance conservation efforts (Wong and Linkie, 2013).

Indirect signs of sun bears, including scat, footprints, ripped open tree trunks, claw marks,

dug soil and damaged termite nests (Fredriksson, 2012), are useful for confirming and monitoring

the presence of this difficult to observe species (Ngoprasert et al., 2011), and for estimating

relative abundance, habitat use and requirements (Steinmetz and Garshelis, 2010; Fredriksson,

2012). Bear claw marks on trees are particularly conspicuous (Duckworth et al., 1999; Hwang et

al., 2002; Ngoprasert et al., 2011; Steinmetz et al., 2011) and tend to have comparatively slow

decay rates when compared to other types of bear signs (Steinmetz and Garshelis, 2010;

Fredriksson, 2012). These may result from foraging for fruits (Wong et al., 2002; Augeri, 2005;

Fredriksson, 2012) or insects (Payne et al., 1985; Augeri, 2005; Steinmetz, 2011), territory

marking (Augeri, 2005; Fredriksson, 2012), resting (Wong et al., 2002, 2004; Fredriksson, 2012),

or escaping danger (Payne et al., 1985; Lim, 1998; Yasuma and Andau, 2000). Thus, claw marks

can provide evidence of sun bear feeding behaviour (Wong et al., 2002) and habitat selection

(Ngoprasert et al., 2011), whereas other indirect signs, e.g. scat piles and foot prints, may simply


73

74

75

76

77

78

79

80

81

82

83

84

85

86

87

88

89

90

91

92

93

94

95

96


indicate movement through an area (Augeri, 2005). It is important to note, though, that while

claw marks may be the most prevalent type of sign left by sun bears (Ngoprasert et al., 2011;

Steinmetz, 2011), in Borneo at least, foraging signs for termites appear more abundant, with claw

marks comprising 21-47% of detected signs, depending on habitat disturbance and post-

disturbance recovery time (Fredriksson, 2012).

In this study, we use claw marks and associated signs, specifically ripped open tree trunks, as

an indicator of potential habitat utilisation by sun bears. Specifically, we assess sun bear presence

at the landscape scale using habitat structural attributes, and identify important tree resources for

bears at the plot scale in degraded forest. This will inform and enhance forest restoration

strategies that support the conservation of sun bear in the largest remaining tract of lowland

dryland forest in Sumatra. These approaches and findings may be applicable for conservation

management of degraded landscapes elsewhere within the species’ range.

2. Methods

2.1. Study area and site management

This study was conducted in Harapan Rainforest (HRF), which is situated in the dry

lowlands of southern Sumatra, Indonesia (103°17'49" E, -02°12'94" S). HRF covers 984.6 km2 of

previously logged forest characterised by a largely flat topography and ranging in elevation from

30 to 120 m a.s.l.. Mean annual rainfall is 2,390 mm, with a pronounced dry season from June to

August and highest rainfall in December.

Up to 2006, the site was managed commercially as two production forest concessions for 20-

30 years, with most areas having gone through two logging cycles. These activities have left a

mosaic of largely secondary forest habitats (totalling 933.3 km2; Schweter, 2009; Fig. 1) with the


97

98

99

100

101

102

103

104

105

106

107

108

109

110

111

112

113

114

115

116

117

118

119

120


remainder comprising oil palm, rubber and old Acacia mangium plantations, agricultural land and

scrub. Forest habitats are categorised as ‘high cover secondary forest’ (HSF; covering 37.0% of

the site), ‘medium cover secondary forest’ (MSF; 15.2%), and ‘low cover secondary forest’

(LSF; 42.5%; Schweter, 2009). The least disturbed HSF is characterised by a closed canopy with

a mixed tree species composition lacking in Dipterocarpaceae, but otherwise floristically

characteristic of dry lowland tropical forest in the region. The most degraded LSF typically has

an open canopy largely dominated by pioneer species, including Macaranga spp. and the

invasive non-native Bellucia pentamera, and a dense herbaceous understorey of Marantaceae and

Zingerberaceae (Lee and Lindsell, 2011; Briggs et al., 2012).

Since 2007, the site has been managed as an ecosystem restoration concession, and under

license decreed in response to continual large-scale national forest loss, and a need to protect and

restore degraded forests (Sheil and Meijaard, 2010; Harrison, 2011). This study supported the

management objectives of identifying the habitat requirements of key forest species, particularly

large seed dispersers, and using this empirical information to guide restoration strategies that

support effective species conservation alongside other ecosystem co-benefits.


121

122

123

124

125

126

127

128

129

130

131

132

133

134

135

136


Fig. 1. Land-cover map of the study site in 2009. ‘Non-forest habitats’ comprise small

monoculture plantations, agricultural land and scrub. ‘Plantation’ areas consist of overgrown

Acacia mangium (Schweter, 2009).

2.2. Landscape-scale habitat surveys

Between October 2009 and April 2010, we completed habitat surveys in 262 plots of 25 m

radius (0.2 ha). Plots were positioned every 200 m along parallel transects spaced 1 km apart

using a site-wide systematic sampling design (Lee and Lindsell, 2011). In total, we surveyed 24

transects of 1.8 to 2.4 km length, covering approximately 9% of the site, and proportionally

representing the main habitat types. We categorised plots in the field as HSF, MSF, LSF or ‘non-

forest habitat’ (NFH, which included scrub and cleared areas).

We estimated or counted a number of habitat attributes in each plot quarter:

Percentage cover of the herbaceous layer using a randomly positioned 1 m2 quadrat;


137

138

139

140

141

142

143

144

145

146

147

148

149

150


Understorey cover using a vertical 1 m2 quadrat of 25 points spaced 20 cm apart

(adapted from Bullock, 1996) held at 1.3 m height and 12.5 m from the central point of

the plot;

Canopy openness using a canopy-scope (Brown et al., 2000) and

Numbers of rattans and saplings (<5 cm dbh), and climbers (0, 1, 2-5, and >5 plants).

We recorded the diameter at breast height (dbh; at 1.3 m height), and tree and canopy heights

for the ten largest trees in each plot. We used these metrics to calculate the density (ha-1), basal

area (m2 ha-1), and a measure of above ground biomass (m3 ha-1) of large trees. We checked all

trunk surfaces for bear claw marks and associated signs, including ripped open tree trunks

resulting from foraging for honey (Fredriksson, 2012). Claw mark age, season and tree species

are unlikely to have influenced our assessment of habitat use: ‘old’ signs would represent bear

activity 1-2 years after commercial logging had ceased in the site in 2006; while season and tree

species only influence claw mark decay rates by <1 month (Steinmetz and Garshelis, 2010). We

did not include scat piles or footprints, as they do not necessarily reflect direct dependence upon

the immediate habitat (Augeri, 2005). We also excluded terrestrial feeding signs, as they can be

confused with other mammals (Wong et al., 2002; Fredriksson, 2012) and, compared to claw

marks, can quickly decay beyond accurate identification (Fredriksson, 2012).

2.3. Tree inventory plots

In November-December 2010, we randomly selected 24 forest plots (hereafter referred to as

“bear plots”) across two focal areas (covering 5.3 and 7.0 km2) in which we recorded bear claw

marks during the landscape-scale habitat surveys (see Section 2.2.). We paired each bear plot

with a control plot (n = 24) containing no claw-marked trees or other obvious signs of bear

habitat use. We positioned control plots 100 m away from each bear plot in a random direction.


151

152

153

154

155

156

157

158

159

160

161

162

163

164

165

166

167

168

169

170

171

172

173

174


We searched each control plot thoroughly to confirm with absolute certainty the absence of claw-

marked trees and associated signs. If we found that a control plot contained claw-marked trees,

we randomly selected a different direction and repeated the process until we found a control plot

location with no bear claw marks present.

We tagged all trees >20 cm dbh, which is a general minimum tree size climbed by sun bears

(Augeri, 2005), in each 50 m x 50 m (0.25 ha) tree inventory plot. We identified each tagged tree

to at least genus level, and often to species level with botanical support from the Indonesian

Institute of Sciences (Lembaga Ilmu Pengetahuan Indonesia, LIPI). Tree taxonomy follows that

of The Plant List (2013).

2.4. Landscape-scale habitat analysis

We explored the habitat survey data for outliers, collinearity and missing values (Zuur et al.,

2010), and to test the normality of distributions (Kolmogorov-Smirnov test) and homogeneity of

variances across groups (Levene's test). Where appropriate, we removed extreme outliers and

transformed those variables significantly different from a normal distribution (P < 0.05) to

improve uni- and multivariate normality, and the linearity of any associations between variables.

We used a Principal Components Analysis (PCA) to reduce data dimensionality from the 12

habitat variables into a set of derived factors, and to correct for multicolinearity between the

explanatory habitat variables. We only considered factors with eigenvalues of >1.0 in the final

analysis. We used a Discriminant Analysis (DA) to refine the original habitat classifications of

the plots made in the field. For this, we entered together the PCA factor scores as predictor

variables and grouped plots by the habitat type categories assigned during the habitat surveys.

The predictive value of the DA was cross-validated using a leave-one-out method of k-fold

partitioning. Since the original (subjective) habitat classifications of plots did not necessarily


175

176

177

178

179

180

181

182

183

184

185

186

187

188

189

190

191

192

193

194

195

196

197

198


confer correct classification, we reclassified plots using the predicted habitat group memberships

from the DA. We modelled the presence/absence of bear claw marks in habitat plots as a

binomial response against the retained PCA factors (predictor variables) using binary logistic

regression, with model fit assessed using a Hosmer-Lemeshow goodness-of-fit test (Hosmer and

Lemeshow, 2000). To test univariate hypotheses, we used two-sample tests (for differences

between habitat plots with and without bear signs), and four-way tests of difference (ANOVA;

between habitats) and association (preferential habitat selection/avoidance across habitats). All

analyses were carried out using SPSS v24.0 (IBM Corp., 2016).

2.5. Tree community analysis

We analysed tree community data in the paired control and bear plots at species, genus and

family levels in PRIMER v6.0 (Clarke and Gorley, 2006). The techniques and their application in

PRIMER are described fully in Clarke (1993), and Clarke and Warwick (2001). For bear and

control plots, we calculated diversity (logeH') indices, and estimated predicted species richness

from accumulation plots with 999 permutations using a Jack-knife 2 non-parametric richness

estimator (Colwell and Coddington, 1994). We used two-sample tests to investigate univariate

hypotheses (tree diversity and taxonomic abundances between bear and control plots). Mean

values are reported with S.E., unless otherwise stated.

We factorised each plot in the community data as either a bear or control plot. We pre-treated

tree abundances with a square root transformation to down-weight the importance of the most

abundant tree species (Clarke and Warwick, 2001). We constructed similarity matrices at each

taxonomic level using the Bray-Curtis coefficient on the transformed tree abundance data. We

performed analyses of similarity (ANOSIM), which tested permutations of the similarity

matrices, to determine whether there were any spatial differences between tree communities of


199

200

201

202

203

204

205

206

207

208

209

210

211

212

213

214

215

216

217

218

219

220

221

222


bear and control plots. If an ANOSIM was significant, we carried out a similarity of percentages

analysis (SIMPER) to identify which trees contributed most to any differences in tree

communities between the two plot types.

3. Results

3.1. Habitat classification

Four factors extracted using PCA accounted for 70.0% of the original variation in habitat

data (Table 1). Habitat plots with high scores along Factor 1 (F1; 32.9% of the original variability

in habitat data) were characteristic of high biomass forest with a high density of large and tall

trees. Factor 2 (F2; 16.0%) described the relationship between numbers of understorey stems and

the presence of large trees. High scores represented plots dominated by a high density of

comparatively short trees with an open understorey of few saplings and rattans. These plots were

indicative of previously logged forest where a few large trees remained standing. Factor 3 (F3;

12.8%) described forest increasing in structural complexity and density, being dominated by high

numbers of saplings and climbers, and with little herbaceous cover. Factor 4 (F4; 8.3%) described

increasingly sparse ground cover and fewer rattan stems, and an associated increase in

understorey cover as the density of any larger trees decreased.

Table 1. Extraction results of PCA of the habitat variables (Rotation method: varimax with Kaiser

normalisation). Only factor loadings (measure of the correlation between the variable and the

factor) > 0.2 are displayed.

Variable Factor 1 Factor 2 Factor 3 Factor 4


223

224

225

226

227

228

229

230

231

232

233

234

235

236

237

238

239

240

241

242

243

244

245


Eigenvalue 3.95 1.92 1.54 1.01

Percentage variance 32.9 16.0 12.8 8.3

Large tree above ground biomass +0.465

Basal area of large trees +0.435 +0.225 +0.200

Mean tree height +0.363 -0.202 -0.425

Density of large trees +0.356 +0.255 +0.319 -0.264

Mean dbh +0.353 -0.319

Mean canopy height +0.352 -0.267 -0.356

Canopy openness +0.381 -0.231

Understorey cover +0.344 +0.540

Herbaceous ground cover +0.200 -0.348 -0.601

Number of rattans -0.422 -0.425

Number of saplings -0.517 +0.246

Number of climbers +0.423

After cross-validation, DA correctly classified 77.5% (n = 203) of the original habitat

classifications using the four PCA factors as predictor variables (Table 2). No habitat plots were

misclassified by more than one habitat category (disturbance or intactness level). We reclassified

the remaining 59 plots using the predicted habitat group memberships from the DA. This resulted

in 22 plots classified as NFH, 67 as LSF, 99 as MSF, and 74 as HSF.

Table 2. Summary matrix of cross-validated DA classification results. Percentages of cases are in

parentheses.

Original

membership

Originally

assigned

habitat

type

Predicted group membership

TotalNFH LSF MSF HSF


246

247

248

249

250

251

252

253

254

255


Group

NFH 14 (73.7) 5 (26.3) 0 0 19

LSF 8 (10.9) 51 (69.9) 14 (19.2) 0 73

MSF 0 11 (10.5) 79 (75.2) 15 (14.3) 105

HSF 0 0 6 (9.2) 59 (90.8) 65

Total 22 (8.4) 67 (25.6) 99 (37.8) 74 (28.2) 262

There were significant differences in scores of all four PCA factors across habitat types

(One-way ANOVA: F(3, 251) F1 = 130.51; F(3, 255) F2 = 38.422; F(3, 256) F3 = 10.152; F(3, 256) F4 = 15.633;

all at P < 0.001; Fig. 2). Increasing habitat intactness was characterised by significantly higher F1

and F2 scores compared to areas that were more disturbed. LSF had significantly higher F3

scores than HSF, while MSF had significantly lower F4 scores than all other habitat types

(Games Howell post-hoc test, P < 0.02).


256

257

258

259

260

261

262

263


HabitatHSFMSFLSFNFH

Fact

or sc

ore

4.5

3.0

1.5

0.0

-1.5

-3.0

-4.5

-6.0

Factor 4Factor 3Factor 2Factor 1

Fig. 2. Boxplots of factor scores across the four habitat types, as reclassified by DA.

3.2. Habitat use at the landscape-scale

We recorded sun bear claw marks in 69 (26.3%) of the 262 plots surveyed. Based on the DA

habitat classification derived from the PCA factors, 37.8% (n = 28) of HSF plots, 24.2% (n = 24)

of MSF plots, and 25.4% (n = 17) of LSF plots contained claw marks (Fig. 3). Habitat plots with

bear claw marks had significantly higher F1 scores and lower F4 scores than those plots without

(t-test: t258, F1 = 3.341, P = 0.001; t258, F4 = 1.991, P = 0.048). F2 and F3 did not explain any

differences between habitat plots with or without claw marks.


264

265

266

267

268

269

270

271

272

273

274


Factor 1 scores7.56.04.53.01.5.0-1.5-3.0-4.5-6.0

Fact

or 2

scor

e s

3.0

2.0

1.0

.0

-1.0

-2.0

-3.0

-4.0

-5.0

HSF (+)HSF (-)MSF (+)MSF (-)LSF (+)LSF (-)NFH (+)NFH (-)Habitat

Fig. 3. PCA ordination of DA habitat plot classifications based on their F1 and F2 scores. The

presence (+) or absence (-) of claw marks are included for each habitat plot.

Sun bear habitat use, based on the presence of claw marks, was significantly associated with

broad habitat type (X23 = 13.810, P = 0.003). Using percentage deviations, the observed

frequency of HSF plots with claw marks present was 43.7% greater than expected, with an

absence of claw marks 15.6% less than expected, suggesting preferential selection of this habitat.

Sun bears appeared to disproportionality avoid NFH, with no claw marks present in any NFH

plots (an absence of claw marks had an observed count 35.8% more than expected). Percentage

deviations for LSF and MSF did not vary by more than ± 7%. Phi and Kendall’s tau-b correlation

coefficients were similar (Phi = 0.230, P = 0.003; tau-b = 0.174, P = 0.001), suggesting an

ordinal correlation between habitat type and the presence of claw marks.

As predictors of bear habitat use, the PCA factors fitted the data adequately in the logistic


275

276

277

278

279

280

281

282

283

284

285

286

287

288


regression model (Hosmer-Lemeshow test: Ĉ8 = 9.834, P = 0.277). Two of the four factors had a

significant effect on the probability of bears utilising the habitat. F1 had a positive effect (ß =

+0.263, Z = 3.29, P = 0.001) and an odds ratio of 1.30 (1.11-1.52 95% CIs): for a one-unit

increase in F1, there was a 30% increase in the likelihood of bears utilising the habitat (Fig. 4).

Based on F1 scores, the mean probability of bears using NFH was 0.047 compared to 0.410 in

HSF. F4 had a negative effect (ß = -0.267, Z = -2.08, P = 0.037) with an odds ratio of 0.77 (0.60-

0.98 95% CIs): a one-unit increase in F4 resulted in a 23% decrease in the odds of bears using

that habitat. F2 and F3 did not influence the probability of bears utilising the habitat (95% CIs of

their odds ratios included one).

Factor 1 scores7.56.04.53.01.5.0-1.5-3.0-4.5-6.0

P (S

un b

ear h

a bita

t use

)

0.6

0.5

0.4

0.3

0.2

0.1

0.0

HSFMSFLSFNFH

Habitat

Fig. 4. Predicted probability of sun bear habitat use in relation to habitat F1, with regression line

fitted.


289

290

291

292

293

294

295

296

297

298

299

300

301


3.3. Tree community composition and habitat use

Seven plots (four bear plots, three control plots) were illegally clear-felled during the process

of measuring and identifying tree species so were removed from our analyses. In total, 452 trees

were tagged in the remaining 41 plots (16–64 trees ha-1; mean = 44.0 ± 8.36 ha-1).

Of the 600 tree species from 107 families recorded in the site (Harrison and Swinfield,

2015), 33 tree families comprising 71 genera and 83 species were identified in the plots. Of these

families, five (Apocynaceae, Celastraceae, Centroplacaceae, Oleaceae and Symplocaceae) were

only recorded in control plots, while Cardiopteridaceae was the only family unique to bear plots.

Of the 60 tree species recorded in control plots, 18 (30.0%) were unique to these areas, while 23

(35.4%) of the 65 tree species recorded in bear plots were unique to these plots. Predicted tree

species richness was higher in control plots (SPredicted = 111) than bear plots (SPredicted = 96), and

while tree diversity was lower in control plots (HControl = 1.77 ± 0.412 SD) than in bear plots (HBear

= 1.89 ± 0.549 SD), it was not significantly different across plot types (Paired t-test: t39 = 0.774,

P = 0.444).

We recorded claw marks on 11.7% (n = 53) of trees in bear plots, and representing 51.5% of

families (n = 17) and 28.8% of species (n = 23) recorded. Claw marks were significantly

associated with stems of the family Olacaceae (37.7% of claw-marked stems; Chi-square test for

association: χ21 = 76.673, P < 0.001; Fig. 5). Of the other families, 15.1% (n = 8) of claw-marked

stems were Leguminosae, and 5.7% (n = 3) each from Clusiaceae, Euphorbiaceae and Lauraceae.

For those tree species with >10 stems recorded in plots, the most commonly utilised by bears

were Ochanostachys amentacea (Olacaceae; claw marks were recorded on 74.1% (n = 20) of

stems in this family, and 38.8% of total stems with claw marks) and Falcataria moluccana

(Leguminosae; 37.5% (n = 6) of stems had claw marks).


302

303

304

305

306

307

308

309

310

311

312

313

314

315

316

317

318

319

320

321

322

323

324

325


Fig. 5. Percentage of total tree stems (n = 452) with and without bear claw marks in the key tree

families. Total counts of stems and dominant genera (>20% of stems by family) are attached to

the bars. Chi-square test results are based on the number of stems of all other families in habitat

plots with and without claw marks; * P < 0.001.

Tree community compositions of bear and control plots were significantly different at the

family and species levels (ANOSIM: RFamily = 0.065, P = 0.049; RSpecies = 0.066, P = 0.048), but

not at the genus level (ANOSIM: RGenus = 0.050, P = 0.080). At the family level, tree communities

of control plots had an average similarity of 26.4%, while bear plots were 29.8% similar. Six of

the 33 families contributed >40% of the 73.6% dissimilarity between the tree family composition


326

327

328

329

330

331

332

333

334

335

336

337


of control and bear plots (Table 3). At the species level, tree communities of control plots had an

average similarity of 14.1% and bear plots were 16.1% similar to each other. Macaranga

gigantea (Euphorbiaceae), O. amentacea (Olacaceae) and Callicarpa pentandra (Lamiaceae)

contributed most to the 86.3% dissimilarity in tree species between control and bear plots (Table

3). Of these, the density of Olacaceae stems, represented solely by O. amentacea, was

significantly lower in control plots than bear plots (Mann-Whitney: U39 = 77.50, P < 0.001).

Table 3. Key tree families and species contributing to differences in the community composition

of bear and control plots (SIMPER). Only families and species contributing >5% to the overall

dissimilarity are presented. Cumulative percentage contributions are in parentheses. * P < 0.05;

significant difference in abundance across plot types.

Taxonomic level Mean abundance (± SE) Average

dissimilarity

(± SD)

%

Contributio

n

Control plots Bear plots

Family:

Euphorbiaceae 1.6 ± 0.39 1.5 ± 0.42 5.7 ± 1.16 7.7

Leguminosae 0.7 ± 0.21 1.4 ± 0.76 5.4 ± 0.75 7.4 (15.1)

Lamiaceae 1.3 ± 0.45 0.4 ± 0.27 5.1 ± 0.82 7.0 (22.1)

Olacaceae * 0.1 ± 0.05 1.1 ± 0.23 5.0 ± 1.30 6.8 (28.9)

Dipterocarpaceae 0.7 ± 0.25 0.9 ± 0.26 4.4 ± 1.01 6.0 (34.9)

Burseraceae 0.7 ± 0.19 0.9 ± 0.23 4.3 ± 1.12 5.9 (40.8)

Species:

Macaranga gigantea 1.1 ± 0.41 1.0 ± 0.43 5.2 ± 0.93 6.0

Ochanostachys amentacea * 0.1 ± 0.05 1.1 ± 0.23 4.8 ± 1.30 5.5 (11.5)

Callicarpa pentandra 1.3 ± 0.45 0.4 ± 0.27 4.7 ± 0.77 5.5 (17.0)


338

339

340

341

342

343

344

345

346

347

348

349

350


4. Discussion and conclusions

4.1 Landscape-scale habitat use

Sun bear claw marks were recorded in 26.3% of habitat plots, which represented all three

forest types, indicating widespread use of these disturbed habitats (Wong, 2002; Wong et al.,

2004; Fredriksson, 2005; Meijaard et al., 2005; Linkie et al., 2007; Lindsell et al., 2015). Based

on the presence of claw marks and associated signs on trees, evidence of sun bear habitat use

increased with forest intactness or recovery stage post-logging, with bears appearing to select

preferentially the least disturbed forest areas (HSF), and with broadly neutral selection of low and

medium secondary forest types. This reflects those areas that were disturbed least during logging

and/or have had more time to recover to a more mature forest structure after logging (Sethy and

Chauhan, 2016). Claw marks were not recorded in non-forest habitat. While these areas did have

fewer trees than forested habitat, they retained approximately 46% of the large tree density

recorded in the more degraded forest. If sun bears selected these areas similarly to the forested

landscape, we would have expected 5-7 of the 22 non-forest plots to include claw marks. It

appears that sun bears disproportionately avoided utilizing these most heavily impacted habitats,

despite contiguity with forested land (Augeri, 2005).

The significant effects of habitat F1 and F4, positive and negative respectively, on the

probability of bears utilising less degraded forest support this broad pattern in habitat use or

avoidance. High F1 scores were associated with characteristics of least degraded forest (taller,

more closed canopy, higher density of large trees and saplings, greater prevalence of climbers

less herbaceous ground cover), while higher F4 scores reflected habitat with low densities of

large trees, low ground cover, dense herbaceous understory and few rattans, most indicative of

the non-forest habitat. This is congruous with other studies, which identified greater habitat use


351

352

353

354

355

356

357

358

359

360

361

362

363

364

365

366

367

368

369

370

371

372

373

374


by bears in older forest, and specifically related to higher canopy, escape and ground covers

(Augeri, 2003; Sethy and Chauhan, 2016), and a low probability of detecting bears in young

disturbed forest (<5 years old; Augeri, 2005), which could be comparable to areas of LSF in our

study site. An absence of bear habitat use in areas with very few rattans (Arecaceae), high F4

scores, may also reflect the dietary importance of their fruit (Fredriksson et al., 2006).

Increasing habitat use by sun bears along a post-logging gradient of forest recovery may be

due to greater availability of key resources (Steinmetz et al., 2011; Sethy and Chauhan, 2016),

with evidence of bear habitat use strongly linked to fruit availability (Wong et al., 2002; Linkie et

al., 2007; Steinmetz et al., 2011; Wong and Linkie, 2013). More intact forest should support

higher densities and productivity of fruiting trees than more disturbed areas (Wong et al., 2012),

providing greater fruit availability and resulting in higher detection probabilities of sun bears

(Steinmetz et al., 2011; Wong et al., 2012). From a species conservation perspective at the global

scale, the impacts of logging on sun bear habitat use may be greater in Borneo than in Sumatra,

where fruit production is higher (Wich et al., 2011). Sun bears are also dependent on large trees

of certain species, e.g. Shorea spp., O. amentacea, for resting or sleeping (Wong et al., 2002;

Meijaard et al., 2005; Padmanaba et al., 2013; see Section 4.2). These are increasingly sparse in

the more degraded forest of the site through more recent or intensive logging activities (Lee and

Lindsell, 2011; Harrison and Swinfield, 2015). Complementary data on sun bear relative

abundance support the spatial relationship between the likelihood of habitat use and gradient of

forest intactness in the site (Lindsell et al., 2015).

4.2 Plot-scale habitat use

Tree community composition was significantly difference between areas where there was no

evidence of bear habitat use and those with clear evidence of habitat use. These differences were


375

376

377

378

379

380

381

382

383

384

385

386

387

388

389

390

391

392

393

394

395

396

397

398


driven mainly by trees of Euphorbiaceae, Leguminosae, Lamiaceae, Olacaceae, Dipterocarpaceae

and Burseraceae, and the species M. gigantea, O. amentacea and C. pentandra.

Sun bears most frequently climb, for multiple reasons, large trees of Anacardiaceae,

Bignoniaceae, Burseraceae, Combretaceae, Dilleniaceae, Dipterocarpaceae (e.g. Shorea),

Euphorbiaceae, Fagaceae (e.g. Lithocarpus), Labiatae, Lauraceae, Leguminosae, Moraceae,

Myrtaceae (e.g. Eugenia, Syzygium), Sapindaceae and Tiliaceae (Wong et al., 2002; Augeri,

2003, 2005; Steinmetz, 2011). Where present in the study plots, we recorded claw marks on trees

of all these families, except Burseraceae and Sapindaceae.

We found that sun bears in this previously logged, dry lowland forest preferentially select

stems of Fabaceae and Olacaceae. Olacaceae was represented by a single species, O. amentacea,

an animal-dispersed species that is known to fruit throughout the year (World Agroforestry

Centre, 2018), and climbed repeatedly in HRF. Like the Leguminosae genera Dialium and

Koompassia, O. amentacea is also more abundant in less disturbed forest in HRF, indicative of

its late successional condition (Swinfield et al., 2016). There are no records of sun bears feeding

on the fruits of Olacaceae (Fredriksson, 2012), but O. amentacea is a preferred nesting tree

(Padmanaba et al., 2013). Coupled with a naturally scattered distribution (World Agroforestry

Centre, 2018), it appears the importance of O. amentacea for sun bears is related to its nesting

suitability, especially in logged forest with a reduction in large dipterocarps, such as Shorea spp.

which are often selected for nesting sites (Wong et al., 2002).

We recorded claw marks on 15.4% of Moraceae stems (Artocarpus spp., Ficus spp.) and

11.8% of Myrtaceae stems (Syzygium spp.). These families, along with Burseraceae, provide the

majority of fruit in the bear’s diet (McConkey and Galetti, 1999; Wong et al., 2002; Augeri,

2003, 2005; Fredriksson et al., 2006; Steinmetz, 2011), with Ficus species of particular important

(Leighton, 1990; McConkey and Galetti, 1999; Wong et al., 2002; Fredriksson et al., 2006). The


399

400

401

402

403

404

405

406

407

408

409

410

411

412

413

414

415

416

417

418

419

420

421

422


frequency bears used these stems in this logged forest is commensurate with their recorded use

elsewhere (Wong et al., 2002; Fredriksson et al., 2006), although there appears to be no

preferential selection of these stems.

Of other important tree families, no Burseraceae stems (Dacryodes, Santiria, Canarium spp.)

had claw marks on them. The fact that Dacryodes and Canarium spp. are dioecious (Fern, 2016),

and characteristic of inter-annual phenological variation in the region (Thomas and LaFrankie,

1993), may explain an absence of bear usage in logged forest, while bears do not necessarily

climb Santiria spp. to access the fruit (Wong et al., 2002). Euphorbiaceae appear under selected,

possibly due to higher proportions of pioneer Endospermum and Macaranga trees, and reflecting

a lower likelihood of bears being recorded in these areas.

4.3. Study considerations

While we present strong evidence for landscape and plot-scale differences in sun bear habitat

use in degraded forest based on claw marks, we do not consider possible temporal shifts in

habitat and resource use (Augeri, 2005; Fredriksson et al., 2006) or capture information on other

foraging strategies (Fredriksson et al., 2006) and their associated signs (Fredriksson, 2012). This

may have resulted in an incomplete understanding of habitat use by sun bears (Steinmetz, 2011;

Fredriksson, 2012) across a gradient of forest intactness. In more degraded forest and secondary

growth, where fruit availability is lower, sun bears may spend more time travelling (Wong et al.,

2012) and foraging terrestrially (Wong et al., 2002; Fredriksson, 2012), reducing claw mark

prevalence (Ngoprasert et al., 2011) and evidence of habitat use based on this type of sign. Also,

in the Bornean region of the species’ geographic range at least, claw marks may only represent

21% (to 47%) of signs left by sun bears, depending on habitat disturbance and recovery time,

with the majority of signs (51-71%) associated with foraging for above and belowground termites


423

424

425

426

427

428

429

430

431

432

433

434

435

436

437

438

439

440

441

442

443

444

445

446


(Fredriksson, 2012). There do appear to be some positive associations between the density of

termite nest signs and claw marks (Fredriksson, 2012), which may further support the use of claw

marks as an overall proxy for habitat use, especially when focusing on presence/absence at the

plot or landscape scale rather than sign density and relative abundance. Overall, our conclusions

on sun bear habitat use in degraded forest landscapes of Sumatra are couched in these limitations.

However, this does not diminish the conservation importance of less disturbed logged forest for

sun bears (Wong et al., 2002; Augeri, 2003; Linkie et al., 2007; Steinmetz et al., 2011; Wong and

Linkie, 2013; Sethy and Chauhan, 2016), and especially in a landscape where degraded forest is

vulnerable to land-use conversion (Gaveau et al., 2012).

4.4. Forest management and restoration implications

Our study supports the 20-year site management plan, helping inform restoration strategies

that benefit a species of conservation concern. It emphasizes the importance of less disturbed

forest habitats for sun bears and a lack of use of more heavily disturbed areas in logged forest. At

the site level, this provides strong evidence for protecting existing least disturbed forest cover to

allow natural regeneration to progress and, through collaboration with local partners and

stakeholders, prevent any further habitat degradation.

Indicators of restoration success should include changes in forest intactness attributes, as

described by F1, the strongest predictor of sun bear habitat use. Management approaches that

encourage structural restoration of NFH, with only a 4.7% average likelihood of bear habitat use,

to forest habitat will, on average, result in 19.7% (LSF), 32.4% (MSF) and 36.3% (HSF)

increases in habitat use by bears across the gradient of forest intactness described at HRF.

Improving LSF to MSF or HSF should increase the likelihood of habitat use by 12.7% and

16.5%, respectively. Marginal gains in bear habitat use are made if MSF returns to HSF (3.9%).


447

448

449

450

451

452

453

454

455

456

457

458

459

460

461

462

463

464

465

466

467

468

469

470


Consequently, perhaps the most cost and resource-effective approach for enhancing sun bear

habitat use is the restoration of habitat (NFH or LSF) to forest that is structurally similar to MSF.

The benefit to sun bears diminishes with further commitment of resources to actively restoring

MSF to HSF; at this point, natural regeneration becomes more valuable.

Structurally, in disturbed forest with average tree heights <18 m, canopy height <11 m, the

largest trees with girths of <100 cm, basal area <9.5 m2 / ha, and <7.5 rattan stems / ha, the

probability of bear use drops below 0.25. Conversely, the probability of bear use rises to >0.40 in

previously logged forest with average tree heights >26 m, canopy height >17 m, the largest trees

with girths of >115 cm, basal area >13 m2 / ha, and >15 rattan stems / ha. These forest condition

thresholds provide measurable targets against which restoration success can be monitored

alongside the likelihood of bears using the habitat, or against which the impact of any

degradation to bear habitat use can be evaluated (e.g. Augeri, 2004).

Landscape-scale management and restoration approaches that help push habitat quality

further along a post-logging gradient of recovery, e.g. reducing interior edge-area ratios, should

enhance sun bear access to more productive foraging areas, and increase the diversity and

abundance of key resources (Meijaard and Sheil, 2008). Complementary studies to determine

how different restoration strategies may improve fruit availability and distribution would be

beneficial. In turn, as seed dispersers (Leighton, 1990; McConkey and Galetti, 1999), sun bears

are potentially important contributors to natural forest regeneration, benefitting the recovery of

MSF to HSF.

Direct seeding or enrichment planting strategies (e.g. Harrison and Swinfield, 2015) should

consider including tree species preferentially utilised by sun bears in these forests, assuming that

mother trees can be located, and that those seeds/saplings can become established under the

conditions of different levels of forest degradation. In this instance, O. amentacea appears to be a


471

472

473

474

475

476

477

478

479

480

481

482

483

484

485

486

487

488

489

490

491

492

493

494


particularly important tree species for sun bears, and most likely for resting or nesting

(Padmanaba et al., 2013). While natural regeneration of this slow-growing species tends to be

sparse and scattered (Asian Regional Workshop, 1998), it can be seed-propagated and, assuming

successful germination, would probably take at least 40 years (World Agroforestry Centre, 2018)

to reach a minimum size for sun bears to climb (Augeri, 2005). Direct seeding or planting of

saplings could also help reduce herbaceous cover, enhancing growth of other late successional

species and reducing silviculture costs (World Agroforestry Centre, 2018). Other potentially

important tree species for direct seeding/enrichment planting include late successional species of

Dialium, Falcataria and Koompassia.

Silvicultural trials aimed at assisting natural regeneration (ANR) at HRF have selectively

removed from the understorey two dominant pioneer species, M. gigantea (Euphorbiaceae) and

the invasive, non-native B. pentamera (Melastomataceae), and lianas (Swinfield et al., 2016).

While only over a short time period, the slightly increased canopy openness and increased growth

of late successional stems (≥ 2 cm dbh) at lower intensity thinning (Swinfield et al., 2016) are

promising indicators of natural forest regeneration for sun bears. Such ANR should enhance

habitat suitability for sun bears as the forest understory becomes more open and with an increase

in growth of important late successional species, while M. gigantea is under-selected by bears.

Trialling the removal of pioneer Callicarpa species may also prove beneficial, as they appear to

provide no value to sun bears in this degraded forest.

Other site restoration strategies include planting experiments of Aquilaria malaccensis

(gaharu), a high-value non-timber forest product (Harrison and Swinfield, 2015). Sun bears are

sensitive to, and avoid minor disturbances such as gaharu harvesting (Augeri, 2004). While this

should have minimal impact on bears as it will replace unsuitable Acacia mangium plantation in

the site, it is important to consider how this may displace bears from the immediate area and


495

496

497

498

499

500

501

502

503

504

505

506

507

508

509

510

511

512

513

514

515

516

517

518


reduce landscape permeability.

We recommend the use of surveys of new sun bear claw marks to monitor changes in sun

bear distribution and occupancy within HRF, particularly alongside active site management. They

are low-cost and easy to conduct (Fredriksson, 2012), while in Sumatra at least and with observer

training, bear claw marks should not be confused with any sympatric tree-climbing species. It is

also worthwhile considering the inclusion of termite feeding signs, as a potentially prevalent and

spatially dependent indicator of habitat use (Fredriksson, 2012). This overall approach could be

used to help validate the most compatible and economically viable large-scale restoration

approach(es) (Harrison and Swinfield, 2015) for effective sun bear conservation management in

recovering forest landscapes. It also lends itself to a standardised approach for monitoring the

population and occupancy of the species throughout its range (Linkie et al. 2007; Steinmetz et al.

2011; Wong et al. 2012).

While sun bears are capable of utilising a range of habitat types, including oil palm

plantations (Servheen, 1999; Augeri, 2002; Normua et al., 2004) and farmland crops (Linkie et

al., 2007), some monoculture landscapes appear to be entirely unsuitable, at least without

proximate forested areas (McShea et al., 2009). If fruit availability is limiting, sun bears may be

forced to forage more beyond forest boundaries (Augeri, 2004). Consequently, meta-population

conservation management of sun bears in modified landscapes, such as southern Sumatra, must

consider movement of individuals through permeable anthropogenic matrices (McShea et al.,

2009) alongside forest management. In the case of HRF, as an exemplar of ecosystem restoration,

increasing site isolation may ultimately undermine any empirically informed forest restoration

strategies for sun bears and, indeed, other large forest mammals.

It is important to emphasise that in a landscape of heavily impacted forest cover (Gaveau et

al., 2009, 2012), areas such as HRF tend to be representative of available lowland forest in


519

520

521

522

523

524

525

526

527

528

529

530

531

532

533

534

535

536

537

538

539

540

541

542


Sumatra. Therefore, while these degraded forests may not provide optimal habitat for sun bears,

our study helps highlight what their adapted distribution is within the habitat available to them

(Wong and Linkie, 2013). Furthermore, degraded lowland forests potentially provide better

habitat quality for sun bears than undisturbed higher elevation forest (Linkie et al., 2007),

highlighting the importance of protecting and enhancing these threatened landscapes (Wong and

Linkie, 2013; Fredriksson, 2012).

Author contributions

DL and VP carried out the field research and collected the data; DL performed the statistical

analyses; DL, VP and JL wrote the paper.

Acknowledgments

Our research supports the forest conservation and restoration activities of HRF, a

collaborative initiative of the Royal Society for the Protection of Birds, BirdLife International

and Burung Indonesia. Funding was provided by a Research and Conservation Grant from the

International Association for Bear Research and Management and the Darwin Initiative of the

Department for Environment, Food and Rural Affairs (Reference no. 162/16/005). Yayasan

Konservasi Ekosistem Hutan Indonesia support of this work was co-financed by the Federal

Republic of Germany within the framework of the International Climate Protection Initiative of

the Federal Ministry for the Environment, Nature Conservation and Nuclear Safety (BMU)

through KfW Development Bank. We thank the research staff of HRF for assisting in the surveys

and Mr D. Girmansyah from the Indonesian Institute of Sciences (LIPI) for his botanical

expertise. We are also grateful to two anonymous reviewers for their valuable comments on this

manuscript.


543

544

545

546

547

548

549

550

551

552

553

554

555

556

557

558

559

560

561

562

563

564

565

566


References

Asian Regional Workshop, 1998. Ochanostachys amentacea. Conservation & Sustainable

Management of Trees, Vietnam, August 1996. The IUCN Red List of Threatened Species

1998: e.T33238A9770332.

<http://dx.doi.org/10.2305/IUCN.UK.1998.RLTS.T33238A9770332.en>. Downloaded on 17

June 2018.

Augeri, D.M., 2002. Effects on sun bear (Helarctos malayanus) habitat selection, ecology and

landscape use. Paper presented at the International Bear Association Annual Meeting 2002,

Steinkjer, Norway.

Augeri, D.M., 2003. Conservation of the Malayan Sun Bear (Helarctos malayanus) in Indonesia:

Mitigating potential bear/human conflicts and disturbance effects on sun bear ecology and

landscape use. Unpublished report for the Indonesian Institute of Sciences.

Augeri, D.M., 2004. Effects of disturbance on Malayan sun bear habitat use. Paper presented at

the International Conference on Conservation Science, Cambridge, UK.

Augeri, D.M., 2005. On the biogeographic ecology of the Malayan sun bear. PhD Dissertation,

University of Cambridge, Cambridge, UK.

Briggs, M., de Kok, R., Moat, J., Whaley, O., Williams, J., 2012. Vegetation mapping for

reforestation and carbon capture in the Harapan Rainforest. Royal Botanic Gardens, Kew,

London, UK.

Brown, N., Jennings, S., Wheeler, P., Nabe-Nielsen, J., 2000. An improved method for the rapid

assessment of forest under-storey light environments. Journal of Applied Ecology 37, 1044–

1053.

Bullock, J., 1996. Plants. In: Sutherland, W.J. (Ed.), Ecological census techniques – A handbook.


567

568

569

570

571

572

573

574

575

576

577

578

579

580

581

582

583

584

585

586

587

588

589

590


Pp. 186–213. Cambridge University Press, Cambridge, UK.

Clarke, K.R., 1993. Non-parametric multivariate analyses of changes in community structure.

Australian Journal of Ecology 18, 117–143.

Clarke, K.R., Gorley, R.N., 2006. Primer v6: User Manual/Tutorial. Primer-E, Plymouth, UK.

Clarke, K.R., Warwick, R.M., 2001. Change in marine communities: An approach to statistical

analysis and interpretation. Second edition. PRIMER-E, Plymouth, UK.

Colwell, R.K., Coddington, J.A., 1994. Estimating terrestrial biodiversity through extrapolation.

Philosophical Transactions of the Royal Society B: Biological Sciences 345, 101–118.

Duckworth, J.W., Salter, R.E., Khounboline, K., 1999. Wildlife in Lao PDR: the 1999 status

report. IUCN-The World Conservation Union/Wildlife Conservation Society/Centre for

Protected Areas and Watershed Management, Vientiane, Lao PDR.

Fern, K., 2016. Useful tropical plants database. < http://tropical.theferns.info/>. Last updated on

16 August 2016. Downloaded on 17 June 2018.

Fredriksson, G., 2005. Human-sun bear conflicts in East Kalimantan, Indonesian Borneo. Ursus

16, 130-137.

Fredriksson, G., Wich, S.A, Trisno, 2006. Frugivory in sun bears (Helarctos malayanus) is linked

to El Niño-related fluctuations in fruiting phenology, East Kalimantan, Indonesia. Biological

Journal of the Linnean Society 89, 489–508.

Fredriksson, G., 2012. Effects of El Niño and large-scale forest fires on the ecology and

conservation of Malayan sun bears (Helarctos malayanus) in East Kalimantan, Indonesian

Borneo. PhD Dissertation, University of Amsterdam, Amsterdam, Neth.

Gaveau, D.L.A., Epting, J., Lyne, O., Linkie, M., Kumara, I., Kanninen, M., Leader-Williams,

N., 2009. Evaluating whether protected areas reduce tropical deforestation in Sumatra.

Journal of Biogeography 36, 2165–2175.


591

592

593

594

595

596

597

598

599

600

601

602

603

604

605

606

607

608

609

610

611

612

613

614


Gaveau D.L.A., Curran, L.M., Paoli, G.D., Carlson, K.M., Wells, P., Besse-Rimba, A., Ratnasari,

D., Leader-Williams, N., 2012. Examining protected area effectiveness in Sumatra:

importance of regulations governing unprotected lands. Conservation Letters 5, 142–148.

Harrison, R.D. 2011. Tropical forests: still vital when degraded. Nature 479, 178–179.

Harrison, R.D., Swinfield, T. 2015. Restoration of logged humid tropical forests: an experimental

programme at Harapan Rainforest, Indonesia. Tropical Conservation Science 8, 4–16.

Hosmer, D.W., Lemeshow, S., 2000. Applied logistic regression. John Wiley & Sons, Inc., New

York, New York, USA.

Hwang, M.-S., Garshelis, D.L., Wang, Y., 2002. Diets of Asiatic black bears in Taiwan, with

methodological and geographical comparisons. Ursus 13, 111–125.

IBM Corp., 2016. SPSS Statistics for Windows, v24.0. IBM Corp., Armonk, NY.

Lee, D.C., Lindsell, J.A. 2011. Biodiversity of Harapan Rainforest: summary report on baseline

surveys of mammals, birds, fish, herptiles, butterflies and habitat. The Royal Society for the

Protection of Birds, Sandy, UK.

Leighton, M. 1990. Seed dispersal syndromes of Bornean fruits: The dependence of plants on

birds and mammals for seed dispersal and forest regeneration. A preliminary report on

research results from Gunung Palung. Center for Research and Development in Biology-LIPI

and the Sub-directorate of Nature Conservation (PHPA), Ministry of Forestry, Jakarta,

Indonesia.

Lindsell, J.A., Lee, D.C., Powell, V.J., Gemita, E., 2015. Availability of large seed-dispersers for

restoration of degraded tropical forest. Tropical Conservation Science 8, 17–27.

Linkie, M., Dinatab, Y., Nugrohob, A., Haidir, I.A., 2007. Estimating occupancy of a data

deficient mammalian species living in tropical rainforests: Sun bears in the Kerinci Seblat

region, Sumatra. Biological Conservation 137, 20–27.


615

616

617

618

619

620

621

622

623

624

625

626

627

628

629

630

631

632

633

634

635

636

637

638


Lim, B.L., 1998. The sun bear. In: Yong H.S. (Ed.), The encyclopaedia of Malaysia-Animals. Pp.

24–25. Archipelago Press, Kuala Lumpur, Malaysia.

McConkey, K., Galetti, M., 1999. Seed dispersal by the sun bear Helarctos malayanus in Central

Borneo. Journal of Tropical Ecology 15, 237–241.

McShea, W.J., Stewart, C., Peterson, L., Erb, P., Stuebing, R., Giman, B., 2009. The importance

of secondary forest blocks for terrestrial mammals within an Acacia/secondary forest matrix

in Sarawak, Malaysia. Biological Conservation 142, 3108–3119.

Meijaard, E., Sheil, D., 2008. The persistence and conservation of Borneo’s mammals in lowland

rain forests managed for timber: observations, overviews and opportunities. Ecological

Research 23, 21–34.

Meijaard, E., Sheil, D., Nasi, R., Augeri, D., Rosenbaum, B., Iskandar, D., Setyawati, T.,

Lammertink, M., Rachmatika, I., Wong, A., Soehartono, T., Stanley, S., O’Brien, T., 2005.

Life after logging: Reconciling wildlife conservation and production forestry in Indonesian

Borneo. The Center for International Forestry Research (CIFOR), Bogor, Indonesia.

Miettinen, J., Shi, C., Liew, S.C., 2011. Deforestation rates in insular Southeast Asia between

2000 and 2010. Global Change Biology 17, 2261–2270.

Ngoprasert, D., Steinmetz, R., Reed, D.H., Savini, T., Gale, G.A., 2011. Influence of fruit on

habitat selection of Asian bears in a tropical forest. Journal of Wildlife Management 75,

588–595.

Normua, F., Higashi, S., Ambu, L., Mohamad, M., 2004. Notes on oil palm plantation use and

seasonal spatial relationships of sun bears in Sabah, Malaysia. Ursus 15, 227-231.

Padmanaba, M., Sheil, D., Basuki, I., Liswanti, N., 2013. Accessing local knowledge to identify

where species of conservation concern occur in a tropical forest landscape. Environmental

Management 52, 348–359.


639

640

641

642

643

644

645

646

647

648

649

650

651

652

653

654

655

656

657

658

659

660

661

662


Payne, J., Francis, C.M., Phillipps, K., 1985. A field guide to the mammals of Borneo. World

Wildlife Fund Malaysia, Kuala Lumpur, Malaysia.

Schweter, M., 2009. Forest cover analysis – Harapan Rainforest, Sumatra, Indonesia. The Royal

Society for the Protection of Birds, Sandy, UK.

Scotson, L., Fredriksson, G., Augeri, D., Cheah, C., Ngoprasert, D., Wai-Ming, W. 2017.

Helarctos malayanus (errata version published in 2018). The IUCN Red List of Threatened

Species 2017: e.T9760A123798233. http://dx.doi.org/10.2305/IUCN.UK.2017-

3.RLTS.T9760A45033547.en. Downloaded on 10 December 2018.

Sethy, J., Chauhan, N.S., 2016. Assessing habitat use by sun bears in Namdapha Tiger Reserve,

Arunachal Pradesh, India. Applied Ecology and Environmental Research 14, 215–236.

Servheen, C., 1999. Sun bear conservation action plan. In: Herrero, S., Peyton, B., Servheen, C.,

(Eds.), Bears: status survey and conservation action plan. Pp. 219–224. IUCN-World

Conservation Union, Gland, Switzerland.

Sheil, D., Meijaard, E., 2010. Purity and prejudice: deluding ourselves about biodiversity

conservation. Biotropica 42, 566–568.

Steinmetz, R.G., Garshelis, D.L., 2010. Estimating ages of bear claw marks in Southeast Asian

tropical forests as an aid to population monitoring. Ursus 21, 143–153.

Steinmetz, R.G., 2011. Ecology and distribution of sympatric Asiatic black bears and sun bears in

the tropical dry forest ecosystem of Southeast Asia. In: McShea, W., Davies, S.,

Bhumpakphan, N. (Eds.), Dry forests of Asia: conservation and ecology. Pp. 249–273.

Smithsonian Institution Press, Washington, D.C.

Steinmetz, R.G., Garshelis, D.L., Chutipong, W., Seuaturien, N., 2011. The shared preference

niche of sympatric Asiatic black bears and sun bears in a tropical forest mosaic. PLoS ONE

6, e14509. doi:10.1371/journal.pone.0014509.


663

664

665

666

667

668

669

670

671

672

673

674

675

676

677

678

679

680

681

682

683

684

685

686


Swinfield, T., Afriandi, R., Antoni, F., Harrison, R.D., 2016. Accelerating tropical forest

restoration through the selective removal of pioneer species. Forest Ecology and

Management 381, 209–216.

The Plant List 2013. Version 1.1. Published on the internet; http://www.theplantlist.org/

(accessed 17th June 2018).

Thomas, S.C., LaFrankie, J.V., 1993. Sex, size and inter-year variation in flowering among

dioecious trees of the Malayan rain forest. Ecology 74, 1529–1537.

Uryu, Y., Purastuti, E., Laumonier, Y., Sunarto, Setiabudi, Budiman, A., Yulianto, K., Sudibyo,

A., Hadian, O., Kosasih, D.A., Stüwe, M., 2010. Sumatra’s forests, their wildlife and the

climate. Windows in time: 1985, 1990, 2000 and 2009. Technical Report to the National

Forestry Council (DKN) of Indonesia. WWF-Indonesia, Jakarta, Indonesia.

Wich, S.A., Vogel, E.R., Larsen, M.D., Fredriksson, G., Leighton, M., Yeager, C.P., Brearley,

F.Q., van Schaik, C.P., Marshall, A.J., 2011. Forest fruit production is higher on Sumatra

than on Borneo. PLoS ONE 6(6), e21278. doi:10.1371/journal.pone.0021278.

Wong, S.T., 2002. The ecology of Malayan sun bear (Helarctos malayanus) in the lowland

tropical forest of Borneo. MSc. Thesis, University of Montana

Wong, S.T., Servheen, C., Ambu, L., 2002. Food habit of Malayan sun bears in lowland tropical

forest of Borneo. Ursus 13, 127–136.

Wong, S.T., Servheen, C.W., Ambu, L., 2004. Home range, movement and activity patterns and

bedding sites of Malayan sun bear Helarctos malayanus in the rainforest of Borneo.

Biological Conservation 119, 169-181.

Wong, W-M., Leader-Williams, N., Linkie, M., 2012. Quantifying changes in sun bear

distribution and their forest habitat in Sumatra. Animal Conservation 16, 216–223.

Wong, W-M., Linkie, M., 2013. Managing sun bears in a changing tropical landscape. Diversity


687

688

689

690

691

692

693

694

695

696

697

698

699

700

701

702

703

704

705

706

707

708

709

710


and Distributions 19, 700–709.

World Agroforestry Centre, 2018. Tree functional attributes and ecological database.

<http://www.worldagroforestry.org/output/tree-functional-and-ecological-databases>.

Downloaded on 19 June 2018.

Yasuma, S., Andau, M., 2000. Mammals of Sabah, Part 2, habitat and ecology. Japan

International Cooperation Agency and Sabah Wildlife Department, Kota Kinabalu, Malaysia.

Zuur, A.F., Ieno, E.N., Elphick, C.S., 2010. A protocol for data exploration to avoid common

statistical problems. Methods in Ecology and Evolution 1, 3–14.


711

712

713

714

715

716

717

718

pure.southwales.ac.uk€¦ · Web view1 Lee et al.: Habitat utilisation by sun bears *...

Documents

Transcript of pure.southwales.ac.uk€¦ · Web view1 Lee et al.: Habitat utilisation by sun bears *...