Background/Question/Methods   In the northeastern US, eastern hemlocks (Tsuga canadensis) are severely threatened by the invasive hemlock woolly adelgid Adelges tsugae. Although other hemlock species are resistant to the adelgid, researchers have as yet found no evidence for adelgid resistance in eastern hemlock trees. During a series of landscape-level surveys of hemlock forest health, a few surviving eastern hemlock trees were discovered in otherwise adelgid-devastated forests. We took branch cuttings from these trees, along with cuttings from adelgid-susceptible eastern hemlocks and cuttings from western hemlocks (Tsuga heterophylla, an adelgid-resistant hemlock species), and propagated them under controlled conditions to induce root formation. Successfully-growing rooted cuttings were transferred to unfertilized forest soil and grown for a total of two years. In spring 2009, we used HWA-infested branches to inoculate 37 putatively-resistant cuttings from eight parent trees (representing three sites and three states), 28 control cuttings from six parent trees (representing three sites and two states), and five cuttings from a single western hemlock parent. We recorded adelgid density (#/ mm hemlock growth) in July and October (assessing sistens crawlers and adults, respectively), and used repeated-measures ANOVA to test for both within- and among-group differences in adelgid density and survival over time.

Results/Conclusions   Adelgid density differed among groups and across time, and there was a significant group*time interaction. Although crawler densities in the resistant and control groups did not differ, settled adult densities were significantly lower in the resistant group than in the control. The steep decline in adelgid density between the crawler and adult phases in both the resistant and control groups is typical of this species, since even under ideal conditions only a small fraction of crawlers survive to adulthood. There were no within-group differences among sites in either the resistant or control groups. Western hemlock cuttings had very low rates of both adelgid settlement and survival; in contrast, cuttings from putatively-resistant trees had high adelgid settlement but very poor survival. This suggests that although resistant trees are poor hosts, adelgids cannot differentiate between them and control trees during the settlement process. Work is currently underway to assess the mechanistic basis of plant resistance in this interaction.