Who Is Eating Our Agents?
When new agents establish themselves on host plants, they enter existing food webs and their effectiveness can be influenced by the presence of other species, either native or exotic.
Interactions can occur between different insect species or can involve larger predators, and because they are not always obvious, they have not been well studied. Recent work by Quentin Paynter found that the likelihood of parasitism can be predicted when biocontrol agents have analogues in the native biota. He is now interested in looking at whether the chance of predators (such as spiders, predatory bugs, beetle larvae, soldier flies, lacewing larvae, and birds) hindering the establishment and effectiveness of biocontrol agents is also predictable.
“Some agents that have been introduced to New Zealand are well concealed – stem and seed borers for example, while others leave themselves more open to attack when browsing on leaves,” said Quentin. “The degree of predation on agents appears to depend on how well concealed they are on a plant, but having said that, some unconcealed agents, such as the St John’s wort beetles (Chrysolina spp.), are still hugely successful, so we have a lot of questions to answer yet.”
One reason why we know so little about predation of agents is that we have relied primarily on casual field observations for information because the more detailed research required to investigate properly can be very time consuming. Quentin adds that studying predation is even more challenging than studying parasitism. Until recently you either had to catch predators in the act of feeding on their prey or use predator exclusion techniques to indirectly determine what is feeding on a biocontrol agent. A good example of the latter was the exclusion experiments we conducted on the boneseed leaf roller (Tortrix s.l. sp. “chrysanthemoides”). They showed that the leafroller does poorly when scale insects are present on boneseed (Chrysanthemoides monilifera monilifera) as predatory ants and wasps are attracted to their honeydew.
However, experiments on this scale would be prohibitively expensive to perform on all weed biocontrol agents.
“Now that molecular biology techniques are becoming more sophisticated we should soon be able to routinely perform rapid tests for predation of biocontrol agents instead,” revealed Quent. We will be able to collect all potential predators found living close to weed biocontrol agents and test their gut contents for biocontrol agent DNA. It won’t be totally straightforward as there is a high chance of specimen contamination between predators and prey when they are processed. Also, it will be important to collect predators individually to reduce the chance of getting ‘false positive’ results from predators eating agents in the collection tube that they would rarely feed upon in more natural circumstances. Also the DNA only remains viable in the gut of predators for a few hours so the samples will have to be fresh and carefully stored to preserve their integrity. Quent is in the process of putting together a sampling protocol to maximise the utility of the specimens collected.
Preliminary laboratory-based DNA work conducted by co-workers Simon Connell and Zhi-Qiang Zhang has confirmed that native mites, which have adapted to living in broom galls, feed on the broom gall mite (Aceria genistae). Current research is investigating the interactions between these two mite species and other fungus-feeding mites that make up a surprisingly complex community in these galls. Further investigations to test the viability of the DNA approach in the field are likely to begin with predation of the gorse spider mite (Tetranychus lintearius) by an endemic beetle (Stethorus bifidus). Predation of the spider mites has been reasonably well documented, with beetles feeding more rapidly when the density of mites is higher. “If we find spider mite DNA in field-collected Stethorus beetles, that should confirm that our technique is working,” said Quentin. “We can then study other important predator–prey interactions that we think may be occurring. For example, we suspect the accidentally introduced mirid bug (Sejanus albisignatus) may feed on the broom psyllid (Arytainilla spartiophila) and broom leaf beetle (Gonioctena olivacea).”
If all goes according to plan, it seems inevitable that this study will generate useful data for helping to select the most effective weed biocontrol agents in the future.
This project is funded by the Ministry of Business, Innovation and Employment as part of the Beating Weeds Programme.