Why sloths defecate on the ground: rejection of the mutualistic model

Introduction: Sloths are arboreal mammals that defecate on the ground, increasing the risk of predation. There are several hypotheses that try to explain why they undergo this risk. Objective: To critically review all the hypotheses and to propose a new hypothesis that is compatible with all known data. Methods: I verified the assumptions and implications of five hypotheses against the literature available February, 2021. Results: Previous hypotheses either lack reliable supporting data, or are contradicted by published data. Here I propose that defecation on the ground is an ancestral behavior that persists in all sloth species because there has not been enough natural selection against it. Conclusions: Current knowledge of sloth biology is compatible with the hypothesis that there has not been enough selective pressure for sloths to abandon defecation on the ground.

Sloths are arboreal mammals that harbor a complex community of organisms in their fur, including moths and algae (Aiello, 1985;Vaughan, Ramírez, Herrera, & Guries, 2007;Ramírez, Vaughan, Herrera, & Guries, 2011). Some sloths descend to defecate and urinate on the ground, increasing the risk of predation by ground animals. At least, five hypotheses have been proposed to explain this behavior: to fertilize trees, because feces are deposited at the tree base (Montgomery & Sunquist, 1975); to avoid predation, by covering the feces and reducing smells, as occurs in some other mammals (see Bailey, 1974;Liberg, 1980;Pauli et al., 2014); to communicate with chemical messages because direct interactions among individuals are rare, as implied by Chiarello (2008); to pick trace nutrients from the ground when they lick mud from their claws, as observed by Voirin, Kays, Wikelski, & Lowman (2013); and to increase moth populations, a mutualistic model (Voirin et al., 2013) for which Pauli et al. (2014) published some experimental support.
The current version of the mutualistic model states that sloths of the species B. variegatus lick and digest algae from their hair to obtain nutrients (algae might be fertilized by sloth fur moths, Voirin et al., 2013;Pauli et al., 2014). This model proposes that sloths defecate on the ground, and cover their feces to benefit moth larvae; these larvae develop inside fecal pellets in the "latrines". Pauli et al. (2014) add that B. variegatus descends to the ground, with higher risk of falling prey to predators. According to those authors, another species, C. hoffmanii, is less affected by predation because it defecates from the relative safety of branches, and has less algae and moths.
Here I summarize scientific reports that contradict the mutualistic hypothesis and propose a different explanation for why some sloths defecate on the ground.

The mutualistic model is contradicted by the scientific literature:
The key aspect of the mutualistic hypothesis is that sloths lick and digest the algae from their hair to obtain nutrients. However, Bradypus variegatus do not lick themselves, or each other (Aiello, 1985). Additionally, no noticeable algal remains are found in their digestive systems (Dünner & Pastor, 2017) and they cannot reach most of their own fur because of their short tongues and necks (Dünner & Pastor, 2017). Pauli et al. (2014) reported that 83 % of their B. variegatus digesta samples did not have any algal remains, but explained this absence, which is lethal to the mutualistic hypothesis, as the result of rapid digestion. The same algae may also grow on the leaves that sloths eat (Suutari et al., 2010) and this can explain why Pauli et al. (2014) detected some algal remains in a few stomachs. Furthermore, the test with cow ruminal inoculum, used to propose that sloths can digest fur algae, does not take into account differences in cow and sloth digestive systems (Clauss, 2004;Dünner & Pastor, 2017).
Secondary aspects of the mutualistic model lacking evidence in the scientific literature: Like Aiello (1985), I could not find any reports that sloth algae are fertilized by sloth moths, either through any secretions, or through their corpses as proposed by Pauli et al. (2014). Pauli et al. (2014) also stated that B. variegatus, which descends to the ground, suffers more predation than C. hoffmanii, which often drops feces from the canopy. However, a study in Costa Rica reported the opposite: 5.5 times less predation of B. variegatus than of C. hoffmanni (see Table 2 in .

Recommendation for new studies:
The following hypotheses could be studied in captive sloths: that algae are fertilized by the sloth fur moths (by keeping algae in cultures with and without the moths); that algal nutrients might be absorbed through the sloth skin (by chemical analysis of skin samples from areas with and without algae), and that moths feed on secretions from the sloth skin (by keeping moths in small containers attached to living sloths, with bottoms with or without holes allowing contact with the fur).

New hypothesis:
The new hypothesis that I propose here is that sloths defecate on the ground because they maintain their ancestral defecation behavior and there has been no selective pressure for them to defecate from the canopy.

Supporting references for the new hypothesis:
The six extant species of sloths are the only surviving descendants of a much larger group of mammals that defecated on the ground (Slater et al., 2016;Hunt & Lucas, 2018). Both families adapted independently to life on trees, and instead of convergently evolving ground defecation to increase their moth populations as proposed by, among others, Pauli et al. (2014); or by habit and to obtain additional nutrients, as proposed by Voirin et al. (2013); I propose that they retained the ancestral behavior of ground defecation in all species, as reported in the literature (Sunquist & Montgomery, 1973;Waage & Best, 1985;Hayssen, 2009;Hayssen, 2011;Slater et al., 2016;Dünner & Pastor, 2017), because there was no natural selection pressure (significant predation pressure or other) to stop defecating in the ground.
Maybe we should pay more attention to B. tridactylus, which sometimes defecates from the canopy (Waage & Best, 1985;Hayssen, 2009), and may be in the process of evolving a more fully arboreal life.