Skip to main content
ARS Home » Pacific West Area » Reno, Nevada » Great Basin Rangelands Research » Research » Publications at this Location » Publication #273995

Title: Direct and indirect effects of two biocontrol insects and pollinators on an invasive thistle

item Swope, Sarah
item PARKER, INGRID - University Of California

Submitted to: Ecological Applications
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 1/23/2012
Publication Date: 6/14/2012
Citation: Swope, S.M., Parker, I.M. 2012. Direct and indirect effects of two biocontrol insects and pollinators on an invasive thistle. Ecological Applications. 22:2122-2134.

Interpretive Summary: Biological control agents are often our best hope of controlling the most widespread and problematic invasive plants. Most weed species in western North America are subjected to attack by more than one biocontrol agent but we rarely know how these agents interact with one another, directly and indirectly. We used structural equation modeling, a novel approach to dissecting these direct and indirect interactions among agents and pollinators, to identify the interaction pathways and quantify their relative strength. We found that bud herbivory by one agent had a larger impact on final seed output in the highly invasive annual weed yellow starthistle than did seed predation – the direct consumption of seeds. We also found that the agent with the smaller impact on the plant deterred the agent with the larger impact – a risk of the multi-agent approach to biocontrol. Finally, we found that one of the agents had a larger than expected impact on the weed because it also indirectly deterred pollinators from visiting the plant which is dependent on pollinators to set seed. This study demonstrates the value of using structural equation modeling to dissect and quantify interactions among several species and their net impact on the target weed. This approach also identifies pathways that could be exploited by biocontrol practitioners, e.g., the indirect reduction in pollinator visitation, as they search for new agents to control our most problematic weeds.

Technical Abstract: Herbivores, seed predators and pollinators can simultaneously exert strong impacts on their host plants, with opposing impacts on plant fitness. They can also affect each other’s impact on the plant via their indirect interactions. In some cases, indirect interactions can lead to synergistic outcomes in which the impact on the plant is super-additive and in other cases, they can interfere with one another such that the impact on the plant reduced. These questions are especially relevant to the use of the use of biological control agents to manage invasive plants. Most invasive plants are subjected to control by more than one and often several agents but how these agents interact with one another is rarely studied. Here we used Structural Equation Modeling (SEM) to quantify the impact two biocontrol agents and a generalist pollinator on their shared host, the invasive annual weed Centaurea solstitialis. SEM also allows us to identify and quantify the direct and indirect interactions among agents and pollinators. The weevil Eustenopus villosus acts as both a bud herbivore (as an adult) and a seed predator (as a larva) and the fly Chaetorellia succinea also acts as a predispersal seed predator. We conducted this work at three study sites arranged across the entire latitudinal range of C. solstitialis in California from the coast to the high Sierra. Bud herbivory had the largest impact on the weed’s final seed output; in fact, its impact was between three and ten times larger in magnitude than the impact of both seed predators combined. SEM also revealed that oviposition by E. villosus deterred C. succinea, which is unfortunate because seed predation by C. succinea had a larger impact on the plant than did seed predation by E. villosus. C. succinea’s direct impact (seed predation) was enhanced by the fact that it also deterred pollinators, upon which the plant depends for reproduction, and the plants in this study experienced mild pollen-limited seed set. Ultimately, the agents’ direct and indirect interactions had the largest net impact on the Coast plants, an intermediate impact on the Interior plants and the smallest impact on the Sierra plants. This study demonstrates the importance of considering multiple direct and indirect interactions among agents as well as identifying pathways that could be exploited by biocontrol practitioners seeking new agents to control invasive plants.