2020 ESA Annual Meeting (August 3 - 6)

PS 64 Abstract - Induction of extrafloral nectaries in tree of heaven (Ailanthus altissma) due to spotted lanternfly (Lycorma delicatule) attack in Pennsylvania

Tanner J. Steltz, Kutztown University, Kutztown, PA and Christopher W. Habeck, Environmental Science/Biology, Kutztown University, Kutztown, PA
Background/Question/Methods

The spotted lanternfly (Lycorma delicatula; hereafter SLF) is an invasive phloem-feeding planthopper in eastern North America. SLF is a major pest to many trees and agricultural crops, particularly fruit trees and vines. The tree-of-heaven (Ailanthus altissima; hereafter TOH) is also invasive in North America and is a phloem resource for SLF. TOH has extrafloral nectaries (EFN) at the base of its leaflets, presumably as a trophic reward for ant-assisted defense against attack. We predicted an induced increase in the number of EFN per leaflet with increased SLF density on TOH. To test our hypothesis, we collected five leaves each from 8-10 trees from five sites, three sites had SLF present and two sites did not. For each tree, we counted all SLF on the tree and recoded EFN abundance per leaflet for each leaf. We used ANOVA to understand the influence of SLF presence or absence on EFN production. For the three sites where SLF were present, we used multiple regression to understand the relationship of EFN production to SLF abundance.

Results/Conclusions

We found no difference in ELF production in the three sites with SLF compared to the two sites where SLF was absent (F1,3 = 1.969, p = 0.255). For the three sites with SLF, however, there was a weak positive relationship of EFN production with SLF abundance (β = 0.019, R2 = 0.13, p = 0.045). If we ignore site in our model, we find a strong difference in EFN with and without SLF (F1,46 = 15.71, p < 0.001). This suggests to us that given more replicate sites, we may find that SLF are in fact inducing defense by TOH, manifested as EFN production. We hope to increase the power of our models in the future by increasing our site sample size and the number of trees within sites that we measure. In addition, we recognize that multiple environmental factors might be influencing site level responses. We hope to include more variables to explain this potential variance.