Reducing the mortality of natural enemies of pests caused by insecticides is essential for greater sustainability in rice pest management (Heong & Hardy, 2009; Gurr et al., 2011b). Amongst the most important pest species in terms of driving insecticide use, particularly in the early season, is the rice leaffolder Cnaphalocrocis medinalis (Gueneé) (Lepidoptera: Pyralidae) (Heong et al., 1994; Heong & Escalada, 1997). Losses to this pest and a second leaffolder Marasmia patnalis can be significant if attacks occur in the middle or late stage of rice development (Litsinger et al., 1987a). Work aiming to set an action threshold for insecticidal control of this pest found that azinphos-ethyl and the bacterium Bacillus thuringiensis (Bt) gave poor results, and even the best performing products (i.e. 2-sec-butylphenyl methylcarbamate, endosulfan and monocrotophos) gave only 53% control. Moreover, it was concluded that an Integrated Pest Mangement (IPM) approach was appropriate for exploiting the activity of natural enemies and the ability of rice to compensate for early-season foliar damage by tillering (Litsinger et al., 2006). A single leaffolder larva can consume approximately 25 cm2 of leaf tissue, constituting less than 40% of a normal leaf of indica rice (Heong, 1993). Rice plants have a strong ability to compensate for early-season foliar damage. For example, japonica rice can compensate for as much as 67% of leaffolder-damaged leaves at tillering (Miyashita, 1985). Computer simulations show that leaffolder densities must reach 15 per hill before any detectable yield loss results (Fabellar et al., 1994). The normal larval densities are usually well below three per hill (Guo, 1990; de Kraker, 1996), mainly as a result of of natural biological control. Unfortunately, early-season foliar damage from C. medinalis is very conspicuous and farmers often feel compelled to respond with insecticide application. This is despite the fact that a simulation model by Graf et al. (1992) indicated that normal field populations of leaffolders cause insignificant yield losses. This fact is recognized by the International Rice Research Institute (IRRI, 2003). The major effect of these early-season insecticide applications, usually with broad spectrum active constituents, is to kill most members of the enemy community (Bandong & Litsinger, 1986), including the hymenopteran parasitoids, compromising any available scope for biological control during the rest of the growing season and locking the farmer into the ‘pesticide treadmill’ (van Den Bosch, 1978). Accordingly, insecticides applied into rice fields during the early crop stages to control leaffolders are unlikely to benefit farmers economically (Way & Heong, 1994; Heong & Schoenly, 1998). Instead, they can cause ecological disruptions to the herbivore–predator relationships by shortening the mean food web chain length, favouring an increase in the population of some herbivorous species (e.g. delphacids) and causing secondary brown planthopper pest problems (Chelliah & Velusamy, 1985; Wu et al., 1986; Nadarajan & Skaria, 1988; Panda & Shi, 1989). When these insecticide applications can be avoided, effective biological control is (in principle) achieved via the ‘IPM treadmill’ (Tait, 1987). The scope to achieve this transition in rice is reviewed via a comprehensive integration of the English language literature with that published in foreign language sources. Although detailed information is available on many natural enemies of C. medinalis, much of this exists only in the non-English language (especially Chinese and Vietnamese) publications or in the ‘grey literature’, including institutional reports. These factors make much of the useful information inaccessible to the English-speaking scientific community and inhibit the flow of important information between differing non-English-speaking countries. The need for this analysis is great because persuading farmers not to apply insecticides in response to C. medinalis has proven very difficult, despite some succes in other aspects of rice pest management in recent decades (Ooi, 1996). The particular aim of this review is to assess whether there is scope for enhancing the parasitoid control of C. medinalis to a level that would convince farmers there is no need to spray. This conservation biological control strategy has been researched and deployed in many crop systems (Landis et al., 2000) and has matured into an information-rich approach termed ecological engineering (Gurr et al., 2004).
Parasitoids of C. medinalis and the potential role of nectar sources
Overview of impact. A rich fauna of parasitoids attacks C. medinalis and their impact can be great. Indeed, leaffolder populations are normally held in check by a rich fauna of natural enemies (Barrion et al., 1991). Predators too can be an important mortality factor for this pest (de Kraker et al., 2000), although the focus of the present review is the other important guild, parasitoids. Litsinger et al. (1987b) documented the percentage parasitism of leaffolders across several rice environments: 9–13% in dry-lands, 5–61% in the rain-fed wetlands and 7–33% in the irrigated wetlands. Most of the parasitoids of C. medinalis are Hymenoptera, comprising 91 identified to species level (plus 27 identified only to genus) from 12 families (Table 1). The dipteran parasitoid fauna is smaller, with 14 species (and one report with genus-level identification of one taxon) from three families (Table 2). In China, the hymenopterans Apanteles cypris, Apanletes ruficrus, Elasmus albipictus, Trichogramma japonicum, Itoplectis narangae and the dipteran Pseudoperichaeta nigrolinea are considered important (NPPS & ZAU, 1991; Cheng, 1996). In the Philippines, parasitoids have been reported attacking C. medinalis and leaffolders from the genus Marasmia (Pyralidae) (Barrion et al., 1991). Amongst these, the hymenopterans Cardiochiles philippinensis, Macrocentrus philippinensis, Goniozus nr. triangulifer, Brachymeria spp., Copidosomopsis nacoleiae, Ischnojoppa luteator, Temelucha stangli, Trichomma cnaphalocrocis and Trichogramma spp. are considered particularly to be important parasitoids of rice leaffolder (Barrion et al., 1991). Of the dipteran parasitoid species attacking C. medinalis, Megaselia sp. and Zygobothria ciliata are considered to be moderately important in the Philippines (Barrion et al., 1991) (Table 2). The overall impact of this enemy fauna is an important factor in the suppression of the Philippine leaffolder complex that includes C. medinalis (de Kraker et al., 1999a). There, egg parasitism (by Trichogramma) was in the range 0–27%, whereas larval parasitism (particularly by M. philippinensis) tended to be still more important, at 14–56% of the leaffolder complex community.
|Megaselia scalaris||India||Randhawa et al. (2006)|
|Megaselia sp.||Philippines||Barrion et al. (1991)|
|Pierretia caudagalli||China||NPPS and ZAU (1991)|
|Argyrophylax fransseni||Malaysia (Peninsular)||van Vreden and Ahmadzabidi (1986)|
|Argyrophylax nigrotibialis||Philippines||Barrion et al. (1991)|
|Chaetexorista javana||India||Randhawa et al. (2006)|
|Exorista japonica||China||NPPS and ZAU (1991)|
|Halidaya luteicornis||China||NPPS and ZAU (1991)|
|Lydella grisescens||China||NPPS and ZAU (1991)|
|Nemorilla floralis||India||Randhawa et al. (2006)|
|Nemorilla maculose||China||NPPS and ZAU (1991)|
|Pseudoperichaeta nigrolinea||China||Zhang et al. (1981) and NPPS and ZAU (1991)|
|Pseudoperichaeta nigrolinea||China||NPPS and ZAU (1991)|
|Thecocarcelia oculata||China||NPPS and ZAU (1991)|
|Zygobothria ciliata||Philippines||Barrion et al. (1991)|
Parasitoids also play an important role in controlling this pest in Vietnam (Lam & Thanh, 1983) (Table 1). Apanteles cypris and A. ruficrus were the most important parasitoids in the most recent Vietnamese study (Dung, 2006). Attack rates can be high; overall parasitism of larvae during 1978 at Tien Giang was in the range 33–90% (Lam, 1980). Parasitism of C. medinalis larvae in Vietnamese rice varies markedly with location, year, time of year and agronomic conditions. During the winter–spring rice crop of 1981 and 1982, for example, overall parasitism peaked at 9.5% and 35.2%, respectively (Lam, 1980). Generally, parasitism is higher in the October rice crop than that in the winter-spring rice crop. Accordingly, maximum parasitism in the October rice crop in the same 2 years was 41.4% and 54.9%, respectively (Lam, 1980). Apanteles cypris, Cardiochiles sp. (near C. philippinensis), Goniozus hanoiensis, Temelucha philippinensis and Copidosomopsis coni are common parasitoids (Table 1).
Species of the larval parasite Apanteles (Braconidae) are reported to be important in various parts of Asia other than Vietnam. For example, Apanteles opacus and A. cypris are common in Peninsular Malaysia (van Vreden & Ahmadzabidi, 1986). Work in Hubei Province China indicated that A. cypris is responsible for up to 50% of parasitism (Zhang & Huang, 2000), although rates were generally in the range 20–30% in Jiangsu Province (Zhang et al., 1981).
Attack of leaffolders by at least 34 species of Ichneumonidae has been reported widely from Asian countries (Table 1). Significantly, these include Xanthopimpla spp.. These pupal parasitoids complement the activity of egg and larval parasitoids and may be important in the overall impact of parasitoids on C. medinalis (de Kraker et al., 1999a). Similarly, the larval–pupal parasitoids Brachymeria spp. (Chalcididae) (Table 1) are important for targeting those host individuals that escape attack by egg and early-instar larval parasitoids.
It is clear from the overview outlined above that high C. medinalis parasitism is achieved in some locations at some times. Yet leaffolders, most notably C. medinalis, still frequently cause foliar damage that is considered by farmers as sufficiently severe (despite evidence for a lack of effect on grain yield) to warrant spraying. Pest monitoring by farmers may also be less rigorous than that of trained researchers, with fields being entered only as a last resort and decisions being influenced by factors such as seeing neighbouring farmers applying pesticide (Bandong et al., 2002). Nectar sources of the appropriate type might consistently help raise parasitoid performance for the more effective control of this pest and thereby reduce insecticide use.
The potential role of nectar sources
Enhancing the performance of parasitoids with adult foods such as honey solution (in laboratory rearing) and nectar (for enhancing field impact) is now widely practised (Wäckers et al., 2005).
Despite such conservation biological control leading to instances of enhanced pest mortality (Landis et al., 2000), there is a dearth of studies on enhancement by leaffolder parasitoids. The sole available example involving a parasitoid of C. medinalis is A. cypris, where the potential for adult nutrition to be improved is illustrated by an increase in adult longevity when given access to honey and water. In a study conducted in the Chinese autumn (September to November), adult longevity was 14.4 days with honey water, 8.4 days with water and 3.8 with neither honey nor water (Cheng, 1984). The results obtained were based on a total of 209 individuals, although they need to be interpreted with some caution because no statistical analysis was reported. The magnitude of this parasitoid's response to honey is modest compared with the effects of honey and plant nectar in experiments with other relevant parasitoids species. For example, C. medinalis is attacked by several species of Apanteles/Cotesia (Braconidae) (Table 1) and, although none of these have been investigated for enhancement of fitness by nectar, other species in the genus are well studied, particularly C. glomerata (Wäckers, 2001; Lee & Heimpel, 2008). Amongst such studies, adult feeding on plant nectar by C. glomerata was found to significantly increase flight capacity in female wasps (Wanner et al., 2006). For the congeneric C. congregata, longevity was increased approximately 8.5-fold when provided with buckwheat Fagopyrum esculentum flowers compared with wasps with only water (Witting-Bissinger et al., 2008). Similarly, at least two species of the encyrtid genus Copidosoma have been reported attacking C. medinalis in several Asian countries (Table 1) and the congeneric C. koehleri exhibits dramatically enhanced adult longevity and fecundity when feeding on the nectar from a wide range of plant species, an effect that was exploited to achieve better field control of the host insect, potato moth Phthorimaea operculella (Baggen & Gurr, 1998).
Several species of Trichogrammatidae egg parasitoids are reported from C. medinalis in Asia (Table 1). Of these, Trichogramma japonicum is important in Vietnam (Lam & Thanh, 1989; Lam, 1996, 2000, 2002) and is the most important egg parasitoid in China (Guo & Zhao, 1992) and Peninsular Malaysia (van Vreden & Ahmadzabidi, 1986). Trichogramma japonicum was the only parasitoid species to be found attacking the eggs of leaffolders C. medinalis and M. patnalis in a large study in the Philippines (de Kraker et al., 1999b). In that study, the mean parasitism rates for laboratory-reared and field-laid eggs were 15% and 18%, respectively. Trichogramma spp. have also been used for the inundative biological control of rice leaffolder in India (Pasalu et al., 2004). Trichogramma chilonis (Ishii) and T. japonicum released at rates of 100 000 wasps per hectare provided better control than did lower release densities (Kumar & Khan, 2005). However, the use of parasitoids in an inundative biological control strategy does not preclude there being a benefit of simultaneously using conservation biological control methods; an ‘integrated biological control’ approach sensuGurr et al. (2003). As an illustration of this effect, Begum et al. (2004) showed that Trichogramma carverae used in inundative releases against Epiphyas postvittana (Tortricidae) in Australian vineyards exhibited improved longevity and parasitism performance when provided access to the nectar of the white flowering alyssum Lobularia maritima. Parasitism of the target pest was successfully increased by establishing alyssum in vineyards after an inundative release of the parasitoid (Begum et al., 2006). Gurr and Nicol (2000) showed that T. nr brassicae adults also benefitted from food. Given these findings, it is likely that the release of Indian trichogrammatids into rice crops would similarly benefit from nectar availability, as would the naturally-occurring egg parasitoids present.
Several parasitoid species from Bethylidae have been reported from C. medinalis (Table 1) and, although none of these appear to have been studied from the perspective of adult nutrition, bethylids of the coffee berry borer Hypothenemus hampei have been investigated (Damon et al., 1999). Such work illustrated the magnitude of the benefit to parasitoids from this family of access to nectar. Both Cephalonomia stephanoderis and Prorops nasuta took nectar from Euphorbia hirta (Euphorbiaceae) and, under laboratory conditions, survived 55 and 17 days, respectively compared with the control treatment (longevity of 7 and 4 days, respectively). Although these differences are considerable, the study highlights a recurring problem in some studies of this topic: the control treatment consisted not only of no food, but also of no water. Although this may be a realistic baseline treatment for some arid agricultural systems where access to free water may be limited for parasitoids, it is not the norm and certainly is not applicable to irrigated rice. The same methodological caveat applies to a study of the effects of plant nectar on the stinkbug parasitoid Trissolcus basalis (Hymenoptera: Scelionidae) by Rahat et al. (2005). Only one member of this family, Telenomus dignus, has been reported attacking C. medinalis (Table 1), although future laboratory studies of this and other parasitoids need to include a comparison of performance with available free water.
Dipteran parasitoids in general have been the subject of far less research related to biological control enhancement and none of this has directly addressed the phorid and tachinid species recovered from C. medinalis (Table 2). However, other members from both of these families have been demonstrated to benefit from aphid honeydew (Fadamiro & Chen, 2005) or be attracted to flowering plants (Tooker et al., 2002).
Risks and challenges: choice of nectar plant species
Work carried out in Australia with T. carverae referred to above (Begum et al., 2004, 2006) illustrates the importance of the laboratory screening of potential nectar plants. This allowed the identification of plant species that provide nectar to the parasitoid but did not benefit the pest by being a host to the larvae or providing nectar to the adults. Such ‘selectivity’ (Baggen & Gurr, 1998) is a critical consideration in rice where C. medinalis and other Lepidoptera, such as yellow stem borer Scirpophaga incertulas (Pyralidae), are important pests (Litsinger et al., 2004) because lepidopteran adults often feed on nectar. It is well established that, under laboratory conditions, adult C. medinalis benefit from the availability of sugars (Waldbauer, 1979; Waldbauer & Marciano, 1979), including leafhopper honeydew (Waldbauer et al., 1980). Increased longevity and fecundity in the laboratory has naturally led to speculation that field biology is influenced by nectar (Miyahara, 1997). Although little is known about C. medinalis nutrition in the field (Waldbauer et al., 1980), the significance of nectar is borne out by a report of C. medinalis adults feeding on flowers of the Chinese chaste tree Vitex cannabifolia (Verbenaceae) (Wu, 1991). Large numbers were reported to be taking the nectar of this plant, especially between 21.00 and 10.00 h, although some adults remained in the trees at dawn. This observation was made in early summer (8–16 June) in the sub-tropical Chinese province of Jiangxi when any build-up of moth performance would be especially deleterious to the avoidance of foliar damage that tends to lead to early-season insecticide applications. A second, small Chinese study suggests that C. medinalis densities in a rice crop might be influenced by the availability of nectar from nearby plants (Zhejiang Agricultural University, 1982: 147). Larval densities in a crop in Jiangxi Province declined markedly with an increased distance from an adjacent area of Chinese milk vetch (Fig. 1). This pattern of spatial enhancement of a lepidopteran pest by available nectar is broadly consistent with the situation in a study reported by Baggen et al. (2000), where an inappropriate plant species (buckwheat) was sown adjacent to potatoes to enhance parasitoid performance. By contrast, when a selective food plant (borage Borago officinalis) was used, feeding of the moth was prevented and its parasitoids were enhanced.
Despite the foregoing cases of nectar use by C. medinalis adults, it remains to seen whether the use of nonselective nectar plants to favour parasitoids would also benefit the moths. Laboratory evidence for carbohydrates in the form of hemipteran honeydew affecting the performance of C. medinalis adults (Waldbauer et al., 1980; Pu, 1984) suggest that the availability of this ubiquitous carbohydrate in the field means there is little extra risk from nectar plants sown on rice bunds to favour parasitoids. A Chinese laboratory study by Pu (1984) used batches of six pairs of newly-eclosed adult C. medinalis, each caged on five rice plants, and examined the effect of 25% sucrose solution; peach potato aphid (Myzus percicae) honeydew (replaced daily) and the presence of ten brown plant hopper adults (Nilaparvarta lugens (Stål)) per cage. Although the study had only two replicates and no statistical analysis was conducted on the data, there were conspicuous treatment effects. Longevity and fecundity (but not egg hatchability) were strongly enhanced by carbohydrate availability, supporting the laboratory findings of Waldbauer et al. (1980).
The use of brown planthopper honeydew by C. medinalis under the highly artificial conditions of a laboratory experiment does not, however, mean that this phenomenon occurs in nature. Adult C. medinalis tend not to occupy young rice seedlings other than for nocturnal oviposition, favouring denser vegetation (J. A. Cheng, personal observations). Surprisingly little is known generally about honeydew use by lepidopteran adults, although this phenomenon has been reported from the field for two species of Noctuidae (Johnson & Stafford, 1985) and reports of use by Lycaenidae date back over a century (Bingham, 1907). It may be that, in nature, C. medinalis adults do not generally derive much benefit from pest honeydew. This notion is supported by the large volume of work showing that honeydew is less beneficial than nectar to parasitoids as a result of lower detectably, accessibility and nutritional quality (Wäckers, 2000). An additional factor that would limit the availability of honeydew to lepidopteran adults (with suctorial mouthparts for fluid feeding) compared with parasitoids (with more versatile mandibulate mouthparts) is that raffinose and melezitose (two common sugars in honeydew) rapidly crystallize (Wäckers, 2000). Johnson and Stafford (1985) suggest that, if honeydew use was common among lepidopteran adults, it would be well recognized because they tend to be large and conspicuous. Furthermore, the unpredictable temporal and spatial availability of honeydew is suggested as a reason for it not being widely used. Brown planthopper and other delphacids (significant pests of rice) produce honeydew but are often absent or rare early in the season, so that honeydew availability at the time of the usual peak of C. medinalis infestations is likely to be low and unpredictable. Taken with its sub-optimal qualities, this suggests that honeydew is unlikely to be important in the field nutrition of C. medinalis and parasitoids. In turn, this indicates that the addition of nectar sources might be significant for boosting the performance of lepidopterans and natural enemies in the tropical rice system. Because it is important to avoid benefiting the moth, studies are needed to identify plant species that can selectively provide nectar only to parasitoids.
An additional aspect to be considered in future work aiming to identify the most appropriate nectar plant species to support C. medinalis parasitoids is the possible enhancement of enemies of the parasitoids. This phenomenon was reported by Stephens et al. (1998) in a study where the major effect of nectar plants in a New Zealand apple orchard was reported to enhance the impact of Anacharis sp. Rather than attacking pest species, this parasitoid targets the brown lacewing (Micromus tasmaniae), which is an important predator of pests. Accordingly, it is important that nectar sources used in rice systems maximize benefit to primary parasitoids of C. medinalis and not to enemies of these biological control agents. This is a real danger because many such hyperparasitoids have been reported. For example, the primary parasitoids Apanteles flavipes and Apanteles ruficrus are attacked by Mesochorus discitergus and Elasmus sp. (Zhang et al., 1981) whereas A. cypris is attacked by Ceraphron manilae, Eurytoma verticillata, Niponohockeria sp., Elasmus sp., Eupteromalus parnarae and M. discitergus (Table 1). In Vietnam, A. cypris is attacked by Elasmus anticles and the congeneric Elasmus claripennis. Several members of the parasitoid family Ceraphronidae have been reported from C. medinalis (Table 1) but, similar to C. manilae, these are almost certainly hyperparasitoids rather than primary parasitoids of the lepidopteran.
The types of laboratory experiments used to identify nectar plants that selectively provide nectar to primary parasitoids and not to lepidopteran pests (Baggen & Gurr, 1998; Begum et al., 2006) might be extended to search for those that do not support hyperparasitoids. This is a challenging objective, however, because complex multitrophic interactions can occur and researchers are at an early stage with respect to developing approaches aiming to elucidate their nature. Jonsson et al. (2009) studied such interactions in the laboratory using arthropods that are important in a temperate fruit production system. They found that a predatory lacewing benefited only when aphid prey densities were low (presumably because lacewings were starving and needed the plant food). At higher prey densities, however, the effect of plant food was negative for the predator because the performance of its parasitoid (Anacharis zealandica) was enhanced. Future work of this nature needs to be undertaken for arthropod species of importance in rice and the scope of study will also have to be extended to allow experimentation under field conditions. Work in the field will allow phenological aspects to be covered, particularly to ensure that nectar plants flower in an indeterminate manner, thus providing nutrition for extended periods.