Biological control of torch cactus in South Africa: finding a suitable agent for a non-native weed with an unknown indigenous distribution
- Authors: Griffith, Tamzin Camilla
- Date: 2024-10-11
- Subjects: Trichocereus spachianus , Invasive plants Biological control , Dactylopius , Hypogeococcus , Host specificity
- Language: English
- Type: Academic theses , Doctoral theses , text
- Identifier: http://hdl.handle.net/10962/466781 , vital:76778 , DOI https://doi.org/10.21504/10962/466781
- Description: Trichocereus spachianus is an invasive cactus species in South Africa and poses challenges as a target for biological control due to confusion around its taxonomy and origin. Adapted to arid environments, this cactus is of particular concern in dry savannah and Karoo biomes, where its invasion of rangelands reduces grazing capacity for both indigenous wildlife and livestock. While previous records indicate that T. spachianus is indigenous to Argentina, recent field surveys have failed to verify its presence. Determining the origin of the target weed was important in developing a biological control programme since it enables the collection of potential agents directly from native populations of the target plant. Successful biological control programmes against cactus species in South Africa have often involved utilising both new associations and oligophagous insects, made possible because of the lack of indigenous and valued cacti in the region. Lack of T. spachianus locations in the native distribution, meant direct collection of insects from the target weed was not possible. Efforts to find biological control agents were focused on new associations between closely related cacti and their oligophagous herbivores. Suitability of multiple Hypogeococcus (mealybug) entities and a cochineal species, Dactylopius confertus, were investigated for their efficacy on various South African weedy cactus species, including T. spachianus. Findings revealed that none of the Hypogeococcus entities (species or lineages) were effective biological control agents, but D. confertus demonstrated potential as a biological control agent with a relatively high fecundity and survival rate on T. spachianus. Efficacy trials indicated that D. confertus could reach population densities sufficient to cause mortality of T. spachianus plants. Limited host specificity trials revealed that D. confertus was suitably host specific for release in South Africa, provided host specificity testing was conducted on additional plants. Approval and release of D. confertus has the potential to reduce the invasiveness of this damaging cactus in a sustainable and environmentally friendly manner. , Thesis (PhD) -- Faculty of Science, Zoology and Entomology, 2024
- Full Text:
- Date Issued: 2024-10-11
Developing biological control agents for the management of the invasive tree Robinia pseudoacacia
- Authors: Wolmarans, Abigail
- Date: 2024-04-05
- Subjects: Black locust South Africa , Invasive plants Biological control , Insects as biological pest control agents , Prioritization , Vegetative propagation , Locust leafminer
- Language: English
- Type: Academic theses , Doctoral theses , text
- Identifier: http://hdl.handle.net/10962/435717 , vital:73181 , DOI 10.21504/10962/435717
- Description: Robinia pseudoacacia (Fabaceae) is a deciduous tree native to the Appalachian Mountains of North America but has become naturalised and invasive in other countries such as temperate North America, Europe, Australia, and Southern Africa. In South Africa the tree is classified as a category 1B invasive alien under the National Environmental Management Act (NEMBA), which stipulates the species requires some form of control as it has already caused extensive negative ecological and economic impacts. In the invaded range the tree creates monocultures that displace native species and spreads rapidly from suckering roots, making it a proficient invader. The South Africa plant prioritisation system suggests R. pseudoacacia is in the top three species which should be considered for classical weed biological control in South Africa. This thesis investigates which insects known to be associated with tree should be prioritised as candidate agents, as well as offering interesting insights into prioritising insects for weed biological control and using plant phylogenies and available literature to predict insect specificity. To ensure that no candidate biological control agents were already present in South Africa as well as to prioritise which guilds of the tree to prioritise for potential biological control, pre-release surveys were conducted across nine sites where the tree has invaded South Africa. It was found that no insects from the native range of R. pseudoacacia were present in South Africa. Seed surveys revealed that generalist insects attack a sizable proportion (68 %) of the seeds on the trees. In combination with a low seed soil bank (15.8 %) this suggests that seed- feeding agents may be helpful, however, candidate agents which damage leaves should be prioritised due to R. pseudoacacia relying heavily upon vegetative reproduction and much less on sexual reproduction. Leaves may therefor reduce the spread of these invasive trees. The insect assemblages in the native range of R. pseudoacacia are well understood. In addition, several associated insects have unintentionally followed the tree on its global spread, where they are often regarded as pests. The third chapter is therefore aimed at prioritising the known insects associated with the tree in both the native and invaded range. Literature surveys and Harris (1973) prioritisation systems were used to prioritise close to 64 candidate biological control agents down to three foliage- feeding agents, namely Odontota dorsalis (Coleoptera, Chrysomelidae), Macrosaccus robiniella (Lepidoptera, Gracillariidae) and Obolodiplosis robiniae (Diptera, Cecidomyiidae). To further prioritise the six selected agents in Chapter 4, species distribution were modelled with known climatic variables. This was done by using the species known occurrence localities, from both the native and where applicable invaded ranges, to identify which species that would best match with South Africa’s climate. The study showed that O. dorsalis is best suited to survive in South Africa, followed by O. robiniae and then M. robiniella. Collected data aided in the introduction of the prioritised O. robiniae into South African quarantine facilities. Unfortunately, despite several attempts, cultures could not be established, making conventional host specificity testing impossible. Thus, in Chapter 5 we aimed to determine the potential host range of the midge using information from experts in the field of galling insects, literature surveys, agricultural pest lists, and social science platforms coupled with native and invaded range surveys. The list of non-target species to consider as potential hosts was refined by developing phylogenetic trees of closely related Fabaceae that share the same distribution (native and invaded) as R. pseudoacacia. Through the available information gathered, and field surveys of these species, O. robiniae has not been shown to utilise any species, other than those from the Robinoid clade. In addition, the midge has also never been recorded on a number of closely related leguminous fodder and horticultural species growing in close proximity at high densities to R. pseudoacacia - suggesting negligible risk to South African growers of the same species. Host-specificity assessments through field surveys can be regarded as one of the best indicators of the ecological host range, however, this information is difficult to quantify and infrequently available, thus seldom used when determining the safety of a candidate biocontrol agent. In this unique study, using the extensive data collected we are able to show that O. robiniae would be potentially safe for release in South Africa. However, open field tests exposing closely related non-target plant species under natural and semi-natural conditions are planned in Switzerland over the next two years aiming to confirm these conclusions. By identifying and prioritizing potential biological control agents, this research contributes to the development of a targeted and sustainable solution for managing R. pseudoacacia in South Africa. The economic implications of successful biological control include reduced costs associated with invasive species management and ecosystem restoration. Additionally, by mitigating the negative ecological impacts, the agricultural systems in affected regions stand to benefit from enhanced resilience and productivity. Furthermore, the study's approach of using plant phylogenies and available literature to predict insect specificity offers a valuable methodological contribution to the broader field of weed biological control. This methodology can be adapted and applied to other invasive species, providing a framework for efficient and informed decision-making in weed management strategies. In essence, this research not only addresses the specific challenges posed by R. pseudoacacia in South Africa but also provides a template for tackling similar issues in different geographical contexts, thereby contributing to global efforts in sustainable agriculture and environmental conservation. , Thesis (PhD) -- Faculty of Science, Zoology and Entomology, 2024
- Full Text:
- Date Issued: 2024-04-05
Developing a community of practice to promote the use of biological control in the integrated management of Prosopis in South Africa
- Authors: Van Staden, Gretha
- Date: 2024-04-04
- Subjects: Mesquite South Africa Northern Cape , Community of practice , Stakeholder participation , Mesquite Control , Invasive plants Biological control , Prosopis
- Language: English
- Type: Academic theses , Master's theses , text
- Identifier: http://hdl.handle.net/10962/434998 , vital:73122
- Description: Prosopis spp. are non-native species present in the arid parts of southern Africa. These trees originated in the Americas and were first introduced as a source of fodder, shade, and wood, but have become invasive, negatively impacting local biodiversity, and disrupting the delivery of ecosystem services. Some species of Prosopis hybridise freely, complicating identification and subsequent control. The control of Prosopis in the Northern Cape Province of South Africa is still contentious, because of the perceived benefits to some landowners. The effectiveness of biological control agents to control the spread of Prosopis has been quantified, more damaging agents have been considered in recent years because of the continued increase in Prosopis density. Research into the establishment of a Community of Practice (CoP) to address the differences in perceptions regarding the control of Prosopis in the Northern Cape Province is considered an appropriate starting point to conceptualise the challenges to the successful integrated management of Prosopis. As farmers are the main stakeholders involved with Prosopis control, understanding the role of farmers and the functioning of farming enterprises in the Northern Cape is imperative for the development of both a CoP as well as the appropriate management of Prosopis. To foster engagement with as many stakeholders as possible, workshops dealing with biological control as part on an integrated approach to the control of Prosopis were held in towns in the Northern Cape Province: Groblershoop, Brandvlei, Kenhardt, Upington, Prieska and Williston. Some of the main concerns of the landusers include the host specificity of the released biocontrol agents, as well as increased transparency and communication. This is especially in regards to the research on present and future biological control agents that will possibly improve the perceptions of stakeholders. The mechanical harvesting and use of biomass as a control method was supported, especially where farmers were removing Prosopis from their properties, leaving large amounts of biomass in the veld. Continued research into Prosopis in the Northern Cape needs to consider the possibility of secondary invasions, especially in areas where native bush encroachment is already a concern. The impact of clearing in terms of natural grazing and animal improvement needs to be quantified, to conceptualise the importance of maintaining better veld quality versus using Prosopis for fodder. Biological control targeting not only the reproductive output of the trees, but also the standing biomass has support from the majority of the land users in the Northern Cape Province following these workshops and this study shows the importance of a socio-ecological approach to the control of landscape scale invasion. , Thesis (MSc) -- Faculty of Science, Zoology and Entomology, 2024
- Full Text:
- Date Issued: 2024-04-04
A molecular investigation of stem-galling Tetramesa Walker (Hymenoptera: Eurytomidae) on African grasses: applications to biological control
- Authors: Van Steenderen, Clarke Julian Mignon
- Date: 2023-10-13
- Subjects: Grasses Africa , Tetramesa , Invasive plants Biological control , DNA barcoding , Weeds Biological control , Eragrostis curvula , Sporobolus pyramidalis , Sporobolus natalensis
- Language: English
- Type: Academic theses , Doctoral theses , text
- Identifier: http://hdl.handle.net/10962/432565 , vital:72881 , DOI 10.21504/10962/432565
- Description: South Africa is a larger donor than receiver of alien grasses, where approximately 15% (_ 165 spp.) of the country's native grass species have become naturalised elsewhere. Many of these grasses have become serious invaders, causing significant damage to native species, habitat structure, and ecosystem functioning. Biological control is a sustainable and cost-effective method for the control of invasive weeds, but its application to invasive grasses has been approached with trepidation in the past due to the fears of a lack of host-specific herbivores that may cause non-target damage to agriculturally-important crops. The Tetramesa Walker genus (Hymenoptera: Eurytomidae) is one of three genera in the family that feed exclusively on grasses, and have a record of being host-specific to a particular species, or complex of closely-related congeners. There are over 200 described Tetramesa species, but this taxonomic effort has occurred almost exclusively in the Northern Hemisphere. Only about 2% of the described species are from Africa, with none from southern Africa despite the high diversity of grasses in the region. The low morphological variability between many Tetramesa groups has made identification difficult, where there may in fact be multiple undiscovered cryptic species. This thesis generated genetic sequence data (mitochondrial COI and nuclear 28S) that revealed at least eight native southern African Tetramesa taxa that are new to science, focusing particularly on the assemblages associated with Eragrostis curvula Nees (African lovegrass) and Sporobolus pyramidalis Beauv. and S. natalensis Steud. (giant rat's tail grass) which are alien invasive pests in Australia. Approximately 200 eurytomid wasps were collected and sequenced from 19 grass species across six South African provinces. Additionally, 27 grass species were sequenced using four gene regions (rps16-trnK, rps16, rpl32-trnL, and ITS), which were added to existing sequence data to build a dataset comprising over 700 sequences. Field host ranges and the use of host grass genetic proxies were important in making inferences about the host-specificity of eurytomid wasps of interest. Nine Tetramesa groups appeared to be host-specific to a single grass species, while six Tetramesa were associated with multiple species in a single grass genus. Since S. pyramidalis, S. natalensis, S. africanus, Hyparrhenia hirta, E. trichophora, and Andropogon gayanus are weeds elsewhere, there are at least six potential Tetramesa biological control agents that have been identified. A high diversity of Tetramesa on grasses within the Eragrostis genus was reported, with at least four taxa associated with E. curvula. It is currently uncertain whether these taxa represent different cryptic species or intraspecific populations that are the result of geographic sub-structuring. No-choice host-specificity testing using Tetramesa sp. 4 on E. curvula revealed that the wasp could complete its lifecycle on two non-target African grasses; namely E. plana and E. planiculmis. The wasp was also recorded on other Eragrostis species in the field (namely E. biflora and E. capensis). Using grass genetic sequences obtained in this study, it was found that there are four native Australian Eragrostis species that are more closely related to target E. curvula than to the non-target E. plana and E. planiculmis. This suggests that Tetramesa sp. 4 may not be suitably host-specific for use as a biological control agent. Further host-specificity testing on these native Australian species is required, however, before this insect is ruled out completely. The Tetramesa on S. pyramidalis (Tetramesa sp. 1), and the unidentified Sporobolus species presumed to be S. africanus, were suitably host-specific to be used as biological control agents. Since it was unclear whether some phylogenetic clades were true species or intraspecific populations, which is essential to understand when selecting agents for biological control, a new piece of software, SPEDE-sampler", was developed. It offers users of the Generalised Mixed Yule Coalescent (GMYC) species delimitation model a means of assessing the degree to which sampling effects such as data size and parameter choice can influence species diversity estimates. When applied to the Tetramesa data set, the software assisted in identifying which groups may contain cryptic species, uncovering that the COI marker is affected more by singletons than the 28S marker (i.e. species diversity tends to be overestimated), and confirming putative Tetramesa taxa that could be useful for biological control programmes going forward. This thesis has provided evidence that South Africa contains a diverse assemblage of Tetramesa and other eurytomids that are closely associated with their grass hosts, and that many of these taxa hold promise for grass biological control. This work has also highlighted the importance of integrative taxonomy in the discovery of novel taxa, and that biological control practitioners need to be aware of the caveats of each line of evidence used in the delimitation of putative species. , Thesis (PhD) -- Faculty of Science, Zoology and Entomology, 2023
- Full Text:
- Date Issued: 2023-10-13
The effects of elevated CO2 on feeding guild responses of biological control agents of Pontederia crassipes Mart. (Pontederiaceae)
- Authors: Paper, Matthew Keenan
- Date: 2022-04-06
- Subjects: Carbon dioxide , Pontederia crassipes , Biological pest control agents , Invasive plants Biological control , Pontederiaceae Climatic factors
- Language: English
- Type: Academic theses , Master's theses , text
- Identifier: http://hdl.handle.net/10962/455338 , vital:75422
- Description: Elevated CO2 (eCO2) and rising global temperatures have the potential to alter plant-insect interactions with important implications for plant community structure and function. Previous studies on plant-insect interactions have shown that eCO2 will affect insect feeding guilds differently, impacting negatively, positively or having very little effect. The implications of this on the global invasive plant biological control programme is largely unknown. This study investigates the response of one of the world’s most invasive aquatic plants, Pontederia ( = Eichhornia) crassipes Mart. (Pontederiaceae), to predicted eCO2 conditions of 800 ppm and how this may affect the feeding response of two biological control agents representing different feeding guilds; the leaf chewing Cornops aquaticum Brüner (Orthoptera: Acrididae) and the phloem-feeding Megamelus scutellaris Berg (Hemiptera: Delphacidae). A factorial eCO2 x feeding impact study was conducted at the Rhodes University Elevated CO2 Facility in the Eastern Cape Province of South Africa over 13 weeks in the growing season of 2019. The effect of insect herbivory by C. aquaticum and M. scutellaris at two atmospheric CO2 concentrations, representing current and future predicted concentrations (400 ppm and 800 ppm) on P. crassipes was examined through both biomass and ecophysiological measures. Assimilation rates, C:N ratio, total dry weight and relative growth rate of P. crassipes were unaffected by eCO2 conditions, and plants experienced no CO2 fertilization in eutrophic water conditions representative of South African waterways. Effects of eCO2 on insect herbivory varied depending on the feeding guild. Pontederia crassipes showed compensatory growth responses when exposed to C. aquaticum herbivory regardless of CO2 treatment, but chewing herbivory damage remained constant, and the agent maintained efficacy. Pontederia crassipes showed down-regulation of photosynthesis when exposed to M. scutellaris due to eCO2-related feeding responses by M. scutellaris increasing substantially through a significant (30%) increase in adult population density under eCO2 conditions. These results indicate that the plant-insect interactions that underpin biological control programmes for P. crassipes should remain successful under future CO2 conditions. Phloem-feeding insect damage (M. scutellaris) was significantly greater than chewing damage in this study, suggesting that invasive plant biological control programmes will need to shift focus away from the charismatic chewing insect herbivores and onto the often-neglected phloem-feeding biological control agents due to their overwhelmingly positive response to eCO2. , Thesis (MSc) -- Faculty of Science, Zoology and Entomology, 2022
- Full Text:
- Date Issued: 2022-04-06