The effects of sheep grazing and gap creation on vegetation change in a species-poor grassland in Oxfordshire, England

Background

In the UK, increased fertilizer application to many grasslands since the 1940s has lead to dense agriculturally productive swards dominated by a few perennial grasses. Cessation of fertilizer application and introduction of an intensive grazing regime is considered  one way of reinstating a herb-rich sward. An experiment was undertaken on a nature reserve in southern England to study how plant abundance and diversity in a species-poor grassland where fertilizer application had recently ceased, changed over time with different grazing intensity. Artificial gaps in the sward were also created as a possible way of encouraging seedling recruitment.

Action

Study site: The experiment was set up in 1986 on species-poor grassland at Little Wittenham Nature Reserve, Oxfordshire. The site was last cultivated in the 1940s then sown with an agricultural grass mix, sheep-grazed and received regular fertilizer applications. The site was purchased by the Northmoor Trust in 1984, when application of agrochemicals ceased.

Experimental design and grazing intensity: A sheep grazing experiment (2 x 2 x 2 factorial design with two blocks in 16 (50 x 50 m) paddocks) was set using three recognized grazing seasons: winter (1 November-21 March), spring (21 March-21 May) and summer (21 May-1 November). In winter and spring, sheep grazing in each paddock was undertaken by two ewes, with an ungrazed control. During summer, the sward height was maintained at either 3 cm or 9 cm by adjusting the stocking rate.

Vegetation surveys: Surveys were undertaken between 5-31 January and 1-10 August 1990. In each paddock, 64 locations 4 m apart were sampled using a point-quadrat. Annual meadow-grass Poa annua, smooth meadow-grass P.pratensis and rough meadow-grass P.trivialis were indistinguishable when young and therefore grouped as Poa spp. As they were scare, dicotyledonous (dicot) species were recorded more fully. Presence/absence of rooted plants of each dicot species was recorded in 1m² quadrats at 100 locations 3 m apart. Four surveys were undertaken (mid-February, April, July and November) in 1990 and 1991.

Experimental gaps: On 26 September 1990, two transects were laid out in each paddock. Ten areas (each 10 cm diameter) 3 m apart were cleared of vegetation (including shoot bases). Half of the 20 gaps in each paddock selected at random had soil removed to 15 cm depth and then infilled with sterile loam i.e. equivalent to a seed rain only treatment. The other gaps were left intact i.e. seed bank and seed rain. Commencing two weeks later, any emerging seedlings were identified and regularly removed along with any clonal ingrowth, until 10 January 1992.

Soil nutrients: In 1986 and 1991, soil samples were taken (15 cm deep soil cores) and nutrient concentrations in each paddock analysed .

Consequences

Point-quadrat surveys: Vegetation was dominated by perennial grasses and four (perennial rye-grass Lolium perenne, creeping bent Agrostis stolonifera, Poa spp. and meadow barley Hordeum secalinum ) accounted for 80% of touches. Dicotyledonous species ('dicots') were very rare (overall frequency 0.4%). The frequencies of eight of the 10 dominant grasses (only 10 were abundant enough for analysis) were significantly affected by grazing intensity although these effects depended on the season, were species-specific and were generally small.

Forty species of dicots were recorded; most were rare. Six (daisy Bellis perennis, common mouse-ear Cerastium fontanum, creeping thistle Cirsium vulgare, cut-leaved crane's-bill Geranium dissectum, buttercup Ranunculus spp., dandelion Taraxacum offinale, white clover Trifolium repens and stinging nettle Urtica dioica) accounted for 80% of touches. The dicots exhibited stronger and more consistent responses than grasses, the number of dicot species and abundance being significantly increased by increased grazing intensity in all three periods. In only one species, T.officinale, did increased grazing intensity in one season (spring) have a detrimental effect. The pattern however was not consistent between the two years, probably due to the weather: species abundance generally increased in the summer of 1991, and was lower in the summer of 1990 when there was a drought.

Experimental gaps: Comparison between the species composition of seedlings emerging in gaps where the soil had been replaced with a loam and that of gaps over the original soil showed no evidence of a persistent seed bank and indicated that seeds were derived from recent seed rain. Over the 15 month survey, 14,720 seedlings of 27 species were recorded in the gaps, with 70 % of seedlings being of two species (Lolium perenne and Hordeum secalinum). Sixteen dicot species were recorded but made up only 4.2% of the seedlings. No species new to the sward emerged in the gaps and the species composition of seedlings in the gaps was correlated with that of the vegetation in a majority of the paddocks. However, some species differences in the contribution to the seed rain were noted. In particular, the dicots were over-represented. The number of grass seedlings in the gaps was decreased by increased summer grazing.

Soil nutrients: Despite cessation of fertilizer application there was no drop in fertility over five years, with high concentrations of phosphorous (35 mg/l), potassium (424 mg/l) and magnesium (113 mg/l) recorded.

This study suggests that changes in plant composition solely through grazing is likely to be slow, especially while soil fertility remains high.

Note: If using or referring to this published study, please read and quote the original paper.