Nephotettix virescens (green paddy leafhopper)

Small numbers of insects may not visibly damage the plant. Virus-diseased plants may show various symptoms, such as stunted, deformed leaves, increased tillering, gall formation, or yellowing, depending on the virus (for further details, see virus datasheets and Brunt et al., 1990).

Integrated Pest Management

IPM programmes for the control of N. virescens and some other rice pests are increasingly advocated to reduce the use of pesticides (for a review, see Way and Heong, 1994). The manipulation and encouragement of predators and parasitoids is likely to be of considerable importance in most countries where tungro disease is important. The tungro viruses (RTBV and RTSV) transmitted by N. virescens are more damaging than the insect itself.

Cultural Control and Sanitary Methods

Ratoon rice may serve as a reservoir for both insects and virus diseases. Grass weed species may also harbour viruses. Sanitation in and around seedbeds and fields, covering wet seedbeds with nylon screens at a height of 60 cm, and simultaneous cropping are used against N. virescens (Mochida et al., 1986). Tungro outbreaks can be triggered by the presence of early infected paddy fields in asynchronously transplanted areas. For the control of Rice bunchy stunt virus in China, since the disease attacks rice plants most easily in the seedling stage and green recovering tiller stage, control was effected by regulating sowing and transplanting times to escape the migrant peak of the vectors (Xie et al., 1984). The vectors should also be controlled at the time of sowing and transplanting.

Kiritani et al. (1983) concluded that changes in cropping practices in Japan, such as an increase in the cultivation of early rice and the earlier planting of some mid-season rice, facilitated disease spread and led to serious epidemics. However, the duration and frequency of these epidemics have been reduced by subsequent changes in cultural practices, including the decreased cultivation of wheat and barley, the use of new rice cultivars and insecticides, and winter and spring ploughing of fallow ricefields.

Host-Plant Resistance

Resistant varieties are used as part of an integrated control programme (see, for example, Heinrichs et al., 1986; Myint et al. 1986). Current breeding programmes include screening for resistance to N. virescens (Pathak and Khan, 1994). Myint et al. (1986) investigated the levels of resistance to N. virescens of nine rice varieties in greenhouse and insectary trials in the Philippines. On the basis of seedbox screening, survival and oviposition, and feeding and population growth tests, IR29 and IR56 were resistant, IR8, IR36 and IR42 moderately resistant, and IR22, IR46, Utri Rajapan and TN1 susceptible. The degree of predation of leafhopper by Lycosa pseudoannulata and Cyrtorhinus lividipennis was the same on all varieties, regardless of the level of varietal resistance to N. virescens. Mortality of N. virescens was highest and populations lowest in treatments where the resistant varieties IR29 and IR56 were combined with predators.

Control by Fungal Pathogens

The use of fungal pathogens for control is at a preliminary stage of investigation; fungi are more prevalent at times of high humidity (Nayak and Srivastava 1979; Li 1988). In the Changsha region of China, Entomophthora delphacis generally infected from 37 to 64% in Nilaparvata lugens and the same fungus also attacked N. virescens (Li, 1988; Ambethgar, 1997). Beauveria bassiana spores were found on nymphs and adults of N. virescens in the ricefields of Cuttack, India (Nayak and Srivastava, 1979). Hirashima et al. (1979) found the same fungus in Thailand. In India, Beauveria bassiana attacks N. virescens and N. nigropictus during September-November (Gupta and Pawar, 1989).

Field Monitoring/Economic Threshold Levels

Thresholds are employed as part of the strategy to control tungro disease. Suzuki et al. (1992) investigated the build-up of rice tungro disease infection in paddy fields in Bali, Indonesia and found that practical control thresholds could be established for monitoring 2-5 weeks after transplanting, on the basis of percentage of diseased hills. The areas which might be infected in the first half of the wet season could be predicted from the number of infected locations during the second half of the dry season. Increase in virus infection was preceded by population build-up of the vector. Increased migration in the first-generation adults accounted for population build-up of the vector. Conditions for severe outbreaks were: tungro intensity in paddy fields that are under young plants reaches over four times the economic control threshold; and, at the time of transplanting, the mean infective vector index in migrant-producing fields in the area exceeds (15 insects per 25 strokes) per 100 hills.

Economic losses due to direct feeding damage by N. virescens may sometimes be masked by losses caused by the viruses it transmits. An added complication is that, for some historical data, it is not clear whether yield loss has been attributed to direct feeding damage or to a virus disease. N. virescens is amongst the most widespread and abundant canopy arthropods in irrigated rice ecosystems (Heong et al., 1991). However, populations are rarely large enough to cause significant yield loss through direct feeding damage. Although N. virescens is a phloem feeder on susceptible rice varieties, it does less damage than planthoppers such as Nilaparvata lugens and Sogatella furcifera. These planthoppers, which tend to feed on rice stems rather than leaves, can cause blockage of the vascular system by leaving stylet sheaths in the plant tissue after withdrawing their stylets (Sogawa, 1982). Blockage of the vascular system has not been demonstrated for cicadellids such as N. virescens, though feeding by large numbers of insects can cause wilting and drying of rice plants through excessive removal of assimilates.

Numbers of N. virescens are sometimes very high on irrigated rice crops in South India, with large numbers attracted to lights even in urban areas (Anjaneyulu et al., 1994). It has been estimated that in five states in southern India between 17 and 23 kg/ha of grain yield was lost to N. virescens in 1990-1991 and an additional 15 to 24 kg/ha to tungro (Ramasamy et al., 1996). In rainfed lowland areas in East India, an estimated 15 kg/ha of grain yield was lost due to direct feeding by N. virescens in 1990-91 (Ramasamy and Jatileksono, 1996). The same authors reported that yield losses of 12 kg/ha occurred in Thailand. Setboonsarng (1996) reproduced data from the Department of Agricultural Extension in Thailand in which the average area damaged by N. virescens feeding in 1986-1991was shown as 12,997 and 19,554 ha in rainfed and irrigated areas, respectively.

However, all these figures on production losses due to direct feeding by N. virescens should be treated with caution. Some appear to be based on estimates linked to economic thresholds for the insect. One such threshold used as a basis for insecticide recommendations by a government research institute in eastern India is two insects per rice plant (Chancellor et al., 1997). This number of leafhoppers would have a negligible effect on the rice plant through direct feeding but could spread tungro rapidly through a field if there was a strong source of virus inoculum present (Shukla and Anjaneyulu, 1982; Chancellor et al., 1996). Most economic thresholds, such as the one for eastern India, do not take into account the presence or absence of virus inoculum, perhaps because of the difficulties in identifying inoculum sources. An exception, is the threshold developed for tungro control in Bali, Indonesia which relates insect numbers to the proportion of diseased plants in fields at a particular growth stage (Suzuki et al., 1992).

The major economic impact of N. virescens is through the rice viruses it transmits. The most important of the diseases caused by N. virescens-mediated virus transmission is rice tungro. Tungro is a composite disease caused by Rice tungro bacilliform virus (RTBV) and Rice tungro spherical virus (RTSV) (see respective data sheets for each virus) and has a complex epidemiology (Hibino et al., 1978). Yield losses due to tungro are primarily due to infection with RTBV. The disease has a high profile because of its ability to cause total loss of both grain and straw yield and because there are no effective control measures once rice plants have become infected. Irrigated areas are most at risk, although tungro may also occur in rainfed lowland rice.

A commonly cited figure for annual economic losses due to tungro throughout Asia is US$ 1500 million (Herdt, 1988). However, this is likely to be an overestimate as the frequency and extent of large-scale outbreaks of tungro has declined since the period of peak occurrence in the 1960s and 1970s (Chancellor and Thresh, 1997). In a recent multiple pest and disease survey conducted in four Asian countries, tungro incidence was low, illustrating the sporadic nature of the occurrence of the disease (Savary et al., 2000). However, the results of parallel experimental manipulations demonstrated that tungro was one of the most harmful of the biotic stresses on rice. The relatively low risk of occurrence of tungro, but the high level of losses caused in affected fields, was reflected in farmer perceptions recorded in a household survey conducted in West Bengal, India in 1992 (Saha et al., 1996). A total of 4.5% of respondents reported losses from tungro in 1992. They estimated that the probability of tungro occurring in their field was 0.05, the lowest of the biotic stresses mentioned. However, the production loss in affected areas due to tungro was estimated to be 2.6 tonnes/ha, which was the highest value of any of these stresses.

Although major epidemics of tungro are less common than in previous decades, major outbreaks still occur. In 1990-1991, an outbreak occurred in India in which a total of 134,000 ha of rice in Andhra Pradesh and 49,000 ha in Orissa were affected by tungro (Ramasamy and Jatileksono, 1996). In 1995, an epidemic was reported from Central Java in Indonesia where 12,340 ha of rice were affected causing an estimated loss of US$ 1.87 million (Daradjat et al., 1999). In 1998, 40,000 ha of rice in Punjab, India were damaged by an outbreak of yellow stunt syndrome that has similar symptoms to tungro (Azzam et al., 1999). This outbreak, in which yield losses were estimated at 30-100% in the affected area illustrates one of the difficulties associated with assessing the economic impact of tungro. Symptoms of tungro may be confused with other plant disorders and in some regions appropriate diagnostic techniques are not readily available to confirm the cause of the problem.

In some areas, tungro is endemic and regularly causes economic losses. In certain districts in Bali, Indonesia tungro is a problem in most years even though the total area damaged may not exceed a few hundred hectares (Astika, 1999). On Mindanao in the Philippines, between 900 and 2700 ha were affected annually by tungro during 1993-1998 with the disease occurring every year in the North Cotabato district (Truong et al., 1999). Disease incidence on this scale may not represent significant economic loss when viewed on a national basis. Teng and Revila (1996) cited a study in Malaysia covering the period 1981-1984 when outbreaks of tungro occurred in the Muda irrigation scheme and caused much concern amongst policy makers. These authors pointed out that surveys revealed that when losses were averaged across all rice-producing areas in Malaysia the mean yield loss was less than 1% (Heong and Ho, 1986). However, this perspective does not take into account the serious effects on the livelihoods of those farming communities affected. Many Asian rice farmers rely on credit to purchase inputs such as fertiliser and to pay for labour costs. One crop failure can very seriously affect the economic security of a farm household.

Other virus and phytoplasma diseases transmitted by N. virescens include Phytoplasma oryzae, Rice bunchy stunt virus, Rice gall virus and Rice yellow stunt virus. However, each of these pathogens is transmitted by other cicadellid vectors and with a greater degree of efficiency than by N. virescens (Ou, 1985). Experimental field data on rice yellow dwarf disease in Taiwan showed that 93% yield loss was possible when rice plants were infected at the seedling stage and losses declined to 4% when infection occurred at 50 days after sowing (Chen and Ko, 1975). A similar association between age of infection and yield loss was also demonstrated for rice transitory yellowing which caused significant crop losses in Taiwan in the early 1960s (Ou, 1985).

An assessment of the economic impact of N. virescens should take into account direct costs to farmers as a result of spraying insecticide to control the insect. Farmers in the Philippines who had fields affected by tungro disease continued to spray against the vector, even though they knew from experience that it was unlikely to be effective (Warburton et al., 1997). Much unnecessary insecticide spraying also takes place in situations where tungro is not present. This spraying carries significant costs to human health, particularly to those applying the chemicals, as well as wider environmental costs (Rola and Pingali, 1993). Costs of resistance breeding programmes should also be considered when evaluating the overall economic impact of N. virescens.