Arrestment responses to stimuli from Schizotetranychus nanjingensis on bamboo leaves by the predatory mites Typhlodromus bambusae and Amblyseius cucumeris were examined in the laboratory using two-choice tests. Both predator species responded to spider mite females and web-nests, but not to their feces. Furthermore, T. bambusae responded more strongly to intact web-nests than to empty web-nests, whereas A. cucumeris responded more strongly to empty web-nests.
The spider mite Schizotetranychus nanjingensis Ma & Yuan is one of the serious pests of the moso bamboo (Phyllostaychs pubescens), which is a major forestry crop in southern China (Cui 1989; Yu & Shi 1991; Zhang et al. 1998a, b). This species has a specialized life type that is similar to S. longus Saito, a closely related species inhabiting dwarf bamboo plants in Japan (Saito & Ueno 1979). Females of S. nanjingensis construct web-nests with a very dense webbing roof on the underside of bamboo leaves. The web-nest may function as a shelter in which all spider mites live and feed on leaf tissue (Zhang et al. 1999a). The intact webbing of S. nanjingensis is also effective in preventing the invasion by non-specialist predators such as A. longispinosus (Zhang et al. 1999c) and A. cucumeris (Zhang et al. 2000b), although it could not prevent the attacks of the specialized predator of S. nanjingensis, T. bambusae Ehara, which is adapted to the web-nests of S nanjingensis (Y.-X. Zhang, unpublished data).
Although T. bambusae is the dominant predator of S. nanjingensis in bamboo forests in Fujian and is capable of responding to densities of S. nanjingensis in laboratory functional response experiments, S. nanjingensis has reached pest status in most bamboo forests in Fujian, presumably because of the low density of T. bambusae in disturbed monoculture forests (Zhang et al. 1999b; Zhang 2000c; Zhang et al. 2001). One approach for the biological control of this pest mite is to use the augmentative releases of predatory mites (e.g. A. longispinosus and A. cucumeris) that are relatively easy to produce on a large scale (Zhang et al. 2001).
A series of studies on the potential of these predators as biocontrol agents of S. nanjingensis has been conducted. Studies on the predation of A. longispinosus, A. cucumeris and T. bambusae on S. nanjingensis were published elsewhere (Zhang et al. 1999b,c; Zhang et al. 2000b). Zhang et al. (2000b) observed that A. cucumeris can invade web-nests of S. nanjingensis that have openings and prefers to lay its eggs inside these nests. Zhang et al. (2000a) examined the arrestment response of A. longispinosus to stimuli from S. nanjingensis on bamboo leaves and showed that this predator can respond to stimuli from web-nests of S. nanjingensis. In this paper, we compare the responses of T. bambusae and another predator (Amblyseius cucumeris Oudemans) to stumuli from S. nanjingensis on bamboo leaves to evaluate the searching ability of these mite predator species.
Materials and methods
Bamboo leaves and mites
Bamboo leaves of different ages without S. nanjingensis web-nests and T. bambusae were collected from bamboo forests in Nanping City, Fujian, China. Leaves were brought back to the laboratory in plastic bags and used in the experiments.
Schizotetranychus nanjingensis and T. bambusae used in this experiment originated from the same forests where bamboo leaves were collected and were reared in the laboratory for experiments.
The predatory mite A. cucumeris was introduced from Biological Crop Protection Limited in England and reared in the laboratory.
The experimental set-up, procedure and data analysis were described in detail in Zhang et al. (2000a). All experiments were carried out at 25°C. A piece of water-soaked foam plastic (10cm in diameter and 1cm thick) was placed in a petri dish (12cm in diameter) and a sheet of filter paper was placed onto the foam plastic. Pairs of detached leaves of moso bamboo were placed on the substrate and subjected to the following four different treatments. In treatment A, one leaf contained a webnest with all the stages of S. nanjingensis and the other leaf was a control without any mites, nor mite damage. In treatment B, the treated leaf contained an empty web-nest of S. nanjingensis, whereas the other leaf was an untreated control. In treatment C, feces of S. nanjingensis were placed on one leaf and the other leaf was an untreated control. In treatment D, 5 females of S. nanjingensis were introduced on one leaf, whereas the other leaf was an untreated control.
During each behavioural test, a female of T. bambusae or A. cucumeris was released onto the bottom of the T-shaped bridge using a fine brush and allowed to select between the two leaves freely. The location of each predator female was then checked at 1-hour intervals for 5 hours. The behaviour of the predator was also observed from time to time. There were at least 30 replicates for each treatment.
Responses of T. bambusae
In treatment A, one hour after the start of the experiment, 53.3% of predators were found on the leaves with the web-nests of S. nanjingensis, whereas 43.3% were on the other leaves (Fig 1). The proportion of predators on leaves with web-nests increased to 60% at the end of 2–3 hours and to >80% at the end of 4–5 hours. The difference in the proportion of predators between the two leaves at the end of 4–5 hours was highly significant (t-test, P <0.001).
In treatment B, the response of predators to treatment became significant after three hours, when 63.0% of the predators were found on the leaves with an empty web-nest, whereas 26.6% were on the clean, control leaves (Fig. 2; P<0.001). The observation showed that T. bambusae females selectively stayed on leaves with web-nests and they often stayed inside the empty nests. On the contrary, they frequently moved about on clean leaves and often escaped (Fig. 2).
In treatment C, the proportion of predator females on leaves with S. nanjingensis feces was not significantly different from that on clean leaves and some of the mites frequently moved about and often escaped (Fig. 3).
In treatment D, the proportion of predator females on leaves with spider mite adults was over 60% at the end of the first and second hour but was half as much on control leaves (Fig. 4; t-test, p<0.001). The difference between the two became less significant at the end of the 3rd hour and non-significant at the end of 4–5 hours, because the spider mites on treated leaves had been consumed by . bambusae.
Response of A. cucumeris
In treatment A, one hour after predator introduction, 70% of the predators were found on the leaves with a web-nest of S. nanjingensis, whereas only 20% were on the control leaves; the rest of the predators (10%) were on the bridge and no mites escaped from the system (Fig. 5). During the next 4 hours, 16–26% of the mites escaped, but the preference of predator females for leaves with web-nests over those without web-nests was still significant (P <0.001), although its strength was not as great as during the first hour (Fig. 5).
In treatment B, the preference of predator females for leaves with empty web-nests was consistently significant for the five hours (t-test, P<0.001) (Fig. 6).
In treatment C, predators showed no apparent preference for leaves with feces of S. nanjingensis (Fig. 7). They frequently moved off the leaves and often escaped from the system.
In treatment D, the proportion of predators on the leaves with 5 females of S. nanjingensis increased from 50% at the end of the first hour to 70% at the end of the fifth hour, whereas the proportion of predators on the control leaves fluctuated around a quarter. As a result, the preference of predator females for leaves with live mites of S. nanjingensis over those of the control ones was very significant (t-test, P <0.001) (Fig. 8).
Arrestment responses to tactile stimuli from webbing and other products of spider mites have been shown in several species of Phytoseiidae (e.g. Schmidt 1976; Hislop & Prokopy,1981; Hoy & Smilanick 1981; Rasmy et al. 1991). In a previous paper, we showed that females of A. longispinosus preferred bamboo leaves with empty web-nests of S. nanjingensis (Zhang et al. 2000a). This study shows that T. bambusae and A. cucumeris also have similar preferences (Figs. 2 and 6). Furthermore, we showed in this study that feces alone did not arrest the movement of T. bambusae and A. cucumeris (Figs. 3 and 7), and that spider mite females themselves also arrested these two predators (Figs. 4 and 8), as did intact web-nests containing spider mites (Figs. 1 and 5). This indicates that these two predators responded to stimuli associated with webbing and the body of S. nanjingenses, but not their feces. The chemical nature of these stimuli should be interesting topics for future studies.
Although T. bambusae and A. cucumeris both responded to spider mites and their web-nests, they differed in the way they responded. T. bambusae, a specialist predator of S. nanjingensis, responded more strongly to intact web-nests than to empty nests and their response to intact nests became stronger with time, but their response to spider mites themselves decreased with time. In contrast A. cucumeris, a more generalist feeder, responded more strongly to empty web-nests and their response to intact nests become less strong with time, but their responses to spider mites increased with time.
In an integrated control programme, we released A. cucumeris into moso bamboo forests in early May–June for the control of S. bambusae, Aponychus corpuzae and Aculus bambusae, and also used it to disrupt the migration of S. nanjingensis from the ground to new leaves, especially in one year shoots. The results from this study are encouraging in that A. cucumeris, although it has no natural association with S. nanjingensis, can respond to stimuli from female spider mites and also their webnests. Coupled with the results of a previous study on the predation of A. cucumeris on S. nanjingensis, A. cucumeris shows promising potential as a biocontrol agent for S. nanjingensis.
This project was funded by the Science and Technology Commission of Fujian Province, China and the State Administration of Foreign Expert Bureau, Beijing China. Dr Z.-Q. Zhang's visits to China in 2000 and 2001 were funded by Fujian Yingzhiban, a provincial office for promoting exchange and collaboration between scientists in China and elsewhere. This paper was finalized when the senior author was funded by the Foundation for Research, Science and Technology, New Zealand. We thank Dr Anne Baker (The Natural History Museum, London, UK) a critical review of the manuscript.