Failure of Pheromone Traps in Detecting Incipient Populations of Boll Weevils (Coleoptera: Curculiondae): Investigation of Two Potential Contributing Factors
The boll weevil, Anthonomus grandis Boheman, has been eradicated from >90% of the cotton, Gossypium hirsutum L., acreage in TX, but economically-damaging populations still exist in South TX (Anonymous 2013; http://www.txbollweevil.org/). Although eradication efforts have substantially reduced boll weevil populations in this area, eradication progress has been at a standstill during the past several years. The lack of progress has been attributed to several factors, including the inconsistent performance of pheromone traps in detecting incipient boll weevil populations. In 2008 substantial infestations of boll weevils were detected in several cotton fields in Medina Co., TX, but adjacent traps failed to capture weevils during the preceding weeks. Because eradication programs rely almost exclusively on pheromone traps to detect incipient boll weevil populations and to indicate the need for insecticide treatments, factors that adversely affect captures of weevils in traps need to be identified and resolved. We investigated the possible existence of a boll weevil population in Medina Co., TX, that produces a unique blend of pheromone and, subsequently, no longer responds to the standard grandlure formulation. We also examined the quality of pheromone lures used in Medina Co. during the 2008 trapping season. Presented, herein, are the results of our investigation.
The existence of a boll weevil population in Medina Co., TX, that produces a unique pheromone blend was investigated by examining the ratio of the 4 pheromone components produced by boll weevils collected from traps and cotton fields. Pheromone traps (Technical Precision Plastics, Mebane, NC), baited with lures containing 10 mg of grandlure (Hercon Environmental Inc., Emigsville, PA), were established along brushlines in areas where detection failures were previously reported. Captured weevils were collected weekly between 14 April and 12 May, and lures were replaced biweekly during that period. Field-collected adults were initially obtained by sampling 5 near-by cotton fields with a Keep-It-Simple-Sampler (Beerwinkle et al. 1997; pp. 1330 - 1333, In Proc. Beltwide Cotton Conf.) between 10 and 22 June 2009. Because only 6 male weevils were collected with this method, adults also were reared from oviposition-punctured squares using procedures described by Spurgeon and Suh (2007; J. Entomol. Sci. 42: 250 - 260). Trap- and field-collected adults were sexed using the tergal notch method (Sappington and Spurgeon 1998, Ann. Entomol. Soc. Am. 93: 610 - 615), and 24 - 30 males from each collection were held individually in 100 × 15-mm Petri plates. Each male was provided a 2.5-cm section of cotton wick saturated with distilled water and was fed a freshly-picked square (6 - 9 mm diam. with bracts intact; greenhouse-grown plants) daily for 6 - 8 d prior to pheromone collection. Weevils were held at 29.4 ± 1°C and a photoperiod of 14:10 (L:D) h during the feeding period, and only those that fed on squares every day during the feeding period were selected for pheromone production assays. Following each feeding period, up to 16 weevils were randomly selected and transferred individually into pheromone collection vessels, and pheromone was collected over a 24-h period using methods described by Spurgeon and Suh (2009; J. Entomol. Sci. 99: 323 - 330). Weevils that produced <10 μg of pheromone were considered “abnormal” and were excluded from the analysis. Pheromone composition for each source of weevils was expressed as the proportion of total pheromone represented by each of the 4 pheromone components. Data for each component were arcsine square root transformed, and both the transformed and nontransformed data were each initially examined for normality and heterogeneity of variance using the Shapiro-Wilk test of PROC UNIVARIATE and residual plots, respectively (SAS 2010; SAS 9.3 Help and Documentation, SAS Institute, Cary, NC). Although the variation among weevils was relatively small for each component, significant deviations from normality were detected. Consequently, additional analyses to compare the pheromone composition of field- and trap-collected weevils were not conducted. Instead, the mean percentage of total pheromone represented by each of the 4 components was expressed as a ratio (I:II:III:IV) for each source of weevils using the output previously generated by PROC UNIVARIATE. The component ratios were informally compared with those previously reported for boll weevils collected from Hildago, Hill, and Limestone counties of TX (Spurgeon and Suh 2007, J. Entomol. Sci. 42: 250 - 260) to determine whether boll weevils in Medina Co., TX, produced a unique blend of pheromone.
The quality of pheromone lures was assessed by quantifying the dose of grandlure in individual lures. Four unopened packages of pheromone lures, each containing 100 lures with a nominal dosage of 10 mg of grandlure, were obtained from the Texas Boll Weevil Eradication Foundation. These packages of lures were stock leftover from the 2008 trapping season in Medina Co., TX. Because these lures were also needed for another experiment, 18 lures each were randomly selected from 3 packages, and 9 lures were selected from the remaining package to examine the pheromone content of lures. Pheromone was extracted from each lure using procedures described by Armstrong et al. (2006; J. Econ. Entomol. 99: 323 - 330), and the amount of pheromone in each lure was quantified by gas chromatography (GC) using the GC methodology and equipment described by Spurgeon and Suh (2009; J. Entomol. Sci. 44: 209 - 221). Because it was apparent that the physical dimensions of lures varied considerably among packages, all the lures within a package were individually weighed to examine the variation in lure weight within and among packages. Data for the grandlure content and lure weight were each subjected to ANOVA using PROC GLM, and differences among packages were separated (α=0.05) using the TUKEY option of the MEANS statement of SAS (2010; SAS 9.3 Help and Documentation, SAS Institute, Cary, NC). Additionally, the relationship between pheromone content and lure weight was examined on the original lures used to assess pheromone quality (PROC REG, SAS 2010; SAS 9.3 Help and Documentation, SAS Institute, Cary, NC).
On average, field-collected and trap-captured boll weevils released 107 and 87 μg of pheromone, respectively, during the 24-h collection period. These amounts are comparable to the quantity of pheromone previously reported for boll weevils that were similarly fed and tested (Spurgeon and Suh 2009; J. Entomol. Sci. 44: 209 - 221). Based on the small standard deviation values associated with the individual components (Table 1), the composition of pheromone was remarkably consistent among weevils. The mean pheromone component ratio of field-collected (44:43:2:11) and trap-captured boll weevils (43:43:3:11) were nearly identical (Table 1), and both were essentially the same as the ratio (45:42:3:10) previously reported for boll weevils collected from Hildago, Hill, and Limestone counties of TX (Spurgeon and Suh 2007; J. Entomol. Sci. 42: 250 - 260). Thus, we found no evidence of boll weevils in Medina Co., TX, producing a unique blend of pheromone.
Suh et al. (2013; J. Entomol. Sci. 75 - 78) evaluated the attraction of boll weevils to an experimental formulation of grandlure (≈44:43:3:10) at 3 sites in Mexico and 2 sites in South TX. Those authors found boll weevils at all the evaluation sites were equally attracted to the experimental and standard formulation (≈32:38:15:15) of grandlure. Similarly, Hardee et al. (1974; Environ. Entomol. 3: 135 - 138) reported boll weevils responded to a range of grandlure component ratios. In light of our findings and the apparent attraction of boll weevils to a broad range of component ratios, the existence of a boll weevil population in Medina Co. that no longer responds to the standard grandlure formulation seems highly unlikely. Instead, our results suggests other factors were likely responsible for the trap detection failures observed in Medina Co., TX.
Analysis of the lure quality data revealed the actual amount of grandlure in lures varied considerably among packages (F=634.18; df=3, 59; P < 0.001). The pheromone lures in each package were supposed to contain 10 mg of grandlure, but the average grandlure content of lures in half of the sampled packages was less than 7.4 mg (Table 2). Although the minimal acceptable release rate for boll weevil pheromone lures has not been established, the rate and total amount of pheromone released from lures is related to the initial pheromone content of lures (C. Suh; unpubl. data). Westbrook et al. (2010; pp. 994 - 998, In Proc. Beltwide Cotton Conf.) also indicated that the daily rate of grandlure released from commercial dispensers held under field conditions declined rapidly from the first to second week of aging. Considering active eradication programs typically use lures with a nominal dosage of 10 mg of grandlure and replace lures in traps every 2 wks, lures containing a substandard amount of grandlure may not release enough pheromone to attract weevils to traps throughout the entire 2-wk lure replacement interval.
Examination of the lure weights confirmed that the size of lures varied significantly among packages (F= 3,312.89; df=3, 391; P < 0.001; Table 2). Also, a significant relationship was detected between the weight and respective pheromone content of lures (F=149.34; df=1, 61; P< 0.001; r2=0.71). In general, larger-sized lures contained more pheromone than smaller lures. Thus, the differences in grandlure content among lures from different packages appeared to be simply related to the manner in which the manufacturer cut and packaged the lures, and not because grandlure was unevenly distributed among lures.
In light of our results and prior findings, the possible existence of boll weevil populations that no longer respond to the standard formulation of grandlure seems highly unlikely. Instead, our findings suggest the use of lures containing a substandard amount of grandlure may be, at least partially, responsible for the trap detection failures observed in Medina Co., TX. As such, eradication program success may be facilitated by a specific standard of lure quality from manufacturers, by routine monitoring of the pheromone content of lures, and by reducing the replacement interval of lures in areas where weevil detection has been problematic.
Contributor Notes