Submitted to: Meeting Abstract
Publication Type: Proceedings
Publication Acceptance Date: 7/14/2009
Publication Date: 10/1/2010
Citation: Wheeler, G.S. 2010. Selection of test plant list for weed biological control with molecular and biochemical data. Meeting Abstract. Vol. 10, Article 166.
Interpretive Summary: The compiling of a list of plants to test is one of the first steps in a biological control project. These plant species are offered to prospective biological control agents under quarantine conditions to determine if the candidate agent will be sufficiently safe to field release. The plant species thought to be most vulnerable to non-target damage are those that are close relatives and those that have similar chemistry. Insects and plants have sequentially evolved creating a close association of plant feeding insects and defensive plant chemistry. This has resulted in insect groups feeding on plants with similar ancestry. This presentation examines the historical evidence that supports this doctrine and reviews recent research that questions the general acceptance of these assumptions.
Technical Abstract: The initial steps of weed biological control programs involve the determination of the host range of a prospective agent prior to consideration for release. Accurately predicting the host range of a potential agent is fundamental to this process. This may be conducted first in the country of origin in open field testing (Briese et al. 2002) and later under controlled environmental conditions in quarantine (Zwolfer and Harris 1971; McFadyen 1998). Initially a plant test list is established composed of species that are taxonomically related to the weed and species of economic and ecologic importance from the area where the weed is a problem (Wapshere 1974). This centrifugal / phylogenetic testing procedure involves “testing plants of increasingly distant relationship to the host until the host is circumscribed” (Wapshere 1974) and is based upon the assumption that host shifts occur to plants of similar taxa (Ehrlich and Raven 1964; Mitter and Farrell 1991). Typically rare species are also included in the plants tested. As useful as this process is it potentially overlooks unrelated plant taxa that share similar secondary plant metabolites. Recent evidence indicates that chemical similarity may be a better predictor of host use than are phylogenetic relationships (Becerra 1997; Wahlberg 2001). Although little evidence may exist from weed biological control projects (Schaffner 2001), species with secondary metabolites similar to the target weed should be included in the test list as they may contain the behavioral cues used by these specialized herbivore species to locate hosts and initiate feeding (Wheeler 2004). As useful as the centrifugal / phylogenetic testing procedure may be it potentially overlooks distantly unrelated plant taxa that share similar secondary plant metabolites.