Consequences of global change for socially-mediated seed dispersal
Our framework illustrates that in many systems where fruit is a supplemental resource for animals (e.g., carnivores and folivores), subordinate individuals are likely to be the most effective seed-dispersal agents, and thus may be critical for understanding patterns in plant populations in the Anthropocene. First, since subordinate individuals have less restricted home ranges and are more likely to disperse beyond current conspecific territories (Aycrigg & Porter 1997; Henry et al. 2005; Dorning & Harris 2017; Kamleret al. 2019), subordinate individuals may be more likely to provide long-distance seed-dispersal events. Long-distance seed dispersal is critical for plant species range expansion to track changing climates (Dyer 1995; Cain et al. 2000; Corlett & Westcott 2013). Hence, subordinate individuals may facilitate range expansions that allow plant species track changing climates (Fig. 3A). Second, since subordinate individuals are more likely to utilize less-preferred habitats (Aycrigg & Porter 1997; Wittemyer et al.2007; Ward et al. 2018), and urban habitats are often demographic sinks (Vierling 2000; Lamb et al. 2017, 2020), subordinate individuals may be more likely to transport seeds into urban habitats, playing a potentially unappreciated role in promoting the spread and persistence of some plant populations in urban landscapes (Fig. 3B). Differences in movement across fragmented landscapes due to social status may also have important, yet unappreciated implications for the efficacy of conservation corridors (i.e., thin strips of habitat connecting otherwise isolated patches). While corridors have been shown to increase plant species diversity, including animal-dispersed plants (Damschen et al. 2006, 2019), there is limited knowledge of how an animal’s social status affects its propensity to move through corridors (but see Box 1 for discussion of work by Ford et al. (2017)). Since subordinate individuals typically forage in a larger number of patches in a landscape and are more likely to disperse from natal habitats, we predict that subordinate individuals may disproportionately contribute to seed movement through corridors and the consequent benefits of corridors to plant diversity (Fig. 3C). Initiatives to restore plant communities through corridor implementation may therefore require wildlife management that maintains or restores social structure in animal populations. Finally, if subordinate individuals have significant contributions to the quantity and spatial spread of seed dispersal, then subordinate individuals may facilitate the spread of invasive plants.
Our framework also highlights how environmental changes that modify animal social structure and behavior may have indirect consequences for plant populations through changes in seed dispersal effectiveness. For example, rank-dependent differences in diet and space use often only occur during seasons when the preferred resource is limited and can be monopolized (Pazol & Cords 2005; Wittemyer et al. 2007; Tsujiet al. 2020). Events that cause an overabundance of a preferred resource (e.g., masting or human subsidies) could therefore lead to a cryptic function loss of seed dispersal by subordinates (McConkey and O’Farrill 2015; Box 2). It has been shown that animal social network structure is relaxed near anthropogenic habitats (Belton et al.2018; Morrow et al. 2019), indicating that areas of high human activity may be hotspots of cryptic function loss of seed dispersal if high-quality foods are no longer monopolized by dominant individuals. While it is appreciated that animal species extinction can lead to catastrophic loss in plant biomass due to seed-dispersal limitation (Peres et al. 2016), we posit that shifts in animal social behavior may have similar deleterious effects on plant biomass. A study by McConkey and Drake (2006) on seed dispersal by flying foxes illustrated that seed dispersal services may be lost when animal population densities fall below a threshold for density-dependent behaviors that lead to seed dispersal (Box 2). Consequently, the functional role of animals as seed-dispersal agents can be lost in the early stages of species decline when social behaviors diminish, long before species become rare (McConkey & Drake 2006; McConkey & O’Farrill 2015, 2016). Finally, while it is known that harvesting frugivorous animals causes seed-dispersal limitation (Peres et al. 2016), we posit that selective harvesting of dominant individuals may also lead to seed-dispersal limitation (Box 1). By removing dominant males or matriarchs from populations, selective harvesting can disrupt social systems (Milner et al. 2007), which may lead to reductions in seed-dispersal efficacy by subordinate individuals if dominant individuals are no longer able to establish territories or defend preferred resources.
It is unclear if changes in the plant communities and the distribution of resources may indirectly affect seed dispersal by changing the identity and monopolizability of resources for social animals. For example, human disturbances may modify the types of resources available, which may alter consumer resource preferences (e.g., dominant individuals that used to monopolize fungus patches switch to consuming agricultural crops) and the quantity of seeds dispersed by individuals. Encounter rates with resources should also change the nature of resource preferences as well as the economics of resource defense. Global-change forces, such as introduced plant invasions, that reduce encounter rates with preferred resources may change the realization of preference as well as the opportunities to monopolize preferred resources. Moreover, it is possible that in cases where consumer preference does not change, the monopolizability of preferred resources is diminished, which may change the quantity of seeds dispersed by subordinate individuals. For example, recent work suggests that carnivores may abandon carcasses more quickly in the presence of human activity (Suraci et al. 2019), indicating that spatially clumped resources may be less defended by dominant individuals when located near human activity, possibly leading to greater access to preferred resources for subordinate individuals.