Insect Pests of Rainfed Wetland Rice

International Journal of Pest Management Vol. 55, No. 3, July–September 2009, 221–242

Insect pests of rainfed wetland rice in the Philippines: population densities, yield loss, and insecticide management
J.A. Litsingera*, B.L. Canapib, J.P. Bandongb, M.D. Lumabanb, F.D. Raymundoc and A.T. Barrionc
a c

1365 Jacobs Place, Dixon, CA 95620, USA; bInternational Rice Research Institute, DAPO Box 7777, Metro Manila, Philippines; Philippine Rice Research Institute (PhilRice), Maligaya, Science City of Mun˜oz, Nueva Ecija 3119, Philippines

(Received 21 June 2008; final version received 13 January 2009) Rainfed wetland rice (RWR) had more species in common with irrigated wetland than dryland rice agroecosystems. Across ecosystems, higher pest densities and losses were recorded in RWR sites. We hypothesise that under low pressure from natural enemies, vegetative stage losses became particularly high due the combination of whorl maggot and caseworm damage combined with the physiological stress of transplanting shock. Both of these pest groups benefited from an expanded vegetative period common in RWR agroecosystems. Losses in older rice were probably due to stemborers. RWR is more prone to an array of physiological stresses than irrigated rice that we believe minimises crop compensation to accentuate insect losses. Chemical control is uneconomical mainly due to the low yield potential of RWR and the poor efficacy of applied insecticide. Keywords: rice insect pests; yield loss; natural enemies; action thresholds; chemical control; crop compensation; crop stresses; integrated pest management

1. Introduction Domesticated rice (Oryza sativa L.) probably evolved first as a perennial, terrestrial species which then became adapted to an aquatic existence in rainfed wetland rice (RWR) ecosystems (Catling 1992). RWR is the traditional system of wetland rice culture where farmers utilise accumulated rainfall to puddle the soil to enable hand transplanting. RWR is normally a single rice crop grown in the rainy season, thereafter fields are either left fallow or sown to a drought-tolerant, short-maturing, non-rice crop over the dry season. Due to irregular rainfall, rice bunds tend to be tall as farmers attempt to store large amounts of water after transplanting. Three main rice agroecosystems developed from the RWR cultural archetype to reach most habitats within a watershed (Garrity et al. 1986): (1) deepwater rice which occupies large river floodplains where ponding exceeds 1.5 m for more than a month, (2) dryland rice, located in the uppermost landforms where rice is dry-seeded without ponding in non-puddled soils, and (3) irrigated rice which is the most recent agroecosystem replacing RWR in the more favourable landforms near sources of water. Irrigated rice greatly expanded after the advent of the Green Revolution modern rices in the late 1960s (IRRI 1985a). RWR covers about 38 million ha, representing 28% of rice area worldwide (Garrity et al. 1986). Approximately two-thirds of the total RWR environment is unfavourable insofar as frequent drought, submergence, and adverse soils contribute to low yields.

Farmers often hedge their bets by having fields in both low- and high-lying locations in a toposequence (land fragmentation) to increase the probability of a harvest in any rainfall contingency (Fujisaka 1990). The RWR plant archetype is attuned to the Asian monsoon climate. Cultivars are tall (1–1.5 m) to tolerate flooding but have less tillering ability (DeDatta 1981). They are photoperiod-sensitive and longmaturing thus have more time to compensate from biotic and abiotic stresses, a trait that ensures stable yields. While transplanting may last for over 3 months to cover all field locations, flowering and crop maturity become synchronised by the short day lengths that occur at the end of the monsoon season. Thus, the crop matures in dry weather to provide optimal grain quality. Unfortunately production by the low yielding traditional RWR cultivars is insufficient to feed the rice-eating world, a dilemma which motivated the development of modern, high-yielding, semi-dwarfs. There are few entomological research papers that relate specifically to the RWR ecosystem. Mainly lists of pests, prioritised by importance in most Asian countries, have been compiled (Litsinger 1979; Mackill 1986; Heinrichs et al. 1986; Kalode et al. 1986). Farmers in Thailand, Cambodia, Laos, and Nepal identified key insect pests, along with control practices that generally included cultural methods and botanical pesticides (Fujisaka 1990). The Philippines has 1.2 million ha of RWR which represents 40% of its rice area (Garrity et al. 1986).

*Corresponding author. Email: [email protected]
ISSN 0967-0874 print/ISSN 1366-5863 online Ó 2009 Taylor & Francis DOI: 10.1080/09670870902745070 http://www.informaworld.com

222

J.A. Litsinger et al. the latter has a history of pest epidemics leading to severe losses. Parasitism levels over a range of RWR pests and stages was also measured. A more thorough investigation in RWR ecosystems which historically have been less affected by pest epidemics may lead to insights that may aid in the management of more pestprone, rice ecosystems. 2. Methods and materials 2.1. Study sites On-farm research was carried out for 3–5 years in each of three sites where comparisons were made in similarly operated dryland and irrigated sites (four sites each) spanning a 15-year period 1976–1991. 2.1.1. Iloilo province Iloilo averaged 1.5 m/year rainfall during the study period 1976–1979 (Morris et al. 1986). The sloping terrain predisposed the watershed to drought. The principal land forms in Oton and Tigbauan towns were sideslopes, plateaus, and low plains. RWR occurred in a parallel strip along the coast sandwiched between uplands and rice lands irrigated from small rivers. There was no second rice crop before the project, but indeterminate cowpea was commonly relay-sown before the single rice crop harvest. 2.1.2. Pangasinan province

Generally, RWR farmers use fewer agroinputs than irrigated rice farmers, particularly in unfavourable habitats (Fujisaka 1990). Surveys of RWR farmers in Iloilo and Pangasinan provinces in the late 1970s, representing favourable and unfavourable RWR sites, revealed that 54 and 29% used insecticides averaging 2.5 and 1.3 applications per crop among users (Litsinger et al. 1980b). In Cagayan province, 47% of farmers used insecticides, averaging 1.9 applications among users (range 1–4) (Litsinger et al. 1982). Another survey in Central Luzon, also in a favourable site by Pineda et al. (1984), noted that RWR farmers adopted insecticides at a slower rate than nearby irrigated farmers, but lagging some 5 years behind. Indeed the adoption rate slope mirrored that of irrigated farmers where 70% of both groups used insecticides averaging 3–4 applications each among users. Farmers will spray upon seeing some moths flying when flushed from the fields or by observing damage only on a few hills (Bandong et al. 2002). The surveys also revealed that farmers severely underdosed at about half of the recommended rates. The current study was carried out by a joint project between the International Rice Research Institute (IRRI) and the Philippine Department of Agriculture that formed multi-disciplinary teams at three RWR sites (Morris et al. 1986). The task was to increase rice production by boosting cropping intensity via replacing traditional rices with modern photoperiod-insensitive ones that mature within 4 months. The teams tested different cropping patterns and management practices with a view to optimise input usage. RWR environments are classified as favourable or unfavourable (Garrity et al. 1986). The latter are under chronic drought stress and subjected to submergence at depths 450 cm for several days which is represented by Cagayan province. Sites in Iloilo and Pangasinan provinces were favourable, as submergence and drought were less frequent and enduring. In these sites, double-cropped rice was feasible using early maturing rices and direct-seeding methods, while in Cagayan we tested growing mungbean before rice as well as early planted, direct seeded rice designed to escape anticipated flooding later in the season. We documented the arthropod fauna and quantified yield loss using the insecticide check method. As insecticide had already been adopted by farmers, our objective was to rationalise usage by testing action thresholds as decision tools (Litsinger et al. 2005). Testing of the different cropping patterns allowed ecological information to be gathered on the interaction of cropping patterns and site on insect pest abundance. Rainfall patterns also differed each year, introducing yet another dimension of insect pest ecology. Results are compared to Philippine dryland and irrigated rice ecosystems managed by similar applied research teams. The former has mainly soil and seed pests, uncommon in wetland culture, while

Manaoag, Pangasinan in NW Luzon, is a large, lowlying, rice bowl with a high water table and was subject to submergence after heavy rains. Rainfall averaged 1.7 m/year during the study 1976–1980 (Morris et al. 1986). Mungbean was broadcast-sown after rice harvest in tilled fields. Farmers alternated between deep ponding and draining their fields to cope with a dual micro-nutrient problem of zinc deficiency and iron toxicity. Less field water led to weediness and exacerbated drought stress. The project introduced zinc application so that farmers no longer had to drain their fields. More detailed site descriptions for the Iloilo and Pangasinan sites can be found in IRRI (1976, 1977, 1980a). 2.1.3. Cagayan province The third site was Solana located in Cagayan province in NE Luzon where annual rainfall from 1980 to 1982 averaged 1.8 m. Rice lands lie along sloping terraces formed by the Cagayan river but too high to receive irrigation. Some farmers store canoes under their homes because, due to deforestation in the upper watershed, flooding is common during the frequent typhoons. On the well-drained upper wetland slopes, some farmers plant mungbean at the beginning of the rainy season; otherwise, no other non-rice crops are

International Journal of Pest Management grown in the rice fields. Maize is commonly grown in the adjacent uplands. The nearest irrigated rice was 20 km away in Gadu. Further site description can be found in IRRI (1981a). 2.2. Research teams

223

Field offices were established in each site. Local assistants acquainted with the farmers and fluent in the regional languages were recruited. This gave the local teams acceptance in the target farm communities. The annual on-farm research programme was conducted by multi-disciplinary teams with researchers specialising in varietal testing, agronomic practices, weed control, plant pathology, and economics as well as entomology. Laboratory assistants were trained in rice arthropod identification and each was equipped with a dissecting microscope. 2.3. Rice crop management

All trials were superimposed on farmers’ fields managed under the teams’ agronomic recommendations carried out by the farmers with the exception of insect pest control practices. Because the project involved introducing modern rices into these areas that were responsive to better management, agroinput levels were being tested at levels above those used by farmers on traditional varieties. Farmers immediately began adopting the improved varieties in the favourable sites along with increased use of inorganic fertiliser usage. Almost no farmer adopted the research teams’ action thresholds, however, as this would require specialised extension programmes which did not materialise. Trials with traditional varieties were managed under the farmers’ low input systems. In Iloilo and Pangasinan, rice variety IR28 was first tested and then replaced with IR36; both are early-maturing, modern semi-dwarfs (105–110 days). In Cagayan, IR36 was first tested and then replaced by IR52, a 116-day variety with some drought tolerance. All three rice varieties were resistant to rice green leafhopper Nephotettix virescens (Distant), while IR36 and IR52 were also resistant to rice brown planthopper Nilapar˚ vata lugens (Stal). The major traditional variety in Pangasinan and Cagayan was Wagwag. In Pangasinan it was Inano while BE3 and Kapopy dominated in Iloilo. Except the wet-seeded first-crop fields in the double rice crop pattern, all crops were transplanted. The seeding rate for transplanted rice was 30 kg/ha and 100 kg/ha for wet-seeded rice which was soaked for 2 days before being broadcasted on puddled soil. Only inorganic fertiliser was used with NPK rates (kg/ha) of 70-40-0 in Iloilo, 80-0-0 in Pangasinan, and 30-0-0 in Cagayan. In Pangasinan, zinc oxide was applied at 25 kg/ha to wet-seeded rice and by seedling root dip if transplanted. Weeds were controlled by a combination of ponding, herbicides, and hand

weeding. Grain yield involved five 5-m2 samples taken in a stratified grid within each 100–200-m2 plot. Grain was dried to 14% moisture content. The site teams were encouraged to interact with farmers to improve on current insect control practices. This farming systems approach involved an annual updating of recommended farming practices based on the results of each year’s on-farm trials (Zandstra et al. 1981). For insect control this meant improving the action thresholds and choice of insecticide. Improved practices included identification of better varieties, more efficient inorganic fertilisers, and improved weed and insect control. All things being equal, the improved practices would have steadily increased yields with the consequence of increasingly augmenting the crops’ ability to compensate from insect pest damage with a resulting increased lowering of measured yield losses. Sites in the other two rice agro-ecosystems underwent this applied research process. The results did not show this trend, however, due to the more powerful random weather events on yield common in the sites. This will be elaborated more in the discussion section. 2.4. Treatment descriptions

2.4.1. Partitioned-growth-stage yield loss trials The yield loss trials involved five treatments employing the insecticide check method (Litsinger 1991). Three crop stages were recognised: (1) vegetative (transplanting to panicle initiation), (2) reproductive (panicle initiation to flowering), and (3) ripening (flowering to maturity 10 days before harvest) (Yoshida 1981). In a typical 110-day modern variety, the reproductive stage begins about 40 days after transplanting (DT) and ends about 30 days later. In traditional rices only the vegetative stage is prolonged. The seedbed had been included in experiments prior to the current study, but as no significant yield loss was ever measured (Litsinger 2009), this portion of the vegetative growth stage was eliminated from experimentation. The first treatment termed ‘full protection’ attempts to achieve the crop’s yield potential with the least possible amount of insect damage. Weekly insecticide sprays at the manufacturers’ recommended dosages were applied to each plot with inter-plot spray drift minimised by a mosquito cloth-covered 1 6 3 m wood frame held downwind by two assistants. Insecticides (monocrotophos, chlorpyrifos, or g-BHC sprayed every 10 days at 0.75 kg a.i./ha each in the respective three rice growth stages) were selected for efficacy as well as proven phytotoxic or phytotonic neutrality (Venugopal and Litsinger 1984). The insecticides were applied at 300–500 l/ha spray volume (increased with crop growth) with 19-l, lever-operated, knapsack sprayers fitted with hollow cone nozzles. From the second to fourth treatments, insecticide was withheld from each successive growth stage; the fifth treatment was the untreated check.

224

J.A. Litsinger et al. yield losses were greatest in the vegetative stage and that stemborer whiteheads were often high. More timely application occurred via prophylactic applications than from monitoring which may provide better control. Soil incorporation of low dosage carbofuran granules had been found to be highly effective (dela Cruz et al. 1981). Crops were monitored weekly, and when a threshold was reached the currently recommended insecticide was applied within a day. Insecticide technology was also developed in an iterative process; thus if efficacy was low, adjustments were made, normally involving a change in the character, chemical, or dosage. Tests to determine the minimum effective dosage were undertaken in separate on-farm trials with a view to cost saving (IRRI 1985b). Foliar sprays were applied as described in the yield loss trials. The degree of control was based on comparisons between the full protection and the untreated check in the insecticide check treatments. Cost and return economic analyses were carried out for each treatment that led to a significant yield increase. The protocol of Smith et al. (1988) was followed based on 1986 prices for insecticides, unmilled rice, and labour. Interest on materials was levied at 60% per season, spraying labour was set at 8 h, and granular application at 4 h/ha. The 60% interest charged was a realistic rate for rural Philippines and labour costs/ha were assessed 33% interest. Pest monitoring was assessed at 60 h/ha per season for the action threshold treatment. 2.5. Arthropod sampling methods Pests were sampled weekly in the action threshold, prophylactic, and untreated check plots, but only once per growth stage in the other yield loss treatments. The total number of tillers and leaves per hill was recorded along with the percent damaged by whorl maggot

All treatments were arranged in randomised complete block design (Litsinger et al. 1980a). Within each co-operator’s field (replication), a 0.2-ha research area was demarcated with plastic string tied to bamboo stakes. Plot sizes were 100–200 m2. Side by side placement of insecticide treated and untreated plots had been found to have unbiased effects on pest abundance (Litsinger et al. 1987b); total yield loss, calculated both in terms of grain weight and percentage, was the difference between the ‘full protection’ and untreated check. To calculate percentage, total loss was divided by the full protection yield. The losses in the three separate growth-stage omission treatments were calculated as the respective differences from the full protection plot. These were summed and adjusted upwards or downwards proportionally so that the total equalled the total yield loss in each crop. 2.4.2. Chemical control treatments

Two types of insecticide control practice were randomised among the yield loss plots in the superimposed trials, so as to generate recommendations for farmers. Their inclusion each season followed an iterative process of technology development for each site. These treatments included: (1) the use of action thresholds as a guide to insecticide application and (2) a prophylactic insecticide regime. An action threshold is composed of the pest character to be measured as well as the triggering level of abundance and an insecticide response at a specified dosage. The action threshold practices were reviewed each year and changed if the trials pointed to a better practice (Table 1). The prophylactic treatment involved soil incorporation of 0.5 kg a.i. carbofuran granules/ha before transplanting followed by two foliar sprays of 0.4 kg a.i. chlorpyrifos/ha 10 days apart during the late reproductive stage to prevent stemborer damage. This practice was based on the results of earlier studies that showed

Table 1.

The evolution of action thresholds for rainfed wetland rice insect pests in the Philippines, 1976–83.1 Action threshold2

Growth stage Vegetative

Pest Whorl maggot Caseworm Stemborer

Year 1976 1982 1976 1982 1976 1982 1976 1982 1976 1982 1976 1982

Character 25% damaged leaves 15% damaged leaves 15% defoliation 10% cut leaves 10% deadhearts 15% deadhearts 3% deadhearts 5% deadhearts 10% damaged leaves 15% damaged flag leaves 4 bugs/m2 8 bugs/m2

Insecticide (dosage kg ai/ha) monocrotophos EC (0.75) triazophos EC (0.40) monocrotophos EC (0.75) malathion EC (0.40) monocrotophos EC (0.75) chlorpyrifos EC (0.40) monocrotophos EC (0.75) chlorpyrifos EC (0.40) monocrotophos EC (0.75) BPMC WP (0.40) monocrotophos EC (0.75) diazinon EC (0.40)

Cost ($/ha) 23 12 23 3 23 7 23 7 23 6 23 7

Reproductive

Stemborer Leaffolder

Ripening
1 2

Rice seed bug

Action thresholds were re-evaluated each year in an iterative process based on field results. Only the beginning and final thresholds are shown. EC ¼ emulsifiable concentrate (liquid), WP ¼ wettable powder.

International Journal of Pest Management Hydrellia philippina Ferino and various defoliators at 21 DT or 35 days after sowing (wet seeded crops). Damage caused by each of the four common defoliating insect ´ pest species [caseworm Nymphula depunctalis (Guenee), armyworm Mythimna separata (Walker)/cutworms Spodoptera spp., green semilooper Naranga aenescens Moore, and green hairy caterpillar Rivula atimeta (Swinhoe)] could be readily distinguished. Stemborer larvae feed internally in tillers, severing vascular tissue and so causing wilting which is distinguished on younger tillers as ‘deadhearts’ and in panicles as ‘whiteheads’. Stemborer-damaged tillers/panicles were pulled to confirm that larval feeding was responsible. Deadhearts and whiteheads were sampled during tiller elongation and ripening stages. Leaffolder larvae fold the leaves in a distinctive fashion. Damage was recorded in the flag leaf stage, the most important period in terms of yield loss (Heong 1990). Rice seed bug adults and nymphs were sampled by visual counting within three 5-m2 areas at milk and soft dough stages. Mechanical tally counters were used to record the number of leaves, tillers (including deadhearts and whiteheads), and insects as appropriate. One staff member served as a recorder during sampling. Rice plant- and leafhoppers and their key predators were sampled by sweep net and visual counting on a per-area basis in Pangasinan and Iloilo in insecticide untreated fields. A standard 38-cm diameter sweep net was used for the more mobile species with the operator making pendulum swings while walking through the field. In each field, 20 sweeps were taken in the vegetative, reproductive, and ripening growth stages. The contents were placed in large plastic bags and chloroform-soaked cotton balls were added to kill the insects which were then identified and counted. For the more sedentary species, 1 m2 was marked off and all tillers within the sampling area were inspected for homopterans, herbivores, and predators which were counted by a team of three people, two doing the counting and one doing the recording. This was done once in the three major growth stages in five samplings of 1 m2 per sampling date per field. Insect parasitism rates were also measured by weekly visits to farmers’ fields untreated by insecticide. The various stages of common insect pests were collected live and brought to the field office and held in vials, fabricated plastic cages, or Petri dishes for parasitoid emergence (Barrion and Litsinger 1985). Plant-/leafhopper eggs were detected from tiller dissection and held on moist seed germination cards in Petri dishes with an added fungistatic agent. Larvae held in rearing containers were checked regularly and refreshed with new leaves. Stemborer larvae, detected from stem dissection and damaged tillers, were reared on tillers in long glass vials. Adult plant-/leafhoppers were inspected for Strepsiptera under a stereomicroscope or held for dryinid or pipunculid parasitoid emergence in special plastic cages (Chandra 1978).

225

2.6. Light trap collections Kerosene light traps (Litsinger and Entomology Department 1979) were operated daily to census rice insect pests in one village per each of the three sites. A farmer was trained to manage each light trap by collecting and placing nightly catches in vials of 70% alcohol while trained staff identified and counted. Light traps were placed in pairs per village in open rice fields at least 100 m apart. Enough kerosene was used so that the lamps burned all night. The height of the flame was standardised by wick length. Collections in the three RWR sites were compared to four established in dryland rice sites (Siniloan, Quezon; Claveria, Misamis Oriental; Tupi, South Cotabato; Tanauan, Batangas) (Litsinger et al. 2009) and three irrigated sites (Koronadal, S. Cotabato; Victoria/Santa Maria, Laguna; Zaragoza/Cabanatuan/Jaen, Nueva Ecija) (Litsinger et al. 2005) where similar research teams worked and paired light traps were managed by farmers. Two irrigated sites were subdivided into synchronously and asynchronously planted doublecropped rice areas in Koronadal and Jaen (Loevinsohn et al. 1988). Counts of all light trap collections were summed over a 6-month period to represent the duration of each seasonal rice crop. This was done so that RWR single crop sites could be compared to irrigated double crop sites on an equal per-crop basis. In irrigated sites, data for each species were summarised for each wet and dry season crop based on the planting pattern each year. 2.7. Ecosystem comparisons RWR arthropod fauna and yield losses in the three RWR sites were compared to the irrigated and dryland rice sites which also employed the same arthropod sampling and partitioned-growth-stage insecticide check methods. On-farm insect pest sampling and yield loss trials were conducted in the same three irrigated, double-crop and four dryland rice areas mentioned above. Farm record keeping was conducted in each of the sites during the crops where light traps were operating. Accordingly 20–40 farmers were interviewed in each site by the local staff to determine applied kg N/ha and number of insecticide applications per crop. Mathematical models of best fit correlating each of the two input levels to seasonal light trap totals for selected insect pests over a range of sites in each of the three agro-ecosystems were analysed by regression analysis. 2.8. Statistical analysis

All statistical analyses were performed by SAS with P  0.05 as the criterion for significance. Results were subjected to one-way ANOVA and regression analyses where appropriate. Treatment means were separated

226

J.A. Litsinger et al. numbers particularly after a drought (Mochida et al. 1987). The panicle-cutting caterpillar Mythimna separata (Walker) can cause significant loss as it often aggregates, while Spodoptera litura F. cutworm was observed in Cagayan after building up on weeds then shifting to rice after the farmer hand weeded. Defoliation levels from all of these species, save caseworm, were insignificant in the trial fields. Despite its common name M. separata rarely was noticed affecting panicles. The highest stemborer damage occurred in traditional varieties in Iloilo and Pangasinan averaging 11– 16% deadhearts and 9–11% whiteheads (Table 2). The next highest level occurred in the single crop in Pangasinan on modern rices with 11% deadhearts, but 52% whiteheads. Least damage occurred in the single transplanted crop in Cagayan on modern rices (51%) followed by the first wet seeded crop in Pangasinan (54%). Tiller dissections showed that Scirpophaga stemborers predominated in all three sites with Pangasinan registering only yellow stemborer S. incertulas (Walker) (Table 3). The principal species in Iloilo was white stemborer S. innotata (Walker) which along with yellow borer totalled five species including striped stemborer Chilo suppressalis (Walker), pink stemborer Sesamia inferens (Walker), and gold fringed stemborer Chilo auricilius Dudgeon, while Cagayan had four. Three leaffolder species were encountered in the RWR sites with Marasmia exigua (Butler) being more ´ prevalent than Cnaphalocrocis medinalis (Guenee), as the former has a wider host plant range (Barrion et al. 1991). M. patnalis Bradley was also noted. The greatest abundance was recorded on the traditional varieties of Iloilo (10% damaged flag leaves) and the first rice crops in both Pangasinan (9%) and Iloilo (7%) (Table 2). The lowest defoliation was recorded in the second crops of Iloilo and Pangasinan (1–3%). Rice seed bug was dominated by Leptocorisa oratorius (F.), with L. acuta (Thunberg) being rarely found. Highest incidence (bugs per m2) was in the second crop of Iloilo (7 bugs), followed by traditional rice in Iloilo and the single transplanted crop in Cagayan (4–5 bugs) (Table 2). Least incidence (52 bugs) occurred in the first and single crops of Pangasinan and the first crop in Iloilo. Farmers do not receive a price deduction for pecky rice thus the rice seed bug is less of a pest than in other countries that value grain quality. The Iloilo and Pangasinan sites were sampled for rice hoppers and their key predators both on traditional and modern rices (Table 4). Nephotettix ˚ nigropictus (Stal) was the most dominant green leafhopper on both traditional and modern varieties in both sites. Six-fold more N. nigropictus were censused from traditional varieties than on modern ones in Iloilo, while in Pangasinan the ratio was 3-fold. The modern rices IR28 and IR36 had resistance to both Nephotettix species. A disease flare-up of tungro occurred in Iloilo on traditional varieties in 1976 but

using the paired t-test for two variables or Least Significant Difference (LSD) test for more than two variables. Means are shown with standard errors of the mean (SEM) using a pooled estimate of error variance. 3. Results

3.1. Insect pest and natural enemy fauna Rice whorl maggot was prevalent in all RWR locations. The ephydrid fly’s larva tunnels into newly developing leaves producing a characteristic feeding injury along the edges when they unfold (Ferino 1968). Loss results from reduced photosynthetic area and disruption of vascular flow. Percent damaged leaves was highest on all crops in Pangasinan among sites and including traditional (29%) and modern cultivars on the single crop (36%), as well as first (23%) and second (25%) crops of the double crop pattern (Table 2). Only in Cagayan was incidence significantly higher on traditional (21%) than modern (5%) varieties. Lowest incidence (6–7% damaged leaves) was in Iloilo in both the first and second crops. Rice caseworm, the most prevalent lepidopteran defoliator in the RWR sites, is a semi-aquatic pyralid whose larva cuts off the tips of young rice leaves that fall onto the water surface before naturally rolling into tubes in which it seeks shelter (Litsinger et al. 1994a). Most feeding occurs at night, but during the day the floating larvae in their cases are swept toward the leeward side of fields by wind or by water coursing to lower-lying fields in the watershed. The greatest concentration of damage occurs after rains flush ‘upstream’ larvae to bottomland fields, particularly during the vegetative stage. In the sloping topography of Iloilo, rice fields drain from a larger watershed than occurs in other sites resulting in a greater concentration effect where a high of 29% damaged leaves was recorded in the first crop of the double crop pattern (Table 2). Moderate damage levels (17–18%) occurred in traditional varieties in Iloilo and Pangasinan with least (4%) in the single crop in Cagayan with modern varieties. The green hairy caterpillar was more encountered than the green semilooper (both noctuids) in the three RWR sites, but both vegetative stage defoliators occurred at very low levels. Other common defoliators were the satyrid greenhorned caterpillar Melanitis leda ismene (Cramer) and the hesperiid rice skipper Pelopidas mathias (F.). The former blends into the bright green foliage, thus even though large, the larvae are often overlooked (Litsinger et al. 1993). The hesperiid also is less noticed as its larvae live within tightly folded leaves (Litsinger et al. 1994b). Both of these species are most often encountered in the reproductive rice stage. A wide range of grasshoppers can be found in RWR, but like the butterflies are low in density. A range of noctuid armyworms and cutworms can attack the vegetative stage, at times in large

Table 2. Insect pest densities from single and double crop rainfed wetland rice from three sites comparing high yielding modern varietial types to traditional rices. Results from rainfed rice are compared to four dryland and four irrigated rice sites. Philippines, 1976–91.1
No. crops Variety Deadhearts Whiteheads No. fields Whorl maggot (% DL) 21–35 DT/DAS Defoliators (% DL) 21–35 DT/DAS Stemborer damage (%) Leaffolders (% damaged flag leaves) Rice seed bug (no./m2) milk stage

Culture

Site

Province

Years

Planting method2

1976–78 1980–82 1976, 79–80 15.9 + 6.8 a 0.6 + 0.3 c 11.2 + 6.7 ab 0.7 + 0.4 c 10.8 + 5.6 ab 3.1 + 2.2 c 1.6 + 1.3 c 8.3 + 4.1 b 9.9 + 4.1 b 0.0 c 1.6 + 0.5 c 0.6 + 0.3 c 0.5 + 0.3 c 0.5 + 0.2 c bc5 bc5 bc5 c5 bc c ab c bc c bc c 11.0 + 2.4 7.5 + 2.8 9.7 + 2.8 0.9 + 3.0 3.6 + 2.6 1.5 + 3.0 6.0 + 2.6 0.8 + 2.8 50.003 2.62 86 c5 c5 bc5 c5 19.2 + 2.6 11.5 + 3.1 19.5 + 2.7 5.8 + 2.7 50.0001 9.15 93 1.7 + 0.4 1.7 + 0.5 1.5 + 0.5 1.6 + 0.5 1.1 + 0.5 4.7 + 1.0 1.8 + 0.5 2.9 + 0.5 50.0001 6.81 113 c c c c c bc c c 0.6 + 0.2 c 1.6 + 1.0 c 1.3 + 1.1 c 3.5 + 1.6 bc 2.7 + 1.7 c 4.4 + 2.3 bc 0c 2.6 + 0.8 c 0.5 + 0.2 c 0.7 + 0.3 c 1.5 + 0.6 c 2.7 + 0.7 4.7 + 1.0 2.5 + 1.0 2.3 + 0.9 2.0 + 0.8 2.4 + 1.1 2.4 + 0.9 1.6 + 0.9 0.0006 2.73 116 c bc c c c c c c 9.0 + 4.5 a 1.4 + 0.7 c 11.1 + 5.8 a TP TP WS WS TP TP DS DS DS DS DS TP TP TP TP TP TP TP TP 11 6 8 8 IR36, 42, 52, 56 IR58, 64 IR60 C1 69 44 57 37 P F df 12 7 7 9 IR36, 42, 52, 56 IR58, 64 IR60 C1 72 44 52 44 20.4 + 2.2 7.4 + 2.9 25.0 + 2.9 11.4 + 3.1 7 3 3 UPLRi5 UPLRi5 UPLRi5 37 19 13 03 03 03 03 03 03 2 5 Benernal Dagge 8 25 03 03 03 03 6 4 IR28, 36 IR28, 36 30 19 24.6 + 18.2 ab 6.7 + 9.0 c 6.2 + 3.6 bc4 8.4 + 8.7 bc4 3 4 IR28, 36 IR28, 36 12 21 23.3 + 19.1 ab 5.9 + 1.1 c 7.8 + 4.4 bc4 29.4 + 27.3 a4 3 6 IR36, 52 IR28, 36 16 33 5.1 + 3.2 c 35.8 + 13.5 a 3.7 + 2.5 c4 11.7 + 6.9 bc4

TP TP TP

3 3 3

BE3, Kapoypoy Wagwag Wagwag, Inano

13 13 15

7.9 + 4.8 c 21.3 + 8.3 b 28.7 + 4.1 ab

16.5 + 11.4 ab4 11.0 + 9.2 bc4 17.9 + 2.8 ab4

9.5 + 4.8 ab 3.7 + 3.8 bc 4.4 + 1.3 bc 3.0 + 0.9 bc 4.8 + 3.5 bc 8.9 + 10.5 ab 7.2 + 5.6 ab 3.3 + 1.7 c 1.4 + 1.1 c 0c 17.3 + 5.1 a 1.2 + 0.3 c 6.5 + 2.8 bc 9.6 + 4.0 ab 6.2 + 1.2 2.3 + 1.4 3.5 + 1.4 1.0 + 1.3 4.2 + 1.2 1.0 + 1.5 7.5 + 1.3 0.4 + 1.3 50.0001 3.52 114 bc c c c bc c ab c

3.8 + 2.0 ab 2.2 + 1.1 bc 2.6 + 1.8 bc 5.2 + 2.4 ab 1.4 + 0.9 c 1.9 + 1.6 c 2.1 + 1.7 c 2.5 + 1.4 bc 6.8 + 4.3 a 4.9 + 2.7 ab 1.3 + 1.6 c 1.0 + 2.1 c 1.0 + 0.9 c 1.3 + 0.9 c 0.2 + 0.3 1.3 + 3.1 1.0 + 0.2 0.6 + 0.6 0.1 + 0.1 0.2 + 0.1 0.9 + 0.6 1.3 + 0.7 50.0001 3.64 105 c c c c c c c c

1980–82 1976–80

1976,79–80 1976–79

1976–80 1976–79

1984–85 1976–80

1984–90 1988–90 1978–80

Rainfed wetland Single crop - traditional varieties Oton/Tigbauan Iloilo Solana Cagayan Manaoag Pangasinan Single crop - modern varieties Solana Cagayan Manaoag Pangasinan Double crop - first crop modern varieties Manaoag Pangasinan Oton/Tigbauan Iloilo Double crop - second crop modern varieties Manaoag Pangasinan Oton/Tigbauan Iloilo Rainfed dryland Single crop - traditional varieties Siniloan Quezon Tanauan Batangas Single crop - modern varieties Claveria Misamis Oriental Tupi S. Cotabato Tanauan Batangas Irrigated wetland Double crop - wet season modern varieties Zaragoza Nueva Ecija Guimba Nueva Ecija Koronadal S. Cotabato Calauan Laguna Double crop - dry season modern varieties Zaragoza Nueva Ecija Guimba Nueva Ecija Koronadal S. Cotabato Calauan Laguna

1979–91 1984–91 1983–91 1984–91

1979–91 1984–91 1983–91 1984–91

International Journal of Pest Management

DL ¼ damaged leaves, DT ¼ days after transplanting, DAS ¼ days after sowing. In a column, means separated by different letters are significantly different (P  0.05) by LSD test, data averaged per crop. 2 TP ¼ transplanted, DS ¼ direct seeded in dry soil, WS ¼ wet seeded, direct sown on puddled soil. 3 Not observed in the environment. 4 Mainly caseworm with 51% Rivula þ Naranga. 5 Mainly Rivula þ Naranga with 51% caseworm.

1

227

228

J.A. Litsinger et al.

Table 3. Stemborer species prevalence determined from tiller dissections at the reproductive stage in three rainfed wetland sites, Philippines, 1976–82. Tiller dissections1 Relative abundance (%) Site Oton/Tigbauan Manaoag Solana Province Iloilo Pangasinan Cagayan Years 1976–78 1977–79 1980–82 Variety IR36 IR36 IR36, IR52 Crops (no.) 3 4 3 Scirpophaga spp.2 60.1 100 77.0 Chilo suppressalis 24.1 0 4.1 Sesamia inferens 8.3 0 13.3 Chilo auricilius 7.5 0 5.6 n 150 hills 200 hills 150 hills

1 One field was sampled per crop that had not been treated with insecticide. Sample size was 50 hills/field. In each site the percentages add to 100%. 2 White and yellow stemborers are indistingishable in the larval stage thus they are not separated by species. White stemborer does not occur in Luzon thus in Pangasinan and Cagayan all Scirpophaga specimens were yellow, while in Iloilo there was a mixture.

not in modern rices. The white leafhopper Cofana spectra (Distant) achieved higher densities on traditional cultivars in Iloilo, while in Pangasinan there was no difference. On the other hand, the zigzag leafhopper Recilia dorsalis (Motschulsky), another vector of tungro, showed no difference in numbers by varietal type or location. High densities of brown planthopper occurred in Iloilo on both traditional and modern rices but three times more on the former which are highly susceptible (Table 4). Whitebacked planthopper Sogatella furcifera (Horvath) occurred in greatest numbers in Iloilo on modern rices. Isolated patches of hopperburn from planthoppers only occurred in Iloilo in 1977 on IR28 and IR30. Despite the high incidence of green leafhoppers and brown planthoppers in the RWR sites, the mirid predator Cyrtorhinus lividipennis Reuter showed no response in density either by site or cultivar. Spiders and ladybeetles, however, were more numerous on traditional rices in Iloilo where leafhopper and planthopper numbers were higher as well as on modern rices in Iloilo attacked by brown planthopper. Thrips Stenchaetothrips biformis (Bagnall) and mealybugs Pseudococcus saccharicola Takahashi were noted only during drought spells in all three sites. Across the sites, lowest egg parasitism rates occurred on whorl maggot (0%) and caseworm (0.3%) (Table 5). Whorl maggot eggs have a calcareous chorion that no doubt retards parasitoid ovipositors, while caseworm eggs are laid underwater making locating them difficult (Litsinger et al. 1994a). In Pangasinan, where whorl maggot was highly prevalent, larval parasitism averaged 19%, with only 1% in the other two sites. The semi-aquatic caseworm larvae also are rarely parasitised as shown by one larva in a sample of 119 parasitised in Pangasinan. Other early-season defoliating Lepidoptera such as the green hairy caterpillar and cutworms registered low levels of parasitism across the three sites. In Cagayan, where S. litura was most abundant, larval parasitism reached a high of 8%, but only 2% in Iloilo. M. separata larval parasitism attained only 2–5% in the three sites.

Parasitism rates were higher on mid- and late-season defoliating Lepidoptera. Greenhorned caterpillar egg parasitism reached 15% in both Iloilo and Pangasinan but was nil in Cagayan. Greenhorned larval parasitism averaged 22% in Pangasinan but only 7–8% elsewhere. Rice skipper egg parasitism was nil in both Iloilo and Cagayan. Larval parasitism was highest in Iloilo 26% with only 7–8% in the other two sites. Leaffolder eggs are very small thus were too difficult to assess, while larval parasitism was very low, averaging 2–5%. Turning to stemborers, some of the highest parasitism rates occurred in Pangasinan with Scirpophaga eggs 33% but only 6–8% in the other sites. Scirpophaga larval parasitism was also high in Pangasinan 32% but 2–6% elsewhere. Larval parasitism for the other stemborer species ranged from 0 to 7%, the highest being of pink stemborer in Cagayan. Leafhopper egg parasitism was high in all three sites ranging from 17 to 35%. The white leafhopper C. spectra revealed 17% egg parasitism, while surprisingly zigzag leafhopper R. dorsalis registered only 0–1%. Leafhopper nymphs and adults are parasitised mainly by pipunculid flies which reached 4–7% on Nephotettix spp. and 0–1% on zigzag. Results were similar for planthoppers. Egg parasitism of N. lugens registered 14–31% while S. furcifera showed similar numbers 18–29%. Nymphs and adults are parasitised mainly by Strepsiptera and dryinids where together on N. lugens they attained a high of 12% in Pangasinan but 2–4% elsewhere. S. furcifera produced somewhat lower numbers: 7% in Pangasinan and 0–2% elsewhere. 3.2. Yield losses Traditional tall varieties in Cagayan and Pangasinan yielded just under 2 t/ha in the unprotected plots but registered a mean 19% yield loss (range 13–25%) or 0.4 t/ha (range 0.1–0.7 t/ha) (Table 6). But only the mean loss in the higher yielding Pangasinan site was significant (P ¼ 0.02, F ¼ 6.70, df ¼ 15), with greatest loss (16%) in the vegetative stage. In Cagayan in the same single-cropped, transplanted culture, yields and

International Journal of Pest Management
Table 4. Comparison of rice hoppers and their predators in two rainfed wetland sites for both traditional and modern varieties using two sampling methods: 1) net sweeps for fast moving insects that inhabit the top of the canopy and 2) per area for more sedentary species lying below, Oton and Tigbauan, Iloilo and Manaoag, Pangasinan, 1976–79.1 30.6 + 17.2 b 11.6 + 7.8 c 0.01 6.54 16 53.7 + 19.0 a 12.0 + 8.5 c Spiders

229

In a column, means separated by different letters are significantly different (P  0.05) by LSD test, data averaged per crop. 20 net sweeps per crop, sampled three times per crop at vegetative, reproductive, and ripening stages. 3 Three samples per crop once each in the vegetative, reproductive, and ripening stages, five samples of 1 m2 were taken each date. 4 BE3 and Kapopoy in Iloilo, Wagwag and Inano in Pangasinan. 5 IR28, IR36.

16.7 + 7.0 a 1.7 + 1.2 b

6.3 + 8.6 ab 3.4 + 5.4 b 0.03 4.37 16

C. spectra

losses with modern varieties were similar to those of traditional rices. Mean losses were higher on modern (22%) than traditional (13%) rices, but not significantly due to the high yield variability in this risk prone site. However, yield potential of modern varieties increased substantially in Pangasinan in the single crop from 1.9 to 3.6 t/ha in untreated plots in a less risky environment. In Pangasinan the percentage yield losses (also significant P ¼ 0.0004, F ¼ 14.07. df ¼ 55) were similar for both traditional and modern rices (24– 25%), but tonnage loss was higher in the latter (0.7 vs. 1.1 t/ha) with the greatest loss occurring in the vegetative stage (11%). In the first wet-seeded crop of the double-crop in both Pangasinan and Iloilo, losses were insignificant, averaging only 2 and 16%, respectively. Losses in Iloilo were higher in years where planting was delayed beyond May as earlier plantings tended to escape pest attack. First crop plantings in Pangasinan were more timely than in Iloilo over years. Highest crop stage losses in Iloilo again were in the vegetative stage (8%). In the second transplanted crop, untreated yields were over 1 t/ha lower than in the first crop in both sites, underscoring the higher risk of late season drought in the double crop system. Losses in the second crop were higher than in the first (18–28 vs. 2–16%) in both sites. Significant losses were 0.7 t/ha in Pangasinan (P ¼ 0.01, F ¼ 7.83, df ¼ 44) and 1 t/ha in Iloilo (P ¼ 0.02, F ¼ 6.87, df ¼ 35) where the untreated averaged 2.6– 3.0 t/ha. Losses were higher in the single vs. the second crop in Pangasinan (24 vs. 18%) even though yield in the unprotected plots was higher in the former (3.6 vs. 3.0 t/ha) showing the greatest stress was in the second crop. Once more losses were highest (13%) in the vegetative stage of the second crop. 3.3. Chemical control

149.3 + 148.0 a 8.7 + 8.1 b 17.9 + 6.9 a 2.8 + 1.9 b 7.7 + 6.7 a 2.0 + 1.0 a 81.3 + 54.9 a 60.8 + 21.7 a 7.7 + 5.9 b 4.0 + 2.6 b

Coccinellids

Per 10 net sweeps2

N. nigropictus

6.7 + 4.2 b 5.6 + 7.8 b 50.0001 8.55 16

1.8 + 1.3 a 1.2 + 0.4 a ns 1.21 16

R. dorsalis

23.2 + 22.7 a 3.0 + 2.1 b 0.04 4.88 16

S. furcifera

7.4 + 8.4 ab 1.9 + 2.1 b 0.03 4.01 16

48.6 + 41.7 a 9.2 + 7.2 b 0.02 4.32 16

N. lugens

3.6 + 4.2 a 12.6 + 21.0 a ns 2.06 16

Cyrtorhinus

(Per m2)3

3.3 + 2.9 a 3.3 + 5.8 a

Prophylactic insecticide application led to significant yield increases in the single rice crop with both traditional and modern rices in Pangasinan and in Iloilo in both the first and second crops with modern rices (Table 7). Almost half of the fields (47%) exceeded an action threshold over the three sites and four categories of crops, with most of the applications directed at the vegetative stage against whorl maggot. However, significant yield increases in the threeapplication, prophylactic regime occurred in four of the seven crop cultures 6 sites 6 varieties tested. In Pangasinan and Iloilo, yield gains were higher (0.5– 0.6 t/ha) in the prophylactic treatment than for action thresholds (0.2–0.3 t/ha). In Pangasinan significant gains occurred in single-crop rice for both traditional and modern rices. Between 38 and 50% of traditional rice fields surpassed action thresholds for whorl maggot and armyworm in Pangasinan and Cagayan, respectively, but only in Pangasinan was there a significant yield response of 0.2 t/ha.

13.2 + 8.6 a 20.3 + 17.2 a 11 11

N. virescens

No. fields

No. crops

Traditional varieties4 Iloilo 3 Pangasinan 3 Modern varieties5 Iloilo 5 Pangasinan 6

Site

26 35 P F df

1.8 + 5.2 b 1.4 + 5.1 b 50.0001 7.65 16

1

2

230

Table 5. Parasitism of the common insect pests of irrigated double cropped rice sites in tropical Asia as reported in the literature compared to single-crop, rainfed, wetland rice sites in three provinces, Philippines, 1976–83. Parasitism (%)1 Rainfed wetland ecosystem Pangasinan3 n 0 19.4 0.8 3.9 1.8 4500 475 14.9 22.4 10.4 5.1 32.5 31.8 34.6 7.0 480 1,411 133 130 8.4 78 15.6 146 39 40 857 243 119 119 109 111 0 35.0 3 3 % n % n All sites (mean %) % n no. crops Cagayan4 Irrigated wetland ecosystem Citation van den Berg et al. 1988 Ferino 1968

Iloilo2

J.A. Litsinger et al.

Pest

Stage

%

Hydrellia philippina

Nymphula depunctalis Rivula atimeta Mythimna separata Spodoptera litura Melanitis leda ismene

Pelopidas mathias

Cnaphalocrocis medinalis, Marasmia spp. Scirpophaga spp. 201

Egg Larva Larva Larva Larva Larva Egg Larva Egg Larva Larva

0 1.1 0 0 5.4 2.3 14.7 7.5 0 26.0 1.7

132 89 43 369 175 97 119 301 21 37 418

0 0.9 0 1.2 2.9 7.9 0 6.9 0 7.8 2.4

53 71 35 56 57 126 86 163 21 118 67

0 7.1 0.3 1.7 3.4 5.1 9.9 12.3 0.0 14.7 3.1

Egg

5.9

Chilo spp. Sesamia inferens Nephotettix spp. 186 200 876 850 539 79 28.7 6.5 332 390 27.4 2.3 340 156 13.8 12.0 520 395

Larva Larva Larva Egg Adult

1.5 1.8 0 16.5 3.6

401 33 34 1253 850

4.5 3.2 7.1 30.4 6.0

87 55 39 454 567

12.6 2.5 3.6 27.2 5.5

20.8 10.1 14.2 13.5 40 35.9 50 30.3 59.55 0.4–3 12 19 22.0 20.5 21.5

844 2361 762 2077 4904 41000 41000 41000 41000 147 29 50 41000 41000 41000

410 410 410 410 2 6 2 3 1 3 3 3 1 2 2

Litsinger et al. 1997 Litsinger et al. 1997 Litsinger et al. 1997 Litsinger et al. 1997 Arida & Shepard 1990 de Kraker et al. 1999 Shepard & Arida 1986 Rothschild 1970 Litsinger et al. 2006c Rothschild 1970 Rothschild 1970 Rothschild 1970 IRRI 1978 IRRI 1978, 1979 Pena & Shepard 1986 ˜

Recilia dorsalis

0 0

Cofana spectra Nilaparvata lugens

Egg Adult Egg Egg Adult

19.7 1.8

1.2 0.9 17.0 31.3 4.3

345 334 37 628 546

18.5 0.5 17.0 21.6 6.0 24.7 3.1

Sogatella furcifera

Egg Adult

18.1 0.4

8.0 26.3 8.5 12 25 8.3 18

41000 41000 41000 41000 41000 41000 41000

1 3 2 2 2 3 2

IRRI 1983 IRRI 1978, 1980b IRRI 1978, 1979 Pena & Shepard 1986 ˜ IRRI 1978, 1981b IRRI 1978, 1979, 1981b Pena & Shepard 1986 ˜

1

2

Eggs/egg masses and larvae collected from the field on excised leaves were held in vials and petri dishes for parasitoid emergence. Mean of three crops 1976–78. 3 Mean of two crops 1976–77. 4 Mean of three crops 1981–83. 5 Synchronous planted area only.

Table 6. Yield loss calculated by the partitioned growth stage insecticide check method for three single and double cropped rainfed wetland rice sites involving traditional and high yielding modern varieties. Further comparison was made with four dryland and four irrigated double cropped sites, Philippines 1976–91. Yield loss Yield (t/ha) Province Crops Protected Untreated t/ha
1

Total % Vegetative Reproductive

By growth stage (%) Ripening

Crop agro-ecosystem

Site

Rainfed wetland 1.84 2.58 2.21 1.99 4.72 3.36 4.52 4.10 4.31 3.65 3.70 3.68 0.48 2.90 1.69 3.43 3.46 4.01 3.63 5.09 5.16 4.39 4.61 6.23 4.85 4.80 4.79 4.99 4.42 4.55 3.67 4.27 5.50 4.10 4.03 4.38 4.37 2.64 2.96 3.61 3.07 0.06 2.85 1.46 0.42*** 0.05 ns 0.24 0.79* 0.50 ns 0.40 ns 0.56 0.70*** 0.60*** 0.72*** 0.30*** 0.63*** 0.75*** 0.77*** 0.39*** 0.61 2.56 3.02 2.79 1.09* 0.68* 0.89 28 18 23 88 2 45 23 15 18 19 13 11 22 6 10 15 18 8 13 14 7 12 11 5 4 9 4 4 7 5 2 5 7 8 4 6 5 4 7 1 4 4 8 4 5 2 0 2 1 3 3 5 1 2 3 5 2 3 3.84 4.00 3.92 0.68 ns 0.10 ns 0.39 16 2 9 8 0.1 4 13 12 13 1.65 3.61 2.63 0.34 ns 1.12** 0.73 22 24 23 8 11 10 8 7 8 4 1.3 3 6 3 5 1.73 1.94 1.84 0.11 ns 0.65* 0.38 13 25 19 6 16 11 5 5 5 2 4 3 6 6 6 4 70.2 2 9 3 6

Single crop transplanted - traditional varieties Solana Cagayan 3 Manaoag Pangasinan 2 Average Single crop transplanted - modern varieties Solana Cagayan 3 Manaoag Pangasinan 5 Average First crop wet seeded - modern varieties Oton/Tigbauan Iloilo 4 Manaoag Pangasinan 3 Average Second crop transplanted - modern varieties Oton/Tigbauan Iloilo 4 Manaoag Pangasinan 5 Average 2 5 Average 7 4 3 Average 12 7 7 9 11 8 6 8 Average

Rainfed dryland

Traditional varieties Siniloan Quezon Tanauan Batangas Misamis Oriental S. Cotabato Batangas

Modern varieties Claveria Tupi Tanauan

Irrigated wetland

International Journal of Pest Management

Wet season transplanted - modern varieties Zaragoza Nueva Ecija Koronadal S. Cotabato Guimba Nueva Ecija Calauan Laguna Dry season transplanted - modern varieties Zaragoza Nueva Ecija Koronadal S. Cotabato Guimba Nueva Ecija Calauan Laguna

1

Levels of significance between protected and untreated: ns (P 4 0.05), *(P 5 0.05), **(P 5 0.01), ***( P 5 0.0001).

231

232
Table 7. Performance in the use of action thresholds for insecticide decision making in three rainfed wetland rice sites compared to use of a prophylactic application, Philippines, 1976–82.

J.A. Litsinger et al.
Prophylactic

TP ¼ transplanted, WS ¼ pre-germinated direct wet seeded rice. The full protection treatment shows the yield potential from insecticide application while the untreated acts as a control. 3 WM ¼ whorl maggot. AW ¼ armyworm, DEF ¼ Naranga/Rivula defoliators, LF ¼ leaffolders, CW ¼ caseworm, RB ¼ rice seed bug, SB ¼ stemborers. 4 Levels of significance: ns (P 4 0.05), *(P 5 0.05), **(P 5 0.01). 5 An economic analysis was performed only on those crops and treatments where there was a significant increase in yield.

Action threshold

70.01 0.29

Only 24% of fields in the single crop rice with modern rices in Pangasinan exceeded thresholds which involved whorl maggot, Naranga/Rivula defoliators, and leaffolders. In the double crop pattern, action thresholds were most frequently surpassed in Iloilo, 86% in the first crop and 95% in the second from whorl maggot, caseworm, stemborers, and rice seed bug. Fewer fields in Pangasinan (13 and 25%) exceeded the action thresholds in each crop of the double crop pattern which was more or less equal to that in the single crop pattern. Leaffolders were the only insect pests in the first crop, whereas whorl maggot and rice seed bug occurred in the second crop in Pangasinan. Economic analyses showed that in no case was there an acceptable profit from chemical control, as in only one crop (the single crop with modern rices in Pangasinan) was there not a loss, but only produced a return of 5$3/ha. The prophylactic regime fared even worse with losses per crop in the order of $38–50/ha. Therefore the benefit: cost ratios in all cases are unacceptable to farmers. 3.4. Ecosystem comparisons RWR arthropod populations were compared to sites in dryland and irrigated environments based on light trap collections, rice field sampling, and yield loss assessments. 3.4.1. Light trap collections Light trap seasonal totals of daily collections give an indication of relative abundance within a radius of 0.5– 1 km, the mean dispersal distance of most common rice insects (Loevinsohn et al. 1988). Light traps were less attractive to leaffolders, caseworm, and other lepidopterous defoliators thus they were not included. First, averaging the crop totals over each agroecosystem and planting synchrony type (see the rice agroecosystem classification averages in the bottom half of Table 8), brown and whitebacked planthoppers, zigzag leafhopper, and the Cyrtorhinus predator registered highest collections in the asynchronous irrigated agroecosystem, while there were no differences among RWR, dryland, and synchronous irrigated agroecosystems. Notably, the RWR agroecosystem was statistically equal to the asynchronous, irrigated system with regard to the abundance of green leafhoppers, and more than the two other agroecosystems. With regard to Scirpophaga stemborers, densities were higher in the asynchronous irrigated agroecosystem than in the dryland sites, with RWR and synchronous irrigated systems being intermediate. There were no differences among agroecosystems regarding white leafhopper and non-Scirpophaga stemborers. The two asynchronous multi-rice cropped sites were: (1) Zaragoza which was at the

746.70

740.01

737.45

Return ($/ha)5

Action threshold

Yield difference from untreated (t/ha)4

Significance

ns **

79.08

**

2.40

ns **

Prophylactic

0.22 0.50

0.56

0.02 0.58

Significance

ns *

ns ns

0.17 0.23

0.32

WM, DEF, LF

LF WM, CW, RB

Target pests3

WM, AW WM

% fields treated for action threshold

Untreated

0.82 1.73

4.02

3.71 3.92

Yield (t/ha)2

Full protection

1.27 2.22

4.93

3.76 4.56

Wagwag Wagwag

IR28, 36

Variety

IR28,36 IR28,36

Planting method1

WS WS

TP TP

TP

Single crop - traditional varieties Solana Cagayan Manaoag Pangasinan Single crop - modern varieties Manaoag Pangasinan Double crop - first crop modern varieties Manaoag Pangasinan Oton/Tigbauan Iloilo Double crop - second crop modern varieties Manaoag Pangasinan Oton/Tigbauan Iloilo Average

Culture

Site

Province

TP TP

IR28,36 IR28,36

2.77 3.76

2.45 2.56

13 95 47.3

38 50

24

25 86

WM, RB SB, RB

0.16 0.27

ns ns

*

0.30 0.48

ns *
1

750.25

2

Table 8.
Seasonal total per light trap per location4

Abundance of insects and factors in rice cropping intensity by rice agroecosystem as determined from kerosene light traps set up in twelve sites in the Philippines, 1979–91.

Rice ecosystem1

Town

Province

No. Area in crops/ rice year (%)2 Brown No. planthopper crops N. lugens C. lividipennis mirid Scirpophaga predator spp. stemborers5 Green Whitebacked leafhoppers planthopper S. furcifera Nephotettix spp. Zigzag leafhopper R. dorsalis White leafhopper C. spectra

No. insect-cide kg N/ applications/ crop3 ha3

Other stemborers6

Site averages Dryland 3 9 30 720 + 294 c 312 + 103 ab 305 + 79 c 1,016 + 612 bc 458 + 185 a 195 + 95 c 595 + 269 b 455 + 314 b 410 + 168 b 0 3 2,908 + 864 c 498 + 115 b 1,449 + 398 c 1,802 + 161 b 80 + 22 b 1,780 + 272 b 40 0 2 262 + 75 d 130 + 37 b 75 + 29 c 190 + 73 c 37 + 19 b 5 0 2 185 + 78 d 451 + 144 c 425 + 158 b

Siniloan

1.0

41 + 7 c 45 + 11 c 170 + 43 c 178 + 98 c 315 + 87 bc

Claveria

1.0

Tupi

1.0

14

Rainfed wetland

Laguna/ Quezon Misamis Oriental South Cotabato Batangas Cagayan Pangasinan Iloilo 60 0 20 30 40 60 30 60 30 4,043 + 1,117 a 50.0001 6.76 40 776 b 953 b 957 b 4,043 a 0.04 6.15 8 709 b 2,612 ab 650 b 6,183 a 0.02 6.30 11 50.0001 5.54 49 P F df 50.0001 37.01 49 871 b 868 b 682 b 13,196 a P F df 50.0001 40.95 11 4.5 3.0 2 4 15,224 + 238 a 11,168 + 1,849 b 9,207 + 2,561 a 3,158 + 501 b 3.0 4 1,677 + 345 cd 549 + 152 b 2,279 + 617 c 855 + 169 c 4.5 2 154 + 14 d 361 + 77 c 93 + 17 b 706 + 252 b 3.0 3 214 + 79 d 1,366 + 896 b 629 + 426 c 0 0.1 0.4 1.3 2 4 2 2 130 + 78 d 491 + 217 d 289 + 94 d 1,824 + 528 cd 1,700 + 367 b 1,527 + 527 b 321 + 81 b 1,012 + 215 b 860 + 452 c 3,218 + 2,320 b 2,179 + 538 c 2,440 + 770 c

1.0 1.0 1.0 1.0

20 60 85 85

50 + 23 c 437 + 133 c 475 + 66 c 997 + 108 b 413 + 88 c 305 + 17 c 662 + 83 bc

131 + 36 c 0c 38 + 30 c 35 + 35 c 0c 405 + 46 b

Irrigated (synchronous)

1.9

90

2.0

80

2.0

70

Irrigated (asynchronous)

Tanauan Solana Manaoag Oton/ Tigbauan Victoria/ Laguna Sta Maria Cabanatuan/ Nueva Ecija Zaragoza Koronadal South Cotabato Jaen Nueva Ecija Koronadal South Cotabato 2,961 + 389 a 145 + 21 ab 50.0001 24.02 40 904 b 661 b 551 b 2,961a 0.02 12.09 7 0.04 3.12 32 59 385 93 145 0.12 ns 7.33 5

2.0 2.4

80 70

7,169 + 720 a 50.0001 17.46 36 1,103 b 525 b 558 b 7,169 a 0.007 37.03 6

137 + 59 c 2,519 + 430 a 50.0001 19.91 49 77 b 636 ab 460 ab 1,328 a 0.04 3.04 11

0c 751 + 123 a 50.0001 17.92 45 247 56 147 376 0.07 ns 0.76 9

Rice agroecosystem classification average Dryland Rainfed wetland Irrigated (synchronous) Irrigated (asynchronous)

1

International Journal of Pest Management

Synchrony refers to farmers’ planting dates and synchrony is defined when farmers plant within a period of one month, the average generational period of most insect pests. Rice area within the circumference of 1 km2 around each light trap siting. 3 Based on formal and informal farmer surveys during the seasons of light trap collection. 4 Daily counts from kerosene light traps of similar design. No data indicates that the insect in question was not measured. Means (+ SEM) in columns, followed by a common letter are not significantly different (P 5 0.05) by LSD analysis. 5 S. innotata, S. incertulas. 6 Chilo, Sesamia, Maliarpha.

2

233

234

J.A. Litsinger et al. has been shown to be significantly influenced by nitrogen level and subject to insecticide resurgence (Litsinger 1994). Regression analyses of nitrogen rates and insecticide frequency shown in Table 8 were not significantly correlated with abundance by linear or non-linear models. The reason could be that the previous confirming studies had been conducted at the field level, whereas in the current study light traps sampled populations at the landscape level. 3.4.2. Field sampling

tail-end of a large irrigation system and (2) Koronadal which had communal irrigation systems fed year-round by free-flowing, artesian wells. Even though in both of these sites many farmers planted resistant rices, genetic resistance had broken down, probably due to a combination of multiple rice cropping which extended pest generations and high insecticide usage which caused resurgence (Gallagher et al. 1994). On the other hand, non-Scirpophaga stemborers and white leafhopper showed no differences between rice cultures. Among the RWR sites (see top half of Table 8), Iloilo had higher populations of brown planthopper, zigzag and white leafhoppers, and Scirpophaga stemborers, while Cagayan had higher green leafhopper densities. The three RWR sites were equal for whitebacked planthopper, Cyrtorhinus predator, and non-Scirpophaga stemborers. The two RWR and dryland sites with highest brown planthopper counts were those nearest to irrigated areas having extensive plantings of susceptible varieties (Iloilo and Tupi). Focusing on brown planthopper, Iloilo was grown to a combination of susceptible (IR28 and traditional rices) and resistant (IR36) varieties over time. Pangasinan had a similar mixture of resistant and susceptible cultivars but was not near to an irrigated rice area. Cagayan, like dryland sites, was planted mainly with susceptible varieties but was a single crop pattern and not situated near to irrigated sites. Two of the irrigated rice sites in Laguna province and Cabanatuan/Zaragoza, Nueva Ecija were dominated by resistant cultivars which resulted in low densities as genetic resistance held. Green leafhoppers seemed to be highly favoured by wetland sites where long-maturing, traditional varieties dominated. Cagayan was essentially a pure traditional variety site, while Iloilo and Pangasinan were sown to mixtures of modern rices and traditional types. Dryland environments were planted to susceptible, shortseason, traditional, japonica varieties planted highly synchronously thus allowed only 2–3 generations compared to 5–6 in Cagayan. Cyrtorhinus, a mirid predator of both planthopper and leafhopper eggs and young nymphs, responded more to planthopper than leafhopper densities across ecosystems. The Central Luzon sites in Pangasinan and Nueva Ecija were pure yellow stemborer sites, whereas Iloilo had a mixture of white and yellow stemborers. Scirpophaga stemborers were dominant in RWR ecosystems, whereas in dryland sites non-Scirpophaga stemborers dominated. Iloilo and Cagayan were near to large maize areas which fostered other species, whereas Pangasinan had only small maize areas early in the wet season as it was less elevated. Dryland rice sites, with the exception of Tupi, had the smallest arthropod catches, probably due to the small rice area and limited growing season, characteristic of this ecosystem (Loevinsohn et al. 1988). Planthopper, stemborer, and leaffolder abundance

Whorl maggot, caseworm, and Naranga/Rivula were observed only in both wetland ecosystems (Table 2). The highest incidence of whorl maggot in the RWR ecosystem consistently occurred in all crops in Pangasinan (23–36% damaged leaves) as well as only the traditional rice in Cagayan (21%). In the irrigated ecosystem, high incidence (19–25% damaged leaves) occurred in Zaragoza and Koronadal in both wet and dry season crops. Caseworm density was much higher in RWR (9–11% damaged leaves for both varietal types) than in irrigated rice on modern varieties (50.5% damaged leaves). In Iloilo, caseworm damage was consistently higher than whorl maggot damage in all crops. Naranga/Rivula were more abundant in irrigated (5% damaged leaves) than in RWR (51% damaged leaves) ecosystems. Caseworm damage was twice as high on average in the RWR agroecosystem than Naranga/Rivula damage was in the irrigated ecosystem. The two sites with the highest stemborer damage were in the RWR agroecosystem (Iloilo and Pangasinan) especially on traditional varieties. Looking among crops grown to modern rices, the Pangasinan single crop had the highest stemborer incidence of all sites. The next highest sites were Iloilo second crop and Guimba, an irrigated site. Three of the sites with lowest stemborer damage were in the drylands which matched the levels observed in Cagayan. Leaffolder damage was greatest in Batangas, a dryland site, ranging from (10 to 17% damaged leaves) but was also equally high in Iloilo on traditional rice, the first crop both in Iloilo and Pangasinan, as well as in irrigated Koronadal in the dry season. Rice seed bug abundance was highest in more crops in the RWR environment but least in the four irrigated sites and three of the four dryland sites (the exception was the slash and burn rainforest site in Siniloan). Although there were sites and crops in the RWR agroecosystem that had low densities for some pest classifications (which was also true of all agroecosystems), RWR registered the highest proportion among the 6 pest 6 8 crop 6 variety categories in Table 2 with overall statistically highest pest densities in 26% of occasions (14 of 54 categories), than the 10% in dryland (3 of 30) or 4% in irrigated (2 of 48) ecosystems.

International Journal of Pest Management 3.4.3. Yield losses Yield losses also were compared across agroecosystems. The least biased method of measuring losses across sites and agroecosystems is percentage rather than tonnage. There was no difference in losses between wet (0.58 t/ha and 13%) and dry (0.64 t/ha and 13%) season, irrigated rice thus the two seasons were combined (Table 6). Across agroecosystems and varietal types, the highest percentage yield losses from insect pests were traditional varieties of dryland rice. However this classification, composed of two sites, was highly influenced by the slash and burn site Siniloan with an ‘off-the-chart’ 88% yield loss. The other site, Batangas, registered only a 2% loss; therefore the difference was more a site effect than an ecosystem effect. The dominant cropping pattern in RWR ecosystems is the single crop, and losses there ranged from 19% for traditional and 23% for modern rices with fairly consistent agreement within each site. This matches the mean yield loss in dryland rice with modern varietal types (23%) and is greater than that for irrigated rice (13%). However, some irrigated rice sites such as Guimba were similar with 18–22% loss for both wet and dry seasons. Comparing the three main agroecosystems for modern varieties, losses in terms of tonnage ranged from 0.56 to 0.73 t/ha based on single crop RWR, the highest loss. However the 0.73 t/ha figure was high because of Pangasinan at 1.12 t/ha, the highest loss for any site 6 crop 6 variety category. The dryland agroecosystem had the highest and lowest yield losses while irrigated rice had a range of sites from very low losses (6–8%) up to a maximum of 22%. All of the RWR sites equalled or exceeded the maximum losses in irrigated rice. Thus, both the rainfed ecosystems had on average higher losses than the more stable irrigated ecosystem. 4. Discussion

235

4.1. Insect pest and natural enemy fauna As irrigated rice replaced RWR in favourable environments, it is not surprising that the insect pest fauna are highly similar. Flooding in both wetland ecosystems kills off soil pests which inhabit the dryland rice agroecosystem and attracts species that are adapted to rice in standing water. Ponding restricts soil pests to rice bunds. Of the common insect pests in the Philippines in the more studied irrigated rice ecosystem (Litsinger et al. 2005), only green semilooper was less encountered in the three RWR study sites. It is unknown why this is so as it has a wide host range (Pantua and Litsinger 1984). In terms of differences in pest abundance between irrigated and RWR, the results of Table 2 show that site rather than ecosystem differences are more important. As RWR is more prone to drought, insects favoured by those conditions such as thrips and mealybugs were occasionally noted

in the three study sites. These pests flourish in rice in all rice agroecosystems when under drought stress (Mochida et al. 1987). A set of 10 characteristics inherent in RWR that defines this ecosystem in terms of pest prevalence can be enumerated based on this study as well as the literature: (1) large rice bunds that create dryland habitat for soil pests, (2) more diverse natural floral habitats and cropping patterns that provide alternative hosts and refugia, (3) low insecticide usage that allows more pest species to survive, (4) low inorganic fertiliser usage that engenders low yield potential, (5) more fluctuating paddy water levels that reduce performance of granular pesticides and inorganic fertilisers, (6) more pest susceptible traditional varieties present, (7) being tall, traditional varieties elongate over longer periods making them more vulnerable to stemborers, (8) longmaturing traditional varieties promote more pest generations, (9) due to land fragmentation, sowing and transplanting are normally staggered over many months fostering pest build-up, particularly if rains are delayed, but (10) a long rice-free dry season restricts generational build-up of both pests and natural enemies to only the wet season. Irrigated rice is normally earlier maturing so cultural practices of harvesting and land preparation for the dry season crop break up pest cycles to a greater degree than in RWR for pests which reside in the stubble. With a shorter fallow period, most natural enemies tend to be more favoured in irrigated multi-crop rice. The context in which these factors affect densities of RWR pest guilds is elaborated in the following discussion. Flooding in both wetland ecosystems kills off soil pests which inhabit the dryland rice ecosystem and attracts species that are adapted to rice in standing water. Ponding restricts soil pests to rice bunds. RWR fields often have large bunds, some 41.5 m, as farmers seek to maximise rainwater storage allowing dryland rice soil pests such as ants, termites, mole crickets, and white grubs to become occasional pests (Heinrichs et al. 1986). Infestations tend to occur in high-lying portions of uneven fields particularly on light soils. In a normal wet season, transplanting the single RWR crop is delayed after the first rains while farmers wait for their fields to become saturated, a necessary step for puddling the soil. Farmers first sow seedbeds, but as transplanting labour is normally scarce, farm families can take many weeks to establish their fragmented fields resulting in staggered plantings. Mechanisation which speeds up these operations in irrigated rice is not economical for a single rice crop, rainfed or irrigated. But if heavy rains arrive early, farmers can direct sow pre-germinated rice. Both volunteer rice and weeds emerge with the first heavy rains to provide habitat for rice arthropod development. Long delays in crop establishment encourage more pest generations to develop. Cutworms and

236

J.A. Litsinger et al. prior to the modern rices in the Philippines (Cendana ˜ and Calora 1967). A preferred caseworm oviposition site is under leaves that lie flat on the water surface, common in the vegetative stage of a rice crop (Litsinger et al. 1994a). This stage bears succulent leaves that facilitate leaf rolling, thus traditional rices with their long vegetative periods would also favour caseworm development. Caseworm also is highly susceptible to insecticides (Litsinger and Bandong 1992), thus area wide populations become depressed from the extensive chemical usage common in irrigated rice bowls. The reason why caseworm was so abundant in the first crop of double crop rice in Iloilo was due to the late (July), and therefore staggered, planting that would allow more generations to develop. In the 2 years of late plantings, defoliation averaged 51%, whereas the 2 years with earlier plantings (May and June) the figure was only 7%. Thus, early plantings normally escape damage. The second crop of rice should have had highest densities among all crops as caseworm has few natural enemies (Litsinger et al. 1994c). However, low rainfall typical of late wet seasons results in intermittent ponding, which perhaps proved unfavourable to the gill-bearing, semiaquatic larvae (Litsinger et al. 1994a). The greenhorned caterpillar, rice skipper, and grasshoppers are more prevalent in RWR sites probably because of lower insecticide pressure than in irrigated rice as was noted in Japan (Kiritani 1992). Both butterfly species diapause thus are highly adapted to a RWR environment. Greenhorned caterpillars, which aestivate as adults with wings folded during the crop-free dry season in litter under trees near rice fields, mimic dead leaves and are readily flushed when one walks past. However, RWR can readily tolerate the moderate defoliation levels they cause during the reproductive rice stage. Traditional rainfed wetland rices exhibited a higher incidence of stemborer damage than on modern varieties (Table 2). Because they are photoperiodsensitive, traditional rices normally sustain five or more generations per crop. Also, being tall, they undergo an extended period of stem elongation and thus longer vulnerability to first instar larval tiller penetration as protective silica bodies become less compact with growth (Bandong and Litsinger 2005). Traditional rices are also low tillering, which concentrates attack particularly among Scirpophaga species where only one larva per tiller normally occurs. In the RWR sites, stemborer damage to modern rices was low to moderate thus equivalent to that of irrigated rice levels but probably for different reasons than stated above. Litsinger et al. (2006b) found that, in the asynchronous portion of Koronadal, stemborers were suppressed due to the sustained pressure from egg parasitoids and predators that also benefited from the continuous availability of rice. On the other hand, the long rice-free period in RWR depressed stemborer

armyworms commonly build-up under these circumstances and thus are more common in RWR than irrigated rice. On the other hand, if rice is direct seeded, the early planting reduces time for pest reproduction. RWR farmers commonly state that late plantings lead to increased pest densities. Natural enemies are slower to colonise, allowing pests to outpace them. Close proximity to irrigated areas often leads to increased pest abundance for the more dispersive species such as planthoppers, leafhoppers, and mirid predators. In RWR areas it is common for N. nigropictus to outnumber N. virescens probably due to the former’s wider host range and thus survival ability over the fallow period (Ishii-Eiteman and Power 1997). Sogawa (1976) found that green leafhopper peaks, just like brown planthopper in single crop RWR, were generally more distinct. Long-maturing varieties, common in RWR, allow exponential leafhopper build-up for five or more generations under less natural enemy pressure than occurs in irrigated double crops. Based on this evidence and the field sampling in Table 4, the high populations of green leafhoppers reported in S and SE Asia, where direct feeding damage was reported (Alam 1967; Viswanathan and Kalode 1984), can be attributed to extensive single-crop RWR areas sown to highly susceptible traditional varieties. Singh and Singh (1985) found that such varieties can become ‘hopper burned’ without any other provocation. Green leafhoppers have become less important with replacement of large tracts of traditional varieties with resistant modern ones. RWR culture tends to favour vegetative stage pests such as whorl maggot and caseworm that undergo more generations from the combination of more prolonged vegetative periods of longer maturing rices plus staggered planting. Rice whorl maggot was first described in the Philippines as a new pest of irrigated rice in the early 1960s soon after modern rices were developed and was assumed to be a new pest spawned by irrigated rice culture (Ferino 1968). Modern rices, being higher tillering, provide greater leaf densities to encourage higher populations. Also whorl maggot survival is highly benefited by irrigation as field drainage is a recommended cultural control method (Litsinger 1994). The possible reason for this is discussed in Jahn et al. (2007). Irrigated rice area greatly expanded with the success of modern varieties; this has created more juxtaposed RWR and irrigated agroecosystems allowing whorl maggot easy colonisation of RWR. Like whorl maggot, Rivula was identified only after irrigation expanded (Malabuyoc 1977). For both Rivula and whorl maggot, their new discoveries seem surprising in light of their abundance in the three widely separated RWR sites which means that they either: (1) were overlooked, (2) were recent introductions from another country, or (3) seasonally migrate to RWR from irrigated areas. None of these three rice pests was mentioned in the early rice pest literature

International Journal of Pest Management populations particularly if rice stubble is ploughed under prior to the rainy season as both Scirpophaga species lie dormant in the crowns. S. innotata aestivates throughout its distribution (Litsinger et al. 2006a), while S. incertulas enters quiescence in locations with cold winters (Islam 1993). In all three RWR sites there was always a higher ratio of Scirpophaga to nonScirpophaga stemborers, pointing to Scirpophaga borers as being better adapted to a wetland environment. In dryland rice sites where early maturing rices are preferred, rice stubble is often ploughed up to establish a follow-on maize crop killing aestivating larvae, but the rice area is smaller than is typical for wetland environments, allowing more non-Scirpophaga borers that multiply on maize and sugarcane to colonise and concentrate. Rice leaffolder populations are often highest in fields where high N rates were applied as noted in Batangas, a dryland site (Table 8). Ovipositing moths are attracted to the most vigorous growing fields (de Kraker et al. 2000), but the long rice-free dry season limits natural enemy population buildup (Barrion et al. 1991). Thus, the high leaffolder damage levels in the first RWR crop in Iloilo and Pangasinan (Table 2) were probably due to the low initial populations of natural enemies. Most likely beneficials then increased over the season to suppress leaffolders in the second crop where additionally drought stress limited survivorship from suboptimal dietary moisture (Mochida et al. 1987). Tall traditional rices are often sown in the lowest-lying, flood-prone fields that accumulate greater amounts of alluvium and thus have higher levels of N which could be the reason why traditional rice in Iloilo had high leaffolder damage levels. The highest rice seed bug densities occurred in RWR environments and the lowest in irrigated rice. There are two possible reasons for this: (1) RWR fields typically have larger bunds which allow more alternative hosts to become established and (2) shaded wooded areas, more prevalent in RWR than irrigated habitats, serve as a refuge necessary to aestivate. Brown planthopper numbers were highest in Iloilo where the mean density of 149/m2 converts to 0.4/tiller (based on 25 hills/m2 and 15 tillers/hill), near to an action threshold level of 0.5–1/tiller (Litsinger et al. 2005). Spider densities of 54/m2 were one-third those of brown planthopper. As a spider can kill a mean of five planthoppers per day (Kenmore et al. 1984), natural enemies were sufficiently abundant to contain the population, thus there was no planthopper-caused hopperburn in our trial plots. The beneficial role of spiders was underscored in Cagayan after the flood waters from a large typhoon carried spiders away, resulting in localised hopperburn from immigrating brown planthoppers (Litsinger et al. 1986). Data from the literature were used to compare parasitism rates of common rice pests in irrigated sites in the Philippines and elsewhere in tropical Asia with

237

our results from the three RWR sites. Of the 17 pest 6 life stage comparisons in Table 5 only in five were the mean rates of parasitism in RWR higher than in irrigated rice while in nine the parasitism rates of irrigated rice were more than twice that of RWR. Interestingly greenhorned and skipper larval parasitism rates in RWR sites exceeded those in irrigated locations while those for the egg stage did not. The same growth stage reversal of parasitism was true for Scirpophaga larvae where rates were higher in RWR (particularly in Pangasinan), but those of the egg stage were substantially lower than irrigated sites. Rates of leaf-/planthopper egg parasitism, as a guild, matched and occasionally exceeded those of irrigated rice. The explanation for this is egg parasitoids build up on the many hopper species feeding on weeds and volunteer rice that emerge at the beginning of the rainy season in RWR sites due to delays in transplanting. As egg parasitoids have shorter life spans than parasitoids that attack nymphs and adults, this does not hold true for pipunculid, Strepsiptera, and dryinids. These results support the generalisation that natural enemies are more prevalent in irrigated than RWR ecosystems. 4.2. Yield losses

Yield losses were both low and high between different sites within each ecosystem (Table 6). Such differences can be explained in terms of cultivar type, yield potential and crop compensation as well as insect pest abundance. Traditional rices have considerable ability to tolerate defoliation, which under certain circumstances, can even cause a yield increase (Litsinger 2009). That modern rices were unable to out-yield traditional cultivars in Cagayan shows how unfavourable the environment was, oscillating between drought and submergence, often repeatedly, within the same season. But in the two favourable RWR sites, modern rices were successfully grown, achieving much of their yield potential. Remarkably modern rices even appear to have greater powers of compensation than traditional cultivars from not only leaf removal but also tiller removal (Litsinger et al. 2005, 2006c). As traditional varieties tiller poorly, loss of tillers cannot be readily compensated. During the vegetative stage, modern rices have shown no yield loss from up to 50% defoliation (van Haltern 1979) or removal of 30% of tillers (Rubia et al. 1990). Compensation from stemborer whiteheads (Rubia and Penning de Vries 1990) and rice seed bug damage (Litsinger et al. 1998) can also be high from a reallocation of photosynthate from damaged to undamaged plant parts. Litsinger (1993) showed that increasing levels of crop management resulted in progressively lower losses in irrigated sites with modern rices. However, compensation can be compromised by: (1) early maturing rices which do not allow enough time to outgrow the injury,

238

J.A. Litsinger et al. under heavy pressure from rats and blast disease. The insecticide check method protected the crops from soiland sown-seed pests, the main insect pest guilds (Litsinger et al. 2009). Applied inorganic fertiliser had little benefit on the poor soils thus compensation was limited. The one irrigated site in Guimba was served by a dysfunctional small irrigation system with an electric pump which often failed from power outages thus drought stress was ever-present and stemborer infestation was perennially high. In addition farmers chose the earliest maturing varieties such as IR58 to sow. The other three crops were from the RWR ecosystem with all three study sites represented. In Pangasinan and Cagayan, the main single transplanted crops registered highest losses with both traditional and modern varieties in the former. The second crop in Iloilo also suffered high losses. All of these crops suffered from a combination of high pest pressure and drought stress. Other RWR and irrigated crops suffered high insect pest infestations but did not register highest losses. Turning to the four crops with lowest yield losses (510%), the Batangas dryland site sown to the traditional variety was on highly fertile, young volcanic soils under high inorganic N while both crops in Calauan, an irrigated site, were sown mainly to longer maturing varieties (Litsinger et al. 2005). The direct seeded first crops in Pangasinan and Iloilo, when planted early, escaped both transplanting shock and insect pest pressure and were able to avail of the natural soil fertility from mineralised soil nitrogen built up over the dry fallow. From these examples we see that compensation is favoured by minimal crop stresses counterbalanced with good crop management and longer maturing varieties. 4.3. Chemical control trials

(2) suboptimal agronomic management, (3) unfavourable weather, or (4) the number of stresses. Poston et al. (1983), Pinnschmidt et al. (1995)], Litsinger (1991, 1993) and Boling et al. (2004) found that losses in modern rices are exacerbated when more than one stress (biotic or abiotic) affects the same growth stage. Within irrigated sites such as Guimba, high losses were associated with the combination of stemborer damage and drought (Litsinger et al. 2006c), while Savary et al. (1994) found the combination of stemborer damage and weediness to be associated with low yields. RWR inherently suffers from more stresses than irrigated rice such as drought and submergence. Another reason is that RWR is grown only during the wet season where rain is intrinsically linked to cloudy weather that retards solar radiation slowing photosynthesis. Kenmore et al. (1984) noted hopperburn from brown planthopper damage occurred more on cloudy days. In addition nutrient stress is common. RWR farmers in the Philippines use about half the fertiliser dosage of irrigated rice farmers due to increased risk of crop failure (Barker and Herdt 1979; Pineda et al. 1984). Fluctuating paddy water levels, typical in RWR environments, result in greater losses of applied N (Yoshida 1981). RWR has relatively high losses during the vegetative stage which in Pangasinan and Iloilo we can attribute to the combined damage from whorl maggot injury and caseworm defoliation plus physiological stress from a number of causes that may differ each crop. Aside from those already mentioned we can add transplanting shock (Jahn et al. 2007). Both whorl maggot and caseworm attack the newly transplanted crop at a time when seedling root systems are recovering from losing root hairs and rootlets. Often, seedlings older than 30 days are transplanted and as transplanters are in a hurry, when the seedlings are jammed into the soil, the long roots invert into a Ushape, being turned upwards, a process which not only takes several weeks for the plants to recover but also limits tillering that lowers yield potential. We hypothesise that the pest abundance 6 crop stress interaction is the reason for the often high vegetative losses reported in this study. Losses in the other two growth stages are probably the result of stemborer damage which was high in many crops, and again compensation would be compromised from any coterminous stresses. From these results we can make some generalisations regarding the characterisation of losses based on the dynamic relationship between the intensity of stresses present that pull yields down versus the crop’s ability to compensate from them. Of the six crops in Table 6 with yield losses 420%, two were dryland and one was irrigated. One can note that these crops were not necessarily those with the highest insect pest pressure. Siniloan and Claveria dryland sites were characterised with highly acidic and eroded soils and

The results showed that, even when applied at lethal dosages, insecticide usage was unprofitable. In fact, use of action thresholds is only marginally profitable in higher-yielding, irrigated rice (Smith et al. 1988; Litsinger et al. 2005). In RWR, depending on the crop, about half of the losses came in the vegetative stage (Table 6), while in irrigated rice, losses occur more or less equally in all three growth stages that would require more costly multiple insecticide applications to prevent (Litsinger et al. 2005). However, RWR presents a greater challenge than irrigated, double-crop rice in terms of making chemical control profitable even if efforts are concentrated to one stage due to: (1) RWR yield potential is significantly lower than in irrigated rice, thus there are lower potential economic benefits from the same insecticide usage. (2) Nonsystemic granular insecticides are less suited to RWR due to the erratic ponding depths (Bandong and Litsinger 1979). (3) Rainfall can occur at any time to reduce insecticide residual life, whereas in the dry season irrigated crop is essentially rainfall free.

International Journal of Pest Management (4) Farmers using lever-operated, knapsack sprayers do not apply adequate spray volumes and severely underdose. A previous study showed that Filipino farmers have adopted insecticides because they feel that they are needed in the same sense that fertiliser is needed to obtain high yields (Kenmore et al. 1985). Farmers overestimate the threat posed by insect pests thus are fearful of epidemics that they have seen or experienced (Escalada and Kenmore 1988; Heong 1998). These perceptions need to be overcome given that insecticide usage in RWR is uneconomical. 4.4. Integrated pest management (IPM)

239

A greater proportion of RWR crops registered highest pest densities than crops in either dryland or irrigated ecosystems (Table 2). We already discussed some of the characteristics of RWR culture that foster greater proliferation of some of the more common chronic insect pests based on adaptive strategies of dormancy, polyphagy, and vagility to overcome the long dry fallow of the Asian monsoon climate (Litsinger et al. 1987a). Many pests exhibit more than one strategy. Based on comparative abundance, pests show better adaptation than most of their natural enemies. Among the least adapted pests were brown planthopper and green leafhoppers [except when juxtaposed to irrigated rice] that achieved only minor pest status as compared to their acute pest status in irrigated multicrop rice. Tungro was long known before the advent of modern rices, but it was irrigated rice not RWR that suffered more during the hopper/virus epidemics of the 1970s and 1980s. Such epidemics were due to the combined shrinkage of the rice-free fallow allowing pest populations to carry over from crop to crop (Loevinsohn et al. 1988) combined with increased use of insecticides that killed off natural enemies leading to insecticide resurgence, secondary pest outbreaks, and a rapid breakdown of resistant varieties (Gallagher et al. 1994). The brown planthopper and tungro/green leafhopper flare-ups in RWR in Iloilo were probably due to its close proximity to irrigated areas that exhibited these causes rather than from indigenous causes. Based on the analysis of the RWR ecosystem, a control strategy for double-cropped, irrigated rice was developed to create a 2-month, pest suppressive, ricefree period between the dry and wet season crops and to discourage planting more than two rice crops per year (Litsinger 2008). The three key pests of RWR in the Philippines were whorl maggot, caseworm, and Scirpophaga stemborers. Based on the results of this study, an IPM strategy for favourable RWR areas would be to increase the compensatory ability of rice by selecting high-tillering, medium-maturing varieties and undertaking good agronomic management to reduce stresses and minimise the use of insecticides to respond only to crop-

threatening attacks. A different strategy is needed for unfavourable areas where varieties are needed that share more characteristics with traditional cultivars in order to best tolerate the extreme stresses that define this environment. One of the reasons why traditional rices can provide a reasonable yield even under severe stress is their longer maturity which allows more time for yield compensation to occur. However, as traditional rices lodge when inorganic nitrogen is added, their response to better management is limited. Traditional rices are tall thus are better adapted to perennially flooded rice fields where shorter varieties more readily succumb after submergence for a few days. But longer maturity also means more insect pest generations and higher pest damage. Now breeders are improving RWR rices in unfavourable environments using taller and longer maturing cultivars (Xu 2006). Bt-rice (Cohen et al. 2000) should increase yield potential in both favourable and unfavourable environments against stemborers and caseworm. Stemborers in particular damage the crop over several growth stages and the type of damage is least able to be compensated in low tillering cultivars by agronomic management. The benefits of Bt-rice in unfavourable environments should be higher than recovering just the loss caused directly by stemborers due to synergistic compensatory effects (Litsinger et al. 2006c). Acknowledgements
We are indebted to the field work performed by our site staff in Oton Iloilo, Manaoag Pangasinan, and Solana Cagayan without whose help this research would not have been possible. We are grateful for the assistance of Nonnie Bunyi and Josie Lynn Catindig at IRRI in providing references and technical information. In addition the assistance provided by three anonymous reviewers is gratefully appreciated.

References
Alam MZ. 1967. Insect pests of rice in East Pakistan. In: The major insect pests of rice. Baltimore (MD): Johns Hopkins Press. p. 643–655. Arida GS, Shepard BM. 1990. Parasitism and predation of rice leaffolders, Marasmia patnalis (Bradley) and Cna´ phalocrocis medinalis (Guenee) (Lepidoptera: Pyralidae) in Laguna Province, Philippines. J Agricultural Entomol. 7:113–118. Bandong JP, Canapi BL, dela Cruz CG, Litsinger JA. 2002. Insecticide decision protocols: a case study of untrained Filipino rice farmers. Crop Protect. 21:803–816. Bandong JP, Litsinger JA. 1979. Evaluation of granular insecticides for rainfed wetland rice in the Philippines. Int Rice Res Newslett. 4(2):15. Bandong JP, Litsinger JA. 2005. Rice crop stage susceptibility to the rice yellow stemborer Scirpophaga incertulas (Walker) (Lepidoptera: Pyralidae). Int J Pest Manag. 51:37–43. Barker R, Herdt RW. 1979. Rainfed lowland rice as a research priority – an economist’s view. p. 3–35. In: Rainfed lowland rice: selected Papers from the 1978 International Rice Research Conference. Los Banos ˜ (Philippines): IRRI. p. 341.

240

J.A. Litsinger et al.
IRRI (International Rice Research Institute). 1976. Lowland rice cropping systems, Cropping Systems Program. p. 375–388. In: Annual Report for 1975. Los Banos ˜ (Philippines): IRRI. p. 479. IRRI. 1977. Environmental description. Cropping Systems Program. p. 311–322. In: Annual Report for 1976. Los Banos (Philippines): IRRI. p. 418. ˜ IRRI. 1978. Control and management of insects. p. 215– 216. In: Annual Report for 1977. Los Banos (Philip˜ pines): IRRI. IRRI. 1979. Control and management of insects. p. 193–196. In: Annual Report for 1978. Los Banos (Philippines): IRRI. ˜ IRRI. 1980a. Environmental description. Cropping Systems Program. p. 379–395. In: Annual Report for 1979. Los Banos (Philippines): IRRI. p. 538. ˜ IRRI. 1980b. Control and management of insects. p. 221– 222. In: Annual Report for 1979. Los Banos (Philip˜ pines): IRRI. p. 538. IRRI. 1981a. Environmental description. Cropping Systems Program. p. 221–222. In: Annual Report for 1980. Los Banos (Philippines): IRRI. p. 467. ˜ IRRI. 1981b. Control and management of insects. p. 197– 198. In: Annual Report for 1980. Los Banos (Philip˜ pines): IRRI. p. 467. IRRI. 1983. Control and management of insects. p. 192– 193. In: Annual Report for 1982. Los Banos (Philip˜ pines): IRRI. IRRI. 1985a. International rice research: 25 years of partnership. Los Banos (Philippines): IRRI. p. 188. ˜ IRRI. 1985b. Insecticide evaluation for. 1984. Entomology Department. Los Banos (Philippines): IRRI. p. 232. ˜ Islam Z. 1993. Diapause in rice yellow stem borer, Scirpophaga incertulas (Walker) (Lepidoptera: Pyralidae), in Bangladesh. J Plant Protect Tropics. 10:163– 175. Ishii-Eiteman MJ, Power AG. 1997. Response of green rice leafhoppers to rice-planting practices in Northern Thailand. Ecol Appli. 7:194–208. Jahn GC, Litsinger JA, Chen Y, Barrion AT. 2007. Integrated pest management of rice: ecological concepts. p. 315–366. In: Koul O, Cuperus GW, editors. Ecologically based integrated pest management. Wallingford (UK): CAB International. p. 462. Kalode MB, Karim ANMR, Pongprasert S, Heinrichs EA. 1986. Varietal improvement and resistance to insect pests. p. 241–252. In: Progress in rainfed lowland rice. Los Banos (Philippines): IRRI. p. 446. ˜ Kenmore PE, Carino FO, Perez CA, Dyck VA, Gutierrez ˜ AP. 1984. Population regulation of the rice brown ˚ planthopper (Nilaparvata lugens Stal) within rice fields in the Philippines. J Plant Protect Tropics. 1:19–37. Kenmore PE, Heong KL, Putter CA. 1985. Political, social and perceptual aspects of integrated pest management programmes. In: Lee BS, Loke WH, Heong KL, editors. Proceedings of a Seminar on Integrated Pest Management in Malaysia. p. 47–67. Malaysian Plant Protection Society, Kuala Lumpur. Kiritani K. 1992. Prospects for integrated pest management in rice cultivation. Jap Agri Res Quart. 26:81–87. Litsinger JA. 1979. Major insect pests of rainfed wetland rice in tropical Asia. Int Rice Res Newslett. 4(2):14–15. Litsinger JA. 1991. Crop loss assessment in rice. p. 1–65. In: Heinrichs EA, Miller TA, editors. Rice insects: management strategies. New York: Springer-Verlag. Litsinger JA. 1993. A farming systems approach to insect pest management for upland and lowland rice farmers in tropical Asia. p. 45–101. In: Altieri MA, editor. Crop protection strategies for subsistence farmers. Westview studies in insect biology. Boulder (CO): Westview Press. p. 197.

Barrion AT, Litsinger JA. 1985. A method for holding eggs of rice insect pests for parasite emergence. Int Rice Res Newslett. 10(2):19. Barrion AT, Litsinger JA, Medina EB, Aguda RM, Bandong JP, Pantua PC, Jr, Viajante VD, dela Cruz CG, Vega CR, Soriano JS, Jr, et al. 1991. The rice Cnaphalocrocis and Marasmia (Lepidoptera: Pyralidae) leaffolder complex in the Philippines: taxonomy, bionomics and control. Philippine Entomol. 8:987–1074. Boling A, Tuong TP, Jatmiko SY, Burac MA. 2004. Yield constraints of rainfed lowland rice in Central Java, Indonesia. Field Crops Res. 90:351–360. Catling D. 1992. Rice in deep water. Los Banos (Philippines): ˜ IRRI. p. 542. Cendana SM, Calora FB. 1967. Insect pests of rice in the ˜ Philippines. p. 591–613. In: The major insect pests of the rice plant. Baltimore (MD): Johns Hopkins Press. p. 729. Chandra G. 1978. A new cage for rearing hopper parasites. Int Rice Res Newslett. 3(1):12. Cohen MB, Gould G, Bentur JS. 2000. Bt rice: practical steps to sustainable use. Int Rice Res Notes. 25:4–10. de Kraker J, Rabbinge R, van Huis A, van Lenteren JC, Heong KL. 2000. Impact of nitrogenous-fertilization on the population dynamics and natural control of rice leaffolders (Lep.: Pyralidae). Int J Pest Manag. 46:225–235. de Kraker J, van Huis A, Heong KL, van Lenteren JC, Rabbinge R. 1999. Population dynamics of rice leaffolders (Lepidoptera: Pyralidae) and their natural enemies in irrigated rice in the Philippines. Bull Entomol Res. 89:411–421. DeDatta SK. 1981. Principles and practices of rice production. New York: John Wiley and Sons. p. 618. dela Cruz CG, Litsinger JA, Paragna F. 1981. Tillage implements for soil incorporation of carbofuran granules in rainfed wetland fields. Int Rice Res Newslett. 6(1):17. Escalada MM, Kenmore PE. 1988. Communicating integrated pest control to rice farmers at the village level. p. 221–228. In: Teng PS, Heong KL, editors. Pesticide Management and Integrated Pest Management in Southeast Asia, College Park, MD, US: Consortium for International Crop Protection. p. 664. Ferino M. 1968. The biology and control of the rice leafwhorl maggot, Hydrellia philippina Ferino (Ephydridae, Diptera). Philippine Agri. 52:332–383. Fujisaka S. 1990. Rainfed lowland rice: building research on farmer practice and technical knowledge. Agri Ecosys Environ. 33:57–74. Gallagher KD, Kenmore PE, Sogawa K. 1994. Judicial use of insecticides deter planthopper outbreaks and extend the role of resistant varieties in Southeast Asian rice. p. 599– 614. In: Denno RF, Perfect TJ, editors. Planthoppers their ecology and management. London: Chapman and Hall. p. 772. Garrity DP, Oldeman LR, Morris RA. 1986. Rainfed lowland rice ecosystems: characterization and distribution. p. 3–23. Progress in rainfed lowland rice. Los Banos ˜ (Philippines): IRRI. p. 446. Heinrichs EA, Katanyukul W, Karim ANMR, Misra BC. 1986. Management of insect pests in rainfed lowland rice. p. 349–358. In: Progress in rainfed lowland rice. Los Banos (Philippines): IRRI. p. 446. ˜ Heong KL. 1990. Feeding rates of the rice leaf-folder, Cnaphalocrocis medinalis (Lepidoptera: Pyralidae) on different plant stages. J Agri Entomol. 7:81–90. Heong KL. 1998. IPM in developing countries: progress and constraints in rice IPM, p. 68–77. In: Zalucki M, Drew R, White G, editors. Proceedings Sixth Australasian Applied Entomological Research Conference, University of Queensland, Brisbane, Australia.

International Journal of Pest Management
Litsinger JA. 1994. Cultural, mechanical, and physical control of rice insects. p. 549–584. In: Heinrichs EA, editor. Biology and management of rice insects. New Delhi: Wiley Eastern Ltd. p. 779. Litsinger JA. 2008. Areawide rice insect pest management: a perspective of experiences in Asia. p. 351–440. In: Koul O, Cuperus G, Elliott N, editors. Areawide pest management: theory and implementation. Wallingford (UK): CABI. p. 590. Litsinger JA. 2009. When is a rice insect a pest yield loss and the Green Revolution. p. 391–498. In: Peshin R, Dhawan AK, editors. Integrated pest management: innovationdevelopment process. Berlin: Springer Science þ Media B.V., Vol 1, p. 670. Litsinger JA, Alviola AL, Canapi BL. 1986. Effects of flooding on insect pests and spiders in a rainfed rice environment. Int Rice Res Newslett. 11(5):25–26. Litsinger JA, Alviola AL, dela Cruz CG, Canapi BL, Batay-an EH, III, Barrion AT. 2006a. Rice white stemborer Scirpophaga innotata (Walker) in southern Mindanao, Philippines. I. Supplantation of yellow stemborer S. incertulas (Walker) and pest status. Int J Pest Manag. 52:11–21. Litsinger JA, Alviola AL, dela Cruz CG, Canapi BL, Batay-an EH, III, Barrion AT. 2006b. Rice white stemborer Scirpophaga innotata (Walker) in southern Mindanao, Philippines. II. Synchrony of planting and natural enemies. Int J Pest Manag. 52:23–27. Litsinger JA, Bandong JP. 1992. Response of the rice ´ caseworm Nymphula depunctalis (Guenee) to insecticides. J Plant Protect Tropics. 9:169–177. Litsinger JA, Bandong JP, Chantaraprapha N. 1994a. Mass rearing, larval behaviour, and effects of plant age on the ´ rice caseworm Nymphula depunctalis (Guenee) (Lepidoptera: Pyralidae). Crop Protect. 13:494–502. Litsinger JA, Bandong JP, Canapi BL, dela Cruz CG, Pantua PC, Alviola AL, III, Batay-an E. 2005. Evaluation of action thresholds against chronic insect pests of rice in the Philippines: I. Less frequently occurring pests and overall assessment. Int J Pest Manag. 51:45–61. Litsinger JA, Bandong JP, Canapi BL, dela Cruz CG, Pantua PC, Alviola AL, III, Batay-an E. 2006c. Evaluation of action thresholds against chronic insect pests of rice in the Philippines: IV. Stemborers. Int J Pest Manag. 52:194–207. Litsinger JA, Barrion AT, Bumroongsri V, Morrill WL, Sarnthoy O. 1997. Natural enemies of the rice greenhorned caterpillar Melanitis leda ismene Cramer (Lepidoptera: Satyridae) and rice skipper Pelopidas mathias Fabricius (Lepidoptera: Hesperiidae) in the Philippines. Philippine Entomol. 11:151–181. Litsinger JA, Barrion AT, Dandi Soekarna. 1987a. Upland rice insect pests: their ecology, importance, and control. IRRI Research Paper Series No. 123:1–41 p. Litsinger JA, Bumroongsri V, Morrill WL, Sarnthoy O. 1993. Rearing, ovipositional biology, and plant host range of the rice greenhorned caterpillar Melanitis leda ismene Cramer (Lepidoptera: Satyridae). J Plant Protect Tropics. 10:205–218. Litsinger JA, Bumroongsri V, Morrill WL, Sarnthoy O. 1994b. Mass rearing, developmental biology, and host plant range of the rice skipper Pelopidas mathias F. (Lepidoptera: Hesperiidae. Insect Sci Appli. 15:9– 17. Litsinger JA, Canapi BL, Alviola AL. 1982. Farmer perception and control of rice pests in Solana, Cagayan Valley, a pre-green revolution area of the Philippines. Philippine Entomologist. 5:373–383.

241

Litsinger JA, Canapi BL, Bandong JP, dela Cruz CG, Apostol RF, Pantua PC, Lumaban MD, Alviola AL, III, Raymundo F, Libetario EM, et al. 1987b. Rice crop loss from insect pests in wetland and dryland environments of Asia with emphasis on the Philippines. Insect Sci Appli. 8:677–692. Litsinger JA, Chantaraprapha N, Bandong JP, Barrion AT. 1994c. Natural enemies of the rice caseworm Nymphula ´ depunctalis (Guenee) (Lepidoptera: Pyralidae). Insect Sci Appli. 15:261–268. Litsinger JA, Entomology Department. 1979. An inexpensive kerosene light trap to monitor rice insects. Int Rice Res Newslett. 4(2):17. Litsinger JA, Gyawali BK, Wilde GE. 1998. Feeding behaviour of the rice bug Leptocorisa oratorius (F.) (Hemiptera: Alydidae). J Plant Protect Tropics. 11:23– 35. Litsinger JA, Libetario EM, Barrion AT, Apostol RP. 2009. Insect pest complex, yield loss, and response to improved management of dryland rice in the Philippines. Int J Pest Manag. 55:129–149. Litsinger JA, Lumaban MD, Bandong JP, Pantua PC, Barrion AT, Apostol RF, Ruhendi. 1980a. A methodology for determining insect control recommendations. IRRI Research Paper Series No. 46:1–31. Litsinger JA, Price EC, Herrera RT. 1980b. Small farmer pest control practices for rainfed rice, corn, and grain legumes in three Philippine provinces. Philippine Entomol. 4:65– 86. Loevinsohn ME, Litsinger JA, Heinrichs EA. 1988. Rice insect pests and agricultural change. p. 161–182. In: Harris MK, Rogers CE, editors. The entomology of indigenous and naturalized systems in agriculture. Boulder (CO): Westview Press. Mackill DJ. 1986. Varietal improvement for rainfed low land rice in South and Southeast Asia: results of a survey. p. 115–144. In: Progress in rainfed lowland rice. Los Banos (Philippines): IRRI. p. 446. ˜ Malabuyoc LA. 1977. A new lepidopterous pest of rice. Int Rice Res Newslett. 2(2):6. Mochida OM, Joshi RC, Litsinger JA. 1987. Climatic factors affecting the occurrence of insect pests. p. 149–164. In: Weather and rice. Los Banos (Philippines): IRRI. p. 323. ˜ Morris RA, Patanothai A, Syarifuddin A, Carangal VR. 1986. Cropping systems for rainfed lowland rice environments. p. 59–76. In: Progress in rainfed lowland rice. Los Banos (Philippines): IRRI. p. 446. ˜ Pantua PC, Litsinger JA. 1984. Life history and plant host range of the rice green semilooper. Int Rice Res Newslett. 9(1):26. Pena N, Shepard M. 1986. Seasonal incidence of ˜ parasitism of brown planthopper Nilaparvata lugens (Homoptera: Delphacidae), green leafhopper Nephotettix spp. (Homoptera: Cicadellidae), and whitebacked planthoppper Sogatella furcifera (Homoptera: Delphacidae) in Laguna province, Philippines. Environ Entomol. 15:263–267. Pineda R, Duff B, Heinrichs EA, Carbonell P. 1984. Insect control practices on irrigated and rainfed rice farms in Nueva Ecija, Philippines. Working Paper No. 102. The Consequences of Small Rice Farm Mechanization Project, International Rice Research Institute, Los Banos, Philippines. p. 24. ˜ Pinnschmidt HO, Batchelor WD, Teng PS. 1995. Simulation of multiple species pest damage in rice using CERES-rice. Agric Sys. 48:193–222. Poston FL, Pedigo LP, Welch SM. 1983. Economic injury levels: reality and practicality. Bull Entomol Soc Am. 29:49–53.

242

J.A. Litsinger et al.
van den Berg H, Shepard BM, Litsinger JA, Pantua PC. 1988. Impact of predators and parasitoids on the eggs of Rivula atimeta, Naranga aenescens (Lepidoptera: Noctuidae) and Hydrellia philippina (Diptera: Ephydridae) in rice. J Plant Protect Tropics. 5:103–108. van Haltern P. 1979. The insect pest complex and related problems of lowland rice cultivation in South Sulawesi, Indonesia. Mededelingen Landbouwhogeschool, Wageningen, Netherlands. 79(1):1–111. Venugopal MS, Litsinger JA. 1984. Effect of carbofuran on rice growth. Protect Ecol. 7:313–317. Viswanathan K, Kalode MB. 1984. Comparative study on varietal resistance to rice green leafhoppers Nephotettix ˚ virescens (Distant) and N. nigropictus (Stal). Proc Ind Acad Sci (Animal Sci.). 93:55–63. Xu K, Xu X, Fukao P, Canlas R, Maghirang-Rodriguez R, Heuer S, Ismail AM, Bailey-Serres J, Ronald PC, Mackill DJ. 2006. Sub1 A is an ethylene-response-factor-like gene that confers submergence tolerance. Nature. 442:705– 708. Yoshida S. 1981. Fundamentals of rice crop science. Los Banos (Philippines): IRRI. p. 269. ˜ Zandstra HG, Price EC, Litsinger JA, Morris RA. 1981. A methodology for on-farm cropping systems research. Los Banos (Philippines): IRRI. p. 149. ˜

Rothschild GHL. 1970. Parasites of rice stemborers in Sarawak (Malaysian Borneo). Entomophaga. 15:21–51. Rubia EG, Penning de Vries FWT. 1990. Simulation of yield reduction caused by stem borers in rice. J Plant Protect Tropics. 7:87–102. Rubia EG, Shepard BM, Yambao EB, Ingram KT, Arida GS, Penning de Vries F. 1990. Stem borer damage and grain yield of flooded rice. J Plant Protect Tropics. 6:205–211. Savary S, Elazequi FA, Moody K, Litsinger JA, Teng PS. 1994. Characterization of rice cropping practices and multiple pest systems in the Philippines. Agri Sys. 46:385– 408. Shepard BM, Arida GS. 1986. Parasitism and predation of yellow stem borer, Scirpophaga incertulas (Walker) Lepidoptera: Pyralidae) eggs in transplanted and directseeded rice. J Entomol Sci. 21:26–32. Singh BN, Singh R. 1985. Hopper burn and varietal susceptibility of different rice varieties to green leaf hopper under field conditions. Ind J Entomol. 47:220–222. Smith J, Litsinger JA, Bandong JP, Lumaban MD, dela Cruz CG. 1988. Economic thresholds for insecticide application to rice: profitability and risk analysis to Filipino farmers. J Plant Protect Tropics. 6:67–87. Sogawa K. 1976. Rice tungro virus and its vectors in tropical Asia. Rev Plant Protect Res. 9:21–46.