The backbone of this study was a systems thinking based analysis of the management and control of predators as practiced in South Africa since the advent of Europeans in 1652. The first bounties were introduced for a variety of animal species in 1656. Many species became labeled as vermin and were persecuted, often with the intention of eradication. A variety of controls have been applied, of which many have not kept pace with contemporary thinking or technology, and which by simply killing predators fail to address the crux of the issue of predator – livestock conflict. Many of the methods used cannot be applied to specifically remove an individual damage-causing animal. Considerable collateral ecological damage is inflicted by the killing of animals regarded as innocent bystanders. The objective of the analysis was to highlight the futility of temporary solutions which fail to resolve the conflict in the long term. These quick fixes frequently perpetuate an ecological imbalance which exacerbates the predator – livestock conflict. The analysis used raw data from a questionnaire survey conducted by the Poison Working Group of the Endangered Wildlife Trust (EWT-PWG) (2003). The EWT-PWG intended to assess pesticide abuse as toxicant for predators by farmers, and to identify all control methods used. The data forthcoming was Cartesian in nature and a fixed snap-shot in time. This study sought to identify the root cause of the conflict by applying systems thinking which added the dimension of cause and effect interrogation. The study categorised and described predator conflict management methods as lethal or preventative, and assessed each category in archetypal terms from a systems thinking perspective. It also sought to identify leverage points, or small changes which have profound effects, to stimulate a change in approach to humanpredator conflict management. In order to assess and illustrate the positive change brought about by application of preventative methods, a small group of farmers who had initiated changes in their predator conflict management over the five years subsequent to the EWTPWG survey were selected from the original group for reassessment. Through evaluation of predator conflict management methods from a systems thinking perspective, and by probing learning processes, the shortcomings or failure of inappropriate management responses to conflict situations were shown to exacerbate conflicts. Contrarily, it was illustrated that application of systems thinking and a process of addressing the root cause of conflict issues in predator conflict management, was a longer term solution. The study illustrated that application of long term proactive prevention and conflict avoidance principles, can offer long term solutions for predator conflict managers. / Thesis (M.Env.Dev.) - University of KwaZulu-Natal, Pietermaritzburg, 2008.

This thesis has two focuses. The first problem facing the olive industry in the Eastern Cape is the growers’ perceptions of both what the industry will provide them and what a pest management program might entail. The second focus is the biology of olive pests in the Eastern Cape in terms of understanding their populations and their natural enemies on private farms, with future hopes of understanding how Integrated Pest Management strategies can be developed for this crop. Eastern Cape private farmers, small-scale farmers and workers from agricultural training institutions were interviewed regarding the history and cultivation of the local olive crop. Only one commercially viable olive grove was identified; other groves were small, experimental pilot ventures. The introduction of olives to small-scale farmers and agricultural training schools was generally a top-down initiative that led to a lack of sense of ownership and the trees being neglected. Other problems included poor human capital; poor financial capital; lack of adequate support; lack of knowledge transfer and stability; lack of communication and evaluation procedures of the project; miscommunication; and finally, olive pests. Apart from hesitancy to plant at a commercial scale, the main problem facing private farmers (Varnam Farm, Hewlands Farm and Springvale Farm) was pests. Therefore an investigation of pests from private farms was conducted ranging from collection of cultivated and wild olive fruit and flea beetle larvae for parasitism, trapping systems both for fruit flies and olive flea beetle adults. A survey of olive fruits yielded larval fruit flies of the families Tephritidae (Bactrocera oleae (Rossi), B. biguttula (Bezzi) and Ceratitis capitata (Wiedemann)) and Drosophilidae (Drosophila melanogaster (Meigen)) from wild olives (O. europaea cuspidata (Wall. ex G. Don) Cif.) but none from cultivated olives (O. e. europaea L.). Braconid wasps (Opiinae and Braconinae) were reared only from fruits containing B. oleae and B. biguttula. This suggests that B. oleae is not of economic significance in the Eastern Cape, perhaps because it is controlled to a significant level by natural enemies, but B. biguttula may be a potential economic pest. A survey of adult fruit flies using ChamP traps baited with ammonium bicarbonate and spiroketal capsules and Sensus trap baited with methyl eugenol and Questlure confirmed the relative importance of B. biguttula over B. oleae. ChamP traps were over 50 times better than Sensus traps for mass trapping of B. biguttula but both were ineffective for trapping B. oleae and C. capitata. Six indigenous flea beetles of the genus Argopistes Motschulsky (Chrysomelidae: Alticinae) were found, three described by Bryant in 1922 and 1944 and three new species. Their morphology was investigated by scanning electron microscopy and mutivariate morphometric analysis. The leaf-mining larvae are pests of wild and cultivated olives in South Africa and threaten the local olive industry. At Springvale Farm, A. oleae Bryant and A. sexvittatus Bryant preferred the upper parts of trees, near new leaves. Pseudophanomeris inopinatus (Blkb.) (Braconidae) was reared from 23 Argopistes larvae. The beetle larvae might not be controlled to a significant level by natural enemies because the rate of parasitism was low. The olive flea beetles showed no attraction to traps containing various volatile compounds as baits. The lace bug, Plerochila australis Distant (Tingidae), was sometimes a pest. It showed a preference for the underside of leaves on the lower parts of the trees. A moth, Palpita unionalis Hübner (Crambidae), was reared in very low numbers and without parasitoids. A twig-boring beetle larva, chalcidoid parasitoids and seed wasps of the families Eurytomidae, Ormyridae and Eupelmidae were also recorded.

The objectives of this study were to establish the effectiveness of chlorpyrifos in suppressing H. armigera population in oranges and to determine the effect chlorpyrifos had on the health of farm pesticide operators. Experiments showed that by applying chlorpyrifos on orange trees, H. armigera larvae population was suppressed significantly. High fruit yields were realized from trees that were sprayed with chlorpyrifos. Visual observation of personnel involved in pesticide related duties, revealed that judicious use of pesticides was not practiced by farm workers in all three farms. Data analysis from questionnaires, health records and interviews proved that farm workers suffered from illnesses that were pesticide related. It was mainly those farm workers in the age group of 31 to 35 years who suffered the most from pesticide related illnesses. / Agriculture and Life Sciences / M. Sc. (Agriculture)

Made available in DSpace on 2014-06-11T19:35:00Z (GMT). No. of bitstreams: 0 Previous issue date: 2004-02-27Bitstream added on 2014-06-13T19:24:07Z : No. of bitstreams: 1 loureiro_es_dr_botfca.pdf: 393758 bytes, checksum: fae8a4e1b680f21d3d03858c5c8608bf (MD5) / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES) / Universidade Estadual Paulista (UNESP) / Com o objetivo de determinar a patogenicidade de diferentes isolados do fungo Metarhizium anisopliae sobre ninfas de Mahanarva fimbriolata foi testada, em laboratório, a patogenicidade de 79 isolados, provenientes de diferentes hospedeiros e regiões do país. Inicialmente, foi realizado um bioensaio para estabelecimento da concentração a ser utilizada nos experimentos de seleção, adotando-se, como padrão, o isolado IBCB 348 de M. anisopliae. Através da Análise de Probit, foi calculada a CL50 obtendo uma concentração de 1,2 x 108 conídios/mL, decorridos 4 dias da pulverização. As avaliações foram realizadas diariamente, observando-se os níveis de mortalidade acumulada (total, confirmada e corrigida) de cada tratamento. Os melhores isolados testados foram IBCB 348; IBCB 351; IBCB 363; IBCB 408; IBCB 410; IBCB 418; IBCB 425 e IBCB 482, os quais apresentaram uma eficiência igual ou superior a 70% de mortalidade confirmada, aos seis dias da pulverização. Os isolados IBCB 348; IBCB 408; IBCB 410 e IBCB 425 foram os que mais produziram conídios em arroz com 2,08 x 108, 1,75 x 108, 2,22 x 108 e 2,30 x 108 conídios/g de arroz pré-cozido, respectivamente, pelo método de bandeja. Foi pulverizado 2kg/ha de arroz+fungo, contendo 1,5 x 1012 conídios/ha, destacando-se os isolados IBCB 348 e IBCB 408 que causaram mortalidade, para ninfas de 100%, respectivamente, aos 15 dias da pulverização. Decorridos 30 dias da pulverização, os isolados IBCB 408 e IBCB 425 apresentaram eficiência de controle de 63 e 62%, respectivamente, para as ninfas e para os adultos foi de 100%, respectivamente. Aos 60 dias da aplicação, o isolado IBCB 425 foi o mais eficiente para controlar as ninfas (48,4%) e para os adultos de M. fimbriolata, os isolados não foram eficientes. Nas avaliações realizadas aos 90 e 120 dias, a população de cigarrinha manteve-se acima do nível de dano econômico... . / This work was carried out evaluate the pathogenicity of different isolates of the fungus Metarhizium anisopliae against nymphs of Mahanarva fimbriolata. The pathogenicity of 79 isolates, provenients from differents hosts and regions of the country, were tested in the laboratory. First, it was performed a bioassay to determine the concentration to be used in the selection experiments, the IBCB 348 isolate of M. anisopliae was used as standard. Evaluations were done daily, observing the levels of mortality (total, confirmed and corrected) of each treatment. The best isolates were IBCB 348; IBCB 351; IBCB 363; IBCB 408; IBCB 410; IBCB 418; IBCB 425 and IBCB 482, which showed an efficiency equal or higher than 70% six days after the pulverization. The isolates IBCB 348; IBCB 408; IBCB 410 and IBCB 425 produced the largest number of conidia with 2,08 x 108; 1,75 x 108; 2,22 x 108 and 2,30 x 108 conidia/g of rice, respectively by the tray method. Was pulverized 2kg/ha de rice+fungus, with 1,5 x 1012 conidia/ha, showing off the isolates IBCB 348 and IBCB 408 that caused a mortality for nymphs of 100%, respectively, 15 days after the pulverization. After 30 days of the pulverization, the isolates IBCB 408 and IBCB 425 showed control efficiency of 63 and 62%, respectively, for the nymphs and for adults the efficiency was 100%, respectively. Sixty day after the application, the isolate IBCB 425 were the most efficients for control nymphs (48%) and for adults of M. fimbriolata the isolates weren't efficients. In the evaluation realized at 90 and 120 days, the speatlebug population was kept above the economic threshold level. The efficiency of the isolates was null for adults and for nymphs, the isolate IBCB 408 showed efficiency of 31%, after ninety days of the pulverization. At 120 days, the efficiency of the isolates was null for nymphs and for adults the maxim efficiency (63%) was... (Complete abstract, click electronic address below).

Orientadores: Stanley Robson de Medeiros Oliveira, Carlos Alberto Alves Meira / Dissertação (mestrado) - Universidade Estadual de Campinas, Faculdade de Engenharia Agrícola / Made available in DSpace on 2018-08-20T23:34:35Z (GMT). No. of bitstreams: 1 Megeto_GuilhermeAugustoSilva_M.pdf: 5556599 bytes, checksum: c545e13e7ec81e96cf9e6a3e46d01de2 (MD5) Previous issue date: 2012 / Resumo: A ferrugem asiática, causada pelo fungo Phakopsora pachyrhizi, atualmente é considerada uma das doenças mais importantes e agressivas da soja. A principal forma de controle é a aplicação calendarizada de fungicidas a qual desconsidera o risco de ocorrência da doença. Estudos epidemiológicos buscam compreender os fatores que influenciam na ocorrência e desenvolvimento das epidemias, especialmente aqueles relacionados ao ambiente tais como condições meteorológicas. Com o avanço da tecnologia da informação e do armazenamento de dados, técnicas de mineração de dados (data mining) apresentam-se promissoras para a descoberta de conhecimento em bases de dados epidemiológicos. Este trabalho tem como objetivo avaliar a influência da chuva e da temperatura na ocorrência da ferrugem asiática da soja utilizando árvores de decisão. Para tal, foram obtidos dados de ocorrências da doença em quatro safras, de 2007/2008 a 2010/2011, oriundos do banco de dados do Consórcio Antiferrugem, e dados meteorológicos, provenientes do sistema Agritempo. A análise exploratória dos dados permitiu obter subsídios para compor o conjunto de dados final e definir o escopo deste trabalho, buscando-se características intrínsecas à doença e sua interação com o ambiente, utilizando apenas variáveis de base meteorológica. As variáveis utilizadas foram relacionadas à precipitação e à temperatura, que deram origem a nove atributos avaliados para cada período temporal...Observação: O resumo, na íntegra, poderá ser visualizado no texto completo da tese digital / Abstract: The Asian soybean rust, caused by Phakopsora pachyrhizi, is now considered one of the most important and aggressive diseases of soybean. The main form of control is the scheduled application of fungicides which disregards the the risk of disease occurrence. Epidemiological studies seek to understand the factors that influence the occurrence and development of epidemics, especially those related to the environment such as weather conditions. With the development of information technology and data warehousing, data mining techniques appear to be promising for knowledge discovery in epidemiological databases. This study aims to evaluate the influence of rainfall and temperature on the occurrence of soybean rust by using decision trees models. To accomplish that, data of the occurence of the disease were collected from four seasons, 2007/2008 to 2010/2011, from the Consórcio Antiferrugem and weather data from the Agritempo system. Exploratory data analysis allowed for obtaining subsidies to generate the final data set and define the scope of this work, seeking intrinsic characteristics of the disease and its interaction with the environment, using only meteorological variables. The variables used were related to precipitation and temperature, resulting into nine attributes evaluated in different periods. Such attributes were related to the event of occurrence (Oc) and non occurrence (NaoOc) of the disease (assumed as the thirtieth day prior to the event of occurrence). The results include a predictive model and an interpretive model for classifying events of occurrences and non occurrences of the disease...Note: The complete abstract is available with the full electronic document / Mestrado / Planejamento e Desenvolvimento Rural Sustentável / Mestre em Engenharia Agrícola

MSc (Zoology) / Department of Zoology / See the attached abstract below

Root rot diseases of bean (Phaseolus vulgaris L.) are a problem wherever they are grown, and are a major constraint to dry edible and snap bean production. Root rot is a primary yield limitation of snap bean production in the US, especially within the top three snap bean producing states of Wisconsin, Oregon and New York. Bean root rot pathogens will be present by the end of the first season even when starting with clean ground. The decline in yield can be relatively slow, so growers might not notice or appreciate the hidden yield cost associated with root rot disease. Traditional methods for disease control such as fungicides, crop rotations, cover crops, seedbed preparations have been proven ineffective (either physically ineffective or economically unviable) against root rot. Therefore, genetic resistance is needed. In order to address the need for genetic resistance to root rot in snap beans, the highly root rot resistant line RR6950, a small seeded black indeterminate type IIIA accession of unknown origin, was crossed with OSU5446, a highly root rot susceptible determinate type I blue lake four-sieve breeding line to produce the RR138 recombinant inbred mapping population. In this study we evaluated the RR138 RI population in the F₆ generation for resistance to Fusarium solani root rot in Oregon and Aphanomyces euteiches root rot in Wisconsin. We also evaluated this population for morphological traits and root structural traits including pod height, pod width, pod length, pod wall thickness, strings, seed color, flower color, tap and basal root diameter, and root angle measurements. The RR138 population was also genotyped on the 10K BeanCAP Illumina Beadchip. The Single Nucleotide Polymorphism (SNP) data was used to assemble a high-density linkage map and Quantitative Trait Loci (QTL) for phenotypic data were evaluated. The linkage map produced from this study contained 1,689 SNPs across 1,196cM. The map was populated with 1 SNP for every 1.4cM, spanning across 11 linkage groups. Three QTL associated with A. euteiches root rot resistance were consistently expressed in 2011 and 2012 trials. A. euteiches QTL were found on Pv02, Pv04, and Pv06 and accounted for 7-17% of total genetic variation. Two QTL associated with F. solani were found in 2011 trial on Pv03 and Pv07, account for 9 and 22% of total genetic variation, respectively. We also found several QTL for morphological traits and root structural traits including QTL for pod fiber and pod height on Pv04, pod length on Pv01, strings on Pv01, taproot diameter on Pv05, and shallow basal root angle on Pv05, accounting for 21, 26, 12, 20, 11, and 19% of total genetic variation, respectively. QTL discovered from Oregon data for F. solani resistance did not cluster with QTL for A. euteiches root rot resistance. "SNP0928_7", was highly associated with F. solani resistance on Pv07 and "SNP0508_2", was highly associated with A. euteiches on Pv02. QTL and markers associated with QTL from this study will be of value to snap bean breeders developing root rot resistant lines with processing traits, and provide more information about targeting the mechanism of resistance. / Graduation date: 2013

The research problem of this mini-dissertation involves the conflicts between human and wildlife populations and the trialing of ‘compensation’ payouts that are emerging as a critical test within the conservancy. Crop raiders such as elephants, buffaloes, hippopotamus, bush pigs and small rodents, diminish farmers’ resource bases and cash crops, while carnivores are responsible for livestock losses. The aim of the research is to assess the level of community understanding of the compensation scheme and the impact of human wildlife interaction within the conservancy. This mini-dissertation investigated the level of community understanding about the Human Animal Conservancy Self-Insurance Scheme (HACSIS), and the impact of human-wildlife conflicts within the Kwandu Conservancy. Some of the research data were obtained from the conservancy game guards’ event book system, and the actual field research data were collected from the 1st August to 20th August 2006. The researcher conducted a total of 35 interviews, whereby 32 involved face-to-face interviews with single individuals, and 3 separate focus group discussions that consisted of four, five and two conservancy members. The interviews averaged 30 minutes in length. Each interview was preceded by a careful explanation of the purposes of the work, stressing that the intent was to evaluate their understanding and perceptions on HACSIS, the impact of human-wildlife conflicts and to explore better management strategies. The researcher has taken into account that the communities might exaggerate the wildlife problem based on his previous experience with the adjacent conservancy, in the hopes of gaining more compensation – they also use the researcher as a way to vent their frustration at the problem. On the assumption that there may be an element of exaggeration verification of these was obtained from the Event Book System (a manual book used by the Community Game Guards for recording both crops and livestock incidents on daily basis). This mini-dissertation reveals that 74.3 % (n = 26) of respondents are aware of the existence of the HACSIS program and its role, while 17.1 % of respondents had no idea about the scheme’s presence and its involvement to minimize the impact felt by communities when they lose livestock to predators. The percentage of respondents who claimed that they had heard of the scheme’s existence but had no knowledge of its role was 8.6 %. In addition, most respondents (n = 15) claimed that the conservancy committee did not explain to them why their claim forms were rejected. In contrast, some respondents (n = 6) did received feedback on rejected claim forms. HACSIS was not formed to compensate livestock losses based on market value, nor was it intended as a ‘compensation’ scheme. Its aim was to test a conservancy-run process – local verification of claims and monitoring by conservancy committee and traditional authority. In addition, the authorization of payments for a type of ‘self-insurance’ is drawn from conservancy income to partially offset the losses of conservancy members versus the overall gains that wildlife brings to the conservancy (direct conservancy income and local jobs through tourism, trophy hunting, own use game harvesting). Conservancy committees and the support NGO, IRDNC, agreed on the amount to be refunded for animal losses before the scheme was started, initially using donor funding in the trial phases. The amount was deliberately kept low as it was acknowledged from the start that conservancies themselves would take over the repayments from their own income. Once the conservancy was used to its own income to finance the scheme, conservancy members could vote to increase amounts paid for predator losses. The crucial aspect, according to IRDNC, was that the process itself be tested and that the scheme be run by the conservancy, with Ministry of Environment and Tourism and IRDNC merely monitoring and providing assistance as needed. Compensation is based on this pre-determined amount that is less than the livestock value. However, the research reveals that respondents (n = 19) were dissatisfied with the amount paid (N$ 800-00 per ox killed), because they claim that the amount paid to relieve the immediate impact from wildlife is too little to sustain the affected member. In contrast, some respondents (n = 8) were satisfied with the amount paid as compensation. Despite criticisms about the amount paid for livestock losses, none of the respondents (n = 22) who were familiar with the scheme wanted it to be abolished. The respondents emphasized the need for the conservancy committee to review the amount paid as compensation, especially for cattle. They suggested an increase from the current N$ 800-00 to N$ 1000-00 per ox loss. The research reveals that community livestock management practices have not changed to deliberately benefit from the compensation. In fact community management strategies have improved because of the condition set by the review committee dealing with the compensation scheme. Wildlife incidents have increased because animals are habituated to techniques used by communities to deter them and this has contributed to high livestock incidents. For human-wildlife conflicts, the research acknowledges that the conflict exists. Between 2003-2005, the Kwandu Conservancy reported 1508 incidents of damage to crops by wildlife. Species that were responsible included elephants with 30.2 % damage, bush pigs (29.8 %), hippopotamus (12.7%), antelopes (12.7 %), porcupine (7.5 %), and baboons/monkeys (7.2 %). Most of the crops destroyed by crop raiders, as suggested by the respondents, were maize (30 %), sorghum (26 %), millet (17 %), groundnuts (14 %), pumpkins (8 %) and beans (5%). During the same period of crop losses, the conservancy reported 98 livestock incidents. Animals responsible for livestock incidents were crocodile with 32 incidents (32.7 %), then hyena (23 incidents, 23.5 %), leopard (22 incidents, 22.4 %) and lion (21 incidents, 21.4%). The role of community game guards was found to be extensive. From a total 35 responses, 74.3 % (n = 26) of members stated that community game guards effectively record incidents, chased problem-causing animals from the community crop fields by shouting or shooting in the air, and assessed or verified killed livestock for compensation purposes. In addition, community game guards conduct crop assessment for record-keeping purposes. The scheme for crop compensation is to be introduced in 2007. Currently there is no proper formula to use in assessing the value of crops and the method to use to compensate the affected members. Other methods used by communities to deter wildlife include sleeping in the field to guard crops, cracking a whip, construction of human statues, hanging tins on the fence, chilli coils, watchtowers and digging trenches. Respondents had different views on the best management practices for problem-causing animals. The response was generally based on the degree of threat that the animal posed. Most (43.8 %, n = 14) preferred the monitoring of problem-causing animals that are sighted in an area as a best practice, while 40.6 % (n = 13) of respondents preferred the animal to be captured and relocated to parks. Only 25.6 % (n = 5) of respondents preferred that the animal be destroyed. The management practices preferred by respondents when an animal kills a person are different from when an animal is simply sighted in the area. If an animal kills a person, only 12.5 % (n = 4) of respondents preferred that the animal be captured and relocated to parks, while 87.5 % (n = 28) of respondents preferred the problem-causing animal to be destroyed. None of the respondents suggested monitoring as the best management option for this degree of threat. In conclusion, the research revealed that Human Animal Conservancy Self Insurance Scheme does not treat the cause of the problem but the symptom. This approach does not decrease the level of the problem given that the cause of the problem is not addressed. Therefore, the researcher stressed the need to fully explore and implement the recent piloted lion, crocodile fencing, and elephant proof fencing and elephants chilli coil to address and reduce the problem within Kwandu Conservancy. In addition, the research revealed that the scheme has very lengthy delays before compensation is paid and the review panel does not arrange meetings on the stipulated dates. This causes a back-log in the number of claims that need to be reviewed and approved. On Human Wildlife Conflict the research findings recommend the need to strengthen and improve existing problem-causing animal management strategies that are in place. Innovative strategies include reducing the number of stray livestock at night and developing static fences. Communities should be advised, as is being done by IRDNC, to herd their livestock during the day and to build strong kraals. This is the most effective and cheapest way to prevent livestock from being taken by predators at night. Further more the research revealed that the combination and rotation of the methods yield high success rather than deploying a single method over a long period, for the prevention of crop losses methods include guarding the crop field, cracking a whip, shooting in the air, watchtowers, human statues and beating drums. / Thesis (M.Env.Dev.)-University of KwaZulu-Natal, Pietermaritzburg, 2006.

Soybeans, Glycine max, are an economically and strategically important crop in South Africa (SA). In order to meet local demands, large imports of soybeans are required, e.g., in the 2005/2006 soybean production period, 842 107 tonnes of oilcake were imported. Due to an increase in soybean production throughout the world, diseases that affect this crop have also increased in incidence and severity. Sclerotinia sclerotiorum, the causal organism of sclerotinia stem rot (SSR), is an important yield limiting disease of soybeans, as well as numerous other crops. The pathogen was first reported in SA in 1979. However, it was only in 2002 that this fungus was considered a major pathogen of soybeans in SA. The research reported in this thesis was conducted to investigate the epidemiology of S. sclerotiorum and examine numerous potential control methods for this pathogen, i.e., resistant cultivars, biocontrol, chemical control and seed treatments. A S. sclerotiorum isolate was obtained from sunflowers in Delmas, Mpumulanga, SA, in the form of sclerotia. This isolate was cultured and sent for identification and deposition in the Plant Protection Research Institute collection. This isolate, in the form of mycelia, was used for the duration of the study. For epidemiology studies, the effect of temperature, leaf wetness duration (LWD) and relative humidity (RH) were examined for their effect on rate of pathogen development. Twenty four combinations of temperature (19°C, 22°C, 25°C and 28°C), LWD (24, 48 and 72 hr) and RH (85 and 95%) were investigated. No interaction between temperature, LWD and RH was found. Temperature alone was the only factor that affected disease development. At 22°C, the rate of pathogen development (0.45 per unit per day) was significantly higher than all other temperatures, indicating that this temperature is optimum for disease development. Thirteen different soybean cultivars, i.e., LS6626RR, LS6710RR, LS666RR, LS555RR, LS6514RR, LS678RR, Prima 2000, Pan 626, AG5601RR, AG5409RR, 95B33, 95B53 and 96B01B, commercially grown in SA were investigated for their reaction to S. sclerotiorum. Prima 2000, 96B01B, 95B33 and AG5409RR were considered to be the least susceptible as they showed a significantly low rate of pathogen development (0.28, 0.28, 0.24, 0.23 per unit per day, respectively) and produced a significantly low number of sclerotia (3.03, 3.42, 3.21, 2.38, respectively). LS6626R and LS666RR may be considered most susceptible because of their significantly high rate of pathogen development (0.45 and 0.42 per unit per day, respectively) and high sclerotia production (8.16 and 7.50, respectively). Regression analysis showed a positive correlation coefficient (R2=0.71) between rate of growth of the pathogen and number of sclerotia produced, indicating that a higher rate is associated with a higher number of sclerotia. In vitro dual culture bioassays were performed to identify the biocontrol mechanisms of the biocontrol agents, EcoT® (a seed treatment) and Eco77® (a foliar treatment), against hyphae and sclerotia of S. sclerotiorum. Ultrastructural studies revealed that mycoparasitism is the probable mode of action as initial signs of hyphae of EcoT® and Eco77® coiling around hyphae of S. sclerotiorum were observed. Surface colonization of sclerotia by hyphae of EcoT® and Eco77® was also observed. In vitro antagonism of EcoT® against S. sclerotiorum on soybean seed was performed to determine pre-emergence and post-emergence disease. There was no significant difference in percentage germination between seeds treated with EcoT® and plated with the pathogen, untreated seeds and no S. sclerotiorum, and the control (i.e. no EcoT® and no pathogen). However, percentage non infected seedlings from seeds not treated with EcoT® was significantly lower, suggesting that EcoT® may be successfully used as a seed treatment for the control of SSR. In vivo trials were performed to investigate the effect of silicon (Si) alone, and in combination with Eco77®, on the effect of the rate of disease development. Plants treated with Eco77® had a significantly lower rate of disease development (0.19 per unit per day for plants treated with Eco77® and S. sclerotiorum and 0.20 per unit per day for plants treated with Eco77®, S. sclerotiorum and Si), compared to plants not treated with Eco77® (0.29 per unit per day for plants treated with S. sclerotiorum and 0.30 per unit per day for plants treated with S. sclerotiorum and Si), regardless of the application of Si. Similarly, plants treated with Eco77® had a significantly lower number of sclerotia (0.46 for plants treated with Eco77® and S. sclerotiorum and 0.91 for plants treated with Eco77®, S. sclerotiorum and Si), compared to plants not treated with Eco77® (3.31 for plants treated with S. sclerotiorum and 3.64 for plants treated with S. sclerotiorum and Si). The significantly lower rate of disease development coupled with a significant reduction in sclerotia showed that Eco77®, and not Si, was responsible for reducing the severity of SSR. A strong positive correlation between rate of disease development and the number of sclerotia produced (R2=0.79) was observed. For the investigation of various fungicides for the control of S. sclerotiorum, in vitro trials to determine the potential of three different fungicides at different rates, i.e., BAS 516 04F (133 g a.i. ha-1), BAS 516 04F (266 g a.i. ha-1), BAS 512 06F (380 g a.i. ha-1) and Sumisclex (760 g a.i. ha-1) were initially conducted. The control (non-amended PDA) had a significantly higher area under mycelial growth curve (243.0) than all fungicides tested. BAS 516 04F (at both concentrations) and BAS 512 06F completely inhibited the mycelial growth of S. sclerotiorum. Sumisclex inhibited the fungus by 89.07%. For in vivo trials, preventative treatments, i.e., BAS 516 04F (133 g a.i. ha-1), BAS 516 04F (266 g a.i. ha-1), BAS 512 06F (380 g a.i. ha-1), curative treatment, i.e. Sumisclex (760 g a.i. ha-1) and a combination preventative/curative treatment, i.e., BAS 512 06F (380 g a.i. ha-1)/Sumisclex (570 g a.i. ha-1) were investigated. No significant difference in disease severity index (DSI) was found between fungicide treatments and the inoculated control. BAS 512 06F and BAS 512 06F/Sumisclex had significantly lower grain yields (6.09 g and 5.96 g, respectively) compared to all other treatments. There was a positive correlation coefficient (R2=0.76), between DSI and grain yield, indicating that a high DSI is correlated with low grain yield. Trials to evaluate the effect of commercially available and currently unregistered seed treatments for the control of S. sclerotiorum on soybean seeds in vivo and in vitro were performed. Seed germination tests were performed to determine if seed treatments had any negative effects on seed germination in vitro. All seed treatments tested, i.e., BAS 516 03F (8, 16 and 32 ml a.i. 100 kg-1 seed), BAS 512 00F (7.5, 15 and 32 ml a.i. 100 kg-1 seed), Celest XL (100, 125, 200 and 250 ml a.i. 100 kg-1 seed), Sumisclex (5 and 10 ml a.i. 100 kg-1 seed), Benomyl (150 g a.i. 100 kg-1 seed), Captan (240 ml a.i. 100 kg-1 seed), Thiulin (180 g a.i. 100 kg-1 seed) and Anchor Red (300 ml a.i. 100 kg-1 seed), showed no negative effect on seed germination. For in vivo trials, BAS 516 03F (16 and 32 ml a.i. 100 kg-1 seed), BAS 512 00F (7.5, 15 and 32 ml a.i. 100 kg-1 seed), Celest XL (100, 125, 200 and 250 ml a.i. 100 kg-1 seed), Sumisclex (5 and 10 ml a.i. 100 kg-1 seed), Benomyl and Anchor Red had significantly similar percent germination and percent seedling survival as the untreated/uninoculated control. These seed treatments should be recommended for the control of S. sclerotiorum, as they protected seed during germination and subsequent seedling development. BAS 516 03F (8 ml a.i. 100 kg-1 seed) should not be recommended for the control of SSR, as it gave the lowest percent germination and percent seedling survival. The results presented in this thesis have helped to identify optimal environmental conditions for the development of S. sclerotiorum, which is important for the development of forecasting models for disease control. The least and most susceptible cultivars of those tested have been identified. Biocontrol using Eco77® as a foliar application showed great potential. The effect of Si needs to be further investigated, including the testing of more frequent applications and higher concentrations. The fungicides tested in this research did not show any potential for the control of SSR. However, the spray programme tested is for the control of soybean rust (Phakopsora pachyrhizi), and was investigated for its potential for the control of SSR. The spray programme, fungicide application and rating scale needs to be modified, to determine the true potential of these fungicides for the control of SSR. Numerous seed treatments have shown potential for the control of seed infection by SSR. Due to difficulties in producing ascospores, which are the primary source of inoculum for this pathogen in the field, all studies in this research were conducted with mycelia and not ascospores. The production, collection and storage of ascospores needs to be thoroughly investigated, and research conducted with ascospores. / Thesis (M.Sc.)-University of KwaZulu-Natal, Pietermaritzburg, 2007.

In the KwaZulu-Natal (KZN) Midlands North region of South Africa, the importance and increased prevalence of endemic scarabaeids, particularly Hypopholis sommeri Burmeister and Schizonycha affinis Boheman (Coleoptera: Melolonthinae), as soil pests of sugarcane, and a need for their control was established. The development of a mycoinsecticide offers an environmentally friendly alternative to chemical insecticides. The identification of a diversity of white grub species, in two Scarabaeidae subfamilies, representing seven genera were collected in sugarcane as a pest complex. Hypopholis sommeri and S. affinis were the most prevalent species. The increased seasonal abundances, diversity and highly aggregated nature of these scarabaeid species in summer months, suggested that targeting and control strategies for these pests should be considered in this season. Increased rainfall, relative humidity and soil temperatures were linked to the increased occurrence of scarab adults and neonate grubs. Beauveria brongniartii (Saccardo) Petch epizootics were recorded at two sites in the KZN Midlands North on H. sommeri. Seventeen different fluorescently-labelled microsatellite PCR primers were used to target 78 isolates of Beauveria sp. DNA. Microsatellite data resolved two distinct clusters of Beauveria isolates which represented the Beauveria bassiana senso stricto (Balsamo) Vuillemin and B. brongniartii species groups. These groupings were supported by two gene regions, the nuclear ribosomal Internal Transcribed Spacer (ITS) and the nuclear B locus (Bloc) gene of which 23 exemplar Beauveria isolates were represented and sequenced. When microsatellite data were analysed, 26 haplotypes among 58 isolates of B. brongniartii were distinguished. Relatively low levels of genetic diversity were detected in B. brongniartii and isolates were shown to be closely related. There was no genetic differentiation between the two sites, Harden Heights and Canema in the KZN Midlands North. High gene flow from swarming H. sommeri beetles is the proposed mechanism for this lack of genetic differentiation between populations. Microsatellite analyses also showed that B. brongniartii conidia were being cycled from arboreal to subterranean habitats in the environment by H. sommeri beetles. This was the first record of this species of fungus causing epizootics on the larvae and adults of H. sommeri in South Africa. The virulence of 21 isolates of Beauveria brongniartii and two isolates of B. bassiana were evaluated against the adults and larvae of S. affinis and the adults of H. sommeri and Tenebrio molitor Linnaeus (Coleoptera: Tenebrionidae). Despite being closely-related, B. brongniartii isolates varied significantly in their virulence towards different hosts and highlighted the host specific nature of B. brongniartii towards S. affinis when compared to B. bassiana. Adults of S. affinis were significantly more susceptible to B. brongniartii isolates than the second (L2) or third instar (L3) grubs. The median lethal time (LT₅₀) of the most virulent B. brongniartii isolate (C13) against S. affinis adults was 7.8 days and probit analysis estimated a median lethal concentration (LC₅₀) of 4.4×10⁷ conidia/ml⁻¹. When L2 grubs were treated with a concentration of 1.0×10⁸ conidia/ml⁻¹, B. brongniartii isolates HHWG1, HHB39A and C17 caused mortality in L2 grubs within 18.4-19.8 days (LT₅₀). Beauveria brongniartii isolate HHWG1 was tested against the L3 grubs of S. affinis at four different concentrations. At the lowest concentration (1×10⁶ conidia/ml⁻¹), the LT₅₀ was 25.8 days, and at the highest concentration (1×10⁹ conidia/ml⁻¹) the LT₅₀ dropped to 15.1 days. The persistence of B. bassiana isolate 4222 formulated on rice and wheat bran and buried at eight field sites in the KZN Midlands North was evaluated by plating out a suspension of treated soil onto a selective medium. All eight field sites showed a significant decline in B. bassiana CFUs per gram of soil over time, with few conidia still present in the samples after a year. Greater declines in CFUs were observed at some sites but there were no significant differences observed in the persistence of conidia formulated on rice or wheat bran as carriers. Overall, poor persistence of B. bassiana isolate 4222 was attributed to suboptimum temperatures, rainfall, which rapidly degraded the nutritive carriers, attenuated fungal genotype and the action of antagonistic soil microbes. Growers’ perceptions of white grubs as pests and the feasibility of a mycoinsecticide market were evaluated by means of a semi-structured questionnaire. The study showed that the reduced feasibility of application, general lack of potential demand for a product, high cost factors and most importantly, the lack of pest perception, were factors which would negatively affect the adoption of a granular mycoinsecticide. Growers however exhibited a positive attitude towards mycoinsecticides, and showed all the relevant attributes for successful technology adoption. It is recommended that because B. brongniartii epizootics were recorded on target pests which indicated good host specificity, dispersal ability and persistence of the fungus in the intended environment of application; that a mycoinsecticide based on this fungal species be developed. What will likely increase adoption and success of a mycoinsecticide is collaboration between various industries partners to increases market potential in other crops such as Acacia mearnsii De Wild (Fabales: Fabaceae).