Mating Disruption of Lobesia botrana (: ): Effect of Formulations and Concentrations Author(s): Dvora Gordon, Tirtza Zahavi, Leonid Anshelevich, Miriam Harel, Shmulik Ovadia, Ezra Dunkelblum, and Ally Rachel Harari Source: Journal of Economic Entomology, 98(1):135-142. 2005. Published By: Entomological Society of America DOI: http://dx.doi.org/10.1603/0022-0493-98.1.135 URL: http://www.bioone.org/doi/full/10.1603/0022-0493-98.1.135

BioOne (www.bioone.org) is a nonprofit, online aggregation of core research in the biological, ecological, and environmental sciences. BioOne provides a sustainable online platform for over 170 journals and books published by nonprofit societies, associations, museums, institutions, and presses. Your use of this PDF, the BioOne Web site, and all posted and associated content indicates your acceptance of BioOne’s Terms of Use, available at www.bioone.org/page/ terms_of_use. Usage of BioOne content is strictly limited to personal, educational, and non-commercial use. Commercial inquiries or rights and permissions requests should be directed to the individual publisher as copyright holder.

BioOne sees sustainable scholarly publishing as an inherently collaborative enterprise connecting authors, nonprofit publishers, academic institutions, research libraries, and research funders in the common goal of maximizing access to critical research. HORTICULTURAL ENTOMOLOGY Mating Disruption of Lobesia botrana (Lepidoptera: Tortricidae): Effect of Pheromone Formulations and Concentrations

DVORA GORDON,1 TIRTZA ZAHAVI,2 LEONID ANSHELEVICH,1 MIRIAM HAREL,1 3 1 1, 4 SHMULIK OVADIA, EZRA DUNKELBLUM, AND ALLY RACHEL HARARI

J. Econ. Entomol. 98(1): 135Ð142 (2005) ABSTRACT The reluctance of Israeli vine growers to adopt the mating disruption technique to control the Lobesia botrana Den. & Schiff. has been attributed to the high cost of this method compared with that of traditional control. In this study, we tested the possibility of reducing the cost, Þrst by testing different pheromone formulations (and thus open the market for competition) and second by reducing the pheromone concentration used in vineyards. Comparisons were made between two pheromone formulationsÑShin-Etsu (Tokyo, Japan) at 165 g/ha and Concep (Sutera, Bend, OR) at 150 g/haÑand between two concentrations of Shin-Etsu, 165 and 110 g/ha. Pheromone dispensers were placed at the onset of the second moth generation. Comparison of the numbers of clusters infested with eggs and larvae of L. botrana showed no signiÞcant differences in the perfor- mance, either between the two formulations, or between the two tested concentrations. The results suggest that 1) the two formulations are equally effective, and 2) a low pheromone concentration is sufÞcient to maintain good control of small populations of L. botrana. However, when the population is high, efÞcacy is not improved by increasing the pheromone concentration. Therefore, in the interest of reducing the relatively high cost of mating disruption, we emphasize that increasing the pheromone concentration does not provide improved control of high populations of L. botrana. The cost of mating disruption can be diminished by reducing the applied pheromone concentration and by using the least expensive pheromone formulations

KEY WORDS Lobesia botrana, mating disruption, pheromone, vineyards

THE TORTRICID MOTH Lobesia botrana Den. & Schiff. is For L. botrana, the active component of the female a key pest of vineyards in Israel. The moth is prevalent sex pheromone has been identiÞed as (E,Z)-7,9-do- over a wide area in the Middle East and Europe (Louis decadienyl acetate (E7,Z9Ð12:Ac) (Roelofs at al. 1973, and Schirra 2001, Varner et al. 2001), and it breeds two Buser at al. 1974), and in the last decade, several to three generations per year in Europe and three to studies in Europe have evaluated the use of mating four in Israel. The damage to the grape yield is two- disruption against this pest. Some were very successful fold: direct damage by feeding larvae and indirect in reducing the pest population and damage (Neu- damage caused by increased susceptibility of the ber- mann et al. 1993, Charmillot and Pasquier 2000, Kast ries to gray mold, Botrytis cinerea (Fermaud and Le 2001), whereas others reported difÞculties (Arias et al. Menn 1992). 1992, Borgo 1992, Perez Martin 1992, Schmid and Sex are used in agricultural systems Raboud 1992). Factors invoked to explain the failure with two main objectives: 1) to bait traps used for of mating disruption against various pests are monitoring ßight activity and population dynamics high population density of the pest (Vickers and Roth- (Anshelevich et al. 1994); and 2) to control pests, as in schild 1991) and the immigration of gravid females mating disruption, and attract-and-kill techniques into the treated plots (Roehrich et al. 1977, Roehrich (Arn and Louis 1997, Charmillot et al. 2000). The and Charles 1982, Vickers and Rothschild 1991). In advantages of using pheromones are that, unlike in- addition, a failure of mating disruption can occur as a secticides, they are generally nontoxic and are species result of application of suboptimal pheromone con- speciÞc. However, despite considerable progress that centrations or formulations (Roehrich et al. 1979, has been achieved the mating disruption technique is Sauer and Karg 1998). not widely applied (Karg and Sauer 1997). One of the main factors affecting mating disruption of L. botrana is the competition between the phero- mone released from the dispensers and the femalesÕ 1 Department of Entomology, Volcani Center, Bet Dagan 50250, natural pheromone (Schmitz et al. 1995). Therefore, Israel. adding more pheromone-dispensing tubes in a given 2 Extension Service, Ministry of Agriculture, Kiryat Shmona, Israel. 3 Carmel Mizrahi Winery, Zichron YaÕacov, Israel. area is expected to improve effectiveness, Þrst by 4 Corresponding author, e-mail: [email protected]. increasing the number of pheromone source points

0022-0493/05/0135Ð0142$04.00/0 ᭧ 2005 Entomological Society of America 136 JOURNAL OF ECONOMIC ENTOMOLOGY Vol. 98, no. 1

Table 1. Plot sizes and grape cluster infestation levels (avg % ؎ SD) in vineyards treated with two different formulations of L. botrana mating disruption pheromone

Cultivar ÔCabernet SauvignonÕ ÔColombardÕ ÔSauvignon BlancÕ Treatment Shin-Etsu Concep Shin-Etsu Concep Shin-Etsu Concep Plot size (ha) 10.5 7.5 6.5 2.8 2.3 1.7 Highest infestation level (%) 26.0 Ϯ 18.97 18.0 Ϯ 13.98 10.0 Ϯ 8.16 7.0 Ϯ 8.03 6.0 Ϯ 6.95 8.0 Ϯ 9.19 Mean infestation level (%) 2.0 Ϯ 5.75 2.95 Ϯ 5.35 0.96 Ϯ 2.18 0.65 Ϯ 1.90 0.93 Ϯ 1.62 2.07 Ϯ 2.46 No. of applications 0 0 1111 and second by increasing the overall pheromone con- Once a week, all traps were checked for captured centration in the vineyard atmosphere (Sauer and males, and 100 randomly selected intact clusters were Karg 1998). examined for the presence of larvae and eggs. Control The cost of mating disruption is relatively high com- measures in pheromone-treated plots were taken pared with that of the traditional insecticide control when two or more clusters were found to be infested (Knight 1995, Arn and Louis 1997), but Arn and Louis with either eggs or larvae. The pheromone dispensers (1997) suggested that pheromone dispensers may in the traps were replaced every 3 wk. have been deployed in greater numbers than neces- The pheromone release rate from each formulation sary. Hence, the cost of mating disruption could be was determined every 2Ð3 wk in the laboratory, by reduced by improving the efÞciency of its use, e.g., by gas-chromatography (GC), each measurement being reduction of the applied pheromone concentration. carried out in three replicates. The dispensers were After the registration of mating disruption as a cut into small pieces and extracted by sonication for means of controlling L. botrana in Israel in 2000, the 1 h with dichloromethane (10 ml for the Shin-Etsu technique was adopted in commercial vineyards. Its tubes and 20 ml for the Concep bags). The samples high cost, however, discouraged many growers from were kept overnight before analysis. Aliquots of 100 ␮l joining the disruption campaign. Accordingly, the aim were diluted with 900 ␮l of solvent, and a 100-␮l of the current study was to examine the possibility of subsample of the diluted sample was further diluted reducing application costs. Comparisons were made with 850 ␮l of solvent and 50 ␮g of 15:Ac in 50 ␮lof between the efÞcacy of two pheromone formulations dichloromethane as an internal standard. GC analysis and two application rates, in disrupting the mating of was performed on a 30 m ϫ 0.25 mm HP5 capillary L. botrana in vineyards, and in limiting the pest pop- column, operated in the splitless mode with 1.5 ml/ ulation and the consequent damage. min ßow helium carrier gas. The column was kept at 60ЊC for 2 min and then programmed at 15ЊC/min to 180ЊC, at which it was held for 15 min. The injector and Materials and Methods detector were maintained at 220ЊC, and the purge Efficacy of Two Pheromone Formulations for Mat- valve was opened after 2 min. Each point in the release ing Disruption. We tested the efÞcacy of two different curves is the average of three replicates. formulations for disrupting L. botrana mating, both Efficacy of Two Pheromone Application Rates for based on the active E7,Z9Ð12:Ac isomer with 15Ð20% Mating Disruption. This experiment was carried out in of the E7,E9Ð12:Ac isomer and trace amounts of the three vineyards; one in the Golan Heights (ÔChardon- other isomers. The two formulations were 1) a poly- nayÕ) and two west of Jerusalem (ÔColombardÕ and ethylene tube dispenser manufactured by Shin-Etsu ÔCheninÕ) (plot sizes are indicated in Table 2). We (Tokyo, Japan) that contained an average of 220 mg of compared two application rates of Shin-EtsuÕs dis- active ingredient per tube, and (2) a bag dispenser pensersÑ110 and 165 g/ha (500 and 750 tubes per manufactured by Concep (Sutera, Bend, OR) con- hectare, respectively)Ñfor their efÞcacy in reducing taining an average of 300 mg of pheromone per bag. the damage caused by L. botrana. The tubes were Shin-EtsuÕs tubes were placed 4.5 m apart in every placed in every row of the test plots, spaced 6.5 and row, to provide a pheromone application rate of 165 4.5 m apart for the lower and higher pheromone ap- g/ha (750 tubes per hectare), and ConcepÕs bags were plication rates, respectively. Control plots were placed 6.5 m apart in every row, to provide an appli- treated with as required. Buffer plots be- cation rate of 150 g/ha (500 bags per hectare). The tween the control and the low pheromone application rows in each vineyard were 3 m apart. rate plots served as wide borders to reduce the mi- The tests were conducted in three vineyards, all gration of gravid females from the control plots to the located in the Golan Heights, and each with a different pheromone-treated plots. The buffer plots were cultivar: ÔCabernet SauvignonÕ, ÔColombardÕ, and ÔSau- equipped with pheromone dispensers at a low appli- vignon BlancÕ (plot sizes are indicated in Table 1). The cation rate and were treated with insecticide as re- dispensers were placed in the vineyards on Þrst May quired. (during the second moth generation of the season), The dispensers were placed in the vineyards at the attached to the higher of the vine-supporting wires, in onset of the second moth generation of the season the vicinity of dense vine foliage. In each treatment, (Þrst May 2001), as described before. Pheromone dis- six pheromone traps (Unitraps) also were placed in rupted plots were treated with according the vineyards. to the above-deÞned threshold of infestation. The February 2005 HARARI ET AL.: MATING DISRUPTION IN VINEYARDS 137

Table 2. Plot sizes and grape cluster infestation levels (avg % ؎ SD) in vineyards treated with two different concentrations of L. botrana mating disruption pheromone

Cultivar ÔChardonnayÕ ÔCheninÕ ÔColombardÕ Treatment 500 750 Control 500 750 Control 500 750 Control (tubes/ha) Plot size (ha) 3 3 6333333 Highest infestation 22.0 Ϯ 13.17 9.0 Ϯ 18.85 37.0 Ϯ 17.03 9.0 Ϯ 11.01 7.0 Ϯ 10.59 10.0 Ϯ 10.54 14.0 Ϯ 11.75 11.0 Ϯ 8.76 8.0 Ϯ 6.36 level (%) Mean infestation 2.41 Ϯ 4.43 1.63 Ϯ 1.37 6.7 Ϯ 10.47 4.63 Ϯ 3.96 2.12 Ϯ 3.23 3.0 Ϯ 3.34 5.5 Ϯ 8.31 3.0 Ϯ 6.89 3.0 Ϯ 5.43 level (%) Fungi infested 25.0 Ϯ 26.77 9.0 Ϯ 11.08 5.0 Ϯ 7.07 16.0 Ϯ 16.47 14.0 Ϯ 16.47 3.0 Ϯ 4.83 clusters (1) Fungi infested 34.0 Ϯ 20.60 21.0 Ϯ 16.63 19.0 Ϯ 11.97 46.0 Ϯ 15.05 28.0 Ϯ 19.89 38.0 Ϯ 20.98 clusters (2) No. of Pesticide 003003226 applications control plots were treated with insecticides according ysis of variance (ANOVA) (SAS Institute 1988). The to the same threshold of infestation and concurrently average numbers of all males captured in each trap with the capture of more than two males per trap per during the second period were compared between night. cultivars by means of GLM software (SAS Institute Six traps (Unitraps) loaded with the female sex 1988), to evaluate the effects of the different cultivars pheromone were placed in each plot at the onset of on the male catches. Because the cultivar effect on the male ßight in March, i.e., 5 wk before the phero- male catches was found to be signiÞcant, the cultivars mone tubes were deployed to measure the initial local were treated separately. population. The same traps were used for monitoring To compare the mean percentages of infested clus- the males after the deployment of the pheromone ters under two different pheromone application rate dispensers. Once a week, all traps were checked for treatments and the control, we Þrst examined whether captured males, and, at the same time, 100 randomly the different grape cultivars had an effect on the per- selected intact clusters were examined for the pres- centage of clusters infested by the immature stages of ence of larvae and/or eggs. L. botrana and whether there was a detectable inter- In addition, in two of the tested vineyards (ÔCheninÕ action between the cultivar and the application rate. and ÔColombardÕ) attempts were made to assess the We tested for such an interaction by means of the effects of the two pheromone application rates on the GLM procedure (SAS Institute 1988) after applying development of decay on berries, because such de- an arcsine modiÞcation to the square root of the in- velopment may serve as an indicator of previous moth festation rate. Because there was no signiÞcant effect damage. For this purpose, we monitored 100 intact of the cultivars on the percentage of infested clusters clusters (10 replicates, each of 10 clusters within a 1-m and an interaction between the treatments (applica- length of a row) for signs of rotten berries on two tion rates) and the cultivars, the different cultivars occasions, 1 mo and 1 d, before harvest. could be treated as replicates. To compare the mean percentages of clusters in- fected by rot-causing organisms (mainly Aspergillus Statistical Analysis niger) under two different pheromone application Pheromone Formulations. We compared the per- rate treatments and in the control, we Þrst tested the centages of clusters infested by eggs and larvae in plots possible effects of the different vine cultivars on the treated with the two different pheromone formula- percentage of decayed clusters and of an interaction tions. We used the GLM software (SAS Institute 1988) between the cultivar and the application rate. We to evaluate the effects of the formulations on the tested this by using the GLM procedure (SAS Institute percentages of infested clusters, the effects of grape 1988) after applying an arcsine modiÞcation to the cultivars, and the possible effect of an interaction square root of the infestation rate. Because there were between the two, after an arcsine modiÞcation of the no signiÞcant cultivar or interaction effects, the cul- square root of the infestation rate. The absence of any tivars could be treated as replicates. signiÞcant cultivar effect on the percentage of infested clusters, and of any interaction between the treat- Results ments (formulations) and the cultivars, indicated that the different cultivars could be treated as replicates. Efficacy of the Two Pheromone Formulations for Pheromone Application Rates. The males captured Mating Disruption. The pheromone residue of the in traps were divided according to two trapping pe- two formulations was determined by gas chromatog- riods: before and after the deployment of the phero- raphy. The release rates were constant so that the mone tubes. The average numbers of all males cap- residual formulation contents diminished linearly tured in each trap during the Þrst period were with increasing time, with R2 values of 0.989 and 0.975 compared between vineyards by application of anal- for Shin-Etsu tubes and Concep bags, respectively. 138 JOURNAL OF ECONOMIC ENTOMOLOGY Vol. 98, no. 1

Fig. 1. Mating disruption pheromone dissipation from dispensers over time as analyzed by gas chromatography. (A) Shin-Etsu. (B) Concep. Each point represents the average of three replicates.

The tubes and the bags lasted 120 and 140 d, respec- tively (Fig. 1a and b), and the isomeric composition of the pheromone in the formulations remained almost constant for Ͼ100 d. In a typical analysis the com- pound ratio of E7,Z9:E7,E9 changed slightly from 83:17 to 78:22 after 120 d. Male Capture. From the Þrst week after the appli- cation of the mating disruption formulation until har- vest, a small number of males were captured in the treated plots (trap shutdown). The total number of males captured in any trap in any cultivar under either Fig. 2. Percentages of grape clusters infested by eggs and larvae of L. botrana in vineyards treated with two different of the two formulations did not exceed Þve during the mating disruption pheromone formulations. (a) ÔCabernet whole season. SauvignonÕ. (b) ÔSauvignon BlancÕ. (c) ÔColombardÕ. Infestation of Grape Clusters by Eggs and Larvae. The effect of the cultivar and that of the formulations used on the percentages of infested clusters were not a lower damage percentage than Concep bags, but 1 signiÞcant (GLM, F ϭ 1.58, df ϭ 2, P Ͼ 0.05 and F ϭ wk before harvest, the damage levels rose to 6% with 0.39, df ϭ 1, P Ͼ 0.05, respectively). The same was true the Shin-Etsu tubes and to only 2% with the Concep for the interaction between the two (F ϭ 1.25, df ϭ 2, bags. Both plots were treated only once with pesti- P Ͼ 0.05). Therefore, the cultivars were treated as cides (24 June) (Table 1; Fig. 2b). replicates. Early in the season, there were slightly more dam- No signiÞcant difference was detected between the aged clusters in Colombard plots protected by Shin- effects of the two pheromone formulations on the Etsu tubes than in those with Concep bags, but toward percentages of clusters infested by eggs and larvae the end of the season, the incidence of infested clus- ϭ Ͼ (ANOVA, F124,1 0.149; P 0.05). ters in Shin-Etsu plots dropped practically to zero, In ÔCabernet SauvignonÕ, during most of the season, whereas that in Concep plots rose to 6%. Plots treated the Shin-Etsu tubes kept the percentage of damaged with each formulation received one pesticide appli- clusters at a slightly lower level than the Concep bags. cation (10 June) (Table 1; Fig. 2c). In spite of the However, at the end of the season, the percentage of above-mentioned variations, the differences between damage with the Shin-Etsu formulation had risen to the two formulations in the mean percentages of in- the high level of 25%, whereas the Concep formulation festation were insigniÞcant in all cultivars (Table 1; kept the damage under 20%. Neither plot was treated Fig. 2) with pesticides (Table 1; Fig. 2a). Efficacy of Two Pheromone Application Rates on A similar pattern was noticed in ÔSauvignon BlancÕ. Male Mating Disruption. Male Capture. The mean During most of the season, Shin-Etsu tubes maintained total numbers of males captured in the traps during 5 February 2005 HARARI ET AL.: MATING DISRUPTION IN VINEYARDS 139 wk before application of the mating disruptive pher- omone were used to estimate the initial population size in each vineyard. The results of the male trapping showed that the early-season pest population in ÔChar- donnayÕ was signiÞcantly higher than those in ÔCheninÕ ϭ Ͻ or ÔColombardÕ (ANOVA, F51,2 7,229; P 0.01) (Table 2). A signiÞcant cultivar effect on the male catches was found when the average total numbers of males captured in traps in the second period were compared (F ϭ 6.73, df ϭ 4, P Ͻ 0.001); therefore, the effect of the pheromone application rate on male capture had to be tested separately in each vineyard. From the Þrst week after the application of the dis- ruptive pheromones until harvest, signiÞcantly more males were captured in the control plots of all vine- yards than in any of the pheromone-treated plots, where only a few males were captured (ANOVA with TukeyÕs honestly signiÞcant difference multiple com- ϭ Ͻ parison test: ÔChardonnayÕ, F15,2 19.152; P 0.001; Ͻ ϭ Tukey, MSE of 36.733, P 0.001; ÔColombardÕ, F15,2 17.956, P Ͻ 0.001, Tukey, MSE of 493.8, P Ͻ 0.001; and ϭ Ͻ ÔCheninÕ, F15,2 34.597, P 0.001; Tukey, MSE of 35.033, P Ͻ 0.001). Percentages of Clusters Infested by L. botrana Eggs and Larvae. When the average infestation percent- ages of clusters protected by the different pheromone application rates and in the control were compared, no signiÞcant effect of the cultivar and of the application rates were detected (F ϭ 0.17, df ϭ 2, P Ͼ 0.05 and F ϭ 1.46, df ϭ 2, P Ͼ 0.05, respectively) nor of any inter- action between the cultivars and the pheromone ap- plication rates (F ϭ 1.186, df ϭ 4, P Ͼ 0.05). Therefore, the results from all cultivars were combined, with all cultivars regarded as replicates. No differences be- ϭ tween the treatments were found (ANOVA, F118,2 0.990; P Ͼ 0.05), but because each vineyard had been Fig. 3. Percentages of grape clusters infested by eggs and treated with a different number of insecticide appli- larvae of L. botrana in vineyards treated with two different cations, we analyzed the results from each vineyard mating disruption pheromone concentrations. (a) ÔChardon- separately. nayÕ. (b) ÔCheninÕ. (c) ÔColombardÕ. For ÔChardonnayÕ, both pheromone application rates signiÞcantly reduced the percentage of infected clusters compared with that found in the control plot, in the plot treated with 500 tubes per hectare than in but no signiÞcant difference was found between the the plot treated with 750 tubes per hectare. The in- ϭ two pheromone-treated plots (ANOVA, F807,2 21.88, festation levels with both pheromone application rates P Ͻ 0.001; Tukey with MSE of 0.31, P Ͻ 0.001) (Fig. were not signiÞcantly different from that in the con- ϭ Ͻ 3a). These percentages of infested clusters were ob- trol plot (ANOVA, F297, 2 4.255; P 0.01; Tukey with served after three insecticide applications in the con- MSE of 0.029, P Ͻ 0.05) (Fig. 3c). However, the con- trol plot (18 June, 1 July, and 26 July) and without any trol plot received six insecticide treatments (11 May, applications in the two pheromone-treated plots. 3 June, 1 July, 16 July, 28 July, and 7 August) compared For ÔCheninÕ, the percentages of clusters infested by with only two chemical applications in each of the eggs and larvae in the pheromone-treated plots were pheromone-treated plots (17 July and 8 August in both relatively low, throughout the season, and no signiÞ- 500 and 750 tubes per hectare). cant differences in the percentages of the infested Percentage of Clusters Infested by Rots. When the clusters between the control plot and pheromone- average percentage of decayed clusters under the dif- ϭ treated plots were noticed (ANOVA, F237,2 2.046; ferent pheromone application rates and in the control P Ͼ 0.05) (Fig. 3b). The control plot was treated three were compared 1 mo before harvest, no signiÞcant times with pesticides (6 June, 26 June, and 12 July, effect of the cultivar was found (F ϭ 0.022, df ϭ 1, P Ͼ whereas the pheromone-treated plots were not sprayed. 0.05), and no interaction between the cultivar and the For ÔColombardÕ, the mean percentage of grape pheromone application rates was detected (F ϭ 0.430, clusters infested by larvae and eggs were relatively low df ϭ 2, P Ͼ 0.05); therefore, all cultivars were com- in all the treatments. However, a signiÞcantly larger bined in the analyses. A signiÞcant effect of the pher- ϭ percentage of clusters was infested by eggs and larvae omone application rates was detected (F57,2 5.284; 140 JOURNAL OF ECONOMIC ENTOMOLOGY Vol. 98, no. 1

When testing the efÞcacy of Shin-EtsuÕs and Con- cepÕs pheromone formulation, we applied similar amounts of pheromone per hectare (165 and 150 g/ha, respectively), and because the release rates were sim- ilar, we assumed that the airborne concentrations in the plots protected by the respective formulations were the same. The shutdown of the traps in the vineyards treated with both formulations suggests that there was some overlap of the areas covered by each of the pheromone dispensers (Sauer and Karg 1998). This may explain the lack of signiÞcant difference between the efÞcacies of the two formulations. Sup- porting insecticide treatments were used when re- quired, applied at the same time in the plots protected by the respective formulations. The pheromone in ConcepÕs bags lasted 140 d, com- pared with 120 d for that in Shin-EtsuÕs tubes. The pheromone dispensers were placed at the onset of the pestÕs second generation and they were expected to last until harvest time; both formulations contain suf- Þcient pheromone for the early-harvested white Fig. 4. Percentage of grape clusters infested by gray mold grapes (usually during August) and that in ConcepÕs in vineyards treated with two different mating disruption pheromone concentrations. (a) ÔCheninÕ. (b) ÔColombardÕ. bags should be sufÞcient to cover the harvest of early- maturing red grapes (during September). However, neither formulation would persist until the harvest of P Ͻ 0.01). Whereas no signiÞcant difference was no- the late-maturing red grapes (in October); therefore, ticed between the percentages of decayed clusters exhaustion of the pheromone in the dispensers may under the two pheromone application rates (TukeyÕs explain the sharp increase in infestation that was ob- test with a model MSE of 0.067 with df ϭ 57, P Ͼ 0.05), served before the harvest of ÔCabernet SauvignonÕ a signiÞcant difference was found between the per- clusters treated with either formulation. centage of infested clusters in the control plots and One of the main factors affecting mating disruption that under the low pheromone application rate (500 in L. botrana is the competition between the phero- tubes per hectare) (Tukey with MSE of 0.067 with mone released from the dispensers, and the natural df ϭ 57, P Ͻ 0.005) (Fig. 4a). When samples were pheromone released by the females (Schmitz et al. taken 1 d before harvest, no signiÞcant effect of the 1995). Therefore, adding more pheromone tubes in a interaction between the pheromone application rates given area should enhance their efÞcacy, because of (treatment) and the cultivars was found (F ϭ 0.371, the increases in the number of pheromone source df ϭ 2, P Ͼ 0.05); therefore, all cultivars were com- points and in the overall pheromone concentration in ϭ Ͻ bined in the analyses. A signiÞcant (F57,2 3.930; P the vineyard atmosphere (Suckling and Angerelli 0.05) effect of the treatment was found, and a signif- 1996, Sauer and Karg 1998). However, when the pop- icant difference was observed between the effects of ulation is large, a male may encounter a female by the two pheromone application rates (Tukey MSE of chance, thus reducing the success of the mating dis- 0.062 with df ϭ 57, P Ͻ 0.05). However, there were ruption method (Roehrich and Charles 1982, Aude- intermediate levels of fungal infestation in the control mard 1987, Vogt 1987). plots leading to an insigniÞcant difference between the It was previously shown by Karg and Sauer (1997) control and the two pheromone application rates (Tukey that the active space of a single dispenser (BASF in with MSE of 0.062 with df ϭ 57, P Ͼ 0. 05) (Fig. 4b). their case) had a radius of 5 m. Therefore, the instal- lation of dispensers at a greater spacing may cause a drop in pheromone concentration in the air, whereas Discussion overlapping of the active volumes of the dispensers The physical characteristics of the pheromone for- would lead to an even distribution throughout the plot. mulation for mating disruption may inßuence the ef- When we compared the efÞcacy of two pheromone Þcacy of the method as a means of control. An uneven application rates, the largest distances between two release rate may cause low pheromone concentrations pheromone sources in the higher and lower phero- in certain areas, which would enable the male to re- mone-concentration treatments were 4.5 m (750 tubes gain his susceptibility to the female pheromone per hectare) and 6 m (500 tubes per hectare), respec- (Rumbo 1981), thus reducing the likelihood of sensory tively. On the assumption that the active volume of the adaptation and bringing about the failure of the mating tube dispenser was 5 m, as found by Sauer and Karg disruption methods (Carde´ 1990). In our study, the (1997), we expected a better performance by the release rates of the two pheromone formulations higher pheromone application rate, as expressed in (Shin-Etsu and Concep) were constant, as was con- reductions in the number of mating events and in Þrmed by the chemical analysis. oviposition of viable eggs, and subsequent reductions February 2005 HARARI ET AL.: MATING DISRUPTION IN VINEYARDS 141 in the percentages of damaged berries. In fact, how- a higher pheromone application rate may not be sufÞ- ever, we found that an increase in the pheromone cient because of chance encounters between the sexes. application rate did not result in better control of the Therefore, with a view to reducing the relatively pest population, leading to a reduction in the damage high cost of mating disruption treatments (Knight to berries. The percentages of clusters infested by 1995, Arn and Louis 1997), we concur with the sug- larvae and eggs were similar under the two tested gestion of Arn and Louis (1997) that increasing the pheromone application rates. pheromone application rate is not instrumental to the Charmillot et al. (1995a, b) have suggested that control of high populations of L. botrana. Mating dis- when the population density is very low, dispensers ruption could be made less expensive by reducing its placed at a greater spacing could still be effective. In cost, by reducing the applied pheromone application the current study, with ÔCheninÕ, for example, the rate, and by choosing the most appropriate phero- infestation level was restrained throughout the season. mone formulation available. No chemical control was applied in the pheromone- treated plots, compared with three chemical applications in the control. The higher pheromone application rate Acknowledgments had been expected to yield better performance, but no This publication is in the memory of the late Moshe Kehat. difference was observed between the effects of the We thank Ruth Marcus and Miri Zarhi for help in analyzing two pheromone application rates, in terms of the mean the data. We thank the chemical companies Agan and Macht- percentages of infested clusters. eshim for supplying L. botrana disruption pheromones (Shin- In ÔColombardÕ, the average percentage of clusters Etsu and Concep, respectively); the grape growers in Odem, infested by eggs and larvae was well below the eco- Gshur, Eshtaol, and Neta for allowing us to use vineyards for nomic threshold (5% of clusters infested at harvest). the experiments; and the Fruit Council of Israel and Israeli However, this level of damage was achieved after two wine board, for Þnancial support. We also thank the late insecticide applications in each of the pheromone- Reneh Modiano for editorial comments and Phyllis Wein- traub for comments on the early draft. treated plots, or six applications in the control plot. The Þnding that the percentage of infested clusters increased with advancing time until midseason, re- References Cited gardless of the insecticide treatments, suggests an ev- er-growing population, due to the inßux of gravid Anshelevich, L., M. Kehat, E. Dunkelblum, and S. Green- females into the treated plots from adjacent vineyards. berg. 1994. Sex pheromone traps for monitoring the Eu- ropean vine moth, Lobesia botrana: effect of dispenser This is consistent with the relatively high male capture type, pheromone dose, Þeld aging of dispenser, and type rates in pheromone traps in the control plot. High of trap on male captures. Phytoparasitica 22: 281Ð290. levels of fungus-infested clusters were subsequently Arias, A., M. Bueno, J. Nieto, M. Valenzuela, A. Perez, B. observed in both the control and the pheromone- Cuenda, F. Gallego, F. Alamada, and M. A. Castillo. 1992. treated plots, 1 d before harvest. Essais de confusion sexuelle de Lobesia botrana Den. et In ÔChardonnayÕ, the average percentage of infested Schiff., pendant 1989 et 1990 dans “Tierra de Barros” clusters in the control was higher than the economic (Espagne). Int. Org. Biol. Control ÐWest Palearctic Sec. threshold (5% of clusters infested at harvest) after Bull. 15: 18. three insecticide applications, which indicates a larger Arn, H., and F. Louis. 1997. Mating disruption in European vineyards, pp. 372Ð382. In R. T. Carde´ and A. L. Minks population of L. botrana in the vineyard. Indeed, a [eds.], Insect pheromone research. Chapman & Hall, signiÞcantly higher level of infestation was observed New York. in the control, where it reached 37%, compared with Audemard, H. 1987. Perspectives et problemes de la lutte 22 and 9%, respectively, in the low- and high-concen- par confusion. Utilisation des pheromones et autres me- tration pheromone-treated plots, but the difference diateurs chimiques en lutte integree. Bull. SROP/OILB, between the pheromone-treated plots was not signif- Neustadt, 3Ð4. icant, despite the better control expected in the plot Borgo, M. 1992. Ulterieurs essais de lutte par confusion sex- treated with higher pheromone application rate. How- uelle contre les tordeuses de la grappe dans les vignobles ever, the initial population in this vineyard was very de lÕest de la Venetie. International Organization for Bi- ological Control ÐWest Palearctic Sec. Bull. 15: 20. large, as was deduced from the high male capture rates Buser, H. R., S. Rauscher, and H. Arn. 1974. Sex Pheromone in pheromone traps before the application of the mat- of Lobesia botrana: (E,Z,)-7,9-dodecadienyl acetate in the ing-disruption treatment. However, it could be that female grape vine moth. Z. Naturforsch. 29c: 781Ð783. the initial moth population was too high for effective Carde´, R. T. 1990. Principle of mating disruption, pp. 241Ð control under either high or low pheromone applica- 265. In R. L. Ridgway, R. M. Silverstein, and M. Inscoe tion rates, because of the possibility of chance sporadic [eds.], Behavior-modifying chemicals for insect manage- meetings between individual males and females ment. Academic, London, United Kingdom. (Roehrich and Charles 1982, Audemard 1987, Vogt Charmillot, P. J., and Pasquier, D. 2000. Vers de la grappe: 1987, Sauer and Karg 1998). technique de confusion, lutte calsique et dynamique des populations. Rev. Suisse Vitic. Arboric. Hortic. 32: 315Ð320. In our current study, either both application rates Charmillot, P.-J., D. Pasquier, N. J. Alipaz, and U. Neuman. were sufÞcient or neither achieved good control of the 1995a. Effet dÕune ceinture de diffuseurs a attractif sexuel pest. This indicates that, when the pest population is low, sur la distribution des captures de males, des accouplements 500 tubes per hectare may be sufÞcient to provide ef- de femelles attachees et de lÕattaque de Lobesia botrana Den. fective control, but when the population is too high, even & Sciff. en vignoble. J. Ecol. Entomol. 119: 206Ð210. 142 JOURNAL OF ECONOMIC ENTOMOLOGY Vol. 98, no. 1

Charmillot, P.-J., D. Pasquier, and A. Scalco. 1995b. Lutte Roehrich, R., J. P. Charles, C. Tresor, and M. A. Vathaire. par confusion contre les vers de la grappe eudemis et 1979. Essais de Ôconfusion sexuelleÕ contre les Tordeuses cochylis a perroy et Allaman: resultats de 1995. Rev. de la grappe lÕEudemis Lobesia botrana Den. et Schiff. et Suisse Vitic. Arboric. Hortic. 27: 347Ð358. la cochylis Eupocilia ambuguella Tr. Ann. Zool. Ecol. Charmillot, P.-J., D. Hofer, and D. Pasquier. 2000. Attract Anim. 11: 659Ð675. and kill: a new method for control the codling moth Cydia Roelofs, W., J. Kochansky, R. Carde´, and H. Arn. 1973. Sex pomonella. Entomol. Exp. Appl. 94: 211Ð216. attractant of the grape vine moth, Lobesia botrana. Fermaud, M., and R. Le Menn. 1992. Transmission of Bot- Schweiz. Entomol. Ges. 46: 71Ð73. rytis cinerea to grapes by grape berry moth larvae. Phy- Rumbo, E. R. 1981. Study of single sensillium responses to topathology 82: 1393Ð1398. pheromone in , Epiphyas pstvittana, Karg, G., and A. E. Sauer. 1997. Seasonal variation of phero- using and averaging technique. Physiol. Entomol. 6: 87Ð98. mone concentration in mating disruption trials against Eu- SAS Institute. 1988. SAS userÕs guide: Statistics. SAS Insti- ropean grape vine moth Lobesia botrana (Lepidoptera: Tor- tute, Cary, NC. tricidae) measured by EAG. J. Chem. Ecol. 23: 487Ð501. Sauer, A. E., and G. Karg. 1998. Variables affecting phero- Kast, W. K. 2001. Twelve years of practical experience using mone concentration in vineyards treated for mating dis- mating disruption against Eupocilia ambiguella and Lobe- ruption of grape vine moth Lobesia botrana. J. Chem. sia botrana in vineyards of the Wuerttemberg region Ecol. 24: 289Ð302. Germany. IOBC WPRS Bull. 24: 71Ð74. Knight, A. L. 1995. Evaluating pheromone emission and Schmid, A., and G. Raboud. 1992. Essai de confusion contre blend in disrupting sexual communication of codling Lobesia botrana dans le vignoble de Sierre (VS) 1989Ð1990. moth (Lepidoptera: Tortricidae). Environ. Entomol. 24: Int. Org. Biol. Control ÐWest Palearctic Sec. Bull. 15: 23. 1396Ð1403. Schmitz, V., R. Roehrich, and J. Stockel. 1995. Disruption Louis, F., and K. J. Schirra. 2001. Mating disruption of Lobesia mechanisms of pheromone communication in the Euro- botrana (Lepidoptera: Tortricidae) in vineyards with very pean grape moth Lobesia botrana Den. & Schiff. (Lep., high population densities. IOBC WPRS Bull. 24: 75Ð80. Tortricidae) II. Inßuence of the population density and Varner, M., L. Roberto, M. Luisa, and F. Flavia. 2001. Ex- the distance between for males to detect the perience with mating disruption technique to control females in atmosphere impregnated by pheromone. grape berry moth, Lobesia botrana, in Trentino. IOBC J. Appl. Entomol. 119: 303Ð308. WPRS Bull. 24: 81Ð88. Suckling, D. M., and N.P.D. Angerelli. 1996. Pheromone Neumann, U., A. Schimd, C. Ioriatti, M. Verner, R. Castillo, point source distribution affects antennal spike fre- A. Lucas, J. L. Perez-Martin, and M. A. Castillo. 1993. La quency and communication disruption of Epiphyas technique par confusion contre les “vers” de la grappe en postvittana (Lepidoptera: Tortricidae). Environ. Ento- Europe aujourdÕhui. Phytoma 456: 15Ð17. mol. 25: 101Ð108. Perez Martin, J. L. 1992. Lutte contre Lobesia botrana de la Vickers, R. A., and G.H.L. Rothschild. 1991. Use of sex pher- vigne par la technique de confusion sexuelle en la Rioja omone for control of codling moth, pp. 339Ð354. In L.P.S. (Espagne). Int. Org. Biol. Control ÐWest Palearctic Sec. van der Geest and H. H. Evenhuis [eds.], Tortricid : Bull. 15: 19. their biology, natural enemies and control. Elsevier, Am- Roehrich, R., and J. P. Charles. 1982. Essai de confusion sterdam. The Netherlands. sexuelle en vignoble contre lÕEudemis de la vigne, Lobesia Vogt, H. 1987. Several years of experience with the mating botrana Schiff. Colloq. INRA. 7: 365Ð371. disruption technique for the control of the European Roehrich, R., J. P. Charles, and C. Tresor. 1977. Essai pre- grape moth Eupocilia ambiguella Hbn. Bull. SROP/OIB liminaire de protection du vignoble contre Lobesia bo- Newstadt. 3: 5Ð6. trana Schiff. au moyen de la pheromone sexuelle de synthese (methode de confusion). Rev. Zool. Agric. Pathol. Veg. 76: 25Ð36. Received 19 November 2003; accepted 7 June 2004.