Search

Seed dispersal mutualisms, including seed dispersal by ants, are critical to the assembly of communities and the function of ecosystems. However, the consequences of human-caused global change factors, such as habitat fragmentation and climate change, for the future of these mutualisms are not fully understood. My dissertation consists of five chapters that investigated the effects of habitat fragmentation and climate change on ants and their seed dispersal mutualisms. Chapter 1 is an introduction that provides background information on habitat fragmentation and anthropogenic climate change and their impacts on biodiversity. In Chapter 1, I also introduce my study system of ant-mediated seed dispersal mutualisms, myrmecochory. My next two chapters (Chapters 2 and 3) explored the effects of reconnecting fragmented habitat patches with corridors in restored longleaf pine savanna systems in South Carolina. We used a landscape scale experiment to investigate how reducing isolation affects the assembly of ant communities over time (Chapter 2) and seed dispersal of the myrmecochorous forb, Piriqueta cistoides (Chapter 3). For Chapter 2, we found evidence that both habitat connectivity and edge effects underly the effects of corridors on ant communities over time. We found that connected patches accumulated ant species faster than isolated patches over time suggesting that corridors function by facilitating colonization. We also found evidence that edge effects play a role with greater ant functional group diversity in patches with higher edge than patches with lower edge amounts. For Chapter 3, we also found evidence of corridor and edge effects with ants dispersing seeds of P. cistoides longer distances in patches connected via corridors than isolated patches, but only in the center of patches. In Chapter 4, we investigated the effects of predicted climate change scenarios for seed dispersal mutualisms in eastern deciduous forests. For this chapter, we conducted a mesocosm experiment in which we crossed temperature with altered precipitation magnitude and frequency. Our mesocosms contained a common spring ephemeral wildflower, Sanguinaria canadensis, and whole colonies of their mutualist seed disperser, Aphaenogaster rudis. This design allowed us to collect high-resolution data on how ants interacted with seeds under different climate change scenarios that incorporated warming temperatures and altered precipitation. We found that warming effects depended on the precipitation treatment with negative effects of warming on the collection of seeds by ants under historical precipitation regimes and positive effects of warming under simulated precipitation conditions altered under predicted climate change. Finally, Chapter 5 describes my general conclusions from this body of work. Taken together, the research making up my dissertation provides valuable insights into how changing environmental conditions under habitat fragmentation and climate change may alter ant seed dispersal mutualisms. Importantly, we often found that the impacts of global change were context dependent and that our experiments were important tools in disentangling that context dependency. Further, this work demonstrates the value of understanding the basic ecology of the interactions among organisms. Understanding the natural history of organisms across changing environmental conditions will benefit the ways in which we conserve and restore ecosystems in a fragmented and warmer world.

Ant-mediated seed dispersal, myrmecochory, is a diffuse mutualism in which ants are rewarded for seed dispersal services with food via a seed-coat derived appendage, the elaiosome. Seeds gain dispersal benefits including escape from distance- and density- dependent mortality agents, protection from seed predators, and a nutrient-rich germination site in or near ant nests. However, microbes, have been essentially overlooked in this mutualism, despite their pathogenicity to the ants and plants involved. The work presented here investigates the effects of microbial mortality agents on the risks and benefits offered to ant and plant partners in myrmecochory. First, I investigate the effects of a plant-produced chemical on the growth of a ubiquitous entomopathogen and on ant survival and foraging. Then, I investigate ant effects on the microbiomes associated with seed dispersal location and the seed coat. While a plant-produced chemical, sanguinarine, promotes the growth of an ant pathogen at certain concentrations, it did not influence ant survival or foraging patterns. These results suggest that though plant chemicals may increase pathogenic risk to ants in myrmecochory, ant social immune behaviors help counteract this threat. Ant activity influences both the ant nest and the seed coat microbiome. Increasing or stable ant activity in ant nests reduces turnover of 1) the plant pathogenic fungal community in soil substrates and 2) the myrmecochore pathogenic fungal community in log substrates. The handling of seeds by ants also alters the seed coat microbiome. Though seed species also influences the seed coat bacterial and fungal communities, ant handling and the removal of the elaiosome decreases the alpha diversity of plant pathogens on the seed coat. These results suggest that seeds are provided with additional benefits in myrmecochory in the form of pathogen defense in the ant nest environment and with ant handling. These findings, overall, suggest the importance of including microbial communities in the consideration of the risks and rewards in ant seed dispersal. Generally, this dissertation advances the field of ecological interactions by extending studies of a pairwise mutualism to include ecological community members that act as mortality agents for mutualistic partners.

Drivers of species rarity can be classified as abiotic (environmental) or biotic (species interactions), and the spatial scale at which factors impact rarity ranges from fine to coarse. This dissertation explores drivers of narrow endemism, a kind of rarity in which species are restricted to geographic regions or habitat types, in the genus Trillium, a group of perennial understory forest herbs in eastern North America. To identify factors characterizing narrow endemism in this system, I used a framework comparing ecological attributes of co-occurring endemic and widespread congeners. At a fine spatial scale, I explored whether seed dispersal effectiveness differed for endemic and widespread species. Trilliums rely on ants for seed dispersal; my work establishes that rates and probabilities of seed dispersal are lower in the field for endemic trilliums, and that ants display a preference for the seeds of widespread trilliums, suggesting that narrowly endemic species may be dispersal limited. To investigate underlying mechanisms, I quantified interspecific differences in the morphology and chemistry of seeds and elaiosomes (i.e. seed-coat appendages that are attractive to ants). I found that seed morphology, concentrations of fatty acids, and aspects of elaiosome phytochemistry were significantly related to dispersal probability for five species of Trillium, suggesting that endemic species produce less-attractive seeds. At a coarse spatial scale, I generated species distribution models (SDMs) using Maxent for 21 species of Trillium; SDMs produced areas of predicted suitability for each species based on abiotic conditions. Within this framework, I found that rarer species of greater conservation concern had lower proportional occupancy of the fundamental niche than common species of lower conservation concern; that SDMs performed best for species of higher conservation concern with larger predicted suitable areas; and finally, that flower type and ovule number were significantly related to model performance. These discoveries provide valuable insight into the ecology and biogeography of species in the genus Trillium, thus advancing the conservation-relevant knowledgebase for this horticulturally-important group of plants, and contribute to our understanding of factors that maintain plant endemism at fine and coarse spatial scales, from both abiotic and biotic perspectives.

Plant species introduced to new locations may lose their natural enemies but can also leave behind important mutualists. Here, I take a novel comparative approach to identify the potential role of mutualistic interactions in determining invasion outcomes. I examine the strength of pollination, seed dispersal and belowground symbioses with nitrogen-fixing bacteria (rhizobia) across three species that vary in invasion success in both their introduced and native range. I used species of Australian Acacia introduced to New Zealand. I hypothesised that if interactions with mutualists are important for plant invasion then species would vary in the strength of interactions with one or more of the groups of mutualists I examined, and that the pattern of variation would correlate with the degree to which they have established and spread in New Zealand. At each stage I also consider the potentially mediating influence of natural enemies. For A. dealbata, a highly invasive species, and A. baileyana, a species that is widely naturalised in New Zealand, I found no differences in any of the variables I examined in relation to pollination and predispersal seed predation. However, for A. pravissima, currently considered a casual species in New Zealand, pre-dispersal seed predation was lower in New Zealand and overall seed production was much higher relative to conspecifics in Australia, and relative to the other two species. In relation to seed dispersal I found that the three species, which are all adapted for dispersal by ants (myrmecochory), were able to form dispersal mutualisms in New Zealand, potentially to the same degree as in Australia. Seed predation following seed fall was also lower for species in New Zealand than in Australia. There was no variation between the three species in seed removal associated with either dispersal or predation. By examining species’ growth and nodulation with rhizobia in both Australia and New Zealand I found that their ability to spread away from introduction sites could be limited by the availability of rhizobia in New Zealand, relative to Australia. However, there were again no differences between species. This is the first study to have directly measured mutualistic interactions across species that vary in invasive success in both their native and introduced range. I demonstrated that species introduced to new locations are able to establish mutualistic interactions with pollinators and dispersers to the same degree as in their native range. I also found the first direct evidence that the availability of rhizobia could limit species’ abilities to colonise new sites in the introduced range. However, mutualistic associations could not explain the variable invasive success of each species. Overall, these findings suggest that mutualistic interactions may be important for alien plant establishment, but alone cannot explain invasion outcomes. Instead, it is likely that invasive success is determined by a combination of biotic, abiotic and human factors, with the ability to establish mutualistic interactions just one component necessary for successful establishment and spread. These findings underline the importance of such broad geographical and comparative studies in attempts to elucidate drivers of invasion.