Prioritization of potential biocontrol agents for the invasive alien weed, Cylindropuntia pallida (Cactaceae) in South Africa and insights into the use of cochineal in future biological control programmes

Abstract

Cylindropuntia pallida F.M. Knuth (Cactaceae) is an invasive alien species in Australia, Saudi Arabia, Spain, Namibia, and South Africa. Like other invasive Cylindropuntia species, it has no commercial value and has had numerous negative impacts on the environment and the economy. Since the biocontrol of invasive cactus weeds using cochineal insects has been successful in the past, it is likely that it will be the best form of control against C. pallida. In this thesis, the aims were to prioritize potential biocontrol agents for C. pallida in South Africa as well as to gain insights into how to select the most damaging agent for other cactus weeds in the future. Two lineages of the cochineal insect Dactylopius tomentosus Lamark, were considered as potential biocontrol agents for C. pallida in South Africa. One lineage, referred to as the ‘californica var. parkeri’ lineage had already been released against C. pallida in Australia, where it is an effective agent. The second lineage was collected off C. pallida in Mexico and was imported into quarantine in South Africa and is referred to as the ‘pallida’ lineage. The compatibility of both cochineal lineages with invasive C. pallida from South Africa was tested by comparing cochineal fitness parameters when reared on South African C. pallida. Surprisingly, the ‘californica var. parkeri’ lineage had much lower fitness on South African C. pallida plants compared with Australian plants, whilst the ‘pallida’ lineage had higher fitness on South African plants. These results indicate that the ‘pallida’ lineage is the preferred candidate and should be considered further as a potential agent but also raises interesting questions as to whether ‘old associations’ or ‘new associations’ are better for cactus biocontrol with cochineal lineages. A ‘new association’ in biocontrol is defined as a biocontrol agent that attacks a target weed which it has not co-evolved with (like the ‘californica var. parkeri’ lineage on C. pallida, which was collected off C. bernardina outside of the indigenous distribution of C. pallida). ‘Old associations’, are when the biocontrol agent and weed have a long-standing evolutionary relationship in the indigenous distribution (like the ‘pallida’ lineage on C. pallida, which was collected off C. pallida in its indigenous distribution). To test whether ‘new’ or ‘old associations’ are more likely to be effective in controlling invasive cactus with cochineal, the fitness and performance of three D. tomentosus cochineal lineages on seven different invasive Cylindropuntia species was tested with the aim to assess whether fitness and performance decrease with more genetically distant plant species from the primary host plant. For every fitness performance metric tested in this study, there was a correlation between phylogenetic distance and performance, with fitness measurements decreasing for each cochineal lineage on less closely related species. This study provides evidence that ‘old associations’ should be prioritized for Cactaceae biocontrol agent selection. Fitness assessments like those present in this thesis contribute to predicting the efficacy of an agent as well as help in the agent selection process. This thesis should contribute towards the successful biocontrol of C. pallida in South Africa and improve the selection process for finding new biological control agents for other cactus weeds. Since C. pallida is a high priority weed in South Africa, the benefits of managing this weed will have numerous positive impacts, such as restoring native biodiversity and helping return profits to landowners and communities.