SPECIES SURVIVAL COMMISSION

TAPIR SPECIALIST GROUP

Tapirs:
Status Survey and Conservation Action Plan
Published 1997

---

Tapirs as Seed Dispersers and Predators

Fábio Olmos

R. Antonio F. Gandra 182
São Vicente, SP, 11390-250 Brazil



Abstract

Tapirs have a fair to high amount of fruit in their diets and are both seed dispersers and seed predators. Although more quantifed information is lacking, current evidence suggests tapirs are important as key seed dispersers for some groups or even communities of plants and are able to disperse species primarily adapted to other dispersers like birds and wind. Nevertheless, acting as seed predators, tapirs may have an impact on the seed crop of some plant species. The relationship between tapirs and seeds needs more research, including basic aspects of percent viabilty of dispersed seeds, distance of dispersion, and the fate of dispersed seeds.


Background

Tapirs have been eating leaves, twigs, and fruits from the world's forests perhaps longer than the existence of some of the plant families they eat today. Such a long history has probably caused relationships to evolve between these animals and their food plants.

In Mesoamerica and South America tapirs are the largest extant forest-dwelling mammals since humans caused the demise of the Pleistocene large mammal fauna (see Diamond 1989, Burney 1993). As Janzen (1982b) noted, tapirs are probable surrogate dispersers for large-fruited species which were once dispersed by the Pleistocene (from 1.5 million to 12,000 years B.P.) megafauna (Janzen and Martin 1982), and may be the key species to assure the survival of these plants. On the other hand, many seeds do not survive the masticatory and digestive apparatus of tapirs.

The study of the role of tapirs as seed predators and dispersers is fairly recent (Janzen 1981, 1982b), and we are still beginning to understand the interactions of tapirs and their food plants. Here I summarize the available information on the subject.


Tapirs as herbivores

Tapirs are browsers/frugivores (Bodmer 1990a), feeding on a diverse array of leaves, undergrowth and forest-edge herbs, shrubs and saplings, and fruit (Medway 1974, Terwilliger 1978, Williams and Petrides 1980, Janzen 1982a, Fragoso 1983, Williams 1984, Bodmer 1990b, Rodrigues et al. 1993, Naranjo 1995, Downer 1996). In addition, lowland tapir (Tapirus terrestris) feed on fair amounts of grasses and aquatic plants in Amazonian flood-plain forests and in the Brazilian Pantanal (Bodmer 1990c, pers. obs.) and on bamboo leaves and twigs in the Atlantic forest of eastern Brazil (Rodrigues et al. 1993, pers . obs.) . Mountain tapirs (Tapirus pinchaque) eat considerable amounts of grasses, bamboo, sedges, and bromeliads in their high altitude habitats (Downer 1996).

Although there is no record of tapirs actively looking for and taking animal food, Downer (1996) has found arthropod remains (coleoptera elytra = beetle wings) in mountain tapir dung, perhaps accidentally ingested.

The factors influencing the choice of some food plants by tapirs seem to be complex and are not well understood. Apparently both individual idiosyncrasies and plant defenses are involved, as young leaves and shoots are often preferred (Terwilliger 1978, Williams and Petrides 1980, Downer 1996). Captive tapirs are susceptible to rectal prolapsus, attributable to the ingestion of coarse and indigestible food (Crandall 1964, Deutsch and Puglia 1988) and this may be related to their selective feeding. Downer (1996) found that nitrogen-fixing plants and silica-rich horsetails are a significant component of mountain tapir diet, so it is quite possible that tapirs select food plants for their nutrient and compound content.

Like other Perissodactyla, tapirs are hind-gut fermenters with an enlarged cecum. Although large hindgut fermenters are considered to cope best with high-fiber, low-quality forage (Demment and Van Soest 1985, Bodmer 1990b), the selective browsing shown by all tapirs, and the fairly large percentage of fruit in their diet shown by some studies (Williams and Petrides 1980, 1984, Bodmer 1990b, 1990c) suggest tapirs do select for high-quality food when available. Indeed fruit patches are actively searched out and determine foraging patterns of tapirs (Bodmer 1990b, Naranjo 1995, Downer 1996).

There is variation of the relative importance of fruits in the diets of tapirs, both among species and among habitats. However, because different methods have been used to evaluate tapir diets it is only possible to qualitatively compare tapir fruit consumption. Fruit accounted for 33% of lowland tapir diet in Amazonian Peru (Bodmer 1990b), while two out of 15 dung piles had fruit remains in an eastern Brazilian forest site (pers. obs.). Large amounts of fruit were eaten by Baird's tapir (Tapirus bairdii) during the dry season in a northwestern Costa Rican site (Williams 1984), but were a minor item in the diet of the same species at Barro Colorado Island (Terwilliger 1978). At a site in southwestern Costa Rica at least 23 species of fruit comprised 3.8% of Baird's tapir diet during the rainy season compared to 12% during the dry season (Naranjo 1995). Mountain tapirs actively seek the berries of three specis of Vaccinium and Pernettya (Ericaceae) in the Andean páramos during the dry season. Williams and Petrides (1980) found that the Malayan tapir (Tapirus indicus) both looks for and includes a significant amount of fruit in its diet, remains being present in all but one of the dung samples he examined. Khan (this book) points out that fruit makes up 8.1% of the Malayan tapir's diet in Thailand, but more quantitative information is needed on this species.

Contrary to ruminant herbivores, which kill most or all seeds they ingest (Bodmer 1989, 1991), seeds of several species can survive the tapir's digestive system (Janzen 1982a, 1982b, Williams 1984, Bodmer 1991, Downer 1992, 1996, Rodrigues et al. 1993). These seeds must overcome both the tapir's mastication and the fairly long passage time (4-23 days in a captive Baird's tapir, see Janzen 1981), which is longer for larger seeds and may initiate germination in the tapir's gut, killing the seed.


What fruits do tapirs eat?

Tapirs are known to look for and eat a diverse array of fruits (Table 1.1). Fruits eaten by tapirs range from small, soft figs and berries 1-3cm in diameter to the 5-12 cm diameter fruits of the gourd tree Crescentia alata, the general impression being that tapirs are opportunistic, eating most fruit which are readily accessible, but still looking for concentrations of favored fruit (Bodmer 1990b, Downer 1996).

Sweet-tasting and/or smelling fruit seem to be preferred both in captivity (Janzen 1982a) and in the wild (Janzen 1982b, Williams 1984, Bodmer 1991, Downer 1996), but bitter-tasting fruits like the wild nutmeg Virola oleifera, the laurel Ocotea catharinensis (adapted to dispersal by birds), or the legume Sweetia fruticosa (a wind-dispersed species) are also eaten in large amounts by lowland tapirs when available (Rodrigues et al. 1993, P. Martuscelli pers. comm., Brisola 1989). Also, hundreds of the astringent acorns of the guanacaste oak Quercus oleoides may be eaten in a single feeding bout by Baird's tapir (Williams 1984).

There is little information on the nutrient contents of fruits eaten by tapirs. The drupes of the palm Mauritia flexuosa, a staple for lowland tapirs in the Amazon, have pulp which is 53.2% fat, 43% carbohydrate, and 3.8% protein (Bodmer 1990b). The aril of the nutmeg Virola oleifera, also eaten by lowland tapirs, is 23% lipids, 1.72% protein, and 12% carbohydrate (per weight, Galetti et al. in press). Mountain tapirs feed on the oil-rich fruit of the quindean wax palm Ceroxylon quindiuense (Downer 1996), but most of the fruits eaten by tapirs do not seem to be especially rich in fat as the examples above. For example, the fruits of the palm Euterpe edulis, eaten by lowland tapirs, are 21% carbohydrate, 6% fat, and 2.3% protein (wet weight, while the ones of Cryptocaria moschata, eaten by the same species, are 12.7% carbohydrate, 1.2% protein, and only 0.63% fat (Galetti et al. in press).

Another category of fruits eaten by tapirs are the ones ingested with forage (Janzen 1983a, Malo and Suárez l995). Downer (1992, 1996) found seeds of an impressive array of grasses and herbs in the dung of mountain tapirs, that were ingested while browsing for leaves and stems, but in some instances the tapirs may have been selecting for the seeds (Downer in litt.). Viable grass seeds were also found in dung and cecal samples from lowland tapirs (Bodmer 1991). Both Janzen (1982b) and Williams (1984) found live seeds in Baird's tapir dung that were most likely ingested while feeding on leaves.

Size is important to determine if a seed will be swallowed, spat, or crushed while chewing. Seeds larger than 2cm are usually spat out or crushed (Janzen 1982a,b; Bodmer 1991). However, hundreds of 2-3cm long nuts of Spondias mobim were found by Williams (1984) in dung piles left by Baird's tapirs, while a dung pile from a lowland tapir had 100 seeds of Virola oleifera, each 2.2-2.4cm long (Rodrigues et al. 1993). Seeds usually spat out are swallowed at times (Janzen 1982a). It is possible that individual idiosyncrasies, so patent in captive tapirs, play a role in whether the seeds are ingested or spat out.

Tapirs as seed predators

Tapirs kill seeds by chewing. Seeds which are not small or hard enough are crushed by tapirs, sometimes in large numbers. Janzen (1982b) and Williams (1984) found tapir dung piles containing mostly the hulls of the oak Quercus oleoides, each pile with the remains of about 500 seeds. Both authors also found that Baird's tapirs chewed the soft seeds of Brosimun alicastrum (Moraceae), Mastichodendrum capiri, and Manilkara sapota (Sapotaceae), being effective predators of those species. Chewing was also found to damage the large, hard seeds of guapinol Hymenaea courbaril, which are mostly spat out after being stripped of the pulp, although some are chewed and swallowed by lowland tapirs (Brisola 1989).

Tapirs also kill seeds by digesting them. Vulnerability of seeds to digestion depends on the hardness of their coat, their size, and whether or not they have been damaged by chewing. Compared to foliage and small objects, which take 1-3 days to pass through the digestive tract of a tapir, large seeds may spend more than 10 days in the tapir's gut, increasing the chances of germinating there and of being killed (Janzen 1981). Also, a hard coat helps to protect the seed both from chewing and digestion, provided it has not been damaged.

Janzen (1981) demonstrated that a Baird's tapir was able to kill 78% of the seeds of guanacaste Enterolobium cyclocarpum and 100% of the seeds of Carao Cassia emarginata fed to it, despite the hardness of their seed coats, apparently because they germinated inside the tapir gut. Nevertheless, Williams (1984) had different results from a study of wild Baird's tapirs, finding a fair amount of live carao seeds in dung piles, so the process is not always efficient and it may vary in different circumstances. Brisola (1989) found lowland tapirs to be predators of the seeds of the legumes Hymenaea courabaril and Sweetia fruticosa, and to damage (and perhaps kill) over 50% of the seeds of some laurel species it fed on.

Tapirs seem to partially digest seeds. Janzen (1982b) shows that there were about equal numbers of live and digested (as seen by their empty seed coats) seeds of the legume Pithecelobium saman in the wild tapir dung piles he examined, while Williams (1984) points out that for 6 of the 33 species of fruit eaten by Baird's tapir, all seeds passed intact through the tapir. Bodmer (1991) found that lowland tapirs in Amazonian Peru damage fewer seeds than sympatric ungulates, 54% of seed samples he examined being intact. Interestingly, many (12.5%) of the seeds belonged to Sapotaceae, a group killed by Baird's tapir (see above).

Tapirs may kill large numbers of seeds at a time, depending on the seed characteristics and availability, perhaps destroying the entire seed crop of an isolated individual. The partial efficiency of tapirs in digesting seeds usually allows part of them to survive the digestive tract and be dispersed. However, when compared to sympatric ungulates like deer and peccaries, tapirs are far less efficient seed predators (Bodmer 1991).


Tapirs as seed dispersers

Tapirs may transport intact seeds both by swallowing them and defecating later, and by eating a fruit and spitting its seeds. Although the latter method precludes the possibility of dispersing seeds far from the parent plant, Bodmer et al. (1993) consider the spitting habits of the lowland tapir important for the regeneration and maintenance of stands of the colonial palm Mauritia flexuosa. Similarly, Naranjo (in litt.) indicates that Baird's tapirs are important for maintaining Raphia taedigera swamps in Costa Rica and Nicaragua.

Janzen (1981) noted that tapirs have the potential of generating very complex and remote seed shadows, because of the possibility of seeds being moved many kilometers, and of the spacing of seeds in the tapir's dung, even when eaten at the same time, thus maximizing their scattering. Also, tapir can disperse large numbers of live seeds at a time, which may reach hundreds or even thousands (Janzen 1982b, Williams 1984, Rodrigues et al. 1993). The live seedlings, sometimes in large numbers, found sprouting from tapir dung show that tapirs are successful in transporting live seeds. For example, of 205 species eaten, 86 (or 42%) germinated from mountain tapir feces according to a study conducted in Ecuador's Sangay National Park (Downer 1996).

An important characteristic of tapirs as seed dispersers is their propensity to defecate in water (Janzen 1981, Williams 1984, Bodmer 1991). Indeed, in one Costa Rican study 94% (n=136) of tapir feces were found in water (Naranjo in litt.). This habit may lead to the loss of seedlings due to excessive dampness, but may also put the seeds in a favorable environment for secondary dispersal and recolonization, usually with fewer seed predators (Janzen 1981, but see Goulding 1981 for seed predation by Amazon fish). Although preferring to defecate in water, tapirs also defecate in dry places never reached by flowing water, especially in hilly areas.

Live seeds respond differently to their passage through the tapir's digestive tract. Some species seem to be unaffected, showing delayed germination that allows further dispersion by abiotic or biotic dispersers (Williams 1984). This group is probably best adapted to cope with the uncertainty of where the tapir will defecate, delayed germination allowing them to be secondarily carried to a more favorable habitat. Among the species in this group are several Spondias species, whose large, live nuts are found in tapir dung (Williams 1984, Bodmer 1991), sometimes in large numbers. Spondias nuts have long water-immersion lives and a buoyant fibrous jacket, suggesting adaptations for secondary dispersal by water (Williams 1984).

Other species showed different germination rates when compared to seeds extracted from fruits by researchers. Williams (1984) found that seeds of the legume Pithecelobium saman germinated at a faster rate and higher percentage after passing through a tapir. The same was true for the seeds of the gourd tree Crescentia alata, a species adapted for dispersion by large mammals, and of Alibertia edulis (Rubiaceae), the fruits of which are eaten by a diverse set of vertebrates (Janzen 1982b).

Some species show a faster germination rate when ingested by tapir, probably due to scarification, coupled with a decreased number of germinating seeds. Rodrigues et al. (1993) found that 45% of seeds of the palm Euterpe edulis germinated almost simultaneously in the first month after being swallowed by a lowland tapir. Control seeds taken from palms had an 85% germination success, but most seeds germinated 2-3 months after seeding, with some germinating 5 months later (pers. obs.). The same apparently happens with seeds of the legume Prosopis juliflora swallowed by Baird's tapirs (Williams 1984). In contrast, Rodrigues et al. (1993) observed that seeds of the nutmeg Virola oleifera found in tapir dung germinated later and in greater numbers than control seeds.

As mentioned previously, tapirs are less efficient than ruminants in digesting seeds. For species that escape the tapir's mastication, such digestive incompetency is what makes tapirs both dispersers and predators of their seeds. Although a large percentage of swallowed seeds are killed by the tapir, with seed dispersal a large number of seeds is not always an advantage (Janzen 1981). Perhaps more important than the number of live seeds dispersed by tapirs is their fate.

Many of the seeds taken by tapirs come from fallen fruit found under the mother plants, where they are otherwise likely to be killed by seed predators like rodents and beetles, or suffer the effects of competition (Howe and Smallwood 1982). Tapirs may give a survival edge to at least some of the seeds because they are competitors of those more efficient predators and are able to remove large numbers of seeds at a time (Fragoso 1994). This is not limited to species adapted to mammalian dispersal, because tapirs feed on fruit with adaptations to other dispersal agents, such as birds (Silvius 1995). Tapirs also ingest seeds of grasses and sedges, successfully dispersing species with no special adaptations for endozoochory (Downer 1992, 1996).


Conclusion

Unfortunately, we do not know for certain how important tapirs are as seed dispersers, although the evidence points to a significant role. The fact that lowland tapirs are the most effective dispersers of the palm Maximilliana maripa (Fragoso 1994); that live seeds were found in all but one of the Malayan tapir dung piles analyzed by Williams and Petrides (1980); that 22 out of 33 species known to be eaten by Baird's tapir are dispersed by it (Williams 1984); and that mountain tapirs disperse 86 of 264 tracheophyte plant species in their habitat (Downer 1996) suggest a very important role. It must be considered that tapirs can (and do) disperse hundreds or even thousands of seeds far (perhaps kilometers) away from the parent plants, being long-distance dispersers.

Also, the finding of Downer (1992, 1996) that mountain tapirs disperse the seeds of many of the grasses, sedges, and herbs they feed on, and the possibility of a significant impact of tapir dung deposition on soil formation, especially around dung deposits and feces piles, point to the real possibility of tapirs having a keystone role in shaping their habitats, at least in some situations. In the high Andes, where active volcanoes sometimes cover extensive areas with cinders and ashes, mountain tapirs may be the most important seed dispersers triggering the successional process in those areas (Downer 1996).

It is possible that the mountain tapir is the single most important species in structuring its habitat, due to the large number of plants of the high Andean flora that are adapted to tapir dispersal (Downer 1996). Such a role affects other species. For example, mountain tapirs are considered among the main dispersers of the quindean wax palm Ceroxylum quindiuense, a species threatened by lack of regeneration of their stands, possibly due to the local extinction of mountain tapir (Downer 1996) and competition with introduced grasses (Collar et al. 1992). This decline, together with deforestation, is endangering other species such as the yellow-eared parrot Ognorhynchus icterotis, a wax palm specialist endemic to the high Andes of Ecuador and Colombia, now on the verge of extinction (Collar et al. 1992).

Regrettably there is no detailed study on the fate of seeds found in tapir dung or spat by them to help us determine the impact of tapirs as seed dispersers. Nevertheless, there is much evidence that tapirs do disperse seeds and that they are important for the survival of at least some of the species they feed on. Some, like the Spondias trees, quindean wax palm, and many high Andean species, rely upon tapirs as their primary natural dispersers. Others, like the gourd tree and some legumes, orphaned after the American megafauna vanished, rely upon tapirs as dispersers which enable them to persist. Moreover, tapirs feed on and disperse seeds of species which are primarily adapted to other dispersal agents, thus acting as surrogate dispersers. In this way tapirs are important, and probably even play a critical role in maintaining the biological diversity of the ecosystems and help shape plant communities. This effect is magnified by the fact that some of the species tapirs disperse, like the palms, are keystone species with a broad impact on their ecosystems. The loss of tapirs and of other large mammalian seed dispersers is likely to have important, perhaps irreversible, impacts on the dynamics of the forested ecosystems they live in (Redford 1992).

There are still large gaps in our knowledge about tapir seed dispersal, and important questions still need to be answered. What is the fate of live seeds in tapir dung? What is the impact of tapirs on the species they prey on? Are tapirs real competitors to more efficient seed predators? The answers to these questions are important topics for future research.

• Return to top of page
  
• Return to previous page: Introduction
  
• Table of Contents
  
• Where to obtain printed copies
  
• Home
  

CITATION:
Brooks, Daniel M.; Bodmer, Richard E.; Matola, Sharon (compilers). 1997. Tapirs - Status Survey and Conservation Action Plan. (English, Spanish, Portuguese.) IUCN/SSC Tapir Specialist Group. IUCN, Gland, Switzerland and Cambridge, UK. viii + 164 pp.
Online version: http://www.tapirback.com/tapirgal/iucn-ssc/tsg/action97/cover.htm

Copyright © 1997 International Union for Conservation of Nature and Natural Resources

---

CONTACTS
Chair: Patrícia Medici
Deputy Chair: Sheryl Todd
This page is hosted by The Tapir Gallery
Webmaster: tapir@tapirback.com

---