AbstractTaking an ecological intensification approach to weed management could increase agriculture’s ability to meet human needs for food, fuel and fibre whilst avoiding impacts on the integrity of Earth’s life-supporting ecosystems. Ecological intensification would achieve this by replacing anthropogenic inputs with ecosystem function. In this thesis, ecosystem processes are sought that could replace conventional weed control inputs and actions such as herbicides and tillage, and so reduce the environmental impacts associated with these. First, there is a need to shift thinking away from isolated mechanisms that seek to remove weeds, and to instead aim for ‘agroecosystem resilience to weeds’ in which farming systems would be designed to incorporate ecological processes and properties that inherently limit the negative impacts and promote the positive roles of weeds. To understand how this resilience could be achieved, recent advances in weed ecology are reviewed to identify relevant processes and properties, and to consider how these might be implemented in farm design and management. In sum, practices that could increase the diversity of filters applied to weed communities while decreasing filter strength, and that could reduce resource availability, would be expected to confer agroecosystem resilience to weeds.
This thesis explores the practical implementation of some of these practices in South Africa’s winter rainfall region through field studies and trials. The practices were chosen for their suitability for conservation agriculture systems (the dominant farm management style in the region) and for their potential to harness or enhance ecological processes for weed management. The first practical study, a field survey of weeds in 15 vineyards with differing weed management practices, confirmed that the use of management techniques imposing lower disturbance leads to more diverse weed communities composed of less competitive species, and this effect can be enhanced by using specific management techniques to select for specific weed traits. The second practical study was linked to a long-term crop rotation experiment, and explored the effects of increasing crop diversity and integrating livestock as methods to increase the variability experienced by weeds in these rotation systems. Combining these two practices substantially reduced weed abundance and conserved weed diversity over the twelve years investigated. They also reduced herbicide and fertiliser requirements, and sustained cash crop yields, thus contributing to both profitability and sustainability. The final practical study applied theories of biotic resistance from invasion ecology to investigate how best to design cover crop mixes for weed suppression, that could be used in field crop systems or vineyards. Mixes composed of highly productive species were most effective at resource capture, and thus most effective at reducing resource availability to suppress weeds. In sum, reducing herbicide use in favour of grazing or mowing, increased crop and management diversity, and competitive crops (in mixes or monoculture) are viable practices that constitute first steps toward the ecological intensification of weed management in South Africa’s winter rainfall region. They offer improvements to agricultural sustainability through sustaining yields and farm incomes whilst reducing the environmental impacts and health risks associated with conventional weed control such as herbicides and tillage. Overall, the findings of this thesis suggest that ecological intensification offers a promising direction for future weed management to achieve agricultural sustainability, both in South Africa’s winter rainfall region and around the world. Weed researchers can assist farmers in this challenge by drawing on global advances in weed ecology to design and test locally appropriate weed management techniques and strategies.
|Date of Award||2018|
|Supervisor||Katharina Dehnen-Schmutz (Supervisor), James Bennett (Supervisor) & Julia Wright (Supervisor)|