Organic Control of Pear Psylla in Pear with Trunk Injection

Simple Summary Organic pear production is challenged, in part, by short lived effects of biopesticides when applied as foliar sprays. Trunk injection may enhance their performance by delivering the biopesticides directly to the vascular system of the tree, right where pear psylla feed. The objective of this study is to compare trunk injections to foliar sprayed applications of two insecticides, azadirachtin and abamectin, on their ability to control pear psylla in pear trees. The azadirachtin and abamectin trunk injected treatments performed equally or better than two foliar applications in the control of the pear psylla. The trunk injected trees from the first season provided a moderate level of control into the second season, one year after the injections. This study suggests that trunk injection is a superior delivery system for biopesticides used in organic pear production. Abstract Organic production of pears is challenging in part because OMRI (Organic Materials Review Institute) approved biopesticides are short lived when applied as foliar sprays. Trunk injection is an alternative method of insecticide delivery that may enhance the performance of biopesticides for control of pear psylla. The objective of this study is to compare the efficacy of azadirachtin and abamectin in the control of pear psylla using two different application methods, airblast sprayer and trunk injection. Trunk injections of azadirachtin and abamectin were compared to airblast applications of equal labeled rates on 33-year-old Bartlett Pear trees (Pyrus communis L., var “Bartlett”). The azadirachtin and abamectin trunk injected treatments performed equally or better than the two airblast applications in the control of the pear psylla. The trunk injected trees from the first season provided a moderate level of control into the second season, one year after the injections. This study suggests that trunk injection is a superior delivery system for biopesticides used in organic pear production.


Introduction
Pear Psylla, Cacopsylla pyricola (Forster), is the number one insect pest to the pear industry. In fact, more than half of the money spent to control insect pests in commercial pear orchards are directed specifically at controlling pear psylla [1]. Psylla nymphs feed on sap from the trees and produce honeydew, which drips down onto the leaves and fruit. A black sooty mold grows in the honeydew, and the black color on the pears downgrades the fruit [2]. In cases where trees have been chronically heavily infested, they become stunted, reduce fruit production, and lose their leaves [3]. Pear psylla also transmit a disease that causes pear decline, which renders the tree unable to move nutrients to the roots and can cause the death of the tree [4].

Trunk Injection
Treatment injection applications were made 7 days after petal fall (23 May 2017). Injections were performed using an Arborjet Tree IV with 4 portals equally spaced along the circumference of each trunk. The injection equipment included the Arborjet Tree IV TM kit (Tree IV, #4 arbor plugs, and plug tapper) (Arborjet Cleanjet TM , Arborjet, Inc., Woodburn, MA, USA), hammer, cordless drill, and a 0.95 cm wood drill bit. First, 4 holes were drilled into the pear trunk approximately 5 cm deep, 90 • horizontal from the trunk, and 30 cm above the ground, spaced as equally as possible while strategically placing under the main scaffold branches of the tree to distribute maximum compound volume throughout the canopy. Next, the plugs were tapped into place deep enough thus that the outside rim of the plug was just below the bark. At this stage, the tree was ready for injection.
Before each injection, the Tree IV was cleaned with a sanitizing solution (Arborjet Cleanjet TM , Arborjet, Inc., Woodburn, MA, USA) and water to rinse out residues. The insecticide was measured out and diluted into distilled water, thus that the final volume was 500 mL. The compound was then poured into the Arborjet injector holding tank. The needles were inserted into the plugs, and the compound was injected via a hand-operated pressurized pump (Arborjet, Inc., Woodburn, MA, USA, part number 070-2025) into the tree.

Foliar Application
Foliar applications were performed using an FMC 1029 airblast sprayer (Jonesboro, AK, USA) in 935 L/ha of water (100 GPA, gallons per acre). The 1st applications were made on the 23 May 2017 and 30 May 2018. A 2nd application of the foliar treatments was made on 23 June 2017 and 17 July 2018 after psylla nymphs reached an action threshold of 1 nymph per 3 leaves [1]. Individual trees were sprayed by using the nozzles on one side of the airblast sprayer only to apply the treatment on one side of the tree at a time. Foliage on both sides of the tree were thoroughly covered by the output of the sprayer.

Field Evaluations
To evaluate the C. pyricola numbers, field evaluations were conducted every 2 weeks in the 2017 and 2018 seasons. Samples of 50 leaves were taken (randomly throughout the entire tree-high, low, shielded, exposed, and all 4 quadrant areas) for each replication. Pear psylla eggs and nymphs were counted using a stereomicroscope. In 2017, C. pyricola nymph and egg evaluations were made on the 31 May, 12 June, 19 June, 28 June, 14 July, and 21 July. In 2018, C. pyricola nymph and egg evaluations were made on the 18 June, 16 July, 25 July, and 8 August. All the field evaluations on the 2017 injection trees continued into the 2018 season with field evaluation dates occurring June through mid-August.
At the end of the season, on 17 August 2017, a 2 min count for sooty mold occurred. Each tree was surveyed for 2 min and leaves with sooty mold were counted. This evaluation was used to quantify the negative impact of psylla honeydew on leaves at the end of the season.

Residue Sample Collection and Preparation
Pear leaf and fruit samples were taken for residue analysis. Leaf samples were collected on 1, 7, 14, 28, 56, and 84 days after treatment (DAT). Fruit samples were taken on 7 and 84 days after treatment. For the leaf samples, 40 leaves were randomly sampled from each replication (approximately 25 g). For the fruit samples, 5 pears were randomly sampled from each replicate tree. Each pear was diced into half-inch squares and homogenized in a bowl, and 25 g were taken from this sample. Pears taken at 7 DAT were not as large as the 84 DAT fruit, therefore, after dicing and homogenizing the 7 DAT samples, most of the homogenized sample was used to make 25 g. Samples were weighed and held in 50 mL of dichloromethane (DCM) until processing. Samples were mixed with 4 g magnesium sulfate and 1 g sodium chloride and allowed to sit for 48 h in a 4 • C walk-in cooler. The DCM was then filtered through a funnel lined with filter paper and 10 g of sodium sulfate to remove water and allowed to evaporate under a hood for 4-12 h. We added 2 mL acetonitrile to the evaporated jars and swirled for 90 s to ensure maximum uptake of the dried pesticide residue. The acetonitrile solution was then analyzed on a HPLC (High Performance Liquid Chromatography) utilizing a previously reported method [22,23].

Residue Sample Analysis
The residue levels were quantified using a waters 2695 separator module HPLC equipped with a Waters MicroMass ZQ mass spectrometer detector (Waters, Milford, MA, USA), and a C18 reversed-phase column 50 by 3.0 mm bore, 3.5 µm particle size (Waters, Milford, MA, USA).
The mobile phase, solvent A, was water with 0.1% formic acid, and solvent B was acetonitrile with 0.1% formic acid and was initially held at 80% solvent A and 20% solvent B and followed by a gradient. The column temperature was 40 • C.
Monitored ions for abamectin were 158.3 and 886.7 m/z (Da). The HPLC level of quantification was 0.0023 mg/kg −1 of a.i., and level of detection was 0.001 mg/kg −1 . By using the above described extraction method, the mean parent compound recovery from 4 pear leaf samples (each 100 g) treated only with standard abamectin solution (0.046 mg/kg −1 ), then agitated and left to dry, was 73%. The results have not been corrected for abamectin recovery.
Monitored ions for azadirachtin were 685.4 and 703.4 m/z (Da). The HPLC level of quantification was 0.015 mg/kg −1 of a.i., and level of detection was 0.005 mg/kg −1 . By using the above described extraction method, the mean parent compound recovery from four pear leaf samples (each 100 g) treated only with standard azadirachtin solution (0.030 mg/kg −1 ), then agitated and left to dry, was 83%. The results have not been corrected for azadirachtin recovery.
The normality assumption was assessed by checking normal probability plots and histograms of residuals. The equal variance assumption was assessed by checking plots of residual v. predicted values, side by side box plots, and Levene's test.
The data indicate that as time goes on, the variability in the variances increases, making it necessary to fit a variance-covariance structure with unequal variances. When treatments were significant, all pairwise comparisons among the treatment means were analyzed. PROC GLIMMIX SAS 9.4 (SAS Institute, Cary, NC, USA, 2013) was run to generate a plot of pear psylla least-squares means for treatment by evaluation day sliced by treatment and adjusted for Tukey-Kramer honestly significant difference (p ≤ 0.05).

2017 Season
The overall treatment effect for mean C. pyricola eggs was significant (F = 16.01, Num df = 4, Den df = 23.4, p < 0.0001). Differences of treatment by evaluation day LS (Least Squares) means sliced by treatment indicated a total of 5 significant differences between treatments (Figures 1A and 2A). The overall treatment effect for mean C. pyricola nymphs was significant (F = 39.46, Num df = 4, Den df = 22, p < 0.0001). Differences of treatment by evaluation day LS means sliced by treatment indicated a total of 11 significant differences between treatments (Figures 1B and 2B).  The sooty mold evaluation showed significantly fewer leaves infected on treated trees compared to untreated trees (F = 17.92, NumDF = 4, DenDF = 12, p < 0.0001). All treatments averaged less than 16 leaves with sooty mold, significantly fewer than those in the untreated check (Table 2).   The overall treatment effect for mean C. pyricola nymphs was not significant for azadirachtin (F = 3.12, Num df = 2, Den df = 9.71, p = 0.0897) ( Figure 3B).
The overall treatment effect for mean C. pyricola eggs was significant for abamectin (F = 6.57, Num df = 2, Den df = 12.7, p = 0.0109) ( Figure 4A). Differences of treatment by evaluation day LS means sliced by treatment indicated a total of 2 significant differences between treatments. Abamectin treatments significantly reduced C. pyricola nymphs (F = 13.8, Num df = 2, Den df = 11.2, p = 0.002) ( Figure 4B). Differences of treatment by evaluation day LS means sliced by treatment indicated a total of 2 significant differences between treatments.
Abamectin trunk injection and airblast treatments reduced the number of eggs on 18 June 2018 (p < 0.05) relative to the untreated control but there was no effect of treatment on the number of eggs the rest of the season ( Figure 4A). Without treatment, the number of nymphs peaked in mid-summer, but populations were reduced significantly (p < 0.0001) by both insecticide treatments on 25 July 2018 ( Figure 4B).

Residue Profiling
Azadirachtin-treated fruit samples had only one detection, which was in the trunk injection treatment for samples taken 7 days after application (Table 3). No residues were detected in abamectin treated fruit samples (Table 4). No residues were detected above the MRL(Maximum Residue Level) for fruit (abamectin MRL = 0.02 ppm, azadirachtin was exempt from the tolerance requirement) [24]. Table 3. Mean azadirachtin residue on fruit taken 7 days and 84 days after the first treatment (DAT). Residue (ppm) reported above the level of detection (0.005 ppm), or below the level of detection (nd).

Treatment/Application Method 7 DAT (Days after First Treatment) 84 DAT
Untreated nd nd Azadirachtin/Trunk Injection 1.507 nd Azadirachtin/Airblast nd nd Leaf sample residues for azadirachtin were highest following applications and steadily decreased throughout the season ( Figure 5A). All azadirachtin residues were below 30 ppm, and no residue was detected beyond 28 days after treatment. Leaf sample residues for abamectin were relatively low, with all residue detections below 0.25 ppm, and most detections below 0.1ppm ( Figure 5B).

Discussion
This study contributes new information on how trunk injection of biopesticides may enhance the control of C. pyricola in pear production. A single injection of azadirachtin provided season-long control, and two seasons control following a single injection of abamectin. Most importantly, one trunk injected application of either product resulted in the same or better control than four foliar sprays over two seasons.
Azadirachtin is an antifeedant, repellant, and insect growth regulator (IGR) [25][26][27]. Azadirachtin affects the morphogenesis, ovarian development, fecundity, egg viability, and molting of psylla through the endocrine system [25]. The injection data reflected the IGR effect with significantly lower nymphs starting just a week after injection. The antifeedant and repellant properties of azadirachtin likely played a role in lowering psylla numbers as well.
Leaf residue levels dropped rapidly after the first azadirachtin foliar spray, whereas azadirachtin showed a more uniform pattern of persistence in the trunk injected trees. This makes sense as the tree can store azadirachtin within leaf tissue as a metabolite without changing its biological effect [28,29]. Azadirachtin on the surface of the plant, however, breaks down quickly under ultraviolet light [30,31]. The biological activity is significantly reduced around 200 h of exposure to UV (Ultraviolet) light [30]. The antifeedant potency is rapidly decreased with exposure to sunlight [32]. The half-life of azadirachtin was found to be less than an hour [31]. Sun exposure degradation may explain why the residues for the airblast application declined rapidly, and why airblast applied azadirachtin was not effective for as long as trunk injected azadirachtin.
Abamectin is a neurotoxin, permanently opening the glutamate-gated chloride channels and inhibiting the nerve and muscle cell communication. When ingested, the psylla is affected by uncoordinated movement, paralysis, starvation, and ultimately death [33]. The neurotoxicity effects the nymph stage most, stopping life before reproduction occurs, and our data reflect this. Egg numbers stayed quite low, never ramping up to a peak in egg production, and always staying flat line at a low level. Nymph numbers were consistently below the action threshold (0.3 nymphs/leaf) for the abamectin trunk injected treatments. While airblast treatments were not statistically different from the trunk injected treatments, they were consistently above the nymph action threshold, meaning that farmers would take action at that point to protect their crops by making another insecticide application. If a second application can be avoided, time and money can be saved, and less pesticide introduced into the environment.
Abamectin rapidly degrades when exposed to light [33,34]. Abamectin products used in tree fruit production are formulated for foliar application, for which horticultural oils are recommended to move the active ingredient into the leaf tissue. Foliar application of abamectin with horticulture oil is generally expected to control pear psylla for half the season. Trunk injected abamectin provides season long control and may be further protected from degradation within the tree canopy. This likely explains why the injected treatment was superior in its ability to control pear psylla. This is the first modern study in the United States demonstrating the potential for using trunk injection of biopesticides in pear trees. Our study showed that injected insecticides achieved a high level of control over C. pyricola for two seasons. Likewise, this has been documented in apple trees for similar phloem-feeding insects such as potato leafhopper and rosy apple aphid [19,[35][36][37]. Residues in pear fruit were extremely low and often zero, well below the MRL allowed by the US EPA, similar to other studies in apple [37].

Conclusions
In conclusion, our study shows that the trunk injection of insecticides to control C. pyricola in pear trees has many promising aspects for future pear production. For organic production in particular, which is reliant on biopesticides, trunk injection can enhance the control of C. pyricola. First, one injection provides a high level of control of C. pyricola for two seasons using 75% less insecticide than what is needed with airblast application. Second, the insecticide is delivered directly to the feeding psylla through the sap of the tree, therefore, it is not lost to drift, runoff, or subjected to photodegradation. This reduces production costs for farmers and saves the environment from unnecessary non-target exposure. Thirdly, detections of pesticide residues for abamectin and azadirachtin were zero for fruit at harvest. This reduces exposure and dietary risks to farmworkers and consumers.
Trunk injection works well in reducing C. pyricola in pear. Future efforts to make the injection process more efficient will lead to economic feasibility for farmers to adopt. Currently, the trunk injection systems require 2 to 5 min per tree to deliver crop protection materials. Economic viability may depend on reducing the time and labor needed to cover acres of orchard crops.
Author Contributions: Conceptualization, methodology, and resources, J.C.W., C.V., and C.E.W.; funding acquisition, project administration, and supervision, J.C.W.; visualization C.E.W.; writing-original draft, J.C.W. and C.E.W.; writing-review and editing, J.C.W., C.V., and C.E.W. All authors have read and agreed to the published version of the manuscript.
Funding: This research received no external funding.