Plant effects on biological control of spider mites in the ornamental crop gerbera

IntroductionThe spider mite Tetranychus urticae Koch is an important pest in many greenhouse crops. In vegetables it can be successfully controlled with the predatory mite Phytoseiulus persimilis Athias-Henriot, a specialist predator of spider mites (Helle & Sabelis, 1985). However, on ornamental crops where aesthetic damage is not tolerated, biological control is more difficult. Pest control on ornamentals mainly relies on the use of pesticides (van de Vrie, 1985).In the ornamental crop gerbera, Gerbera jamesonii Bolus, biological control stands a better chance, since gerbera cutflowers are sold without leaves. The leaves stay behind on the plants in the greenhouse, when flowers are harvested. Therefore, zero tolerance for aesthetic damage does apply for gerbera cutflowers but not for the leaves (van de Vrie, 1985).In gerbera, spider mites usually only damage leaves and do not feed on flowers. The spider mites only migrate to the flowers when the leaves become overexploitated. Gerbera growers will interfere with control methods long before overexploitation takes place. If they apply biological control, there will be some damage by spider mites on the plant leaves. But spider mite damage on the leaves can be accepted, as long as it does not lead to an intolerable lower flower production by the plants. Therefore, biological control may be a suitable method to control spider mites in gerbera.Many cultivars of gerbera are commercially available. At present these cultivars are mainly selected for the colour and shape of the flowers. However, the cultivars also differ in several plant characteristics that have not been selected for, such as leaf shape and density of trichomes on the undersurface of the leaves (Sütterlin and van Lenteren, 1997; Krips et al. , 1999a). Earlier studies showed that the success of natural enemies in biological control can be influenced strongly by characteristics of the host plant (Bergman and Tingey, 1979, Boethel and Eikenbary, 1986;van Lenteren and de Ponti, 1990; Hare, 1992; Walter, 1996). If gerbera cultivars differ in such characteristics, biological control may not be equally successful on all cultivars.The objective of this PhD study was to investigate which differences between gerbera cultivars result in differences in the success of biological control of T. urticae with the predatory mite P. persimilis. We have focussed on the following three host plant characteristics:Host plant resistance to T. urticaeDensity of leaf hairs on the under surface of leavesProduction of volatiles by leaves with spider mite damageWe determined to which extent gerbera cultivars differ in each of these characteristics. In addition, we assessed the effect of these characteristics on certain aspects of the interaction between spider mites and their predators on gerbera.1. Host plant resistance to T. urticaeThe intrinsic rate of population increase, r m , of T. urticae was determined on nine gerbera cultivars. The r m differed largely per cultivar and varied from 0.09-0.23/day. Hence, gerbera cultivars differ largely in resistance to T. urticae. However even on the most resistant cultivar additional methods of spider mite control will eventually be necessary. An obvious additional method is biological control with the predatory mite P. persimilis. But this is only possible if plant resistance to spider mites and biological control are compatible in gerbera.To investigate whether resistance and biological control of spider mites are compatible in gerbera, we determined the r m of the predatory mite P. persimilis on four gerbera cultivars on which the r m of T. urticae differed largely. Cultivar differences in r m of the predators were negligible, the r m varied from 0.43/day to 0.45/day. This indicates that a combined use of resistance and biological control of spider mites might be possible in gerbera.In contrast to what is commonly assumed, the speed at which spider mites are exterminated by P. persimilis is not higher on resistant cultivars than on susceptible cultivars. At first this seems to be an argument against breeding for resistance to spider mites in any crop. However, the advantage of resistant cultivars lies more in the time before predator introduction is necessary, so, to delay the moment that addidtional control methods are necessary, than in the success of biological control after predators are introduced.2. Density of leaf hairs on the under surface of leavesGerbera cultivars differ largely in density of trichomes on the undersurface of leaves (Sütterlin & van Lenteren, 1997). We investigated whether prey searching behaviour of P. persimilis is affected on cultivars with a high density of trichomes. We determined the walking speed, walking activity and rate of encounter with prey on three cultivars with trichome densities of 105, 400 and 730 trichomes/cm 2. Furthermore, we assessed the predation rate on these three cultivars at three prey densities.Walking speed of the predators was highest on the cultivar with the lowest leaf hair density. Walking activity, defined as the percentage of time the predators spent walking, was not dependent on leaf hair density of the cultivars. The rate of encounter with prey was inversely related to trichome density. Predation rate of adult female P. persimilis was affected negatively by trichome density when prey density was low. However, at high prey density such an effect was absent, because the predation rate is not limited by the rate of encounter with prey at high prey density.The results of these experiments showed that the effect of trichome density on the predator-prey dynamics between T. urticae and P. persimilis on gerbera plants will be dependent on the density in which T. urticae occurs on these plants. In a small-scale greenhouse experiment we found indications that the density of T. urticae is limiting the population growth rate of the predators and that trichome density affects the growth of the predator populations on gerbera plants.3. Production of volatiles by leaves with spider mite damageMany host plant species produce volatiles when herbivores damage the leaves. Natural enemies of the herbivores can use these volatiles to locate plants that are infested with their prey or hosts (see Vet & Dicke, 1992 and Dicke, 1994 for reviews). Such volatiles also are important in the interaction between T. urticae and P. persimilis (Sabelis and van de Baan, 1983; Sabelis et al . , 1984; Dicke and Sabelis, 1988; Dicke et al . , 1990a,b; Sabelis and van der Weel, 1994; Sabelis and Afman, 1994). Plants of several species produce volatiles when spider mites damage their leaves. Phytoseiulus persimilis is attracted to these volatiles (Dicke and Sabelis, 1988). Once the predators have found spider mite-infested leaves, they stay there until all prey is exterminated, which is most likely the result of attraction of the predators to volatiles from the spider mite patch and arrestment by these volatiles (Sabelis & van de Meer, 1986; Sabelis and Afman, 1994).We have investigated whether gerbera leaves with spider mite damage also produce volatiles that attract P. persimilis. In contrast to undamaged gerbera leaves that had a low production of volatiles, spider mite-damaged gerbera leaves produced many volatiles in large quantities. However, P. persimilis from our standard culture on Lima bean leaves with T. urticae was not attracted to these volatiles. In contrast, the predators responded strongly to the gerbera volatiles when they were given six days of experience with spider mites on gerbera leaves and therefore were exposed to the spider mite-induced volatiles for six days.We have investigated whether the chemical composition of the spider mite-induced volatiles differs between gerbera cultivars. Furthermore, we determined whether cultivar differences exist in attractiveness of the spider mite-induced volatiles to P. persimilis. In a two-choice situation the predators preferred the volatiles of cultivars Rondena and Bianca over those of cultivar Sirtaki. The volatiles from the cultivars Sirtaki and Fame did not differ in attrativeness.The blend emitted by spider mite-damaged Sirtaki leaves consisted of a much lower relative amount of terpenes than the blends of the other three cultivars, which was mostly due to a low production of cis-alpha-bergamotene, trans-alpha-bergamotene, trans-beta-bergamotene and (E) -beta-farnesene. Furthermore, the production of (E) -beta-ocimene and linalool was lower by Sirtaki and Fame leaves than by leaves of Bianca and Rondena. These two compounds attract P. persimilis when offered in a Y-tube olfactometer (Dicke et al. , 1990b) . Whether the poor attractiveness of the spider mite-induced volatiles of Sirtaki and Fame is related to the limited production of these compounds should be investigated in future studies.