The scale of space-use by large mammals is mismatched with the scale of PAs, a challenge that is not limited to migratory species or large carnivores (Wikramanayake et al., 2008; Chundawat et al., 2016; Tucker et al., 2018). This is especially true in Asia, where PAs are orders of magnitude smaller than counterparts on other continents (Chowdhury et al., 2022). Because space-use is dictated by life history, we examined how these attributes might affect PA-use among Asian elephants. We found that PA-use strategies vary among individuals of the same sex and even social community within a given population. Observations also suggest nearly all adult males and at least half of adult females (i.e. ~3/4 of the population) may not be long-term residents. We consider these observations in more detail and the issues they raise for the reliance on PAs for accommodating elephants and other wide-ranging taxa on highly-modified landscapes.
Although males use the PA disproportionately when reproductive, 14% use it primarily for foraging, with a significant tendency for younger males to use it exclusively for foraging. Males following this strategy included a residential minority, but close to 90% also disperse at some point (Table 1). Those that use the PA during both states were more likely to be seen across years. The progression of musth stages (Fig. 2) suggests that more males switch from foraging to musth within the PA than vice versa. However, musth-only males behave differently. They tend to enter the study area during the early- or peak-phase and many stay through the late-phase (but never forage). Repeated sightings across years were rarest for musth-only bulls (Table 1), similar to observations from India (Keerthipriya et al., 2020). Interestingly, we observed a shift in strategy with age (Fig. 3A). Most musth-only males aged 20–40 were highly mobile and thus seen often when present (at which time they had low BSIs), but did not return annually (undefined BSIs). In contrast, males > 40 years appear to have settled into a more regular cycle of visitation during musth, evidencing significantly longer (inter-annual) BSIs relative to other classes. Thus, even fully sexually mature males seem to require decades of exploration before finally settling into a recurrent ranging pattern, perhaps through an extended process of contests eventually leading to temporal-partitioning with other similarly-aged males. However, we find no evidence of spatial partitioning suggestive of competitive exclusion among bulls following different strategies.
The mixture of strategies employed by males in this population demonstrates that remaining in a limited area during both motivational states is not a viable option despite the PA’s relative safety, availability of forage, and presence of many females. The median birth interval for females in this population is 6 years (de Silva et al., 2013), therefore males must leave the PA eventually due to the scarcity of breeding opportunities. Anecdotally, bulls implicated in human fatalities near the study area (3 individuals) were rarely seen in the PA and all exhibited signs of musth when captured (SdS personal observations). It is possible that some fatal encounters resulted from males ranging into unfamiliar areas during musth-induced range expansion. Nevertheless, there have been fewer than five such incidents during the course of this study. Given that males in this population seem largely transient, the rarity of such encounters is striking.
We expected that female PA-use would be related to social relationships because associations are defined in terms of spatiotemporal co-occurrence. However, the fission-fusion process introduces differences among individuals within the same putative community. We found that members of the same community were not necessarily seen at similar frequencies and that the influence of SRI on BSI was significant but weak. We also found a tendency for individuals from larger communities to have lower BSIs, and given that many were seen across multiple years, these individuals were likely to be more residential to the observation area. Earlier studies of this population found that on average, the number of companions an individual had was negatively correlated with the strength of her ties (de Silva et al., 2011a) and that the fission-fusion process undermines assertion of matriarchal dominance (de Silva et al. 2017). Thus, PA-use reflects some combination of individual decision-making and socially-associated movements. Community members may split up owing to local competition and constraints on group size (Nandini et al., 2018). When individuals compete, dominance hierarchies typically function to mitigate conflicts. Strong hierarchies, such as those observed in African savannah elephants, can mediate priority of access to resources, or even safe zones that are more central to PAs (Wittemyer et al., 2007). Because Asian elephants in this population do not exhibit dominance hierarchies, spatiotemporal avoidance may instead buffer against conflicts (de Silva et al., 2017). Individuals may move to less accessible areas within the PA, or outside it entirely (observed anecdotally), without their social companions.
A recent study in Malaysia suggested that agricultural landscapes might be prime habitats for elephants (de la Torre et al., 2021), while studies in Indonesia and Borneo document the propensity of elephants to exploit “degraded” areas associated with forest edges, often outside PAs (Rood et al., 2010; Evans et al., 2020). Although primary forests are often prioritized for biodiversity conservation (Morales-Hidalgo et al., 2015), elephants may prefer to forage in secondary and regenerating landscapes. But we must be extremely careful in interpreting and generalizing from these studies. First, PAs globally are biased towards steeper, higher terrain (Joppa & Pfaff, 2009). In Southeast Asia, PAs often contain of steep slopes, boxing in elephants when lowland valleys are rapidly converted to other land uses (Wilson et al., 2021). Use of “edge” habitat and agricultural areas may reflect the lack of adequate, low-risk, preferred lowland habitat. Much forest already been lost throughout Asia through conversion to intensive agriculture, contrasting with pre-colonial systems of management (Ellis et al., 2021). Thus many populations remaining today may have no choice but to use human-dominated landscapes, despite associated risks and costs (Goswami et al., 2014; de Silva & Leimgruber, 2019).
Many PAs in South Asia run counter to the trend, being frequently located around rivers and other water bodies, largely to protect their catchment zones, but also seasonally host significant populations of elephants and other wildlife. Their attraction to these PAs likely has to do with the presence of both water and monsoon-mediated forage (e.g. floodplains). Our observation that a substantial fraction of the population is non-residential reiterates the importance of permeability between PAs and wider landscapes. Individual-based studies of how other wildlife actually use PAs as well as in human-dominated landscapes (Kumar et al., 2010; Srinivasaiah et al., 2019), though logistically daunting, would further illuminate how strategies vary in response to anthropogenic changes.