Efficacy of Soil and Trunk Applied Systemic Insecticides for Three Years After Application for Control of Emerald Ash Borer (Coleoptera: Buprestidae) on Green Ash*

Abstract

We evaluated the efficacy of two formulations of emamectin ben-zoate applied as trunk injections, and imidacloprid applied as a soil drench for control of emerald ash borer (EAB) (Agrilus plani-pennis Fairmaire) for 3 years following application. A 0.2-ha plan-tation of ash (Fraxinus) species was established in May 2003 at the Michigan State University's Tollgate Education Center in Novi, MI. On 15 May 2012, green ash (Fraxinus pennsylvanica) with trunk diameters averaging 15.2 cm at breast height were treated with 4% emamectin benzoate (TREE-ageV R ; Arborjet, Woburn, MA), a novel formulation of 4% emamectin benzoate (ARBORMECTIN TM ; Rotam Agrochem International Co. Ltd., Hong Kong, China), or imidacloprid (XytectV R 75 WSP; Rainbow Treecare Scientific Advancements, Minnetonka, MN). Emamectin benzoate formulations were applied as trunk injections at a rate of 10 ml/in. trunk diameter using the Arborjet Tree I.V system with #4 ArborplugsV R (Arborjet, Woburn, MA). Imidacloprid was ap-plied as a soil drench (Table 1) at a rate of 1.4 g active ingredient per inch trunk diameter. The experiment was designed as a ran-domized block with five replicate trees per insecticide treatment and six replicates for the untreated control. Treatments were evaluated on 26 Jun 2013 and 15 Aug 2014 by visually rating the degree of canopy dieback on a 1–5 photo-graphic rating scale with 1 showing no signs of decline and 5 be-ing dead (Klooster et al. 2014, Biol. Invas. 16:859–873). Trees were assessed independently by two evaluators and the ratings av-eraged to generate a single value for each tree. Number of adult exit holes were recorded on 4 Jun 2013 and 15 Aug 2014. Trees were harvested on 15 Aug 2014, and a trunk section 183-cm in length from the ground was subsequently debarked with a draw knife, and the number of living larvae were recorded. Data were analyzed via ANOVA; Number of larvae per tree were trans-formed [ln(X þ 1)] to increase homogeneity of variances. Number of larvae was the only variable for which the F-test from ANOVA was significant (P ¼ 0.002) and means were separated using the Protected LSD test (P 0.05). Treatment effects on canopy decline, number of exit holes and larvae per tree are presented in Table 1. All trees expressed very low or no signs or symptoms of infestation in 2012 and 2013. Hence, no conclusions could be reached about product ef-ficacy during the first (2012) or second (2013) year following treatment. Quantifying the number of larvae per tree in Aug 2014 (infestation resulting from oviposition in 2014) provided an indication of efficacy in the third year following treatment. Number of larvae in trees treated with the two emamectin ben-zoate formulations did not differ in 2014, and were very low rel-ative to untreated trees, which indicates that these treatments provided 3 years of control following a single application. However, due to high variation in the control, the difference be-tween number of larvae in trees treated with Arbormectin and the untreated control was not significant. Trees treated with imi-dacloprid in 2012 had high numbers of larvae in 2014, indicat-ing that imidacloprid did not provide control 3 years following application. There was no evidence of phytotoxicity for any treatment on any evaluation date.