Results/Conclusions Twelve terpenoids in addition to DME were identified with each phytochemical class exhibiting a high degree of variability. Nonetheless, diterpene acids dominated the phytochemical profile of most genotypes (averaging 0.75% of fresh mass) with kolavenic acid being the single greatest component. Monoterpene, sesquiterpene and acidic diterpene concentrations were all significantly greater (P< .001) in the elongating rhizomes compared to established rhizomes. Terpene differences between elongating and established rhizomes are in general agreement with Optimal Defense Theory which predicts increased defense of high-value tissues. Approximately 26% of the established rhizomes (> 1 yr) processed for chemical analysis exhibited damage by cortex mining Eucosma sp. (Tortricidae) larvae; however, little herbivore damage to the softer elongating rhizomes was observed. Conversely, mean DME level (± 1SE) of the elongating rhizomes (69.0 ± 46.3 μg g-1 fresh mass) was lower (P< .001) than that of the established rhizomes (370.4 ± 85.6). Furthermore, approximately half of the elongating samples exhibited no detectable DME. Because DME is cytotoxic and known to inhibit radicle growth in other species, the low levels of DME but high levels of terpenoid chemicals in the actively elongating rhizomes may result from a biosynthetic or allocational strategy to limit DME autotoxicity to meristematic tissue.