Associate Professor Queen's University Kingston, Ontario, Canada
Abstract: Background/Questions/Methods
Ixodid ticks (Acari: Ixodidae) are important vectors of zoonotic disease in humans, second only to mosquitos. Three genera comprise most of the medically important species in North America: Ixodes, Dermacentor, and Amblyomma. Ixodes scapularis (blacklegged tick) is of particular concern because it’s rapidly expanding its range along with Borrelia burgdorferi, the spirochete bacteria that causes Lyme disease. Accurately determining the species of a tick often requires the use of high-powered microscopes along with taxonomic expertise that is hard to come by. We developed a landmark based morphometric approach to tick species identification and quantified the morphologic variation between and within species and sex. We surveyed morphometric variation in Ixodid ticks (N = 517) sampled from twenty locations across a Lyme disease hotspot at the leading edge of range expansion in Eastern, Ontario between 2017 and 2021. Sampled ticks were imaged and analyzed using 10 dorsal and 16 ventral landmarks known to vary among and within species. We assessed morphological and diversity to address three questions: (1) Can we use machine learning methods to identify tick species? (2) Which morphological characteristics distinguish different species and sex? and (3) Are there morphologically distinct populations within each species?
Results/Conclusions
Of the 518 ticks sampled, I. scapularis ticks dominated the sample accounting for 71% of all ticks. The remaining ticks (29%) were morphologically identified as D. variabilis. The 10 dorsal and 16 ventral common landmarks were able to distinguish between species and sex with a high degree of accuracy both dorsally and ventrally. Dorsally, PC1 accounted for 85.40% of the morphologic variation and distinguished between sex, while PC2 accounted for 7.73% of the variation and distinguished between species. Ventrally, PC1 accounted for 43.80% of the variation and distinguished between species, while PC2 accounted for 32.89% of the variation and distinguished between sex. Most of the morphometric variation was found in overall width and length, shape of the shield, as well as length and position of feeding structures. Supervised machine learning techniques were able to distinguish geographic location with relatively high accuracies of between 65% and 75%. Knowing which species are present and the morphological variation that defines them is crucial for disease mitigation and management where healthcare practitioners and policy makers can prescribe the best possible treatment or solution for exposure to tick born pathogens.
