Drylands are marginal regions where low and variable rainfall, infertile soils and land degradation often constrain agricultural productivity (Safriel et al. 2005, Hein and De Ridder 2006, Zika and Erb 2009). Livestock grazing plays a major role in land use and income generation in drylands but is also one of the major desertification drivers globally (Asner et al. 2004, Safriel et al. 2005). Hence an acute sustainability challenge in land management is to adjust livestock pressure to the marginal dryland conditions. Adaptive grazing strategies have been developed in drylands including destocking and restocking in response to low and variable rainfall and associated vegetation dynamics. The term ‘adaptive’ indicates that land users base their decisions on ecological and economic considerations such as vegetation cover, pasture productivity, precipitation at the start of the growing season or capital available to purchase supplementary fodder. In adjusting stocking rates, land users take into account a land’s grazing capacity. The stocking rate depicts the ratio of livestock and available fodder while grazing capacity refers to the number of livestock that the vegetation can sustain. Hence, sophisticated balancing of stocking rate and grazing capacity is key to sustainable dryland management. In this modelling strategy, we consider common adaptive management strategies including opportunistic and conservational grazing management in combination with varying degree of risk aversion.

When using opportunistic strategies, pastoralists adjust livestock density to pasture productivity in each year (Westoby et al. 1989, Behnke et al. 1993). This means that the stocking rate is adjusted to the land’s grazing capacity. This allows pastoralists to directly benefit from annual productivity changes yet without considering the time span necessary for the vegetation to recover sufficiently. Other land users may follow conservational strategies excluding livestock from some part of their land in wet years while fully grazing the land in other years (Frank et al. 2007, Müller et al. 2007, Quaas et al. 2007). This ‘resting’ implies that stocking rates remain below the grazing capacity in wet years facilitating vegetation recovery and potentially higher stocking rates in a relatively short time. In addition, management depends on the degree of risk aversion. An important aspect is the perceived likelihood that productive land would degrade or degraded land would recover. Observed vegetation cover serves in this modelling approach as a proxy for degradation risk with lower cover indicating higher risk of potentially irreversible degradation. For example, land users who are not risk averse may continue to graze a given livestock number on a degraded, sparsely vegetated pasture (e.g. 30%) assuming that this vegetation cover could be maintained. Late destocking may however cause long-lasting or irreversible degradation. In contrast, more risk averse land users may reduce stocking rates earlier (e.g. 40% vegetation cover) possibly allowing the pasture to recover. The management scenarios considered in this modelling approach are summarised in Table 1.

Table 1: Scenarios of adaptive land management. (Note: Start conditions at degraded site are 45% vegetation cover (both Cyprus and Crete) and at restored site 73% (Cyprus) and 52% (Crete)).

Taken together, the management scenarios are designed to test the relevance of land users’ risk aversion and resting periods. Outcomes are evaluated over a 10-year period assuming that land users would expect improvements to materialise within this range. In a land users’ perspective, a major question is: how effective is a management strategy in safeguarding sufficient pasture in the following year while ensuring the restoration of a degraded pasture in the next 10 years? Or: Is there a risk that a healthy pasture may degrade under a certain management strategy in the next 10 years? Extreme environmental conditions can significantly alter ecological conditions and the potential for restoration and degradation including regime shifts (see »Conceptualising socio-ecological effectiveness of land management). Hence, management scenarios include windows of opportunities and risks, which consist of wet and dry years, both of which on average occur twice in 10 years in the Mediterranean. This frequency reflects observed climate variability in the Mediterranean region (Sousa et al. 2011, Vicente-Serrano et al. 2014).

All management scenarios imply annual decisions on livestock destocking or restocking. In the case of destocking, part or all of the livestock is kept in stables, rather than being sold, and entirely fed with supplementary fodder, whereas restocking implies that those animals are brought back to graze on the pastures. Ecological impacts may manifest themselves only gradually and longer adjustment periods may be required to yield significant changes. For example, low livestock pressure in a single year may not be sufficient for a severely degraded vegetation to recuperate. Longer reduction of livestock pressure may be necessary to trigger the desired vegetation recovery demanding longer-term expenditures, e.g. to purchase supplementary fodder when keeping livestock in stables.

Note: For full references to papers quoted in this article see

» References