Defaunation increases the spatial clustering of lowland Western Amazonian tree communities

Abstract

Declines of large vertebrates in tropical forests may reduce dispersal of tree species that rely on them, and the resulting undispersed seedlings might suffer increased distance‐ and density‐dependent mortality. Consequently, extirpation of large vertebrates may alter the composition and spatial structure of plant communities and impair ecosystem functions like carbon storage. We analysed spatial patterns of tree recruitment within six forest plots along a defaunation gradient in western Amazonia. We divided recruits into two size cohorts (“saplings”: ≥1 m tall and <1 cm diameter at breast height [dbh], and juveniles, 1–2 cm dbh) and examined the spatial organisation of conspecific recruits within each cohort (within‐cohort) and around conspecific reproductive‐sized trees (between‐cohort). We used replicated spatial point pattern analysis to quantify relationships between recruit clustering and cohort, defaunation intensity, each tree species reliance on hunted dispersers and the interactions among these three covariates. Within‐cohort clustering of conspecific saplings increased with reliance of tree species on hunted dispersers, and this trend strengthened significantly as defaunation increased, probably because of reduced dispersal. Within‐cohort clustering of conspecifics declined from saplings to juveniles, suggesting density‐dependent mortality of saplings. However, the positive relationship between sapling clustering and defaunation did not lead to greater reductions in within‐cohort clustering during the sapling–juvenile transition, suggesting that higher conspecific densities did not translate into increased mortality. Instead, the increased spatial clustering associated with defaunation was retained for juvenile recruits. Between‐cohort clustering was unrelated to defaunation and did not change during the sapling–juvenile transition. Synthesis. Defaunation increased spatial aggregation of saplings of tree species reliant on hunted dispersers. The increase in sapling clustering did not increase density‐dependent thinning, and persisted into older recruit cohorts, suggesting that hunting may initiate long‐term spatial reorganisation of Amazonian tree communities. The lack of increased density‐dependent thinning indicates that reduced dispersal did not increase mortality of large‐vertebrate dispersed tree species that contribute disproportionately to forest biomass. We, therefore, caution against the fait accompli acceptance of the prediction by recent modelling studies that overhunting will precipitate a collapse in carbon sequestration by tropical forests.

Detecting plant spatial pattern change after disperser loss: A simulation and a case study

“…There is growing evidence, largely from tropical forest ecosystems, that defaunation and the loss of zoochorous seed dispersal services profoundly affects plant populations (Terborgh, ). A number of studies have documented changes in plant spatial pattern (Bagchi et al, ), plant germination and abundance (Egerer, Fricke, & Rogers, ), and altered interaction networks (Fricke, Tewksbury, & Rogers, ) following disperser loss. However, the relationship between disperser decline and seed dispersal can be complex (McConkey & O'Farrill, , ); for example, dispersal services can fail long before the dispersal agent is completely lost, and loss of frugivores may actually increase dispersal in some cases due to shifts in interaction networks (e.g., reduced competition for fruit).…”

“…The loss of a key disperser can fundamentally change plant spatial patterns, with implications for the demographic processes that shape plant populations (Bagchi et al, ). Thus, we expected to see a shift in spatial pattern following the immediate loss of a disperser in a model system without the confounding impacts of spatial effects such as resource gradients and microhabitat variability.…”

“…This representation means that density dependence is applied at the population level, rather than locally, and so it is a null model for local density effects. While strengthening aggregation might be expected to drive more intense negative density dependence, Bagchi et al () found that despite local aggregation increasing under defaunation, there was no commensurate increase in self‐thinning. Likewise, Sugiyama et al () found that although bat dispersal leads to high local seedling densities the benefits of deposition in favorable sites outweighs negative density dependence.…”

“…For example, Holbrook and Loiselle () and Forget and Jansen () describe dispersal limitation in tree populations where disperser abundance (large vertebrates and scatter‐hoarding rodents, respectively) has been reduced. In some cases, disperser loss has led to juveniles becoming more aggregated around adults (Bagchi et al, ; Harrison et al, ), but in others it has not (Terborgh, Nuñez‐Iturri, Pitman, et al, ). These differences may represent site‐level effects, but different studies have used different methods to quantify spatial pattern and these characterize different facets of spatial pattern (Perry, Miller, & Enright, ).…”

Intertwined effects of defaunation, increased tree mortality and density compensation on seed dispersal

“…In humanaltered habitats, these frugivorous vertebrates have frequently suffered a history of persecution and exploitation that perpetuates contemporarily, causing marked population declines and local extinctions (Peres andDolman 2000, Ripple et al 2014). As a consequence, plant populations relying on frugivorous vertebrates for seed dispersal experience reductions of their seed rain, impoverishments of the diversity of habitats of seed arrival (Wright et al 2000, Carlo et al 2013, declines in recruitment and establishment (Caughlin et al 2015, Pires et al 2018, and eventually modifications of their ranges and spatial distributions (Rotllan-Puig and Traveset 2016, Bagchi et al 2018). Furthermore, these same plant populations inhabiting defaunated areas are often seriously impacted by human-induced mortality (Lewis et al 2015, Trumbore et al 2015.…”