H. Verheyden-tixier et P. Duncan, Selection for small amounts of hydrolysable tannins by a concentrate-selecting mammalian herbivore, J CHEM ECOL, 26(2), 2000, pp. 351-358
Although herbivores usually avoid plants with high concentrations of tannin
s, roe deer in natural habitats select tannin-rich plants. We tested the hy
potheses that: (H1) roe do not seek tannins, rather they select for other (
unmeasured) factors such as minerals or vitamins; and (H2) roe do select fo
r tannins. Tame roe deer were offered a choice of food pellets to which hyd
rolyzable chestnut tannins had been added or control pellets. The daily con
sumption of pellets containing a high level of tannins (9.1%) was lower tha
n consumption of control pellets. However, when offered medium level tannin
pellets (3.8%), roe deer ate more tannin than control pellets. Moreover, t
he rue regulated their intake of these tannins closely. The regression of t
annin intake versus total pellet intake across all the tests was linear and
precise (r(2) = 0.94): intake was about 28 g of tannin per kilogram of pel
lets eaten. We therefore reject H1 and suggest that roe can monitor tannin
concentrations in their food, and regulate their intake precisely. The resu
lts suggest that a certain level of tannins induces no costs or that they p
rovide a nutritional benefit for roe deer.