Insects

Green lacewings, C. rufilabris, were purchased as eggs or 2^nd instar larvae (Evergreen Growers, Portland, OR). Predator larvae were shipped in a hexcel unit, and eggs were either shipped loose in a jar filled with rice hulls, or glued on cardboard tags with approximately167 eggs per card. Predator larvae and eggs were stored at 10 C (50 F) until experimentation. Azalea lace bugs were obtained from a colony maintained by periodically adding wild-caught lace bugs from Benton Co. onto azalea and rhododendron plants enclosed in a vented cage (BugDorm 2400, Megaview Science Co., Ltd., Taiwan). The colony was maintained in a greenhouse at ambient light and temperature.

Study 1: Life stage predated

Feeding assays tested whether green lacewings predate on various life stages of lace bugs. One 2^nd instar predator larva was placed in a 44 mm lockable Petri dish with moist filter paper with lace bugs: six 3^rd instar, seven 4^th instar, five 5^th instar nymphs, or 3 to 5 adults. The 3^rd instar lace bug develops quickly (Braman et al. 1992, Neal and Douglass 1988), which prevented us from using more in assays. Dishes were held at 21 C (69.8 F), 60% RH, and 16L:8D. After 1 day, the number of individuals that had their body contents removed were counted under the microscope. They could be distinguished from plump individuals that died naturally. Each lace bug life stage was replicated 17 to 26 times and tested in July 2015 (Table 1). Because this experiment was observational, only averages and standard errors were calculated.

Table 1 Average number and standard error (SE) of azalea lace bugs at various life stages consumed by one predatory C. rufilabris larvae in a Petri dish for 1 day.

View Fullscreen

Study 2: Predator application

The application efficiency of tapping predator larvae from a hexcel unit was tested on two cultivars of rhododendron plants: ‘Anna Rose Whitney' in 4 L (#1) and 19 L (#5) pots, and ‘Purple Jem' in 4 L (#1) pots. Plants were 30.5 to 45.7 cm (12 to 18 in) and 61 cm (24 in) tall in #1 and #5 pots, respectively. ‘Anna Rose Whitney' has large leaves whereas ‘Purple Jem' has small leaves with dense foliage. Fifteen or 30 larvae were tapped per plant about 15 cm (6 in) above plants in 4 L and 19 L (#1 and #5) pots, respectively. Any predator larvae that that did not fall on foliage were counted on the cardboard collar fitted around the trunk of a plant extending to the pot rim, or the white sheet below. Five replicates were tested for each cultivar and size in August 2015. For analysis, an ANOVA compared the proportion of predator larvae landing on the three treatments (SAS 2016).

The application efficiency of shaking predator eggs was tested on dry or wet rhododendron plants. Six potted rhododendrons from a single cultivar (not recorded) were used (11 L or 3# pots). Three plants were dry, and three were sprayed with water beforehand. Fifty eggs were shaken over each collared plant over a white sheet in August 2015. Eggs that did not land on the foliage were counted on the collar or white sheet. A t-test compared the proportion of eggs landing on dry or wet leaves.

Study 3: Potted rhododendron

The efficacy of lacewing larvae on reducing lace bug populations was tested on rhododendron plants in hoop houses of a commercial nursery in Yamhill Co., Oregon. Rhododendron plants were in 11 L (3#) pots and 61 cm (24 in) tall. On Aug. 21, 2014, ten 2^nd instar predator larvae were transferred to each plant via a camel hair brush (Shrewsbury and Smith-Fiola 2000). No predators were transferred onto control plants. Each treatment was replicated in 6 plants. Both treatments were tested in each cultivar, this included two ‘Lee's Dark Purple', two ‘Cunningham's White', and eight ‘Anna Rose Whitney' plants, all cultivars with large leaves. Control and treated plants were 10 m apart to prevent any cross-movement of predators. To assist the grower with general biological control, 2 g methyl salicylate lures (PredaLure, AgBio, Westminister, Colorado) were hung on plants. Methyl salicylate has attracted various predators and parasitoids into agricultural crops (Lee 2010, Kaplan 2012). While this may have increased predation on lace bugs, this volatile chemical was present in each of the four hoop houses.

Prior to larval release, each plant was visually inspected using 1.75X Opti-visors (BioQuip, Rancho Dominguez, California) to count nymphal and adult lace bugs. Plants were assigned such that treated and control plants started with similar lace bug densities, and their infested leaves were tagged. Because accurate nymphal counting required a microscope at 6 to 15 X and leaf removal affected plant appearance, a limited number of ‘infested' leaves, selected by visual inspection for obvious lace bug damage was collected weekly. On week 1, previously tagged leaves plus two additional infested leaves per plant were collected. On weeks 2, 3, 4, and 6, three infested leaves per plant were collected.

For analyses, lace bug counts from leaves were averaged per leaf per plant; each leaf was the observational unit and each plant was the experimental unit. Response variables were the number of nymphs, adults, or nymphs + adults, and tested in three separate models. To confirm that initial lace bug densities were equivalent among treatments, counts at week 0 were compared between treatments. Next, counts from weeks 1 to 6 were analyzed in a repeated measures model with treatment, week, and treatment by week interaction as fixed effects, each plant as the random subject effect, with autoregressive correlation. Because treatment by week interactions were significant, the two treatments were compared each week by lsmeans comparisons. These and analyses for Studies 4 and 5 used a generalized linear mixed model with the appropriate distribution (normal, lognormal, Poisson), Proc Glimmix in SAS 9.3 (SAS 2016).

Study 4: Landscape rhododendron sole treatment

Here, one treatment type (sole treatment) was applied repeatedly per plant. The efficacy of green lacewing releases and water sprays were tested on landscape rhododendrons in a private garden in Washington Co., Oregon, that did not use insecticides. Plants were approximately 2 m tall and 2 m wide, and received either: 1) 6 predator egg cards (approx. 1000 eggs total), 2) predator larvae tapped from 4 hexcel rows which is 1/5 of the hexcel unit (approx. 88 larvae total), 3) a 5 min pressurized water spray on the underside of leaves, or 4) nothing as the control. For pressurized water sprays, a CO₂ backpack sprayer (R & D Sprayers, Opelousas, Louisiana) sprayed 2 L distilled water per plant at 38 PSI (2.67 kg/cm²) (XR 8002 VS TeeJet Technologies, Glendale Heights, Illinois). Treatments were initiated July 21, 2016 on 5 plants per treatment, and repeated every 2 weeks into the end of August. Selected plants were of different cultivars since it was not possible to block by cultivar; instead, all plants were pre-checked to have an adequate infestation where lace bugs were found on most sampled leaves. Plants were spaced at least 5 m apart. Previously, azalea plants spaced 1.5 m apart showed differences with releases of another lacewing species, C. carnea (Shrewsbury and Smith-Fiola 2000), so 5 m was considered sufficient.

Sampling occurred from week 0 to week 8 on September 8, 2016. Two types of sampling were done. First, weekly samples of 5 to 20 ‘infested' leaves were removed per plant, depending on leaf size and plant canopy density. On weeks 0 and 2, five infested leaves were removed per ‘water' plant before and after water sprays to assess lace bug removal. On other weeks, leaves were removed only after water sprays. In the lab, nymphs and adults on leaves were counted under the microscope while labelling each leaf. Then, frass from the same leaves were washed off with warm water to count eggs at 20X. Lastly, the leaf area was measured with a leaf area meter (LI-3000, LI-COR, Lincoln, Nebraska). Because this sampling focused only on damaged leaves, a second sample involved visually examining 40 ‘random' leaves per plant at the end of the experiment. We counted the number of live adults per leaf. Since dead lace bugs remain on leaves, lace bugs were determined to be alive if they moved when poked. We recorded whether the leaf topside was either ‘visibly damaged', or ‘not damaged'. Visibly damaged leaves resembled leaves with 2 to 50% of the leaf area injured as illustrated by Klingeman et al. (2000, see Fig. 1); this level of damage was recognized by consumer, managers and growers.

View Figure

• Download as Powerpoint
• Download Figure

The first analysis tested whether spraying a leaf with water removed lace bugs and only included data from water-sprayed plants. Counts of nymphs, adults, nymphs + adults on weeks 0 and 2 were compared before and after a water spray with each infested leaf as the experimental unit. Treatment was a fixed effect, and the week sprayed combined with the plant the leaf was taken from was a random effect. The remaining analyses compared all four treatments together. Lace bug counts from ‘infested' leaves were standardized by leaf area (cm²) with each plant as the experimental unit. The second analyses confirmed that initial lace bug densities were similar among treatments. Lace bug counts at week 0, including pre-water-sprayed leaves, were compared among treatments. Third, counts from weeks 1 to 8 were analyzed in a repeated measures model with treatment, week, and treatment by week interaction as fixed effects, plant as the random subject effect, with autoregressive correlation. Fourth, ‘random' leaves at the end of the experiment were compared. Adult counts were averaged per leaf or the proportion of visibly damaged leaves were compared by treatment with each plant as the experimental unit. Proportional data fit a binomial distribution in the model.

Study 5: Landscape rhododendron dual treatment

Here, two control measures were applied together per plant (dual treatment). The efficacy of water sprays followed by predator egg releases was tested on landscape rhododendrons in the same garden. Plants were either sprayed with pressurized water and then received 6 egg cards, or not treated with water or egg cards as the control. Because plants were of different cultivars and some were taller than 2 m, the area under 2 m was treated and flagged for weekly sampling. Treatments were initiated on July 21, 2017 on 5 plants per treatment, and treatments were repeated on August 4, and August 17.

Sampling occurred weekly starting 3 weeks prior to the study, and from weeks 0 to 6, with a break in sampling until week 9 on September 18, 2017. Prior to the experiment, 5 ‘infested' leaves were collected per plant per week, and counted for eggs, nymphs and adults. Starting at week 0, 10 ‘infested' leaves were collected per plant, and counted for eggs, nymphs, and adults, and measured for leaf area in the lab. In Study 5, leaves were collected before water sprays were applied. Collecting dry leaves prevented us from including lace bugs already dead in the field in our counts. On weeks when sprays were not done, weeks 1, 3, 5, 6 and 9, 40 ‘random' leaf samples were also visually checked per plant for live adults, and rated for damage as in Study 4.

For ‘infested' leaf samples, the number of eggs, nymphs, adults, and nymphs + adults were counted on a per leaf basis or divided by leaf area with each plant as the experimental unit. To confirm that initial lace bug densities were not different among treatments, counts per leaf for the three weeks preceding the experiment were compared between treatments, and counts per leaf area at week 0 were compared between treatments. Next, counts from weeks 1 to 9 were analyzed in a repeated measures model with treatment, week, and treatment by week interaction as fixed effects, each plant as the random subject effect, with autoregressive correlation. For ‘random' leaf samples, adult counts per leaf and proportion of visibly damaged leaves were tabulated with plant as the experimental unit. These counts or proportions were analyzed with a similar repeated measures model with appropriate distributions. While treatment by week interactions were not significant, control and water treatments were compared each week by t-test for discussion purposes.

Study 1: Life stage predated

The feeding assay confirmed that 2^nd instar green lacewing larvae, C. rufilabris, consume azalea lace bugs at the 3^rd, 4^th and 5^th instar, and adult stage (Table 1). The predator consumed on average fewer lace bug nymphs when given larger instars, the predator may have gotten more satiated with larger nymphs. Previously, 2^nd instar C. rufilabris had only been confirmed to feed on 4^th and 5^th instar lace bugs in a laboratory functional response study (Stewart et al. 2002). Adult lace bugs do not appear as suitable for 2^nd instar predators since less than one was eaten in a day in our study, and 11 out of the 24 predators tested fed on the adult in a confined arena. In the field, predation may be even lower since adult lace bugs have been observed to drop off the plant to elude predators (Robin Rosetta, personal observations).

Study 2: Predator application

Tapping predatory C. rufilabris larvae from a hexcel unit resulted in 88 to 99% of the larvae landing on the rhododendron plant (Fig. 1). Landing rates were 9 to 10% lower on Anna Rose Whitney in #1 pots than in #5 pots or Purple Jem in #1 pots (ANOVA F_2,14 = 10.6, P = 0.0022). This suggests that with larger-leaved Anna Rose Whitney rhododendrons, predators are more likely to land on larger than smaller potted plants possibly due to a thicker canopy. Among small potted plants, predators are more likely to land on Purple Jem with smaller leaves and denser canopy than on Anna Rose Whitney with large leaves. Shaking loose predator eggs over wet leaves resulted in 79% landing on the plant, which was twice the rate on dry leaves (Fig. 1, t-test df = 4, P = 0.0019). Two hours later after leaves dried, eggs were still observed to stick to the foliage.

Study 3: Potted rhododendron

For simplicity, results are reported as nymphs + adults and only reported separately if outcomes differed. All differences mentioned are statistically significant unless otherwise noted.

To our knowledge, this is the first demonstration that releases of C. rufilabris larvae reduce lace bugs in whole-plant trials, as well as on rhododendrons. Starting densities of lace bugs on assigned plants were similar before treatments were applied at week 0 (Fig. 2; F_1,10 = 0.33, P = 0.581). Later, rhododendrons that had 10 predator larvae added per plant had a lower number of lace bug nymphs + adults on infested leaves compared to controls. A treatment by time interaction occurred, treatment effects were short-term (next paragraph), with a 70 to 86% reduction 1 to 2 weeks after predator release. No differences were observed 3 to 6 weeks after release. Both predator and control treatments had low lace bug counts by week 3 which may have been due to naturally occurring predation, or normal mortality among lace bugs during fall.

View Figure

• Download as Powerpoint
• Download Figure

A short-term impact is expected since released predator larvae would have pupated, stopped feeding, and potentially fly away as an adult. Our released species, C. rufilabris, can develop from 1^st instar to pupa in just 8 to 10 days on high quality foods in the laboratory (Cohen and Smith 1998). Adult C. rufilabris are not predaceous and feed on nectar and pollen. Adults of another species, C. carnea, migrate before they lay eggs (Duelli 1980). Trials with C. carnea in azaleas reduced lace bug densities 4 to 7 days post-release, and the authors commented that season-long control would require multiple releases and integration with other tactics (Shrewsbury and Smith-Fiola 2000, Shrewsbury et al. 2004).

Study 4: Landscape rhododendron sole treatment

First, pre- and post-water spray comparisons confirmed that water sprays dislodged azalea lace bugs. Leaves that were sprayed with water on the underside had 71% fewer lace bug nymphs + adults than unsprayed leaves (F_1,82 = 57.4, P < 0.0001). Unsprayed leaves had on average 5.20 ±1.85 lace bugs per leaf, and sprayed leaves had 1.53 ±0.53 per leaf.

Starting densities of lace bugs appeared similar before treatments were applied at week 0. Specifically, the number of eggs, and nymphs + adults per leaf area on ‘infested' leaves did not differ (eggs F_3,16 = 1.48, P = 0.258; nymphs + adults F_3,16 = 0.51, P = 0.683). From weeks 1 to 8, ‘infested' leaves had a marginally significant difference in egg and adult counts between treatments (P < 0.06), but not nymph counts (Table 2); counts were standardized per leaf area.

Table 2 Statistical results from predatory C. rufilabris releases and water sprays on azalea lace bug counts and damage to rhododendron plants.

View Fullscreen

Lastly, treatments had a noticeable impact on ‘random' leaves sampled at 8 weeks after four bi-weekly applications (Table 2). Plants with predator eggs had 70% fewer adults per leaf than control plants, and plants with water spray had 59% fewer leaves visibly damaged than control plants (Fig. 3). These results suggest that treatments impacted the overall plant based on random leaf samples. However, there were marginal impacts on the infested leaves suggesting that treatments partially remedied the most afflicted plant parts. Since results with water sprays and predator egg releases were encouraging, they were studied in combination the following year.

View Figure

• Download as Powerpoint
• Download Figure

Study 5: Landscape rhododendron dual treatment

When a water + predator egg treatment was applied bi-weekly three times, a consistent decrease in azalea lace bug counts and damage was observed from both ‘infested' and ‘random' sampled leaves starting 1 week after the first application. First, starting lace bug densities did not differ for 3 weeks prior to the experiment with counts standardized per leaf (eggs F_1,28 = 0.723, P = 0.402, nymphs + adults F_1,27 = 0.252, P = 0.620), nor at week 0 with lace bug counts standardized per leaf area (Fig. 4a,b; eggs F_1,8 = 0.36, P = 0.566, nymphs + adults F_1,8 = 0.001, P = 0.993). Later among ‘infested' leaves, fewer lace bug eggs and adults were found in treated plants than controls (Table 2, Fig. 4a,b). Among ‘random' leaves, fewer adults and visibly damaged leaves were found in treated plants than controls (Table 2, Fig. 4c,d). Treated plants had 44 to 90% fewer lace bugs than control plants at various weeks. A 44 to 68% difference still appeared by week 9, which was 5 weeks after the last application.

View Figure

• Download as Powerpoint
• Download Figure

Our studies demonstrate that C. rufilabris can suppress azalea lace bugs in rhododendrons, and water sprays can dislodge nymphal and adult lace bugs. Ideally, treatments should be targeted when lace bug nymphs have recently hatched in spring and mid-summer. This requires detailed monitoring to detect the presence of small nymphs. Treatments may be most effective in spring if overwintering lace bug eggs hatch synchronously into nymphs, and predators suppress populations at the start of the growing season. In our final trial, the combined water + predator egg treatment showed a reduction 5 weeks after the last treatment. Also, treated plants had consistently lower lace bug counts on both ‘infested' and ‘random' leaves, and lower visible leaf damage each week. These control methods might be appropriate when there is some tolerance for lace bugs, such as in landscapes. When there is little tolerance for pest damage, these control measures would need to be integrated with other measures to provide season-long control, and water sprays might be set-up in the irrigation system if possible to dislodge lace bugs without labor input.