Problem Context

Hemlock woolly adelgid (Adelges tsugae) is an invasive insect pest that exclusively parasitizes hemlock and spruce trees in its native range in Asia (Havill et al., 2006). Hemlock woolly adelgid (HWA) was first discovered in North America in 1916 in British Columbia, where it had likely been present for thousands of years (Havill et al., 2006). It spread along the native range of Western hemlock (Tsuga heterophylla) from Alaska to California but did not cause significant damage to Western hemlock (Emilson et al., 2018). However, a distinct population of HWA was discovered in the Eastern United States in 1951 in Virginia, which was found to be a recent introduction sourcing from Japan (Havill et al., 2006). HWA infested native Eastern hemlock (Tsuga canadensis) and quickly spread across Eastern North America to 21 different states and Nova Scotia (Emilson et al., 2018).

Eastern hemlocks are very long-living, shade-tolerant, coniferous trees with a unique characteristic where mature trees keep the needles on their lower branches. Many species, such as the black-throated green warbler (Setophaga virens) strongly rely on hemlock for protection and nourishment (Tingley et al., 2003), and some species of fish are much more likely to be found in streams shaded by Eastern hemlock (Ross et al., 2003). Eastern hemlock stands provide a unique habitat for many species and are considered a keystone species because of this (Brantley et al., 2016).

Unfortunately, Hemlock Woolly Adelgid was found to be very destructive to Eastern hemlock, causing up to 100% mortality in hemlock stands and preventing them from regenerating (Ford et al., 2011). HWA is a tiny insect that feeds in great numbers by inserting its stylet, an elongated mouth, into the base of hemlock needles. Hemlock trees can die as soon as four years after infestation but may survive up to 10 years in more northern ranges. Currently, there are no known predators of HWA in its Eastern North American range (Evans and Gregoire, 2006), and even healthy Eastern hemlock have little to no natural defense against it (Ford et al. 2011).

Currently, HWA is found within all five states that border eastern Canada (Maine, New Hampshire, Vermont, New York, and Michigan) and 5 counties in Nova Scotia. In 2017, HWA was detected in Erie, a New York county that borders Ontario near Fort Erie (USDA FS: HWA Predator Database, 2018). HWA infestations have been found and destroyed in Ontario on two occasions, and a management plan for HWA has been created (Emilson et al., 2018). However, this management plan could benefit from a GIS-based identification of the most at risk hemlock stands to prioritize for monitoring.

HWA is thought to spread in the United States naturally using wind and birds as vectors at a rate of 12.5 km/yr, however this rate is variable depending on temperature (Evans and Gregoire, 2006) and hemlock density (Morin et al., 2009) and so will likely spread at a slower rate as infestation cuts into Canada. Many studies have been done on how temperature limits or slows the spread of HWA. 90% or higher HWA mortality is thought to be enough to stop the spread of HWA; this mortality can be predicted by average winter temperatures (Paradis et al., 2007), lowest minimum temperature that winter (McAvoy et al., 2017; Paradis et al., 2007), mean temperature of the three days following that minimum (McAvoy et al., 2017), the number of days where the temperature is below -1°C (McAvoy et al., 2017), and various other measures of temperature (Paradis et al., 2007). HWA may currently be limited by winter temperatures, however, climate change will likely result in a much greater potential range in the future as average temperatures increase. Furthermore, HWA is thought to be evolving better resistance to cold (Hart, 2008). Exceedingly warm summer temperatures are thought to limit the spread of HWA as well (Mech et al., 2018), however, this is at the southern extent of HWA’s range and is unlikely to play a major role at the northern end. As temperature varies across space, using GIS to analyze how HWA could spread and where its spread is limited to would be useful for directing management efforts.

Although much of the spread of HWA has been modeled using temperature and natural spread rates, there have been many occasions of HWA jumping huge distances through human-mediated vectors, such as during the transport of firewood or hemlock for ornamental purposes that are often not accounted for. For example, one of the two occurrences of HWA in Ontario was thought to happen due to the transportation of infested trees into a hemlock nursery (Emilson et al., 2018) and much of the initial dispersal of HWA in Virginia was thought to have been from the accidental transportation of infested hemlock (Morin et al., 2009). A GIS approach that involves road networks would be perfect to identify potentially vulnerable areas due to this type of spread.

The most obvious limiting factor of the spread of HWA is that it cannot spread outside the range of Eastern hemlock, therefore an accurate catalog of hemlock distribution would prove useful by representing the absolute limit of HWA spread. Furthermore, fragmented hemlock populations exist that have been isolated from the core population either naturally or by development; identifying these could assist in conservation as they are not very likely to become infested due to their disjunction (Hart, 2008). Unfortunately, a complete Canadian hemlock inventory has not yet been completed (Emilson et al. 2018). Given this limitation, we expect to contribute by showing where hemlock stands will be most vulnerable whether or not they are actually present there.

Purpose of Research 

To determine the vulnerability of Eastern hemlock groves in Eastern Canada to the spread of Hemlock Woolly Adelgid from the United States of America into Canada based on changing climate using GIS analysis.

Research Objectives

Objective 1: To identify factors and variables related to the spread of Hemlock Woolly Adelgid

Objective 2: To develop a GIS-based model evaluating the current spread of Hemlock Woolly Adelgid and Eastern Hemlock

Objective 3: To apply the model and predict the future spread of Hemlock Woolly Adelgid

Objective 4: To compare the worst case and best case scenarios for the spread of Hemlock Woolly Adelgid and vulnerability of Eastern Hemlock

Objective 5: To identify the strengths and weaknesses of the model and the strategy/ processes used for the purpose/user of the model