Thesis (doctoral)--Swedish University of Agricultural Sciences, 2003. / Appendix includes reprints of five manuscripts, four co-authored with others. Includes bibliographical references. Also partially issued electronically via World Wide Web in PDF format; online version lacks appendix.

Woody plant encroachment is a phenomenon whereby trees and shrubs invade grasslands or increase in an already wooded area, resulting in lower yields of herbaceous plants and a reduction in the carrying capacity of rangelands. It is not only the extent of woody plant encroachment, but also the rate at which it occurs, that is a major concern for livestock ranchers interested in herbaceous production. The question of what causes woody plant encroachment still remains unanswered. Animals consume a considerable amount of woody plant seeds during the dry season and could be responsible for spreading undesirable species to new areas or they could enhance the dispersal of species already established. However, under natural conditions, seed dispersal does not guarantee seed germination, which depends on appropriate environmental and seed survival conditions. To better understand the mechanisms involved in woody plant encroachment, we studied the effects of livestock (goats, sheep and cattle), diet quality (high-quality (Medicago sativa hay) vs. low-quality (Digtaria eriantha hay)), seed characteristics (size, shape and hardness), grass competition and fire on germination and seedling recruitment of Dichrostachys cinerea and Acacia nilotica seeds. These objectives were achieved by conducting the following trials: a) recovery and germination of D. cinerea seeds fed to goats, b) diet quality on germination of Dichrostachys cinerea and Acacia nilotica seeds fed to ruminants, c) the effects of gut passage and dung fertilization on seedling establishment of Dichrostachys cinerea and Acacia nilotica seeds and d) the effects of gut passage, dung fertilization, trampling, grass competition and fire on seedling recruitment of the two woody plant species. During the first trial, seed recovery rate of D. cinerea seeds consumed by goats, either voluntarily after mixing them with feed pellets (mixed), or by force-feeding (gavaged) and germination percentage were measured. Seed recovery for the gavaged treatment (32.7%) was significantly higher than for the mixed treatments (9.9%; P < 0.001). Seeds that passed through the digestive tract (mixed (35.5%)) and gavaged (31.2%) treatments) had a significantly higher germination percentage than untreated seeds (19.0%). A non-negligible proportion of D. cinerea seeds remained intact after chewing and passage through the digestive system, and their germination percentage increased. In the second trial, I tested the effects of animal species (goats and sheep, goats and cattle), diet quality (Medicago sativa hay and Digitaria eriantha hay), and seed characteristics (size, shape and hardness) on the effectiveness of animal seed recovery and germination of D. cinerea and A. nilotica seeds. The trial was divided into two experiments. In the first experiment, a significant interaction effect of animal species (goats, sheep), diet (high-quality hay, low-quality hay) and seed species (A. nilotica seeds, D. cinerea seeds) was found on percentage germination (P < 0.0001). There was also a higher percentage seed recovery (P < 0.009) when animals were offered high-quality hay (47.4% + 4.65) compared to low-quality hay (30.2% + 3.24). In goats fed D. eriantha hay, A. nilotica seed germination (9.38% + 3.66) was higher (P < 0.05) than goats fed D. eriantha hay and D. cinerea seeds (6.78% + 1.13). A greater germination percentage was observed in goats fed M. sativa hay with D. cinerea seeds (6.71% + 1.53) than goats fed M. sativa hay with A. nilotica seeds (2.50% + 0.97) (P < 0.05). In the second experiment, animal species had a positive impact, both on seed recovery (P < 0.0325; goats 32.0% + 6.44; cattle 50.3% + 4.27) and germination percentage (P < 0.055; goats 14.1% + 1.48; cattle 9.3% + 0.94; control: D. cinerea 0.64 + 0.06; A. nilotica 0.59 + 0.07). Animal species in experiment one (goats and sheep) and two (goats and cattle) was most important for seed recovery and germination. However, diet (M. sativa hay and D. eriantha hay) and seed species (D. cinerea, A. nilotica seeds) also had important effects on germination of seeds retrieved from experiment one. The interaction of animal species and size, diet quality, and seed characteristics (size, hardness) all played a major role in recovery of viable and scarified seeds either alone or in combination. I also studied the effects of seedling emergence, seedling establishment and recruitment of dispersed A. nilotica and D. cinerea seeds by goats and cattle under natural conditions. The interaction effect of animal species, seed recovery day and seed germination treatment/planting method was significant on seedling recruitment. Seeds retrieved from goats in the last four days and planted 2 cm in the soil with dung (25.85% ± 0.46) and seeds planted 2 cm in the soil with no dung (24.77% ± 0.35), recruited significantly better than seeds planted on top of the soil (16.98% ± 0.46). The results also indicated significant differences in percentage seedling recruitment among goats, cattle and control, with goats and cattle having the highest percentage recruitment than controls or untreated seeds. Overall, seeds can potentially germinate and recruit following passage through the gut, thereby facilitating woody plant encroachment. The results of the last trial showed that seed passage through the digestive tract of goats and cattle compared to untreated seeds (i.e. not ingested) played an important role in improving germination through seed scarification. However, seed recovery by livestock does not guarantee seedling establishment survival, survival and recruitment. Fire and grass mowing treatments affected seedling emergence, seedling survival and recruitment, most probably because of reduced grass competition, thereby reducing competition for resources (light, water and nutrients) between grasses and seedlings. In conclusion, this study indicated that animal species (goats, sheep and cattle), associated diet (low-quality vs. high-quality), seed species (D. cinerea and A. nilotica) and seed characteristics (size, shape, hardness) all played an important role in seed germination. The interactions of animal species, fire, dung, and season either directly or indirectly were pivotal in the emergence, survival and recruitment of D. cinerea and A. nilotica seedlings. Thus, acid scarification in the gut of herbivores in combination with their indirect effects (dung fertilization) and removal of grass competition (either by fire or mowing) can facilitate seedling emergence, seedling survival and recruitment of woody plant species, which may lead to woody plant encroachment. / Thesis (Ph.D.)-University of KwaZulu-Natal, Pietermaritzburg, 2014.

Plant-herbivore interactions play a significant role in the structure and functioning of ecosystems. Co-evolutionary theory suggests that plant defenses evolved due to herbivores and herbivore pressure can shape the genetic composition of their food resources. We used interactions between North American porcupines (Erethizon dorsatum) and trembling aspens ( Populus tremuloides) as a system to investigate this theory's important assumption that herbivores select food sources based on genetically controlled traits. We confirmed that porcupines exhibit intra-specific food selection and that this is linked to the genetic composition of the aspens. We also demonstrated that variation in phenolic glycosides and condensed tannins are strong components of this selection, thereby creating an important link between genetics, plant chemistry, and mammalian herbivory. We investigated potential impacts of porcupine herbivory on aspen using fluctuating asymmetry, however we did not detect any stress on heavily eaten trees, thereby questioning the validity of this tool for this study system.

Thesis(M.S.)--Auburn University, 2005. / Abstract. Vita. Includes bibliographic references.

Thesis (Ph.D. (Zoology & Entomology)) - Rhodes University, 2008.

No description available.

Orientador: Paulo Sergio Moreira Carvalho de Oliveira / Tese (doutorado) - Universidade Estadual de Campinas, Instituto de Biologia / Made available in DSpace on 2018-08-09T07:36:37Z (GMT). No. of bitstreams: 1 Christianini_AlexanderVicente_D.pdf: 2579789 bytes, checksum: 2736086f93095fb3c874eabe47217321 (MD5) Previous issue date: 2007 / Resumo: Nosso conhecimento a respeito dos sistemas de dispersão de sementes tem aumentado consideravelmente nos últimos anos. Um número crescente de estudos tem mostrado que a regeneração de plantas é freqüentemente muito mais complexa do que pensávamos, incluindo vários agentes ao longo de etapas subseqüentes do processo de dispersão de sementes. Por exemplo, formigas podem rearranjar a sombra de sementes que cai ao solo, o que pode influenciar as probabilidades de transição do estágio de semente para plântula durante o recrutamento. Neste trabalho reportamos informações mostrando que formigas são importantes agentes de dispersão secundária de sementes na maior savana da América do Sul, o cerrado. Formigas interagiram com diásporos caídos de muitas plantas dispersas primariamente por vertebrados frugívoros. Formigas freqüentemente limparam as sementes da polpa dos frutos, o que aumentou sua germinação. As sementes de Erythroxylum pelleterianum (Erythroxylaceae), Xylopia aromatica (Annonaceae) e Miconia rubiginosa (Melastomataceae) são dispersas primariamente por aves, mas a maior parte de seus frutos cai ao solo sob a planta-mãe. Formigas removeram grande parte destes diásporos caídos, e promoveram dispersão direcionada a microsítios ricos em nutrientes onde houve maior sobrevivência de plântulas, como demonstrado para E. pelleterianum. Contudo, este benefício por vezes foi alcançado à custa de perdas significativas de sementes para formigas granívoras, como em Xylopia aromatica. Aves são responsáveis pela dispersão de sementes a longas distâncias e colonização de novos sítios, enquanto formigas rearranjam a sombra de sementes numa escala menor, depositando-as em sítios onde a sobrevivência das plântulas é aumentada. Embora a maioria das formigas foi generalista em relação às características dos diásporos, formigas cortadeiras mostraram algumas preferências, especialmente por diásporos ricos em carboidratos. Nós sugerimos que estas preferências podem ser devidas às defesas químicas da folhagem de plantas do cerrado, que forçariam as formigas a depender de frutos carnosos para o cultivo de fungo no interior dos ninhos. Finalmente, nós mostramos que as interações formiga-diásporo são suscetíveis a efeitos de borda, que diminuem os benefícios obtidos por plantas dispersas secundariamente por formigas. Esta informação é especialmente relevante, uma vez que o cerrado está sendo convertido para agricultura a taxas alarmantes e efeitos de borda não haviam sido reconhecidos para o cerrado até o momento / Abstract: Our knowledge about seed dispersal systems has been improved considerably in the last few years. An increasing number of studies has shown that the process of plant regeneration is often much more complex than we realize, including several different agents across subsequent steps of seed dispersal. For instance, ants may reshape the seed shadow after seeds fall to the ground, and this may influence the transition probabilities from seed to the seedling stage in plant recruitment. Here we report data showing that ants are important agents of secondary seed dispersal in the largest South American savanna, the cerrado. Ants interacted with fallen diaspores of many plants primarily dispersed by vertebrate frugivores. Ants often cleaned the seeds from fruit matter, what increased seed germination. The seeds of Erythroxylum pelleterianum (Erythroxylaceae), Xylopia aromatica (Annonaceae) and Miconia rubiginosa (Melastomataceae) are primarily dispersed by birds, but most fruits fall to the ground under the parent tree. Ants removed a considerable number of fallen diaspores of these plants, and provided directed dispersal to nutrient-enriched microsites where seedling survival was increased, as shown for E. pelleterianum. However, this benefit sometimes is attained at the cost of significant seed loss to granivorous ants, as in the case of Xylopia aromatica. Birds are likely responsible for long-distance dispersal and colonization of new patches, while ants reshape the seed shadow at a finer scale, delivering seeds to specific sites where seedling survival is more likely. Although most ant taxa were generalist in relation to diaspore traits, leaf-cutter ants showed a preference pattern for some diaspores, particularly carbohydrate-rich ones. We suggest that such preference may be driven by the chemically-protected plant leaves of the cerrado, which would constrain leaf-cutter ants to rely on fleshy fruits for fungus culturing inside their nests. Finally, we showed that ant-diaspore interactions are susceptible to edge effects, which decrease benefits obtained by plants secondarily dispersed by ants. This information is particularly relevant, since the cerrado is currently being converted to cropland at an alarming rate, and so far edge effects had not been recognized in the cerrado / Doutorado / Ecologia / Doutor em Ecologia

In this study, I investigated the interaction between two megaherbivores, the African elephant and white rhinoceros, that has the potential to impact grazing lawns of which white rhino are the creators and maintainers and elephants are potentially the modifiers. I hypothesized that as elephants browse, they discard a variety of coarse woody debris onto the ground; should this woody debris (of varying amounts and sizes) fall onto grazing lawns, white rhino either move them, consume grass around the woody debris or abandon the lawn entirely. If high levels of woody debris are deposited here, grazing by white rhino is likely to be prevented, at which time I predicted that mesoherbivores would have a competitive advantage in accessing forage that white rhino cannot. I examined the mechanistic links between different levels of elephant-deposited woody debris and grass response at a point scale and feeding patch spatial scale of grazing lawns in an African savanna. In addition, I assessed the response of mesoherbivores in terms of vigilance behaviour with increasing levels of predation risk posed by increasing levels of woody debris. I present the first evidence of an indirect effect of elephant on white rhino foraging behaviour. I demonstrate how increasing levels of woody debris lead to a decreasing probability of foraging by white rhino. I also demonstrate how the probability of foraging by mesoherbivores increases as the amount of forage increases. However, since this study took place during a severe drought where resources are extremely limited, I was unable to properly separate the effects of elephant-deposited woody debris from the severe lack of rainfall on grass response and subsequently herbivore foraging behaviour. Due possibly to the drought, mesoherbivores responded less or not at all to risk factors such as woody debris therefore woody debris was not a predictor of vigilance behaviour in my study. This study contributes to our understanding of how the impacts of elephants, as ecosystem engineers, have cascading effects on savanna ecosystems. My study showed that elephant impact mediates the foraging behaviour of white rhino during a drought. However, under average rainfall periods, my original hypothesized effect of the indirect impacts of elephants on white rhino foraging and grazing lawn dynamics could still hold. This key hypothesis that I was unable to test under ‘normal’ conditions due to the drought is still valid and functionally important for understanding the ecosystem processes driving grazing lawn formation, persistence and composition in African savannas where elephants and white rhinos coexist.

A dissertation submitted to the Faculty of Science in fulfilment of the requirements for the Master of Science degree. in School of Animal, Plant and Environmental Sciences, University of The Witwatersrand, Johannesburg July 2017. / Fire is known to affect spatial patterns of grazing by altering the amount and quality of forage. Animals select the post-burn green flush that remains palatable until the grass recovers its biomass. How quickly the grass regrows depends on the rainfall and grass growth rates, and also grazing intensity. Theoretically, highly concentrated grazing can maintain short (relatively more palatable) grasslands throughout a growing season. Therefore this study aimed at; i) determining how long different grazer species were attracted to the burn, and whether this increased grazing pressure (as a result of concentrating grazers on a small burnt patch) maintained a short, palatable grass sward throughout the growing season, ii) investigating the long-term impacts of herbivore attraction to small burns on grass community and landscape function in a Highveld grassland. We therefore monitored grazer utilization of an experimentally applied small (5ha) burnt patch using dung counts and camera traps, and also measured the structural changes of the burnt patch over a period of 12 months. To test whether this process of attraction to small burns could have long-term impacts on grass community composition and landscape function we quantified species composition, infiltration rates, soil compaction, soil moisture, and ANPP in another landscape which had received 10+ years of small annual burns (a firebreak). A novel finding was that indeed grazers especially the short-grass specialist stayed on the burn and kept the grass short (<10cm) for the duration of the study post fire: the burn only treatment on the 5ha burn recovered its biomass within 2 months of the first rains. This result was due to the fact that it was a drought year with half the normal rainfall (and lower grass regrowth rates). However, the long-term study indicated that the attractive effect of small fires in this ecosystem alters both community composition and ecosystem properties. The firebreak had more bare ground and less water infiltration than the surrounding grassland – but was more diverse and had higher grass productivity. It also continued to attract the short-grass specialist species (blesbok, wildebeest and hartebeest). This counter-intuitive result indicates that perhaps these grasslands are not as severely degraded as we think. This study therefore, showed that coupling small burns with appropriate grazer species has a great potential for creating palatable grazing “hotspots”, in sourveld grassland without obvious damage to ecosystem function. / LG2018

Plant-based defence mutualisms involve aspects of plant morphology that influence the performance of plant parasites, their natural enemies and trophic interactions. Leaf domatia, small indentations on the underside of leaves, can be structurally complex, and are often inhabited by potentially beneficial mites and other arthropods. Plant morphological traits such as domatia that enhance mutualistic relationships may result in increased plant growth rates, and reproductive success. New Zealand supports ~60 plant species that have domatia, the most speciose genus being Coprosma. The aim of this thesis was to examine factors that affect the production of leaf domatia and their relationship with foliar mite assemblages. The three main objectives of this thesis are: First, to investigate the production of foliar domatia and their susceptibility to limited resources, particularly to carbon availability. Second, to test if domatia are inducible structures during leaf ontogeny in the presence of foliar mites and/or fungi. Finally, to explore the effect of domatia availability on foliar mite assemblages on leaves with and without resident mites. This thesis tested the stated objectives using C. lucida, C. ciliata, C. foetidissima and C. rotundifolia, with a combination of field investigations and controlled manipulative experiments. The cost of domatia production was investigated using two field surveys and two controlled experiments. Under natural conditions the relationship between leaf morphology and domatia were measured in situ and across an altitudinal gradient. The experimental manipulations used carbon and nutrient stress, induced by temperature, light and fertilizer application. The second objective was experimentally tested under field conditions by manipulating foliar mites and fungal densities on C. rotundifolia. The third objective was investigated by manipulating domatia availability on C. lucida shrubs across three different vegetation types. Under field conditions, the number of domatia per leaf was associated with leaf morphology in C. lucida and C. foetidissima, but not C. rotundifolia. Foliar carbon showed a positive, but weak association with domatia production in C. foetidissima and C. ciliata. Altitudinal induced-carbon stress on domatia production was ambiguous. Domatia production in C. foetidissima was positively associated to altitude in field survey (1), and negatively associated in the second survey, with no correlation found between carbon and altitude. Experimental C. rotundifolia shrubs held under elevated night-time temperatures showed a 2.5 fold increase in respiration, a 34% to 91% decrease in daily carbon gain, and 38% decrease in domatia per leaf mass. Domatia production showed no significant differences under nutrient stress. The results showed little evidence to support a role for induction of domatia. Domatia production in new leaves was similar across all experimental treatments. Diverse vegetation types supported 60% higher mite species. Leaves with domatia supported ~22 to 66% higher mite densities, greater colonisation success and more diverse mite assemblages, than those without domatia. In the pastoral vegetation, the absence of predatory mites on experimental shrubs resulted in no differences in fungivorous mite densities regardless of domatia availability. Plant investment in foliar domatia appears associated with the number of available sites on the leaf under field conditions. The role of carbon availability during leaf ontogeny suggests a complex and highly variable association with domatia production. Domatia are constitutive defence structures that influence mite assemblages, mediating both beneficial and antagonistic relationships. This thesis concludes that domatia are in part, carbon-based non-inducible structures that influence mite assemblages, plant-mite and mite-mite interactions, and increase the probability of successful colonisation.