Biology and ecology of stem-boring insects associated with the invasive weed Senecio madagascariensis (Asteraceae) and related species in their native range in KwaZulu-Natal, South Africa.

Abstract

Senecio madagascariensis Poiret (Asteraceae), commonly known as fireweed, is an herbaceous plant that is native to South Africa and Madagascar and has become a major agricultural weed in many countries around the world, most notably Australia where it poisons livestock, hybridizes with native plants and reduces pasture productivity. Consequently, it has been targeted for biological control. This study formed part of the collaboration between the University of KwaZulu-Natal and Commonwealth Scientific and Industrial Research Organisation in Australia to source suitable control agents for fireweed. The aims of the study were to: (1) determine the seasonal abundance of stem-boring insects associated with fireweed; (2) use DNA barcoding to determine the identity of important insect species; (3) determine the field host range of these insects by surveying related native Senecio species and comparing the insect taxa found; (4) determine the impact of larvae of the stem-boring weevil, Gasteroclisus tricostalis (Thunberg) (Curculionidae), on fireweed plants and; (5) determine the laboratory host range of G. tricostalis to determine its suitability for release in Australia. Stem-boring taxa that were considered to have biocontrol potential included Coleoptera (specifically Curculionidae) and Lepidoptera (specifically Tortricidae and Pterophoridae). Season played a significant role in determining the abundance of insect taxa, with Curculionidae larvae displaying two peaks in abundance (May and January), while Lepidoptera larvae displayed a single peak in April. DNA barcoding of the COI gene region revealed 19 weevil species associated with native Senecio species, with G. tricostalis being restricted to the S. madagascariensis species complex. DNA barcoding also revealed six Lepidoptera species, with two species restricted to the S. madagascariensis species complex. Following these results, the weevil G. tricostalis was prioritized as the most promising candidate agent due to its narrow field host range. During impact trials involving varying larval loads, arising from differential oviposition densities, on the growth and reproductive traits of S. madagascariensis, larvae of G. tricostalis were able to significantly reduce the floral productivity and shoot production of fireweed plants, both of which influence the abundance and spread of fireweed. Following these results, further host-range testing was conducted on G. tricostalis in the laboratory. Although the weevil displayed a narrow field host range, these tests revealed that it was capable of surviving on some non-target Australian Senecio species. Gasteroclisus tricostalis larvae were recovered on seven native Australian Senecio species, whilst adults were reared on four species, in numbers that were not significantly different to those recorded on fireweed. This host range was considered to be unacceptably broad and the weevil was rejected as a potential agent for Australia. Although S. madagascariensis is a challenging target for biological control, other invaded countries that have fewer native species within the genus Senecio (such as New Zealand and Hawaii, USA), could consider G. tricostalis, and other previously discounted candidate agents, for release, due to lower chances of non-target impacts.