Nezara viridula
(green stink bug)

N. viridula has been widely investigated over a protracted period. The life cycle, ecology and behaviour of N. viridula have been reviewed (Todd and Herzog, 1980; Panizzi and Slansky, 1985; Waterhouse and Norris, 1987; Todd, 1989). N. viridula is classed by Todd (1989), as "... one of the most important pentatomid insect pests in the world ... It is cosmopolitan and highly polyphagous on many important food and fiber crops".

The ecology of N. viridula varies with locality, but the scale at which such variation manifests itself remains unclear. Interpretation of N. viridula ecology in any particular area requires considerably more information about the specific use made by both nymphs and adults of the various plant species available.

Life Cycle

The life cycle of green vegetable bugs is mostly multivoltine, with the extent of voltinism related to local differences in climate and the availability of host plants suitable for reproduction (Todd, 1989; Velasco and Walter, 1992; Cividanes and Parra, 1994a; Velasco et al., 1995; Chang and Chen, 1997).

Most green vegetable bugs undergo winter diapause, invariably in the adult stage, but in southern Brazil they continue reproducing on alternative host plants (Panizzi and Meneguim, 1989). Overwintering individuals generally, but not invariably, assume a russet colour (Seymour and Bowman, 1994). During warm periods they may emerge from their hiding places under bark, in litter, in thick vegetation, or behind panels of buildings, to feed (Kiritani et al., 1966; Todd and Herzog, 1980), which enhances survival. Cold tolerance has not been extensively investigated, and would be of differential importance for bugs at different latitudes. The survival of green vegetable bugs in different geographical localities, and the factors that influence their survival, have been described by Todd (1989) and Elsey (1993).

Adults leave their hibernation sites permanently in spring, start feeding, mainly nocturnally (Shearer and Jones, 1996b) and soon mate and oviposit. Eggs are deposited in the upper canopy of the herbaceous host plants, mostly under leaves or fruiting structures. They may take as long as 2-3 weeks to hatch in spring and autumn, but as few as 5 days in summer.

The first-instar nymphs do not feed, and form tight clusters at their natal site. Second- and third-instar nymphs also cluster, perhaps for protection, but they disperse if disturbed. Fourth- and fifth-instar nymphs do not aggregate. These older nymphs, along with adult green vegetable bugs, bask on the outer surface of the canopy or on the sunny side of those fruiting structures that emerge from the canopy. In the middle of the day, most bugs retreat into the canopy or to the shaded side of fruiting structures.

Nymphs in the fifth instar are sensitive to day length, which (in combination with temperature) determines entry into adult diapause (Todd, 1989). The developmental rate of nymphs is dictated by temperature and nutritional quality. Some studies indicate optimal temperatures are about 30°C, when the nymphal stage lasts about 23 days, but almost 8 weeks is needed to complete this stage at 20°C (Todd, 1989). A daytime temperature of 30°C for 14 h and a 10 h reduction in night-time temperature to 20°C resulted in an intermediate nymphal duration of 5 weeks in Australian bugs (Velasco and Walter, 1993a).

On primary host plant species (for example, soyabean), nymphs may develop up to twice as fast as on other species (for example, clover) at 27±5°C (Velasco and Walter, 1992). Brazilian nymphs, in contrast, do not survive well at 30°C, and oviposition by Brazilian bugs was greatest at 20°C, which led Cividanes and Parra (1994b) to conclude the species is better adapted to lower temperatures. Harris and Todd (1980a) discussed possible causes of such differences among studies of N. viridula.

The performance of N. viridula on some other host plants, and on various combination diets, has become available (Velasco and Walter, 1993b; Koymen and Karsavuran, 1995; Panizzi et al., 1996; Roychoudhury and Joshi, 1996; Shearer and Jones, 1996a; Panizzi and Mourao, 1999). Tests on macadamia demonstrate that N. viridula may be a pest on a crop, despite that crop species being a poor host and not supporting successful development of the nymphs (Shearer and Jones, 1996a) and though a host may be good for reproductive development of adult females, it may be relatively poor for nymphal development, as has been documented for Brazilian bugs on Japanese privet (Panizzi et al., 1996; Panizzi and Mourao, 1999).

Populations of N. viridula may have host relationships in one area that are significantly different from those of other populations, sometimes in the same general area (Todd and Herzog, 1980; Panizzi and Meneguim, 1989; Panizzi and Slansky, 1991). It is not known whether these are genetically different sibling species of N. viridula or whether these differences are imposed by host plant availability. Until recently, differences between populations in sound production and pheromone composition have been assumed to be produced by local adaptation within that locality. However, an alternative to that view has been suggested; the taxon N. viridula may contain unrecognized sibling species (Ryan et al., 1996) and recent tests support this interpretation (Jeraj and Walter, 1998).

That sibling species and not local adaptation may be involved is supported by the considerable migratory potential of the bugs, which means that the potential for gene flow among localities is thus considerable. In parts of Africa, they migrate in large groups under the influence of the Inter-Tropical Front (Bowden, 1973). Anecdotal evidence in Australia suggests that mass migration to overwintering sites may take place (Gu and Walter, 1989). Females in Japan reputedly fly up to a kilometre per day when moving between feeding and oviposition sites (Kiritani et al., 1965). Records of extended flights of hundreds of kilometres by individuals have also been documented (Hokkanen, 1986; Gu and Walter, 1989; Aldrich, 1990).

Pheromones

Nymphs have dorsal abdominal glands that secrete n-tridecane, which functions in nymphal aggregation and, at high concentrations, in dispersal (Lucchi and Solinas, 1990). Adults secrete n-tridecane from their paired metathoracic glands, as well as some (E)-2-hexenal and (E)-2-hexenyl acetate, which are adult alarm pheromones. These three compunds play a defensive role against certain potential predators including fire ants (Lucchi and Solinas, 1990; Pavis et al., 1994). Nymphs in aggregations suffer less predation than isolated nymphs (Todd, 1989). Another adult metathoracic gland compound, (E)-2-decenal, also attracts female egg parasitoids (Trissolcus basalis) (Mattiacci et al., 1993).

Questions remain, however, about the sex pheromones and the structures that secrete them. The pheromone gland is not discrete, but made up of numerous minute glands in the ventral subcuticular tissues (Lucchi, 1994). Several compounds are secreted by sexually active males (Aldrich et al., 1989; Borges, 1995). Evidence conflicts over their precise function. Field tests show that they attract adults of both sexes as well as fifth-instar nymphs and are thus said to be aggregation pheromones (Harris and Todd, 1980b).

In laboratory experiments, only sexually mature adult males are attracted by the compounds (Mitchell and Mau, 1971; Brezot et al., 1993). Most attention has been given to the cis- and trans-forms of bisabolene epoxide (Brezot et al., 1994) because bugs from different parts of the world were recorded having different ratios of these enantiomers (Aldrich et al., 1989, 1993). Two lines of evidence question this interpretation. Within-population variation in ratios is as great as that recorded among populations (Ryan et al., 1995) and different ratios of the cis- and trans-epoxides did not affect the attraction of females to males (Brezot et al., 1994).

Sound

The sexes communicate by means of substrate-borne vibrations in the form of a duet (Ota and Cokl, 1991; Ryan et al., 1996), which takes place after they arrive on a plant and have to locate one another (Ota and Cokl, 1991). Different strains have been documented (Harris et al., 1982; Kon et al., 1988) but recent research suggests that sibling species are involved. At least two species are known in the taxon N. viridula; bugs in Australia have significantly different songs and song alternation sequences from Slovenian bugs, and bugs from the two areas do not communicate effectively with one another even in confined spaces (Jeraj and Walter, 1998).

Due to the variable regulations around (de)registration of pesticides, your national list of registered pesticides or relevant authority should be consulted to determine which products are legally allowed for use in your country when considering chemical control. Pesticides should always be used in a lawful manner, consistent with the product's label.

IPM Programmes

Participation in soyabean IPM programmes in Georgia, USA, have been low, probably because of the risk involved (Szmedra et al., 1990). In contrast, a soyabean IPM programme developed in Brazil is seen as a 'spectacular success story of IPM implementation for a major crop over a wide area', and has returned substantial economic, ecological and social benefits (Moscardi, 1993). The N. viridula component involves release of Trissolcus basalis (at least in the state of Paraná), which are produced centrally. See section on Egg Parasitoids for more information on these releases. Insecticides are applied at half the recommended doses with 0.5% NaCl added to the spray tank (Corso, 1993).

In the southern states of Brazil, early-maturing soyabean varieties are planted to escape N. viridula attack. Such varieties have also been used successfully as trap crops to concentrate bugs before the true crop matures, when they are treated chemically or with mass-released T. basalis. But this tactic is still at the experimental stage in relation to IPM in Brazil. Biological control efforts against N. viridula are likely to be enhanced if carried out in association with selective plantings of species that encourage both N. viridula and its parasitoids (Bennett, 1990). Different tillage treatments for soyabean plantings did not influence N. viridula population dynamics, so the tilling operations can be selected for predator conservation (Funderburk et al., 1990).

The possible use of neem seed extract (azadirachtin) in IPM programmes in pecans has been investigated by Seymour et al. (1995). This plant-derived compound seems to reduce feeding by the bugs even at low concentrations, so may well prove useful.

Cultural Control and Sanitary Methods

Crop sanitation has been advised (Waterhouse and Norris, 1987).

Limited use has been made of trap crops to reduce the impact of N. viridula on soyabean, but the approach has good potential (Drake, 1920; Todd and Herzog 1980; Todd, 1989). Early-maturing varieties can be used as trap crops to protect the later maturing, true crop, but insecticides should be applied to the trap crop before the main crop reaches pod set (Kobayashi and Cosenza, 1987; Todd and Schumann, 1988). Adjustment of the planting date of crops, as well as row width (Hussain and Saharia, 1993), allows a degree of manipulation of N. viridula numbers (Rizk et al., 1990; McPherson and Bondari, 1991). Sesbania rostrata has been tested as a trap crop for soybean in Indonesia (Naito, 1996). Intercropping of soyabean with other crops has been tested. The most promising combinations of crops are non-related ones (Das and Dutta, 1996).

In orchards, planting species more attractive to N. viridula than the crop itself shows some promise (Umana et al., 1993). Weed control is thought to reduce catfacing damage to peaches (Killian and Meyer, 1984) but does not reduce damage in soyabean (Altieri et al., 1981). The recommendation for transgenic cotton in the USA is to plant early and avoid areas where alternative hosts are growing (Turnipseed and Greene, 1996).

Todd (1989) lists three further approaches to cultural control of N. viridula that offer substantial benefit:

- use of alternatives crops that are less attractive than soyabean;
- attraction of N. viridula bugs to localized areas of highly attractive hosts late in the soyabean season, with their subsequent eradication with insecticides;
- elimination of preferred overwintering sites.

Host-Plant Resistance

Soyabean germplasm with resistance to N. viridula has been identified, but little progress has been made in the development of agronomically acceptable resistant cultivars or in quantifying the impact of resistant varieties on N. viridula abundance and pest status (Todd, 1989). In Brazil, a variety (IAC-100) resistant to stink bugs in general has been released to growers, and transfer of resistant genes into otherwise more suitable genotypes shows promise (Moscardi, 1993). Soyabean genotypes with resistance to N. viridula may influence parasitism rates negatively, at least to some extent, and warrants consideration (Orr et al., 1985). Recent emphasis has been on the local testing of different soyabean genotypes for resistance in Brazil (Lourencao et al., 1997, 1999), the USA and Thailand (Suwanpornskul and Khadkao, 1996). Correlated measures have also been sought to facilitate selection for resistance (Lopes et al., 1997).

Other crops are also being tested such as spine gourd and summer green gram in India (Sarma and Dutta, 1997; Shaw et al., 1998). Different pecan and macadamia cultivars show differential susceptibility to green vegetable bug attack, which in turn is related to shell thickness of the nuts (Dutcher and Todd, 1983; Jones and Caprio, 1992).

Chemical Control

Fourth- and fifth-stage N. viridula nymphs and adults bask outside the plant canopy until about mid-day, so application of insecticides is most effective at that time. Damage to nut crops may not be restricted to the time they are on the tree. For example, full-sized macadamia nuts may be damaged whilst on the tree, but even more so up to 1 week after they drop to the ground, and this needs to be taken into account when applying chemical treatment (Jones and Caprio, 1994).

A range of carbamates and organophosphates may control N. viridula, but their persistence is too low to prevent subsequent outbreaks (Waterhouse and Norris, 1987; Martins et al., 1990).

Alternative insecticides (tralomethrin, lambda-cyhalothrin and acephate) are less toxic and may reduce production costs, increase yield and improve soyabean quality (Chyen et al., 1992; Le Page, 1996; McPherson, 1996). Some of the synthetic pyrethroids tested gave greater residual control than acephate; however, permethrin, did not control N. viridula in soyabean (McPherson et al., 1995).

Peaches have been protected effectively with chlorpyrifos (Johnson et al., 1985).

The permissible amount of stained rice grains (including pecky rice) is too low to establish reliable control thresholds, and prophylactic applications of insecticides tend to be made against hemipteran rice pests in Japan, with N. viridula being the principal species in southern Japan (Ito, 1986).

Trials on insecticidal plant extracts such as azidarachtin do not compare in efficiency with synthetic insecticides (Ivbijaro and Bolaji, 1990). Some research has been conducted on insect growth regulators but they do not appear to be effective (Canela et al., 1995; McPherson and Gascho, 1999).

Disease transmission by N. viridula to citrus fruits has resulted in the use of insecticidal control methods in Cuba (Grillo and Alvares, 1983).

In South America, the addition of NaCl to insecticides is recommended because it lowers the required dosage (see IPM Programmes).

Resistance and chemical control in soyabean were evaluated by Gazzione (1995) and Rosso et al. (1995).

Field Monitoring and Economic Threshold Levels

Control strategies against N. viridula in soyabean should be related to the stage of pod development (Brier and Rogers, 1991). Early pod fill (stage R5: Kogan and Turnipseed, 1980) is the most sensitive stage, and the only one in which yield, seed weight and oil content was significantly reduced (Brier and Rogers, 1991). Bugs should be controlled before this stage is reached, i.e. towards the end of pod elongation. Once pod fill is completed, soyabeans are not at risk and control is not warranted unless planting seed or edible seed is being grown (Brier and Rogers, 1991; Suzuki et al., 1991). For further specific information on when control methods are warranted on soyabean, see Kogan et al. (1977), Kobayashi (1981), Panizzi and Slansky (1985) and McPherson et al. (1993).

For sorghum, the damage threshold is about four N. viridula adults per panicle for infestation from the milk stage of grain development to maturity, but about 16 adults from the soft dough stage to maturity (Hall and Teetes, 1982b).

When milk-stage wheat kernels are infested at levels of more than two adults per 20 spikes, control measures are warranted (Viator et al., 1983). In maize, two N. viridula adults per V15 stage plant causes significant reduction in yield (Negron and Riley, 1987). In contrast, the permissible levels of stained rice grains are too low for reliable thresholds to be established (Ito, 1986).

Cowpeas suffer significant yield loss only when N. viridula density reaches 12 per metre of row (Nilakhe et al., 1981), whereas passion fruit is treated at one pentatomid per metre (Neethling, 1992). In transgenic cotton, an arbitrary threshold of 1 bug/6 feet of row has been used (Greene and Turnipseed, 1996).

Prediction of population peaks has been attempted with a day degree temperature model (Cividanes and Figueiredo, 1997).

Biological Control

The geographical origin of N. viridula is a matter of debate, which warrants attention because accurate knowledge of its original distribution may yield specific parasitoids useful to control efforts. The most likely origin is the Mediterranean area and/or the African mainland (Hokkanen, 1986; Jones, 1988). Biological control efforts across the world have been summarized in detail by Waterhouse (1998).

Egg parasitoids

N. viridula has long been a target for biological control, mainly through the introduction of parasitic wasps and flies (see Waterhouse and Norris (1987), Bennett (1990), Clarke (1990)). Although success has been widely claimed, especially in Australia and Hawaii (Caltagirone, 1981; Waterhouse and Norris, 1987; Todd, 1989; Bennett, 1990) the degree of that success has been seriously questioned through the research of Clarke (1990, 1992) in Australia, and Jones (1995) in Hawaii.

The problems with the claims of success are as follows:

- The parasitoid said to have achieved successful control in Australia was the so-called Pakistan strain of Trissolcus basalis. However, T. basalis does not exist in Pakistan. The wasps sent originally from Pakistan to Australia are morphologically distinct from T. basalis and represent a new species, T. crypticus (Clarke, 1993). These are evidently parasitoids of species other than N. viridula in Pakistan (although they parasitize N. viridula eggs successfully in the laboratory) and have probably not established in Australia (Clarke, 1993).

- Parasitism of N. viridula eggs by T. basalis is relatively low in many situations, for example, in Hawaiian macadamia orchards (Jones, 1995) and soyabean in south-east Queensland (Clarke and Walter, 1993b).

- In many studies, most of the destruction of N. viridula eggs in the field was attributable to predators, mainly ants (Jones (1995) in Hawaii; Van den Berg et al. (1995) in Sumatra, Indonesia).

- Adequate post-release studies were never conducted on N. viridula populations, so the impact of the released natural enemies was never quantified. Any decline in N. viridula numbers at the time of natural enemy releases (at least in Australia) can be explained by changing agricultural practices, including the widespread introduction and prophylactic use of synthetic organic insecticides. Reduced use of synthetic organics in Australian vegetable crops has seen the re-emergence of N. viridula as a pest in areas where it was considered to have been under good biological control, but this has not been quantified.

The widely accepted success of T. basalis as a biological control agent led to it being distributed around the world for release, despite inaccurate claims of its inability to parasitize N. viridula eggs efficiently on soyabean (Turner, 1983) and its being restricted to coastal areas (Jones, 1988). The species achieves high levels of parasitism in soyabean in several parts of the world and is sometimes also abundant inland (Seymour and Sands, 1993).

As a consequence of its reputed success as a biological control agent in Australia and Hawaii (Waterhouse and Norris, 1987) T. basalis has been imported for release into places where it was already present, including Brazil (from Australia: Corrêa-Ferreira and Zamataro, 1989). The impact of such multiple strain introductions has never been accurately assessed (Clarke and Walter, 1995).

That the claimed results of these biological control projects are being questioned does not imply that T. basalis is not worth importing, and it has, since 1980, been introduced into Argentina from Australia (de Crouzel and Saini, 1983) and into California from France, Spain and Italy (Hoffmann et al., 1991). The species evidently causes high levels of parasitism in parts of South America (Corrêa-Ferreira and Moscardi, 1995, 1996), Pohnpei (Esguerra et al., 1993) and in Australian pecan orchards (Seymour and Sands, 1993).

The influence of T. basalis as a biological control agent will possibly be enhanced by the development of mass-release programmes for situations in which N. viridula populations breed up in a crop and thus attain pest status. Release programmes in Brazilian soyabean are reputedly successful (Moscardi, 1993; Corrêa-Ferreira and Moscardi, 1996) and the strategy has been tested elsewhere on other crops (Justo et al., 1997; Cameron and Rea, 1998).

Techniques to facilitate the production of egg parasitoids are quite advanced. In addition to mass-rearing techniques (Oi, 1991; Awadalla, 1996), storage of frozen hosts (Corrêa-Ferreira and Moscardi, 1993; Corrêa-Ferreira and de Oliveira, 1998), identification and development of kairomones (Sales, 1985; Bin et al., 1993; Mattiacci et al., 1993) and development of artificial hosts (Volkoff et al., 1992) have all been tackled, but none is yet available for commercial use.

Biological control by egg parasitoids is virtually never hampered by the influence of hyperparasitism, because hyperparasitoids of egg parasitoids are unusual. However, researchers should be aware that two species of pteromalids in the genus Acroclisoides parasitize T. basalis in the field in Australia (Clarke and Seymour, 1992).

Parasites of stages other than eggs
Parasitoids that attack nymphs and adults of N. viridula have also been used in biological control, and these organisms are being increasingly investigated for biocontrol purposes (see Parasitoids). These organisms show some potential but, to date, none has proved as successful as the egg parasitoids. Species from the area of origin of N. viridula may worth trying. In particular, of the species present in Africa, the most promising may be Bogosia antinorii, as it is apparently specific to N. viridula (Waterhouse, 1998).

Predators
An increasing number of publications claim that ants have a strong impact on N. viridula eggs and nymphs, both in orchards (Yang, 1984; Seymour and Sands, 1993; Jones, 1995) and in soyabean fields (Krispyn and Todd, 1982; van den Berg et al., 1995). The thoughtful manipulation of ant populations could contribute considerably to achieving acceptable levels of biological control. Spiders may also contribute significantly to reducing N. viridula populations, but the data are not convincing (Kiritani and Hokyo, 1962) and they are still being investigated in the laboratory (Sarma and Dutta, 1996b).

Pheromones
To date, no practical use has been made of N. viridula pheromones, either for monitoring or control purposes. Disagreement still exists as to the identity of the functional components in the pheromone blend.