Introduction

The process of seed dispersal is of fundamental importance to conserving habitats and biodiversity (Fricke et al. 2017; McConkey et al. 2012). Primates are dominant frugivores and herbivores in many habitats and play important, possibly essential, roles in seed dispersal (Chanthorn et al. 2017; Corlett 2017). Asian primates are a diverse group, ranging from the 300-g exudate-feeding loris (Nycticebus borneanus) to the 50–100-kg frugivorous orangutan (Pongo species). This diversity is reflected in how much fruit they consume, what they do with the discarded seeds and, therefore, the seed dispersal roles they perform. Asian primates also differ in their vulnerability to disturbance (IUCN 2017) and they inhabit ecosystems that are currently more threatened by anthropogenic pressure than most other regions in the world (Almeida-Rocha et al. 2017; Sodhi et al. 2004). Hence, it is imperative we understand the functional roles performed by primates and how vulnerable these roles are under current rates of disturbance.

The objective of this review was to provide a multilayered account of seed dispersal by Asian primates. My aim was to document the information that is currently available and emphasize the most important gaps in our knowledge. I examine seed dispersal at different levels: 1) seed dispersal by functional groups of primates, 2) contribution of primates to seed dispersal at a community level, and 3) intrapopulation variation in seed dispersal roles. In addition I summarize our knowledge of 4) understudied processes related to seed dispersal by Asian primates and 5) effects of disturbance on seed dispersal by Asian primates.

Methods

I conducted an exhaustive literature review on all seed dispersal studies on Asian primates using Google Scholar and Scopus®. I used the search term “seed dispersal” in conjunction with the “genus” of each Asian primate, with the exception of the carnivorous tarsiers (Tarsiidae). Because name changes have occurred for some taxa, I also used the common name for the broad groups: orangutan, gibbon, macaque, langur, leaf-eating monkey, loris. I supplemented this information with studies reported in the gray literature. I include studies specifically addressing seed dispersal by primate taxa, as well as broader community studies. I broadened the search to include accounts of frugivory when information on seed dispersal could not be found.

To visually compare the contributions of different primate taxa I displayed their contributions on a Seed Dispersal Effectiveness Landscape (SDE; Schupp et al. 2010, 2017). An animal’s contribution to seed dispersal is typically evaluated according to “quantitative” and “qualitative” components (Schupp et al. 2010).“Quantity” is a measure of the amount of seeds dispersed, while “quality” is a measure of how likely a dispersed seed is to establish as an adult plant, taking into account seed handling, seed deposition, and germination (Schupp 1993); however, since measuring survival to adulthood is not possible for most studied plants “quality” is practically measured to an earlier dispersal stage. The SDE of an animal is the product of these quantitative and qualitative components, and can be visualized on an SDE Landscape (Schupp et al. 2010, 2017). Here, I define Quantity as “the percentage of fruit recorded in the diet.” I use “dispersal distance” to represent Quality since this parameter has significant influences on seed survival and seedling establishment (Swamy and Terborgh 2010; Wang and Smith 2002) and is the most consistently measured parameter in the available studies. For the quantity and quality values, I used the mean and standard deviation across available studies. I constructed isoclines that represent the SDE (quantity × quality) of the taxa. Since orangutans and macaques (Cercopithecinae) spit larger seeds, I include different size classes of seeds (less or greater than 10 mm in width) in the SDE landscapes; 10 mm overestimates the contributions of orangutan and macaques, but it provides the most conservative comparison with gibbons (Hylobatidae). Colobines (Colobinae) are included on the same landscape, but do not defecate seeds longer than 5 mm.

Understanding Seed Dispersal Within “Functional Groups”

Around 90 plant-eating primate species currently live in Asian habitats (IUCN 2017). The only primates not recorded to consume some fruits in this region are the carnivorous tarsiers. Documenting the seed dispersal roles of individual species is the first step in understanding primate–seed dispersal mutualisms, and this has been the traditional approach of most studies. Because primates can often be habituated and followed, these data are (usually) relatively easy to obtain for primates compared to other animal groups, yet we currently have seed dispersal data for only 11 species (about 12% of all species) (Table I) and dietary information for several more. However, species within families, or subfamilies, have similar physiologies and will behave similarly with respect to seed dispersal. Hence, a realistic approach to assessing the roles of primates—without performing studies on all species—is to assume that species within families, or subfamilies are functionally similar in their seed dispersal role and consider these higher taxonomic categories as a functional group (Dennis and Westcott 2006). The functional groups are the orangutans, gibbons, macaques, colobines, and lorises.

Table I Published seed dispersal studies for Asian primates arranged by functional group

More important than ensuring a large number of primate species have been studied is ensuring that each functional group has been studied across a representative range of habitats. The habitat and associated distribution of resources play a large role in determining the ranging and foraging behavior of the animals within it and, therefore, the range of fruits eaten and the subsequent distribution of the seeds (McConkey and O’Farrill 2015, 2016; Tsuji et al. 2013a). For example, the eastern hoolock gibbon (Hoolock leuconedys) that occupies montane forest in Yunnan, China consumes the lowest number of fruit species ever recorded for a gibbon species and it is folivorous during the coldest months (Fan et al. 2013). In comparison, the western hoolock gibbon (Hoolock hoolock) that inhabits Bangladesh is among the most frugivorous recorded, with 90% of its diet as fruits and figs (Hasan et al. 2005). Given that much more minor habitat changes have been shown to influence seed dispersal distances (McConkey and Chivers 2007; Phiphatsuwannachai et al. 2017), it might be misleading to generalize seed dispersal roles across such divergent habitats. At present, fewer than half the habitat types occupied by Asian primates have had any seed dispersal studies (Fig. 1), and no functional groups have had detailed seed dispersal studies conducted within even half the habitats occupied (as specified by the IUCN, 2017).

Fig. 1
figure 1

Habitats occupied by different primate functional groups in Asia, showing the number of frugivory or seed dispersal studies conducted. Studies arranged by a functional groups, and b habitat. Where seed dispersal studies have been completed, I assume frugivory papers are also available. Occupation of habitats was determined from primate listings on the IUCN Red List. Habitats are grouped into broad categories.

Hominidae

Given their large body sizes and wide ranges, orangutans have the potential to play important roles as seed dispersers (Corlett 2017), and this has generally been assumed, despite the minimal studies that have been conducted (Table I). Two orangutan species (Pongo abelii and Pongo pygmaeus) are listed as Critically Endangered (IUCN 2017), and a newly described species (Pongo tapanuliensis) is probably the world’s most endangered great ape (Nater et al. 2017). Our lack of an understanding of their seed dispersal roles could be a severe limitation to assessing the resilience of habitats following their decline. They may play significant roles in moving large-seeded fruits, which have few others dispersers (Corlett 2017)—especially in Bornean habitats that lack elephants (Elephas maximus) and rhinoceros (Rhinocerotidae). However, the feeding data that are available suggest that orangutans might play just as dominant roles as seed predators as they do seed dispersers (Fig. 2).

Fig. 2
figure 2

Plant families dispersed by the main functional groups of Asian primates. Figures show the number of genera in each family that are dispersed (swallowed and defecated), spat, dropped, destroyed, or eaten but the fate is unrecorded. Only plant families that have at least two genera recorded are included. If a plant genus had two dispersal modes, then a value of 0.5 was applied to each category. If the seeds were consistently destroyed and not dispersed by any primate group, the plant taxa are not shown; if they were dispersed by at least one primate group then records of seed eating are shown for the other groups.

Orangutans select for meal size, in terms of large crop size and high pulp or fruit weight, rather than other fruit attributes (Leighton 1993), and they respond spatially and temporally to increases in fruit abundance, by consuming more fruit or moving into areas where fruit is most available (Kanamori et al. 2010, 2017). However, they are inconsistent in their seed handling, often spitting larger seeds (>15 mm wide; Galdikas 1982; Mohd-Azlan et al. 2015) and they are frequently seed predators (Mohd-Azlan et al. 2015; Nakashima et al. 2008); often they spit, swallow, and destroy seeds of a single plant species (Mohd-Azlan et al. 2015). Dispersal distances have the potential to be long for swallowed (small) seeds (>1 km) given the large ranges occupied by orangutans (Corlett 2009), and defecated seeds retain the capability to germinate (Nielsen et al. 2011), but even spat seeds can probably be regularly dispersed away from parent crowns (up to 75 m; Corlett 1998; Galdikas 1982). The role of orangutans as seed dispersers and/or seed predators remains one of the largest gaps in our knowledge of primate seed dispersal in Asia, despite decades of research (Fig. 3).

Fig. 3
figure 3

Overview of the seed dispersal roles of the four main functional groups of fruit-eating primates in Asia (excluding lorises, which have no available seed dispersal information); strengths are indicated in black font, and weaknesses in gray.

A second, larger ape once occupied parts of Asia. Giganthipithecus blacki weighed up to 270 kg (three to five times heavier than an orangutan) and lived in northern Southeast Asia and southern China from the Pliocene to ca. 100,000 years ago (Corlett 2017). Stable isotope analysis suggested it had a vegetarian diet that included fruit, and that it was a forest dweller (Bocherens et al. 2017). Orangutans also occupied parts of the range of Giganthopithecus in prehistoric times (Corlett 2013), suggesting these regions are missing potentially dominant contributors to seed dispersal mutualisms.

Hylobatidae

The four genera of gibbons are probably among the most important seed dispersers in the habitats they occupy (Fig. 3) (Corlett 2017; McConkey 2009). Gibbons are generalist frugivores, consuming and dispersing a large array of taxa (Brockelman 2011; McConkey 2000; Suwanvecho et al. 2017). They exhibit preferences for certain fruit types, which are primarily medium-sized (6–30 g), with soft-juicy pulp, and are thin-skinned or with a rind (McConkey et al. 2002); in some habitats their preferred fruit are available over long time frames or fruit asynchronously (Dillis et al. 2015). However, gibbons consume fruits of diverse morphologies, including dehiscent fruits predominantly dispersed by hornbills (Bucerotidae) (Savini and Kanwatanakid-Savini 2011) and tough-skinned fruits more easily accessed by large animals (Corlett 2017; McConkey et al. 2015; Vogel et al. 2009), and appear to have no clear selection for fruit color (Corlett 2017; D’Agostino and Cunningham 2015). While gibbons consume and can disperse some of the largest fruits in the forest (McConkey et al. 2015), their capability is limited by the size of the seed that can be swallowed (a maximum of ca. 20 mm, but influenced by size of individual gibbons) (McConkey 2000) and by access to the fruit and pulp. This means that the very largest seeds—such as those of Mangifera—and very well-protected fruit—such as Durio—cannot be dispersed by gibbons (McConkey et al. 2002).

Gibbons are consistently effective dispersers as they swallow and defecate seeds of almost all species, and drop or chew seeds only very rarely (Fig. 2). Because they are highly territorial, dispersal distances are confined by the boundaries of their home range and are influenced by the shape of the home range, habitat occupied, and the behavior of the individuals (McConkey and Chivers 2007; Phiphatsuwannachai et al. 2017). Distances of up to 1300 m have been recorded for gibbons, but more frequently average 200–400 m, and the majority of seeds are dispersed beyond 100 m (McConkey and Chivers 2007; Phiphatsuwannachai et al. 2017; Hai et al. unpubl. data). Gibbons have traditionally been considered to be poor dispersers across habitats, but this can occur where individual territories encompass a mosaic of habitats (Phiphatsuwannachai et al. 2017). Given that gibbons usually cannot maintain populations in habitats in which the canopy is not continuous (Cheyne et al. 2013), their dispersal role is probably confined to relatively undisturbed habitats.

Cercopithecidae, Subfamily Cercopithecine

The family Cercopithecidae contains two subfamilies that exhibit very different seed dispersal capabilities, suggesting that the functional group for this family should be considered at the level of subfamily. The cercopithecines are represented solely by the genus Macaca (macaques) in Asia. The 21 species of macaques in Asia are spread across a very diverse range of habitats (Fig. 1), which is reflected in a very diverse range of plant species consumed and dispersed (Fig. 2) (Tsuji et al. 2013a) but, like gibbons, macaques prefer fruits with juicy-soft pulp (Sengupta and Radhakrishna 2015; Ungar 1995). There are seed dispersal studies for only three species, but these include studies in tropical, temperate, dry, and even urban habitats (Table I; Fig. 3). Macaques show consistency in seed treatment across these habitats; ca. 60–85% of feces contain small seeds (Table I), and the major exception was when the macaques were provisioned by humans (27%, Sengupta et al. 2015). In tropical regions the majority of seeds dispersed are spat after processing in the cheek pouches (59% and 70%; Lucas and Corlett 1998; Sengupta et al. 2014), while rates of seed spitting are often lower in temperate regions (Tsuji this issue), where seeds tend to be smaller (Moles et al. 2007). Macaques are also seed predators, often spitting, dispersing, and destroying seeds of a single plant species (Otani 2004). There appears to be no absolute threshold for seed sizes to be swallowed, but most species consistently swallow seeds <5 mm in diameter (Albert et al. 2014; Tsuji this issue) and inconsistently up to ca. 25 mm (Albert et al. 2013; Yumoto et al. 1998). The prevalence of seed spitting appears to depend more on how easy the pulp can be separated from the seed than seed size (Yumoto et al. 1998; Tsuji this issue). Terrestrial and semiterrestrial species can also disperse seeds by carrying the fruit in the hands (Albert et al. 2014).

Information on dispersal distances for macaques is scarce, considering that they are widely distributed and relatively common in some regions. Spat seeds frequently end up under parent crowns (McConkey and Brockelman 2011), but can reach distances of ca. 200–400 m (Albert et al. 2013; Tsujino and Yumoto 2009; Tsuji this issue). The actual distances achieved by seed spitting can be influenced by the size of the foraging group and the size of the fruit resource (which influences how long the macaques stay within the source) (Albert et al. 2014). Because macaques frequently forage in large groups, the sheer number of seeds dispersed at close distances to the parent tree can imply an important role for this primate in seed dispersal (McConkey and Brockelman 2011). Dispersal distances for defecated seeds have averaged 97–486 m, with maximums of 265–1300 m (Sengupta et al. 2014; Terakawa et al. 2009; Tsuji and Morimoto 2016; Tsuji this issue), but given the large home ranges occupied by many species (Albert et al. 2014), it seems likely that some populations might achieve greater dispersal distances.

Cercopithecidae: Subfamily Colobinae

Colobines (Colobinae: langurs, leaf monkeys and proboscis monkeys) have been considered as seed dispersers only very recently and the available data are very limited (Fig. 3) (Matsuda et al. 2013; Tsuji et al. 2017). They digest their food in a similar way to ruminants (Kay and Davies 1994) and consume more foliage, unripe fruits, and seeds than other Asian primates (Corlett 2017; Matsuda et al. 2014; Tsuji et al. 2013a). At present, only dispersal of small seeds (<5 mm in length) has been confirmed, and for three genera: Nasalis (Matsuda et al. 2013), Trachypithecus (Nguyen et al. 2013; Tsuji et al. 2017; L. Ong unpubl. data), and Presbytis (K. R. McConkey unpubl. data) (Table I, Fig. 2), but all genera have been recorded to consume “fruit” (e.g., Dela 2007; Erb et al. 2012; Grueter et al. 2009; Hoang et al. 2009; Kool 1993; Ma et al. 2017; Marshall et al. 2014; Matsuda et al. 2009; Ungar 1995; Zhou et al. 2006). These studies have limited usefulness for assessing seed dispersal, because ripe fruits and unripe fruits are frequently grouped together even though the seeds will have different fates. While only small seeds appear to be dispersed internally by colobines, larger-seeded fruits might be dispersed when they are carried away from the parent crown; we lack any information on this potential role but langurs (Semnopithecus entellus) were reported to drop seeds “close to” as well as under parent trees in a dry forest in Northern India (Prasad et al. 2004).

Lorisidae

Several species of loris have been noted to consume fruits, but they are usually considered obligate exudativores (Starr and Nekaris 2013), or their diet is dominated by insectivory (Radhakrishna and Kuamara 2010). Fruits can form as low as 0.3% of the diet (Nycticebus bengalensis: Swapna et al. 2009), but several species have ca. 20% of their diet as fruit (N. pygmaeus, 19%: Starr and Nekaris 2013; N. coucang, 22.5%: Wiens et al. 2006; Loris lydekkerianus, 24%: Radhakrishna and Kumara 2010). It is likely that they eat more fruit when their preferred food items become scarce (Radkakrishna and Kuamara 2010). Fruits of several plant species have been recorded in the diet of lorises (two species, Swapna et al. 2009; six species, Radhakrishna and Kuamara 2010; nine species, Wiens et al. 2006), with 19 species recorded in the scats of a single species (including broken seeds) (Wiens et al. 2006); however, some of these fruits are eaten while unripe (S. Radhakrishna pers. comm.), and it is unclear whether lorises can disperse viable seeds. Consumption of small-seeded plants, such as figs and Dillenia, suggest they could have minor contributions to the seed dispersal of some plants.

Comparing Seed Dispersal Roles Across Taxa

The four main functional groups of frugivorous primate (orangutans, gibbons, macaques, and colobines) show different contributions to seed dispersal, when presented on a Seed Dispersal Effectiveness Landscape (Fig. 4). More plant species are potentially dispersed by macaques than any other taxa (Fig. 2), which probably reflects their wide distributions and diversity of habitats that can be occupied (Fig. 1).

Fig. 4
figure 4

Comparison of functional groups of Asian primates on seed dispersal effectiveness (SDE) landscapes. Quantity is the percentage of fruit in the diet, and Quality is the dispersal distance. Where multiple studies have been performed the mean and standard error are shown. SDE landscapes are shown for small seeds, and large seeds, as orangutans and macaques exhibit different seed handling behavior for different sized seeds. Isoclines represent the SDE value of the taxa.

For small seeds (Fig. 4a), the contributions of macaques and the gibbons are clustered together, suggesting they might play similar dispersal roles for these plants (at least in terms of the simplified view presented here). If their roles do differ it might be in terms of seed handling /digestion and factors associated with microhabitat. Colobines are unlikely to be able to match the seed dispersal contributions of the more frugivorous taxa, where they occur sympatrically. Orangutans might offer a complementary service to macaques and gibbons, as they could disperse seeds consistently longer distances, but there are no published studies on dispersal distances for orangutans and the value given here is an estimate based on movements (Corlett 2009). For larger seeds (Fig. 4b), gibbons disperse seeds consistently longer distances than orangutans and macaques (which spit seeds of this size). Overall, this suggests that gibbons are consistently reliable seed dispersers, while we might be overestimating the role of orangutans and underestimating the role of macaques, based on information that is currently available.

Increasing Scales: The Importance of Primates in the Community

Studies of individual primate taxa are essential for understanding seed dispersal roles, but the importance of these taxa (or functional groups) cannot be evaluated until these roles are considered at the level of communities. Asian habitats still lack published community-wide network analyses, which would help assess the relative contributions of primates to seed dispersal mutualisms within communities. The most broadly studied seed dispersal communities in Asia are dry forests in India (Prasad 2010), seasonal evergreen forests in Thailand (e.g., Brodie et al. 2009; Chanthorn et al. 2017; Kitamura et al. 2002; McConkey and Brockelman 2011) and temperate forests in Japan (e.g., Masaki et al. 2012; Noma and Yumoto 1997). Community-wide information also exists for a mosaic of habitats in Kalimantan, Indonesia (e.g., Marshall et al. 2014), tropical semievergreen forest in Arunachal Pradesh, India (Datta and Rawat 2008), evergreen forests, Southern India (Ganesh and Davidar 2005), and dipterocarp forest in Peninsular Malaysia (L. Ong unpubl. data). These very different habitats exhibit interesting differences in the communities of fruit-eating animals, and it is unlikely primates play equivalent roles across habitats.

Wet Forests vs. Dry Forests

Primates play central roles in seed dispersal in wet forest systems, such as tropical rain forests (Chanthorn et al. 2017) and the temperate forest habitats of Japan (Masaki et al. 2012). In tropical rainforests, primates often occur at relatively high diversity and density, which is matched by a high diversity and abundance of medium–large fleshy fruited species (Kitamura et al. 2002). Fruits such as these, which are mostly unavailable to birds owing to their physical structure, are most often dispersed by primates (McConkey and Brockelman 2011; McConkey et al. 2014, 2015). In Thailand, primates generated species-rich patterns of seed rain around their preferred food species and this pattern was still prevalent in saplings; similar patterns were not observed in bird-dispersed fruits (Chanthorn et al. 2017). However, habitats that have both a high primate density and a high diversity of terrestrial dispersers have not yet had published accounts. In Japan, macaques dominated the much simpler communities in terms of their contribution to seed dispersal; they dispersed 23 species compared to 13 species by the bulbul Hypsipetes amaurotis, while all other birds dispersed fewer than 7 species (Noma and Yumoto 1997). Masaki et al. (2012) found that the traits of Japanese fruits had evolved to suit mammals and not migratory birds, confirming a central role for primates in these systems.

In contrast, the dry forest communities that have been studied in India are dominated by terrestrial herbivores such as deer (Cervidae) and elephants (Prasad 2010; Sekar et al. 2017; Sridhara et al. 2016). The primate communities are limited to two taxa: langurs and macaques and the other key dispersers include frugivorous birds (Prasad and Sukumar 2010; Ramaswami et al. 2016; Sekar and Sukumar 2013, 2015 Sengupta et al. 2014). Community studies in these regions have indicated that primates occupy a less central role in seed dispersal (Prasad and Sukumar 2010; Sekar and Sukumar 2013, 2015) than has been shown in wet systems (Albert et al. unpubl. data). The peat swamp forests of Kalimantan appear to be another region where the contribution of primates to seed dispersal might be limited; primate-dispersed species were the most dispersal limited while bird and bat species were the most common (Freund 2012). This might be expected given that volant animals have a clear advantage, compared to primates, in terms of mobility across a water-logged landscape.

Primates vs. Other Taxa

Birds and primates are often the most conspicuous seed dispersing animals, and can be the easiest to collect data for. Hence, many comparisons of broad animal groups have focused on these two groups and have usually shown significant diet overlap. In Arunachal Pradesh, 45% of fruits dispersed by primates were also dispersed by hornbills (Datta and Rawat 2008); gibbon and hornbill diet overlap was 31% in Kalimantan, Indonesia (Marshall et al. 2009), while in Sulawesi and Sumatra they shared 45% and 41% of diet species, respectively (Kinnaird and O’Brien 2005). Although these figures help us understand frugivory—the first stage of the seed dispersal process—they do not describe actual dispersal overlap within communities. Fruit-eating animals are often generalists and consume a wide range of species to ensure their dietary requirements are met (Suwanvecho et al. 2017), so a high overlap between birds and primates could be expected. However, the preferences of these groups can be very different and they are important as dispersers for different subsets of these fruits (Table II); the “bird” fruits tend to be dispersed by birds, while the “mammal” fruits tend to be dispersed by mammals. Hence, “dispersal overlap” needs to be measured in terms of the relative numbers of fruits each taxa takes, and the chances of the seeds surviving and not based solely on frugivory information.

Table II Plant-focused studies that compare primates with other seed dispersing animals

Diet and dispersal overlap between primates and other mammals might be more prevalent than with birds (Table II), but this has been considered more rarely. A good reason for the lack of comparisons until recently has been the sheer difficulty in obtaining information on these other animals, which tend to be nocturnal and are often terrestrial (and are therefore more difficult to approach). Camera traps have indicated that civets (Viverridae), bear (Ursidae), deer, and even elephants (all effective dispersers for some species) often consume similar species to some primate taxa and these are fruits not generally eaten by birds (Table II). Tree watches have also confirmed consumption of these fruits by squirrels (Sciuridae) (50% diet overlap with gibbons: Marshall et al. 2009), but squirrels are generally ineffective as seed dispersers (Corlett 2017; McConkey and Brockelman 2011; McConkey et al. 2014). Determining diet and dispersal overlap with bats has proven more difficult in this region because flying foxes (Pteropodidae)—which leave the most easily recognizable fruit remains—are often heavily hunted and/or persecuted (Aziz et al. 2015; Mildenstein et al. 2015) or presence of elephants and tigers prevent nocturnal field studies. Bats also prove difficult to capture on camera traps. However, the limited information we have about the diet of Asian fruit bats suggests overlap could be reasonable (Aziz et al. 2017; Boon and Corlett 1990; Hodgkison et al. 2003).

Understanding Redundant and Complementary Dispersal

While a network approach that assesses the links between fruits and frugivores can provide a picture of how important a primate population might be within a community in terms of seed dispersal, the complexity within individual roles is rarely considered. When two taxa consume and disperse the same plant species they might be redundant as seed dispersers (so that only one taxon is required to ensure the species can regenerate) or they might play complementary roles (so that regeneration would be negatively impacted if one taxon was removed) (McConkey and Brockelman 2011; Rother et al. 2016). Determining whether a species plays a redundant or complementary role is quite difficult, in practice, but efforts have been made to address this at both landscape and local levels.

Traditionally, the “best” seed dispersers were considered to be those that dispersed seeds across large distances, but both local and far seed dispersal are probably required for a plant species to regenerate (Nathan 2007; Schupp et al. 2010); local seed dispersal ensures regeneration of the immediate population, while dispersal across long distances helps promote genetic mixing and colonization of new areas. Hence, two taxa that consume the same fruit species but disperse seeds at different average distances could have important complementary roles, rather than the taxa with more constrained dispersal being redundant (McConkey and Brockelman 2011; Rother et al. 2016). Hornbills and gibbons play potentially complementary roles because gibbons have a better capacity for local dispersal, while hornbills move seeds over large areas and across habitat fragments (Savini and Kanwatanakid-Savini 2011).

Redundancy and complementarity can be investigated by focusing on the seed dispersal ecology of selected plant species (Table II). These studies emphasize the importance of moving beyond species-focused approaches to measuring seed dispersal. A focus on species can produce an incomplete picture of a primate’s importance and it is essential to evaluate their role in relation to other dispersal agents. For example, macaques are often considered to be ineffective, redundant dispersers because they spit most seeds close to parent trees (Lucas and Corlett 1998); yet, the sheer volume of seeds that a large macaque group can disperse made them the most important local seed disperser for Prunus javanica in Thailand, with gibbons and hornbills playing complementary roles as medium- to long-distance dispersers (McConkey and Brockelman 2011). For another plant species in this forest, macaques and gibbons played complementary roles in terms of plants selected to feed and their actual effectiveness was almost equivalent despite dispersing seeds in very different ways (Salacia chinensis: McConkey et al. 2014). Other studies have shown that the role of primates might be primarily as a back-up dispersal mechanism when more efficient, but also more sporadic dispersers, are unavailable (Dillenia indica and macaques: Sekar and Sukumar 2013; Garcinia bethamii and gibbons: McConkey et al. 2015). In contrast, gibbons had a most likely redundant role in the dispersal of Choreospondias axillaris despite being regular consumers of the fruit; muntjacs (Muntiacus muntjak) were the only disperser that consistently moved seeds to the best microhabitats (Brodie et al. 2009). Overall, these studies are too few to understand the roles of different primate species at a community level. However, they suggest we should not underestimate what we do not know and consider that at times short-range, high-volume seed dispersal might be as important as long-distance, low-volume seed dispersal, or that even good seed dispersers might not provide equivalent services across all plant species.

Decreasing Scales: Intrapopulation Variation

Behavior is Important to Seed Dispersal

Behavioral studies are integral to primatology, providing a wealth of information with which to interpret seed dispersal behavior; yet studies of seed dispersal usually incorporate only the bare minimum of available behavioral information. Understanding the motivations of primate frugivory and seed dispersal behavior would give us a broader understanding of why variation in seed shadows exists and allow us to make predictions about how seed dispersal roles might be altered by changes such as anthropogenic disturbance.

Social and antagonistic behaviors can have important influences on frugivory and ranging of animals, and could be included more in seed dispersal studies (McConkey and O’Farrill 2016). In Asia, these behaviors could influence the seed dispersal role of the highly social macaques. Macaques often occur in large groups but may forage in small subgroups or as part of the large group (Albert et al. 2014), and this dichotomy could have quite different impacts on seed dispersal roles. In large groups, hierarchies among individuals can force individuals to process fruit away from the parent tree, resulting in scattered seed shadows close to the parent tree crown (McConkey and Brockelman 2011). In contrast, when subgroups forage in scattered fruit resources they might collect a few fruits as they pass through, store them in their cheek pouches, and spit seeds while roaming to the next food resource, resulting in longer dispersal distances (McConkey et al. 2014). Interactions with conspecifics can also influence the seed dispersal behavior of primates in smaller groups, such as gibbons. Gibbons are highly territorial and their movement patterns are influenced by the density, location, and aggressiveness of surrounding groups (Mitani and Rodman 1979). This could have repercussions on food choice and seed shadows (Phiphatsuwannachai et al. 2017) and how they might be altered by disturbance and reductions in gibbon density (McConkey and O’Farrill 2016).

Individual animals within populations often make consistent differences in fruit choices (McConkey and O’Farrill 2016). This might be driven by personality or learning and means that the collective seed shadow generated by a group might be very different from that observed in individual animals. This has rarely been studied, but in Sumatra, individual female orangutans made different fruit choices, and these preferences might have been transmitted from mother to offspring by social learning (Hardus et al. 2013).

Variation Within Habitats

Within landscapes, species and even individual plants do not offer identical rewards, owing either to their nutritional content or their spatial distribution (Carlo and Morales 2008; Jordano 2000; Prasad and Sukumar 2010; Worman and Chapman 2005). Clumped resources tend to attract more frugivores than scattered resources (Carlo and Morales 2008; Prasad and Sukumar 2010), while plants of a single species can vary in nutritional content across individuals, and within individuals across time (Houle et al. 2007; Worman and Chapman 2005). Individual primates also differ in their nutrient requirements according, for example, to their age and reproductive status (Vogel et al. 2017). Pioneering work on nutrition has been conducted in Africa (e.g., Raubenheimer et al. 2015; Rothman et al. 2011), with far fewer studies in Asia (Vogel et al. 2017). We do know that spatial and temporal patterns of fruit selection can have a persistent effect on seed shadows of birds (Carlo and Morales 2008), ungulates (Prasad and Sukumar 2010), and bats (Chen et al. 2017); hence, understanding the motivation for what to eat and when to move is an important gap in our knowledge on the seed dispersal roles of Asian primates.

Understanding More of the Seed Dispersal Process

Seed Dispersal Beyond Endozoochory

There is a tendency among seed dispersal researchers to focus on frugivory and seed deposition, but truly understanding the function of a seed disperser requires a broader appreciation of what seed dispersal involves. Many (if not most) primates play multiple roles: as seed predators or seed dispersers, or they might have neutral roles. Primates can also act as seed dispersers by endozoochory (swallowing the seed) or by synzoochory (carrying the seed by hand or in the mouth) and we need to consider the joint impacts of all these roles. In Asia, this is particularly relevant to macaques and orangutans, which do not treat seeds consistently (Albert et al. 2014; Mohd-Azlan et al. 2015; Tarsisz 2016) (Figs. 2 and 3). They show four different types of seed handling behavior (predation, swallowing, spitting, carrying) and often treat single plant species in multiple ways (Albert et al. 2013; Mohd-Azlan et al. 2015; Sengupta et al. 2014; Tarsisz 2016). Macaques, and other cercopithecines, are unique among primates in the presence and use of cheek pouches (Albert et al. 2013) in which they can store fruit and process it slowly, before spitting out the seeds. This is possibly reminiscent of regurgitation in birds, and can result in short- to medium-range dispersal distances (depending on movement patterns) (Albert et al. 2013; Tsujino and Yumoto 2009). Macaques can have pivotal roles in seed dispersal via seed spitting (McConkey et al. 2015; Sengupta et al. 2014; Tsujino and Yumoto 2009). Orangutans also spit seeds but almost nothing is known of how effective this is as a seed dispersal method. Even carrying seeds can promote seed dispersal (Albert et al. 2013), especially for large-seeded fruits that have fewer alternative dispersers (Kitamura et al. 2002). This dispersal mode is available only to terrestrial or semiterrestrial primates because arboreal primates require their hands and feet to move through the canopy; hence, again it is the macaques and orangutans that might use this method most effectively.

Seed predation can have just as profound impacts on plants as seed dispersal (Genrich et al. 2017; Janzen 1971). Colobines are major seed predators in Asian forests (Ganesh and Davidar 2005; Hanya and Bernard 2013; Sun et al. 2007), but the impact of this function on the forests is almost unexplored. In Pasoh, Malaysia, leaf monkeys were the major predispersal predator of Shorea seeds and might play a critical role in the evolution of masting in these plants (Sun et al. 2007). Orangutans and macaques are both seed predators and seed dispersers, and the balance between this mutualism and antagonism is similarly unexplored, yet is potentially very important (Genrich et al. 2017; Otani and Shibata 2000). For example, orangutans eat dipterocarp seeds when non-dipterocarp fruits (which they can disperse) are not widely available (Curran and Leighton 2000), while in other cases they might preferentially target species for their seeds (Hayna and Bernard 2013).

Secondary Processes

Although secondary processes on dispersed seeds are recognized as very important (Culot et al. 2015; Janzen 1971), relatively few studies have assessed what happens to seeds dispersed by Asian primates (Enari and Sakamaki-Enari 2014; McConkey 2005a, b). Dispersed seeds might be removed by secondary dispersers, destroyed by seed predators, be attacked by pathogens, or germinate, and the likelihood of these different fates will be strongly species-specific. In a tropical forest in Thailand, rates of seed survival in macaque and gibbon-dispersed seeds ranged from 2% to 66% and 7% to 70%, respectively, depending on the plant species and where the seeds were dispersed to (McConkey and Brockelman 2011; McConkey et al. 2014, 2015). These secondary stages of seed dispersal can profoundly alter seed shadows (Culot et al. 2015); only 11% of seeds in gibbon seed shadows in dipterocarp forest germinated and almost all dungs were visited by seed predators or dispersers (McConkey 2005a). Primate feces can act as a strong attractant for seed-eating animals, particularly rodents (McConkey 2005b). Hence, understanding how seed spitting vs. seed swallowing impacts subsequent seed survival is an important topic to explore for primate species that handle seeds in different ways.

Secondary seed dispersal can occur through seed hoarding by rodents (Muridae and Sciuridae) or seed burial by dung beetles (Scarabaeidae). The function of dung beetles in removing seeds in primate feces has been studied intensively in the Neotropics (e.g., Andresen 2002; Culot et al. 2015), but Asia lacks studies on a comparative scale. In the heavy snowfall forests of Japan, between 28% and 40% of seeds defecated by macaques were incorporated into the seed bank by dung beetles with rates influenced by seed size (Enari and Sakamaki-Enari 2014) and some beetle species preferentially targeted macaque feces over those from other mammals (Enari et al. 2016). In a rain forest in Thailand, 10% of large seeds (length > 20 mm) in gibbon dungs were removed by dung beetles (Jadejaroen 2003). Similarly, seed hoarding by rodents is an important means of secondary seed dispersal in Neotropical forests (e.g., Forget et al. 2002; Jansen et al. 2012). Hoarding is also common for some rodent–plant interactions in Asia (e.g., Wang and Ives 2017; Xiao et al. 2005), and rodents are common visitors to primate dungs (McConkey 2005b), but the extent of seed hoarding from primate dungs is not known.

Choosing a Fruit

Fruit selection is another aspect of seed dispersal that has received little research attention. Selection can occur at differing scales—from choosing species, individual trees, and individual fruits within trees—and is influenced by multiple factors, including nutritional requirements; olfactory, visual, tactile, and possibly auditory, acuity; and behavioral and ecological constraints (Corlett 2011; Dominy et al. 2001), as well as the range of fruits actually available at any time (McConkey et al. 2002; Suwanvecho et al. 2017). Langurs select different fruit types than orangutans, macaques, and gibbons (Fig. 3), but even within the three monogastric primates differences in fruit selection were observed (Ungar 1995) and there has been little follow-up work. All Old World monkeys and apes have full trichromatic color vision, which promotes easier detection of green-yellow-red fruits (Dominy et al. 2001); this has possibly occurred at the expense of reduced olfactory ability (Gilad et al. 2007), but we know very little about the role other senses actually play in fruit selection. Both orangutans and gibbons have been shown to track fruit supplies; orangutan density alters with fruit availability (Marshall et al. 2009) and social transmission also influences fruit choices (Hardus et al. 2013), while gibbons make targeted visits to certain trees, presumably to check fruiting status (Asensio et al. 2011). Location of fruiting trees by macaques and langurs has not yet been assessed, but this has been studied in cercopithecines (baboons) in Africa (Noser and Byrne 2010).

Understanding Change

Asian primates are at high conservation risk and are among the least adaptable of the world’s primate species to the extreme changes that are affecting their habitats (Almeida-Rocha et al. 2017); this could have serious repercussions for the ecosystems they help to maintain. Most primate taxa already inhabit severely reduced distributions (although macaques have established introduced populations in some regions; Brodie et al. 2017) and reduced populations, and taxa are at varying degrees of threat from hunting, habitat loss, the pet trade, and “pest” control (Corlett 2007). All orangutan species are listed as Critically Endangered (IUCN 2017), along with 25% of gibbon species and species from all other functional groups (macaques, colobines, and lorises) (IUCN 2017). Although orangutans can maintain populations in degraded forests, the severe threat from hunting, pest control, and the pet trade (Corlett 2007) ensures they are unlikely to have long-term roles in seed dispersal in these habitats. Most gibbon species are listed as Endangered (69%) and their inability to maintain populations in degraded areas in which the canopy is discontinuous suggests their excellent seed dispersal capabilities are useful only in the increasingly limited intact forests. Macaque and colobine species are listed across the range of threat levels, but they also have species considered to be at low risk, and in some instances these species are commensal with humans (Sengupta et al. 2015; Tsuji 2011). These taxa, particularly macaques, are perhaps the most likely primates to maintain seed dispersal services in degraded areas (Albert et al. 2013).

The most serious conservation concern, in terms of seed dispersal, is what might happen to habitats in which primates are locally extinct. The quantity of seeds of Myrica rubra dispersed on Tanegashima Island, Japan—where Macaca fuscata are locally extinct—was 20 times less (all dispersed by bulbuls [Pycnonotidae]) than the quantity of seeds dispersed on Yakushima Island, where the macaque still occurs (Terakawa et al. 2008). Although primates are among the most hunted animals in Asian forests and have been extirpated or exist in reduced populations in many forests (Corlett 2007), this is the only study to specifically address the impact the loss of primates might have; however, evidence for significant effects on seed dispersal exists at a community level. Lambir Hills National Park in Sarawak, Malaysia is severely defaunated and has lost 50% of its primate species (and 22% of all mammals); overhunting has caused pervasive changes in the spatial structure of tree populations, with animal-dispersed species showing more clustering around adult trees than seeds dispersed abiotically (Harrison et al. 2013). Although these changes could not be attributed to specific animal species, the Bornean gibbon (Hylobates muelleri) was among the primate species lost (Harrison 2011). Similarly, the likelihood of extinction of a gibbon-dispersed species, Miliusa horsfieldii, was predicted to increase 10-fold if all dispersers were to become extinct from the Thailand forest (Caughlin et al. 2014); however, again, the contribution of gibbons compared to other dispersers (civets and bears) was not distinguished.

Many other factors can alter how primates contribute to habitat maintenance, but these have been studied in only a handful of primate species in Asia. Reductions in primate density, changes in resource availability and habitat structure, and alteration in community composition can all change the frugivory and seed dispersal behavior of primates (McConkey and O’Farrill 2015, 2016). Resource availability, for example, is a strong determinant of primate diet and ranging. Provisioning by humans has changed the movement patterns, intergroup competition, and seed dispersal role of at least two macaque species (José-Domínguez et al. 2015; Sengupta et al. 2015). Under provisioning, rhesus macaques (Macaca mulatta) ate less fruit, moved less, and deposited most seeds on the road where germination was impossible (Sengupta et al. 2015). Macaques might be particularly vulnerable to changes in seed dispersal behavior given the behavioral flexibility of most groups (Albert et al. 2013), and any habitat or resource change could potentially enhance or decrease their role (McConkey and O’Farrill 2016). Changes in community composition have also been shown to alter the role of macaques. In Singapore, macaques have shifted a trophic level, eating more fruit now than before, possibly as a response to a reduction in competitors for that fruit (Gibson 2011).

Conclusions

There are five functional groups of seed-dispersing primates in Asia. Gibbons are consistently effective dispersers, while colobines are predominantly seed predators, with orangutans and macaques playing dual roles (Fig. 3). Very little is known about the role of lorises, but they may play dual roles, or be predominantly seed predators. The importance of these roles is likely to be habitat specific, so it is essential that studies incorporate a range of habitat types and consider the seed dispersing “niche” of the primate within the wider frugivore community. Orangutans have the potential to act as long-distance seed dispersers for small to medium seeds, while gibbons are consistently medium-range dispersers for all swallowed seeds; dispersal distances achieved by macaques are extremely variable but they have the capacity for consuming large quantities of fruit. There are many gaps in our knowledge on primate seed dispersal in Asia; these include understanding the role of behavior in seed dispersal, fruit selection at different scales, postdispersal processes, and the combined impacts of dual functions (predators and dispersers). In the future, we must consider the role primates play at the level of the frugivore community and as a dynamic system that can be altered by natural and anthropogenic disturbances; only then can we determine how critical primates are for habitat persistence and the extent to which their roles are threatened by disturbance.