Ecological communities are collections of species bound together by their influences on one another. Community structure, therefore, refers to the way in which these influences are organised. As a result, ecologists are mainly interested in the factors driving the structure, functioning, and persistence of communities. The traditional focus, however, has been on the feeding relationships among species (direct trophic interactions), whereas relationships mediated by a third species or the environment (indirect interactions) have been largely overlooked. I investigated the role of indirect interactions in structuring communities through a series of field experiments in a diverse assemblage of arthropods living on a Brazilian shrub species. I experimentally reduced the abundance of the commonest galler on the shrub and found that the perturbation resonated across the food web, affecting its structure and robustness. Since there was no potential for these effects to be propagated directly or indirectly via the documented trophic links, the effects must have spread non-trophically and/or through trophic links not included in the web. Thus, I investigated non-trophic propagation of effects in the system. I demonstrate that hatched galls of the commonest galler, which serve as habitat for other species, can mediate non-trophic interactions that feedback to the galler modifying its interactions with parasitoids and inquiline aphids. I performed further manipulative experiments, excluding ants, live galls and hatched galls, to reveal mechanisms for the non-trophic interaction modifications observed in this system. Finally, I explored how non-trophic interaction modification could affect the structure and stability of a discrete ecological community in the field. I investigated how the densities of certain pairs of groups relate to each other, and how their relationship changes in relation to a third group. Then, I assembled an "effect network" revealing, for the first time in an empirical community, a hidden web of non-trophic indirect interactions modifying the direct interactions and modifying each other. Overall, the thesis presents evidence that communities are strongly interconnected through non-trophic indirect interactions. This is one of the first empirical demonstrations of the context-dependent modification of interactions via non-trophic interactions. However, determining the mechanisms behind such interaction modifications may be unfeasible. Understanding how the observed effects relate to community structuring requires shifting our focus from bipartite interaction networks to a more holistic approach.

Mopane (Colophospermum mopane) trees are browsed upon by two key species, namely mopane caterpillars (Imbrasia belina) and African elephants (Loxodonta africana), which each inflict a different type of damage while feeding, namely defoliation (leaf removal) and pruning (branch and/or stem breakage). Damage type can have a significant influence on plant responses, and these induced changes in morphological and chemical characteristics of regrowth can influence the subsequent feeding behaviour by each species. The objective of this study was therefore partly to investigate the differential effect of defoliation by mopane caterpillars and pruning by elephants on mopane trees, and then to investigate whether these two taxonomically different species interact through their shared food resource, by looking specifically at the effect of elephant utilisation of mopane on mopane caterpillar abundance. To determine the comparative effect of each browsing type, mopane trees were subjected to simulated mopane caterpillar or elephant utilisation treatments, at various frequencies and times within the year. Regrowth characteristics were then measured on treatment and control trees, as well as on naturally utilised and unutilised trees. Reproductive investment was also recorded on naturally utilised and unutilised trees. Additionally, the impact of mopane caterpillar defoliation and elephant pruning on plant stress was investigated by measuring the level of fluctuating asymmetry (FA) in leaves. Then, to determine whether there is an interaction between elephants and mopane caterpillars, mopane caterpillar egg mass abundance in areas of high elephant impact was compared to that in areas of low elephant impact. Firstly, however, in areas without elephant damage, those tree characteristics determining host tree preference by ovipositing mopane moths were identified. From this, an understanding of how elephant utilisation may influence mopane caterpillar abundance could therefore be gained. Defoliation and pruning had a significant different effect on mopane regrowth responses. Shoot and leaf length were significantly longer on pruned trees than control trees, for both naturally utilised and simulated elephant treatment trees, while there was no difference in shoot density. Defoliation, however, resulted in shorter shoots and leaves, particularly on naturally defoliated trees, which also had leaves of a higher nutritional value (tannin:protein ratio and total polyphenolic content) than control trees. A similar increase in leaf nutritional value was recorded in areas of high elephant impact in the Kruger National Park, but not after simulated or natural elephant damage in Venetia, where natural elephant utilization was less intense. Time since damage (i.e. first versus second flush) had a significant influence on regrowth after pruning, as shoot and leaf length were significantly longer on trees flushing for the first time, while within-season timing of damage was important for defoliation, as late-season defoliation had a greater negative impact than mid-season defoliation. Late-season defoliation also had a negative effect on leaf carriage into the dry season, while pruning appeared to aid leaf retention. Reproductive investment was found to be unaffected by mopane caterpillar defoliation or elephant pruning, as mean pod density and pod mass on utilised trees was no different to unutilised trees. Defoliation also had no influence on a plant’s likelihood of flowering that same season, with flowering being determined more by tree height. Unlike pod production, however, mean leaf density was significantly reduced in the regrowth of defoliated trees, presumably due to the use of stored resources for reproduction prior to the onset of regrowth. Neither simulated nor natural defoliation by mopane caterpillars and pruning by elephants was found to affect the level of leaf FA in mopane trees, even though the degree of damage inflicted on trees was considerably higher than in studies on other species where increases in FA were observed. Mopane therefore appears to be extremely tolerant of herbivory in comparison to other species. A positive relationship between leaf nutritional value (higher protein and lower tannin and polyphenolic content) and FA was detected, but only when trees from all study areas (i.e. a wide range of environmental conditions) were considered simultaneously. Environmental conditions, rather than herbivory, therefore appear to have a greater stressing affect on mopane. In the absence of heavy elephant utilisation of mopane trees, tree size, rather than shoot length, leaf length, leaf FA or leaf nutritional value, was found to have the greatest influence on oviposition behaviour of mopane moths. Ovipositing moths showed a preference for the tall riverine habitat over the shorter woodland and scrub mopane. This preference for large trees was, however, not evident at the individual tree level, as even though egg mass number per tree was positively related to tree height, large trees were not utilised more than expected according to the available canopy volume in each size class (resource availability). Heavy elephant utilisation of mopane had a negative impact on the density of tall trees within an area, due to branch and stem breakage while feeding. Unsurprisingly then, mopane caterpillar egg mass abundance was also significantly reduced in these areas, even though the nutritional value of leaves was higher than in non-elephant impacted areas. Elephants therefore appear to have a negative effect on mopane caterpillar abundance, primarily due to their negative impact on the density of tall mopane trees. This megaherbivore and invertebrate do therefore interact through their shared food resource, mopane trees. / Thesis (PhD (Zoology))--University of Pretoria, 2007. / Zoology and Entomology / unrestricted

Cyclic and extreme ecological disturbances have the capacity to alter resources and thereby animal populations. Interactions between disturbance and resource availability can influence predator-prey interactions. Predator-prey responses to ecological disturbance may be more pronounced in herbivores and their predators as herbivores track food resources that are often augmented by ecological disturbance. My objective with this dissertation was to examine how various forms of ecological disturbance influence predator-prey interactions through the lens of a case study – white-tailed deer (Odocoileus virginianus) and Florida panther (Puma concolor coryi) in southwestern Florida public and conservation lands. I quantified species-specific behavior of deer to an extreme disturbance event (i.e., Hurricane Irma), examined behavior of females with different fate outcomes to varied ecological disturbances and predation, investigated how ecological disturbance mediates the influence of human disturbance on predator-prey interactions, and quantified deer spatial ecology in response to fire, hydrology, panther and human activity. I found that deer behaviorally mediated the negative fitness impacts of Hurricane Irma. Further, I found that female deer with different fate outcomes selected areas of different ecological disturbance and the ecological disturbance type conferred different fitness costs (through differences in predation risk). Finally, I found that South Florida deer utilize diurnal times when humans are the most active to temporally reduce predation risk as panthers were more nocturnal in response to humans. My work here suggests that ecological disturbance regimes have the capacity to influence predator-prey interactions through nuanced mechanisms. Outcomes of these nuanced species-specific and predator-prey responses should be examined further. More practically, if disturbance influences aspects of animal fitness, a deeper understanding of species-specific and predator-prey responses to disturbance will improve management and conservation efforts as some regimes can be manipulated (e.g., prescribed fire). More broadly, consideration of ecological disturbance when examining predator-prey interactions may yield novel insight that deviates from predictions based on inference suggested in systems without disturbance. Highlighting nuanced predator-prey interactions mediated by ecological disturbances will improve predictions regarding species and community responses to global changes such as climate change and ecological restoration. / Doctor of Philosophy / Events that cause temporary changes to ecosystem structure and function (ecological disturbances) have the capacity to influence resources (i.e., food and shelter) for wild animals. Changes in resources as a function of ecological disturbance has the capacity to influence prey and predator species interactions. Predator-prey responses to ecological disturbance may be more pronounced in plant-eating animals (herbivores) and their predators as herbivores utilize food resources that are often altered by ecological disturbance. My objective with this dissertation was to examine how various forms of ecological disturbance influence interactions between predators and prey by using the white-tailed deer (Odocoileus virginianus) and Florida panther (Puma concolor coryi) in southwestern Florida wildlands as a case study. I quantified species-specific behavior of deer to an extreme climate event (i.e., Hurricane Irma). Next, I quantified and compared behavioral differences in responses to ecological disturbance (flooding and fire) and panther predation risk between female deer that survived and those killed by panthers during the offspring rearing season. I also investigated how ecological disturbance and human use of wildlands influenced predator-prey interactions. Finally, I characterized deer behavior in response to fire, hydrology, and panther and human activity. I found that deer changed their behavior during Hurricane Irma, presumably to offset the negative impacts of the storm as all our monitored deer survived the event. Further, I found that different ecological conditions generated by fire and flooding, respectively, influenced female behavior during the offspring rearing season in response to predation risk, and those behavioral differences may explain differences in mortality outcomes. Finally, I found that South Florida deer utilize daylight hours when humans are the most active to minimize encounters with predators as panthers were shown to be more active at night in areas with greater human use. My findings suggest that ecological disturbances have the capacity to influence predator-prey interactions in novel ways not suggested elsewhere. Outcomes of novel predator and prey interactions in response to ecological disturbance should be investigated further. More practically, if disturbance influences aspects of animal livelihood, a deeper understanding of species-specific and predator-prey responses to disturbance will improve management and conservation efforts as some disturbances can be manipulated (e.g., prescribed fire). More broadly, consideration of ecological disturbance when examining predator-prey interactions may yield novel insight that deviates from predictions based on inference suggested in systems without disturbance. Highlighting novel predator-prey interactions that is changed as a result of ecological disturbances will improve predictions regarding species and community responses to global changes through climate change and ecological restoration.

No description available.