Activity of Stem-Injected and Soil Applied Imidacloprid Against Hemlock Woolly Adelgid in the Great Smoky Mountains

INTRODUCTION

Eastern hemlock (Tsuga canadensis [L.] Carrière) is an important component of the riparian ecosystem. Riparian zones are the transitional area between land and water which include soils near streams, rivers, and wetlands. Due to the widespread establishment of hemlock woolly adelgid (Adelges tsugae Annand) (HWA) across the range of eastern hemlock, woodland trees may be infested for extended periods (years), resulting in tree decline. Recently HWA has spread into Michigan and Nova Scotia, outside the contiguous range of hemlock (Limbu et al. 2018). HWA has a bi-voltine, parthenogenetic life cycle, where the sistens generation occurs over the winter months and the progrediens occur in the spring months. HWA undergoes aestivation as sistens in the summer months. In the southern Appalachian, the sistens occurs on hemlock from mid-July through mid-March, and the progrediens are present from mid-March through mid-June (Coots et al. 2013). Symptoms of prolonged infestation include extensive dieback and thinned canopies.

As the HWA infestation increases, hemlock growth is inhibited. McClure (1991) reported that HWA densities > 4/20 mm² of branch completely inhibit new growth in the following year. Further, as populations increase, HWA are forced to feed on older growth. When the resource is depleted, HWA mortality increases. Doccola et al. (2007) reported that HWA densities 2 to 4/linear cm were associated with decreased twig growth. In this study, we evaluated HWA per cm, cm of twig growth, and percentage of tips with growth as metrics for tree health and treatment effectiveness.

Imidacloprid, a systemic neonicotinoid insecticide, is an effective product for managing HWA (Joseph et al. 2011). Previous work demonstrates that imidacloprid is an effective product for managing HWA when used as a soil drench (Cowles et al. 2006; Cowles 2009), but there are per-acre limits to its use for the protection of eastern hemlock. Tree injection is an option to consider when treating hemlock in environmentally sensitive areas (e.g., near streams and ponds) and is exempt from per-acre use restrictions. Trees in the Greenbrier area of the Great Smoky Mountains National Park (GRSM), Gatlinburg, TN (UTM E: 281559, N: 3957949), USA were selected for the study. The Greenbrier site is typical of the riparian zone. In an independently evaluated study at the Cornell Botanic Gardens (CBG), IMA-jet (Arborjet, Inc., Woburn, MA, USA) was successfully injected and effective in control of HWA-infested hemlock in Cascadilla Gorge, a riparian zone (Graziano 2010). However, in contrast to the CBG study [located in USDA hardiness zone 5b (−26 to −23 °C)], the Greenbrier site is in USDA hardiness zone 7a (−18 to −15 °C). Winter low temperatures and rapid temperature change have effects on the survival of HWA (Costa et al. 2004). We’d expect lower winter sistens mortality in the Southeast, and therefore see more direct effects of imidacloprid applications. This study was designed to consider the utility of tree injection in riparian areas where soil drench cannot be used and to evaluate the comparative effectiveness of two methods of tree injection over a 4-year period.

Imidacloprid and its two primary metabolites, 5-hydroxy and olefin, were detected in sap and tissue 15 and 36 months post-treatment, respectively (Coots et al. 2013). In another study, concentrations of imidacloprid and olefin were evaluated following a basal drench of imidacloprid in the GRSM, where the concentrations decreased below the LC₅₀ for HWA in year 5 (Benton et al. 2015). Long-term activity therefore is attributed to imidacloprid and olefin in hemlock. In forested settings, HWA suppression has been observed at 120 ppb (0.12 μg/g) 2 years post-treatment (Cowles et al. 2006). In a 3-year study conducted at the Biltmore Estate (Asheville, NC), imidacloprid residues in tree-injected hemlock were detected at 0.2, 1.8, 2.0, and 1.4 μg/g, at 70, 435, 800, and 1165 days, respectively, compared to soil-injected imidacloprid where residues were 0.14, 0.33, 3.1, and 2.4 for the same periods (Doccola et al. 2012). In that study, trees were in similarly poor condition as in the present study. Normal resumption of growth was not observed in the Biltmore study until 2 years post-treatment, where 72.3% and 41.7% of tip growth was observed in the tree and soil injections, respectively. Although the LC₅₀ was met or exceeded at 70 days, trees recovered slowly. This observation was consistent with those reported by Webb et al. (2003), where they found that trees with significant dieback recovered only slowly, compared to hemlock that were infested, but in better condition. Untreated hemlock remained sparsely foliated and exhibited dieback.

HWA feeds in the xylem ray parenchyma cells (McClure 1991). Parenchyma is vascular tissue that grows as axial and radial tissues as a reticulated network within the xylem. It is a component of the symplast (or living tissue) and plays a role in growth, metabolism, reproduction, storage, and defense (Shigo 1991). The xylem tissue is apoplastic (nonliving) and functions to move liquid and dissolved solutes; movement of liquids is bi-directional and acropetal, with net movement upward. Long-term storage of solutes is associated with the symplast. The initial movement upward into the canopy occurs within the xylem. In deciduous trees, much of the imidacloprid is shed at leaf fall. In a study by Mota-Sanchez et al. (2009) using ¹⁴C-labeled imidacloprid, the researchers found that the imidacloprid accumulated in the foliage during the first season following injection into ash. However, in the second year they found imidacloprid primarily in the trunk tissues. They concluded that the movement of imidacloprid in ash trees was through the xylem. Green ash (Fraxinus pennsylvanica Marsh) and white ash (F. americana L.) shed their leaves annually, and as a consequence, much of the imidacloprid is lost from the tree. Conifers are comprised of uniseriate parenchyma characterized by a single layer (von Arx et al. 2015). Both parenchyma and foliage are the storage tissues in evergreen conifers, including hemlock. Imidacloprid is likely stored within the xylem axial and radial parenchyma. Dilling et al. (2010) found a concentration gradient within the canopy that decreased with tree height. This observation may be in part due to tree condition (tree transpiration) and to formulation used and method of application. For example, transpiration from the tree canopy is negatively affected by dieback, a sparse canopy, or in trees in a suppressed condition. These conditions may also result in imidacloprid binding to cellulose, particularly if formulations are already saturated, such as a 10% imidacloprid used in that study. However, Benton et al. (2015) did not find significant differences in strata, but treatments in that study were applied by soil drench alone. Xylem cellulose is 44% carbon (Heukelekian and Waksman 1925), constituting an extensive adsorptive surface. The carbon adsorption coefficient of imidacloprid is ~350 (Cox et al. 1997), which suggests moderate adhesion to xylem. Pressure bomb extractions such as were conducted by Dilling et al. (2010) extract sap from the xylem tissues; whereas needle and twig tissue milling (Dilling et al. 2010; Benton et al. 2015) extract from the entire vascular tissues (xylem and parenchyma). However, Dilling et al. report that sap from needles and twigs (milled) met or exceeded the LC₅₀ for HWA, whereas pressure bomb extraction (sap from xylem) did not. This differential analysis supports symplastic absorption of imidacloprid. The application of imidacloprid in an evergreen is retained over the leaf’s life span and protected from photolytic degradation (Scholz and Reinhard 1999). In hemlock, needles are retained perennially (i.e., from 3 to 6 years)(Ford and Vose 2006), and may play a role in the extended activity reported.