Antagonistic and mutualistic interactions alter seed dispersal of understory plants at forest edges

Abstract

Seed dispersal by ants is an important interaction in North American eastern deciduous forests, where 30-40% of understory plants are myrmecochores, with seeds that possess lipid-rich appendages (elaiosomes) that attract seed-dispersing ants. Contemporary forests are fragmented and have regenerated from being previously cleared (secondary forests). In secondary forests, and especially along edges, myrmecochores are a depauperate component of forests. Here, we assess if seed dispersal of myrmecochores by ants, particularly the keystone disperser, Aphaenogaster sp., is intact in forest interiors compared to edges. In three North American northeastern deciduous forests, we compared myrmecochore cover and richness, seed dispersal by ants, and the abundance and richness of ants and other forest floor invertebrates between interiors and edges. We also conducted a seed removal experiment, excluding either ants, rodents (seed predators), or ants and rodents, to test their effects on seed dispersal between interiors and edges. We found differences in the composition of understory plants between interiors and edges, with edges lacking myrmecochores. We also found that seed dispersal by ants was lower in two of the three edges and corresponded to Aphaenogaster sp. abundance. Antagonistic interactions with an invasive slug that is an elaiosome robber, Arion subfuscus, was more apparent than damage by rodents in the experiment, negatively affected dispersal by ants more at edges than in interiors. Here we show that changes in mutualistic and antagonistic interactions affect seed dispersal at forest edges and that not all forest edges are the same, with some more intact than others. Understanding how seed dispersal is impacted in contemporary forests is important towards the goal of conserving and restoring depauperate forest understory communities.

Historical forest disturbance results in variation in functional resilience of seed dispersal mutualisms

“…(Umphrey 1996;Ellison et al 2012;DeMarco & Cognato 2016). In our sites, we found A. rudis and A. picea (that are polyphyletic and challenging to delineate morphologically) and monophyletic A. fulva (Parker et al 2021;Buono et al 2022;Quartuccia & Buono, unpublished data) and refer to this group as Aphaenogaster sp. in our study.…”

“…This suggests that overall, other ant species do not have a direct antagonistic interaction with seeds or mutualist partners. This is an interesting finding, as it is generally predicted that other ants negatively affect dispersal directly (by being antagonistic or low quality partners) or indirectly by outcompeting the good disperser (Ness 2004;Giladi 2006;Ness et al 2009;Prior et al 2020;Parker et al 2021). Habitat factors in both the combined and remnant path analysis influenced other ant species abundance, which suggests that the other ant species were more abundant in microhabitats that also favored Aphaenogaster sp.. During the seed removal trials, we only observed two other ant species interacting with, but not removing, seeds: the native species Lasius americanus and the invasive species Nylanderia flavipes, with the latter occurring at high abundances and removing parts of elaiosomes.…”

“…High-quality dispersers find seeds quickly, do not harm seeds, and deposit seeds outside of nests (Giladi 2006; Canner et al 2012; Prior et al 2014; Prior et al 2015; Gordon et al 2019; Meadley Dunphy et al 2020). Poor-quality dispersers interact with seeds but fail to disperse seeds and might damage seeds (Christian 2001; Bronstein et al 2006; Giladi 2006; Stuble et al 2011; Warren & Giladi 2014; Parker et al 2021). In NAEDF, ants belonging to the genus Aphaenogaster are considered “keystone” dispersers, responsible for majority of dispersal events (Ness et al 2009).…”

“…Agricultural disturbances leave legacy effects in soils that impact herbaceous plants such as elevated pH and nutrients and lower organic matter (Koerner et al 1997; Dyer 2010). Less is known about whether functionally important interactions, such as seed dispersal, are resilient to historical land use change (HLUC) and if reduced function contributes to low understory recovery (except see Mitchell et al 2002; Kiel et al 2020; Parker et al 2021).…”

“…Reduction in ant partner presence, abundance, or interactions could slow understory recovery in secondary forests (Mitchell et al 2002; Kiel et al 2020; Parker et al 2021). It is well established that Aphaenogaster presence and abundance positively influences seed dispersal, distribution and community structure of myrmecochores (Kalisz et al 1999; Ness et al 2009; Warren & Giladi 2014; Prior et al 2015).…”