Presence/absence Sequential Plans for Pest Management …pertanika.upm.edu.my/Pertanika PAPERS/JTAS Vol. 20 (1) Apr. 1997/07... · PRESENCE/ABSENCE SEQUENTIAL PLANS FOR PEST MANAGEMENT

PertanikaJ. Trop. Agric. Sci. 20(1): 51-63(1997) ISSN: 0126-6128© Universiti Putra Malaysia Press

Presence/absence Sequential Plans for Pest Management Decision-making,for Arthropods of Wet Rice Ecosystem in Malaysia

S.T.S. HASSAN and M.M. RASHIDDeparment of Biology

Universiti Putra Malaysia43400 UPM, Serdang, Selangor, Malaysia

Keywords: Insecta, rice, pest management, sequential sampling, Malaysia

ABSTRAKPelan pensampelan berjujukan disediakan bagi 11 kategori artropod untuk membantu pengurusan populasinyadalam sawah padi yang mengandungi pelbagai spesies perosak di Malaysia. Data dari pengamatan visualterhadap 204 sampel, dengan menggunakan 40 dan 100 rumpun dalam setiap sampel, telah diperolehi bagimerumus pelan tersebut. Ambang tindakan bagi setiap satu daripada 11 kategori (lima perosak, enam predator)artropod berkenaan diperolehi melalui regresi polinomial kuasa keempat perkadaran infestasi melawan mindensiti populasi, dengan menumpu pada nilai titik permulaan ketepuan infestasi. Species perosak adalahNephotettix spp., Nilaparvata lugens, Recilia dorsalis, Sogatella furcifera dan Cnaphalocrocis medinalis(Pyralidae) manakala predator pula diwakili oleh Cyrtorhinus lividipennis, Anatrichus pygmaeus (Diptera),spiders, Odonata, Paederus fuscipes dan Casnoidea spp. Tahap risikofenis I (a) andjenis 11 (ft) ditetapkanpada nilai 0.3, kerana nilai yang lebih rendah memerlukan bilangan sampel yang berganda banyaknya. Pelanberjujukan dapat dihasilkan dengan menggunakan program komputer SEQUAN (Talerico dan Chapman1970). Dalam penggunaan setiap pelan ke atas kawasan yang tidak melebihi 50 ha} adalah dinasihati supayasekurang-kurangnya 10 sampel diteliti dahulu sebelum membuat apa-apa rumusan pengurusan. Pensampelanserentak perosak dan predator membolehkan status populasi predator diambilkira dalam ketetapan pengurusanpopulasi spesies perosak.

ABSTRACTPresence-absence sequential sampling plans are presented for 11 arthropod categories to assist in management oftheir populations in the multipest-infested rice crop in Malaysia. Data from visual inspection of 204 samples, with40 and 100 hills per sample, were used to develop the plans. Action threshold for each of the 11 (5 pests, 6predators) arthropod categories was obtained through a fourth-order polynomial regression of proportion ofinfestation against mean population densities, at the point of saturation of infestation. The pest species are:Nephotettix spp., Nilaparvata lugens, Recilia dorsalis, Sogatella furcifera and Cnapholocrocis medinalis(Pyralidae), and the predators: Cyrtorhinus lividipennis, Anatrichus pygmaeus (Diptera), spiders, Odonata,Paederus fuscipes and Casnoidea spp. Risk levels of Type I (a) and Type II error (ft) were prefixed at 03,since lower levels entail taking a larger number of samples. The sequential plans were then generated using theSEQUAN computer program of Talerico and Chapman (1970). During field operation on not more than 50 haat a time, it is suggested that at least ten hills should be examined visually before recommending any pestmanagement action. Simultaneous sampling of pests and predators enables status of predators' populations to beconsidered before recommending any decision.

INTRODUCTION One major factor that obviates development of aDespite claims to the contrary, many pest con- nationwide comprehensive pest information-trol decisions on rice in Malaysia are still based based management decision system (e.g. Song eton short term ad hoc considerations, compared al 1992) is the inability to continually monitorto relatively long term ecological based the status of pest populations on a nationwideweightings for sustainable agricultural output. scale. However, on a regional basis, surveillance,

S/T.S. HASSAN AND M.M. RASHID

monitoring and forecasting systems have beenattempted (Ooi 1982a, b; Ooi and Heong 1988)and are still operational to a limited extent insome major rice planting areas in the country.To date, determination of pest status is stilllargely based on the fixed-sample-size samplingsystem. The inefficiency and limitations of thefixed-sample-size decision-making system havebeen well elaborated and are applicable to anyecosystem (Sterling and Pieters 1979).

In sequential sampling, the sample size isnot prefixed; its rationale, prerequisites for de-velopment of the plans, advantages and disad-vantages have been described (Onsager 1976;Pieters 1978). The plans can be developed fromformulae presented by Waters (1955), and easilyproduced using the SEQUEN computer pro-gram (Talerico and Chapman 1970).

Operational sequential plans for manage-ment of pests of major world crops such ascotton are well established (e.g. Sterling 1976;Pieters 1978; Rothrock and Sterling 1982; Plantand Wilson 1985). For rice, the operation ofsequential decisions on planthoppers on an ex-perimental basis has been attempted in coun-tries such as Madagascar (Bianchi et al 1989)and the Philippines (Shepard et al 1986). Therehas been no attempt reported on developingthe experimental operation into an area-wideactual farm decision-making strategy. Moreover,the scattered sequential plans developed for ricepests focused on certain species at a time(Nishida and Torii 1970; Kuno 1977, 1986;Shepard et al 1986; Ferrer and Shepard 1987;Shepard et al 1988 a, b) with little emphasis,except in Shepard et al (1989), on developingplans for the predators. In Malaysia, more thanone species of pest and predator are usuallyfound simultaneously in a rice crop, thus neces-sitating sampling plans for those economicallyimportant species.

In Malaysia, too, concise and precise eco-nomic thresholds and economic injury levelsbased on insect-caused damage functions arenon-existent. In fact, Benedict et al (1989) main-tained that most economic thresholds for insectsare nominal and not based on damage func-tions. Consequently, action thresholds (Hassanand Wilson 1993; Hassan 1997) are used in thisstudy based on point of saturation of infestation,on regressing proportion of infestation againstmean population density (Hassan and Ibrahim1987). Although it is not a substitute for eco-

nomic threshold, it is a reliable numerical spe-cies-specific characteristic featuring inherentspatial-temporal distribution entity of the par-ticular arthropod (Sterling 1976; Taylor 1984;Wilson et al 1989).

In this report, for each arthropod categorystudied, pest or predator, a sequential decisionplan that can be operated by professional pestmanagers and trained farmers was developedbased on positive binomial (presence/absence)distribution. Analyses in another paper (Hassan1996) clearly indicate high fits (r2 > 0.90) onregressing proportions of infestation (P(I)) cal-culated from binomial family distribution mod-els of Wilson and Room (1983) against P(I)observed. The presence/absence recordingscheme is more practical and more efficient,especially for large acreages, than other sequen-tial sampling plans which involve counting ofthe actual arthropods (e.g. plans based on nega-tive binomial and Poisson distributions).

MATERIALS AND METHODS

Data Collection

Data from 204 sampling occasions (= 204 sam-ples, each containing 100 and 40 sampling unitsat the first, and second and third locationsrespectively as described below) were used todevelop the sampling plans. Of the 22 categoriesof arthropod recorded, only 11 categories yieldedsufficient data enabling calculation of thresh-olds and other sampling attributes to be used informulating sampling plans (Hassan and Ibrahim1987). Visual counts of arthropods on a per hillbasis were recorded from three locations: a paddyestate at Bukit Cawi village, Seberang Perak,Perak (4° 7 N, 101° 4'E) (1986), experimentalplots at Universiti Pertanian Malaysia (UPM),Serdang, Selangor (3° 2 N, 101° 42 E) (1992),and a farmer's plots at Sawah Sempadan, TanjungKarang (SSTK), Selangor (3° 20 N, 101° 12 E)(1992). The 11 categories whose data were usedto develop the sampling plans are the pestsNephotettix spp. (Homoptera: Cicadellidae),Nilaparvata lugens (Homoptera: Delphacidae),Cnaphalocrocis medinalis (Guenee) (LepidopteraPyralidae), Recilia dorsalis (Motschulsky)(Homoptera: Cicadellidae), Sogatella furcifera(Horvath) (Homoptera: Delphacidae) and thepredators: Cyrtorhinus lividipennis (Reuter)(Heteroptera: Miridae), Anatrichus pygmaeus(Lamb) (Diptera: Chloropidae), spiders,Odonata, Paederus fusdpes (Curtis) (Coleoptera:

52 PERTANIKAJ. TROP. AGRIC. SCI. VOL. 20 NO. 1, 1997

PRESENCE/ABSENCE SEQUENTIAL PLANS FOR PEST MANAGEMENT DECISION-MAKING

Staphylinidae) and Casnoidea spp. (Coleoptera:Carabidae).

Action Thresholds

Each sample produced mean density (x ) andproportion of infested hill (P(I)) data for eacharthropod category. Values of P(I) were alsogenerated from two distribution models (Wilsonand Room 1983): (1) derivation of negativebinomial, (2) Poisson,

P(I )= l - e

P(I) = 1-

(1)

(2)

where s2 is the variance. These two models yieldedthe highest fit (r2) when for each arthropodcategory, the calculated P(I) were regressedagainst observed P(I) (Hassan and Ibrahim1987). The P(I) values versus x were plottedusing firstly the actual field data, secondly usingthose derived from model 1 and lastly frommodel 2. A polynomial regression up to thefourth power was fitted for each plot. For eacharthropod category, the most fit (usually thefourth order) polynomial equation (highest r2)was then differentiated to obtain a value of xwhere the first dy/dx=0 occurred (point of satu-ration of infestation) (Hassan and Ibrahim,1987). For each arthropod category, the threevalues of x obtained (field data, model 1, model2) were subsequently averaged to obtain themean action threshold in terms of x /hill andhence P(I) (Hassan 1997). These values werethen used in developing the sampling plans.

Development of Plans

The binomial sequential plans were generatedusing the SEQUAN computer program ofTalerico and Chapman (1970). This programenabled optimization of each plan for samplingerror, width of indecision zones and minimumsample size required to make decisions on ac-tion or no-action taken. Action threshold valuesin terms of P(I); Pu and P lower (with respectiveequivalents mx 8c m2; mt is the mean populationdensity at which management treatment isneeded and m2 is the no treatment level) usedwere; 0.96, 0.91 (Nephotettix spp.); 0.99, 0.96 (N.lugens); 0.95, 0.61 (C. medinalis) (Pyralidae); 0.93,

0.92 (R dorsalis); 0.96, 0.95 (5. furdfera); 0.94,0.93 (G lividipennis); 0.92, 0.80 (A. pygmaeus)(Diptera); 0.79, 0.65 (spiders); 0.87, 0.76(Odonata); 0.80, 0.54 (P. fuscipes); and 0.82,0.78 (Casnoidea spp.). Their respective thresh-olds in terms of mean density per hill were; 3.76,2.96 (Nephotettix spp.); 6.33, 6.06 (JV. lugens);1.97, 1.10 (C. medinalis) (Pyralidae); 2.49, 2.33(R dorsalis); 4.04, 3.87 (S. furdfera); 4.01, 3.68(Cyrtorhinus sp.); 2.13, 1.85 (Odonata); 1.73,1.53 (C. lividipennis); 2.39, 0.85 (A. pygmaeus)(Diptera); 1.75, 1.42 (R fusdpes); and 1.58, 1.56(spiders). In the sequential plans, the abovethresholds are converted to uninfestation pro-portion (the reverse of infestation proportion),since in practice it is more efficient to countuninfested sample units (Sterling 1976). Conse-quently, all plans are tailored towards monitor-ing of uninfested samples. The assigned risks(Type I & Type II errors) of making an incor-rect decision a & p respectively were given thevalues of 0.30 each. Equations for calculatingdecision lines (dj and d2) are as follows;

d, = bn - h,

d2 = bn + h2

where dj (smaller number column) is the lowerthreshold of the number of sample unitsuninfested, when the running total of uninfestedsample units is below d( the need for treatmentis indicated, hence dY forms the lower decisionline; d2 (large number column) is the upperthreshold of the number of sample unitsuninfested, when the running total of uninfestedsamples is greater than d2, the status of notneccesary for treatment is indicated, hence d2

forms the lower decision line (between djanddt,is the indecision zone), n is the number ofsample units examined, h( and h2 are the inter-cepts, and b is the common slope of both lines(Waters 1955).

The slope and intercepts are calculated as fol-lows;

b =

m , 1 - i r u

PERTANIKAJ. TROP. AGRIC. SCI. VOL. 20 NO. I, 1997 53

S.T.S. HASSAN AND MM. RASHID

log1 - a

m, [ 1 - m 2

log!h2 =

a

hm, | 1 - m 2

with m] & m2 as defined earlier.

In the implementation of the uninfestedplans, absence of the appropriate arthropodcategory would be recorded as +1 in the run-ning total, whereas a presence would be notedas 0 (zero). The cumulative total would be up-dated and examined continually as more sampleunits are examined.

RESULTS AND DISCUSSION

The presence/absence sequential plans are pre-sented in Table 1. In terms of the equivalentx values, the thresholds used in this paper arelower than those usually used as guidelines forinitiating field treatment (Hassan and Ibrahim1987). As an example, the mean threshold forNilaparuata lugens used here is 6.28 per hill. Inthe Philippines, a threshold of up to 23 N, lugensper hill has been used in formulating sequentialplans (Shepard et al. 1986), although the criticaleconomic injury level may be much lower (2 -5per hill) (Sogawa and Cheng 1979). Similarly inMalaysia, slightly higher values of threshold havebeen proposed for Nephotettix spp. and N. lugens,and other species not considered in this paper(Hassan and Ibrahim 1987). However, counts ofN. lugens per hill of 25 - 40 only occurred inMalaysia under outbreak situations, such as thatobserved in the Tanjung Karang area in 1984/85 (Hassan, unpubl. data). Counts of less than 10per hill are more frequent, hence suitable forsampling based on the positive binomial theory(Sterling and Pieters 1979). Thresholds pro-posed in this paper should then be regarded asprovisional guidelines until better defined onesare available. Nevertheless, since we are propos-ing here the simultaneous usage of pest andpredator plans, the possible inadequacy of theplans for pests due to their conservativeness

would be compensated by the similarly conserva-tive plans for the predators. Shepard et al (1989)compensated for the presence/absence of preda-tors by having different thresholds for binomialplans with and without predators, in N. lugensand S. furdfera. However their plans were aimedat preliminary rather than simultaneous sam-pling of predators.

It is worth noting that the plans presentedare based on uninfested sample units, i.e. oneuninfested sample unit is given the value of +1in the running total and one infested unit thevalue of 0. There is a time-saving advantage inusing this procedure (Sterling 1976). These plansare also designed for pest populations that peakin cycles. Many pests of rice in the tropics, suchas N. lugens, exhibit distinct generations in thefield (Dyck et al 1979). Otherwise, the continu-ous presence of below-threshold level populationsfor extended periods of time may lead to thecumulative survival exceeding action thresholdsand management action may be needed.

Operating the Plans

Since rice is often attacked by a number of pestssimultaneously (Kisimoto 1984), plans for theappropriate species of pests and predators shouldbe used simultaneously. These plans can beprinted on shirt-pocket-sized cards. A set of plansshould be operated to cover not more than 50ha of paddy field at a time. Otherwise accuracyand precision may be sacrificed (Sterling andPieters 1979). The manner of walking throughthe field should be varied on subsequentsamplings to enable reasonable coverage of theentire field. Sampling units chosen for visualinspection should be selected at random andshowed be a fair representative of the popula-tion of plants in the entire field. It is suggestedhere that one should sample a minimum of tenunits at random before making a managementdecision. Since these plans are for presence/absence of infestation, visual examination of thehill can be done quickly. Absence of the appro-priate arthropod category would be recorded as+1 in the running total, whereas a presencewould be noted as 0 (zero). The cumulativetotal would continually be compared with thecontrol and no-control column totals.

Prior to commencing the sampling opera-tion, the operator fills in the concomitant infor-mation e.g. locality, plot, date and growth stageof the crop, in the sampling plan cards. After



TABLE 1Positive binomial (presence-absence) sequential sampling plans for (i) Nephotettix spp., (ii) Nilaparvata lugens,

(iii) Cnaphalocrocis medinalis (Pyralidae),(iv) Recilia dorsalis, (v) Sogatella furrifera, (vi) Cyrtorhinus lividipennis,(vii) Anatrichus pygmaeus (Diptera), (viii) Spiders, (ix) Odonata, (x) Paederus fuscipes

(ix) Casnoidea sp., with a and p each equal to 0.30

(0 (ii)

+VE BINOMIAL, INFESTATION • 0, NO INFESTATION • +1

UX.'Al XT Y . DATE : UNINFRSTAT1OX

f t o r : riANTCRowmsTAGB T W E S I I O L D •.

PEST : Ntphouttix ipp. a -0.3 p - 0.3 Upper: 4% Lower : 9 *

I M M IMUMttt

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

com*.

0

1

2

3

4

5

6

7

8

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

tUKftftttW A I

MOeofmoL

2

3

4

5

6

7

8

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

23

24

UttftJHUMtt

26

27

28

29

30

31

32

33

34

35

36

37

38

39

40

41

42

43

44

45

46

47

48

49

50

CftKTlOl

24

25

26

27

27

28

29

30

31

32

33

34

35

36

37

38

39

39

40

41

42

43

44

45

46

mcownoL

25

26

27

28

29

30

31

32

33

34

35

36

37

38

39

39

40

41

42

43

44

45

46

47

48

+ VE BINOMIAL, INFESTATION • 0t NO INFESTATION = +1

LOCALITY : DATB : UTflKTBTTATlONnot : PLAKTO*O*THJTAGI- TnftttlKMl) :

PEST : NiUparvata ip. a - 0.3 p » 0.3 Upper : 1* Lower : 4%

ftMNUMIAMI •

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

catrncn.

0

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

W A L cowriM.

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

UMTUL

26

27

28

29

30

31

32

33

34

35

36

37

38

39

40

41

42

43

44

45

46

47

4 8

49

50

ctmoi

25

26

27

28

29

30

31

32

33

34

35

36

37

38

39

40

41

42

43

44

45

46

4 6

47

48

T0T*L COMTtOL

26

27

28

29

30

31

32

33

34

35

36

37

38

39

40

41

42

43

44

45

46

46

47

48

49

PERTANIKAJ. TROP. AGRIC. SCI. VOL. 20 NO. 1, 1997 55

S.T.S. HASSAN AND MM. RASHID

Table 1 (Contd)

(iii) (iv)

+VE BINOMIAL, INFESTATION * 0, NO INFESTATION « +1

LOCALITY DATB | UMHreTTATIQNfUft 1 rLAKTOttOwm 8TAGB •. 71WB5HOU) :

PEST : Py«lid»e a - 0,3 P - 0.3 Upper 15* Loww: 39*

MMru

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

IS

19

20

21

22

23

24

25

comoi.

o1

2

3

4

5

5

6

7

8

9

10

10

I)

12

13

14

14

15

16

17

18

19

19

20

******VMM.

MCOKTtOt.

1

2

3

4

4

5

6

7

8

9

9

10

11

12

13

14

14

15

16

17

18

18

19

20

21

UMMJtKlfMttt

26

27

28

29

30

31

32

33

3 4

35

3 6

3 7

38

39

40

41

42

43

44

45

46

47

48

49

50

21

22

2 3

24

2 4

25

2 6

2 7

28

28

29

30

31

32

33

33

34

35

3 6

37

38

38

3 9

4 0

41

>UNHXP

MMco&oc

22

23

23

24

25

26

27

28

2 8

2 9

3 0

31

32

32

33

34

35

36

37

37

38

39

4 0

41

4 2

+VE BINOMIAL, INFESTATION = 0, NO INFECTION * +1

JJXALSTt : DAT* : UMNPBlrrATKJN

tun : turn WIOWTH .nAOE THRESHOLD S

PEST : RttiUatbnalU a -0 .3 p -0 ,3 Upper:7* Low«r:$*

MMtlMMM

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

cmmaL

-

-

-

0

I

2

2

3

4

5

6

7

8

9

10

11

12

13

14

14

15

16

17

•wmn«cVMM

noCONTBOl

7

8

9

9

10

11

12

13

14

15

16

17

18

19

20

21

21

22

23

24

25

26

27

28

29

MMKJ

26

27

28

29

30

31

32

33

34

35

3 6

37

38

3 9

4 0

41

42

43

44

45

46

47

4 8

4 9

50

coK-not

18

19

20

21

22

2 3

2 4

2 5

2 6

26

27

28

2 9

30

31

32

33

34

35

36

37

37

38

39

4 0

TOtU COKT1M.

30

31

32

33

33

34

35

36

3 7

3 6

3 9

40

41

42

43

44

44

45

46

47

48

49

50

51

52


PRESENCE/ABSENCE SEQUENTIAL PIANS FOR PEST MANAGEMENT DECISION-MAKING

Table 1 (Cont'd)

(v) (vi)

+VE BINOMIAL, INFESTATION • 0, NO INFESTATION • +1

LOCALITY : OAT! ; UNTNrjWTATfON

vucrr : njwroaowTHSTAOE THRESHOLD :

PEST : Sogat€lia tp. a * 0.3 p*0.3 Upper: 4% Lower 5 *

MMK4MMtM

1

2

3

4

5

6

7

8

9

10

II

12

13

14

15

16

17

18

19

2 0

21

22

2 3

24

25

comet.

0

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

2 0

2 1

21

T0T*tM

cotmot

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

13

19

20

21

22

22

2 3

24

25

26

t M N UMMMM

2 6

2 7

2 8

2 9

3 0

31

32

3 3

3 4

3 5

3 6

3 7

3 8

3 9

4 0

41

4 2

4 3

4 4

45

4 6

4 7

4 8

49

5O

covnot

22

2 3

24

2 5

2 6

27

28

2 9

3 0

31

32

33

34

3 5

3 6

3 7

3 8

3 9

4 0

41

4 2

4 3

4 3

44

4 5

• - « ! % . . M

27

28

29

3 0

31

32

3 3

3 4

35

3 6

3 7

3 8

39

4 0

41

42

4 3

4 4

4 4

45

4 6

4 7

4 8

4 9

5 0

+VE BINOMIAL, INFESTATION * 0, NO INFESTATION =

UMNPR3TAT10WTHM3M0U)

PEST I Cyrtorhitu* *p. a -0 ,3 $ - 0.3 Upper : 6% Lower ; 7%

JAMrtSwtMiU

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

2 0

21

22

2 3

2 4

25

COMTlOt

-

0

1

2

3

4

5

5

6

7

8

9

10

11

12

13

14

15

16

17

18

*bXtt*«TOTAL

NO

7

8

9

10

10

11

12

13

14

15

16

17

18

19

2 0

21

22

2 3

2 4

2 5

2 5

2 6

27

2 8

2 9

uMnjKVM»M

26

27

28

2 9

3 0

31

32

3 3

34

35

3 6

37

3 8

39

40

41

4 2

4 3

4 4

4 5

46

47

4 8

49

5 0

COWTJQL

19

20

2 0

2 1

2 2

2 3

24

25

2 6

2 7

2 8

2 9

3 0

31

3 2

3 3

3 4

35

35

3 6

3 7

3 8

3 9

40

41

toCOKTtOL

3 0

31

32

3 3

34

3 5

3 6

3 7

3 8

3 9

4 0

4 0

41

42

4 3

4 4

45

4 6

4 7

4 8

4 9

50

51

5 2

5 3


S.T.S. HASSAN AND M.M. RASHID

Table 1 (Cont'd)

(vii) (viii)

+VE BINOMIAL, INFESTATION = 0» NO INFESTATION = +1

PREDATOR :Dipu*i o«0.3

UNWPBSTATION!I STAGE: THRE51KXX>

p-0.3 Upper :7%

KlftttM

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

16

19

20

21

22

23

24

25

CCMTkOL

0

1

2

3

4

4

5

6

7

8

9

10

11

11

12

13

14

15

16

17

18

18

19

20

21

l u m M CTOTAL

*©CONTtOl

2

2

3

4

6

6

7

8

9

9

10

11

12

13

14

15

16

16

17

18

19

20

21

22

23

uwrutKtM*Ct

26

27

28

29

30

31

32

33

34

35

36

37

38

39

40

41

42

43

44

45

46

47

48

49

50

cotmoL

22

23

24

25

25

26

27

28

29

30

31

32

32

33

34

35

36

37

38

39

39

40

41

42

43

• UMNIMTOTAL COXTkOl

23

24

2 5

26

27

28

29

30

30

32

32

33

34

35

3 6

37

37

38

39

40

41

42

43

44

44

+VE BINOMIAL, INFESTATION * 0, NO INFESTATION «+/

LOCAUTY : DAT* : UNIKPOTATKIHK J O T : fLAXT OHOWTH STAOB : TMWWH0LD 1

PREDATOR ; Spidoi a » OJ p • 0.3 Upper : 21 % Lower ! 33*

•MM!MMM

1

2

3

4

5

6

7

8

9

10

II

12

13

14

15

16

17

18

19

20

21

22

23

24

25

COXTtOt

-

0

1

2

2

3

4

5

5

6

7

7

8

9

10

10

11

12

12

13

14

15

15

16

17

T0T*LXO

CCMTKH.

2

3

3

4

6

6

7

8

8

9

10

11

11

12

13

13

14

15

16

16

16

17

18

19

19

•AHftJMUMItt

26

27

28

29

3 0

31

32

3 3

34

35

36

37

38

39

40

41

42

43

44

45

46

47

48

49

50

18

18

19

20

20

21

22

23

23

24

25

25

26

27

28

28

29

30

31

31

32

33

33

34

35

wr*LMO

catmot

20

21

21

22

23

24

24

25

26

26

27

28

29

29

30

31

32

32

33

34

34

35

36

37

37

58 PERTANIKA J. TROP. AGRIC. SCI. VOL. 20 NO. 1, 1997


Table 1 (Cont'd)

(ix) fx)

+VE BINOMIAL, INFESTATION =: 0, NO INFESTATION = +1

LOCALITY : DATS : U W M T A T I W

run : n ANT OHOWTH STAGS : nouumou>

PREDATOR rCWoruu a-0.3 £ - 0 . 3 Upper: 13% Lower: 24%

Ntt1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

2 0

21

22

23

24

25

cwrtot

0

0

1

2

3

4

5

5

6

7

3

9

9

10

11

12

13

14

14

15

16

17

18

18

19

•«¥ff MO

cantot.

2

3

4

4

5

6

7

&

9

10

11

12

13

13

14

15

16

17

18

18

19

20

21

22

22

**Mtt4lNUtflM

26

27

28

29

3 0

31

3 2

33

34

35

36

37

38

39

40

41

42

43

44

45

46

47

48

49

50

ewnot

20

21

22

23

23

24

25

26

27

27

28

29

30

31

32

32

33

34

35

36

36

37

38

39

40

•BSff •t

22

23

24

25

26

27

27

28

29

30

31

31

32

33

34

35

36

36

37

38

39

40

40

41

42

+VE BINOMIAL, INFESTATION = 0, NO INFESTATION = +1

L0CAUTY: DATS : UWWMTATKW

run- : FLAW GROWTH STAGS I TKMSHOU) I

PREDATOR :?a*<Ums*p. a-0.3 3 -0 .3 Upper : 20* Lower : 46*

I

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

2 3

24

25

cawnoi.

0

1

1

2

3

3

4

5

5

6

7

7

8

9

10

10

11

12

12

13

14

14

15

16

16

•B3ff woCOWTtOi.

1

2

3

3

4

5

5

6

7

7

8

9

10

10

11

12

12

13

14

14

15

16

16

17

18

26

27

28

29

3 0

31

32

33

34

35

36

37

38

39

40

41

4 2

4 3

4 4

4 5

46

47

48

49

50

corntK.

17

18

18

19

20

20

21

2 2

22

2 3

24

24

25

26

27

27

28

29

29

30

31

31

32

33

33

MO

18

19

20

2 0

21

22

22

23

24

25

25

26

27

27

28

29

29

30

31

31

32

33

33

34

35

PERTANIKA J. TROP. AGRIC. SCI. VOL. 20 NO. 1, 1997 59


Table 1 (Cont'd)

(xi)

+VE BINOMIAL, INFESTATION = 0, NO INFESTATION = +1

LOCALITY : DAT* : UMNFBITAT1OHPUTT : PLAIT OMOWTH I T A O B TtMSHOLD :

PREDATOR : Casnoidea ip. o - 0.3 p - 0.3 Upper : 1 S% Lower : 2 2 *

HUM H I

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22.

23

24

25

MMMI

-

-

0

1

2

2

3

4

5

6

6

7

8

9

10

10

11

12

13

14

14

15

16

17

iss? 90

4

5

6

6

7

8

9

10

10

11

12

13

14

14

15

16

17

18

18

19

20

21

22

23

23

MM26

27

28

29

30

31

32

33

34

35

36

37

38

39

40

41

42

43

44

45

46

47

48

49

50

comot

18

18

19

20

21

22

22

23

24

25

26

26

27

28

29

30

31

31

32

33

34

35

35

36

37

•ssr MCOKTlOt

24

25

26

27

27

28

29

30

31

31

32

33

34

35

35

36

37

38

39

39

40

41

42

43

43

60 PERTANIKA J. TROP. AGRIC. SCI. VOL. 20 NO. 1,1997


examining ten sample units, the running totalshould be compared with the control and nocontrol columns. If the running total equals oris less than the figure in the control column,stop sampling for that particular arthropod cat-egory, and management action should be con-sidered. On the contrary, if the running totalequals or is more than the figure in the nocontrol column, stop sampling; no treatment isneeded against that particular pest. A similaroperational procedure is performed for thepredators, though in this case the respectivedecisions would be classifying the populationlevel of a particular predator as high or low.However, in both plans for pest and predators,if no decision can be made even at the 50thsampling unit, stop sampling and sample againtwo or three days later. After completion ofsampling a particular area, the need for man-agement action against some particular pest spe-cies should be weighted by the levels of preda-tors' populations. In the presence of high levelsof predator populations, control actions to sup-press pest populations may be left to the naturalpredators. To date there is ample evidence indi-cating the regulatory role of natural enemies inpaddy fields (e.g. Kenmore et al 1984; Heong etal 1991; Way and Heong 1994).

Advantages of Presence/absence PlansIt is proven that a good pest management sys-tem can be achieved by using a sampling tech-nique that is rapid, reliable, reproducible, real-istic, and has a low level of risk (Sterling andPieters 1979). Presence/absence plans can meetthese criteria. In addition, binomial sampling isthe most feasible field sampling method formany organisms (Binns and Nyrop 1992); also,it is usually faster and therefore less costly on aper-sample unit basis (Wilson 1982). Moreover,the sequential sampling probabilities a and (3can be preset as desired, especially consideringthe damaging status of a particular pest. The aand p levels presented here each equals 0.30.This 30% level of risk has been practically ac-ceptable in management of cotton pests in Aus-tralia (Sterling 1976; Wilson 1982). Lower levelsof risk entail taking larger number of samples.Although various researchers have further con-vincingly elaborated on the practicality, espe-cially for small arthropods since no actual count-ing is necessary, and on the virtues of binomialsampling plans (e.g. Shepard et al. 1989; Wilson

et al 1989;), the plans presented need to betested in Malaysian field conditions to deter-mine its practicality, validity and reproducibility.

ACKNOWLEDGEMENTS

We are most grateful to the Bukit Cawi PaddyEstate, Seberang Perak, Perak, and the FarmDivision, Universiti Putra Malaysia, and thefarmer at Sawah Sempadan, Tanjung Karang,for their various assistance and cooperation.Research grant was provided by the MalaysianNational Scientific Research and DevelopmentCouncil (IRPA Project 1-07-05-064).

REFERENCES

BENEDICT, J.H., K.M. EL-ZIK, L.R. OLIVER, P.A. ROBERTS

and L.T. WILSON. 1989. Economic injury levelsand thresholds for pests of cotton. In IntegratedPest Management and Cotton Production, ed. R.E.Frisbie, KM. El-Zik and L.T. Wilson, p. 121-154. New York: Wiley.

BIANCI, G., J. BAUMGARTNER, V. DELUCCHI, N.

RAHALILIVAVOVOLONA, S. SKILLMAN and P.H.

ZAHNER. 1989. Sampling egg batches ofMaliarpha separatella Rag. (Lep., Pyralidae) inMadagascan rice fields. Trop. Pest Management35(4): 420-424.

BINNS, M.R. and J.P. NYROP. 1992. Sampling insectpopulations for the purpose of IPM decisionmaking. Ann. Rev. EntnmoL 37: 427-453.

DYCK, V.A., V.C. MISRA, S. ALAM, C.N. CHEN, C.Y.

HSIEH and R.S. REJESUS. 1979. Ecology of thebrown planthopper in the tropics. In BrownPlanthopper: Threat to Rice Production in Asia,p. 61-98. Manila, Philippines: International RiceResearch Institute.

FERRER, E.R. and B.M. SHEPARD. 1987. SamplingMalayan black bugs, Scotinophra coarctata F.(Heteroptera: Pentatomidae) in rice. Environ.Entomol 16: 259-263.

HASSAN, S.T.S. 1996. Population and distributionparameters of arthropods of wet paddy ecosys-tem, and their fits to distribution models.Malays. Appl Bioi 25(2): 61-68.

Hassan S.T.S. 1997. Action thresholds of wet ricearthropods for pest management decision-making in Malaysia. PertanikaJ. Trop. Agric. Sri.20(1): 65-74.

HASSAN, S.T.S. and R.C. IBRAHIM. 1987. Satu kaedahbaru menentukan nilai ambang tindakan se-rangga perosak tanaman padi. Pertanika 10(3):381-383.



HASSAN, S.T.S. and L.T. WILSON. 1993, Simulatedlarval feeding damage patterns of Heliothisarmigera (Hiibner) and H. punctigera (Wal-lengren) (Lepidoptera: Noctuidae) on cottonin Australia. International Journal of Pest Man-agement 39(2): 239-245.

HEONG, K.L., G.B. AQUINO and A.T. BARRION. 1991.

Arthropod community structures of rice eco-systems in the Philippines. Bull Entomoi Res.81: 407-416.

KENMORE, P.E., F.O. CARINO, C.A. PEREZ, V.A. DYCK

and A.P. GUTIERREZ. 1984. Population regula-tion of the rice brown planthopper (Nilaparvatalugens) within rice fields in the Philippines. /PL ProL Tropics 1: 19-37.

KISIMOTO, R. 1984. Insect pests of the rice plant inAsia. ProL Ecoi 7: 83-104.

KUNO, E. 1977. Distribution pattern of the ricebrown planthopper and sampling techniques.In The Rice Brown Planthopper. Taipei, Taiwan:Food and Fertilizer Technology Center for theAsian and Pacific Region.

KUNO, E. 1986. Evaluation of statistical precisionand design of efficient sampling for the popu-lation estimation based on frequency of occur-rence. Res. Pop. Ecol 28: 305-319.

NISHIDA, T. and T. TORII. 1970. A handbook of fieldmethods for research on rice stemborers andtheir natural enemies, IBP Handbook No. 14.London: IBP.

ONSAGER, J.A. 1976. The rationale of sequentialsampling, with emphasis on its use in pestmanagement. USDA Agr. Res. Serv. Tech. Bull1526.

Ooi, P.C. 1982a. A surveillance system for riceplanthoppers in Malaysia. In Proc. Int. Conf PLProL in Tropics, pp. 551-565. Kuala Lumpur:MAPPS.

Ooi, P.C. 1982b. Attempts at forecasting ricep l an thoppe r popula t ions in Malaysia.Entomophaga 27 (Special Issue): 89-98.

Ooi, P.C. and K.L. HEONG. 1988. Operation of abrown planthopper surveillance system in theTanjung Karang Irrigation Scheme in Malay-sia. Crop Prot. 7: 273-278.

PIETERS, E.P. 1978. Bibliography of sequential sam-pling plans for insects. Bull. Entomoi. Soc. Amer.24: 372-374.

PLANT, R.E. and L.T. WILSON. 1985. A Bayesian

method for sequential sampling and forecast-

ing in agricultural pest management. Biometrics41: 203-214.

ROTHROCK, M.A. and W.L. STERLING. 1982. Sequen-

tial sampling for arthropods of cotton: its ad-vantages over point sampling. Southwestern En-tomologist 7(2): 70-80.

SHEPARD, B.M., E.R. FERRER, P.E. KENMORE and J.P.

SUMANGIL. 1986. Sequential sampling:planthoppers in rice. Crop Prot. 5(5): 319-322.

SHEPARD, B.M., E.R. FERRER and P.E. KENMORE. 1988a.

Sequential sampling of planthoppers andpredators in rice. / PL ProL Tropics 5: 39-44.

SHEPARD, B.M., D.R. MINNICK, J. SORIANO, E.R. FERRER

and O. MAGISTRADO. 1988b. A simple peg-boardfor use in sequential sampling leaffolders,planthoppers and major predators. Int. RiceRes. NewsL 13(6): 40-41.

SHEPARD, B.M., E.R. FERRER, J. SORIANO and P.E.

KENMORE. 1989. Presence/absence sampling ofplanthoppers and major predators in rice. / .PL ProL Tropics 6(2): 113-118.

SOGAWA, K. and C.H. CHENG. 1979. Economic thresh-olds, nature of damage, and losses caused bythe brown planthopper. In Brown Planthopper:Threat to Rice Production in Asia, p. 125-142.Manila, Philippines: International Rice Re-search Institute.

SONG, Y.H , K.L. HEONG, A. LAZARO and K.S. YEUN.

1992. Use of Geographical Information Sys-tems in analyzing large area movement anddispersal of rice insects in South Korea. InProceedings of Workshop on the Role of Pest Surveil-lance Systems in Rice Pest Management Decisions,MADA Headquarters, Alor Setar, Malaysia, 21December 1992.

STERLING, W. L. 1976. Sequential decision plans forthe management of cotton arthropods in south-east Queensland. Aust. J. Ecol 1: 265-274.

STERLING, W.L. and E.P. PIETERS. 1979. Sequential

decision sampling. In Economic Thresholds andSampling of Heliothis Species on Cotton, Corn,Soybeans and Other Host Plants, p. 85-101. USA:Southern Cooperative Series Bulletin 231.

TALERICO, R.L. and R.C. CHAPMAN. 1970. SEQUAN. A

computer program for sequential analysis, p. 1-6.USA: USDA For. Serv. Res. Note NE-116.

TAYLOR, L.R. 1984. Assessing and interpreting thespatial distributions of insect populations. Ann.Rev. Entomoi 29: 321-327.

WATERS, W.E. 1955. Sequential sampling in forestinsect surveys. Forest Sci. 1: 68-79.



WAY, MJ. and K.L. HEONG. 1994. The role of

biodiversity in the dynamics and managementof insect pests of tropical irrigated rice - areview. Bull. Ent. Res. 84: 567-587.

WILSON, L.T. 1982. Development of an optimal

monitoring program in cotton: emphasis onspidermites and Heliothis spp. Entomophaga 27:45-50.

WILSON, L.T. and P.M. ROOM. 1983. Clumping pat-

terns of fruit and arthropods in cotton, with

implications for binomial sampling. Environ.EntomoL 12: 50-54.

WILSON, L.T., W.L. STERLING, D.R. RUMMEL and J.E.

DEVAY. 1989. Quantitative sampling principlesin cotton IPM. In Integrated Pest ManagementSystems and Cotton Production, ed. R.E. Frisbie,K.M. El-Zik and L.T. Wilson, p. 85-120. NewYork: Wiley.

(Received 3 July 1996)(Accepted 22 May 1997)


Documents

Presence/absence Sequential Plans for Pest Management …pertanika.upm.edu.my/Pertanika PAPERS/JTAS Vol. 20 (1) Apr. 1997/07... · PRESENCE/ABSENCE SEQUENTIAL PLANS FOR PEST MANAGEMENT