In Ky. this northeastern shrub is restricted to cool calcareous slopes in a few localities, perhaps only in the Cliff Section of the Appalachian Plateaus (especially to the north). It is locally dominant in the ground vegetation on lower north-facing slopes where cool air drains from limestone caves in deep narrow ravines (Campbell et al. 1989, 1991); these microsites are difficult to model with current GIS tools (Pace 2015). Reports from EDMO (Hussey 1876) and OWSL (Gonsoulin 1975) may be supported by colls., but remain unverified. The record from Louisville (JEFF) area by McMurtrie (1819) remains highly dubious. Disjunct plants also occur in c. Ind. (D, K). As in Thuja, the restriction of Taxus to remote protected ravines across the southern part of its range may have been enhanced by winter-browsing of deer and other herbivores, despite its potential for toxicity (Allison 1990; Fargione et al. 1991). As FNA 2 states: "The foliage, bark, and seeds---but not the fleshy red aril---of most Taxus species are toxic due to the presence of taxine... [but] the fresh foliage of T . canadensis is browsed by deer," Pace (2015) found similar taxane composition, including taxine and related alkaloids, in canadensis, cuspidata and baccata; paclitaxel (used as "taxol" for anticancer uses) is unknown in canadensis, but its chemical precursors are present. Toxic effects can be rapid, especially on mammalian hearts, and can be transmitted through pollen, fallen leaves, and wood dust. T. canadensis is generally monoecious, but with much variation in the proportion of male versus female cones (Allison 1991, 1993).