Differential Effects of Pesticide Applications on Liriomyza huidobrensis (Diptera: Agromyzidae) and its Parasitoids on Pea in Central Kenya

Journal of Economic Entomology Advance Access published February 25, 2015 HORTICULTURAL ENTOMOLOGY

Differential Effects of Pesticide Applications on Liriomyza huidobrensis (Diptera: Agromyzidae) and its Parasitoids on Pea in Central Kenya M.M. GUANTAI,1,2 C.P.K.O. OGOL,1 D. SALIFU,2 J.M. KASINA,3 K.S. AKUTSE,2 AND K.K.M. FIABOE2,4

J. Econ. Entomol. 1–10 (2015); DOI: 10.1093/jee/tov006

ABSTRACT Three Liriomyza species [Liriomyza huidobrensis (Blanchard), Liriomyza trifolii (Burgess), and Liriomyza sativae Blanchard] have been reported as the most important leafminer pests in vegetable production systems in Africa. In Kenya, farmers rely on indiscriminate synthetic insecticides use. On-farm field investigations were set up at three different locations (Sagana, Kabaru, and Naromoru) in central Kenya to determine the effect of pesticide application on the abundance of leafminers and their parasitoids under three management practices, namely: farmer practice (FP), reduced pesticide use (RP), and a control with no use of pesticides (CO). In addition, laboratory experiments were designed to test the effect of commonly used pesticides in pea production systems in central Kenya—Dimethoate, Dynamec, Thunder, Cyclone, Bestox, Folicur, Milraz, and Bulldock—on L. huidobrensis and two of its parasitoids, Diglyphus isaea Walker and Phaedrotoma scabriventris Nixon. The mean numbers of leafminer flies in control treatment were higher than in RP and FP in both first and second seasons across all sites, but RP and FP did not differ significantly. Parasitoid numbers were very low and there was no much variation between treatments at each location in both first and second seasons. No significant differences were observed between the three management practices with regards to the yield measurements. In the laboratory, the estimated LD50 values for L. huidobrensis larvae were all more than two times higher than the recommended dosages, while the LD50 of adults were below the recommended dosages. The estimated LD50 values for the parasitoids were much lower than recommended dosages for all pesticides except Thunder. This study, therefore, demonstrates that the pesticides currently used do not control the Liriomyza leafminer larvae that constitute the most destructive stage of the pest, but are rather detrimental to their parasitoids. In addition, the current low level of parasitoids recorded under field conditions even where no pesticide was used during this study, warrants consideration of classical biological control programs. KEY WORDS Leafminer parasitoid, insecticide, fungicide, LD50, LD90

Production of high value horticultural crops such as peas (Pisum sativum) has been identified as a key pathway out of poverty and a prospect for development because of high land productivity compared with the production of staple crops (Asfaw et al. 2009, McCulloch and Ota 2002). Kenya’s horticulture industry is very important in meeting the country’s employment needs and improving livelihoods (McCulloch and Ota 2002). Overall, the horticultural industry contributes 23% of the gross domestic product and employs 4.5 million people directly and 3.5 million indirectly both on- and off-farm activities in Kenya (Fresh Produce

1 Department of Zoological Sciences, Kenyatta University, P.O. Box 43844, Nairobi, Kenya. 2 International Centre for Insect Physiology and Ecology (ICIPE), P.O. Box 30772-00100. 3 Kenya Agricultural Research Institute (KARI), P.O. Box 1473300800, Nairobi, Kenya. 4 Corresponding author, email: [email protected].

Exporters Association of Kenya [FPEAK] 2012). The total production of horticultural crops in central Kenya has been far below the expected levels. Among the important causes is the invasion of the crop by insect pests such as Liriomyza leafminers [leafminer flies; Liriomyza huidobrensis (Blanchard), Liriomyza sativae Blanchard, and Liriomyza trifolii (Burgess)], which also happen to be on the European Union list of quarantine pests (Kedera and Kuria 2003). Liriomyza leafminer adult females cause damage by puncturing leavesfor feeding and laying eggs (Weintraub and Horowitz 1997), while the larvae cause damage by tunneling into leaves of plants, creating “mines” that result in severe yield reduction (Weintraub and Horowitz 1995). The pea leafminer, L. huidobrensis, is the most important pest of snow peas and sugar snaps in Kenya (Njuguna et al. 2001, Kedera and Kuria 2003) and has been reported to pose a worldwide threat to horticultural field crops (Murphy and La Salle 1999, Weintraub and Horowitz 1995) with yield losses

C The Authors 2015. Published by Oxford University Press on behalf of Entomological Society of America. V

All rights reserved. For Permissions, please email: [email protected]

2

JOURNAL OF ECONOMIC ENTOMOLOGY

ranging between 50 and 100% (Weintraub and Horowitz 1997). Being highly polyphagous, L. huidobrensis spreads quickly in greenhouses and outdoors even in the absence of pea, and if not controlled, it results to economically important losses. Three factors make L. huidobrensis a pest of unique economic importance in snow peas production: it attacks the marketed part of the crop, the damage created by the larvae is often unnoticeable until the produce has reached the destination market, and the pest has been included in the European Union list of quarantine pests as an A2 pest (Anderson and Hofsvang 2010, Gitonga et al. 2010). In Kenya, yield losses have been documented to be between 20 and 100%, depending on crop species, cultivar, and crop developmental stage, which subsequently affects the level of infestation (Chabi-Olaye et al. 2008). The most widely reported reason for the first leafminer outbreaks in their adventive ranges was the indiscriminate use of insecticides, which adversely affect their natural enemies (Murphy and LaSalle 1999). Several pesticides including early organophosphates (Methyl parathion), carbamates (oxamyl), pyrethroids (permethrin), and triazines (cryomazine) were identified for Liriomyza leafminer control (Price and Nagle 2002). However, resistance against pesticides is one of the consequences of indiscriminate pesticide use (Price and Nagle 2002). According to Weintraub and Horowitz (1997), there are currently no effective adulticides and only few larvicides (Abamectin and Cryomazine) against this pest, which if used exclusively could quickly generate resistance. In addition, insecticides use has harmful effects on the parasitoids, making natural control ineffective (Tong-Xian et al. 2009). Field studies in central Kenya indicated that a range of pesticides are being routinely used by farmers against pea pests (Gitonga et al. 2010). Many horticultural growers in Kenya have been using avermectins (abamectin), triazines (cyromazine), carbamates, organophosphates, and pyrethroids to control leafminers (Kabira 1985). Kotzee and Dennill (1996) reported resistance of L. trifolii to cyromazine and triazine in South Africa. Resistance of Liriomyza leafminer to most carbamate, organophosphate, and pyrethroid insecticides has also been reported in the United Kingdom (MacDonald 1991). In this study, we evaluated the effect of commonly used pesticides in pea production systems of central Kenya on L. huidobrenis and its parasitoids under laboratory and field conditions to obtain information for planning and development of strategies for effective and environmental friendly management of leafminer flies.

Materials and Methods Field Experiments. On-farm trials were conducted in Nyeri County, central Kenya at three locations; Naromoru (0 270 S; 23 220 E, 2,036 m above sea level [a.s.l]) Kabaru (0 28.30 S; 37 210 E, 2,530 m a.s.l) and Sagana (0 350 S; 37 190 E, 1,208 m a.s.l). The climate of the area is cool and wet with annual temperatures of

23 C (maximum 26 C and minimum 19 C). The annual rainfall of 1,700 mm is concentrated between March and July and between October and December growing seasons. We selected small-scale farmer fields that produce vegetables for both export and domestic markets to study three management practices, namely: (1) farmer practice (FP), where farmers followed their routine practice of pest control (FP), (2) reduced pesticide use (RP), and (3) no pesticide application as control (CO). In the reduced pesticides management system, pesticides were used only when the leafminer flies numbers were high (>10 flies per step-count). The list and frequency of chemicals used is given in Table 1. At each location, four experimental fields were established and each field was divided into three plots to accommodate the three different management systems. Each plot was 100 m2 and the distance between fields was >200 m apart. The study was conducted in two cropping seasons under irrigation, from February to April 2009 (Season 1: wet cool) and then from May to August 2009 (Season 2: dry cold). The number of parasitoids present in different plots was assessed by sampling infested leaves from each plot, kept in separate plastic bags, and taken to the laboratory for further assessment. In the laboratory, the leaves were placed in large plastic container and stored for 1 wk. After 1–2 wk, the total number of leafminer flies pupae and emerging natural enemies were recorded. The leafminer flies pupae were placed in a glass vial and stored till the flies emerged. Emerging Liriomyza flies and parasitoids were preserved in 95% ethanol þ 5% glycerin for identification. Sampling was done weekly until harvest (12 wk). Laboratory Experiment. Commonly used chemicals in central Kenya (Nyeri) were adopted from a study by Gitonga et al. (2010; Table 2). The concentrations evaluated for each chemical were determined based on the recommended dosages provided by the supplier, where two dose levels were below the recommended dose and two above the recommended dose (Table 2).

Liriomyza Leafminer Larvae Chemical Bioassay Peas were planted in 10 pots each week for 8 wks in the greenhouse. Three weeks after planting, the 10 plants (one plant per pot) were exposed to 50 adult L. huidobrensis (obtained from laboratory cultures at International Centre for Insect Physiology and Ecology [ICIPE]) for 24 h for oviposition to take place in a cage. This was done in an exposition cage at the greenhouse. The infested plants were transferred to another holding cage until about six to eight mines had formed in 10 leaves. The leaves were excised using a pair of scissors and taken to the laboratory. In total, 100 live larvae were counted on the leaves under a light microscope. The leaves (with 100 larvae) were put together in a glass Petri dish and treated with the chemicals. Five different Petri dishes each holding 100 larvae (in the leaves), were used for each insecticide and replicated four times. Six insecticides, namely; Bulldock,

2015

GUANTAI ET AL.: PESTICIDES ON LEAFMINER FLIES AND THEIR PARASITOIDS

3

Table 1. Frequency of chemical application in reduced pesticide plot (RP) as compared with farmers plot (FP) Location

Chemicals used

Sagana

Cyclone, Milraz, Ridomil, Dimethoate, Methane, Folicur, Dithane, Thunder, Bestox, and Dynamec Antracol, Bulldock, Triger, Thunder, Milraz, Dynamec, Folicur, and Dimethoate Dimethoate, Milraz, Cyclone, Cumulus, Thunder, Bestox, Folicur, and Dynamec

Kabaru Naromoru

Frequency of chemical application FP

RP

14

10

13

8

13

6

Table 2. List of pesticides used in the laboratory experiment at ICIPE on the leafminer flies and its parasitoids Trade name Active ingredients

Pesticide Mode of action type

Target pests

Bulldock Beta-cyfluthrin Dimethoate Dimethoate

I I

Cutworm, aphid, leafminer flies 0.75 ml Aphids, cutworm, leafminer flies 1.5 ml

1.0 ml 1.6 ml

Milraz

Propineb/cymoxanil

F

Blight, P. mildew

2.0 g

2.95 g

Dynamec

Abamectin

I

Liriomyza spp.

0.5 ml

0.75 ml

Folicur

Tebuconazole

F

P. mildew, blight

0.75 ml

0.5 ml

Bestox

Alphacypermethrin

I

Cutworm, thrips, leafminer flies

0.5 ml

0.9 ml

Thunder

Imidaclopride/betaI cyfluthrin Cyclone (Cypermethrin I 10% þ Chlorpyriphos 35%)

Thrips, leafminer flies

0.5 ml

0.75 ml

Cutworm, aphid, leafminer flies

1.5 ml

1.3 ml

Cyclone

Non systemic/neurotoxicity Contact and systemic/Cholinesterase inhibition after metabolism Inhibitor of oxidizing reaction of pyruvate Neurotoxicity, stomach poison and limited contact activity Systemic action with Preventive, Curative & Eradicative Action Non systemic insecticide with contact and stomach action Contact activity and systematic action Nonsystemic insecticide with quick knockdown

Recommended Farmers rate rate (ml/liter) (ml/liter)

Source: Gitonga et al. (2010). I, insecticide; F, fungicide; P. mildew, Powdery mildew.

Dimethoate, Dynamec, Cyclone, Thunder, and Bestox plus two fungicides Milraz and Folicur were evaluated, each with four concentrations and distilled water treatment as control. Mined leaves placed in Petri dish one at a time were immersed in the pesticide concentration for 5 s. The treated leaves were then placed in trays and air-dried in the open for 15 min, after which they were put back in the Petri dish lined with a paper towel and taken to the laboratory at 25 6 2 C. After 5 d, total number of pupae were counted and recorded from each Petri dish. The pupae were kept under the same temperature conditions until the adult leafminer fly emerged (a week later). These too were counted and recorded. Larval mortality was determined based on the total number of larvae that failed to pupate. The pupae that failed to emerge were also recorded. Adult Leafminers Chemical Bioassay. Adult Liriomyza leafminers were obtained from the mass rearing stock at ICIPE. All the pesticides recorded in Table 1 (fungicides and insecticides) were used. Each pesticide application included four dosages and one control (distilled water). For each dosage, 25 adult leafminer flies were used. About 15 ml of each of these concentrations were poured separately in a 20-ml glass vial swirled and then emptied. The vial was then air-dried for 6 h before

use to prevent the flies from drowning in the chemical. The 25 leafminer flies adults were placed into the glass vial that was then covered with a perforated cork to avoid suffocation because of lack of oxygen supply. Cotton swabs saturated with honey–sugar solution were placed at one corner of the cork to provide food to the flies during testing. The treated vials were taken to the laboratory immediately. Mortality was recorded every 24 h until all the adults were dead (100% mortality obtained at 48 or 72 h). Each concentration was replicated four times. Adult Leafminer Parasitoid Chemical Bioassay. Two parasitoids D. isaea (obtained from Dudutech Kenya Ltd.) and P. scabriventris (obtained from the laboratory cultures at ICIPE) were used. All the pesticides (insecticides and fungicides) recorded in Table 1 above were used. The same concentrations, methods, and replicates as the ones used for the adult Liriomyza leafminer bioassay were used. Mortality was recorded every 24 h until all the parasitoids were dead (100% mortality obtained at 48 or 72 h). Data Analysis. Mortality data on leaf miner larva, leafminer adults, D. isaea, and P. scabriventris under dose–response experiments were corrected for mortality in control treatment using Abbott’s formula (Abbott 1925). The corrected mortality data were then

JOURNAL OF ECONOMIC ENTOMOLOGY

4

subjected to probit analysis to obtain LD50 and LD90 estimates, slopes and 95% confidence limits. The data analysed were for mortality at 24 h postexposure. Weekly leafminer abundance data from field experiments were averaged over the weeks and then analyzed using a generalized linear model (glm) with logarithmic link assuming Poisson distribution error. In cases where there was overdispersion, the negative binomial distribution error was assumed for the counts (O’Hara and Kotze 2011). The effect of treatment factor in a glm is reflected in the deviance that has an approximate chisquare distribution; hence, the chi-square values have been presented as test statistic. Weekly emerged parasitoids from leafminer infested leaves from field experiments were summed and analysed using a glm, as was the case with the leafminer flies. The analysis on leafminer and parasitoids’ abundance was done for each site to evaluate the effect of management systems. Analysis of variance was used to analyze pea yield data and means together with SEs are presented. The analysis was performed using R 2.13.1 (R Development Core Team 2011).

Results Leafminer and Parasitoid Abundance in Pea Fields. Analysis of leafminer flies abundance for the first season showed that there were significant differences between management systems in Sagana (v2 ¼ 10.6; df ¼ 2; P ¼ 0.005) and Kabaru (v2 ¼ 7.2; df ¼ 2; P ¼ 0.028), but not in Naromoru (v2 ¼ 0.12; df ¼ 2; P ¼ 0.94). In Sagana, leafminer infestation was reduced by 3% in RP treatment and 48% in FP treatment, whereas in Kabaru, infestation reduced by 29% in RP treatment and 48% in FP treatment, but the difference between RP and FP was not significant (Fig. 1a). The results in the second season were similar to those obtained in season 1, namely, there were differences in treatments at each location except for Naromoru. In both Kabaru and Sagana, leafminer infestation was reduced by 39 and 36% for RP and FP treatments, respectively. Again in both sites, RP and FP did not differ significantly (Fig. 1b). Generally, Naromoru had lower infestation than Kabaru and Sagana. As for the parasitoid abundance, there was no significant variation between treatments in Sagana (v2 ¼ 1.3; df ¼ 2; P ¼ 0.52) and Naromoru (v2 ¼ 2.99; df ¼ 2; P ¼ 0.22), while there was significant variation in Kabaru (v2 ¼ 7.42; df ¼ 2; P ¼ 0. 02) in the first season. Inasmuch as the parasitoids were very low in numbers, the control treatments had more parasitoids than the RP and FP treatments (Fig. 1c). In the second season, there was no significant variation in parasitoid abundance in the management systems at Kabaru and Naromoru, but significant difference was obtained at Sagana (v2 ¼ 6.3; df ¼ 2; P ¼ 0.04; Fig. 1d). In general, parasitoids population abundance was low in all the locations and during both seasons. Pea Fresh Yields. Pea fresh yields were not significantly different between treatments during both seasons. However, just as with leafminer infestation, pea

yields were higher in the second season than in the first season (Table 3). Dose Response of Leafminer Larvae and Adults. The estimated LD50 values for insecticides against leafminer larva were all above the recommended dose of the insecticides studied and this is also confirmed by the slope values (Table 4). Of all the chemicals tested, Dynamec showed the highest effects on leafminer flies’ larval mortality based on the LD50 value estimated as 1.11 ml/liter with a 95% confidence level [0.99, 1.26] and a slope of 1.54. Thunder and Dimethoate had the least effect, as reflected by the high estimated values of LD50 (Table 4; Fig. 2). None of the pesticides tested was effective against the leafminer flies’ larvae. For instance, LD50 obtained for Bestox, Bulldock, Dimethoate, Dynamec, and Thunder were 20.8, 6.2, 2.6, 2.2, and 33.1 times, respectively, higher than their respective recommended doses (Table 4). Dynamec (LD50 ¼ 0.10 ml/liter), Bestox (LD50 ¼ 0.16 ml/liter), and Bulldock (LD50 ¼ 0.35 ml/liter) were more effective on L. huidobrensis adults than the rest of the chemicals studied, while Thunder was the least effective (LD50 ¼ 6.29 ml/liter; Table 4). The slope of Dynamec (1.48), Bestox (1.23) Bulldock (1.83), and Thunder also confirmed the toxicity level of the various insecticides (Table 4). In addition, the values of the slopes of the various pesticides are small, indicating that the response of the populations submitted to the test was heterogeneous. Apart from Folicur and Milraz that have their estimated LD50 higher than their recommended doses, the LD50 of all other pesticides tested were below the recommended dose, denoting the effective control of leafminer adults by the majority of pesticides studied. The estimated LD90 values for insecticides against L. huidobrensis adults were all far above the recommended dosage (Table 4). The average mortality (corrected for control) of L. huidobrensis adults after 24 h of exposure to different dosages of Dimethoate was 8–12%, and thus the dosages had no significant effect on mortality and so the probit model could not estimate the LD50 and LD90 values. However, at 48 h after exposure, the LD50 and LD90 values were 1.11 and 3.33 ml/liter, respectively, with a slope of 2.65 (Table 4). Dose–Response of Diglyphus isaea and Phaedrotoma scabriventris. The LD50 values for the dose–response of D. isaea and P. scabriventris for the pesticides studied were all far below the recommended dose except for Thunder. Diglyphus isaea Walker and Phaedrotoma scabriventris Nixon (Braconidae: Opiinae) were most susceptible to Dynamec, which had the lowest LD50 and LD90 values of 0.004 and 0.27 ml/liter for D. isaea and LD50 and LD90 values of 0.02 and 0.24 ml/ liter for P. scabriventris. The slope values were 0.69 and 1.22 for D. isaea and P. scabriventris, respectively (Table 5). For the other insecticides, the LD50 values were all between 0.06 and 0.24 ml/liter except for Thunder, which LD50 values were 1.27 and 0.98 ml/liter for D. isaea and P. scabriventris (Table 5). From the result of the probit analysis, the LD50 of the various pesticides were between 0.004 and 0.24 for D. isaea and 0.02 and 0.19 for P. scabriventris (Table 5). These values

2015

GUANTAI ET AL.: PESTICIDES ON LEAFMINER FLIES AND THEIR PARASITOIDS

5

Fig. 1. Mean leafminer abundance in season 1 (a) and season 2 (b), and parasitoids abundance in season 1 (c) and season 2 (d) under different management systems in pea fields; FP, farmer practice; RP, reduced pesticides; CO, control (no pesticide use) at different locations in central Kenya. Means and SEs were obtained from the generalized linear model that was fitted to the data.

Table 3: Mean fresh yield of peas (tons/ha) under different treatments Treatment

RP FP CO f2,13 P value

Yield (tons/ha) Season 1

Season 2

0.85 6 0.21 1.28 6 0.41 0.92 6 0.25 0.65 0.57

3.26 6 0.41 2.07 6 0.46 4.17 6 0.87 6.67 0.08

show that the used pesticides were more injurious on parasitoids than on the host pest L. huidobrensis at lower concentrations. The toxicity level of the pesticides tested on the two parasitoids was also confirmed by the slope values, which ranged between 0.18 and 1.93 for D. isaea and 0.46 and 1.84 for P. scabriventris (Table 5). As the values of these slopes are small, it can then be deduced that the response of the populations submitted to the test was heterogeneous for the various pesticides. The results of the study imply that Thunder is 423 times

less toxic than Dynamec on D. isaea and 25 times less toxic than Dynamec on P. scabriventris. The LD50 values of the pesticides tested on the parasitoids were 2–33-fold lower than the LD50 values for the same insecticides tested on L. huidobrensis adults. Average mortality of D. isaea exposed to Bestox was 60–77% after 24 h of application (Table 5). Discussion In the first season, where relatively lower numbers of leafminer flies were reported in the field, the use of pesticide, whether at reduced level or routinely used, as in the case of farmers fields, did not significantly reduce leafminer flies and its parasitoids populations compared with the control, where no pesticides were used. However, during the second season when leafminer flies’ attack was more pronounced, significantly higher numbers of leafminer flies were observed in the control compared with RP and FP treatments across all localities, with no difference found between RP and

a

0.16 [0.05, 0.25] 0.35 [0.24, 0.43] 1.14 [1.08,1.22] 1.11 [0.61, 1.43] 0.10 [0.03, 0.17] 0.85 [0.80,0.89] 1.10 [0.99,1.20] 6.29 [2.12,6.50] Bestox Bulldock Cyclone Dimethoatea Dynamec Folicur Milraz Thunder

Cyclone, Folicur and Milraz were not tested on L. huidobrensis larva. At 24 h the dosages had no significant effect (slope was not significantly different from zero) and therefore the LD50 and LD90 values are probit model estimates at 48 h after exposure.

0.5 ml/liter 0.75 ml/liter 1.5 ml/liter 1.5 ml/liter 0.5 ml/liter 0.75 ml/liter 2.0 g/liter 0.5 ml/liter 0.70 [0.35, 1.06] 1.03 [0.57, 1.50] – 2.34 [1.70, 2.99] 1.54 [1.31, 1.77] – – 0.85 [0.55, 1.16]

Slope LD90

700.6 [66, 925 838] 80.01 [19.7, 3128] – 13.94 [9.85, 23.40] 17.29 [10.54, 35] – – 523.5 [98, 16 889] 10.38 [3.9, 184.9] 4.62 [2.71, 18.04] – 3.93 [3.33, 5.16] 1.11 [0.99, 1.26] – – 16.55 [7.5, 82.86] 1.23 [0.62, 1.84] 1.83 [1.30, 2.35] 3.33 [0.48, 7.15] 2.65 [1.25, 4.04] 1.48 [0.79, 2.16] 2.83 [1.11, 6.77] 2.35 [2.22, 6.91] 1.09 [0.54, 10.73]

L. huidobrensis larvae

LD50 Slope LD90 LD50 Chemical

L. huidobrensis flies

1.70 [1.32, 2.52] 1.75 [1.52, 2.10] 2.79 [2.57, 3.09] 3.33 [2.95, 3.9] 0.74 [0.65, 0.89] 2.45 [2.10, 3.01] 3.79 [3.28, 4.46] 6.79 [5.04, 9.15]

Recommended application rate

JOURNAL OF ECONOMIC ENTOMOLOGY

Table 4. LD50 and LD90 values and their 95% confidence limits for pesticides against adult leafminer flies and leafminer larvae estimated from mortality at 24 h after exposure

6

FP. This showed that the pesticides had some level of control on leafminer flies; however, such reduction in RP and FP was at the expense of parasitoids. Such an outcome could easily lead to usage of high doses of pesticides in an effort to control the pest, while extremely low doses are detrimental to the parasitoids. Similar cases of great diminution in biological control potential in the management of leafminer flies were reported for cryomazine, ethylparathion, ethophenprox, methylparathion, methomyl, permethrin, and prothiofos (Oatman and Kennedy 1976, Trumble 1985, Saito et al. 1996, Djoko et al. 2004, Hidrayani et al. 2005). Murphy and LaSalle (1999) reported that the most important cause of Liriomyza leafminer high infestations in the United States is the indiscriminate use of insecticides and their negative effects on natural enemies. Similarly, Hidrayani et al. (2005) reported that insecticide applications reduce parasitism rates of Hemiptarsenus varicornis (Girault) (Hymenoptera: Eulophidae) and Opius chromatomyiae and decrease the predatory of muscid fly Coenocia humilis, effects that significantly reduce the control of leafminer. In addition, Herna´ndez et al. (2011) showed that when spraying pepper fields with novaluron, abamectin, spinetoram, and lambda-cyhalothrin against leafminer infestation, not only high Liriomyza population densities were recorded but also low number of parasitoids per leafminer larva and low parasitoid diversity index were obtained. To increase parasitoids population and boost natural control levels in the field, farmers would minimize the use of chemicals, particularly considering that the high level of pesticide use in this study did not provide any better results compared with reduced level of pesticide. The parasitism rates recorded in both seasons were extremely low, indicating that natural control by indigenous parasitoids was not effective under current field conditions. Conservation, augmentative, and classical biological control initiatives need to be considered to increase the efficiency of the existing natural enemies. Some indigenous natural parasitoids such as Opius dissitus Muesebeck (Hymenoptera: Braconidae), D. isaea, Neochrysocharis formosa (Westwood) (Hymenoptera: Eulophidae), and H. varicornis have been identified (Chabi-Olaye et al. 2008) and could then be used to boost the parasitism rates of leafminers. D. isaea is widely used locally and overseas as augmentative biocontrol agent and can be responsible for high levels of parasitism (Tong-Xian et al. 2009). Other promising exotic parasitoids such as P. scabriventris, Halticopera arduine (Walker) (Hymenoptera: Pteromalidae) and Chrysocharis flacilla Walker (Hymenoptera: Eulophidae), reported to cause mortalities of up to 80% of Liriomyza leafminer flies in the pest’s original zone, need to be introduced into the vegetable production systems in Kenya to enable environmentally friendly control of the pest (Mujica and Kroschel 2011, Chabi-Olaye et al. 2013). P. scabriventris is found to be the most important leafminer parasitoid in Argentina, Brazil, and Peru, where it has been reported to cause mortality of up to 52% (Valladares et al. 1999, 2001; Kroschel 2008). In Italy, parasitoids reduced L. huidobrensis populations significantly on lettuce (Masetti et al. 2010). However,

2015

Fig. 2. 2009.

GUANTAI ET AL.: PESTICIDES ON LEAFMINER FLIES AND THEIR PARASITOIDS

7

Percentage leafminer larval mortality in the laboratory under different chemical concentrations, May–August

Table 5. LD50 and LD90 values and their 95% confidence limits or pesticides sprayed on D. isaea and P. scabriventris estimated using mortality at 24 h after exposure D. isaea adults

P. scabriventris adults

Chemical

LD50

LD90

Slope

LD50

LD90

Slope

Bestox Bulldock Cyclone Dimethoate Dynamec Folicur Milraz Thunder

0.015 0.15 0.24 [0.06,0.42] 0.16 [0.008,0.32] 0.004 [0,0.027] 0.064 0.141 [0.01,0.35] 1.27 [1.08,1.63]

250.846 0.24 1.53 [1.30,1.79] 0.71 [0.29,0.95] 0.27 [0.06,0.41] 1.67 2.47 [1.95,4.76] 6.83 [4.29,14.51]

0.30 [0.08, 1.41] 0.18 [0.11, 0.75] 1.63 [0.45, 3.71] 1.93 [1.07, 4.93] 0.69 [0.27, 8.64] 0.92 [0.09, 2.83] 1.02 [0.77, 2.81] 1.21 [0.70, 3.13]

0.17 [0.05, 0.27] 0.14 [0.03, 0.24] 0.17 [0.02,0.35] 0.00 085 0.02 [0.002,0.05] 0.07 [0.003,0.15] 0.195 [0.04,0.37] 0.977 [0.76,1.60]

0.85 [0.74,1.01] 0.82 [0.72,0.98] 1.13 [0.81,1.34] 0.18 0.24 [0.12,0.32] 0.47 [0.25,1.72] 1.3 [1.03,1.52] 51.56 [13,1579]

1.84 [0.84, 2.84] 1.65 [0.81, 2.50] 1.57 [0.43, 3.56] 0.46 [0.09, 2.85] 1.22 [0.22, 2.67] 1.50 [0.62, 3.62] 1.57 [0.32, 3.46] 0.74 [0.16, 1.65]

the effectiveness of these parasitoids will depend on the development of an effective integrated pest management that will enhance the natural populations of parasitoids (Salvo et al. 2005). Pea fresh yields were not significantly different between the various management practices during both seasons. However, just as with leafminer infestation, pea yields were higher in second season than the first season. This is an indication that there are no gains in yield, despite various pesticides sprays targeting leafminer flies that ultimately reduced parasitoid populations. Under laboratory conditions, none of the pesticides studied could control the larval stage of L. huidobrensis at recommended dosages of the studied pesticides. This then explains the observation by Gitonga et al. (2010), who noted that farmers used higher rates than the recommended doses of the chemicals in their fields. Schuster and Everett (1983), Ferguson (2004), and Reitz et al. (2013) reported the effectiveness of pesticides (spinosad, abamectin, and cyromazine) against the larva with no resistance found. This is in

contrast with our results, where all the tested pesticides, including those that have systemic effects (Cyclone and Dynamec) and stomach poison effects (Bestox and Thunder) were not able to kill the leafminer flies larvae at their recommended doses. Liriomyza leafminers have high ability to develop resistance to pesticides. Cross-resistance to multiple classes of insecticides is also found in Liriomyza spp. (Mason et al. 1987, Lamport et al. 2001, Varela et al. 2003, Gitonga et al. 2010). Populations of invasive L. trifolii obtained from greenhouses in Canada treated intensively with the organophosphate pyrazophos for