ARTICLE IN PRESS

0261-2194/$ - se

doi:10.1016/j.cr

�CorrespondE-mail addr

Crop Protection 26 (2007) 1031–1035

www.elsevier.com/locate/cropro

Integrated weed management practices in garlic crop in Pakistan

Tariq Mehmood, Khalid Mahmood Khokhar�, Muhammad Shakeel

Vegetable Crops, HRI, NARC, Islamabad, Pakistan

Accepted 24 September 2006

Abstract

Weed management studies in garlic crop were conducted during 2000–2001 and 2002–2003 at the National Agricultural Research

Center, Islamabad Pakistan. During first year, pendimethalin was sprayed at 0.80 l a.i. ha�1 pre emergence 2 days after first irrigation in

moist condition followed by different mechanical weeding regimes. During second year, pendimethalin, oxadiazon, glyphosate, and

metribuzin were sprayed at 0.80 l, 0.25 l, 0.6 l, and 0.45 kg a.i. ha�1, respectively, pre emergence 2 days after first irrigation in moist

conditions followed by one hoeing at 80 days after herbicide application. Herbicide treatments were compared with weed free and weedy

check. The most dominant weed species were Phalaris minor, Cronopus didymus, Medicago denticulata, and Rumex dentatus. All herbicide

treatments followed by hoeing except metribuzin gave bulb yield at par with weed free treatment. Metribuzin resulted in minimum bulb

yield (0.59 t ha�1) because of its extreme phytotoxicity to garlic crop, which resulted in the survival of a few plants. Pendimethalin in

combination with manual hoeing gave the highest bulb yield and monetary returns.

r 2006 Elsevier Ltd. All rights reserved.

Keywords: Pendimethalin; Oxadiazon; Glyphosate; Metribuzin; Weed fresh biomass; Bulb yield; Garlic; Pakistan

1. Introduction

Garlic (Allium sativum L.) is very popular and grown inall parts of Pakistan. It is widely used after onion andconsidered as a valuable spice for food. It is cultivated overan area of 6586 ha producing 55.9,000 t at an average yieldof 8.49 t ha�1 (Anonymous, 2004). Garlic has very littlecanopy, which necessitates weeding as an importantcultural practice for higher marketable bulb yield. Manualweeding is tedious, expensive and often damages the crop.Qasem (1996) studied the effect of 18 herbicides on weedcontrol in garlic in Jordan and claimed that post-emergence application of oxidiazon at 0.24 l a.i. ha�1 at3–4 leaf stage gave better yield as compared to weed freecrop.

Vora and Mehta (1998) studied the efficacy of differentherbicides for weed control in garlic and observed highestbulb yield in weed free treatment while herbicide applica-tion alone did not control weeds effectively. However,herbicides application followed by one hand hoeing gave

e front matter r 2006 Elsevier Ltd. All rights reserved.

opro.2006.09.013

ing author.

ess: tokmahmud@yahoo.com (K.M. Khokhar).

the best results (Vora and Mehta, 1999). Tewari et al.(1998) compared the efficacy of different weedicides with orwithout manual weeding and reported highest net mone-tary returns with hand weeding followed by pendimethalinat 0.80 l a.i. ha�1 supplemented with one hand hoeing.Sandhu et al. (1997) conducted a trial on the efficacy ofherbicides for the control of weeds in garlic and claimedthat pendimethalin with one hoeing after 105 days ofsowing resulted in significant increase in bulb yieldcompared to weed free control. Mahmood et al. (2002)studied efficacy of two herbicides i.e. pendimethalin at0.80 l a.i. ha�1 and oxadiazon at 0.25 l a.i. ha�1 were testedand compared with weed free (four manual hoeing) andweedy check. They reported that single application of anyof the herbicides was not sufficient to obtain yield equal toweed free treatment.In Pakistan, chemical weed control in garlic has received

little attention and weeds are mostly managed manually,which is expensive. Sometimes due to shortage of labor andunexpected rains, hand weeding or mechanical weedoperations are delayed or left altogether. In such situations,herbicides offer the most practical, effective, and economic-al method of weed control for increasing crop yield.

ARTICLE IN PRESST. Mehmood et al. / Crop Protection 26 (2007) 1031–10351032

The present investigation was undertaken to evaluate thecomparative efficacy of four herbicides alone and incombination with different mechanical weeding regimesin garlic.

2. Materials and methods

The experiments were conducted at Vegetable CropResearch Programme, National Agricultural ResearchCenter, Islamabad, and Pakistan during 2000–2001 and2002–2003. The experiment was laid out in randomizedcomplete block design with three replications. The plot sizewas 5.0m2. Cloves of garlic variety ‘Lehson Ghulabi’ wereplanted in the field in the second week of October duringboth seasons, keeping inter and intra row spacing of 10 and25 cm, respectively. The crop was immediately irrigatedafter planting. During 2000–2001 and 2002–2003, sixtreatments (see Tables 1 and 2) were assessed forgermination, weed biomass, bulb weightm�2, number ofbulbsm�2, and bulb yield t ha�1. Pendimethalin, oxidia-zon, and glyphosate at 0.80, 0.25, and 0.6 0 l a.i. ha�1 whilemetribuzin at 0.45 kg a.i. ha�1 respectively, were sprayedpre-emergence 2 days after first irrigation in moistcondition.

Table 1

Yield and other parameters of garlic crop as affected by different weed contro

Treatments Germination

(%)

Fresh biomass

of weeds

kgm�2

Number

bulbs m�

Pendimethalin alone 88.67 NS 1.000 b 30.18 a

Pendimethalin+1

hoeing

86.67 0.054 c 31.93 a

Pendimethalin+2

hoeing

86.67 0.020 c 32.58 a

Pendimethalin+3

hoeing

85.00 0.010 c 32.58 a

Weed-free 84.67 0.010 c 31.08 a

Weedy check 81.00 3.290 a 15.83 b

Values followed by the same letters in the respective column do not differ sig

Table 2

Yield and other parameters of garlic crop as affected by different weed contro

Treatment Germination

(%)

Fresh biomass

of weeds

kgm�2

Average

weight (g

Oxadiazon+1 hoeing 84.07 NS 0.467 b 29.25 a

Glyphosatee+1 hoeing 83.52 1.067 b 26.67 a

Pendimethalin+1

hoeing

80.81 0.733 b 31.97 a

Metribuzin+1 hoeing 87.04 0.801 b 29.50 a

Weed-free 84.44 0.333 c 31.03 a

Weedy check 87.04 3.100 a 12.52 b

Values followed by the same letters in the respective column do not differ sig

In weed free treatment, first manual hoeing was doneafter 25 days of crop sowing while rest of three manualhoeing were done at 30 days interval to keep the crop weedfree for entire growth period. The recommended doses ofNPK at 100:90:60 kg ha�1 in the form of urea, single superphosphate and potassium–sulphate, respectively, wereapplied. Full dose of phosphorus and potash along withhalf amount of nitrogen was applied at the time of landpreparation while remaining half dose of nitrogen wasapplied at six-leaf stage of the crop.The data were subjected to analysis of variance and the

test of significance was applied following Duncan’s multi-ple range test (Steel and Torrie, 1980).

3. Results and discussion

3.1. Crop germination (%)

Germination percentage in all treatments was recordedafter 15 days of sowing (Tables 1 and 2). The germinationvaried from 81% to 89% during 2000–2001 and 81% to94% during 2002–2003. However, there was no significantdifference in germination percentage in all treatmentsduring both the years. These results indicate that

l methods in garlic crop (2000–2001)

of2

Average bulb

weight (g)

Yield

(t ha�1)

Yield increase

over weedy

check (t ha�1)

Percent

increase

29.86 b 7.29 b 4.12 130

42.30 a 13.38 a 10.21 322

42.66 a 13.83 a 10.66 336

43.49 a 14.17 a 11.00 347

42.39 a 12.21 a 9.04 285

21.36 c 3.170 c � �

nificantly at 5% level of significance according to DMR test.

l methods in garlic crop (2002–03)

bulb

)

Number of

bulbs m�2Yield

(t ha�1)

Yield increase

over weedy

check (t ha�1)

Percent

increase

30.26 a 8.85 a 7.32 479.57

27.74 a 7.39 a 5.87 384.41

29.74 a 9.51 a 7.98 522.79

2.0 c 0.59 b �0.93 �61.17

30.94 a 9.60 a 8.073 528.68

12.20 b 1.52 b � �

nificantly at 5% level of significant according to DMR test.

ARTICLE IN PRESST. Mehmood et al. / Crop Protection 26 (2007) 1031–1035 1033

pre-emergence application of all herbicides did not affectthe germination of the garlic crop but in case of metribuzin,after 25 days of application, the leaves of the plants starteddrying and only a few survived.

3.2. Phytotoxicity

The phytotoxic effect was visually observed regularlyafter the application of pendimethalin, oxadiazon, glypho-sate and metribuzin at 0.80 l, 0.25 l, 0.6 l, and 0.45 kga.i. ha�1, respectively, when used before emergence 2 daysafter first irrigation in moist condition. All doses of theherbicides except metribuzin caused no phytotoxicity ongarlic plants as no burning of leaves and mortality wasobserved. Similarly, Qasem (1996), Vora and Mehta(1998), Tewari et al. (1998) and Mahmood et al. (2002)reported that pendimethalin at 0.7–0.80 l a.i. ha�1 andoxadiazon at 0.25 l a.i. ha�1 were non-toxic to garlic crop.The extreme phytotoxic effect of metribuzin at 0.45 kg a.i.ha�1 was observed after 25 days of application as leaves ofgarlic plants began to dry and ultimately all the plants died.

3.3. Weed prevalence

The experimental field was infested mostly with broad-leaves weeds. Based upon visual observation, the mostdominant weed species were Phalaris minor, Cronopus

didymus, Medicago denticulata, and Rumex dentatus

(Table 3). Chenopodium species, Euphorbia helioscopia,and Fumaria indica were medium in prevalence. Weedprevalence of Avena fatua, Malva neglecta, Melilotus indica,and Papaver somniferum was very low.

3.4. Weed fresh biomass (kg)

All the weed control treatments proved effective inreducing the weed population and fresh weight comparedwith weedy check (Tables 1 and 2). In different weedcontrol treatments, manual hoeing over the application ofherbicide resulted in lower fresh biomass of weeds as

Table 3

Weed species found in garlic crop

Common name Botanical name Weed prevalence

Wild oats Avena fatua L. Low

Canary grass Phalaris minor Retz. High

Common lambsquarters Chenopodium album L. Medium

Missouri goosefoot Chenopodium murale L. Medium

Swine cress Coronopus didymus L. High

Sun spurge Euphorbia helioscopia L. Medium

Fumitory Fumaria indica Pugsley. Medium

Bur clover Medicago denticulata

Willd.

High

Sheeps sorrel Rumex dentatus L. High

Common mallow Malva neglecta Wallr. Low

Indian sweet-clover Melilotus indica L. Low

Common poppy Papaver somniferum L. Low

compared to herbicide application alone followed by nohoeing. Vora and Mehta (1998) also reported thatherbicide application alone did not control weeds affec-tively. During 2000–2001, each additional hoeing after theapplication of pendimethalin resulted in further reductionin fresh biomass of weeds. Control plots having no weedcontrol produced maximum fresh biomass (3.29 and3.10 kgm�2) during 2000–2001 and 2002–2003, respec-tively. Similarly, Khokhar et al. (2006) reported thatmanual hoeing after application of herbicide reduced freshbiomass of weeds in onion crop. The variability in weedfresh biomass in different herbicide treatments may beattributed to the fact that some herbicide treatments weremore effective for weed control than the other was. Thehighest fresh weed biomass was recorded in weedy checkbecause of the prolonged growth period of weeds.

3.5. Number of bulbs (m�2)

The weed control methods significantly affected thenumber of bulbs m�2 (Tables 1 and 2). The minimumnumber of bulbs (2.0m�2) was observed with the applica-tion of metribuzin at 0.45 kg a.i. ha�1. After 25 days ofmetribuzin application, plants started drying and ulti-mately died, significantly reducing the number of bulbs. Inweedy check, the number of bulbs recorded during2000–2001 and 2002–2003, were 15.83 and 12.52m�2,respectively. The decrease in number of bulbs m�2 in weedycheck could be due to the weed competition for the entiregrowth period of garlic crop. The maximum number ofbulbs m�2 was recorded in weed free and herbicidetreatments during both the years, as there was lesscompetition between crop plants and weeds due to lessnumber of weeds.

3.6. Average bulb weight (g)

In all weed control treatments, higher bulb weight wasrecorded as compared to weedy control during both theyears (Tables 1 and 2). During 2000–2001, significantly lessbulb weight (29.86 g) was recorded in the treatment wherependimethalin alone was used followed by no hoeing.However, pendimethalin suppressed weeds growth butlater on the weeds, which survived the spray, began toflourish again and competed with the crop plants. Theminimum average bulb weight of 21.36 and 12.52 g wasrecorded in weedy check during 2000–2001 and 2002–2003,because of the prolonged growth period of weeds thatremained in the field until the time of harvest.

3.7. Bulb yield (t ha�1)

The minimum bulb yield (0.59 t ha�1) was recorded whenmetribuzin at 0.45 kg a.i. ha�1 was used (Table 2). Thislowest yield is due to extreme phytotoxicity of metribuzinto garlic crop, which resulted in the survival of a few garlicplants until harvest. During 2000–2001, significantly lower

ARTICLE IN PRESS

Table 4

Economic analysis (2000–2001)

Treatments Gross income (US

$ ha�1)

Value of increased yield

(US $ ha�1)

Variable cost

(US $ ha�1)

Increased net benefit over

weedy control (US $ ha�1)

Pendimethalin alone 2431 1374 23 1351

Pendimethalin+1 hoeing 4460 3403 51 3352

Pendimethalin+2 hoeing 4610 3553 79 3474

Pendimethalin+3 hoeing 4723 3666 65 3601

Weed-free 4070 3013 112 2901

Weedy check 1057 — — —

Price of garlic bulbs t�1 US $333.33, cost of variables: pendimethalin 330 EC at US $8.0 l�1, sprayer rent US $1.0, labour charges of herbicide application

US $2.0 ha�1, cost of hand weeding 14 men day at US $2.0 man�1.

Table 5

Economic analysis (2002–2003)

Treatments Gross income (US

$ ha�1)

Value of increased yield

(US $ ha�1)

Variable cost

(US $ ha�1)

Increased net benefit over

weedy control (US $ ha�1)

Oxadiazon+1 hoeing 3690 3054 57.55 2996.45

Glyphosatee+1 hoeing 3084 2448 50.25 2397.75

Pendimethalin+1 hoeing 3965 3329 59.75 3269.25

Metribuzin+1 hoeing 247 �389 49.42 �438.42

Weed-free 4003 3367 140.00 3227.00

Weedy check 636 — — —

Price of garlic bulb t�1 US $417.0, cost of variables: oxadiazon 120 EC at US $7.4 l�1, glyphosate 40.67 EC at US $7.4 l�1, Pendimethalin 330 EC at US

$10.0 l�1, metribuzin 70% G at US $20 kg�1, sprayer rent US $1.5, labor charges of herbicide application US $2.5 ha�1, cost of hand weeding 14 men day

at US $2.5 man�1.

T. Mehmood et al. / Crop Protection 26 (2007) 1031–10351034

bulb yield (7.29 t ha�1) was recorded in the treatment wherependimethalin alone was used followed by no hoeing(Table 1). However, pendimethalin suppressed weedsgrowth for approximately a period of 2 months but lateron, the weeds began to flourish and competed with the cropresulting in lower bulb yield. These results are in agreementwith Vora and Mehta (1998) and Mahmood et al. (2002)who found that pendimethalin and oxadiazon applicationalone did not control weeds effectively in garlic crop. Thehighest bulb yield of 12.21–14.17 t ha�1 was recordedduring the year 2000–2001 when pendimethalin wasapplied followed by different mechanical weeding regimes.During the year 2002–2003, bulb yield of 7.39–9.76 t ha�1

was recorded when herbicides were applied followed by onehoeing. The highest yield in these treatments could be dueto less weed competition. Our results are in partialagreement with Sandhu et al. (1997), who reported thatpendimethalin application when supplemented with onehoeing after 105 days of sowing resulted in significantincrease in bulb yield compared to weed free treatment.

3.8. Monetary returns

Farmers adopt only those improved practices, which aremore profitable. On the basis of current market prices ofherbicides and the wholesale prices of garlic, the economicsof different weed control methods were calculated to findout the most profitable (Tables 4 and 5). During the year,

2000–2001, all the weed control treatments gave consider-ably higher net income ha�1 over weedy check. Pendi-methalin in combination with different mechanical weedingregimes recorded net return of US $ 3352–3601 ha�1 overweedy check. Weedy check resulted in gross income of US$ 1057 ha�1. During the year 2002–2003, application ofmetribuzin resulted in loss of US $ 438.42 ha�1 over weedycheck. Among the weed control treatments, pendimethalinin combination with one hoeing recorded the highestnet return of US $ 3269 ha�1 over weedy check. Theseresults are contrary to those of Tewari et al. (1998), whoreported the greatest net monetary returns with handweeding and with pendimethalin application followed byhand hoeing.These studies suggest that pendimethalin in combination

with manual hoeing can be successfully applied to boost upthe bulb yield in garlic and to harvest the maximum profitfor raising the economic status of farmers.

References

Anonymous, 2004. Agricultural Statistics of Pakistan, 2003–04, Ministry

of Food, Agriculture and Livestock (Economic Wing), Islamabad,

Pakistan, p. 73.

Khokhar, K.M., Mahmood, T., Shakeel, M., Chaudhry, M.F., 2006.

Evaluation of integrated weed management practices for onion in

Pakistan. Crop. Prot. 25, 968–972.

ARTICLE IN PRESST. Mehmood et al. / Crop Protection 26 (2007) 1031–1035 1035

Mahmood, T., Khokhar, K.M., Hussain, S.I., Laghari, M.H., 2002. Weed

control in garlic crop in relation to weedicides. Asian J. Plant Sci. 1 (4),

412–413.

Qasem, J.R., 1996. Chemical weed control in garlic in Jordan. Crop. Prod.

15 (1), 21–26.

Sandhu, K.S., Singh, D., Singh, J., 1997. Weed management in garlic

(Alliums sativum L.). Vegetable Sci. 24 (1), 7–9.

Steel, R.G.D., Torrie, J.H., 1980. Principles and Procedures of Statistics.

McGraw Hill Book Co. Inc, New York.

Tewari, A.N., Rathi, K.S., Singh, B., 1998. Integrated weed manage-

ment in garlic (Allium sativum). Indian J. Agric. Sci. 68 (5),

281–283.

Vora, V.D., Mehta, D.R., 1998. Integrated weed management in winter

garlic. Agric. Sci. Digest Karnal 18 (4), 237–239.

Vora, V.D., Mehta, D.R., 1999. Studies on growth yield and yield

attributes of garlic as influenced by herbicides and weeds. Agric. Sci.

Digest Karnal 19 (2), 129–133.