Management of Gypsy Moths Using Sticky Trunk Barriers and Larval Removal

Materials and Methods

The study was conducted in 1993 at forested sites near Harrisonburg, VA, Pomona, MD, and Vienna, MD. The mean bole diameters and mean dripline areas of the experimental trees at each site are given in Table 1. At least 80% of the trees used in the study at each site were white oak (Q. alba). Preseason egg-mass densities varied from about 4,000 per hectare at the Pomona site to over 16,000 per hectare at the Harrisonburg site. Ten groups of four oak trees at both the Pomona and Vienna sites, and six groups of four oak trees at the Harrisonburg site, were selected to serve as replicates. The canopies of all trees in the experiment were touching those of adjacent trees. Each of the four trees within a group received a different treatment. The four treatments were 1) sticky trunk barrier, 2) sticky trunk barrier + burlap skirt, 3) sticky trunk barrier+four burlap skirts, and 4) untreated control. Prior to egg hatch, sticky barriers were applied to tree trunks by wrapping them with duct tape (52 mm wide), then applying by gloved hand a 10-20 mm wide, 1-5 mm thick band of Tanglefoot™ (The Tanglefoot Company, Grand Rapids, Ml) to the center of each band of tape. Small gaps between the tape and the tree were filled with Poly-fil™ polyester fiber (Fairfield Processing Corp., Danbury, CT) to prevent larvae from passing beneath the barrier. The effectiveness of these barriers in preventing gypsy moth larvae from crossing was confirmed by direct observation throughout the study. For treatments 1 and 2, the trunk barrier was applied to the bole at a height of approximately 1.5 m.

For treatment 2, a burlap skirt (40 cm wide) was folded in half lengthwise and wrapped around the bole 1-2 cm above the trunk barrier. For treatment 3, the trunk barrier was applied to the bole at a height of approximately 0.5 m, and four burlap skirts were applied to the bole above the trunk barrier. Each burlap was positioned 1-2 cm above the trunk barrier or burlap directly below it.

On each of 4 (Pomona and Vienna) or 6 (Harrisonburg) dates, all larvae beneath the burlap skirts were counted, removed, and destroyed. After the completion of adult eclosion, the number of pupae and new egg masses was determined for each burlap. On all trees, the number of pupae and egg masses on the lower 2 m of bole was recorded. Defoliation was estimated subjectively with the aid of binoculars in 10% increments on each of the experimental trees by a single experienced worker after larval feeding had ended but before refoliation occurred.

Egg masses on the upper bole and canopy of each experimental tree were counted with the aid of binoculars before egg hatch and again after oviposition. The egg-mass density at each site was estimated from ten 0.01 ha (1/40th acre) eggmass surveys, using the procedures described by Kolodny-Hirsch (3).

Larval population density prior to the onset of pupation was estimated for each tree using the frass drop/frass yield method (5, 6). Frass falling from the canopy was sampled with ten plastic buckets (20.0 cm diameter x 15.5 cm high) positioned in random locations beneath the canopy of each experimental tree. The frass was collected from the buckets approximately 24 h after they were placed beneath the trees. The number of frass pellets falling into the buckets was counted and used to estimate the amount of frass falling per m² of ground surface beneath the canopy. Frass yield (the amount of frass produced per larva during the sampling period) was determined by collecting 50 larvae from each site at the same time that frass drop was sampled, and placing them individually in 177 ml plastic cups with cardboard lids. The cups were each provisioned with one or two oak leaves, and were then set in an area near the experimental trees but protected from the sun. These larvae were removed from the cups at the same time that the frass was collected, so that the sampling duration and temperature conditions experienced by larvae in the cups and in the canopy were similar. The mean density of larvae in the canopy of each tree (number of larvae per m² of ground surface) was estimated using the equation (6):

• where C = 1/(area sampled by each bucket);

• x_d = mean drop (frass/bucket);

• x_y = mean yield (frass/larva).

The number of larvae per tree was estimated by measuring the perimeter of the dripline of each tree, calculating the area contained within the dripline, and multiplying this area by the estimated larval density. Samples were conducted at the Pomona and Harrisonburg sites on June 10, when larvae were predominantly in the fifth instar. Because of the extremely high gypsy moth population density at the Vienna site, the frass sample was conducted on May 25, when larvae were predominantly in the fourth instar. This was necessary to avoid sampling trees that had already received greater than 50% defoliation.

Treatment effects were analyzed by analysis of variance (ANOVA) using the GLM procedure of the SAS statistics package (11). The data were analyzed as a split-plot design, with site as the mainplot factor, using groups within sites as the error term to test the significance of the site effect, and with trunk barrier/burlap treatment as the subplot factor. The site x treatment interaction effect was tested for each variable to determine the appropriateness of combining the three sites into a single analysis. Functional relationships between selected variables were tested using the REG procedure of the SAS statistics package (11). When needed to stabilize the variance, the data were transformed to logarithms prior to analysis.

Results

Gypsy moth population density varied among the three sites (Table 2). Larval density was low to moderate at the Pomona and Harrisonburg sites (94.1 and 83.9 larvae per square meter of canopy, respectively) and high at the Vienna site (215.0 larvae per square meter of canopy) (F= 19.6; df = 2,23; P < 0.0001). Average number of larvae per tree ranged from 4,780 at the Harrisonburg site to 17,238 at the Vienna site (F= 17.4; df = 2,23; P< 0.0001). Defoliation was low at the Harrisonburg and Pomona sites (14.8 and 20.8%, respectively), and high (68.4%) at the Vienna site (F = 93.4; df = 2,23; P< 0.0001). The average number of larvae per tree removed from the burlap skirts above the sticky barrier bands varied from 5.7 at the Vienna site to 35.4 at the Harrisonburg site (F= 34.3; df = 2,23; P=0.03).

For all variables measured, the site x treatment interaction was either non-significant, or if it was significant, then the ordering of the treatment means was consistent for all sites. Therefore, data from the three sites were combined to show the overall effects of the treatments (Table 3). The number of gypsy moth larvae beneath the burlap bands on any given date ranged from 0 to 49 on trees with a single burlap and from 1 to 109 on trees with four burlaps. On average, a total of 9 and 25.8 larvae per tree were removed from trees with single and multiple burlaps, respectively (F= 20.2; df = 1,23; P= 0.0002). A regression analysis indicated that there was no relationship between the density of larvae in the canopy and the number of larvae removed from under burlap (F< 0.01; df = 1,50; P= 0.99; R²=0.01).

The estimated average number of larvae per square meter of canopy differed significantly among the treatments (F = 7.3; df = 3,68; P = 0.0003). Density was highest in the untreated trees (161.5 per m²) and ranged from 103.3 to 126.5 per m² among the trees treated with sticky trunk barriers. Density was significantly lower on trees with the trunk barrier + single burlap treatment. The estimated average number of larvae per tree also differed significantly among treatments (F= 5.0; df = 3,68; P= 0.003), and varied from 4,169 for trees with the trunk barrier + single burlap treatment to 7,943 for untreated trees. Because of the greater variability in these estimates, the trunk barrier alone and the trunk barrier + multiple burlaps treatments were not significantly different from the untreated trees. Defoliation was significantly lower in trees with sticky trunk barriers, but there were no differences among the treatments with trunk barriers (F= 3.8; df = 3,69; P = 0.01).

The number of gypsy moth pupae and egg masses on the lower 2 m of the boles of untreated trees averaged 9.6 and 36.5, respectively (Table 4). On trees with sticky trunk barriers but no burlap skirts, there were an average of 39.2 pupae and 48.8 egg masses. On trees with a trunk barrier and one burlap there were an average of 64.9 pupae and 59.2 egg masses. On trees with a trunk barrier and four burlaps there were an average of 143.7 pupae and 90.5 egg masses. Significant differences occurred among the treatments for both numbers of pupae (F= 55.9; df = 3,60; P< 0.0001) and egg masses below 2 m (F= 46.4; df = 3,56; P< 0.0001).

The total number of postseason egg masses per tree ranged from 145.9 on trees with a trunk barrier and one burlap to 235.9 on untreated trees (Table 4). However, the means for the treatments were not significantly different (F= 1.9; df = 3,60; P= 0.14). The number of egg masses on all trees, regardless of treatment, increased by approximately 5-fold.

The functional relationship between gypsy moth larval density in the canopy of individual trees and percent defoliation is shown in Figure 1. The relationship is non-linear, so a polynomial model was f it to the data (F= 121.2; df = 2,100; P < 0.0001; R² = 0.70).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1.

Relationship between gypsy moth larval density and defoliation in the canopies of individual oak trees. The center line indicates predicted values and the outer lines indicate 95% prediction intervals.

Discussion

Despite the wide range of gypsy moth larval densities encountered in this study, the number of gypsy moth larvae available for removal from beneath burlap skirts was very low at each of the sites. While more frequent and persistent collection of larvae would probably have resulted in higher numbers removed, it is unlikely that enough larvae could be removed in this way to impact population density in the tree canopy. For example, based on the average number of larvae per tree (Table 2), to reduce larval density by 10% would require the removal of an average of 478, 491, or 1,724 larvae per tree from the trees at the Harrisonburg, Pomona, and Vienna sites, respectively. Given the low numbers of larvae found beneath the burlap skirts, the removal of this many larvae would not be possible. Thorpe et al. (14) found a similarly low number of larvae beneath burlap skirts positioned above sticky trunk barriers relative to larval density in the canopy. However, in a 1988 study, Webb et al. (unpublished data) report the removal of over 4,000 larvae from beneath burlap bands positioned above sticky trunk barriers. It is likely that the number of larvae that can be removed in this way is highly variable and unpredictable, but that under certain circumstances it may be possible to remove significant numbers of larvae from the tree canopy. It should be noted that much greater numbers of larvae would be expected to occur beneath burlap skirts in the absence of trunk barriers. However, the source of most of these larvae would be from the ground. Since there is no barrier to invasion of the canopy by additional larvae, the removal of larvae from burlaps in the absence of trunk barriers is unlikely to impact canopy density.

Larval density in the canopy of trees with a sticky trunk barrier plus one burlap skirt was significantly lower than in trees with four or no burlaps. Since the average number of larvae removed was lower on trees with one burlap than on trees with four burlaps, and since the number removed was less than 1 % of the total number of larvae per tree, it is unlikely that the reduced density of larvae in the canopy resulted from larval removal. Also, given the low rate of visitation of burlap by larvae, it is unlikely that increased mortality of larvae beneath the burlap skirt was the cause of the reduction in density. Therefore, it seems probable that this apparent difference between these treatments is a statistical artifact.

The sticky barrier bands reduced larval density in the canopy by an average of 25%. The magnitude of this effect is very similar to that seen in our previous studies (27% [13], 35% [14], 28% [15]). In the Thorpe et al. (15) study the gypsy moth population was too low to cause measurable defoliation. In the Thorpe et al. (14) and Thorpe and Ridgway (13) studies, defoliation was reduced by sticky trunk barriers 25 and 28%, respectively, but neither of these reductions was statistically significant. In the present study, defoliation was reduced 30% by trunk barriers, and this was a statistically significant reduction (Table 3). If the defoliation reduction results of Thorpe and Ridgway (13), Thorpe et al. (15), and this study are combined into a single statistical test (12), the effect is significant (x² = 22.5; df = 6; P= 0.001). Therefore, while the effect of sticky trunk barriers on foliage protection is small, it does appearto be a consistent and experimentally reproducible effect.

While the sticky trunk barriers had no effect on the total numbers of egg masses in treated trees, the treatments did affect the number of pupae and egg masses deposited on the lower 2 m of the bole. This may have practical significance for homeowners who want to remove and destroy gypsy moth pupae and egg masses from the portions of their trees that they can easily reach. The combined use of sticky trunk barriers and burlap skirts provided the highest number of pupae and egg masses on the lower 2 m of bole. The number was increased by using multiple burlaps. However, according to a study employing this technique (2), the removal of pupae and egg masses from the lower boles of trees did not affect subsequent local populations.

The use of sticky trunk barriers has been shown to reduce the density of gypsy moth larvae in the canopy by approximately 25-35%. While the degree of foliage protection achieved through this reduction in larval density is more variable, it appears to average approximately 25-30%. The experiments upon which these values are based were conducted over a range of population densities, with both closed-canopy and isolated oak trees, yet the results were consistent. Therefore, it appears that sticky trunk barriers can provide homeowners with a low-cost method to reduce gypsy moth density in the canopy of protected oak trees and to provide some foliage protection. However, the amount of protection is relatively small, and trunk barriers alone should not be relied upon to adequately protect trees from gypsy moth defoliation.