Monitoring resistance of Helicoverpa armigera to different insecticides used in cotton in Spain

1 Title: 1 Monitoring resistance of Helicoverpa armigera to different insecticides used in cotton 2 in Spain. 3 4 Authors: Carlos Avilla, José E. González-Zamora 5 6 Affiliations: Department of Ciencias Agroforestales, University of Seville. Carretera 7 de Utrera, km 1. E-41012-Seville (Spain) 8 9 Corresponding author: José E. González-Zamora 10 [email protected] 11 Phone: +34-954 486 459 12 Fax: +34-954 46 436 13 Published in: Crop Protection 29 (2010) 100–103 14 doi:10.1016/j.cropro.2009.09.007 15 2 Abstract 1 Helicoverpa armigera is the key pest of cotton in Spain, resulting in many 2 insecticide treatments against it. The resistance status of Helicoverpa armigera to 3 different insecticides currently used in cotton was evaluated in Spain in two different 4 seasons, 1999 and 2004. Four populations were tested in total, two in each season. 5 Toxicological bioassays were conducted in the laboratory, and performed on third 6 instar larvae by topical application of the insecticides. LD50’s were estimated by 7 probit analysis and resistance factors (RF) were calculated at the LD50 level. Four 8 insecticides were evaluated, but only endosulfan reached a moderate resistance level 9 (RF=11.4), and the others (methomyl, chlorpyrifos and lambda-cyhalothrin) showed 10 low resistance (RF between 1.9 and 6.0). Such results indicate the generally low 11 resistance of H. armigera to most of the insecticides used against this pest in cotton 12 in Spain. Possible explanations for this situation are discussed. 13 14 Key words: insecticide resistance, cotton, Helicoverpa armigera , endosulfan, 15 methomyl, chlorpyrifos, lambda-cyhalothrin. 16 3 1. Introduction 1 Cotton is an important crop in the Guadalquivir river valley (Andalucía, southern 2 Spain), especially in the provinces of Seville, Córdoba and Cádiz. One of its main 3 economic costs comes from insecticide treatments against pests like Helicoverpa 4 armigera Hübner (Lepidoptera: Noctuidae), the key pest of cotton in southern Spain 5 (Durán, 1999). H. armigera can severely reduce cotton yields, mainly because it 6 feeds on flower buds, flowers and on the boll. Its control is based mainly on an IPM 7 schedule, which involves estimating populations of the pest and its natural enemies: 8 predadory bugs (Heteroptera) and lacewings (Neuroptera), and recommending 9 insecticides only when H. armigera populations surpass a threshold and natural 10 enemies are scarce (Durán, 1999; 2003). 11 12 The control of this pest in Spanish cotton may require several applications 13 through the season (the IPM cotton is sprayed with insecticides between 2 and 9 14 times, range obtained from 23 seasons, Durán (1999; 2003)), mainly of endosulfan 15 and methomyl (2-6 treatments), complemented with other products including 16 pyrethroids (e.g. lambda-cyhalothrin), organophosphates (e.g. chlorpyrifos) or other 17 carbamates (e.g. thiodicarb) (Durán, 2003). The continuous use of these products 18 can lead to the development of resistant populations of H. armigera , and in the 2003 19 season important control failures of H. armigera were observed in cotton fields. The 20 failure reports involved mainly the Seville province and the two products that were 21 most used against this caterpillar, endosulfan and methomyl. There are many reports 22 of the resistance of H. armigera (and other Heliothinae) to insecticides around the 23 world, especially to pyrethroids, but also to other groups of insecticides, like 24 4 cyclodienes (such as endosulfan), carbamates (such as methomyl and thiodicarb) 1 and organophosphates (such as clorpyriphos) (McCaffery, 1999; Bues et al., 2005; 2 Ahmad et al., 2007; Pietrantonio et al., 2007; Ugurlu and Gurkan, 2007; Saleem et 3 al., 2008). 4 5 Cotton is a good crop in which to study the presence of resistance to insecticides 6 due to the high number of treatments that may be applied (Deguine et al., 2008), 7 but until now no specific study has been done in this crop in Spain, even though, 8 with Greece and Bulgaria, it is one of the few countries in the European Union where 9 cotton is cultivated. On the other hand, our study complements others carried out in 10 Spain to analyze the status of resistance in H. armigera in different crops (Torres-Vila 11 et al., 2002a,b). This study covered two seasons: the 1999 season, in which the 12 baseline resistance to the insecticides that are most frequently used against H. 13 armigera in cotton was established with a laboratory strain, and compared with a 14 field population, and the 2004 season, in which the evolution of susceptibility to the 15 same insecticides over the period of 5 years, and control failures of H. armigera in 16 cotton reported in the previous 2003 season, were analysed. 17 18 2. Material and methods 19 The H. armigera populations used in this study were obtained from cotton fields 20 in Seville and Córdoba (Andalucía, southern Spain) from larvae collected in different 21 plots. In 1999, the laboratory strain was created from a number of pupae from a 22 colony that had been maintained in the University of Córdoba by Professor E. Vargas- 23 Osuna since 1996, with periodical reintroductions of individuals from cotton fields, 24 5 and with no contact with any kind of insecticide. The 1999 field strain was obtained 1 in June by collecting larvae from different cotton fields from the province of Seville. 2 In the 2004 season, two populations were collected, the first in June (prior to 3 application of insecticides against H. armigera in that season) from larvae from 4 different cotton fields in the province of Seville in which control failures of H. 5 armigera had been observed in the previous 2003 season, and the other in 6 September, from the same or nearby cotton plots. 7 8 Larvae were brought to the laboratory and reared following the methodology of 9 Poitout and Bues (1974), at 26±1 ºC and 55±10 %RH with a 16:8(L:D) photoperiod. 10 Adults were reared in vertical cylindrical tubes of 11 cm in diameter and 26 cm in 11 height, constructed from laboratory drying paper and covered with paper. Three to 12 five pairs of adults were placed in each tube for mating and oviposition in the 13 conditions described above, and were fed with a 10% honey solution. Eggs were 14 collected in cotton bundles and placed in containers. After hatching, larvae were fed 15 with a semi-synthetic diet (Poitout and Bues, 1974) and reared in the conditions 16 described above. Third instar larvae were transferred to individual cages of 4 cm in 17 diameter and 2 cm in height, with a portion of diet, for insecticide bioassays. 18 19 Laboratory testing of the insecticides was carried out with the first or second 20 generation of larvae reared in the laboratory. Toxicological bioassays were performed 21 on third instar larvae using the standard topical application procedure as 22 recommended by the Entomological Society of America (Anonymous, 1970) and 23 following the guidelines of Robertson and Preisler (1992). Four products were used: 24 6 endosulfan (Thiodan-35, 350 g l-1 a.i. from Bayer CropScience), methomyl (Bonsul, 1 200 g l-1 a.i. from Bayer CropScience), chlorpyrifos (Cuspide-48, 480 g l-1 a.i. from 2 Comercial Química Massó), and lambda-cyhalothrin (Karate King, 25 g l-1 a.i. from 3 Syngenta). For each insecticide, serial dilutions in acetone [or 0.075% humectant 4 (Agridexa, from Bayer CropScience) in water for lambda-cyhalothrin] were prepared 5 so that each was one-half of the previous dose. A drop of solution (2 µl) was applied 6 with a 0.5–10 µl micropipette to the thoracic dorsum of each third instar larvae. 7 Controls for each replicate were treated with acetone (or water with Agridexa) alone. 8 The cages were closed with a lid ventilated by a hole covered with mesh, and left in 9 the same rearing conditions described above. For each strain, at least 240 larvae 10 were treated per insecticide, usually with 4 replicates of 10 larvae at each of 5 or 11 more insecticide concentrations. After 48 hours mortality was assessed: larvae were 12 considered dead if they were desiccated, or unable to move when disturbed with a 13 brush. 14 15 Dose-mortality regressions, LD50’s and their fiducial limits were estimated by 16 probit analysis (Finney, 1971) using the POLO Plus 1.0 software (LeOra Software, 17 Berkeley, CA, USA). Resistance factors (RF) were estimated at the LD50 level as RF= 18 LD50 collected strain/ LD50 reference strain (either the 1999 laboratory strain or the 19 strains collected in the field in 1999 or June 2004). The insecticide resistance level 20 was classified according to the criteria reported in Torres-Vila et al. (2002a,b): 21 susceptibility (RF=1), low resistance (RF=2–10), moderate resistance (RF=11–30), 22 high resistance (RF=31–100), and very high resistance (RF=>100). 23 24 7 3. Results and Discussion 1 The resistance in Helicoverpa armigera to different insecticides that are usually 2 applied in Spain on cotton was tested in two seasons (Table 1). In all cases, the t- 3 ratio of the slope was significant (P<0.05), so that a significant dose-response line 4 was obtained. 5 6 Both inter-season and intra-season comparisons can be made with the data. In 7 the inter-season comparison, the results showed that the LD50 of all products tested 8 had increased at the end of the 2004 season (in September), when compared 9 against the LD50 of the 1999 laboratory strain, with the fiducial limits not 10 overlapping, except in chlorpyrifos. This increment was most obvious with 11 endosulfan, which reached the category of moderate resistance (RF=11.4), whereas 12 the other products were placed in the category of low resistance (RF range between 13 1.9 and 6.0). Comparison of the data from September 2004 with the field population 14 of 1999 produced RFs that ranged from 1.1 to 3.1, i.e., from susceptibility to low 15 resistance, but only endosulfan had fiducial limits that did not overlap, indicating an 16 increment in its LD50 in this period. 17 18 The second kind of comparison is intra-season. The only comparison that is 19 meaningful is for the year 2004, when field populations were collected at the 20 beginning and at the end of the season. The September population showed 21 susceptibility/low resistance (RF between 0.6 and 1.7) when compared with the June 22 population, but no increment of the LD50 was detected, with the fiducial limits 23 overlapping for all the insecticides. 24 8 1 The results obtained in this work are similar to those obtained in a study by 2 Torres-Vila et al. (2002a,b) during 1995 to 1999, which included the same 3 insecticides tested here, but in different crops (excluding cotton). They found the 4 LD50 of endosulfan to be between 0.93 and 6.70 µg/larva, which includes our result. 5 In terms of the RF, the value obtained in this study was higher (RF=11.4, included in 6 the moderate resistance category), but this is probably because it was compared 7 with a more susceptible laboratory strain. If the less susceptible strain of Torres-Vila 8 et al. (2002a) had been used instead, then the RF would be 3 (included in the low 9 resistance category). 10 11 The other insecticides tested in this study (i.e., methomyl, chlorpyrifos and 12 lambda-cyhalothrin) followed the same pattern as endosulfan: their LD50 is included 13 in the range of values reported in Torres-Vila et al. (2002a,b), and with RF values 14 that are always within the susceptibility to low resistance category. 15 16 A clear increment of the LD50 has been detected in most of the insecticides tested 17 between 1999 and 2004, compared with the laboratory strain. Only a small change 18 was observed in the LD50 in the period from 1999 to 2004 in the field strains (only 19 significative in the case of endosulfan), and no significant increment was detected in 20 the 2004 season. Overall, our results showed little or no difference from the results 21 obtained in the general study carried out from 1995 to 1999 by Torres-Vila et al. 22 (2002a,b) in Spain, even though the study presented here included two seasons, 23 1999 and 2004, that were very separate in time. Such results indicate a general 24 9 scenario of low resistance of H. armigera to almost all the products tested in Spain, 1 except for some pyrethroids (cypermethrin, lambda-cyhalothrin and deltamethrin), 2 for which Torres-Vila et al. (2002b) found high or very high levels of resistance, 3 although only in 4 out of 111 insecticide–strain combinations tested, involving 2 out 4 of 35 H. armigera strains. 5 6 The level of resistance in H. armigera to endosulfan in our study can be 7 considered as low to moderate. This is similar to the situation in other cotton- 8 growing regions such as those in Turkey (Ugurlu and Gurkan, 2007) and India 9 (Kranthi et al., 2002; Suryawanshi et al., 2008). In terms of other insecticides, like 10 methomyl, several organophosphates, and especially pyrethroids (such as 11 cypermethrin, deltamethrin, lambda-cyhalothrin, and others), there are many studies 12 from different countries reporting high levels of resistance in this species (Kranthi et 13 al., 2002; Pietrantonio et al., 2007; Ugurlu and Gurkan, 2007; Suryawanshi et al., 14 2008). Studies of insecticide resistance in Helicoverpa/Heliothis in Europe have not 15 been limited to Spain, and in France there is also evidence of significant levels of 16 resistance to pyrethroids (deltamethrin), and moderate resistance to methomyl in H. 17 armigera , but no resistance has been detected against endosulfan (Bues et al., 18 2005). 19 20 The development of insecticide resistance is primarily the result of the selection 21 pressure exerted on sprayed populations, which increases the frequency of resistant 22 individuals. However, the cotton crop in Spain is sprayed with insecticides against H. 23 armigera from 2 to 6 times per season (Durán, 2003), which does not help to explain 24 16 Table 1. Toxicity of topically applied insecticides, currently used in cotton in 1 Spain, to different strains of Helicoverpa armigera larvae (F1 and F2, third instar), in 2 two years. 3 Insecticide Year–Strain Intercept Slope±S.E. LD50 (µg/larva) 95% fiducial limits RFa Endosulfan 1999–Laboratory 6.31 2.19±0.34 0.25 0.18–0.32 -- 1999–Field 5.05 1.81±0.26 0.93 0.72–1.28 3.7 2004–June 4.54 2.11±0.27 1.64 1.30–2.05 6.6 2004–September 4.10 1.97±0.33 2.86 1.44–5.03 11.4 Methomyl 1999–Laboratory 6.57 1.43±0.28 0.08 0.05–0.12 -- 1999–Field 6.14 1.92±0.27 0.26 0.19–0.33 3.3 2004–June 5.46 1.01±0.16 0.35 0.22–0.50 4.4 2004–September 5.47 1.48±0.22 0.48 0.19–0.90 6.0 Chlorpyrifos 1999–Laboratory 7.83 3.06±0.40 0.12 0.09–0.14 -- 1999–Field 7.16 3.21±0.46 0.21 0.17–0.25 1.8 2004–June 5.88 2.21±0.28 0.40 0.32–0.50 3.3 2004–September 6.41 2.22±0.20 0.23 0.11–0.41 1.9 Lambda-Cyhalothrin 1999–Laboratory 7.88 1.63±0.28 0.02 0.00–0.03 -- 1999–Field 7.12 1.40±0.25 0.03 0.02–0.04 1.5 2004–June 6.87 1.64±0.19 0.07 0.06–0.09 3.5 2004–September 7.08 1.85±0.27 0.08 0.05–0.12 4.0 In all strain–insecticide combinations the probit dose-response lines were significant ( t -ratio 4 test, P <0.05). 5 a Resistance factor (RF) estimated as RF= LD50 specific strain/LD50 1999 laboratory strain. 6 7