Since becoming a pest in graminaceous crops in Africa, the African sugarcane stalkborer Eldana saccharina Walker has been the subject of much study. Its very cryptic habits have precluded more commonly available control measures being used against it. Biological control is regarded as a viable control option, but the apparent paucity of parasitoids in graminaceous crops leads to E. saccharina being regarded as lacking parasitoids, and thus not a good candidate for biological control in the classical sense. In contrast, this project argues that interactions in indigenous hosts of E. saccharina had been ignored, and that classical biocontrol principles and basic ecological theory could be applied by the discovery, collection and introduction of parasitoids from its indigenous habitat to its newly adapted habitat, sugarcane. The habitat offered by Cyperus papyrus L. was shown to be heterogenous both temporarily and spatially. Umbels, from young through mature to senescent, were available in the same proportion for colonisation throughout the year. Umbels with sexual reproductive stages (seeds) were present from early spring into late summer, and provided an additional component to the already heterogenous environment. Young umbels, in addition, developed from rhizomes in an environment regarded as sub-optimal for photosynthesis, until they reached the canopy. All stages of umbels were attacked by E. saccharina, but larvae were only found in rays of umbels and in the apex of the culm, which was the meristematic area for rays, both high nutrient areas. Young umbels were never found with borer pupae, only smaller larvae, indicating that E. saccharina development matched growth of young umbels until they reached canopy height. Also, the majority of borings found were occupied, indicating that infestation of young umbels was recent. All stages of E. saccharina development were found in mature umbels, which were also most abundant at anyone time. Numerous empty borings were found in addition to those occupied, indicative of past occupation by E. saccharina. Very few young larvae were found in old umbels, the majority of life stages found being pupae or empty pupal cases, and also many empty borings, showing that old umbels were not suitable for E. saccharina development. A guild of parasitoids which comprised Orgilus bifasciatus Turner, the most common parasitoid of small and smaller medium E. saccharina larvae, Goniozus indicus Ashmead the most common parasitoid of larger medium and large larvae, and an entomogenous fungus Seauveria bassiana (Sals.) Vuill. attacking all life stages of E. saccharina was found. Three uncommon parasitoids of smaller saccharina were also found, viz. Sassus sublevis (Granger), Iphiaulax sp. and Venturia sp. The former three natural enemies were instrumental in depressing a major outbreak within two months of it being observed and then maintaining the host population at a lower level in C. papyrus. G. indicus and B. bassiana were most effective during the summer and autumn months, and O. bifasciatus most effective during the winter months. This study supports the hypotheses that the apparent paucity of parasitoids and lack of biological control success thus far against E. saccharina in sugarcane has been because very little was known about its ecology and biology in its numerous indigenous host plants, and that studies of the latter factors coupled with ecological theory could enhance biological control programmes against this borer. As more indigenous host plants are investigated in the same way as has been done with C. papyrus, more will become known of natural enemies of E. saccharina. Parasitoid guilds could be selected, even from rare parasitoids in the more stable indigenous habitats, which would provide control in the unstable habitat of sugarcane. / Thesis (Ph.D.)-University of Natal, Pietermaritzburg, 1994.

The ability of Nakivubo wetland (which has performed tertiary water treatment for Kampala city for the past 40 years) to respond to pollution and to protect the water quality of Inner Murchison Bay of Lake Victoria was investigated. The aim of this study was to assess the capacity of Nakivubo wetland to remove nitrogen from the wastewater after its recent encroachment and modification, in order to optimize biological nitrogen removal processes using constructed wetland technology. Field studies were performed to assess the hydraulic loading, stability and water quality of this wetland. The distribution and activity of ammonium-oxidizing bacteria (AOB) in Nakivubo channel and wetland were also investigated, and the significance of the different matrices in biological nitrogen transformations within the two systems elucidated. Studies to optimize nutrient removal processes were carried out at pilot scale level both in container experiments and in the field using substrate-free constructed wetlands (CWs) planted with Cyperus papyrus and Miscanthidium violaceum which were adapted to the local ecological conditions. Results showed that Nakivubo wetland performs tertiary treatment for a large volume of wastewater from Kampala city, which is characterised by large quantities of nutrients, organic matter and to a lesser extent metals. Mass pollutant loads showed that wastewater effluent from a sewage treatment plant constituted a larger proportion of nitrogen and phosphorus and biochemical oxygen demand (BOD) discharged into the wetland. The upper section of Nakivubo wetland exhibited high removal efficiencies for BOD, whereas little or no ammonium-nitrogen and metals except Lead were removed by wetland. Studies further showed that nitrifying bacteria existed in the wetland but their activity was limited by oxygen depletion due to the high BOD in the wastewater and heterotrophic bacteria from the sewage treatment plant. Distributional studies indicated the presence of more AOB in surface sediments than the water column of the lower section of Nakivubo channel, an indication that nitrifiers settled with particulate matter prior to discharge into the wetland, and thus did not represent seeding of the wetland. The significant reductions in concentrations of BOD compared to ammonium and total nitrogen in the channel and wetland wastewater confirmed this finding. Whereas suspended nitrifiers upstream of Nakivubo channel equally influenced total nitrogen balance as those in surface sediments, epiphytic nitrification was more important than that of sediment/peat compartments in the wetland, and thus highlighted the detrimental impacts of wetland modification on the water quality Inner Murchison Bay and Lake Victoria as a whole. Performance assessment of pilot-scale container experiments and field-based CWs indicated highly promising treatment efficiencies, notably in papyrus-based treatments. Plant biomass productivity, nutrient storage, and overall system treatment performance were higher in papyrusbased constructed wetlands, and resulted in effluent that met national discharge limits. Thus, papyrus-based CWs were found to be operationally efficient in removing pollutants from domestic wastewater. / QC 20101028

Conservation managers worldwide are increasingly faced with the challenges of managing and protecting fragmented landscapes, largely as a consequence of human activities. Over recent decades, ecological theory has made a significant contribution to the development of landscape-scale conservation and practice. However, recommendations accounting for what is practically achievable in the modern-day landscape are currently lacking, while criteria for conservation planning and prioritisation continue to neglect the role of habitat networks at the required spatial scale for the long-term persistence of biodiversity. In this thesis, I test and apply ideas surrounding the complexities of managing and conserving species in a landscape context, using a suite of bird species endemic to papyrus (Cyperus papyrus) swamps in East and Central Africa as a model system. In the face of large-scale habitat loss and degradation, practical measures that account for the fragmented nature of this system, the needs of multiple specialist species, and the reliance on this habitat by local people, are urgently required. I first review the concepts originating from reserve design theory to provide a decision-making framework for those involved in landscape-scale conservation amid 21st century challenges to biodiversity, highlighting the key principles to be considered for informed choices to be made. Second, I show that the needs of local people can be compatible with conservation planning in the tropics, and may play an important part in maintaining habitat quality for species residing in historically disturbed landscapes. Third, I develop a novel framework to make an explicit link between metapopulation dynamics and conservation planning. Despite differences in the patch-level dynamics of individual species, areas of habitat where populations of multiple species are resistant to extinction, and resilient because of high chances of (re)colonization can be identified, highlighting where resources could be invested to ensure species have the capacity to respond to future change. Finally, I simulate the metapopulation dynamics of the papyrus-endemic birds to demonstrate that the optimal conservation strategy for the long-term persistence of all species residing in a network depends on the characteristics of individual species, and the total area that can be protected. Overall, this thesis develops and tests the ecological theory used in spatial conservation planning, emphasising the importance of habitat disturbance and interspecific ecological differences for the effective management of habitat networks. The results increase the evidence base for the conservation of wetland birds in Africa, as well as for species residing in fragmented landscapes more generally.

Degradation of water resources has become one of the most pressing global concerns currently facing mankind. Constructed Wetlands (CWs) represent a concept to combat deterioration of water resources by acting as buffers between wastewater and receiving water bodies. Still, constructing wetlands for the sole purpose of wastewater treatment is a challenging task. To contribute to this research area, the fundamental question raised in this doctorate thesis was: how do factors such as vegetation and residing water movements (hydraulics) influence wastewater treatment in CWs? Also, effects of different data analysis methods for results of CW hydraulics and wastewater treatment were investigated. Research was focused on  phosphorus (P), ammonium-nitrogen (NH4+-N) and solids (TSS) in wastewater and o n P in macrophyte biomass. Studies were performed in pilot-scale free water surface (FWS) CW systems in Kenya (Chemelil) and Sweden (Halmstad) and as computer simulations. Results from the Chemelil CWs demonstrated that meeting effluent concentration standards simultaneously for all water quality parameters in one CW was difficult. Vegetation harvest, and thus nutrient uptake by young growing macrophytes, was important for maintaining low effluents of NH4+-N and P, especially during dry seasons. On the other hand, mature and dense vegetation growing for at least 4 months secured meeting TSS standards. Phosphorus in above-ground green biomass accounted for almost 1/3 of the total P mass removal, demonstrating high potential for P removal through macrophyte harvest in CWs. Also, results suggested that harvest should be species-specific to achieve high P removal by macrophytes and overall acceptable wastewater treatment in CWs. Still, different methods to estimate evapotranspiration (ET) from the Chemelil CWs showed that water balance calculations greatly impacted estimations of wastewater treatment results. Hydraulic tracer studies performed in the Chemelil and Halmstad CWs showed that mature and dense emergent vegetation in CWs could reduce effective treatment volumes (e-values), which emphasized the importance of regulating this type of vegetation. Also, it was shown that hydraulic tracer studies with lithium chloride performed in CWs with dense emergent vegetation had problems with low tracer recoveries. This problem could be reduced by promoting the distribution of incoming tracer solution into the CW using a barrier near the CW inlet pipe. Computer simulation results showed that the choice of tracer data analysis method greatly influenced quantifications of CW hydraulics and pollutant removal. The e-value could be 50% higher and the pollutant removal 13% higher depending upon used method. Moreover, unrealistic evalues (above 100%) in published literature could to some extent be explained by tracer data analysis method. Hence, to obtain more reliable hydraulic data and wastewater treatment results from CWs, more attention should be paid to the choice of tracer data analysis method. / Konstruerade våtmarker representerar ett koncept för möjligheten att nå en hållbar vattenresurshantering genom att agera som ”filter” mellan föroreningskälla och viktiga vattenresurser såsom sjöar och hav. Mycket kunskap saknas däremot om hur man konstruerar våtmarker med en optimal och pålitlig vattenreningskapacitet. Den här avhandlingen undersöker därför hur vegetation och vattnets väg genom våtmarken (hydrauliken) påverkar avloppsvattenrening i våtmarker. Dessutom undersöktes hur valet av dataanalysmetod av insamlad data påverkar resultaten. Studier genomfördes i Kenya och Sverige i experimentvåtmarker (ca. 40-60 m2) och inkluderadedatainsamling av vattenkvalité, hydraulik (spårämnesexperiment) samt biomassa och fosfor i biomassan av två olika våtmarksväxter. Dessutom genomfördes datorsimuleringar. Resultaten från Kenya visade att växtskörd och efterföljande näringsupptag av nyskördade växter var viktig för att uppnå låga utgående koncentrationer av fosfor och ammonium i en tropisk våtmark, speciellt under torrsäsongen. Däremot var en välutvecklad och tät vegetation viktig för reningen av partiklar. Fosfor i grön växtbiomassa representerade cirka 1/3 av våtmarkernas totala fosforrening, vilket påvisade potentialen i att genom skörd ta bort fosfor från avloppsvatten m.h.a. konstruerade våtmarker. Resultaten pekade också på att skörden bör vara art-specifik för att uppnå en hög fosforrening och generellt bra vattenreningsresultat. Dock visade olika beräkningsmetoder att vattenbalansen i en tropisk våtmark markant kan påverka vattenreningsresultaten. Resultaten från spårämnesexperimenten demonstrerade att den effektiva våtmarksvolymen för vattenrening blev mindre vid hög täthet av övervattensväxter. Detta pekade på att regelbunden växtskörd var viktig för att uppnå god vattenrening i våtmarker. Experiment med spårämnet litium visade att man kan få felaktiga resultat p.g.a. att en del spårämne fasthålls på botten i våtmarken om denna har mycket övervattensväxter. Därför bör spridningen av spårämnet i sådana våtmarker underlättas m.h.a. en spridningsbarriär nära inloppsröret. Simuleringar visade också att valet av dataanalysmetod av spårämnesdata starkt kan påverka resultaten och därmed också vår tolkning av en våtmarks hydraulik och reningskapacitet. Den effektiva volymen kunde vara 50% högre och reningseffekten 13% högre beroende på vilken metod som användes. Likaså kan valet av dataanalysmetod ha bidragit till överskattade och orealistiska effektiva volymer (över 100%) i artiklar publicerade de senaste 25 åren. Genom att fokusera mer på valet av dataanalysmetod och t.ex. jämföra resultaten från två olika metoder kan man minimera risken för bristfälliga resultat och därmed felaktiga slutsatser om en våtmarks vattenreningskapacitet.