Feeding Postures of Cao Vit Gibbons (Nomascus ... - Springer Link

Int J Primatol (2015) 36:1036–1054 DOI 10.1007/s10764-015-9871-z

Feeding Postures of Cao Vit Gibbons (Nomascus nasutus) Living in a Low-Canopy Karst Forest Hanlan Fei 1 & Changyong Ma 1 & Thad Q. Bartlett 2 & Ran Dai 1 & Wen Xiao 1 & Pengfei Fan 1

Received: 12 August 2015 / Accepted: 25 September 2015 / Published online: 4 November 2015 # Springer Science+Business Media New York 2015

Abstract Food acquisition is an important factor in the evolution of primate postural behavior. Gibbons are well known for their ability to exploit terminal branches by means of below branch suspensory feeding, but few studies of gibbon positional behavior have been conducted since the seminal work of the 1970s and 1980s. We studied the feeding posture of three cao vit gibbon groups living in degraded karst forest in Bangliang Gibbon Nature Reserve between August 2008 and December 2009 to determine if body mass, age, and food type affect feeding posture. We found that cao vit gibbons spent most of their time feeding from branches (59.4 %) and twigs (33.2 %) in the middle canopy of the forest (5–10 m). They used suspensory hanging and sitting as their main feeding postures. Large-bodied gibbons spent more time on larger supports than smaller juveniles when feeding on nonfig fruit and leaves. In addition, gibbons of all age–sex classes adopted a suspensory posture more often when using smaller (twigs) or more flexible (lianas) supports. We found little evidence of age–sex differences in the frequency of suspensory feeding. The subtle differences we did detect suggest that intragroup feeding competition or ontogeny may confound the body size effects on feeding posture. Overall our findings conform to the view that within species positional behavior is largely constrained by musculoskeletal anatomy and not by habitat quality because cao vit gibbons showed a similar pattern of canopy and substrate use to gibbons occupying less disturbed forests.

* Wen Xiao [email protected] * Pengfei Fan [email protected] 1

Institute of Eastern–Himalaya Biodiversity Research, Dali University, Dali, Yunnan 671000, PR China

2

Department of Anthropology, The University of Texas at San Antonio, San Antonio, Texas 78249, USA

Feeding Postures of Cao Vit Gibbon

1037

Keywords Age–sex categories . Body mass . Cao vit gibbon . Feeding postures . Suspensory

Introduction Food acquisition is an important factor in the evolution of postural behaviors in primates (Andrews and Groves 1976; Chivers 1974; Garber 2011; Grand 1972; McGraw 1998; Myatt and Thorpe 2011). Typically, the food resources most exploited by arboreal primates, i.e., fruits, flowers, young leaves, and insects, are located in the small peripheral branches of the main canopy (Houle et al. 2007; Sussman 1991; Sussman et al. 2013) and individuals need to access and occupy such areas during feeding while taking safety and comfort into account (Cant 1992; McGraw 1998) and without expending an excessive amount of energy (Grand 1972). Branches taper to the end, and as a consequence of decreasing diameter they present an increasing danger of breakage as they get smaller and pose more of a problem in maintaining balance for individuals that walk on top of them (Cant 1987; Myatt and Thorpe 2011). One solution to the problem posed by large body size in relation to small substrates is below branches suspension. In the New World, northern muriquis (Brachyteles hypoxanthus) access fruit and leaves by hanging assisted by their elongated grasping tails, a feeding strategy that allows them access to both small terminal branches and larger substrates (Iurck et al. 2013). Similarly, gibbons (Hylobatidae) access fruit on terminal braches by suspending below branches by their long arms (Chivers, 1974; Fleagle 1976), though gibbons sit more often than they suspend when feeding on leaves (Chivers 1974; Fleagle 1976). Macaques (Macaca spp.) also access terminal branches, but do so by sitting or standing above branches and pulling or breaking off the food-bearing terminal end (Grand 1972). Within species, body mass and ontogenetic changes in motor control, as well as limb and body proportions, are expected to play an important role in primate positional repertoires including feeding posture (Cant 1992; Fleagle and Mittermeier 1980; Myatt and Thorpe 2011; Zhu et al. 2015). It has long been argued that larger sized animals will use larger supports more frequently and engage in suspensory postures more often (Cant 1987, 1992; Fleagle and Mittermeier 1980; Myatt and Thorpe 2011; Rose 1974). In practice, the strength of the relationship between body size and positional behavior may have been overstated (Garber 2011). Nevertheless, intraspecific variation in feeding postures and support selection among members of a social group based on adult sex differences in body mass have been documented in some species. For example, the larger males of golden snub-nosed monkeys (Rhinopithecus roxellana) tend to use larger and more stable arboreal supports during feeding (Zhu et al. 2015), as do lowland gorilla males (Gorilla gorilla: Remis 1995). Similarly, fully flanged adult male orangutans (Pongo sp.) are reported to use larger supports more frequently than females. Although for the most part orangutan males do not exploit suspensory postures in the sense of hanging directly below a single branch (Cant 1987; Thorpe and Crompton 2009), adult male orangutans do perform forelimb–hind limb suspension and use multiple supports simultaneously to accommodate their large body size (Myatt and Thorpe 2011). Body mass differences between adults and juveniles may also lead to differences in postural behavior (Workman and Covert 2005), but age-based differences in positional behavior have been less well studied (cf. Bezanson 2009; Doran 1997; Dunham 2015; Myatt and Thorpe 2011).

1038

H. Fei et al.

One goal of conservation research is to understand the degree to which anthropogenically altered habitats disrupt the feeding and foraging of endangered species because any obstacles to accessing and occupying feeding sites will have negative consequences for the species. Changes in the age, species diversity, or stem density of trees within a habitat can alter the substrates available for feeding and locomotion with concomitant changes in the frequency of different positional behaviors (Dagosto and Yamashita 1998; Gebo and Chapman 1995; Manduell et al. 2012). Research on langurs living in degraded karst forests has documented distinct patterns of positional behavior and substrate use relative to other langurs (Trachypithecus delacouri: Workman and Schmitt 2012; T. francois: Zhou et al. 2013; and T. leucocephalus: Huang and Li 2005). However, it is not always the case that changes in support availability translate into significant changes in the positional behavior a species exhibits. Many studies have shown that even in highly disturbed areas a species’ typical postural repertoire is largely conserved (Dunham and McGraw 2014; Garber and Pruetz 1995; Manduell et al. 2012). For example, Dunham and McGraw (2014) showed that the positional repertoire of black and white colobus (Colobus angolensis palliatus) was consistent across three social groups despite marked intergroup differences in forest structure and even substrate use. The authors conclude that anatomical adaptations limit the range of variation expressed by individuals within a species, but that this does not necessarily have negative consequences for species survival. Gibbons exhibit a suite of interrelated morphological, behavioral, and ecological traits including relatively small body size, long arms, and hook-like hands, a frugivorous diet, suspensory feeding from twigs, fast brachiating locomotion, generally small territory size, and small group size (Bartlett 2011). Suspensory feeding is hypothesized to offer important benefits to gibbons, because it expands the sphere of area they are able to reach while feeding (Grand 1972). It is estimated that through suspension gibbons can easily double the space they can access during feeding without moving to a new location in the tree crown (Grand 1972; Mittermeier and Fleagle 1976). Although gibbons are well known for their suspensory feeding posture, little work on gibbon locomotion and posture has been done since the seminal work of the 1970s and 1980s (Chivers, 1974; Fleagle 1976; Gittins 1983; Srikosamatara 1984). Nevertheless, what is known is broadly consistent with the patterns described above and is consistent across siamangs (Symphalangus syndactylus: Chivers 1974; Fleagle 1976), agile gibbons (Hylobates agilis: Gittins 1983), and pileated gibbons (Hylobates pileatus: Srikosamatara 1984). No information on feeding posture is available for gibbons of the genus Nomascus or from gibbon populations inhabiting more disturbed forest, such as the degraded low-canopy karst forest to which all living cao vit gibbons are today confined. The cao vit gibbon is a Critically Endangered species (IUCN, 2013), with only one population surviving along the China–Vietnam border. Its habitat has been degraded by various human activities and the mean canopy height within the study area is only 10.5 m, which is much lower than other tropical gibbon habitats (Fan et al. 2011). Most trees in this area were 5–10 m (70.8 %) or 11–15 m high (21.7 %, Fan et al. 2013a). By contrast Gittins (1983) found that gibbons typically fed at 15–35 m high. Low-canopy karst forest poses a number of challenges for cao vit gibbons when feeding and foraging, including abrupt changes in altitude and gradient, a dense understory, small trees with dense branches, small supports, and potentially increased predation risk by feeding on lower trees (see Fei et al. 2012). In addition, food availability showed


1039

considerable seasonality in the habitat of cao vit gibbon. Consequently, gibbons shifted their monthly diet in response to food availability (Fan et al. 2012). Cao vit gibbons do not use rocky outcrops in the way limestone langurs do (Fan et al. 2013a), but the possible impact of karst habitat on gibbon feeding posture has not been investigated. We here provide a general description of canopy use by cao vit gibbons during feeding, including canopy height and support use. We then test the following predictions: 1) If postures are mainly constrained by anatomy (Garber and Pruetz 1995), the dominant feeding posture of cao vit gibbons should be suspension like other gibbons (Chivers 1974; Fleagle 1976; Gittins 1983; Srikosamatara 1984), especially when feeding on nonfig fruit and figs when compared to leaves and buds (Fleagle 1976). 2) Given that suspensory feeding has been tied to fruit feeding in gibbons, we predict that gibbons will suspend more in months when figs and nonfig fruit make up a bigger part of their diet. Given the hypothesized constraints imposed by body size, we further predict 3) adult gibbons will use larger supports than juveniles when feeding on the same food types (Cant 1992); 4) adults will suspend more often than juveniles when feeding on the same food types and using the same sized supports (Cant 1992); 5) individuals of the same age–sex class will increase the use of suspensory feeding postures when feeding on smaller supports compared to larger ones (Cant 1992).

Methods Study Site and Population We conducted this study in Bangliang Gibbon Nature Reserve (22°55′N, 106°29–30′E) in Jingxi County, Guangxi, China. Together with the neighboring Cao Vit Gibbon Conservation Area in Vietnam, they sustain the only known population of cao vit gibbons (about 110 individuals; Le et al. 2008). The karst limestone area inhabited by gibbons is characterized by steep-sloped, sharp-peaked mountains and the forest is dominated by tropical monsoon forest that has been degraded by selective logging, fuelwood collection, charcoal making, and agriculture encroachment (Fan et al. 2011). We studied feeding posture and support use in relation to food class in three cao vit gibbon groups between August 2008 and December 2009. During the study period, each group had one adult male, two adult females, one or two large juveniles, none to two small juveniles, and one or two infants (Fan et al. 2015). We reported group dynamics of these groups in Fan et al. (2015). We were able to recognize all individuals except for the two similar-sized large juveniles and two similar-sized small juveniles in G4. During our study, G2 and G4 had ranges that straddled the border between China and Vietnam, and only G1 lived exclusively in China (Fan et al. 2010). In total, we observed G1 for 1119 h, G2 for 201 h, and G4 for 776 h during the study period (Fan et al. 2013a). Behavioral Observation Gibbons rarely feed at a single location for >5 min without changing posture or moving to a different support. As dense understory and precipitously steep and dangerous karst limestone landscape precluded tracking gibbons, we observed them from one of 12 observation posts using a spotting scope (Leica Apo-Televid 77 20-60) and binoculars

1040

H. Fei et al.

(Fan et al. 2010). To ensure sampling independence, which is crucial in the study of positional behavior (Dagosto 1994; Zhu et al. 2015), we used instantaneous scan sampling at 5-min intervals to record the feeding behavior (feeding posture, support size, food items being consumed) of every visible individual (Altmann 1974). P. Fan trained all observers for ≥1 mo to estimate tree height, support size, and posture mode (see later). Fleagle (1976), Gittins (1983), and Hunt et al. (1996) have described feeding postures of Hylobatidae. Following Gittins (1983, p. 135), we recorded five different feeding postures in the present study: 1) hanging or suspension: “animal suspended by its arms, weight taken by a support above the animal”; 2) sitting: “weight directed through ischial callosities to a support below the animal”; 3) lying: “weight directed through back or side to a support below the animal”; 4) standing: “weight directed through legs to a support below the animal”; 5) squatting: “the body weight is borne solely by the feet/foot, both hip and knee are strongly flexed. Neither forelimbs nor ischia bear substantial body weight.” Within each posture, we were able to distinguish several submodes such as those described by Hunt et al. (1996). The most common sitting posture was sitting with one arm holding a branch mainly for balance. During suspensory feeding, gibbons normally hang by one arm in an orthograde posture while using one or both feet to grasp branches for balance or to pull food-bearing branches close to them for eating. Forelimb–hind limb suspension was very rare and we observed hind limb suspension only during play bouts involving small juveniles and never observed it during feeding. We did not record these submodes mainly because of four reasons. First, we found it was difficult to observe all body parts of gibbons to describe the submodes such as sit-in or sit-out (Hunt et al. 1996). Second, we did not have enough time to describe the details because we also recorded other information such as diet, time budget, and spatial relationships of individuals (Fan et al. 2012). Third, we believe that recording submodes would increase variability between observers. Fourth, ignoring details of feeding postures did not significantly weaken the power of our data to test and answer our research questions. Gibbons normally used only one main support during feeding. We therefore recorded the size of the main support, for example, the branch gibbons were hanging beneath or sitting on. We classified substrates supporting the gibbons within one of four categories using the diameter of an adult gibbon’s finger (ca. 1.5–2 cm) or hind limb (ca. 10 cm) as a point of reference (Fan et al. 2013a). Definitions follow Gittins (1983, p. 135): “(i) bough > 10 cm, these did not bend or sway much under the weight of a gibbon; (ii) branch 2–10 cm, these did not bend very much but swayed under the weight of the gibbon; (iii) twig 10 m (Fan et al. 2013a). Data Analysis As a proxy for body mass, which was not measured in this study, we classified gibbons into the following five age–sex categories in descending order by ostensive weight: 1)


1041

adult females carrying dependent infants, 2) adult males and adult females not carrying infants, 3) large juveniles, 4) small juveniles, and 5) infants. We excluded data on infants from our analysis here because, by definition, they were not fully independent from their mothers and therefore their food choice and posture might be determined by their proximity or contact with their mothers rather than other factors. As sexual dimorphism in body weight in Nomascus is minimal (Ma and Wang 1986) and sex-based differences in positional behavior have not been documented in many of the primate species studied (Garber 2011), we combined adult males and adult females without dependent infants into a single category, while assuming both to be lighter than an adult female with a dependent infant. A number of studies of positional behavior (Bitty and McGraw 2007; Fan et al. 2013a; Grueter et al. 2013; McGraw 1998, 2000; Workman and Schmitt 2012; Youlatos 2002) used conventional χ2-test or G-test based on raw data; however, these methods have been criticized by Dagosto (1994) because most positional data were collected from a few individuals, i.e. small sample size, and successive samples were not independent. Dagosto (1994) suggested that one should analyze data for individuals or for blocks of observations rather than the raw observations to deal with the pooling fallacy. Given this recommendation, we collected data at 5-min intervals to avoid dependence of successive samples, and then pooled data for each individual. Gibbons in this population live in single male–bifemale groups. As a result, more adult females were available for observation than adult males. In total, we observed three adult males, six adult females, three large juveniles, and five small juveniles. Five females carried dependent infants during some portion of the study period, but were scored as adult females with infants only when actively carrying an infant while ingesting food. Data for two large juveniles and two small juveniles in G4 were pooled because we were unable to distinguish them in all scans. For summary statistics we calculated frequencies of food type, canopy use, support use, and feeding postures for each individual across the entire study period and then calculated the group mean for all individuals. Unless indicated, we used the Mann–Whitney U test for pairwise comparisons and Kruskal–Wallis test to compare multiple means simultaneously. To see if cao vit gibbons suspended more often when they consumed figs and nonfig fruit (prediction 1), we used the Kruskal–Wallis test with the proportion of suspensory feeding as the test (or dependent) variable and food type as the grouping variable. To compare monthly use of suspensory feeding posture to monthly diet (prediction 2), we calculated the monthly mean for both feeding posture and diet based on individuals and then used Spearman’s rank correlation test to examine the strength of the relationship between fig and nonfig fruit feeding and suspensory feeding posture. To determine if support use differed among age–sex classes (prediction 3) we calculated the proportion of overall feeding that each individual allocated to each support type, i.e., bough, branch, twig and liana, across the entire study and computed group means for each age–sex category. We then ran a Kruskal–Wallis test with support use as the test variable and age–sex category as the grouping variable. To control for the effect of diet on support choice we then tested the relationship for each food type, i.e., nonfig fruit, fruit, leaves, and buds, separately. We followed the same procedure to examine differences in suspensory feeding among age–sex groups (prediction 4), with the exception that individual means were based on suspensory feeding only. Finally, to see if individuals within each age–sex class alter their use of suspensory feeding based

1042

H. Fei et al.

on support size (prediction 5), we used the Kruskal–Wallis test with suspensory feeding as the test variable and support size as the grouping variable. As previously, we calculated the test statistic for each food type. We conducted all analyses using Excel 2007 and SPSS 13.0.

Ethical Note This research complied with Chinese law and with the Code of Best Practices for Field Primatology edited by the International Primatological Society. Our study was conducted under research permission from the Bangliang Nature Reserve and Guangxi Provincial Forestry Bureau.

Results Feeding Observations In total, we made 8387 observations of feeding behavior. Gibbons fed on nonfig fruit in 3043 records (36.3 %), figs in 1562 records (18.6 %), leaves in 1579 records (18.8 %), and buds in 1262 records (15.0 %). Feeding on flowers, insects, and other food types together accounted for 941 records (11.2 %) that were excluded from analysis of age–sex difference in feeding posture and support use because of the small sample size. Sample sizes of different age–sex categories varied, with 17.2 % being adult females carrying infants, 44.8 % being adult males and adult females not carrying infants, 13.2 % being large juveniles, and 24.8 % being small juveniles. Canopy Use and Support Size While Feeding The middle canopy (6–10 m) was the most frequently used by gibbons during feeding (60.3 % ± SD 7.5 %), while the lower (14.1 % ± SD 5.4 %) and upper canopy (25.6 % ± SD 8.4 %) was used less often. There was no difference in canopy use among age–sex classes (Kruskal–Wallis test: upper: χ2 = 1.878, df = 3, P = 0.598; middle: χ2 = 4.083, df = 3, P = 0.253; lower: χ2 = 1.789, df = 3, P = 0.617). On average during feeding bouts gibbons spent the most time on branches (59.4 % ± SD 14.0 %), followed by twigs (33.2 % ± SD 12.8 %), and lianas (6.0 % ± SD 2.8 %). Gibbons fed from boughs only rarely (1.4 % ± SD 1.3 %). We found no difference in the time that gibbons used branches (Kruskal–Wallis test χ2 = 2.748, df = 3, P = 0.432) or twigs (χ2 = 2.009, df = 3, P = 0.571) when they were feeding. Gibbons used lianas as supports more often when feeding on nonfig fruit (5.7 % ± SD 4.2 %) and leaves (15.1 % ± SD 14.9 %) than feeding on figs (2.8 % ± SD 2.8 %) and buds (3.3 % ± SD 4.8 %; χ2 = 21.798, df = 3, P < 0.001). Feeding Posture and Diet Suspension (55.5 % ± SD 12.2 %) and sitting (44.0 % ± SD 12.2 %) were the most commonly used feeding postures by cao vit gibbons. Other feeding postures, such as


1043

standing, squatting, and lying together accounted for a mere 0.5 % ± SD 0.6 %. Cao vit gibbons suspended more when they were feeding on figs (65.1 % ± SD 21.5 %) and nonfig fruit (58.2 % ± SD 11.6 %) than when feeding on leaves (47.9 % ± SD 12.7 %) and buds (49.1 % ± SD 16.6 %) (Kruskal–Wallis test: χ2 = 12.584, df = 3, P = 0.006). Seasonal Variation in Feeding Postures Cao vit gibbon feeding postures showed obvious seasonal variation (Fig. 1), but we found no correlations between the amount of nonfig fruit and figs in the diet each month and the proportion of time in suspensory feeding (Spearman correlation: nonfig fruit vs. hanging, r = 0.144, P = 0.581; figs vs. hanging, r = 0.339, P = 0.184). Age–Sex Class and Support Size Overall, adults and females carrying infants used branches more often (Kruskal–Wallis test: χ2 = 8.342, df = 3, P = 0.039) and used lianas less often (χ2 = 9.583, df = 3, P = 0.020) than large and small juveniles (Table I). Different age–sex categories showed no differences in the use of boughs or twigs (boughs: χ2 = 0.954, df = 3, P = 0.812; twigs: χ2 = 5.142, df = 3, P = 0.162). When broken down by food type (Fig. 2), adult gibbons spent significantly more time feeding on nonfig fruit from branches compared to juveniles (Kruskal–Wallis test: χ2 = 9.002, df = 3, P = 0.029); however, there were no differences among age–sex categories while feeding from boughs (χ2 = 6.058, df = 3, P = 0.087), twigs (χ2 = 7.080, df = 3, P = 0.069), and lianas (χ2 = 1.250, df = 3, P = 0.741). When feeding on leaves, adults spent more time feeding from branches (χ2 = 8.514, df = 3, P = 0.036) and less time feeding from lianas (χ2 = 9.835, df = 3, P = 0.020) compared to juveniles, but again feeding from boughs and twigs showed no differences among age–sex classes (bough: χ2 = 4.2, df = 3, P = 0.241; twig: χ2 = 4.419, df = 3, P = 0.220). Members of different age–sex classes showed no differences in support use when feeding on figs (bough: χ2 = 0.815, df = 3, P = 0.846; branch: χ2 = 6.494, df

Fig. 1 Seasonal variation in diet and feeding posture of cao vit gibbons in Bangliang between August 2008 and December 2009.

1044

H. Fei et al.

Table I Support use by different age–sex classes of cao vit gibbon in Bangliang between August 2008 and December 2009 Age–sex class Females carrying infant Adults Large juveniles Small juveniles

Bough (%)

Branch (%)

Twig (%)

Liana (%) 4.7

Mean

1.2

66.4

27.7

SD

1.2

10.9

11.2

1.9

Mean

1.4

64.8

29.0

4.9

SD

1.3

7.1

6.9

2.5

Mean

2.2

48.7

40.3

8.9

SD

2.9

6.1

8.6

0.4

Mean

1.3

43.8

46.1

8.8

SD

1.1

19.0

19.0

1.8

= 3, P = 0.09; twig: χ2 = 3.982, df = 3, P = 0.263; liana: χ2 = 2.022, df = 3, P = 0.568) or buds (bough: χ2 = 0.781, df = 3, P = 0.854; branch: χ2 = 5.65, df = 3, P = 0.13; twig: χ2 = 5.316, df = 3, P = 0.15; liana: χ2 = 5.019, df = 3, P = 0.17). Age Class and Suspensory Feeding Overall, age–sex classes showed no differences in the proportion of time spent in suspensory feeding (Kruskal–Wallis test: χ2 = 5.073, df = 3, P = 0.166). There were also no consistent differences among age–sex classes in time spent in suspensory feeding when controlling for food type or substrate size (Fig. 3, Table II). However, further comparison showed that adults suspended less often than small juveniles and

Fig. 2 Mean and SD percentages of support use by different age–sex classes of cao vit gibbon while feeding on different food types in Bangliang between August 2008 and December 2009. Asterisk (*) indicates a significant difference (Kruskal–Wallis Test: P < 0.05) within substrate class.


1045

Fig. 3 Mean and SD percentages of time spent in suspensory feeding by different age–sex classes of cao vit gibbon in Bangliang between August 2008 and December 2009, organized by food type and substrate size. No comparisons were significant.

large juveniles when they were feeding on figs from branches (Mann–Whitney U test: adults vs. small juveniles: Z = –2.315, P = 0.020; adults vs. large juveniles: Z = –2.121, P = 0.034). Suspensory Feeding and Support Size Except for large juveniles, all other age–sex classes increased suspensory feeding when they used flexible supports, twigs and lianas, compared to branches (Table III; Table II Comparisons of time spent in suspensory feeding by different age–sex classes of cao vit gibbons in Bangliang between August 2008 and December 2009, organized by food type and substrate size Food type

Supports

χ2

df

P

Nonfig fruit

Branch

2.052

3

0.562

Twig

0.666

3

0.881

Liana

4.383

3

0.223

Bough

2.133

2

0.344

Branch

7.019

3

0.071

Twig

3.311

3

0.346

Liana

4.479

3

0.214

Branch

2.241

3

0.524

Twig

0.542

3

0.909

Figs

Leaves

Buds

Liana

0.829

3

0.842

Branch

4.833

3

0.184

Twig

0.755

3

0.860

Liana

1.814

3

0.612

1046

H. Fei et al.

Table III Mean and SD percentages of suspensory feeding from different sized supports by different age–sex classes of cao vit gibbon in Bangliang between August 2008 and December 2009 Age–sex class Females carrying infant

Bough

Branch

Twig

Liana

Mean

13.9

38.5

76.8

81.9

SD

24.1

10.4

14.1

18.9

9.5

39.7

74.6

76.9

Adults

Mean SD

10.8

7.7

8.7

16.8

Large juveniles

Mean

35.0

46.6

82.3

70.1

SD

21.2

11.2

3.2

14.3

Small juveniles

Mean

14.1

54.1

84.1

70.1

SD

16.8

15.0

1.7

18.2

Kruskal–Wallis test: females carrying infant, χ2 = 10.059, df = 3, P = 0.018; adults, χ2 = 23.036, df = 3, P < 0.001; large juveniles: χ2 = 5.667, df = 3, P = 0.129; small juveniles: χ2 = 10.743, df = 3, P = 0.013). All age–sex classes spent more time suspending from twigs than from branches when broken down by food type, though in a few cases suspensory feeding from lianas was not greater (Fig. 4, Table IV).

Fig. 4 Mean and SD percentages of suspensory feeding from different sized supports by cao vit gibbon in Bangliang between August 2008 and December 2009, organized by food type and age–sex class. Asterisk (*) indicates a significant difference within age class (Kruskal–Wallis test: P < 0.05).


1047

Table IV Comparisons of time spent suspensory feeding from different sized supports by cao vit gibbon in Bangliang between August 2008 and December 2009, organized by food type and age–sex class Food type Nonfig fruit

Age–sex class

6.996

2

0.030

0.261

3

0.878

55.151

3