Spider mite's use of a refuge during the quiescent stage in the presence of a predator

Abstract

In the presence of predators, some organisms alter their behavior & morphology to reduce the risk of predation (e. g., Dodson, 1989; Tollrian 1995; Wiackowski & Staronska, 1999; Buskirk & McCollum, 2000). Some preys utilize refuges that they do not inhabit in the absence of predators (e.g., Crowder et al., 1997). Such changes are examples of predator-induced defenses. If an inducible defense had no associated costs, it would be expressed regularly rather than only in specific situations (Riessen, 1992). Inducible defenses are therefore assumed to incur fitness costs that are outweighed by the benefit of predator avoidance. The development of the Kanzawa spider mite, Tetranychus kanzawai Kishida (Acari: Tetranychidae), consists of active stages (larva, protonymph, and deutonymph) and quiescent stages (egg, protochrysalis, deutochrysalis, and teleiochrysalis). Individuals are more likely to be attacked by predators during the quiescent stages than during the active stages (see T. urticae[Shimoda et al., 1997], for example). Tetranychus kanzawai constructs highly complicated webs over the surface of a leaf (Saito, 1983). Empirical observations show that T. kanzawai enters its quiescent stages either on the web, or on the leaf surface (hereafter referred to as ‘on the leaf’) (Figure 1). Although the space between the web and the leaf can serve as a refuge from some predators, other predators can access this space (McMurtry & Johnson, 1966). Some predatory mites are attracted to spider mite webbing (Pratt & Croft, 1999), and prefer to reside and oviposit on leaf surfaces with the webbing (Roda et al., 2001); Amblyseius womersleyi Schicha (Acari: Phytoseiidae), a predator of T. kanzawai that is native to Japan, is an example of such a predator (K. Oku, pers. obs.). We hypothesized that T. kanzawai enters into quiescent stages on its web as a predator-induced defense. We discuss the benefit and possible costs associated with such a defense. Positions on a leaf of T. kanzawai females in the teleiochrysalis stage. The study population of T. kanzawai was collected from a strawberry garden in Kyoto, Japan, and maintained on bean (Phaseolus vulgaris) leaf discs pressed on to water-saturated cotton in Petri dishes (90 mm diameter, 14 mm depth). The A. womersleyi population was also collected in Kyoto, and was reared on bean leaf discs that were heavily infested with T. urticae as prey (30–50 female adults and individuals of other stages per leaf). All of the dishes were placed together in a transparent plastic container and kept at 25 °C, with 50% r.h. and a L16:D8 photoperiod. To observe the quiescent stage of T. kanzawai in the presence of A. womersleyi, we transferred 40 mated females, 2 days old, to each of two bean leaf discs; they were allowed to oviposit for 1 day and were then removed. After 7 days, when T. kanzawai had developed into the deutochrysalis stage, we introduced an adult female of A. womersleyi on to one leaf disc (predator present); the other leaf disc served as a control (predator absent). After 24 h, we removed the predator, at which time all T. kanzawai individuals were in an active stage (deutonymph). From that day onward, we recorded whether T. kanzawai females entered the teleiochrysalis stage on the leaf or on the webs that they produced. To compare the survival rate of teleiochrysalis females on webs with that on leaves, we had to use A. womersleyi females to induce T. kanzawai to form teleiochrysalises on webs; in the absence of A. womersleyi, T. kanzawai seldom enter the teleiochrysalis stage on webs (see Results). We transferred 20 2-day-old mated females to each of six leaf discs, allowed them to oviposit for 1 day, and removed them. After 4 days, when all the eggs had hatched, we transferred an A. womersleyi adult female to each leaf disc for 3 days. We removed excess females and males from the discs, so the number of teleiochrysalis females varied among the discs. We then transferred three to nine A. womersleyi female adults to each disc to approximate an A. womersleyi/T. kanzawai ratio equivalent to 27–30%. As T. kanzawai in the quiescent stages delay molting at high humidity (Ikegami et al., 2000), to keep the females in the quiescent stage for observation, we covered the leaf discs with a transparent plastic cup to raise the relative humidity to nearly 100%. In spite of the high humidity, A. womersleyi females consumed T. kanzawai females. After 5 h, we removed the plastic cups and the A. womersleyi females. The number of emerged adults was then recorded. In addition, in order to examine the mortality of teleiochrysalis females due to causes other than predation, we compared the survival of females on webs and leaves in the absence of predators. We prepared another 12 leaf discs with teleiochrysalis females on webs and leaves, using the method described above, and kept them under high humidity conditions without predators for 5 h. The number of emerged adults was then recorded. Male adults of tetranychid mites often guard teleiochrysalis females, because only the first mating is effective for females (see examples in T. urticae[Potter & Wrensch, 1978; Satoh et al., 2001]). To determine whether T. kanzawai adult males can find teleiochysalis females on webs and on leaves, we prepared seven leaf discs with equal numbers of teleiochrysalis females on webs and leaves, using the methods described above. We then transferred T. kanzawai adult males to the discs, at a ratio of one male to two females. After 2 h, we counted the number of adult males that were guarding females that were on webs or on leaves. In the presence of A. womersleyi, a significantly greater proportion of T. kanzawai females entered the teleiochrysalis stage on webs than on leaves (P = 0.0002, t-test on arcsine-transformed data; Figure 2). Furthermore, significantly more teleiochrysalis T. kanzawai females survived on webs as compared with those on leaves (P = 0.027, paired-comparison t-test on arcsine-transformed data; Figure 3a). In the absence of predators, however, nearly 100% of the females on both webs and leaves survived, and the difference was not significant (P = 0.339, paired-comparison t-test on arcsine-transformed data; Figure 3b). Therefore, females that died in the presence of predators (Figure 3a) were considered to have been consumed by them. Amblyseius womersleyi foraged mainly in the space between the webs and the surface of the leaves. Therefore, the surface of the webs may be free of predators (Jeffries & Lawton, 1984) for T. kanzawai during quiescent stages. Quiescence on webs in response to predators is an inducible defense. Teleiochrysalis females on webs in the presence or absence of a predator. (P = 0.0002, t-test on arcsine-transformed data.) Costs and benefits of predator avoidance. (a) Survival of teleiochrysalis females in the presence of predators. (P = 0.027, paired-comparison t-test on arcsine-transformed data.) (b) Survival of teleiochrysalis females in the absence of predators. (P = 0.339, paired-comparison t-test on arcsine-transformed data.) (c) Attraction of male adults. (P = 0.025, paired-comparison t-test on arcsine-transformed data.) If inducible defenses had no associated costs, they would be expressed regularly rather than only in specific situations (Riessen, 1992). Inducible defenses are therefore assumed to incur fitness costs. If the predatory mite cannot easily access quiescent spider mites on webs, then T. kanzawai adult males probably cannot either. In fact, T. kanzawai adult males guarded significantly fewer teleiochrysalis females on webs than on leaves (P = 0.025, paired-comparison t-test on arcsine-transformed data; Figure 3c). A teleiochrysalis female not guarded by a male would lose both an opportunity to mate and the genetic benefits of male-male contests, especially within a small colony (e.g., Promislow et al., 1998). The decreased accessibility of females to males during the teleiochrysalis period may therefore be a cost of predator avoidance. Another possible cost is exposure to other predators that attack from the edges of the webs; some predators that do not intrude into the space between webs and leaves are likely candidates (e.g., Roda et al., 2000; Venzon et al., 2000). The plasticity in predator avoidance during quiescent periods may be maintained in part by trade-offs in defense against different predators. This possibility should be examined in future studies. We thank Dr K. Mori of Ibaraki University, Dr M. Osakabe, Dr K. Fujisaki, and Mr A. Kasai of Kyoto University for their comments and encouragement. We also thank Dr Norton of Colorado State University and an anonymous reviewer for their valuable suggestions.