Top Banner

of 22

Dryland Rice Insect Pests

Apr 05, 2018

Download

Documents

James Litsinger
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
  • 8/2/2019 Dryland Rice Insect Pests

    1/22

    PLEASE SCROLL DOWN FOR ARTICLE

    This article was downloaded by: [Litsinger, J. A.]On: 1 May 2009Access details: Access Details: [subscription number 910450850]Publisher Taylor & FrancisInforma Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,37-41 Mortimer Street, London W1T 3JH, UK

    International Journal of Pest ManagementPublication details, including instructions for authors and subscription information:http://www.informaworld.com/smpp/title~content=t713797655

    Comparison of insect pest complexes in different Philippine dryland riceenvironments: population densities, yield loss, and managementJ. A. Litsinger; E. M. Libetario a; A. T. Barrion a; R. P. Apostol ba Philippine Rice Research Institute (PhilRice), Maligaya, Science City of Muoz, Nueva Ecija, Philippines b

    International Rice Research Institute, Metro Manila, Philippines

    Online Publication Date: 01 April 2009

    To cite this Article Litsinger, J. A., Libetario, E. M., Barrion, A. T. and Apostol, R. P.(2009)'Comparison of insect pest complexes indifferent Philippine dryland rice environments: population densities, yield loss, and management',International Journal of PestManagement,55:2,129 149

    To link to this Article: DOI: 10.1080/09670870802604054URL: http://dx.doi.org/10.1080/09670870802604054

    Full terms and conditions of use: http://www.informaworld.com/terms-and-conditions-of-access.pdf

    This article may be used for research, teaching and private study purposes. Any substantial orsystematic reproduction, re-distribution, re-selling, loan or sub-licensing, systematic supply ordistribution in any form to anyone is expressly forbidden.

    The publisher does not give any warranty express or implied or make any representation that the contentswill be complete or accurate or up to date. The accuracy of any instructions, formulae and drug dosesshould be independently verified with primary sources. The publisher shall not be liable for any loss,actions, claims, proceedings, demand or costs or damages whatsoever or howsoever caused arising directlyor indirectly in connection with or arising out of the use of this material.

    http://www.informaworld.com/smpp/title~content=t713797655http://dx.doi.org/10.1080/09670870802604054http://www.informaworld.com/terms-and-conditions-of-access.pdfhttp://www.informaworld.com/terms-and-conditions-of-access.pdfhttp://dx.doi.org/10.1080/09670870802604054http://www.informaworld.com/smpp/title~content=t713797655
  • 8/2/2019 Dryland Rice Insect Pests

    2/22

    Comparison of insect pest complexes in different Philippine dryland rice environments:population densities, yield loss, and management

    J.A. Litsingera*, E.M. Libetariob, A.T. Barrionb and R.P. Apostolc

    a1365 Jacobs Place, Dixon, CA 95620, USA; bPhilippine Rice Research Institute (PhilRice), Maligaya, Science City of Munoz,Nueva Ecija, 3119, Philippines;

    cInternational Rice Research Institute, DAPO Box 7777, Metro Manila, Philippines

    (Received 12 February 2008; final version received 5 November 2008)

    In the Philippines most of the dryland rice pests are distinct from those of wetland culture. Partitioned-growth-stage yield loss studies revealed the highest losses in dryland rice were due to sown-seed and seedling pests (ants,field crickets, mole crickets, and termites) as well as root feeders (white grubs and root aphids) and early seedlingpests (seedling maggot and flea beetle) more than the common foliar wetland pests. Losses (571%) were highestin the sites with the smallest rice area in which pests were concentrated and the poorest soils (which constrain yield

    compensation) along a continuum of dryland rice habitats. Crop management practices such as overseeding andfertility management can mitigate potential losses to a large degree. Therefore integrated crop management playsa central role in integrated pest management in dryland rice cultivation where the use of insecticides should beminimized for economic and environmental reasons.

    Keywords: rice insect pests; yield loss; slash and burn; upland rice; integrated crop management; pestmanagement; crop compensation

    1. Introduction

    Dryland rice is a cultural type distinct from the

    more common wetland culture where direct rainfall

    is the only source of water and japonica-based

    varieties, rather than indica ones, are normallycultivated in non-bunded fields. Dryland rice,

    representing ca. 13% of rice area worldwide (Gupta

    and OToole 1986), is neither puddled nor ponded

    and the soils are aerobic (Morris 1986). The

    Philippines ranks fifth among SE Asian countries

    in dryland rice area (Gupta and OToole 1986); it

    supports 15% of the Filipino population (Garrity

    et al. 1993).

    Arraudeau and Harahap (1986) conducted a

    worldwide survey among dryland rice scientists and

    produced a long list of constraints beginning with

    poor soil, frequent water stress, fungal diseases,

    nematodes, weeds, and lack of adaptive and manage-

    ment responsive varieties. They also listed a wide

    array of insect and vertebrate pests that contribute to

    low yields, but these were considered of lesser

    importance. Grist and Lever (1969) and Fujisaka

    et al. (1991) also mentioned birds, rodents, wild pigs,

    monkeys, squirrels and even elephants and rhino-

    ceroses as dryland rice pests, particularly near forests

    which are habitats for most of these animals. The

    degree to which the aforementioned pest constraints

    are manifested is mainly linked to the type of dryland

    habitat; the latter can vary dramatically in terms of

    slope, soil type, rainfall pattern, and surrounding

    crops and flora.

    Litsinger et al. (1987b) reviewed the world

    literature on dryland rice insect fauna and noted

    life-history strategies based on polyphagy, dormancy,and/or dispersal abilities, as dryland rice is dom-

    inantly a single rice crop system. Key insect pests of

    the wetlands, i.e. brown planthopper Nilaparvata

    lugens (Sta l), green leafhopper Nephotettix virescens

    (Distant), and yellow stemborer Scirpophaga incertu-

    las (Walker), are specific to Oryza spp. and cannot

    sustain themselves in a dryland rice environment. Due

    to the short rice season, they must re-migrate each

    season from nearby wetland sites. Several soil insects

    common in the drylands have life-cycles of up to a

    year or more in duration. More dryland rice pests

    enter periods of dormancy than wetland pests.

    Armyworms, butterflies, and locusts have greater

    dispersal powers than most wetland insect pests, with

    the exception of rice planthoppers and leaffolders

    (Denno et al. 1991).

    In the Asian literature, there have been few trials

    in which yield losses were measured, but all research-

    ers used the insecticide check method. Other yield loss

    studies have been conducted in Latin America and

    Africa where dryland rice is the most important rice

    culture (Litsinger et al. 1987b). In Thailand, Kata-

    nyukul and Chandartat (1981) recorded losses of only

    5% (range 113%) from 1976 to 1979.

    *Corresponding author. Email: [email protected]

    International Journal of Pest Management

    Vol. 55, No. 2, AprilJune 2009, 129149

    ISSN 0967-0874 print/ISSN 1366-5863 online

    2009 Taylor & Francis

    DOI: 10.1080/09670870802604054

    http://www.informaworld.com

  • 8/2/2019 Dryland Rice Insect Pests

    3/22

    Earlier studies in the Philippines documented the

    insect pest fauna and measured losses by the

    insecticide check method in Batangas and Camarines

    Sur provinces (IRRI 1975, 1976a). Those studies were

    undertaken in habitats with favourable soils and

    more permanent agriculture based on thoroughtillage. Losses incurred by traditional tall Dagge

    and Kinanda varieties were very high, averaging

    31 + 17% (2.4 vs. 1.8 t/ha) over nine fields. Studies

    were also carried out in dryland fields on the border

    of the irrigated experimental farm of the Interna-

    tional Rice Research Institute (IRRI), the conclusion

    being that most insects found in the drylands were

    also present in the wetland area (Pathak and Dyck

    1975).

    There is a continuum of dryland rice habitats that

    represents stages in settlement, economic develop-

    ment, and population growth in a locality (Litsinger

    1993). It begins with pioneering farmers settling intorecently logged rainforests; they then cultivate dry-

    land rice by slash and burn or swidden agriculture

    (Morris 1986). As the area becomes more densely

    populated, animal power is used to pull tillage

    implements. Still further population increase brings

    nearby markets and crop diversity increases where

    cash crops are rotated with rice.

    In this report we argue that each of these

    successive stages in land use leads to different pest

    complexes and population intensities. We also expand

    on earlier trials in the Philippines. Multi-year crop

    loss assessments were undertaken in four farmcommunities, each representing a different environ-

    ment along the continuum. These differed from

    previous studies as losses now were partitioned by

    rice growth stage. Also more attention was given to

    recording pest complexes, as researchers lived at the

    site as opposed to making periodic visits, and so

    sampling could be done more frequently. We also

    report on the results of trials which were also

    undertaken to develop practical crop production

    practices centred upon cultural controls as well as

    minimal usage of insecticides following an integrated

    crop management approach.

    2. Materials and methods

    2.1. Site descriptions

    Two research sites (Claveria and Tupi) were estab-

    lished in Mindanao Island under the influence of the

    Inter-tropical Convergence Zone climate, and a

    further two were established in Luzon Island (Sinilo-

    an and Tanauan) with a monsoon climate. Together,

    all sites represented an environmental and crop

    cultural continuum from subsistence, low input, slash

    and burn systems on sloping, acidic, and eroded soils

    sown with traditional varieties, to high input,diversified agriculture on flat volcanic soil sown

    with modern rice and cash crops.

    2.1.1. Siniloan

    This slash-and-burn site was established in Magsay-

    say village located mid-way between Siniloan and

    Real municipalities in the Sierra Madre mountains of

    the Laguna and Quezon provincial border in highlyeroded, acidic soils (Gonzaga et al. 1986). Traditional

    rice is sown in holes made with a dibble stick in July

    among smouldering tree stumps. Rice fields are small

    (ca. 0.1 ha) due to the high labour demands, as

    perennial grasses quickly invade requiring tedious

    hand-weeding. Farmers do not use either inorganic or

    organic fertiliser.

    2.1.2. Claveria

    Claveria in Misamis Oriental province, near Cagayan

    de Oro City, is located on an escarpment along the

    lower slopes of Mt Balatukan volcano. This highlyeroded and acidic soil site is representative of a

    recently deforested stage; the forest has been replaced

    by perennial grasslands, leaving only scattered trees

    in a relatively steep terrain. The farmers priority is to

    first plant maize in fields prepared by animal-drawn

    implements including the mouldboard plough. Rice is

    sown in rows in May or June made with a furrow

    opener which also serves for inter-row cultivation of

    weeds in the early growth stages. There is more

    uncultivated than cultivated land in the site. Farmers

    have their own tall, traditional japonica type varieties,

    which they cultivate with few purchased inputs.Further site description can be found in Litsinger

    et al. (2002).

    2.1.3. Tupi

    Tupi is a town in South Cotabato province, lying on

    favourable, young, and only slightly acidic volcanic

    soils. This third stage is more populated and on flat

    terrain with few uncultivated areas (IRRI 1990).

    Farmers typically apply 30 kg/ha of N and P;

    uncharacteristically for dryland rice sites, farmers

    grow a mix of modern semi-dwarf varieties along with

    high value traditional types.

    2.1.4. Tanauan

    The principal crops in Cale village, Tanauan town in

    Batangas province are maize and dryland rice as well

    as a wide array of vegetables for the nearby Manila

    market. The nearby Taal volcano erupts regularly to

    spread new ash over the landscape. This fourth stage

    represents a highly favourable soil in a populated area

    where farmers engage in the growing of high-input,

    diversified croppings where fields are intensively tilled

    (IRRI 1976b). The topography is gently sloping,and farmers prevent the slight erosion with living

    fencerows. Farmers are economically well off from

    130 J.A. Litsinger et al.

  • 8/2/2019 Dryland Rice Insect Pests

    4/22

    the sale of vegetables, and can afford an average of

    110 kg N/ha on rice where lodging can occur during

    ripening stage during typhoons. Due to its preferred

    taste and grain type, farmers prefer to grow a

    traditional variety rather than to purchase rice in

    the market.

    2.2. Rice crop management

    UPLRi5 dryland rice variety bred by the University

    of the Philippines was used in all sites except Siniloan.

    This semi-dwarf is a cross between indica and japonica

    types, and is high tillering and of medium maturity

    (130140 d). The traditional rice Benernal in Siniloan

    was grown at a seeding rate of 50 kg/ha whereas in

    other sites the rate was 100 kg/ha. Siniloan farmers

    placed three to four seeds per dibble hole 2025 cm

    apart. Each hole was covered with soil and stepped

    on. All trials were superimposed on farmers fieldsunder their own agronomic management. In addition

    to yield loss treatments, the trials included some

    improved agronomic practices in the experimental

    design. In all but the slash and burn site, land was

    tilled with animal-drawn ploughs. Furrows were

    made with a wooden implement (lithao), and seed

    was broadcast and raked into the furrows with a

    spike-toothed, box-harrow and covered by a final

    levelling. In Tanauan 90 kg N/ha was applied in two

    equal splits as a side dressing during hilling up and at

    panicle initiation. Prior to seeding in Claveria and

    Tupi, P was placed in the open rows at 25 kg/ha andcovered. A side dressing of 25 kg N/ha occurred at

    14 d after crop emergence (DE) before the first inter-

    row cultivation followed up by a second top dressing

    of 25 kg N/ha before panicle initiation. Farmers

    hand-weeded as needed. In Tanauan and Tupi

    farmers undertook dry season ploughing for weed

    control, in the latter site a number of times. No

    cooperating farmer applied any pesticide to dryland

    rice crops. Grain yield was taken from five samples of

    5 m2 cuts taken in a stratified grid within the 100-m2

    plots. The grain was dried to 14% moisture.

    2.3. Treatment descriptions

    2.3.1. Partitioned-growth-stage yield loss trials

    The insecticide check, partitioned yield loss method

    was conducted following Litsinger et al. (1980). There

    were eight treatments in Claveria (seven seasons) and

    Tupi (four seasons) to quantify losses in six growth

    stages. The first treatment was termed the complete

    control where the objective of a regimen of insecticide

    applications was to suppress all groups of insect pests.

    Sown seeds were protected with a systemic insecticide

    (6 g a.i. carbosulfan STD/kg seed) from ants, field

    crickets, and mole crickets as well as seedling maggotand flea beetles that attack young seedlings. Carbo-

    sulfan was added to cassava flour then mixed into a

    paste. Seeds were then added to be coated with the

    mixture, and finally they were dusted with cassava

    flour for easier handling after drying overnight. Next,

    a soil insecticide (1 kg a.i. Lindane (gBHC) granules/

    ha) was placed in the seed furrows directed against

    root-feeding white grubs and termites, and a secondapplication was side-dressed at hilling up. Litsinger

    et al. (1983) showed that this insecticide had high

    efficacy against white grubs. Reproductive stage soil

    protection was 0.75 kg a.i. carbofuran granules/ha

    placed in shallow furrows 40 and 70 DE for root-

    feeding aphids and mealybugs. The reproductive stage

    was also protected by an additional treatment of

    weekly foliar sprays (0.4 kg a.i. monocrotophos/ha)

    for stemborers, plant- and leafhoppers, and leaffolders.

    Ripening stage protection against rice seed bugs was

    10 g a.i. deltamethrin EC/ha sprayed twice weekly

    from panicle emergence to hard dough. The second

    treatment was an untreated check. Foliar insecticideswere applied in 19 l, lever-operated, knapsack

    sprayers fitted with cone nozzles. Spray volume

    increased from 300 to 500 l/ha as the crop grew.

    Each of a succession of treatments eliminated one

    of the complete control treatments in order to

    partition the loss among growth stages. Rice growth

    stages were described by Yoshida (1981). The third

    treatment eliminated the seed treatment, the fourth

    eliminated the soil applications of gBHC granular

    insecticide, the fifth eliminated the reproductive stage

    carbofuran granules, the sixth eliminated the repro-

    ductive stage sprays, and the seventh treatmenteliminated ripening stage sprays.

    In Tanauan, yield loss was calculated in the tall

    traditional Dagge rice over five seasons in separate

    experiments and from the improved UPLRi5 rice in

    experiments over three seasons. The partitioned yield

    loss treatments consisted of a complete control which

    protected five growth stages. The first stage was the

    sown-seed where 4 g a.i. bendiocarb/kg seed made as

    a water slurry without the cassava flour. A soil

    treatment consisted of diazinon granules at 1 kg ai/ha

    applied in the seed furrows. A third was vegetative

    foliar protection with 1 kg a.i. monocrotophos/ha

    sprays at 10-day intervals (25, 35, 45 DE) and the

    follow on to protect the reproductive stage (55, 65, 75

    DE), and then by 1 kg a.i. gBHC EC/ha at milk, soft,

    and hard dough stages.

    The complete control treatment in Siniloan (two

    seasons) included increasing the dosage of the seed

    treatment to 0.4 kg a.i. carbosulfan/ha. This was

    followed by weekly sprays of foliar insecticides:

    0.75 kg a.i. monocrotophos/ha during the vegetative

    and ripening stages and the mixture of 0.75 kg a.i.

    chlorpyrifos BPMC/ha from panicle initiation to

    heading. Only the total yield loss was determined in

    Siniloan where the complete control was compared tothe untreated check, as no growth stage partitioning

    was carried out to keep the trials small.

    International Journal of Pest Management 131

  • 8/2/2019 Dryland Rice Insect Pests

    5/22

    Total yield loss was calculated as the difference

    between the complete control treatment and the

    untreated check. To calculate percentage, total loss

    was divided by the full protection yield and multiplied

    by 100. Yield losses from each of the partitioned-

    growth-stage treatments where insecticide wasomitted were calculated in the same way and then

    added. As the two totals were seldom mathematically

    equal, losses for each pest group or crop stage were

    apportioned based on the total loss from the complete

    control.

    2.3.2. Crop management treatments

    Plots of integrated crop management practices were

    randomised among the yield loss plots in the on-farm

    trials. Their purpose was to generate recommenda-

    tions for farmers, and their inclusion each season

    allowed an iterative process of technology develop-ment for each site. These treatments included varia-

    tions in seeding rate, inorganic and organic fertiliser,

    and insecticide protection either as soil and/or seed

    treatments.

    Six potential recommended practices were com-

    pared in Siniloan: (1) was a high input practice to

    determine the yield potential that involved organic

    fertiliser plus lime mixed into the soil in each dibble

    hole to raise soil pH thus making applied P more

    available in the highly acidic soils. Insect control was

    an insecticide seed treatment (0.25 kg a.i. carbosulfan

    ST/kg seed). The seeds were slightly moistened toallow the seed treatment (ST) formulation to stick to

    the seed coat. Foliar insecticide protection was

    0.75 kg a.i. monocrotophos EC/ha weekly during

    the vegetative stage to flowering, and 0.75 kg a.i.

    Brodan EC (BPMC chlorpyrifos)/ha during milk,

    soft dough, and hard dough. (2) The same as

    (1) except it lacked the seed protection. (3) This

    included only the insecticide seed treatment plus ash

    placed in the seed holes plus 23 kg N/ha from urea

    applied on the soil surface 14 DE. (4) This consisted

    only of the ash and 23 kg N/ha. (5) This received the

    seed treatment plus the ash. (6) This consisted only

    the seed treatment. (7) The untreated check (i.e.

    control).

    In Tupi and Claveria similar treatments were

    compared which involved two seeding rates (50 and

    90 kg/ha) and a soil insecticide treatment 0.25 kg

    a.i. Lindane G/ha and an insecticide seed treatment

    (0.3 kg a.i. carbosulfan ST/ha). Two fertiliser

    treatments also were compared: the first lacked

    inorganic fertiliser; the second was 50-25-0 with the

    N application split at 14-25 DE during inter-row

    cultivation and before panicle initiation, and P was

    applied basally during land preparation. The inclu-

    sion of these treatments varied from season toseason to enable us to find the best performing

    treatment.

    In Tanauan a crop management trial was con-

    ducted in 1976 with 12 treatments replicated in six

    farmers fields. Four of the treatments involved

    incremental increases in the seeding rate from 50 to

    125 kg/ha. Four other treatments were conducted

    at 50 kg seed/ha that tested two seed and twosoil treatments. The seed treatments tested two

    chemicals carbofuran ST and dieldrin WP both

    at 0.5 kg ai/ha dosages. The two soil treatments were

    carbofuran G at 0.5 and 2 kg ai/ha dosages. Two

    other treatments tested the seed treatments at a higher

    seeding rate of 100 kg/ha while the last two treat-

    ments tested the soil insecticide dosages, again at the

    100 kg seed/ha. Data taken were plant stand (ten 3-m

    row samples) and yield (two 10 m2 cuts).

    A follow-up trial was conducted in four farmers

    fields over 3 years (19781980) with UPLRi5 sown at

    100 kg/ha where a seed (0.30 kg a.i. carbosulfan ST/

    ha) and a soil (0.75 kg a.i. diazinon G/ha) treatmentwere compared to an untreated check. Crop manage-

    ment treatments were analyzed economically to

    determine marginal returns and benefit:cost ratio

    (B:C) following the method of Smith et al. (1988)

    using 1986 prices for agricultural inputs and including

    a 12% interest per annum for inputs and labour

    priced at $0.10h, the wage in rural Philippines at the

    time. Unmilled rice was valued at $0.128/kg. A B:C

    ratio of 42 is considered a favourable return on

    investment.

    2.4. Experimental design

    Field trials were conformed to a randomised complete

    block design with farmers as replicates. New farmer

    cooperators were selected each season, as the results

    would become recommendations for the farm com-

    munity as a whole. Farmer cooperators were selected

    on a staggered basis evenly along each seasons

    planting curve. Plot sizes were 100 m2 with the

    exception of the slash and burn site, where they

    were 10 m2, the largest obtainable size of uniform

    field conditions free of tree stumps.

    2.5. Arthropod and crop sampling

    Identification of arthropods was undertaken by one

    of us (ATB). Voucher specimens were deposited in

    the IRRI Entomological Museum. On each sampling

    date a team of at least two staff participated in data

    collection. The stratified grid method was used to

    sample plants or arthropods from each quadrant of

    all plots. Data collection sites were randomly selected

    within each quadrant. One person recorded the data

    on a clipboard while others did the counting with

    mechanical tally counters. Plant stand was recorded

    weekly from the first through the fourth weeks bycounting the number of plants along the row

    measured in 14-m sections. Insect damage was also

    132 J.A. Litsinger et al.

  • 8/2/2019 Dryland Rice Insect Pests

    6/22

    assessed at the same time. Seedling maggot was

    censused by damaged tillers (deadhearts) and mole

    cricket by damaged plants. Root pests such as

    termites, root aphids, and mealybugs were recorded

    in 50 plant samples with each plant being uprooted

    and the root system inspected. Stemborer deadheartsand whiteheads as well as leaffolder damaged leaves

    were censused 50, 70, 90, and 110 DE from 10

    samples of 5-m rows. In these cases the number of

    tillers and leaves was counted in each 5-m row. Plant-

    and leafhoppers were sampled in the crop in Tanauan

    using a motorized D-VAC1 suction machine that was

    swung side to side a distance of 1 m and walked a

    measured 25 m to give a sample size of 25 m2.

    Sampling was done weekly in four fields from early

    vegetative stage to 2 weeks before harvest. Rice seed

    bugs were sampled by taking 25 pendulum sweeps

    with a 38-cm diameter insect net at milk, soft, and

    hard dough stages. Damaged grains were determinedvia the acid fuchsin staining technique (Litsinger et al.

    1981).

    2.6. Light trap collections

    Pairs of kerosene metal light traps were used, made

    by Laguna fishermen with the lamps enclosed in a

    glass housing (Litsinger et al. 1979). Farmers were

    trained to manage the light traps by depositing each

    nights catch in vials of 70% alcohol. Trained staff did

    the identification and counting with the aid of a

    stereomicroscope. Light traps were placed two pervillage in rice fields out of sight from one another and

    unobstructed by trees within a radius of 100 m.

    Enough kerosene was used to enable the lamps to

    burn throughout the night. The height of the flame

    was standardized across sites by wick length. Collec-

    tions in the four dryland rice sites were compared to

    three rainfed wetland and three irrigated sites where

    similar research teams worked. Two irrigated sites

    were separated into areas that represented the

    standard synchronized double-rice cropping and a

    more intensive cropping (2.5 crops in 5 years) planted

    asynchronously (Loevinsohn et al. 1988). Counts

    were summed over six months that represented the

    mean period of a seasonal rice crop. This was done so

    that rainfed single crop areas could be compared to

    irrigated double crop sites on a per crop basis. In

    irrigated sites data were summarized for each wet and

    dry season crop based on the planting pattern each

    year. Data presented are averages of all of the crops

    during the specified years for each species.

    2.7. Statistical analysis

    All statistical analyses were performed by SAS with

    P 0.05 as the criterion for significance. Resultswere subjected to one-way ANOVA. Treatment

    means were separated using the paired t-test for two

    variables or the 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. Results3.1. Insect pests

    3.1.1. Siniloan

    The most injurious group of insects in the slash and

    burn site were sown-seed and seedling pests which

    reduced plant stand 91% in 1984 and 27% in 1985.

    The first group included ants, dominated by the

    ubiquitous Solenopsis geminata (F.), that removed

    seeds, while field crickets fed on the germinating

    seeds. Foliar pests attacking young seedlings were

    seedling maggot Atherigona oryzae Malloch and a flea

    beetle Chaetocnema basalis (Baly). The rice seedling

    maggot caused 42% deadhearts in 1984 but rose to440% deadhearts in 1985 (Table 1). The 2-mm flea

    beetle adults caused shot hole leaf damage to 46% of

    young seedlings leading to death of many in 35 days.

    Its normal hosts are grasses (Barrion and Litsinger

    1986a).

    As rice was planted in such small areas of burned-

    off forest there were fewer species in this pest complex

    compared to other sites. This was due to both the

    small size of fields and lack of alternative hosts.

    Absent were sown-seed and root pests, such as root

    aphids, mole crickets, white grubs, mealybugs, and

    termites. Foliar pests such as plant- and leafhoppers,stemborers, and defoliators were also missing.

    The crop suffered periods of moisture stress

    manifested by leaf rolling. These conditions encour-

    aged a leaf-feeding thrips Stenchaetothrips biformis

    (Bagnall) to multiply causing stippling by removing

    photosynthetic area, but their numbers were soon

    suppressed by heavy rains (Barrion and Litsinger

    1986b). Two leaffolder species were encountered with

    Marasmia exigua (Butler) being more prevalent than

    Cnaphalocrocis medinalis (Guene e) as the former has

    a wider host range (Barrion et al. 1991). Light trap

    collections recorded leaffolders year-round and they

    peaked during rice harvest in the lowlands 1520 km

    away, showing their high migratory capability.

    Damage was non-economic, however. Rice seed

    bugs Leptocorisa oratorius (F.) and L. acuta (Thun-

    berg) were recorded at relatively high densities (4.1/

    m2). The latter is less associated with rice areas and

    has a wider plant host range. This was the highest

    density recorded among 11 research sites representing

    dryland, rainfed, and irrigated wetland ecosystems

    (Litsinger 2008).

    Kerosene light trap collections showed low to

    moderate densities of brown planthopper and green

    leafhoppers along with one of their major predatorsCyrtorhinus lividipennis Reuter (Table 2). Yellow

    stemborers were also caught in low levels. Despite

    International Journal of Pest Management 133

  • 8/2/2019 Dryland Rice Insect Pests

    7/22

    Table1.Effectofincreasinginputlevelsofnutrients,soilamendments,andinsecticideonyieldofBenernaldrylandricegro

    wninslashandburnculture,Siniloan,Lagunaprovince,

    Philippines,1985.a

    Treatment

    T

    reatmentinputs(dosageperhectare)

    Plantstandd

    (no.plants/30hills)

    14DE

    Leafbeetle

    (%

    damaged

    leaves)14DEe

    Seedlingmaggot

    (%

    deadhearts)

    28DEe

    Yield(t/ha)

    f

    Nutrients(kg/ha)

    Soilamendments

    Insecticidetreatment

    1.High

    input,

    fullp

    rotection

    46N,28P,14K

    500kglime

    3tchicken

    manure/ha

    0.4kga.i.carbosulfan

    ST/ha

    weekly

    foliarspraysc

    206+

    24ab

    8.3+

    4.9a

    17.4+

    4.5a

    0.48+

    0.25a

    2.High

    input,

    foliarprotectiononly

    46N,28P,14K

    500kglime

    3tchicken

    manure/ha

    Weeklyfoliars

    praysc

    142+

    15c

    7.8+

    5.1a

    42.4+

    23.2c

    0.47+

    0.26a

    3.Low

    input,seed

    treatmentonly

    23N

    Ash

    b

    0.25kga.i.carbosulfanST/ha

    221+

    26a

    12.4+

    8.6a

    28.8+

    11.3ab

    0.42+

    0.24a

    4.Low

    input,

    noprotection

    23N

    Ash

    b

    162+

    11c

    38.9+

    13.2b

    30.6+

    16.3b

    0.45+

    0.21a

    5.Low

    input,

    seed

    treatmentonly

    Ash

    b

    0.25kga.i.carbosulfanST/ha

    204+

    27ab

    13.5+

    9.9a

    18.9+

    9.0a

    0.20+

    0.08b

    6.Low

    input,

    seed

    treatmentonly

    0.25kga.i.carbosulfanST/ha

    162+

    16c

    11.2+

    7.2a

    15.7+

    5.4a

    0.21+

    0.07b

    7.Untreated

    149+

    19c

    45.7+

    20.3b

    37.0+

    15.2b

    0.06+

    0.05c

    P

    500001

    500001

    500001

    500001

    F

    5.32

    4.84

    5.92

    4.98

    df

    10

    10

    10

    10

    aAverageoffourreplications.Inacolumn,means(+SEM)followedbyadifferentletterares

    ignificantlydifferent(P

    0.05)byLSDanalysis.DE

    daysaftercropemergence.

    b

    cup

    perdibbleholederivedfromforesttrees.

    c0.75kg

    aimonocrotophosEC/haweeklyduringt

    hevegetativestagetopanicleinitationandduringripening,0.75kgaiBPMC

    chlorpyrifos(BrodanEC)/hafrompanicleinitiationtoflowering.

    ECem

    ulsifiableconcentrate(liquid)formulation.ST

    seedtreatment.

    d30dibb

    leholessampledperreplication.

    e50hills

    sampledperreplication.

    f25-m

    2

    samplesperreplication

    134 J.A. Litsinger et al.

  • 8/2/2019 Dryland Rice Insect Pests

    8/22

    Table2.

    Abundanceofinsectsandvariablesofricecroppingintensitybyricecultureasdeterminedfromkerosenelighttrapssetupin12locationsinthePhilippines,19791992.

    Riceculturea

    Town

    Province

    No.

    crops/yr

    Area

    inrice

    (%)b

    Seasonaltotalperlighttrapperlocationc

    Brown

    planthopper

    N.

    lugens

    Whitebacked

    planthopper

    S.

    furcifera

    Green

    leafhoppers

    Nep

    hotettixspp.

    Zigzag

    leafhopper

    R.

    dorsalis

    White

    leafhopper

    C.spectra

    C.

    lividipennis

    miredpredator

    Scirpophaga

    sp

    p.

    Stem

    borers

    d

    Other

    stemborerse

    Dryland

    Siniloan

    Laguna/Quezon

    1.0

    3

    185+

    78d

    451+

    144c

    425+

    158b

    41+

    7c

    Claveria

    MisamisOriental

    1.0

    9

    262+

    75d

    130+

    37b

    75+

    29c

    190+73

    c

    37+

    19b

    45+

    11c

    178+

    98c

    Tupi

    SouthCotab

    ato

    1.0

    14

    2,908+

    864c

    498+

    115b

    1,449+

    398c

    1,802+16

    1b

    80+

    22b

    1,780+

    272b

    170+

    43c

    315+

    87bc

    Tanauan

    Batangas

    1.0

    20

    130+

    78d

    1

    ,700+

    367b

    860+

    452c

    720+29

    4c

    50+

    23c

    Rainfedw

    etland

    Solano

    Cagayan

    1.0

    60

    491+

    217d

    1

    ,527+

    527b

    3,218+

    2,320b

    312+

    103ab

    595+

    269b

    437+

    133c

    131+

    36c

    Manaoag

    Pangasinam

    1.0

    85

    289+

    94d

    321+

    81b

    2,179+

    538c

    305+79

    c

    475+

    66c

    0c

    Oton/Tigbauan

    Lloilo

    1.0

    85

    1,824+

    528cd1

    ,012+

    215b

    2,440+

    770c

    1,016+61

    2bc

    458+

    185a

    455+

    314b

    997+

    108b

    38+

    30c

    Irrigated

    (synchronous)

    Victoria/

    StaMaria

    Laguna

    1.9

    90

    214+

    79d

    1

    ,366+

    896b

    629+

    426c

    195+95

    c

    410+

    168b

    413+

    88c

    35+

    35c

    Cabanatuan/

    Zaragoza

    NuevaEcija

    2.0

    80

    154+

    14d

    361+

    77c

    305+

    17c

    0c

    Koronadal

    SouthCotab

    ato

    2.0

    70

    1,677+

    345cd

    549+

    152b

    2,279+

    617c

    855+16

    9c

    93+

    17b

    706+

    252b

    662+

    83bc

    405+

    46b

    Irrigated

    (asynchronous)

    Jaen

    NuevaEcija

    2.0

    80

    15,224+

    238a

    9,207+

    2,561a

    137+

    59c

    0c

    Koronadal

    SouthCotab

    ato

    2.4

    70

    11,168+

    1,849b4

    ,043+

    1,117a

    3,158+

    501b

    2,961+38

    9a

    145+

    21ab

    7,169+

    720a

    2,519+

    430a

    751+

    123a

    P

    50.0001

    50.0001

    50.0001

    50.0001

    50.0001

    50.0001

    50.0001

    50.0001

    F

    37.01

    6.76

    5.54

    24.02

    3.12

    17.46

    19.91

    17.92

    df

    49

    40

    49

    40

    32

    36

    4

    9

    45

    RiceCult

    ureAverage

    Dryland

    871b

    776b

    709b

    904b

    59

    1,103b

    77

    b

    247

    Rainfedw

    etland

    868b

    953b

    2,612ab

    661b

    385

    525b

    636

    ab

    56

    Irrigated

    (synchronous)

    682b

    957b

    650b

    551b

    93

    558b

    460

    ab

    147

    Irrigated

    (asynchronous)

    13,196a

    4,043a

    6,183a

    2,961a

    145

    7,169a

    1,32

    8a

    376

    P

    50.0001

    0.04

    0.02

    0.02

    0.12ns

    0.007

    0.04

    0.07ns

    F

    40.95

    6.15

    6.30

    12.09

    7.33

    37.03

    3.04

    0.76

    df

    11

    8

    11

    7

    5

    6

    1

    1

    9

    aSynchronyreferstofarmersplantingdatesandsy

    nchronyoccurswhenfarmersplantwithinaperiodofonemonth,theaveragegenerationalperiodofmostinsectpests(Loevinso

    hnetal.1988).

    bCircum

    ferenceof1km

    2

    centeredaroundeachlighttrap.

    cDailycountsfromkerosenelighttraps.Nodatain

    dicatesthattheinsectinquestionwasnotmeasured.

    Inacolumn,means(+SEM).followedbyadiffere

    ntletteraresignificantlydifferent(P4

    0.05)byLSDanalysis.

    dS.innot

    ata.S.incertulas.

    eChilosu

    ppressalis,Sesamiainferens,Maliarphasp.

    International Journal of Pest Management 135

  • 8/2/2019 Dryland Rice Insect Pests

    9/22

    these catches that were equivalent to a number of

    wetland sites where moderate damage levels occur on

    almost each crop, such densities did not translate into

    significant infestations in the slash and burn area, as

    deadhearts or whiteheads were rarely encountered.

    3.1.2. Claveria

    Stand loss averaged 14% in the vegetative stage, more

    from sown-seed and root pests (Table 3), mainly ants

    again dominated by S. geminata. Germinating seed

    pests were mole cricket Gryllotalpa orientalis Burme-

    ister and field crickets, while root pests were white

    grubs, termites, root aphids, and mealybugs. The

    subterranean termite Macrotermes gilvus (Hagen)

    prefers wheat to rice as a 100-m2 trial plot (var.

    Trigo) sown within a rice field was totally denuded in

    the seedling stage. Densities in rice reached 7 termites/

    50 plants with minimal damage to rice.The seedling maggot averaged a modest 6%

    deadhearts. After tillering their numbers declined, as

    the larvae can only attack developing tillers. Defolia-

    tion from general defoliators including C. basalis flea

    beetle was subeconomic. The most common defolia-

    tors were leaffolders that averaged only 1% damaged

    leaves, the most common being M. patnalis Bradley

    followed by M. exigua and C. medinalis. Also noted

    was the armyworm Mythimna separata (Walker)

    which did negligible damage.

    Three species of root aphids built up in numbers

    with crop growth reaching 1.8/plant. Tetraneuranigriabdominalis (Sasaki) was most common followed

    by Geoica lucifuga (Zehntner) and Rhopalosiphum

    rufiabdominalis (Sasaki) in fields where numbers were

    high, plants turned yellow and became stunted. The

    common root mealybugs were Trionymus sp. and

    Pseudococcus sp. where only token numbers occurred

    (1/50 plants).

    Stemborer damage as deadhearts or whiteheads

    was 51% where species composition determined by

    tiller dissection showed striped stemborer Chilo

    suppressalis (Walker) to be the most dominant (72%

    of collections) followed by yellow stemborer (11%),

    white stemborer S. innotata (Walker) (10%) and

    finally pink stemborer Sesamia inferens (Walker)

    (6%). Larvae were reared to distinguish between the

    Scirpophaga species. Rice seed bugs were a mixture of

    L. oratorius and L. acuta averaging 1/m2.

    White grubs also fed on roots having the greatest

    detrimental effect on seedlings averaging 8/10-m row

    after harvest. The two most common species Leuco-

    pholis irrorata (Chevrolat) and Holotrichia mind-

    anaoana Brenske had synchronized 2-year life-cycles

    which on odd-numbered years resulted in the last

    instar larvae being in the field at the time of crop

    planting (Litsinger et al. 2002). Females ovipositduring land preparation predominantly in even

    numbered years when only the less damaging young Table3.

    Insectpestdensitiesonfarmersdryland

    ricefieldssowntoUPLRi5inClaveria,MisamisOrientalandTupi,SouthCotabato,

    Mindanao,Philippines19841990.a

    Site

    Crops

    (no.)

    Years

    Plantstandreduction(%)21DEb

    Seedingmaggot

    (%

    deadhearts)

    21DEb

    No.per50plants40DEd

    Stemborer(%)b

    Leaffolder

    (%

    damaged

    flagleaves)b

    Riceb

    ug

    (no./m

    2)e

    Whitegrubs

    (no./10-mrow)

    atharvest

    f

    Sown

    seedpests

    Root

    pests

    Combinedc

    Ter

    mites

    Root

    aphids

    Rootmealy

    bugs

    Deadhe

    arts

    5070DE

    Whiteheads

    Claveria

    7

    198490

    11.0+

    2.3

    9.3+

    1.6

    14.3+

    1.9

    5.8+

    1.1

    7.3+

    2.9

    88+

    30

    1.0+

    0.6

    0.6+

    0.3

    0.5+

    0.2

    1.2+

    0.3

    1.0+

    2.1

    8.4+

    2.8

    Tupi

    4

    198687

    198990

    8.4+

    3.9

    8.5+

    3.8

    13.6+

    2.8

    8.3+

    2.0

    9.9+

    5.4

    172+

    33

    0

    0.5+

    0.3

    0.7+

    0.3

    6.5+

    2.8

    1.0+

    0.9

    7.2+

    5.6

    aEachyear48fields(replications)weremonitored

    frominsecticideuntreated100m

    2

    plots.DE

    daysaftercropemergence.Dataarem

    eans+

    SEM.

    b0-mrowsample.

    cSampledfromtheuntreatedcheck.

    dPlantsdugupto10cmdepthandrootsinsepected

    .

    eDataar

    emeansfromsampling25m

    2

    duringmilk,soft,andharddoughstages.

    fPlantsd

    ugupandsoilinspectedto20cmdepth.

    136 J.A. Litsinger et al.

  • 8/2/2019 Dryland Rice Insect Pests

    10/22

    larvae affect the crop. Three other species were also

    found feeding on rice but in much lower densities:

    H. flachi Brenske, Adoretus luridus Blanchard, and

    Anomala humeralis Burmeister.

    Claveria had the lowest light trap collections of

    any of the sites (Table 2), as it was the most isolatedfrom wetland areas. Plant-/leafhoppers were rarely

    encountered on the crop. As with Siniloan, stemborer

    species balance favoured genera other than

    Scirpophaga.

    3.1.3. Tupi

    An average stand reduction of 14% resulted from

    both sown-seed and root pests in this favourable

    dryland site (Table 3). The main seed pest was S.

    geminata, along with mole and field crickets. Seedling

    maggot damage averaged 8% deadhearts but other-

    wise the seedling stage was not significantly infestedby insect pests, as flea beetle damage was barely

    noticed. Root pests were white grubs and termites.

    White grubs continued to gnaw away on roots

    throughout the crop season; the same five species as

    found in Claveria were recorded. They averaged 7 per

    10-m row at harvest. Termites averaged only 1 per

    five plants. Other root feeding pests were root aphids

    which averaged 43 per plant and thus were one of

    the most important pest guilds. Root mealybugs,

    however, were not recorded.

    Stemborers were mainly Scirpophaga dominated

    by white stemborer although there was a species shiftin the province prior to the study due to El Nino

    climate change (Litsinger et al. 2006a). Damage

    levels, however, were extremely low with deadhearts

    and whiteheads 51%. Leaffolders averaged 7%

    damaged leaves dominated by M. patnalis (55% of

    field collections) and C. medinalis (45%). Rice seed

    bugs were low in number (1/m2).

    Light trap densities of the major rice pests were

    greatest in Tupi among the dryland rice sites and even

    greater than many other wetland rice sites (Table 2).

    The exception was whitebacked planthopper Soga-

    tella furcifera (Horvath) which was more abundant in

    Tanauan due to its close proximity to Laguna

    wetland areas. Densities of brown planthopper,

    zigzag leafhopper Recilia dorsalis (Motschulsky),

    and Cyrtorhinus mirid were highest in Tupi except

    for one or two of the asynchronously planted

    irrigated rice sites. Green leafhopper numbers were

    high in the dryland sites but lower than the irrigated

    wetland sites. Stemborer densities were also highest

    among the dryland sites but much less than at the

    wetland sites. The ratio of Scirpophaga to non-

    Scirpophaga stemborers favoured the latter, equal to

    the other dryland sites. The most abundant was

    Maliarpha sp. with 63% of the catch followedby S. innotata (26%), C. suppressalis (9%), and

    S. inferens (2%).

    3.1.4. Tanauan

    Insect pest species richness and their densities in

    Tanauan were only higher than in Siniloan. Sown-

    seed and root pests (white grubs and termites) each

    reduced stand 911 and 1110%, respectively, whenmeasured individually and for each varietal type

    (Table 4). When both contributions are combined,

    stand reduction ranged from 14 to 16% for Dagge

    and UPLRi5, respectively. S. geminata was probably

    responsible for most of the loss of seed followed by

    mole and field crickets. Only one species of white

    grub L. irrorata was encountered and is a major pest

    of sugarcane in nearby fields. Farmers cite termites

    dominated by M. gilvus along with Coptotermes spp.

    as major dryland rice pests. Termites are permanent

    residents of the uplands as their nests extend below

    the plough pan and feed mainly on maize stalks and

    rice straw. Farmers, however, overestimate theimportance of termites but do not sufficiently value

    the other pests. Seedling maggot was also present but

    at very low numbers during the vegetative stage.

    Several defoliating Lepidoptera were recorded,

    based on periodic field collections during the 1976

    1980 seasons mainly in the reproductive stage. From

    the total collection of 635 larvae and pupae, the most

    abundant was greenhorned caterpillar Melanitis leda

    ismene Cramer (46% of the total) followed by

    armyworms M. separata (17%) and Spodoptera

    mauritia acronyctoides Guenee (16%) plus rice

    skipper Pelopidas mathias (F.) (13%) and brownsemi-looper Mocis frugalis (F.) (9%). However, the

    amount of defoliation from these species never

    reached economic levels.

    The light trap catches showed that among the four

    dryland sites Tanauan had supported high levels of

    plant- and leafhoppers, second only to Tupi (Table 2).

    The highest light trap densities were of whitebacked

    planthopper followed by green and zigzag leafhop-

    pers which were mirrored in field sampling (Figure 1).

    Densities peaked in the late reproductive stage and

    declined toward crop senescence with the abundance

    of Cyrtorhinus predator trailing that of its prey.

    Abundance of yellow stemborer from light trap

    catches was equivalent to that at other dryland sites

    except Tupi. Cyrtorhinus numbers were also moderate

    despite its major prey (planthoppers) being abundant

    on the crop.

    Plant- and leafhoppers were more abundant on

    Dagge than on UPLRi5 (Table 4). Numbers were

    higher on the traditional variety than on the modern

    variety, and at times small patches of hopperburn

    were noticed in some fields. Of the green leafhoppers,

    most individuals (83%) were N. virescens, as the

    remainder being N. nigropictus (Sta l) indicating that

    the site was still within the dispersal range of riceinsect pests moving from wetland areas. Cyrtorhinus

    predators were also prevalent, their numbers building

    International Journal of Pest Management 137

  • 8/2/2019 Dryland Rice Insect Pests

    11/22

    up over the season and peaking after the planthop-

    pers (Figure 1). Spider densities were at 12/m2

    throughout the crop season. The most common

    spiders were Argiope catenulata (Doleschall) (Aranei-

    dae), Clubiona japonicola Boesenberg et Strand

    (Clubionidae), Pardosa pseudoannulata (Boesenberget Strand) (Lycosidae), Tetragnatha javana (Thorell),

    T. mandibulata Walckenaer, and T. nitens (Audouin)

    (Tetragnathidae).

    Three species of rice stemborer were recovered by

    periodic stem dissections (n 703 larvae) with

    S. incertulas the most abundant (74%) followed by

    S. inferens (19%) and dark headed stemborer

    C. polychrysus (Meyrick) (7%). The combined da-

    mage as deadhearts was minimal 52%, with white-

    heads only slightly 43% (Table 4). Damage was

    higher on Dagge probably because as it was the taller

    variety it elongated more, making it more susceptible

    (Bandong and Litsinger 2005). These levels were alsosubeconomic. S. innotata does not occur on Luzon.

    The high nitrogen usage encouraged three leaf-

    folders, M. patnalis being the most commonly

    encountered while M. exigua colonized first followed

    by C. medinalis. Damage was greatest at the flag leaf

    stage, averaging 17 and 10% damaged leaves on

    traditional and modern varieties, respectively. Few

    living larvae were found upon opening up folded

    leaves indicating high rates of predation. Rice seed

    bug numbers were moderate averaging 1/m2 over the

    ripening stage which damaged 57% of rice grains.

    3.1.5. Comparison of rice agro-ecosystems

    Kerosene light traps operated over standardized crop

    seasons in 12 sites measured the activity of a narrow

    spectrum of rice insect pests, mainly plant-/leafhop-

    pers, stemborers, and Cyrtorhinus mirid that preys on

    hoppers. The overall analysis shows that brown

    planthoppers were just as prevalent in dryland as in

    other synchronously planted wetland sites (Table 2).

    The same was true of whitebacked planthopper for

    which there are no resistant rice varieties. Brown

    planthopper was found at low levels in three of the

    four dryland sites despite susceptible varieties being

    grown. At Tupi the incidence was as high as in

    irrigated areas and only less in the two asynchro-

    nously planted irrigated areas where genetic resis-

    tance had broken down. The low levels in Manaoag

    (IR36), Victoria (IR42) and Cabatanuan/Zaragoza

    (IR42, IR52) indicate that the rice varieties planted at

    the time were resistant to brown planthopper. At

    Iloilo farmers grew a mixture of traditional types and

    IR36 which was resistant. Tupi farmers selected

    modern semi-dwarfs that, however, were not

    resistant.

    Green leafhoppers were most abundant in Tupiwith the lowest incidence in Claveria but in none of

    the dryland sites were varieties resistant to greenTable4.

    Insectpestdensitiesonfarmersdryland

    ricefieldssowntotraditionalandmodernricevarietiesinTanauan,Batangas19761980.a

    Plantstand(%

    reduction)21DEb

    No.per25m

    2

    40DEd

    Stemborers

    b

    Ricebug

    de

    Variety

    Crops

    (no.)

    Sown

    seedpests

    Root

    pests

    Combinedc

    Nep

    hotettix

    spp.

    N.

    lugens

    S.

    furcifera

    R.

    dorsalis

    C.spectra

    Cyrtorhinus

    predator

    Deadhearts(%

    )

    Whiteheads

    Leaffolder

    (%

    damaged

    flagleaves)b

    No./m

    2

    Damaged

    grains(%)

    Whitegrubs

    (no./10-mrow)

    atharvest

    f

    50DE

    70DE

    Dagge

    5

    9.1+

    1.411.4+

    2.214.1+

    4.7

    78+

    47

    143+

    90

    1,318+

    763

    158+

    141

    8.1+

    2.531.4+

    10.10.4+

    1.41.6+

    0.5

    2.6+

    0.8

    17.3+

    5.1

    1.3+

    1.6

    4.1

    +

    0.6

    1.2+

    0.9

    UPLRi5

    3

    10.7+

    3.0

    9.8+

    2.416.4+

    5.3

    34+

    8

    34+

    12

    310+

    218

    27+

    17

    7.3+

    1.723.2+

    2.2

    0.3+

    0.10.5+

    0.2

    1.5+

    0.6

    9.6+

    4.0

    1.3+

    0.9

    3.7

    +

    0.5

    1.7+

    1.5

    aEachyear48field(replications)weremonitoredfrominsecticideuntreated100m

    2

    plotsfor

    eachvariety.

    DEdaysaftercropemergence.Dataaremeans+

    SEM.

    b10-mro

    wsample.

    cSampledfromtheuntreatedcheck.

    dSampledbyD-VAC1

    suctionmachine.

    eDataar

    emeansfromsamplingduringmilk,soft,a

    ndharddoughstages.Damagedgrainswe

    redeterminedbytheacidfuchsinstainingtechniqueofthefeedingsheaths(Litsingeretal.1981).

    fPlantsd

    ugupandsoilinspectedto20cmdepth.

    138 J.A. Litsinger et al.

  • 8/2/2019 Dryland Rice Insect Pests

    12/22

    leafhopper grown. In the case of dryland sites, crop

    apparency was low due to a long dry fallow and in

    irrigated sites genetic resistance prevailed. Modern

    rices are resistant to green leafhoppers, as can be seen

    in Laguna and Cabanatuan/Zaragoza among the

    irrigated sites. However, resistance had broken down

    in the asynchronous areas. Generally the lowest

    incidence was in both dryland and irrigated synchro-

    nous sites with high numbers in rainfed wetland and

    asynchronous irrigated areas. The short season dry-

    land rice environment suppressed green leafhopper

    build up whereas in synchronous irrigated sitesresistant varieties were responsible. There are no

    resistant varieties for zigzag leafhopper and white

    leafhopper Cofana spectra (Distant). The highest

    zigzag leafhopper incidence was the asynchronous

    sites, but was also recorded in Tupi, and there were

    no differences between dryland, rainfed wetland, and

    irrigated (synchronous) cultures. Also, there was no

    significant difference between rice cultures for white

    leafhopper. The mirid predator densities mirrored

    more the brown planthopper than any other prey

    species.

    The lowest incidence of Scirpophaga stemborers

    was recorded in the dryland habitat, whereas the

    highest catches were recorded in Koronadal fromwhite stemborer in the asynchronous area. The ratio

    of non-Scirpophaga to Scirpophaga was positive only

    Figure 1. Abundance of planthoppers, leafhoppers, spiders, and a mirid predator on a traditional dryland rice varietydetermined by weekly suction sampling with a DVAC machine in Tanauan, Batangas, Philippines, 1980. Data are averages of25 m2 samples in each of four fields where rice was sown in June and harvested in October. WBPH, whitebacked planthopperSogatella furcifera; BPH, brown planthopper Nilaparvata lugens; GLH, Nephotettix spp. green leafhoppers; ZLH, zigzagleafhopper Recilia dorsalis, and the mirid predator Cyrtorhinus lividipennis.

    International Journal of Pest Management 139

  • 8/2/2019 Dryland Rice Insect Pests

    13/22

    in the dryland sites. Differences among habitats for

    non-Scirpophaga species were not significant. Skew-

    ing the results were three sites wherein only Scirpo-

    phaga stemborers were found. Koronadal had high

    incidence of Maliarpha sp. stemborers in both

    dryland and irrigated sites.

    3.2. Yield loss

    3.2.1. Siniloan

    The slash and burn crop suffered greatly from both

    biotic and abiotic stresses, foremost of which was the

    eroded acidic soil which reduced root growth, thus

    when periodic water stress occurred the plants were

    not able to extract moisture from deep in the soil.

    Biotic stresses were mainly diseases and vertebrate

    and insect pests to the extent that in 1984 there was

    no grain harvest. Most serious was leaf and neck blast

    disease (Pyricularia oryzae Cav.) followed by rats andbirds and an array of insect pests. Also, those plots

    that grew best from added soil amendments and good

    insect control suffered greater yield loss than the

    poorer-growing ones due to lodging from typhoons.

    At harvest, plant height and biomass were insignif-

    icantly different between treatments. Less blast

    occurred in the second year which produced a grain

    harvest that illustrates how complex quantification of

    yield loss is when multiple stresses are involved. The

    farmers practice was the untreated (treatment 7)

    which yielded 60 kg/ha, barely returning the seed

    sown (Table 1). If insecticide seed treatment isprovided, a yield of 210 kg/ha occurred resulting in

    a loss of 0.15 t/ha or 71% (treatment 6 vs. 7) from

    sown-seed and seedling pests. However, if the farmer

    follows the recommended practice of adding 23 kg N/

    ha there is no yield loss from insect pests (treatment 3

    vs. 4) because of crop compensation. The best

    treatment yielded 480 kg/ha, that is the technical

    yield potential, which was also reached by applying

    only 23 kg N/ha plus ash without insecticide. There

    was no additional yield benefit from foliar sprays,

    thus the measured yield loss from insects came from

    control of the sown-seed and seedling pest guilds.

    Clearly, the degree of loss is highly influenced by

    agronomic practices.

    3.2.2. Claveria

    Claveria resulted in an average loss of 0.79 t/ha or

    23% as the difference between the fully protected

    crop (3.43 t/ha) and the untreated (2.64 t/ha) grown

    under farmer management (Table 5). Using the

    partitioned yield loss method, losses (t/ha) were

    highest from white grubs and termites (0.31) followed

    by root aphids (0.21), sown-seed/seedling pests (0.16),

    rice seed bugs (0.09), and stemborers (0.03). The totalyield loss equalled 22.8%: that breaks down to 8.3%

    for white grubs and termites, 6.6% for root aphids, Table5

    .Yieldlossdeterminedbypartitionedgrowthstage,insectpestguildinse

    cticidecheckmethodinfarmersfieldsinClaveria,MisamisOrientalandT

    upi,S.Cotabato,

    Philippines19851991.a

    Site

    Crops

    Years

    Yield(t/ha)

    Yieldloss(t/ha)

    Yieldloss(%

    )

    Complete

    protection

    Untreated

    Total

    Whitegrubs

    termites

    Seed/

    seedling

    p

    ests

    Rootaphid

    Stemborer

    leaffolder

    Ricebug

    Total

    White

    grubs/

    termites

    Seed/

    seeding

    pests

    Root

    aphid

    Stemborer/

    leaffolder

    Ricebug

    Claveria

    7

    198591

    3.43+

    0.21a

    2.64+

    0.16b

    0.79+

    0.24

    0.31+

    0.18a

    0.16+

    0.21ab

    0.21+

    0.17ab

    0.03+

    0.10b

    0.09+

    0.08ab

    22.8

    8.3

    5.0

    6.6

    0.7

    2.2

    P

    0.01

    0.04

    F

    6.36

    3.17

    df

    73

    73

    Tupi

    4

    1987198991

    3.46+

    0.19

    2.96+

    0.32

    0.50+

    0.16

    0.10+

    0.08

    0.11+

    0.08

    0.19+

    0.15

    0.09+

    0.07

    0.01+

    0.04

    15.0

    3.3

    3.5

    5.5

    2.5

    0.2

    P

    0.21ns

    ns

    F

    1.61

    0.57

    df

    39

    39

    aYieldlo

    sswasdeterminedasthedifferencebetweenafullprotectiontreatmentthatprotectedfivegrowthstagesandassociatedpestgu

    ildswithinsecticidesselectedfortheireffic

    acyandphytotoxic/

    phytotonicneutrality

    Fivetreatmentssucessivelyomittedinsecticideprotectionsothatlossescouldbeattributedtothatgrowthstage/guild

    Seetext

    fordescriptionsofinsecticides,dosagesandtiming.Alltrialswereconductedinfarme

    rsfieldswitheachfieldasareplication.

    Means+

    SEM

    inarownotfollowedbyacommonletteraresignificantlydifferent(P

    0.05)byLSDtest,ns

    notsignificantlydifferent

    140 J.A. Litsinger et al.

  • 8/2/2019 Dryland Rice Insect Pests

    14/22

    5.0% for sown-seed pests, 2.2% for rice seed bugs,

    and 0.7% for stemborers.

    3.2.3. Tupi

    Interestingly the yield potential in Tupi was similar tothat in Claveria despite these being better soils in

    Tupi. The four crops in Tupi recorded a 0.50 t/ha or

    15% loss as the difference between 3.46 and 2.96 t/ha

    for full protection and the check (Table 5). Yield loss

    was lower than in Claveria, even though pest

    complexes and densities were similar. The ranking

    of yield loss differed, in that the greatest loss was from

    root aphids (0.19 t/ha or 5.5%), sown-seed/seedling

    pests (0.11 t/ha or 3.5%), white grubs and termites

    (0.10 t/ha or 3.5%), stemborers (0.09 t/ha or 2.5%),

    and rice seed bugs (0.01 t/ha or 0.2%).

    3.2.4. Tanauan

    In Tanauan the modern variety out-yielded the

    traditional variety under the same management in

    both the comparison of full protection (4.01 vs.

    2.90 t/ha) and untreated (3.61 vs. 2.85 t/ha) (Table 6).

    With Dagge there was an insignificant difference

    between the protected (2.90 t/ha) and the unprotected

    check (2.85 t/ha); total loss was too low (5%) for us

    to have confidence in assigning significance appor-

    tioned across each pest guild. There was, however, a

    significant yield loss in UPLRi5 of 0.40 t/ha or 18%

    between protected and unprotected treatments. Thishigher loss was spread fairly evenly between guilds

    with most due to white grubs/termites and sown-seed/

    seedling pests (5.9% each), and least in the ripening

    stage (2.3%). The 4.1% loss in the reproductive stage

    was probably from whitebacked planthopper.

    3.3. Crop management

    3.3.1. Siniloan

    The slash and burn crop was protected with a seed

    treatment as an initial recommended practice which

    was tested together with several agronomic and

    insecticide interventions including organic and inor-

    ganic fertiliser amendments and foliar sprays. With

    carbosulfan seed treatment, yields increased signifi-

    cantly from 60 to 210 kg/ha, giving a marginal return

    of only $0.60/ha (Table 1). The highest yield with the

    least amount of inputs was obtained from applying

    23 kg N/ha and ash to reach 450 kg/ha and a

    marginal return of $42 and B:C ratio was a highly

    favourable 6.1. Adding the seed treatment did not

    increase yield, neither did adding lime, chicken

    manure, doubling the N rate from 23 to 46 kg/ha,

    or foliar insecticide sprays. This agronomic recom-

    mendation probably increased the crops compensa-tory ability against insect pest damage, but it may

    change if and when the yield potential can be Table6

    .Yieldlossdeterminedbypartitionedgrowthstage/pestguildinsecticidecheckmethodinfarmersfieldsinTana

    uan,Batangas,Philippines,19761980

    .a

    Varietytype

    Crops

    Years

    Yield(t/ha)

    Yieldloss(t/ha)

    Yieldloss(%)

    Complete

    protection

    Untreated

    Totalyield

    loss

    Whitegrubs/

    termites

    Sownseed/

    seedlingpests

    Reproductive

    stage

    Ripening

    stage

    Total

    yieldloss

    White

    grubs/

    termites

    Seed/

    seedling

    pests

    Reproductive

    s

    tage

    Ripening

    stage

    Traditional/

    Dagge

    5

    197680

    2.90+

    0.082

    .85+

    0.06

    0.05+

    0.07

    0.02+

    0.02

    0.01+

    0.02

    0.01+

    0.01

    0.01+

    0.02

    1.7

    0.8

    0.3

    0.3

    0.3

    P

    0.734ns

    F

    0.117

    df

    63

    Modern/

    UPLR

    i5

    3

    197880

    4.01+

    0.163

    .61+

    0.35

    0.40+

    0.38

    0.13+

    0.12

    0.13+

    0.12

    0.09+

    0.10

    0.05+

    0.06

    18.2

    5.9

    5.9

    4.1

    2.3

    P

    0.165ns

    F

    2.02

    df

    35

    aYieldlo

    sswasdeterminedasthedifferencebetweenafullprotectiontreatmentthatprotected

    fourgrowthstagesandassociatedpestguildswithinsecticidesselectedfortheirefficacy.

    Fourtre

    atmentssuccessivelyomittedinsecticidepro

    tectionsothatlossescouldbeattributedtothatgrowthstage/guild.

    Seetext

    fordescriptionsofinsecticides,dosagesandtiming.Alltrialswereconductedinfarme

    rsfieldswitheachfieldasareplication.

    Means+

    SEM

    inarowfollowedbyadifferentletteraresignificantlydifferent(P5

    0.05)by

    LSDtest,ns

    notsignificantlydifferent

    International Journal of Pest Management 141

  • 8/2/2019 Dryland Rice Insect Pests

    15/22

    significantly raised, i.e. with a blast resistant variety,

    better soil management, or improved vertebrate pest

    control.

    3.3.2. Claveria

    The crop management variables tested were inorganic

    fertiliser, seeding rate, as well as insecticide soil and/

    or seed/seedling treatments. The lowest yields resulted

    without addition of either fertiliser or insecticide

    producing equally low levels (2.4 t/ha) at both the 50

    and 90 kg/ha seeding rates (Table 7). The minimal

    input to raise the yield significantly above the

    untreated level was the adding of inorganic fertiliser

    (50-25-0) at the 90 kg seed/ha rate producing 2.6 t/ha

    but a negative marginal return of $9/ha. The next

    yield plateau was reached at a seeding rate of 50 kg/

    ha, 50-25-0 fertiliser level, and an insecticide seed

    treatment which improved to 2.8 t/ha, but only ameagre marginal return of $7/ha with a B:C ratio of

    only 1.2. The addition of a soil insecticide treatment

    to this last practice did not further improve yield. The

    fully protected plots produced the highest yields

    (3.4 t/ha) among all treatment combinations and

    out-yielded the fertilised control at the same seeding

    rate by 0.80 t/ha. This treatment was not a potential

    practice; but it merely showed that there was more

    scope for improvement. Thus the most profitable

    practice for the farmer was to not apply any

    agricultural inputs due to the meagre yield response

    of added N and P.

    3.3.3. Tupi

    Almost the same mix of management practices was

    tested in Tupi as in Claveria. Lowest yields again

    were in the unfertilised and non-insecticide treated

    plots at either 50 and 90 kg/ha seeding rates (2.42.5 t/ha) (Table 7). There was no benefit of

    increasing just the seeding rate from 50 to 90 kg/

    ha without fertiliser or seed treatment. The most

    economical return was at 50 kg seed/ha with

    inorganic fertiliser (50-25-0) and insecticide seed

    treatment with highest marginal return of $90/ha

    and 3.0 B:C ratio. Adding only fertiliser had a

    greater benefit as yields increased 0.53 and 0.66 t/ha

    in the 50 and 90 kg/ha seeding rates, respectively,

    although not significantly. Economically there was a

    marginal return of $44/ha and B:C ratio of 2.8

    from use of inorganic fertiliser at 50 kg seed/ha

    without seed treatment. If the farmer increased to90 kg seed/ha with fertiliser but without seed

    treatment, the marginal return became $64/ha and

    B:C ratio rose to a more favourable 3.1. Profit was

    higher with an insecticide seed treatment however.

    No further benefit was realized from applying the

    more expensive soil insecticide treatments, indicat-

    ing less pest pressure than in Claveria. In Tupi the

    full protection treatment did not yield more than

    the seed treatment or use of fertiliser. Therefore

    there was no significant benefit from any insecticide

    protection on a well fertilised crop at 90 kg seed/ha,

    again indicating crop compensation.

    Table 7. Yield response to crop management practices on dryland rice production, Claveria and Tupi, Mindanao,Philippines, 19841990.

    Insecticidetreatment

    Seedingrate

    (kg/ha)N-P

    (kg/ha)

    Yield (t/ha)a

    Claveria No. crops N Tupi No. crops n

    Completeprotectionb

    90 5025 3.41 + 0.23 a 7 28 3.61 + 0.22 a 4 16

    Seed/seedling soil treatments

    90 5025 2.89 + 0.15 b 4 16

    Seed/seedling

    treatment

    90 5025 2.85 + 0.23 b 6 24 3.20 + 0.26 ab 4 16

    Seed/seedlingtreatment

    50 5025 2.84 + 0.28 b 3 28 3.48 + 0.24 ab 3 12

    Soil treatment 90 5025 2.65 + 0.17 bc 2 8 3.41 + 0.29 ab 2 8Soil treatment 50 5025 2.69 + 0.39 bc 3 26 3.37 + 0.37 ab 3 12Untreated 90 5025 2.61 + 0.18 bc 7 28 3.17 + 0.28 ab 4 16Untreated 50 5025 2.33 + 0.25 c 3 28 2.96 + 0.32 ab 3 12Untreated 90 0 2.44 + 0.32 3 12 2.51 + 0.30 b 3 12Untreated 50 0 2.42 + 0.28 c 3 12 2.43 + 0.27 b 3 12P 0.007 0.03F 5.50 2.22df 294 135

    aEach crop consisted of 46 farms replications (n) and treatments were analyzed on the basis of fields and not crops.

    In a column, means + SEM followed by a different letter are significantly different (P 0.05) by LSD analysis.

    Soil treatment was 0.25 kg al/Lindane G/ha, seed treatment was 0.30 kg ai carbosulfan ST/ha.

    Fertilizer applied was 50250 with the N application split at 1425 days after crop emergence during inter-row cultivation and panicleinitiation while P was applied basally during land preparation.bFrom the yield loss trial which treatments were randomly mixed with the insect control treatments each season.

    142 J.A. Litsinger et al.

  • 8/2/2019 Dryland Rice Insect Pests

    16/22

    3.3.4. Tanauan

    Combinations of seeding rates and insecticide treat-

    ment of seeds and soil were tested in one season with

    Dagge variety. There was a yield benefit (0.50 t/ha) of

    increasing the seeding rate from 50 to 100 kg/haalthough not significantly (Table 8), but it had a

    highly favourable marginal return of $55/ha and a

    B:C ratio of 8.3. Higher seeding rates up to 150 kg/ha

    did not increase yield. At the 50 kg/ha seeding rate,

    insecticide seed treatment resulted in modest but

    insignificant yield gains (0.230.27 t/ha) that pro-

    duced a marginal return of only $15/ha and B:C ratio

    of 1.7. Slightly higher, but again insignificant, gains

    (0.36 t/ha) were obtained with the 0.5 kg a.i. carbo-

    furan/ha soil insecticide treatment at 50 kg seed/ha

    producing a marginal return of $20/ha and B:C ratio

    of 1.8. The highest yield gain (0.64 t/ha) was obtained

    with the highest rate of soil insecticide treatment(2 kg a.i. carbofuran/ha) at the 100 kg/ha seeding

    rate producing marginal return of only $7/ha and B:C

    ratio of 1.1.

    In 2 years trials (19801981) the same seed and

    soil treatments produced no significant yield gain in

    UPLRi5. In the 1979 trial which had more sown-seed

    and root (white grub) pest pressure, also at 100 kg

    seed/ha, that included both a seed treatment of

    0.30 kg a.i. carbosulfan ST/ha (3.68 t/ha) and a soil

    treatment of 0.75 kg a.i. diazinon G/ha (3.82 t/ha)

    resulted in significant yield increase over the un-

    treated of 3.28 t/ha (P 0.03, F 3.54, df 28).

    Both insecticide treatments produced similar margin-

    al returns of $31 and $41/ha, respectively, and B:C

    ratios of 2.6 and 2.4. The average in untreated plotswith UPLRi5 over the 3 years was 3.6 t/ha; thus, in

    1979 there was greater insect pest pressure. The

    farmer practice of 100 kg seed/ha without insecticide

    is probably the most economical practice as farmers

    are hesitant to use insecticides.

    4. Discussion

    4.1. Yield potential

    Except for the very low productivity of the slash and

    burn site, yields in the unprotected treatment of the

    other three sites ranged from 2.6 to 3.6 t/ha; these are

    much higher than the world average of 1 t/ha. Thiseffect is likely attributable to the Philippine climate

    and soils being as a whole more suitable for dryland

    rice compared to those of many other countries. The

    highest yielding site was Tanauan with the most

    favourable soils, good crop management, and highest

    use of inorganic fertiliser where the low tillering

    variety Dagge averaged 2.9 t/ha, which is equal to

    yields obtained from similar traditional varieties in

    the irrigated wetlands. The high tillering UPLRi5

    Table 8. Comparison of different practices on the yield of Dagge rice including increasing the seeding rate, insecticide seedand soil treatment and combinations thereof, Tanauan, Batangas, Philippines, 1976.a

    Seed rate

    Insecticide treatedPlant stand

    b

    no./m-row) 14 DEYield

    c

    (t/ha)Seed Soil

    Seeding rate50 kg seed Untreated Untreated 35 d 2.00 b100 kg seed Untreated Untreated 45 bc 2.49 ab125 kg seed Untreated Untreated 60 a 2.43 ab150 kg seed Untreated Untreated 59 a 2.23 ab

    Seed treatment50 kg seed Carbofuran ST Untreated 38 cd 2.27 ab50 kg seed Dieldrin WP Untreated 41 cd 2.23 ab

    Soil treatment50 kg seed Untreated 0.5 kg ai carbofuran G/ha 38 cd 2.36 ab50 kg seed Untreated 2 kg ai carbofuran G/ha 41 cd 2.47 ab

    Seed treatment increasing seeding rate100 kg seed Carbofuran ST Untreated 51 ab 2.27 ab100 kg seed Dieldrin WP Untreated 52 ab 2.19 ab

    Seed treatment increasing seeding rate100 kg seed Untreated 0.5 kg ai carbofuran G/ha 52 ab 2.51 ab100 kg seed Untreateed 2 kg ai carbofuran G/ha 51 ab 2.64 aP 0.02 0.04F 5.38 4.98df 33 33

    aFarmer practice was 00 kg seed/ha without insecticide usage and N was applied at 60 kg/ha in all treatments.

    ST seed treatment formulation, WP wettable powder, G granule.

    In a column, means + SEM followed by a different letter are significantly different (P 0.05) by LSD analysis. The randomized completeblock trial was replicated across six farmers fields.b20 samples of 1-m rows. DE days after crop emergence.cTwo 10 m2 yield cuts.

    International Journal of Pest Management 143

  • 8/2/2019 Dryland Rice Insect Pests

    17/22

    averaged even more 3.6 t/ha under the same

    management. Its characteristics of high tillering

    combined with blast resistance make UPLRi5 an

    attractive variety for dryland rice farmers. Tupi

    soils are similar to those of Tanauan, but farmers

    applied less nitrogen, 30 vs. 110 kg/ha, andaveraged 3.0 t/ha with UPLRi5 almost equal to

    Dagge under higher N rates. The two sites with the

    least favourable soils achieved the lowest yields.

    Claveria, with its highly eroded acidic soils and low

    organic matter content, averaged 2.6 t/ha. Slash

    and burn culture, exemplified by Siniloan, recorded

    the lowest yields with the farmers only doubling

    their investment in seed.

    4.2. Partitioned-growth-stage yield loss

    There was no partitioned yield loss data for the slash

    and burn site due to the difficulty of carrying outmultiple-treatment trials on small and very patchy

    fields. The mean yield loss figure across the other

    three sites for UPLRi5 was 18.6%, with 5.8%

    attributed to white grubs and termites, 4.8% to

    sown-seed/seedling pests (ants, crickets, seedling

    maggot, and flea beetles), 4.0% to root aphids and

    mealybugs, 2.4% to reproductive stage pests (stem-

    borers, leaffolders and plant- and leafhoppers), and

    1.6% to rice seed bugs.

    The partitioned-growth-stage method has its

    limitations. We noted interactions in the yield loss

    calculations between pest guilds attributed to thebroad spectrum nature of the insecticides used and

    the overlapping growth stages of a number of them.

    The greatest interaction was evident with Lindane

    granules between damage caused by white grubs and

    termites (root pests) that would also have affected

    sown-seed pests such as ants and crickets. The degree

    with which this occurred can be seen in Tables 3 and 4

    where sampling plant stand in the untreated controls

    (combined) was 3.36.0% lower loss than was

    estimated by separate treatments that measured

    sown-seed/seedling and root pest protection. Other

    overlapping stages were between root aphids/mealy-

    bugs and the foliar pest protection for the reproduc-

    tive stage. In the former the granular insecticide

    carbofuran directed at root aphids and mealybugs

    would have had some systemic activity against

    stemborers and leaffolders dwelling in and feeding

    on the above-ground plant parts. The dosage selected

    (0.75 kg ai/ha) for carbofuran was such that only

    trace amounts would have been taken up by the stems

    and leaves from the root zone placement. Any

    statistical interaction would have underestimated

    loss from reproductive pests.

    Additional interactions would have developed

    from use of carbofuran as it has other propertiesthan as an insecticide. In addition to being a

    nematicide it also has been found to have plant

    growth hormone like properties (Venugopal and

    Litsinger (1984). Nematodes are a serious problem

    in dryland rice (Villanueva et al. 1992), causing

    high losses in their own right. In fact, carbofuran is

    the chemical of choice for estimating losses in the

    nematicide check method, but again the dosageused in our study (1 kg ai/ha) was half used by

    nematologists (e.g. Plowright et al. 1990) and at the

    lower end of the dosage range that gives a

    phytotonic effect. Another mitigating effect was

    that the timing of the first of two applications 40

    and 70 DE which, being late in the growth period,

    would have tempered both potential effects. Each of

    these factors would lead to an overestimation of

    yield loss. On the other hand, none of the

    insecticide treatments achieved 100% control of

    insect pests in the trials offsetting the aforemen-

    tioned effects.

    The low yield loss of 5% on Dagge rice contrastsmarkedly with the results of the earlier studies by

    Pathak and Dyck (1975) where 430% yield loss was

    recorded on similar traditional varieties in the same

    location. Such high losses recorded earlier were

    probably due to the high rate of in-furrow carbo-

    furan G (2 kg ai/ha) which would have both induced

    a phytotonic response (Venugopal and Litsinger

    1984) and significantly suppressed nematodes (Villa-

    nueva et al. 1992), resulting in overestimates of

    insect losses. Carbofuran was not used in our study

    in Tanauan.

    4.3. Insect pest complex

    The key ecological factors of dryland sites that

    affected insect species composition were the result

    of: (1) an aerobic soil due to lack of soil puddling and

    ponding, (2) the extensiveness of alternative plant

    hosts (perennial grasslands and presence of maize or

    sugarcane), and (3) nearness to wetland rice bowls.

    We conclude that the most significant species

    associated with loss were adapted to the non-flooded

    drylands (ants, field crickets, mole crickets, white

    grubs, termites) or those that survive well on grassy

    weeds, or on the more dominant maize or on

    sugarcane (seedling maggot, flea beetle, root aphids,

    mealybugs, thrips, rice seed bugs, non-Scirpophaga

    stemborers). The most important among this list were

    undoubtedly ants which was the only taxon of

    economic importance acting across all four sites.

    Root aphids also reached economic numbers in two

    sites. Seedling maggot was economically important in

    all sites except for Tanauan which is devoid of

    grasslands. Siniloan has extensive grassy areas

    comprising the abandoned slash and burn fields

    from previous seasons that only slowly return to

    forest.Termites appear not to prefer living rice plants, so

    are not considered economically important pests as

    144 J.A. Litsinger et al.

  • 8/2/2019 Dryland Rice Insect Pests

    18/22

    dead plants were not seen. Thrips and mealybugs

    were present and would have become more abundant

    if prolonged drought had occurred. Rice seed bugs,

    like flea beetles, reached economic importance only in

    Siniloan due to their concentration in small fields.

    Rothschild (1970) also attributed high numbers ofrice seed bugs in slash and burn cultivation in

    Sarawak to the same factor. Both pest groups have

    good local dispersive powers and can seek out small

    plantings.

    Puddling is a means of transforming a soil into a

    muddy consistency to allow long-rooted seedlings to

    be readily transplanted (DeDatta 1981). This process

    of tillage and eventual ponding would kill all except

    the most adapted soil-dwelling insects such as aquatic

    root weevils (Lissorhoptrus and Echinocnemus) which,

    only occur in flooded habitats. Our results show that

    dryland soil pests are mainly those that cannot

    tolerate sustained flooding. There are only limitedrecords of soil-dwelling pests such as termites, mole

    crickets, and ants occurring in the wetlands. They

    exist primarily when there are large bunds that

    provide an aerobic environment from where they

    can make temporary incursions into the field during

    periods of low water saturation but cannot sustain

    themselves (Way et al. 1998). Flooding is in fact a

    control method for these pests (Litsinger 1994).

    Polyphagy is one of the most distinguishing fea-

    tures of dryland rice insects (Litsinger et al. 1987a) and

    is exhibited by most of the prevalent species. The only

    monophagous insect pest encountered in significantnumbers in dryland rice was the white stemborer.

    Whitebacked planthopper has a wide host plant range

    compared to the monophagous but more infamous

    brown planthopper which it outnumbered in the

    drylands. The host plant range of N. virescens is

    primarily Oryza spp., whereas N. nigropictus could

    persist in the drylands on perennial grasses.

    The dryland site with the highest measured pest

    densities was in Claveria which has large tracts of

    wild perennial grasslands which grow quickly with

    the first rains of the wet season enabling pest densities

    to build up rapidly. In Brazil for example, the

    extensive pasture lands surrounding tracts of dryland

    rice are responsible for build up of a number of pests

    such as spittle bugs which disperse to rice with

    devastating effect (Litsinger et al. 1987a). Both Tupi

    and Tanauan are highly cultivated areas where

    dryland rice occupies a smaller footprint than the

    surrounding maize and sugarcane; thus, immigration

    can be significant by more dispersive species, i.e.

    seedling maggot, flea beetles, white grubs, mole

    cricket, field crickets, root aphids, planthoppers,

    mealybugs, and rice seed bugs. Ho and Kibuka

    (1983) found that in dryland areas of Kenya where

    rice was grown in association with maize andsorghum, the principal rice stemborers were those

    species that fed on all three crops.

    Aestivation is a mechanism enabling survival

    during a long dry season and S. innotata has the

    ability to enter this diapause state for up to a year

    (Litsinger et al. 2006a). No other Asian stemborer has

    this adaptive trait and the ability to diapause is

    probably the reason that in Sarawak, Rothschild(1971) found it to be the most abundant stemborer in

    slash and burn rice. But despite this adaptation, its

    numbers were lower than other stemborer species in

    the Philippines as determined from tiller dissections

    (Jahn et al. 2007). Yellow stemborer enters quiescence

    in cool climates such as N. India and Bangladesh but

    not in the Philippines. Quiescence is not a true

    diapause state as the insect can resume activity

    rapidly from rain or land soaking (Islam 1993).

    Quiescence is different than diapause in that it occurs

    in direct response to an environmental change while

    S. innotata enters physiological inactivity in anticipa-

    tion of the change from short day lengths and cropage (Litsinger et al. 2006a). Dry season tillage in the

    two Mindanao sites acts as a control measure against

    white stemborer survival. Other species with known

    dormancy abilities are Leptocorisa rice seed bugs,

    white grubs, and rice butterflies (Litsinger et al.

    1987a).

    Light trap data revealed that traditional pests

    emanating from wetland culture continually rained

    down on dryland rice fields in all four sites essentially

    year-round (Table 2). Tupi, Tanauan, and Siniloan

    were situated relatively near (1020 km) to large

    wetland rice bowls, whereas Claveria is more distant.Light trap data showed that dispersal of wetland rice

    insect pests, even to Siniloan in the midst of a mature

    tropical rainforest, occurred when rice was not even

    present. The most dispersive species are brown and

    whitebacked planthoppers, Cyrtorhinus, armyworms/

    cutworms, locusts, and leaffolders followed by rice

    butterflies, leafhoppers, and stemborers (Litsinger

    et al. 1987a).

    Inorganic nitrogen usage and drought stress

    accounted for the abundance of some insect pests

    among sites. Tanauan with highest nitrogen usage

    would have favoured planthoppers and leaffolders

    along with stemborers, as noted by studies done in

    wetland rice culture (Litsinger 1994). The beneficial

    effects of nitrogen on planthoppers have been higher

    fecundity, greater survivorship, and increased feeding

    rates while those for leaffolders were egg recruitment

    (ovipositing moths are attracted to the most vigorous

    growing fields) and survival (Kraker et al. 2000).

    Rothschild (1970) also found leaffolders on dryland

    rice to be more abundant in fields of high N