Entomologia Experimentalis et Applicata 91: 155–161, 1999. © 1999 Kluwer Academic Publishers. Printed in the Netherlands. 155 Flowers in tri-trophic systems: mechanisms allowing selective exploitation by insect natural enemies for conservation biological control L. R. Baggen 1 , G. M. Gurr 1,∗ & A. Meats 2 1 Orange Agricultural College, The University of Sydney, Orange, New South Wales, 2800, Australia; 2 School of Biological Sciences, The University of Sydney, New South Wales, 2006, Australia; ∗ Author for correspondence Accepted: November 30, 1998 Key words: nectar, habitat manipulation, selectivity, Copidosoma koehleri, Phthorimaea operculella Abstract Many insects have coevolved with certain angiosperm taxa to act as pollinators. However, the nectar and pollen from such flowers is also widely fed upon by other insects, including entomophagous species. Conservation bi- ological control seeks to maximise the impact of these natural enemies on crop pests by enhancing availability of nectar and pollen-rich plants in agroecosystems. A risk with this approach is that pests may also benefit from the food resource. We show that the flowers of some plants (viz., buckwheat, Fagopyron esculentum Moench and dill, Anethum graveolens L.), and the extrafloral nectaries of faba bean (Vicia faba L.) benefit both Copidosoma koehleri Blanchard (Hymenoptera: Encyrtidae) and its host, the potato pest, Phthorimaea operculella Zeller (Lep- idoptera: Gelechiidae). In contrast, phacelia (Phacelia tanacetifolia Benth) and nasturtium (Tropaeoleum majus L.) benefited only the parasitoid. When adult moths of P. operculella were caged with flowers of phacelia or nasturtium, longevity of males and females, egg laying life, fecundity, average oviposition rate, and number of eggs in ovaries at death were no greater than in the control treatment with access to shoots without flowers plus water. All the foregoing measures were increased compared to the control when the moths were allowed access to dill, buckwheat or faba bean extrafloral nectaries. Such ‘selectivity’ has the potential to make the use of floral resources in conservation biological control more strategic. We present morphometric and observational evidence to illustrate how such mechanisms may operate. Introduction Conservation biological control is the least studied of all biological control approaches (Dent, 1995). Other than withholding the pesticides which may adversely affect natural enemy populations, conservation biolog- ical control also makes use of habitat manipulations to favour predators and parasites. These include pro- vision of shelter, alternative hosts or prey, or food plants from which nectar and pollen may be obtained. The latter technique has been relatively widely used (e.g., Hickman & Wratten, 1996; Matthews-Gehringer et al., 1994; Topham & Beardsley, 1975). A num- ber of workers have sought to maximise the benefit from providing food plants by screening to quantify the benefit to natural enemies of access to different flower types (Idris & Grafius, 1995; Patt et al., 1997; Orr & Plesants, 1996). Generally, plants with exposed nectaries such as are common in the family Apiaceae (formerly Umbelliferae) are recommended. However, applying this method to improve biological control of potato moth (Phthorimaea operculella) by its par- asitoid Copidosoma koehleri proved only partially successful because pest adults apparently fed from the nectaries of the coriander (Coriandrum sativum L.) and/or faba bean (Vicia faba L.) planted alongside the potatoes (Baggen & Gurr, 1998). This observation led to laboratory screening of flower types similar to that carried out by other workers but also included tests to determine which flowers were also fed upon by the pest. Borage (Borago officinalis L.) was found to ben- efit only the parasitoid and the term ‘selective food plant’ proposed (Baggen & Gurr, 1998).