Ecology and social system of northern gibbons living in cold seasonal forests

Gibbons in China represent the northernmost margin of present day gibbon species distribution (around N25∘). Compared to tropical habitats, northern gibbon habitats are characterized by low temperatures and remarkable seasonal variation in fruit abundance. How gibbons adapt to their cold and seasonal habitats and what ecological factors affect their sociality are key questions for understanding their ecology and social system evolution, the elucidation of which will contribute to the conservation of these special populations/species. According to preliminary short-term studies, northern gibbons consume more leaves and use larger home ranges than tropical gibbons. Interestingly, some Nomascus groups consist of more than one adult female. However, these preliminary results are not well understood or incorporated into current socio-ecological theories regarding gibbon species. To better understand northern gibbons, our team has systematically studied three habituated groups of Nomascus concolor, three groups of N. nasutus, and two habituated groups of Hoolock tianxing since 2002. In this paper, we stress the challenges facing gibbons living in northern habitats and summarize their behavioral adaptations to their harsh environments. We also describe the northern gibbon social system and discuss the potential relationships between their ecology and sociality. Finally, we highlight future research questions related to northern gibbons in China.


INTRODUCTION
Gibbons (Hylobatidae) are small apes that live in tropical and subtropical forests in southeastern Asia (Bartlett, 2011). These apes are characterized by several interrelated features, including long arms for brachiation, terminal branch feeding, frugivory, pair-living, and territoriality (Brockelman, 2009). In tropical forests, gibbons rely heavily on ripe fruit pulp (Chivers, 1984), with fruits including figs comprising >60% of their annual diet, though siamang (Symphalangus syndactylus) is more folivorous (reviewed in Bartlett, 2011). Fruit is often distributed in small and defensible patches, resulting in food competition among females. Female gibbons are considered intolerant to each other and live solitarily (Bartlett, 2011;Kappeler et al., 2013;Leighton, 1987;Wrangham, 1987), with males unable to successfully defend ranges occupied by two or more females (Bartlett, 2009). Males may thus benefit from pair-living through 1) infanticide defense against non-resident, unmated males (Borries et al., 2011;Opie et al., 2013;van Schaik & Kappeler, 2003); 2) increased paternal certainty by mate guarding (Palombit, 1996(Palombit, , 1999; 3) limited opportunities for the female to evaluate potential sexual partners (Palombit, 1996(Palombit, , 1999; and 4) enhanced foraging efficiency by sharing knowledge with females (Brockelman, 2009) or by ensuring that resources are not unknowingly depleted by foraging apart (Terborgh, 1983). Females benefit from pair-living through food and territory defense, direct infant care provided by males (reported in siamangs, Lappan, 2008), and enhanced foraging efficiency (Brockelman, 2009). The mated pair use a small home range of on average 42 hm 2 , and actively defend 82% of the area (Bartlett, 2011) through direct intergroup conflicts (Bartlett, 2003) and regularly occurring loud morning songs (Geissmann, 2002;Mitani, 1985). Both sexes benefit from knowledge of this intensively-used territory and goal-directed foraging strategies (Brockelman, 2009). This, in turn, reinforces territory defense and small home range, and consequently often limits gibbon group size (Brockelman, 2009;Brockelman et al., 2014). Bartlett (2009) first integrated seasonality into territory defense. He argued that stable food abundance is a prerequisite for territory defense. However, most information on gibbons comes from only a few studied species (mainly Hylobates lar and S. syndactylus) living in lowland tropical forests (Bartlett, 2009(Bartlett, , 2011Brockelman et al., 1998;Chivers, 1974;Lappan, 2008;Palombit, 1996Palombit, , 1999. Historically, gibbons were distributed widely in China. As recently as 500 years ago, their melodic songs could be heard along The Three Gorges in central China and even further north (close to N35 • ) Zhou & Zhang, 2013). Although the known gibbon range has shrunk considerably from north to south, and is now restricted to Yunnan, Guangxi, and Hainan in southern China (Fan, 2017;Turvey et al., 2015;Zhou & Zhang, 2013), these sites still represent the northernmost margin of present day gibbon distribution (around N25 • , e.g., N. concolor in Wuliang Mountain: Jiang et al., 2006;H. tianxing in Gaoligongshan: Fan et al., 2011b, second only to the Hoolock gibbons in India and Myanmar (Groves, 1967). In this paper, we arbitrarily refer to the populations of N. concolor , N. leucogenys, N. hainanus, N. nasutus, and H. tianxing in China as "northern" gibbons ( Figure 1). It does not mean that these gibbons are distributed in northern areas but is used for comparison between these gibbon populations and gibbons living in "southern" tropical forests.
Based on short-term studies of unhabituated northern gibbons in China, early researchers reported on the differences in ecology and behavior compared to tropical gibbons. For example, northern gibbons were less frugivorous (N. leucogenys: Hu et al., 1990;N. concolor : Chen, 1995;Lan, 1993), occupied larger home ranges (N. hainanus: Liu et al., 1989;N. leucogenys: Hu et al., 1990;N. concolor : Jiang et al., 1994a;, and some groups included more than one adult female (N. concolor : Haimoff et al., 1986Haimoff et al., , 1987N. hainanus: Liu et al., 1989;Jiang et al., 1994aJiang et al., , 1994b. Although early studies reported obvious differences, detailed observations on these gibbons were hampered in China by a lack of habituated wild gibbons due to intense human hunting pressure and difficult terrain (Yin et al., 2016). Systematic, longitudinal research became possible only recently after several N. concolor and H. tianxing groups became habituated to observers in Mt. Wuliang and Mt. Gaoligong (Fan & Jiang, 2008;Fan et al., , 2009aFan et al., , 2013Yin et al., 2016). In 2006, N. nasutus was rediscovered in a small karst forest patch in Bangliang, Jingxi County, Guangxi (Chan et al., 2008). Researchers could observe these gibbons from fixed observation posts located on hill tops or forest gaps due to the steep karst terrain and low canopy height (Fan et al., 2011a. With accumulating data from different species living in different habitats, the picture of northern gibbon ecology and sociality, although still incomplete, is becoming clearer.

CHALLENGES FACING NORTHERN GIBBONS IN CHINA
The human population in China has increased rapidly and currently stands at over 1.3 billion (National Bureau of Statistics of China, http://www.stats.gov.cn/english/). The resulting increase in demand for cropland, together with the steel-making campaign of 1958, has led to the clearance of vast tracts of forest Zhang, 2000). Deforestation and habitat fragmentation caused by expansion of crop cultivation and plantations are the main causes of gibbon decline and extinction in China (Fan et al., 2011a(Fan et al., , 2011bFan, 2017;Jiang et al., 2006;Zhou et al., 2005), with large-scale hunting in the 1970s and 1980s also having a considerable impact. A total of 81 365 primate skins were collected by Chinese government officials between 1974 and 1980 (Quan et al., 1981). While many skins were from macaque species (Macaca mulatta and M. arctoides), gibbons were also hunted during this period (Fan, 2017;Quan et al., 1981). In the late 1980s and early 1990s, many Chinese gibbon populations had declined to such low levels that recovery was considered unlikely. By the late 1980s, the remaining populations of three species had decreased to less than 40 individuals (N. hainanus: Liu et al., 1989;H. lar : Yang et al., 1987;N. leucogenys: Hu et al., 1990), and until recently N. nasutus was considered extinct in China (Tan, 1985). To date, the numbers of H. lar and N. leucogenys have not recovered and both species are now assumed to be ecologically extinct in China Grueter et al., 2009). After more than 20 years of intensive conservation efforts, N. hainanus remains highly threatened with extinction (Fellowes et al., 2008;Zhou et al., 2005) and is the most endangered primate species in the world (Fan, 2017), with only one population of less than 30 individuals surviving in Bawangling. Fortunately, a small population of about 20 individuals of N. nasutus was rediscovered in Bangliang, Jingxi County, Guangxi in 2006(Chan et al., 2008. The remaining gibbon populations in China survive in montane or karst forests in several isolated mountain regions that are unsuitable for cultivation. Gibbon habitats in these northern mountain areas are much colder than tropical forests (Table 1). For example, the annual mean temperature in Dazhaizi, Mt. Wuliang is 16.6 • C (at 2 000 m a.s.l.:  and is even colder at Nankang (13.3 • C, at 2 200 m a.s.l.: Fan et al., 2013) and Banchang, Mt. Gaoligong (13.0 • C, at 2 300 m a.s.l.: Fan et al., unpublished data). At these sites, temperatures can drop below zero and forests are sometimes covered in snow Fan et al., 2013). Thus, the resident gibbons likely face increased thermoregulatory costs compared to tropical gibbon populations. Nakayama et al. (1971) reported that the energy expenditure of outdoor-living captive Japanese macaques (M. fuscata) at 5.2 • C was 2.5 times greater than that at 29.5 • C. Figure 1 Distribution of the gibbon species: Nomascus (7 species); Hoolock (3 species); Hylobates (9 species); Symphalangus (1 species) (According to Thinh et al. 2010a,b;Fan et al. 2017;IUCN 2017) Our gibbon research sites are also marked on the map.   Fruit productivity decreases with increasing latitude and altitude (Hanya & Aiba, 2010), and northern gibbons are known to consume less fruit than tropical gibbons (reviewed in . Fruiting seasonality also increases with latitude, and fruiting seasons are longer in the tropics and shorter toward the poles (Ting et al., 2008). At Mt. Ailao, a site supporting one of the two largest populations of N. concolor in China (Li et al., 2011), figs are absent and other fruit is only available for four months of the year (Chen, 1995;Sun et al., 2007). Haimoff et al. (1987) argued that Mt. Wuliang represents an extreme environment for gibbons due to the unavailability of fruit for much of the year, during which time gibbons must subsist on leaves (also see Fan et al., 2009a). In Mt. Gaoligong, Nankang, fruit availability drops close to zero between December and February . These findings raise the question of how northern gibbons adapt to their cold and seasonal habitats.
Patchily distributed plant species create variable food maps for northern gibbons. To obtain sufficient sustenance, the group of N. concolor in Mt. Wuliang  and H. tianxing in Mt.
Gaoligong (Zhang et al., 2014) shift their monthly home range according to the seasonal availability of food species. Although their monthly home ranges (N. concolor : 37±12 hm 2 ; H. tianxing: 28±15 hm 2 ) are comparable to the home ranges of tropical gibbons (avg. 34±9 hm 2 , calculated from Table 17.3 in Bartlett, 2011, excluding Nomascus), they occupy a larger total home range to satisfy their annual food requirements Zhang et al., 2014). Furthermore, they do not use their large home range evenly, but instead heavily utilize smaller areas with greater food resources . They also do not significantly increase their daily path length compared with tropical gibbons (northern gibbons in Table 1; tropical Hylobates gibbons: avg. 1 300 m), except for some siamang species (e.g., S. syndactylus: 0.8 km, reviewed in Bartlett, 2011).
In response to the overall low availability of fruit, early researchers suggested that northern gibbons ate more leaves and less fruit than tropical gibbons (Chen, 1995;Hu et al., 1990;Lan, 1993). This hypothesis was supported by several later studies with at least one-year duration. For example, the overall diet of N. concolor at Dazhaizi, Mt. Wuliang, was comprised of 46.5% leaves and 44.1% fruit (Fan et al., 2009a), with similar results reported for H. tianxing in Nankang, Mt. Gaoligong (49.1% fruit and 43.3% leaves, Fan et al., 2013) and N. nasutus in Bangliang (50% fruit and 35% leaves, Fan et al., 2012). Comparatively, the diets of tropical gibbons are typically comprised of 64% fruit and 25% leaves, except for the more folivorous siamangs (49% fruit and 38% leaves) (Bartlett, 2011). Although northern gibbons consume less fruit than tropical gibbons, they prefer fruit when it is available . When non-fig fruit is more abundant, H. tianxing feed on fewer leaves . Although Fan et al. (2009a) did not monitor fruit availability at Dazhaizi, Mt. Wuliang, the N. concolor group studied spent more time feeding on fruit in March of the second year when fruit was more available (Fan et al., 2009a). When fruit is less accessible or absent, northern gibbons switch their diet to leaves and buds (Fan et al., 2009a, accounting for 83% of the N. concolor diet in March (Fan et al., 2009a), and increase their feeding time .
When leaves dominant their diets, gibbons decrease their travel time and shorten their daily path length as leaves contain fewer available calories per gram than fruit (Raemaekers, 1978) and are more abundant and evenly distributed in the forest . This travel strategy has also been adopted by tropical gibbons (Chivers, 1974;Bartlett, 2009;Raemaekers, 1978), with two groups of H. lar found to travel on average 1 330 m (range from 620 m to 1 930 m, n=12 months) and 1 160 m (range from 720 m to 1 660 m, n=12 months), they traveled short distance when dominantly feeding on leaves (Bartlett, 2009). However, variation in both time spent traveling and daily path lengths are more dramatic in northern gibbons Zhang et al., 2014). For example, the daily path length of N. concolor in Mt. Wuliang varies monthly from 629 m to 2 356 m (n=14 months) and daily from 300 m to 3 144 m (n=66 days, . The mean daily path length of H. tianxing in Mt. Gaoligong varies monthly from 508 m to 1 926 m (n=14 months) and daily from 345 m to 2 606 m (n=63 days) (Zhang et al., 2014). Furthermore, travel time can show seasonal differences. For example, the H. tianxing group in Mt. Gaoligong has been observed to travel (on average) 3.40 h in June but 1.27 h in February . Seasonal impact on travel time has also been reported for the two groups of N. nasutus in Bangliang (group G1: from 3.92 h in June to 0.52 h in March; group G4: 4.92 h in June to 1.37 h in March; Fan et al., 2012).
Both sleeping and resting behaviors in northern gibbons are strongly influenced by cold temperatures, with more time spent inactive (sleeping and resting) during the 24 h cycle in winter . In Nankang, Mt. Gaoligong, the H. tianxing group spent 20.12 h inactive in February, whereas the two groups of N. nasutus in Bangliang spent 20.23 h (December) and 19.69 h (January) inactive.
Gibbons maintain an energy-conserving posture while sleeping or resting (sitting with hands clasping their legs: Bicca-Marques & Calegaro-Marques, 1998). Northern gibbons usually sit huddled together on tree tops in the sunshine during cold winter days, and even huddle and share sleeping sites on cold nights (Fan et al., unpublished data).
Gibbon groups with more than one adult female were first reported in N. hainanus in Delacour's book On the Indochinese gibbon (Xu et al., 1983), and later in N. concolor from Mt. Wuliang by Haimoff et al. (1986Haimoff et al. ( , 1987. Sheeran (1993) did not observe any multi-adult female groups in Mt. Wuliang or Mt. Ailao; based on brief observations in the same areas, Bleisch & Chen (1991) reported that some N. concolor groups included two females who duetted with the adult male, though they argued that this could be a sub-adult daughter singing with her mother. For a clear demonstration of polygyny, close observation of groups with two or more females (with young sired by the same male) is required (Bleisch & Chen, 1991). Jiang et al. (1994a suggested that the groups they observed were polygynous because the two females presented with offspring of the same body size. They thus proposed several models to explain the evolution of polygyny in N. concolor. First, the large home range occupied by N. concolor and N. hainanus might provide sufficient space for more individuals in one group. Second, mutual tolerance of females or "weak territoriality in females" might facilitate polygyny. Third, phylogeny might play a role in the evolution of polygyny in Nomascus gibbons. They also argued that N. concolor has a relatively large semi-pendulous scrotum (Groves, 1972), which should promote higher sperm production and, consequently, support a higher level of sexual activity. However, the above findings raise questions of why and how these gibbons occupy a large home range and females tolerate each other. These questions remained unanswered at the time as no gibbon groups had been habituated for detailed behavioral observation.
In 2006, Fan et al. (2006) reported that all five groups in a small subpopulation of N. concolor at Dazhaizi, Mt. Wuliang, included two adult females who gave birth or carried infants during the two-year research period. However, females in the same group bred at different times, perhaps to reduce reproductive competition. During the study, the researchers habituated one group and were the first to document the social behavior between two females living in the same group. They reported that the two females fed in the same tree, sang synchronously, groomed each other, and groomed and were groomed by the adult male, but never slept in the same place (Fan et al., 2006). They even shared meat when one female killed a flying squirrel ) and cooperated in evicting a "floating" (unmated) female from their territory (Fan et al., 2006). Further longitudinal observations showed that the two females repeatedly bred with the resident group male  and maintained similar spatial relationships to him (Fan & Jiang, 2010a). More recently, Guan et al. (2013) demonstrated that these groups consisted of two breeding females even after replacement of breeding males and females, and females actively cooperated in maintaining their social relationship, rather than co-existing merely through tolerance or avoidance.
The rediscovery of a N. nasutus population has provided further opportunity to study the evolution of polygyny in gibbons. This population lives in a degraded karst forest patch (Fan et al., 2011a), which is distinct from the habitat of any other gibbon population. Most groups in this population include two adult females. Fan et al. (2015) documented that N. nasutus benefits from living in larger groups because these groups can invade the home range of smaller gibbon groups and access more non-fig fruit, though larger groups also spend more time traveling. Female cooperation to evict unmated females has also been observed in this population but no group consisted of more than two breeding females, suggesting that territories defended by a male could not support a third female (Fan et al., 2015).
Long-term population monitoring can provide insight to better understand the sociality of northern gibbons. Although floating males in N. concolor populations have been observed regularly at Dazhaizi (Fan et al., unpublished data), females in the polygynous groups do not disperse and form new groups with these floating males. Two newly formed N. nasutus groups at Bangliang consisted of two adult females (Wei et al., 2017;Shao-Gan Wei, personal communication), though solitary males were observed frequently (Fan et al., 2010(Fan et al., , 2015. Similar results have been reported from a small population of N. hainanus; although solitary males were observed living in the site, one of the two newly formed groups consisted of two females (Bryant et al., 2016;Deng et al., 2017). These results reject the hypotheses that females are forced to live in their natal groups by delayed dispersal or are forced to immigrate into monogamous groups because suitable habitat and mates are not available (Bleisch & Chen, 1991;Liu et al., 1989). Consequently, Fan et al. (2015) proposed that polygyny is an evolutionarily stable mating system in certain habitats and for some species.
To secure their resource requirements, northern gibbons need to maintain a large territory in their heterogeneous habitats (Liu et al., 1989;Zhang et al., 2014). Within each territory, important food species occur in large patches, which can provide adequate food for two breeding females and their offspring, but perhaps not enough for three females Fan et al., 2015). The ability to digest substantial amounts of leaves in response to fruit scarcity (Fan et al., 2009a) enables these gibbons to minimize feeding competition within groups when fruit is not available and forage in large polygynous groups. Adult Nomascus males have a large scrotum and females in the same group breed at different times (N. hainanus: Zhou et al., 2008;N. concolor : Huang et al., 2013), which both reduce reproductive competition and enables polygyny (Path I in Figure 2). Southern gibbons live in more homogenous habitats (e.g., S. syndactylus : Chivers, 1974;H. lar : Bartlett, 2009;Brockelman et al., 2014), with smaller home ranges (Bartlett, 2011). Although they utilize large fruit trees (Bartlett, 2009), intensive interspecific competition depletes food resources, which may promote within-group competition (Marshall et al., 2009) (Path II in Figure 2). That does not mean food patch size in the tropical gibbon habitat is physically smaller than that for northern gibbons, but that on average, food availability is less.  Males benefit from forming polygynous groups through increased mating opportunities (N. hainanus: Zhou et al., 2008;N. concolor : Huang et al., 2013;N. nasutus: Fan et al., 2015). The cooperative female relationships documented above (N. concolor : Fan et al., 2006;Fan & Jiang, 2010a;Guan et al., 2013;N. nasutus: Fan et al., 2015) suggest that females in polygynous groups also benefit from larger group sizes. The potential benefits to females include resource access from successful intergroup encounters (Fan & Jiang, 2010a;Fan et al., 2015). Resident females in the same group appear to produce similar-aged offspring. After maturing, these offspring can disperse together, which might increase their survival rate and probability of establishing their own territory (Fan & Jiang, 2010a;Fan et al., 2015). In addition, females might benefit from improved inclusive fitness if females born within the same group disperse together (Fan et al., 2015). Even if females do not obtain any direct or indirect benefits from polygyny, bi-female groups could form if the cost of evicting the second female exceeds the cost of tolerance (Fan & Jiang, 2010a;Fan et al., 2015).
We need to note that H. tianxing living in northern montane forest shares many similarities in ecology with the three Nomascus species (reviewed above), but no stable group of more than one adult breeding female has been reported. A possible reason is that food patch size in the habitat of H. tianxing is smaller than that of Nomascus species (Path III in Figure 2). Given that stable polygynous groups have only been reported from Nomascus species living in different habitats, phylogeny may also play an important role in the evolution of polygyny in gibbons (Bryant et al., 2015;Di Fiore & Rendall, 1994;.

Standardized methods
Comparative research using the same or similar methods will be extremely useful for understanding gibbon ecology and social system evolution. First, the same methods should be applied to quantify forest structure, plant abundance and distribution, and food availability (Vogel & Janson, 2011). Second, the same or similar behavioral definitions and methodologies should be used to record behavior.
For example, several researchers distinguished foraging as "the group moved slowly, spread out over a broad front, feeding briefly from many of the trees as they passed through them" from a feeding bout when all group members fed in the same tree at the same time. Kappeler (1981) and Ahsan (2001) recorded foraging, but other researchers did not separate foraging from feeding (Bartlett, 2009;Fan et al., , 2012Fan et al., , 2013McConkey & Brockelman, 2011). Palombit (1997) used focal animal observations to record behavior, whereas other researchers have used scan sampling (Ahsan, 2001;Chivers, 1974;Fan et al., , 2012Fan et al., , 2013Gittins, 1982;Lappan, 2007a). Even for researchers using scan sampling, scan intervals often differ (e.g., 5 min: Bartlett, 2009;Fan et al., , 2012Fan et al., , 201310 min: Ahsan, 2001;Gittins, 1982;30 min: McConkey & Brockelman, 2011). Furthermore, comparisons would be strengthened if researchers used the same method to analyze data. Although most researchers present diet and time budget results as percentages (Ahsan, 2001;Bartlett, 2009;Fan et al., , 2009aGittins, 1982;McConkey & Brockelman, 2011 used time devoted to different activities and food categories because active time showed considerable disparity as a consequence of variations in day length at high latitude areas. It would strengthen the ability to compare results if all authors published their data as both percentages and actual times for both diet and time budgets. Finally, altitudinal range must be considered when calculating home range size and daily travel length because both variables can be impacted at steep study sites (Monterroso et al., 2013).

How do northern gibbons defend their large home range?
Gibbons are territorial and regularly and aggressively defend their home range, which is thought to enable exclusive access to a feeding territory (Bartlett, 2009;Brockelman, 2009). Northern gibbons occupy home ranges several times larger than those of tropical gibbons (N. concolor : N. nasutus: Fei et al., 2012). However, they do not travel any longer than gibbons that occupy smaller home ranges. Gibbons use their loud morning singing to defend their territories. However, northern gibbons do not increase singing to defend their larger home ranges (Fan et al., 2009b;Yin et al., 2016). In fact, hoolock gibbon groups appear to decrease their singing in areas with low group density (Yin et al., 2016). Therefore, the way in which northern gibbons defend their large territories still requires further investigation.

Social relationship in groups with different social systems
Intragroup social behavior in gibbons has received little attention (Bartlett, 2011).
From the limited information available, males have been observed grooming females more often than the reverse in the monogamous groups of S. syndactylus (Chivers, 1974;Palombit, 1996), H. lar (Palombit, 1996), and H. hoolock (Ahsan, 2001). Guan et al. (2013) studied the grooming patterns of two polygynous groups of N. concolor and found that females groomed males more often than the reverse, a pattern also observed in other polygynous primates (Ahsan, 2001). Studying intragroup social relationships in populations with different social systems will help to better understand the evolution and maintenance of social systems in gibbons.

Are northern gibbons adapted to digesting leaves?
Northern gibbons eat more leaves than do tropical gibbons (but see S. syndactylus in Bartlett, 2011, Table 17.2). However, how they resolve the chemical and mechanical challenges associated with leaf eating (Lambert, 1998), especially during cold winters with reduced or no fruit, remains unclear. Thus, comparative research on their digestion system, including their microbiome, will help clarify their ecology and social systems.

Kinship among females
Female kinship may be a factor affecting social relationships (Clutton-Brock & Lukas, 2012). Females living in polygynous groups may benefit through inclusive fitness if they are kin, and is highly possible due to the previous observation of two N. nasutus females dispersing together (Fan et al., 2010). However, as it is difficult to observe group formation or female displacement in the wild (Brockelman et al., 1998), the use of non-invasive genetic methods may help determine kin relationships among females (Lappan, 2007b).

How do northern gibbons find food within their large home ranges?
Do northern gibbons use spatial memory to find food resources in the same manner as tropical gibbons (Brockelman et al., 2014)? In northern habitats, gibbons appear to prefer ripe fruit pulp (Fan et al., 2009a. Scarce but predictable food resources led to the development of the "ecological intelligence hypothesis" (Milton et al., 1981), that is, primates, such as great apes, which rely on ephemeral and scattered fruits need larger ranges and have evolved advanced cognitive capacities that facilitate optimal food finding strategies. Does the "ecological intelligence hypothesis" explain the foraging behavior of gibbons? Long-term comparative research across gibbon species and between gibbons and large-bodied apes will help elucidate northern gibbon foraging behavior in their larger home ranges.

Comparison between foraging behavior
Given the different interspecies food competition at different sites, tree diameter at breast height (DBH) or crown size may not be accurate indicators of food availability. We highly recommend using feeding party size, feeding bout length, and number of feeding patches visited as indices to represent food availability, especially as gibbons often deplete their food patches. However, this prerequisite needs careful investigation. If these indices are validated, then detailed comparison among the foraging behaviors of gibbons living in different social systems can be made.

Life history and cultural behavior
Gibbons born in polygynous groups may develop better social skills for maintaining relationships due to their interactions with more social partners while growing up.
Therefore, ontogeny may also play a role in the maintenance of polygyny. Female gorillas familiar from immaturity usually exhibit friendly relationships in adulthood (Harcourt, 1979). Studying the development of social relationships in gibbons is difficult because of their long lifespans, and thus longitudinal data is required to answer the above question. Only with long-term study and the accumulation of data can we reveal unusual observations, such as "dance"  and "juggling" behaviors (Deng & Zhou, 2016). Current evidence implies that there is a great deal more to learn about gibbon species, and a wide variety of behavior remains to be explored in future studies.

COMPETING INTERESTS
The authors declare that they have no competing interests.