Abstract Much of the UK uplands are dominated by species‐poor grasslands, largely a result of intensive sheep grazing. In some areas, efforts are underway to restore more natural habitats by reducing the number of sheep, with possible benefits to biodiversity, flood alleviation and carbon storage. Restoration aims to improve vegetation diversity, reduce the dominance of grasses and increase vegetation height and structural complexity. This study evaluates whether changing grazing from light sheep grazing to light cattle grazing restored vegetation at an upland calcareous site in the Yorkshire Dales National Park in England. Vegetation surveys were conducted in 19 quadrats of 2 m × 2 m, surveyed every 4 years over a 12‐year period. Generalised linear mixed models were used to understand how a number of key vegetation characteristics, relevant to restoration success, have changed with increasing years since sheep removal. Changing from sheep to cattle grazing led to a change in vegetation species composition. Species richness, calcareous grassland indicators and proportion cover of bryophytes were found to increase with increasing years since sheep removal. In contrast, vegetation height and the proportion cover of grass decreased. Our results show that switching from sheep to cattle grazing was highly successful in restoring diversity and habitat quality in calcareous landscapes. The less selective grazing of cattle led to a more open, less grass‐dominated sward, allowing calcareous grassland wildflowers such as Linum catharticum and Lotus corniculatus to increase in frequency. However, the vegetation became shorter, another important objective for vegetation restoration in the uplands, suggesting a possible trade‐off between diversity and vegetation height at a small scale. Practical implication . Changing from sheep grazing, even at a low intensity, to extensive cattle grazing restores calcareous habitat diversity and quality. Restoration of vegetation height and structure may require other grazing management such as livestock exclusion.
Porton et al. (Thu,) studied this question.