Cookies on Plantwise Knowledge Bank

Like most websites we use cookies. This is to ensure that we give you the best experience possible.


Continuing to use means you agree to our use of cookies. If you would like to, you can learn more about the cookies we use.

Plantwise Knowledge Bank

Your search results

Species Page

brown planthopper

Nilaparvata lugens


You can pan and zoom the map
Save map

Host plants / species affected

Main hosts

show all species affected
Oryza (rice (generic level))
Oryza sativa (rice)
Zizania (wild-rice)

List of symptoms / signs

Leaves - abnormal colours
Leaves - honeydew or sooty mould
Stems - external feeding
Whole plant - discoloration
Whole plant - external feeding


N. lugens damages plants by sucking sap from the mesophyll and blocking the xylem and phloem by laying egg masses in the midribs of the leaf sheath and leaf blade. Affected plants become chlorotic. Older leaves turn progressively yellow from the tip to the midpoint of the leaf, then gradually dry up and die. This feeding damage is commonly referred to as hopperburn. Hopperburn begins in patches but can spread rapidly as the planthoppers move from dying plants to adjacent plants. The physiology of this process is described by Sogawa (1982). Excreted honeydew on infested plants may also become a medium for sooty mould fungus.

In addition to direct feeding damage, N. lugens is also the vector of rice grassy stunt tenuivirus and rice ragged stunt oryzavirus.

Prevention and control


N. lugens, the rice brown planthopper (BPH) was a relatively minor rice pest prior to the Green Revolution and the associated increase in pesticide use in rice. In the 1960s and 1970s, during the early years of the Green Revolution, N. lugens became the number one threat to rice production in many parts of Asia (IRRI, 1979). However, the pest was brought under control through decreased insecticide use and the introduction of resistant varieties and most outbreaks are now sporadic and small in scale. It is widely accepted that existing natural enemies in Asian rice areas are the key to BPH management. In most cases, natural enemies can maintain N. lugens populations below damaging densities (Kenmore et al., 1984; Ooi, 1988).

Organophosphate and pyrethroid pesticides are known to be toxic to most natural enemies. Their widespread use, and price supports, have led to major BPH outbreaks in many countries due to resurgence. Other pesticides, including fungicides, are also known to be highly toxic to natural enemies and are suspected of increasing N. lugens resurgence. 'Preventive sprays' for stemborers and leaf folders may lead to pest resurgence (Heinrichs et al., 1978; Kenmore et al., 1984).

IPM Programmes

Integrated pest management (IPM) programmes should always consider the possibility of BPH resurgence.

IPM programmes for pest control in rice include the following components:

- avoidance of those pesticides (insecticides and fungicides) which are toxic to natural enemies in the rice ecosystem
- restricted application of pesticides in the rice ecosystem, applying only when necessary using narrow-spectrum compounds (including fungicides), and with the knowledge that application may lead to higher N. lugens populations
- regular field observation for rice pests and diseases
- training of field extension staff and farmers to improve their
ecological knowledge, observation skills, and economic analysis of pest management inputs (Matteson et al., 1994).

In the 1980s, it was recognized that the over-use of insecticides was the root cause of outbreaks. Biological control, complemented by host-plant resistance, is now seen as the basis of management of N. lugens (Way and Heong, 1994). IPM programmes emphasize that the routine use of insecticides should be avoided (Gallagher et al., 1994; Matteson et al., 1994).

In temperate areas, where N. lugens does not overwinter, waves of immigration in each cropping season can lead to a sudden build-up of the population. Consequently, insecticide use is more often needed in these areas than in the tropics, but should still be kept to a minimum.

Cultural Control

Increasing nitrogen levels, closer plant spacing, and higher relative humidity are known to increase N. lugens populations, but not to a level that is economically significant when natural enemies are present (Uhm et al., 1985).

Populations of stemborers (Chilo spp., Scirpophaga spp., Sesamia spp.) and leaf folders (Cnaphalocrocis medinalis, Marasmia spp.) may also increase in response to high nitrogen use, so that insecticides may need to be applied, which in turn can lead to BPH resurgence. High nitrogen levels may also have an indirect impact on natural enemies of N. lugens; however, by increasing their population levels. Increased nitrogen levels also provide favourable conditions for pathogens, especially sheath blight (Rhizoctonia solani [Thanatephorus cucumeris]) and rice blast (Magnaporthe grisea). High disease levels usually result in the use of fungicides which are also toxic to natural enemies.

Draining ricefields can be effective for reducing BPH at initial infestation levels (Heinrichs, 1979). Reissig et al. (1986) recommended growing no more than two crops per year and using early-maturing varieties. Judicious use of fertilizer by splitting nitrogen applications can also reduce planthopper outbreaks (Litsinger, 1994). The field should be drained for 3 or 4 days when heavy infestations occur (Reissig et al., 1986).

Synchronous planting, including planting neighbouring fields within 3 weeks of each other and maintaining a rice-free period, may be effective (Reissig et al., 1986) but this approach is controversial (Way and Heong, 1994). Asychronous rice cultivation within areas provides better continuity of natural enemy populations.

Biological Control

Existing species and levels of natural enemies in Asian rice areas are currently regarded as the key to BPH management (Kenmore et al., 1984; Ooi, 1988). N. lugens is normally controlled at low levels by the numerous predators, egg and nymphal parasites, pathogens and nematodes found in ricefield environments. See List of Natural Enemies.

In most rice ecosystems, no classical or indundative biological control is necessary for N. lugens because the naturally occurring predators and parasites are sufficient for economic control in almost all cases. Current research indicates that predators build up their populations on detritivores, and therefore predator populations are not dependent on planthopper populations. Conservation of these natural enemies is a major consideration when managing other insects and diseases in the rice crop. In tropical rice, populations of N. lugens are generally kept below damaging levels in the field by a large and diverse complex of natural enemies. Populations of biological control agents can be severely disrupted by use of broad-spectrum insecticides (Kenmore et al., 1984). Non-rice habitats surrounding rice fields have been shown to serve as refuges for egg parasitoids of BPH (Yu XiaoPing et al., 1996, 1998).

Anagrus spp. and Oligosita spp. are the most important egg parasitoids, while the mirid Cyrtorhinus lividipennis is often the principal egg predator.

The beetles Micraspis and Coccinella, the bug Microvelia, and the spider Lycosa pseudoannulata, are important predators of mobile N. lugens nymphs and adults.

Host-Plant Resistance

After outbreaks became common due to the increased use of pesticides, host-plant resistance became a major control method for N. lugens. Numerous host-plant resistance genes have been identified and incorporated in most breeding lines (Pathak and Saxena, 1980). The first BPH-resistant variety, IR26, which contained the Bph1 gene, was released by IRRI in 1973. IR26 and subsequent resistant varieties provided important short-term protection against BPH outbreaks, but in some areas BPH populations adapted to the new varieties in as little as 3 years (Den Hollander and Pathak, 1983; Gallagher et al., 1994). IR64, one of the most popular varieties in the Philippines, Indonesia and Vietnam, has shown more durable resistance than most varieties, apparently because it contains numerous minor genes (Cohen et al., 1997; Alam and Cohen, 1998a, b). Host-plant resistance breaks down due to the high variation in field populations of N. lugens. Breakdown may happen more quickly when pesticide applications continue at high levels (Gallagher et al., 1994). When pesticides are used, there is greater survivorship (due to less mortality by natural enemies) of pre-adapted individuals in field populations (Gallagher et al., 1994).

Over 50,000 rice accessions from the germplasm collection at the International Rice Research Institute (IRRI), Los Baños, Philippines, have been screened since 1966, and many additional traditional varieties and wild species have been screened at other research institutes throughout Asia. An overview of sources of resistance and techniques for evaluating the level and mechanisms of pest resistance is provided in Heinrichs et al. (1985).

Approximately 10 major genes for resistance to N. lugens have been identified from rice germplasm, and many cultivars also show minor gene resistance. Three resistance genes have been used extensively in modern, semi-dwarf cultivars. The Bph 1 gene was first released by IRRI in 1973, in variety IR26. This gene is also in IR64, currently one of the most popular rice varieties in Asia. The bph 2 gene was first released in 1975 in IR32, but received its most widespread use in IR36, released in 1976. In 1982, the Bph 3 gene was released in IR56. This gene is also present in several recently released IRRI varieties, including IR72 and IR74.

Varieties with major resistance genes at first confer high levels of resistance against N. lugens populations. However, when planted over large areas with little varietal diversity, these varieties have shown resistance breakdown in as little as 3 years. Resistance breakdown proceeds as the proportion of N. lugens adapted to the variety increases in the local population, and can be accelerated by the over-use of insecticides and fertilizers, which leads to a more rapid population growth of surviving N. lugens.

The word 'biotype' is often used to describe N. lugens populations or individuals with virulence to a previously resistant variety. The terms 'biotype 1', referring to N. lugens avirulent to varieties with any major resistance genes; 'biotype 2', referring to N. lugens virulent to varieties with Bph 1; and 'biotype 3', referring to N. lugens virulent to varieties with bph 2, are widely used in the rice literature. N. lugens virulence to the major resistance genes is polygenic, and does not follow the gene-for-gene pattern often seen in interactions between plant pathogens and resistant varieties (Roderick, 1994).

Antixenosis, antibiosis and tolerance have all been observed as mechanisms of resistance against N. lugens in various rice cultivars. No morphological or anatomical features have been found to be associated with N. lugens resistance, but surface and phloem chemistry have been linked to antixenosis and antibiosis. Antixenosis of rice varieties IR36 and IR62, manifest as reduced probing and settling, and movement from the stem to the leaves, has been associated with the chemical composition of surface wax (Woodhead and Padgham, 1988). Probing and sucking stimulants and inhibitors have been identified in various cultivars (Sogawa, 1982; Heinrichs, 1994). Tolerance to N. lugens feeding has been identified in the rice cultivars Mudgo, IR46, Triveni and Utri Rajapan (Velusamy and Heinrichs, 1986).

Wide hybridization of Oryza sativa with wild rice species, and the use of genetic engineering, are providing new sources of resistance against N. lugens. In 1993, varieties with a BPH resistance gene from the wild rice O. officinalis were released in Vietnam (IRRI, 1995; Nirenberg and O'Donnell, 1998). BPH resistance genes have also been transferred to O. sativa from O. australiensis (Multani et al., 1994).

The location of three BPH resistance genes on rice chromosomes are now known in reference to DNA markers, i.e. the genes have been 'mapped'. These genes are Bph1 (Hirabayashi and Ogawa, 1996; Huang Ning et al., 1997; Jeon YongHee et al., 1999), bph2 (Murata et al., 1998), and Bph10 (Ishii et al., 1994). All three genes are located on chromosome 12. Minor genes, or 'quantitative trait loci' (QTLs) for BPH resistance have also been mapped, using a mapping population of doubled-haploid derived from a cross between IR64 and Azucena (Alam and Cohen, 1998a).

Two types of genes have been used for genetic engineering of rice for BPH resistance: the snowdrop (Galanthus nivalis) lectin, also known as GNA, (Powell et al., 1998; Rao et al., 1998) and the soybean Kunitz trypsin inhibitor (SKTI) (Lee SooIn et al., 1999). BPH on GNA-transformed plants showed reduced survival, feeding and growth (Rao et al., 1998), and reduced survival was also found on the SKTI-transformed plants.

Japonica types seem to have some resistance as compared to hybrids in the temperate North-East Asian regions (Uhm et al., 1985).

Although BPH outbreaks can generally be avoided by preventing overuse of insecticides, resistant varieties can provide some 'insurance' against outbreaks caused by insecticide overuse, unusual weather patterns and other factors. Resistant varieties can also help to reduce pesticide use and thus assist in the build-up of natural enemies in areas where pesticides have been heavily used in previous seasons.

Chemical Control

Due to the variable regulations around (de-)registration of pesticides, we are for the moment not including any specific chemical control recommendations. For further information, we recommend you visit the following resources:

This information is part of a full datasheet available in the Crop Protection Compendium (CPC); For information on how to access the CPC, click here.



N. lugens is probably the most serious insect pest of rice in Asia. Its feeding causes plants to wilt and causes a symptom called hopperburn. It also transmits grassy stunt and ragged stunt virus diseases (Reissig et al., 1985). N. lugens was a minor rice pest until the mid-1960s in much of tropical Asia (Pathak and Dhaliwal, 1981). However, it assumed the status of the most destructive pest in the 1970s (Heinrichs and Mochida, 1984).

Colonization and Damage

Hopperburn, which is a characteristic symptom of N. lugens damage, does not usually appear until the crop reaches the milk or dough stages; in its vegetative stage, rice can tolerate a population of 100 to 200 nymphs/hill without showing any outward symptom of injury on the plants. Therefore, hopperburn appears at a very late stage of infestation (Kulshreshth et al., 1976).

When the food resources available to the brown planthopper become limiting, as in a crop nearing maturity or in hopperburned rice fields, a preponderance of the long-winged macropters is produced; these macropters disperse and colonize new verdant rice fields (Saxena et al., 1981). The brown planthopper disperses and colonizes rice fields in relatively calm weather conditions (Ohkubo and Kisimoto, 1971; Ohkubo, 1973; MacQuillan, 1975). Nitrogen fertilizer, high tillering varieties and/or close spacing and good water management are factors that contribute to a population increase. Double/continuous cropping associated with good water management may also have contributed to this gradually increasing pest problem. However, in areas where these contributing factors are present, infestation has not always been serious (IRRI, 1975).

During the 1970s, N. lugens caused severe damage to the rice crop in many countries in tropical Asia (Dyck and Thomas, 1979). Several outbreaks have occurred in the Philippines (Calora, 1974), Solomon Islands (Stapley, 1975), Thailand (Tirawat, 1975), Sri Lanka (Fernando, 1975), Vietnam (Huynh, 1975), India (Kulshreshth et al., 1974), Malaysia (Ooi, 1977) and Indonesia (Mochida, 1979). In the Philippines, N. lugens damaged at least 80,000 ha in 1973-74 (IRRI, 1979).

The attack of rice crops by N. lugens during the vegetative phase resulted in fewer panicles/unit area and lower mean grain weight. Attack between the booting and heading stages and after anthesis reduced the percentage of ripened grains and number of grains/panicle (Chen et al., 1979).

Greenhouse studies at IRRI, Philippines, indicated that large insect populations (above 50/hill) on plants of all ages tested, caused a reduction in the number of tillers, number of panicles and total grain weight per plant. 50 nymphs or 4 adults feeding for a period of 2 days and 2 weeks on plants at 25 days and 50 to 75 DAT (days after transplanting), respectively, did not cause any significant damage. This suggests that although young plants are very susceptible to brown planthopper damage, those at 50 to 75 DAT can tolerate a population below 90 nymphs or 4 adults per plant for several days without any significant damage (IRRI, 1967).

The effects of infestation with N. lugens on rice after flowering and during the milking stage were investigated (Zhang et al., 1997). Though dry matter increased with dry days after flowering, all parameters used to measure dry matter decreased significantly with N. lugens infestation.

Crop Losses

In a 117-day crop period at IRRI in the 1978 dry season it was indicated that at a population of 78 N. lugens/trap, the yield was 3.0 t/ha (IRRI, 1979).

An estimate of losses caused by N. lugens in Malaysia was M$ 10 million (Lim et al., 1980).

The yield loss due to N. lugens in India was 1.1-32.5% (Jayaraj et al., 1974).

In a greenhouse experiment at IRRI in 1972 on plants of IR22, the damage by N. lugens was studied. At a population level of 100, 50 and 10/plant the number of panicles/plant were 4, 18 and 7, respectively, with corresponding numbers of filled grains/plant of 14, 208 and 203, respectively. Plants with no N. lugens produced 18 panicles/plant and the number of filled grains/plant was 373 (IRRI, 1973).

Lim et al. (1980) reported yield losses in the Tanjong Karang Irrigation Scheme area, West Malaysia, attributable to N. lugens in years with and without outbreaks of this hopper. In an outbreak year such as 1977, as much as 25% (870 kg/ha) of the yield was lost, compared with ca 1% (or 34 kg/ha) in the 1976-77 season. Though the yield losses are a consequence of the planthopper outbreak, the application of large quantities of insecticides for control resulted in an upsurge of Chilo polychrysus that caused heavy damage.

Tu (1983) reported that serious yield losses of rice were caused by outbreaks of N. lugens throughout China.

Dyck and Thomas (1979) summarized that some damage has been reported from Bangladesh, Brunei, China, Fiji, Korea, Malaysia, Papua New Guinea, the Solomon Islands, Sri Lanka, Thailand and Vietnam, but extensive losses from the insect and from grassy stunt disease have occurred in India (estimated at US$ 20 million ), Indonesia ($100 million ) and the Philippines ($26 million). Losses from the insect alone are $100 million in Japan and $50 million in Taiwan. The estimated losses due to N. lugens and grassy stunt virus disease total more than $300 million.

In field investigations in Tamil Nadu, India in 1971-72, 31 varieties of rice were tested for resistance to N. lugens by Jayaraj et al. (1974). Six varieties escaped attack, the remainder suffering medium-severe hopperburn with losses in yield of up to 35% of 1000 grain weight. Varieties that flowered in mid-October (a period of high rainfall and humidity) suffered most damage. In further investigations on assessment of damage caused by N. lugens to rice treated with different insecticides, the dry weight of 1000 grains in the most effective treatment was 28.3 g in healthy clumps, 26.7 g in moderately-damaged clumps and 20.3 g in severely-damaged clumps (Jayaraj et al., 1974).

Studies carried out on economic threshold levels by Ding et al. (1981) in China showed that losses due to N. lugens were heaviest when infestations occurred from the booting to milking stages. Plants infested before maximum tillering have fewer panicles per unit area and fewer grains per panicle than uninfested plants, while those infested after the heading stage have a lower percentage of ripened grain and a lower grain weight. It has been recommended that populations of the pest should be kept as small as possible in the 20 days from late booting to full ear stages.

In Korea Republic, the yield loss of rice infested at the booting and heading stages with N. lugens was studied in the field by Kim et al. (1984). Losses were greater from infestation at the booting than at the heading stage, but infestation of plants in either stage reduced the weight of 1000 grains and the numbers of filled grains compared with uninfested plants. A positive relationship was observed between rice yield and the number of days from heading to the appearance of hopperburn on the plants.

In a field study of 29 Korean cultivars, the resistant cultivars Sangang, Gaya, Hangang Chal and Cheongcheong supported low populations of N. lugens, were undamaged without insecticide and fungicide protection and had a relatively low yield increase of 29, 12, 35 and 18%, respectively, when insecticides were used. Milyang 30 and Wonpung, however, had considerable hopper populations despite their resistance gene and showed some hopperburn later than japonica cultivars, among which there were no resistant cultivars (Kim et al., 1985).

In a field trial in India, the population of N. lugens increased to 621/m2 at 10 x10 cm from 203/m2 at 20 x 20 cm. At a lower dose of nitrogen (75 kg N/ha) the population was 355/m2 and rose to 425/m2 at 150 kg/ha nitrogen. In the same trial, the grain yield at 10 x 10 cm spacing with protection from N. lugens was 5056 kg/ha, while it was 4684 kg/ha in unprotected plots (Pillai et al., 1979).

In laboratory and field studies in China after oviposition by N. lugens at the neck node of the rice panicle, 2-11% of panicle stems were damaged by N. lugens and seed set was reduced by 7.23%, 1000 grain weight by 1.98 g and panicle weight by 16% (Wu et al., 1992).

Also in China, Huang and Cheng (1990) showed that yield losses varied significantly with growth stage attacked and pest density, and were greater earlier in the season and at greater pest densities.

Field trials in China indicated that the loss in rice yield associated with attack by N. lugens was linearly related to the peak density of the 2nd generation. Thresholds at which insecticide treatment is required to prevent the pest from causing 5% yield loss were determined for single and 2-treatment strategies. These thresholds were particularly sensitive to transplanting time and to September temperatures (Cheng et al., 1990).

Leo and Goh (1984) observed the yield losses of susceptible rice varieties (Sumjinbyeo and Milyang 23) due to N. lugens in Korea Republic. Grain yields without insecticide treatment decreased from 5.8 to 4.1 and from 6.6 to 5.4 t/ha for Sumihbyeo and Milyang 23, respectively. N. lugens populations, monitored 55, 70 and 90 DAT increased to 503 and 500 insects/plant for Sumjinbyeo and Milyang 23, respectively. When the yield losses of rice attributable to N. lugens were evaluated, they were assessed at 29.4% for the japonica variety Suminbyeo, 19.1% for the tongil hybrid Milyang 23 and 0 for Cheopcheongbyeo, which possesses the PBH 1 gene for resistance. Planthopper populations were high on both the susceptible varieties, but low on Cheongeheonbueo (8.7/plant).

In field trials in 1990-91 on the yield loss of rice caused by N. lugens, the 1000 seed weight, fertility, empty grain rate, and their correlation with yield loss were investigated in different treatments with various densities of delphacid population by Xiao et al. (1993). The yield loss was positively and 1000-seed weight was negatively correlated with the density of the delphacid population (r = 0.853 and -0.6815, respectively).

Based on the assessment of yield losses, a control threshold of 20-25 planthoppers/hill has been tentatively suggested by Sogawa and Cheng (1979) for tropical countries; however, the critical economic injury level may be as low as 2-5 planthoppers/hill.

In the studies by Chen and Cheng (1978) on the relationship between population levels of N. lugens and yield loss of rice, an account was given of greenhouse and outdoor tests carried out in Taiwan in 1975-76. The data showed that percentage yield reduction (Y) of the variety Tainan No. 5 infested with the planthopper at different levels (X nymphs/hill) under outdoor conditions for a 2-week period could be estimated from the following equations: Y = - 2. 5733 + 35. 2878 log (X +1) at the tillering stage; and Y=-1.8253 +22.55611 log (X+1) at the milking stage. The number of unfilled grains and weight of 1000 grains were influenced most by attacks begun during the booting stage, and the number of bearing tillers and 'bad' panicles by attacks begun during the milking stage.

Palmer et al. (1978a) estimated that the area damaged by N. lugens in Indonesia was 51,800, 34,900, 31,700 and 522,000 ha in 1974, 1975, 1976 and 1977, respectively, and the average yield was 2.6, 2.6, 2.7 and 2.7 t/ha, respectively. Except during 1974, the area damaged was more (almost double) in the wet season than the dry season. The total loss in yield was 135,000, 908,000, 856,000 and 1,410,000 t, respectively. Based on a survey by the authors it was estimated that 30% of this loss was due directly to grassy stunt disease.

Lim et al. (1980) reported that infestation of N. lugens reduced rice grain yields in Sungai Burong, Sungai Leman and Sungai Pasir Panjang areas of Malaysia by 53, 75 and 62%, respectively, in 1977. In the 1976-77 wet season, yield losses were 25, 33 and 12%, respectively. An average of 870 and 340 kg grain/ha over the whole area was lost through planthopper infestation in the 1976 and 1977 seasons, respectively.

Krishnaiah et al. (1986) made successful use of IPM against N. lugens in Krishna Delta, Andhra Pradesh, India in 1983-85. As a result the average yield of rice in the area increased from 3438 to 4667 kg/ha, an increase of 36%.

In an insecticide trial against N. lugens it was found that in untreated fields the yield was 4485 kg/ha while in the most effective treatment it was 6121 kg/ha (Sathiyanandam and Subramanian, 1984).

The yield loss from insects in 1974 at IRRI, varied from 0 to 4.8 t/ha on an insect-susceptible selection of IR 20 rice. The major pest was N. lugens (IRRI, 1975).

In a field trial in Karnataka, India, on the susceptible variety Jaya the population of N. lugens was 232.5 and 2782.6/20 hills with and without insecticide protection, respectively. The corresponding yields were 5263 and 3205 kg/ha, respectively (Gubbaiah et al., 1993).

Mixed Damage and Combined Losses

The impact of feeding damage caused by mixed populations of N. lugens and Sogetella furcifera on the vegetative growth of rice plants was studied in screen houses (Wang et al., 1998b). Plants were inoculated at densities of 20, 40, 60, 80 and 100 insects per plant per pot. Adults were recovered after 10 days of treatment and then dried at 80°C for 48 h. The results showed that the dry weight of adults of both planthopper species and the leaf area of rice plants and their shoot dry weight decreased with an increase in nymph density. The ratio of dry weight to leaf area of rice plants increased with increasing damage. A highly significant linear relationship was observed between the total dry weight of mixed populations (X) and the loss of shoot dry weight of rice plants (Y) at three ratios (1:3, 2:2 and 3:1) of N. lugens and Sogatella furcifera (Y= 67.17 + 30.43 X, Y + 172.48 + 36.51 X and Y = 87.59 + 37.67 X, with b values of 30.43, 36.51 and 37.67, respectively) (Wang et al., 1998b).

A similar trend was observed in another experiment on rice plants grown in plastic pots using nylon nets by Wang et al. (1998a). The equations expressing the total dry weight of plant hoppers and the loss of shoot dry weight of rice plants were Y=578.09 +26.01 x (for N. lugens) and Y=21.90 x -81.18 (for S. furcifera).

During 1991-92, the occurrence of S. furcifera, Cnaphalocrocis medinalis and N. lugens and their combined effects on rice growth were studied by Wu et al. (1995) in fields in Jianghua, China, and in cages in a cement pool. The results indicated that the combined infestations of the pests caused economic damage even though the density of each species was below economic thresholds. Values of maximum growth rate, cumulative dry weight during active growth and the turning point of the growth curve decrease as the combined population density increased.

In an experiment on the influence of population density of N. lugens during the booting stage on rice plants already infected by sheath blight conducted in rice fields in Jiangsu, China, Wu et al. (1994) found that the incidence and severity of sheath blight increased with population density of N. lugens. A significant correlation between insect and disease damage, and rice yield loss was observed. The more honeydew secreted by N. lugens, the greater the incidence and severity of sheath blight.

Yang and Chen (1989) conducted field studies in Guangdong, China during 1986-87 which suggested that a decrease in rice yield was greater when the crop was affected simultaneously by N. lugens and Rhizoctonia solani. However, less injury was caused to rice plants when affected by both pests compared with the total amount of injury caused by N. lugens and R. solani separately. The conclusion was that the heavy loss of energy in injured rice plants was the main reason for a decrease in plant assimilates, 1000-grain weight and the poor quality of rice.

Peng et al. (1993) studied the combined damage of S. furcifera, N. lugens and Rhizoctonia solani as well as their mixed action threshold on rice based on field plot experiments and a caged potted rice test. It was found that there was a significant interaction between delphacids and sheath blight on rice yield. A dynamic action threshold model reflecting the combined effect of the number of delphacids and shoot incidence of rice sheath blight on rice yield loss was developed.

Natarajan and Chandy (1979) showed the impact of mixed damage by S. incertulas, O. oryzae, N. lugens and Cnaphalocrocis medinalis on rice yield in a field trial in India. In this insecticide evaluation trial the treatment with the maximum yield of 4018 kg/ha had a population of 8.39% dead heart, 7.57% silver shoots, 6.39 larvae/10 plants and 3.67 N. lugens/10 plants, respectively. In the treatment with the lowest population of N. lugens the yield was 3277 kg/ha. However, between these two, the yield was not significantly different though the pests' populations were. In unprotected crops, the yield was very low (1786 kg/ha).

Pest Outbreaks

Brown planthopper outbreaks are associated with the development of irrigation systems to allow year-round rice cropping (thus continuous planthopper build-up), excessive fertilizer usage that results in higher planthopper populations, and the use of insecticides that kill natural enemies.

The continuous use of insecticides has destroyed the natural equilibrium between N. lugens and its natural enemies in India. Pests which survived, built-up fast due to either the absence of natural enemies or very low populations which were ineffective in preventing a build up of hoppers (Kulshreshth et al., 1976).

Close planting, production of more tillers per unit area, increased use of nitrogenous fertilizers and indiscriminate plant protection are also reported to have increased N. lugens abundance (Kalode, 1974, 1976; Dyck et al., 1977; Oka, 1977).

The population growth of N. lugens is greater on modern varieties without resistant genes than on local varieties (Mochida, 1978; Dyck et al., 1979). In Indonesia, N. lugens epidemics took place on modern rices susceptible to N. lugens (IR 5, IR8, c4 and PelitaI-2) which were planted from 1967 (Mochida and Suryana, 1979).

In India, N. lugens and the green leafhopper became a serious problem after the introduction of Taichung Native 1 in 1964 and IR8 in 1968 (Kulshreshth et al., 1970a).

In the Korean Republic, outbreaks of N. lugens regularly occurred on rice on the south-eastern coast of Jeonnam Province, in the coastal region of Chungnam Province and to a lesser extent in the coastal areas of the province between them, Jeonbuk Province (Kim et al., 1986).

Feng et al. (1992) described an outbreak of N. lugens in rice growing areas of E. Hebai (Tianjin and Luannan County) of China in which yield losses reached 30% in 19.8% of the area and 100% in an area of 4000 mu (1 mu = 0. 067 ha). Simultaneous overcast skies and rain with a strong southwest wind caused the outbreak. The source was adults emerging in mid-July to mid-August from an outbreak in coastal east China and the Sichuan basin. Single cropping rice (with a growth period longer than that of double cropping rice), large areas planted with the susceptible keng rice and high temperatures and humidity in August and September allowed the pest a longer period to reproduce.

Susceptible varieties will not necessarily be damaged by N. lugens even when N. lugens outbreaks occur in adjacent plots. This had been concluded when rice fields treated and untreated with insecticides were monitored by Cuong et al. (1997) for outbreak of pests in the Mekong Delta, Vietnam. Rice varieties susceptible, moderately resistant and highly resistant to N. lugens were treated with deltamethrin. Monitoring for pests indicated N. lugens outbreaks in susceptible varieties treated with insecticides. In insecticide-free plots, the yield of the susceptible variety was lower than those of resistant varieties in only one of four seasons. In plots of a moderately resistant and a highly resistant variety, populations of N. lugens were generally similar, and yields did not differ, under both insecticide-treated and insecticide-free conditions.

In contrast, N. lugens outbreaks and hopperburned plants were observed only in plots of insecticide treated susceptible variety plots when it assumed an epidemic form in India in 1973 and caused extensive damage in Kerala. Since then it is reported to have caused destruction of crops, especially of high yielding varieties, in Andhra Pradesh, Kerala, Tamil Nadu, Orissa, West Bengal, Punjab and Haryana (Kulshreshth et al., 1976). In all the areas, where N. lugens outbreaks occurred monocultures of the susceptible varieties Jaya and IR 8 were grown on an extensive scale. These varieties were most heavily damaged. In all the affected areas intensive cultivation was accompanied by a higher use of nitrogen (80 to 100 kg/ha ) and close spacing (15 x 15 cm) giving rise to a dense crop.

Economic Threshold Levels

A sequential sampling plan for N. lugens was developed by Shepard et al. (1988) in the Philippines and decisions to apply insecticides to plots with threshold levels of plant hoppers resulted in significant yield increases. There was 100% agreement between decisions made by this plan and those made by intensive sampling.

Chen et al. (1979) reviewed and analysed the effects of direct injury by the brown planthopper to rice plants and subsequent yield loss, from the point of view of rice physiology and population.

In an attempt made to define a tolerable level of N. lugens, it was observed that a maximum density of at least 50 hoppers/hill was reached twice in a experimental field without any apparent loss in grain yield (IRRI, 1975). In a greenhouse field-simulated experiment, where the maximum density reached only about 25 insects/hill in either the second or third generation, usually there was no significant yield loss (IRRI, 1975). Differences in the yield between insecticide-treated and untreated plots of IR 20 was monitored during 1973-74 at IRRI to measure the changing pest damage and thus pest density. The differences varied greatly over the experimental period. Damage was greater in the dry season, when N. lugens were abundant, and less in the wet season (IRRI, 1975).

The economic threshold level (ETL) was as low as 3-5 nymphs/hill prior to the milking stage and 5-8/hill thereafter based on greenhouse and field tests in Taiwan (Chen and Cheng, 1978).

Ding et al. (1981) in China considered that the number of insects in the damaging generation should be controlled to less than 8-10/hill and that in the preceding generation to less than 0.3-0.5/hill and that chemical control measures should be taken at peak nymphal hatching for the damaging generation and just before adult emergence for the preceding generation.

Huang and Huang (1986) determined an economic threshold for 3rd to 4th instar nymphs of N. lugens on rice in Fujian, China, under caged conditions each containing 4 hills of a second cropping of rice, at 7 rates from 0 to 180 insects/cage at the end of booting and heading. The permissible damage level (when the yield loss is twice the control cost) was calculated from the formula: L = (C x F)/ (Y x P x E) where L, C, F, Y, P and E are permissible damage level insect control cost, a correlation factor (2), rice yield, rice price and control efficiency, respectively. The permissible damage level was estimated to be 2. 3% under the trial conditions and the economic threshold was set at 40% lower than this. The relationship between the insect population level and rice yield was described with a regression equation and the economic threshold was calculated to be 300,000 insects per mu (1 mu = 0. 067 ha).

Zhang and Zhang (1989) studied the EIL (Economic Injury Level) and ETL of N. lugens on mid- season japonica rice by artificial inoculation and natural infestation in field trials in Jiangsu, China in 1986. The results showed that rice yield loss, which had a significant exponential functional relation to insect number/hill, increased with the insect population. The EIL was calculated as 3.5%, and the ETL was similar to the EIL in the year when the insect occurred early and was lower than the EIL in the year when the insect occurred late.

According to Xiao et al. (1993) the economic threshold for control of the delphacid was 6.5-8.5/hill, during field trials in 1990-91 in the hybrid rice region in Northern Jiangsu, China.

In China, Qian et al. (1996) established ETL levels for rice pests including N. lugens based on their ecology, yield losses, product prices and control costs.

Influence of Cultivation Practices

More than two rice crops per year, lack of a rice-free period during the year with early-maturing varieties and neighbouring fields with early-maturing varieties, staggered planting and the presence of volunteer ratoon after harvest and injudicious use of fertilizer are factors favouring N. lugens build up and subsequent damage. High planthopper populations also occur on high tillering varieties because of the increased plant surface on which to feed (Reissig et al., 1985).

High fertilizer rates are favourable to the development of populations of N. lugens (Kalode, 1971; Dyck et al., 1979).

N. lugens population growth is greater on rice plants grown in pots or fields with standing water (Dyck et al., 1979; Mochida and Heinrichs, 1981). The flooding of rice fields has often been cited as a factor for N. lugens outbreaks (Kulshresth et al., 1974; Stapley, 1975).

Growing more than one rice crop a year influences the degree of pest problems. N. lugens, tungro and grassy stunt epidemics have occurred on continuously cropped fields rather than on those with definite crop seasons (Mochida and Heinrichs, 1981).

The closer spacing of rice plants results in a change in microclimate of the crop, which favours multiplication of insect pests. There are indications that the number of N. lugens nymphs/hill and the number of tillers/hill are often positively correlated (Mochida and Heinrichs, 1981). There are also positive correlations between the number of nymphs and tillers/m2, and between the number of nymphs and tiller density/m2 (Dyck et al., 1979).

High plant population as a result of broadcast seedling rice at high rates has been cited as one of the reasons for N. lugens outbreaks in Kerala, India (Kalode, 1974).

Populations of hoppers in untreated fields were higher in direct seeded rice (DSR) early in the season, probably because of the early establishment of the DSR crop. Hopper populations peaked in March and August in both transplanted rice and DSR (IRRI, 1985).

With a change from transplanting rice to direct-sowing rice, the pests and diseases have changed in rice fields in the Shanghai suburbs of China. In the direct-sown paddy fields the occurrence of N. lugens remained unchanged (Chen et al., 1999).

Natural enemies, particularly rice pest predators in the subsequent rice season, were affected by crop residue-burning in rice fields in Omon, Vietnam. Rice pest populations, particularly N. lugens, was significantly higher in the burnt field from 65 days after sowing onwards compared with unburnt fields. Burnt fields showed statistically lower rice yields than unburnt fields (Loc et al., 1996).

Other Indirect Losses

N. lugens acts as vector for Rice ragged stunt virus (RRSV) and Grassy shoot virus (GSV).

RRSV occurs in Indonesia, the Philippines, India, Sri Lanka and Thailand (Hibino, 1979).

In Japan, N. lugens was able to transmit RRSV and different biotypes of the vector did not vary in their transmitting ability. When macerated diseased plants were injected into the bodies of N. lugens the disease was transmitted by the insects to healthy plants (IRRI, 1978).

In a laboratory test, 14 to 76% of plant hoppers were active transmitters of rice ragged stunt disease (Ling et al., 1977).

When RRSV was maintained for 7 years in vegetatively propagated rice plants without planthopper transmission, the ability of the virus to be transmitted to healthy seedlings by N. lugens was lost after 3 years. The non-vector-transmissible isolate caused less stunting than the original RRSV and galls and ragged twisted leaves appeared less frequently (Maoka et al., 1993).

In 1981, an epidemic of RRSV occurred in deep water rice (DWR) on the Central Plain of Thailand. Virtually all fields were infected and heavy yield losses occurred, especially in the south west region. The disease incidence was lower in 1982 and has been very rare since 1984. There was evidence that infection originated from N. lugens vectors migrating into preflooded DWR from irrigated rice in the southern parts of the central plain (Pattrasudhi and Catling, 1988).

In a pot experiment, the susceptible deep-water rice cultivars KTH 123 (early maturing) and PG 56 (late) were inoculated with viruliferous N. lugens at five plant growth stages. Results indicated that most virus transmission occurred early in the season, before deep flooding (Disthaporn et al., 1983).

Grassy stunt and RRSV transmitted by N. lugens were among the most damaging and widespread rice diseases in Indonesia. Yield losses due to grassy stunt and N. lugens in 1974-77 were in excess of US$ 510 million. Rice ragged stunt virus was a new disease which was widespread in Indonesia and had been reported from several rice growing countries in Asia since 1977; some farmers have suffered losses of 100% in infected fields (Palmer and Rao, 1981).

In a trial by Palmer and Rao (1981) on a crop of variety Pelita 1-1 with or without N. lugens, the incidence of RRSV was 58.3 and 5.3% in 1978 and 45.6 and 3.6% in 1979. On the variety PB 26 with or without N. lugens, the incidence of RRSV was 27.7 and 4.0% in 1978 and 27.3 and 1.6% in 1979 respectively.

In pot studies under simulated field conditions, RSV was transmitted by N. lugens from seedling to early stem elongation stages and reduced yields of two deep water varieties by 50-100% (IRRI, 1984). Yields were not affected by inoculation during the main flooding period (late stem elongation and flowering stages ). Infection caused severe stunting (which led to plant mortality from rising water), distinct foliar symptoms, delayed flowering in one variety, and empty or partially filled panicles. Symptoms resembled those on lowland rice and were most severe during early infection. Symptoms were milder in the late-maturing Pin Gaew 6 than in the early Khas Tah Haeng 17. It was suggested that field infection may be reduced by controlling the vector population before flooding (IRRI, 1984).

Grassy stunt (GS) is caused by a virus or phytoplasma (Palmer et al., 1978a) and is transmitted only by N. lugens (Rivera et al., 1966; Ling, 1972, 1977). The disease was first reported in the Philippines in 1963 and confirmed in 1964 (Rivera et al., 1966 ). GS was observed in Indonesia in 1967, and successfully transmitted in Indonesia in 1971 (Tantera, 1973a, b; Tantera et al., 1973). The disease was identified in India in 1967 (Raychaudhuri et al., 1969), and reached an epidemic level in 1972 (Gopalakrishnan et al., 1973). It also occurs in Thailand, Sri Lanka (Ling et al., 1970, Toriyama, 1975) and Bangladesh (Khush and Beachell, 1972).

In Kerala, India, following an outbreak of N. lugens in 1973-74, heavy losses in yields, as a result of grassy stunt, which was reported for the first time in India, occurred (Anjaneyulu, 1974).

The continuous cultivation of rice over northwest India has virtually eliminated the spatial and temporal hurdles that were so far operating against the movement of vectors (Nagarajan, 1989).

Surveys conducted by Palmer et al. (1978a), Soepriaman and Palmer (1978) and Mochida et al. (1979) annually during 1974-77, determined the distribution of grassy stunt disease in the major rice-growing areas of Java and Bali. Serious losses due to GS, N. lugens, or both, occurred at Serang, Pandeglang, Rangkasbitung and Cirebon regencies in West Java province, Tegal, Pemalong, Pekalongan, Kendal, Semarang, Surakarta, Boyolali and Klaten regencies of central Java Province, Jember and Banyuwangi regencies of East Java province and Denpasar and Tabanan regencies of Bali province. Combined yield losses caused by the plant hopper and the disease in 1974-77 were in excess of 3 million metric tons of grain (Palmer et al., 1978a).

Rice crop loss by N. lugens in Indonesia was about 200 ha in 1951. Losses from N. lugens and grassy stunt, or both, were 7000 ha in 1968-69; 10,000 ha in 1969-70, 8000 ha in 1970-71, 9700 ha in 1971-72 and 56,200 in 1972-73 (Soenardi, 1971; Tantera, 1973; Soehardjan and Tantera, 1975; Soepriaman et al., 1976).

N. lugens and grassy stunt epidemics have occurred on continuously cropped fields rather than on those with definite crop seasons (Mochida and Heinrichs, 1981).