Modern rural policies that incorporate agricultural and environmental aims within the broader framework of sustainable rural development are being formulated to address the problem of declines in grassland biodiversity and the destruction of sensitive landscapes and habitats in Europe. Extensification is the process of reducing fertiliser inputs, management intensity and stocking rates, and is central to these sustainable rural policies. However, research in the Less Favoured Areas of Europe has been fragmented and highly variable reflecting the different uses and requirements of our upland areas. Information is needed to determine the nature and timescale of changes in such systems, and whether extensive management is sustainable in the long-term. This paper presents results from a range of grassland extensification experiments across Europe, mainly within the European Union, over the past 30 years that quantify the impacts on soil, plant and animal components of the system. All have the common theme of changing the focus of land management from solely the agricultural product to include a broader range of ecological and environmental objectives. Beneficial changes in biodiversity resulted from more extensive management treatments, but at the cost of reductions in total animal output, and in some cases a reduction in individual animal performance. However, it is clear that it is a long-term process to achieve many of these changes in biodiversity, and this must be recognised by policy makers. We recommend that future extensification studies adopt an approach that will allow their results to be applied throughout Europe.

No abstract

No abstract

No abstract

Cattle and sheep can create and maintain a mixture of relatively tall and short patches in grass swards through selective grazing. In swards that are grazed by cattle this heterogeneous structure can result in the frequencies of height measurements having a skewed distribution that has variously been better described by the doublenormal distribution the gamma distribution and the Weibull distribution than by the more common normal distribution. The fit of these statistical distributions, and the adequacy of the potentially useful log-normal distribution, to sward height frequencies were tested in sown temperate swards grazed by sheep and compared within a single sward. It was concluded that the single-normal and Weibull distributions were inadequate and that overall the log-normal and gamma distributions had the best fit to the measurements.

Abstract. Due to economic pressures and policy changes Lolium perenne‐Trifolium repens sown swards in upland UK sheep systems are likely to become less intensively managed. We present results from the first 5 yr of a long‐term experiment studying vegetation change under more extensive grazing management at three sites. One treatment was representative of current, intensive management and 5 were unfertilized with different intensities of seasonal grazing. The species composition of unfertilized, ungrazed swards changed dramatically within 2 yr and the sown species had virtually disappeared by year 5. Ranunculus repens, Poa trivialis, Agrostis gigantea, Juncus spp. and Carex spp. became dominant at the wettest site. Grasses were dominant at the other sites. In contrast, the sown species were retained in the unfertilized, grazed treatments; there were small shifts in abundance of the species present initially and few additions or losses of species. Some colonizing species were present in the seed bank whereas others with a transient seed bank appeared to have invaded from neighbouring vegetation. Implications of these results for compensation schemes to reduce animal output and increase biodiversity are discussed.

Question: Predicting the impact of land‐use change on vegetation is vital to understanding how biodiversity and ecosystem function may respond. Is it correct to assume that abandonment is an extreme form of grazing reduction? Location: Borders and central Scotland. Methods: The analysis used data sets from two identical experiments where the impacts of two unfertilized, extensively grazed treatments and one unfertilized abandoned treatment were compared against the species dynamics of a pasture subject to normal, productive grazing management over a 16‐year period. Initial multivariate analysis using Principal Response Curves was used to assess if particular traits were associated with either extensive or abandoned treatments, and was checked using univariate tests of individual traits. RLQ analysis followed by clustering into response groups was used to assess if species behaved in a similar manner between sites. Results: For many traits/attributes the shift in value or proportion was approximately linear across the extensification treatments as grazing was removed. However, certain traits showed step changes and quadratic responses. Leaf dry matter content, an important effect trait, was in the latter group. Most traits/attributes and species behaved similarly at the two sites. However, traits such as regenerative strategy, seed length, longevity and mass and seed bank type behaved differently, indicating that they are not predictable response traits. Conclusion: The results indicate that responses to grazing removal during extensification are largely straightforward and largely independent of species pool. However, there are discrepancies that suggest that simple analyses of the impacts of land‐use changes such as grazing reduction may hide more complex responses.

No abstract