Goat herding in Mediterranean shrubland serves multiple purposes: it is an effective tool for ecologically sound management of the landscape; it helps maintain buffer zones around towns and settlements; it mitigates fire hazards by reducing biomass; and it converts woody and herbaceous vegetation into marketable food products. Improved understanding of the interface between goats and the shrubland vegetation could promote efficient management of this grazing system, and allow a more precise understanding of the nutritional potential provided in this habitat. Most of the research studies conducted in this field have examined the interaction between the goats and the shrubs up to the species level, and from a large-scale perspective that encompasses the entire landscape. To date, little is known about the grazing behavior of goats at much smaller spatial scales and about the patterns of interaction between the goats and individual shrubs within a localized area. The present research focused on relatively small localized areas of shrubland, and attempted to examine the grazing behavior of goats at a high resolution up to the scale of the individual shrub. The purpose of this research was to answer the following question: Is the intake of goats foraging the local population of a given shrub species necessarily distributed uniformly among the individual shrubs? Pistacia lentiscus L., an important shrub species in the Mediterranean shrubland, was chosen as a model species. The research started with a preliminary trial that demonstrated under controlled conditions a particular case of preference by goats between adjacent shrubs of P. lentiscus. These observations suggested that the sense of smell is involved in discrimination between individual shrubs. A negative relation was found between the volatilized quantities of the monoterpene -myrcene and of the total essential oil yield, on one hand, and the preference of goats between shrubs, on the other hand. In the second stage, eight contrasting pairs of shrubs were located in the field showing clear, pronounced evidence of goat preference for one shrub of each pair. A significant negative correlation was found between the volatilized quantities of the monoterpene -myrcene, as well as the monoterpene fraction as a whole, and the preference of the goats. In the third stage, a comprehensive analysis was conducted of the essential oil composition of a local population of P. lentiscus that was found in a representative localized area of 0.9 ha at the south of Mt. Carmel. It was found that the local population comprised four distinct chemotypes, with 88% of the individuals belonging to two main chemotypes: [Germacrene D, Limonene] and [Germacrene D, -pinene]. In the fourth stage, a series of preference trials were conducted under controlled conditions, in which goats were presented branches from pairs of shrubs taken from the localized
population that was characterized above. The trials focused on the preference of the goats when presented two shrubs of different chemotypes, as well as their preference when presented two shrubs of the same chemotype. The goats showed significant preference for shrubs of chemotype [Germacrene D, Limonene] over those of chemotype [Germacrene D, -pinene], and the strength of the preference was medium. Three shrub characteristics were found to be significantly correlated to the observed allocation of intake between the test samples, and explained it quite well: 1. tannin content (P < 0.0001; R2 = 0.34), with a negative correlation; 2. limonene content (P = 0.0002; R2 = 0.27), with a positive correlation; 3. pinene content (P = 0.0015; R2 = 0.20), with a negative correlation. The best correlation to the intake proportion was obtained by a combination of the two factors: tannin content and limonene content (P < 0.0001; R2 = 0.39). This research demonstrated – to the best of our knowledge for the first time – two phenomena: 1) the existence of significant chemical variability within a single woody species population located in a small, localized area of a shrubland; 2) non-uniform distribution of intake by a herbivore browsing on adjacent individuals of a woody species in a Mediterranean shrubland. The findings indicate that the display of preference by goats browsing on adjacent individuals of P. lentiscus is probably not an unusual phenomenon in the field. It seems reasonable to assume that a similar potential for chemical variability exists also in local populations of other woody species in the shrubland. It could also be assumed that preference-based intake from adjacent individuals of the same species exists also in other herbivore species.