Integrated management of Campuloclinium macrocephalum (pompom weed) (Less.) DC. (Asteraceae) in South Africa

Abstract

Campuloclinium macrocephalum (Less.) DC. (Asteraceae) is an invasive perennial forb native to Central and South America and known as pompom weed in South Africa. It has since naturalized and poses a threat to grasslands, wetlands and savannas in many parts of South Africa. The weed is listed as category 1b of NEMBA (National Environmental Management Biodiversity Act), an invasive species which must be controlled and wherever possible, removed and destroyed. It causes significant reduction in grazing capacity of farms and game reserves, as it is unpalatable to livestock and wildlife. The weed has a clustered thick tuber-like root system that enables it to withstand fire and harsh winter conditions such as low rainfall and frost. Pompom has increased its invasion range by 671% between 2006 and 2016. Management strategies to date, such as the application of chemical, fire, mechanical and biological control have not yielded substantial results, hence the increase in invasion range of C. macrocephalum. Application of herbicides alone requires multiple annual applications which may not be feasible, logistically or financially. It is also not ideal to continuously apply herbicides in sensitive environments such as areas that are close to streams, wetlands and dams. Three annual applications of metsulfuron methyl and picloram alone were found to suppress the abundance of C. macrocephalum but failed to control the weed. Mechanical control of C. macrocephalum through hoeing or digging is ineffective as damage to rootstocks by splitting them leads to further vegetative growth resulting in an increase in pompom density. Multiple biological control agents are often released against a single invasive plant to improve the degree of control. There are three biological control agents released against C. macrocephalum in South Africa namely, Liothrips tractabilis Mound and Pereyra (Thysanoptera: Phlaeothripinae), Puccinia eupatorii (Pucciniales: Pucciniaceae) and a flower-feeder, Cochylis campuloclinium Brown (Lepidoptera: Tortricidae). Damage inflicted by L. tractabilis and P. eupatorii is moderate, and the degree of control provided by the two agents has not resulted in sufficient control of the weed. A strategy of combining different control methods such as chemical, mechanical, fire or biological control can improve the impact and effectiveness of any combination of the four methods and improve the degree of control of a particular weed. Integrated management of invasive weeds vary to suit a target plant based on invaded habitat, existing control method and management objectives. The first aim of the study (Chapter 2) was to evaluate how burning alone prior to the growing season of C. macrocephalum and burning alongside herbicide application affects the weed density, regrowth/re-sprouting from rootstocks, flowering density and the seedbank of the weed over two growing seasons at Rietvlei Nature Reserve, Gauteng Province, South Africa. The seedbank was estimated by quantifying the number of newly emerging seedlings on each experimental block. During the first growing season (2020-2021) fire in August significantly reduced the seedbank of pompom by over 85% but failed to reduce the density of mature plants within the invaded areas. Herbicide application in December 2020 post fire of the first growing season resulted in over 98% reduction of the weed density during the first growing season. For the second growing season (2021-2022), there was a 9% regrowth from rootstocks of the weed on experimental blocks previously treated with fire and herbicide, which can be managed by a follow up application of herbicide. Integrating fire and herbicide was found to be a successful management strategy and it is therefore recommended as a management approach for the control of pompom weed in South Africa. The impact of L. tractabilis is variable at different sites invaded by C. macrocephalum. This is because L. tractabilis has established in some open grasslands and failed to establish in more mesic environments particularly seasonal wetlands. A sub-lethal dose of herbicide can be used to suppress the reproductive ability of the weed without killing the plant and sustain the biocontrol population. Therefore, the second aim (Chapter 3) was to identify sub-lethal doses of three herbicides (picloram, impala and metsulfuron methyl) that will not kill but suppress the vegetative growth of C. macrocephalum and sustain the L. tractabilis population. However, the identified sub-lethal doses of the three herbicides were found to be incompatible with the agent L. tractabilis, because of an indirect impact on their feeding sites (young and apical stem sections) which affected the abundance of the agent. Due to lack of feeding sites, L. tractabilis was found to feed on old stems which are the least preferred sites. Consequently, considering the indirect impact of sub-lethal doses on L. tractabilis, the use of such doses is not recommended given the incompatibility. In a study in 2018, early season mowing of surrounding vegetation, led to enhanced effectiveness of L. tractabilis on pompom weed, resulting in increased levels of stem deformation by the agent. Campuloclinium macrocephalum in more sunny vegetation (areas with mown short vegetation) experienced increased light intensity and temperatures compared to areas with unmown long vegetation (more shady sites). However, it remains unclear how the physiology and growth rate of C. macrocephalum were affected by the difference in light intensities and temperature and how subsequent changes in plant growth and ecophysiology affected the biocontrol agents. The third aim of this study (Chapter 4) was to evaluate the influence of two light intensities (full sunlight and reduced sunlight) on chlorophyll content, chlorophyll fluorescence, stomatal conductance, and growth of C. macrocephalum, and to assess the reproduction and feeding impact of L. tractabilis in a laboratory trial. Reduced light intensity (shade plants) resulted in higher chlorophyll content, chlorophyll fluorescence, stomatal conductance, growth rate and chlorophyll fluorescence than in sun plants suggesting increased photosynthetic rates. This resulted in a significant increase in growth rate, with greater height of C. macrocephalum in the shade (108 cm) compared to in the sun (72 cm) over seven weeks. The feeding performance and abundance of L. tractabilis was significantly higher in open sun plants. This suggests that light intensity plays a critical role in the performance and establishment of the thrips. Therefore, mowing or burning the surrounding vegetation prior to and post release of the thrips may aid establishment and improve efficacy of the agent. Liothrips tractabilis is known to prefer young apical shoot stems compared to more mature and old stems. To understand why L. tractabilis preferred young stems and plants in short vegetation, the fourth aim of the study (Chapter 5) was to evaluate and compare the nutritional quality of young apical stems compared with old stems (mid-section) of C. macrocephalum that grew in long and short vegetation by assessing the nitrogen and carbon levels and C:N ratio. Nitrogen content was greater in apical stems than in mid-section stems from short vegetation, suggesting better nutritional quality for L. tractabilis compared to plants from long vegetation. Therefore, the surrounding vegetation has an impact on the nutritional quality of C. macrocephalum which affects the performance and population size of L. tractabilis. In conclusion, neither fire alone nor herbicide alone significantly reduce the population density of pompom weed, but the application of lethal herbicide to C. macrocephalum after fire proved to be the most effective control method.