Grazing ecology and its application in the management of rangelands: The case of shrublands

The goal of this study was to review the existing research data of grazing ecology and management in Greek shrublands with a purpose to better understand how grazing ruminants select specific plants or specific patches in grazing lands and to examine whether this knowledge could be integrated in to land management practices. The specific interest was for the Mediterranean shrublands dominated by kermes oak (Quercus coccifera L.), which are important habitats for grazing ruminants and occupy a large part of the Greek territory. Kermes oak shrublands contain a diverse array of plant species and plant parts that vary in nutritional value, chemical and mechanical defences. Reviewing the research data we concluded that the desired percentages of shrub: grassland cover in shrublands should be 50:50% and the management goal should be the creation of a mosaic of open and semi-open, highly heterogeneous and diversified foraging resource. On such shrubland there is green cover (i.e. forage) for small ruminants throughout the year with shrub providing green forage in dry months. However, plant defences influence the foraging behaviour of herbivores and limit the ability of grazing animals to reach their potential for growth and production. Kermes oak appears with five different grazing types according to leaf morphology and grazing value. Plant defences of these types may be classified in to two categories (1) physical defences e.g. spines and (2) chemical defences e.g. plant secondary metabolites such as tannins. Experimental results were considered to understand how herbivores make choices between forage types in complex ecosystems such as the Greek shrublands with the aim of better predicting herbivore – shrubland interactions. The distributions of nutrients and plant defences in the kermes oak community was such that any preference for a particular type of kermes oak by ruminants was likely to be based on the avoidance of plant defences (leaf spines and/or chemical defences) rather than the selection for nutrient rich kermes oak types. For example, goats avoided kermes oak type V with larger spines and higher concentrations of phenolics compared to kermes oak type II, as both types had similar nutrient content. However, the fact that goats select kermes oak type V at an average level of 37% suggests that the physical and chemical defences of kermes oak do not represent a complete barrier to herbivory and goats can utilise all the kermes oak rangeland types. The diet selection results are discussed in relation to the current shrubland management practices in Greece with purpose to identify the optimal grazing management regime in shrublands.