Tarbush (Flourensiacernua DC.) is a generally unpalatable shrub native to the Chihuahuan Desert and with a resinous leaf surface containing several terpenes. Tarbush epicuticular wax concentration has previously been linked to degree of use by livestock in our laboratory. In 1989 and 1990, cattle, sheep, and goats were densely stocked in tarbush-containing paddocks for 6 to 9 days. Degree of use of 20 permanently marked plants in each paddock (8 paddocks, 2 sites, only 10 plants were monitored in 4 of the paddocks the first year) was recorded daily both years. Thus, no destructive sampling of plants was possible. A high percentage of tarbush in diets and differential use of individual tarbush plants were observed in this companion study. In 1991, leaves were collected from these plants (n = 160) during the same stage of maturity. Leaf surface compounds were extracted with ethanol, and mono- and sesquiterpenes were analyzed using gas chromatography/ion trap mass spectrometry. Our objective was to examine whether degree of tarbush use during the high-density study was related to chemistry of the same plants the following year. Plants were separated into three categories after preliminary data evaluation (Category 1: > 50% use at period midpoint; Category 2: < 50% use at midpoint but > 50% use at period end; Category 3: < 50% use at period end). Multivariate analysis is being used to examine relationships of individual plant use with dry matter, ash, epicuticular wax, and several individual and subgroups of mono- and sesquiterpenes (oxygenated and hydrocarbon monoterpenes and sesquiterpenes). Preliminary evaluation suggests chemical subgroup variables in 1991 were related to degree of herbivory in 1989 and 1990, although statistical analyses are not complete. Monoterpene hydrocarbons, oxygenated monoterpenes, and oxygenated sesquiterpenes were all sign different for the three use categories. We will determine which individual compounds contribute to these differences. Estimated concentrations of unidentified compounds will also be evaluated. One assumption is that heavy browsing during 1989 and 1990 did not cause plant secondary chemistry changes which persisted in 1991. Of the 120 plants with use data for both years, 66 plants in a given use category for 1989 were in the same category in 1990 and 38 plants were in a higher category. Only 15 plants fell to a lower use category and only 1 plant dropped from Category 1 to 3. Based on these observations, degree of use was reasonably consistent for individual plants across years, suggesting that, at least between 1989 and 1990, if induction occurred, it was short term. Leaf surface chemistry appears to be related to degree of use of individual tarbush plants by livestock.