Conservation Evidence strives to be as useful to conservationists as possible. Please take our survey to help the team improve our resource.

Providing evidence to improve practice

Action: Reduce number of livestock Shrubland and Heathland Conservation

Key messages

Read our guidance on Key messages before continuing

  • Two before-and-after trials in the UK and South Africa and one replicated, controlled study in the UK found that the reducing or stopping grazing increased the abundance or cover of shrubs. Two site comparison studies in the UK found that cover of common heather declined in sites with high livestock density, but increased in sites with low livestock density. One site comparison study in the Netherlands found that dwarf shrub cover was higher in ungrazed sites. One replicated, randomized, before-and-after study in Spain found that reducing grazing increased the cover of western gorse. One randomized, controlled trial and one before-and-after trial in the USA found that stopping grazing did not increase shrub abundance. One site comparison study in France found that ungrazed sites had higher cover of ericaceous shrubs, but lower cover of non-ericaceous shrubs than grazed sites. One site comparison study in the UK found that reducing grazing had mixed effects on shrub cover.
  • One replicated, randomized, controlled study in the UK found that reducing grazing increased vegetation height. However, one replicated, controlled, paired site, site comparison study in the UK found that reducing grazing led to a reduction in the height of heather plants.
  • Two site comparison studies in France and the Netherlands found that ungrazed sites had a lower number of plant species than grazed sites. One replicated, controlled, paired, site comparison study in Namibia and South Africa found that reducing livestock numbers increased plant cover and the number of plant species.
  • One controlled study in Israel found that reducing grazing increased plant biomass. However, one randomized, site comparison on the island of Gomera, Spain found that reducing grazing did not increase plant cover and one replicated, controlled study in the UK found that the number of plant species did not change .
  • One replicated, controlled study in the UK found no change in the cover of rush or herbaceous species as a result of a reduction in grazing. Two site comparison studies in France and the Netherlands found that grass cover and sedge cover were lower in ungrazed sites than in grazed sites. One randomized, controlled study in the USA found a mixed effect of reducing grazing on grass cover.

Supporting evidence from individual studies

1 

A before-and-after trial in 1983–1990 at a moorland site in the UK (Anderson & Radford 1994) found that reducing sheep density increased the abundance of shrub and grass species. Cover was higher seven years after sheep removal for the shrubs common heather Calluna vulgaris (16%) and bilberry Vaccinium myrtillus (14%) than before removal (heather: 0%; bilberry: 1%). Cover and biomass of wavy-hair grass Deschampsia flexuosa were also higher seven years after sheep removal (cover: 83%; biomass: 56 g/plot) than before sheep removal (cover: 41%; biomass: 9 g/plot). Between 1982 and 1990 sheep numbers were reduced from 2.5 ewes/ha to 0.3 ewes/ha (with some fluctuations between). Vegetation cover was measured each spring at 1 m intervals along 10 permanent transects per plot.

2 

A replicated, site comparison study in 1994 at three heathlands in the UK (Bullock & Pakeman 1997) found that reducing the abundance of livestock had mixed effects on the cover of shrubs, grasses, and herbaceous plant species. Reducing the abundance of livestock increased cover of shrub species in one of nine comparisons (ungrazed: 80%, grazed: 45%) but reduced cover of shrub species in three of nine comparisons (ungrazed: 0–2%, grazed: 3–8%). For one of eight comparisons cover of grass species was lower in grazed (9%) than ungrazed areas (32%) but for three of eight comparisons cover of grass species was higher in grazed (3–8%) than ungrazed areas (0%). For three of six comparisons cover of herbaceous species was higher in grazed (2–3%) than in ungrazed areas (0%) while for the remaining three comparisons cover did not differ in grazed and ungrazed areas. At each site one area was grazed and another area was not grazed. At each site 5–15 quadrats were located randomly and cover of plant species estimated by eye.

3 

A replicated, randomized, controlled trial between 1989 and 1995 in wet heathland in Northumberland, UK (Hulme et al. 2002) found that reducing the number of sheep increased vegetation height after six years. Vegetation was taller in plots where grazing was light (37 cm) or moderate (37 cm) when compared to heavily grazed areas (12 cm). Heather Calluna vulgaris shoots were also longer in areas where grazing was reduced (63 cm) than in areas that were heavily grazed (44 cm). Four blocks, each composed of two 0.3 plots, were fenced in 1989. Within each block one plot had a density of 0.7 sheep/ha, and one plot had 1.4 sheep/ha. Nearby grazed heath with 2.1 sheep/ha was used for comparisons. Vegetation height was measured twice a year between 1989 and 1995 at 40 random locations in each plot using a sward stick.

4 

A before-and-after study in 1986–1992 in desert scrub in Arizona, USA (Krueper et al. 2003) found that removal of cattle did not affect the abundance of shrubs, herbaceous plants, or trees. Five years after cattle had been removed, the density of shrubs was not significantly higher than in the same area before cattle removal (data reported as a density index). The same pattern was seen for herbaceous plants and trees. In 1987 all cattle were removed from the area. Prior to removal the stocking rate of the area was approximately 0.5 cattle/ha. Two transects, measuring 0.8–1.6 km in length were established in the area. Vegetation was sampled along the transects in September/October in 1986, 1989, and 1992.

5 

A before-and-after trial in 1935–2004 in karoo habitat in Western Cape, South Africa (Rahlao et al. 2008) found that stopping livestock grazing led to increases in the cover of shrubs and trees, decreases in the cover of succulent plants, and no change in grass cover after 67 years. In three of four locations shrub cover was higher 67 years after grazing stopped (33–56%) than when the area was grazed (30–40%). Tree cover increased in two of four cases (before: 1–5%, after: 12–13%), while grass cover did not significantly change in four of four cases (before: 0–1%, after: 1–2%). However, in one of four cases the cover of succulent plants was lower 67 years after grazing was halted (38%) than before grazing stopped (60%). In 1935 all cattle were removed from the site. In 1935 vegetation cover was visually estimated in seven hundred and two 900 m2 quadrats and in 2004 in seventy 900 m2 quadrats distributed across four distinct areas of the site.

6 

A replicated, randomized, before-and-after trial between 2002 and 2006 in a heathland in Northern Spain (Celaya et al. 2010) found that reducing livestock numbers increased the cover of western gorse Ulex galli and the height of plants. After four years cover of western gorse was higher in lightly grazed plots (18%) than in heavily grazed plots (9%). Plants in lightly grazed plots were taller (27 cm) than in more heavily grazed plots (17 cm). There were no differences between lightly and heavily grazed plots in the cover of dwarf heather species (43% vs 34%), tall heather species (9% vs 6%), or herbaceous plant species (25% vs 27%). In addition there were no differences in plant species richness between lightly and heavily grazed plots (27 vs 28 species). In 2002 six 0.6 ha heathland plots were fenced. Each plot was randomly allocated either a high (15 goats/ha) or a low (7 goats/ha) grazing pressure. In each plot 100 random points were used to survey vegetation cover. Plant height was measured using a sward stick every 10 cm along a 50 m transect in each plot.

7 

A randomized, controlled study in 2001–2004 in a sagebrush-steppe habitat affected by wildfire in Idaho, USA (Roselle et al. 2010) found that areas that were not grazed did have higher shrub abundance than areas that were grazed, but there was mixed effect on grass abundance. Cover of the shrubs threetip sagebrush Artemesia tripartita and bitterbrush Purshia tridentata did not differ significantly between ungrazed (sagebrush: 1 plant/10 m2, bitterbrush: 0.3 plants/10 m2) and grazed areas (sagebrush: 2–4 plants/10 m2, bitterbrush: 0–0.6 plants/10 m2). Cover of bluebunch wheatgrass Pseudoroegneria spicata was higher in ungrazed (9%) than grazed areas (7%), but the abundance of drooping brome Bromus tectorum was lower in ungrazed (10 plants/10 m2) than grazed areas (12–60 plants/10 m2). In 2000 a fire burnt 474 ha of the site. Twenty four 2.4–3.3 ha paddocks were established in 2001. Twenty paddocks were grazed by sheep and four were not. Two 15 m x 1 m transects were established in each paddock and surveyed to estimate vegetation cover annually in 2001–2004.

8 

A replicated, controlled, paired side, site comparison study in 2007–2009 in sixteen heathland sites in Scotland, UK (DeGabriel et al. 2011) found that reducing sheep numbers reduced the height of heather (plants of the family Ericaceae). In sites without sheep heather was shorter (13–34 cm) than in sites where sheep were present (17–32 cm). Sheep were removed from eight sites and eight sites were grazed by sheep. All sites were grazed by red deer Cervus elaphus, but deer numbers were higher in areas where sheep had been removed. Six to twelve 10 x 10 m plots were established at each site. Within each plot vegetation height and cover was measured at 40 points.

9 

A randomized, site comparison in 1975–2005 in coastal shrubland on the island of Gomera, Spain (Bermejo et al. 2012) found that reducing the number of goats did not increase plant cover after 30 years. There was no significant difference in plant cover between areas where goats had been removed (67%) and areas where goats were present (67%). Goat abundance in grazed areas was low – 0.1 goat/ha. Twenty 30 m transects were randomly located in grazed areas and 16 in ungrazed areas. Plant cover was assessed every 30 cm.

10 

A replicated, controlled, paired site, site comparison study in 2005–2009 in three karoo shrubland sites in Namibia and South Africa (Hanke et al. 2014) found that reducing livestock numbers increased plant cover in three of three sites and increased the number of plant species in one of three sites. In three of three sites, areas where livestock numbers were low had higher plant cover (13–39%) than areas with high livestock numbers (8–29%). In one of three sites plant species richness was higher in areas with low livestock numbers (15 species) than areas with high livestock numbers (12 species), while at the other two sites there was no significant difference in plant species richness (low grazing: 23–36 species, high grazing: 24–34 species). High and low livestock areas were separated by a fence. Vegetation cover was monitored in 2005–2009 using between sixteen and twenty 100 m2 plots in both areas with low and high livestock numbers.

11 

A controlled study in 1992–2008 in an arid shrubland in Israel (Leu et al. 2014) found that reducing goat numbers increased plant biomass after 16 years. Plant biomass was higher in areas with low goat numbers (0.16 kg/m2) when compared to nearby heavily grazed areas (0.07 kg/m2). In 1992 goat numbers were reduced at the site to <0.5 goats/ha. In 2008 fifteen quadrats measuring 20 cm x 30 cm were placed in both the light and heavily grazed areas and all vegetation was removed. Vegetation was dried for 48 hours at 60°C, then weighed to estimate biomass.

12 

A site comparison in 1967-1987 in 15 heathland sites in the UK (Welch & Scott 1995) found that common heather Calluna vulgaris cover decreased at sites with high livestock density. Over 20 years cover of common heather increased in sites with low livestock density but decreased in sites with high livestock density (no data reported). Ten point quadrats were used to estimate vegetation cover at each sites in July or August every two years. Dung was counted to assess herbivore abundance at each site.

13 

A replicated, controlled, paired study in 1968–1990 in 22 heathlands sites in the UK (Hope et al. 1996) found that reducing the number of livestock increased cover of heather Calluna vulgaris, had no effect on cover of rush and herbaceous species, as well as the number of plant species, and reduced grass cover. Sites where grazing had been stopped had higher cover of heather than sites that were grazed, but cover of the dwarf shrubs cross-leaved heath Erica tetralix, bell heather Erica cinerea, or bilberry Vaccinium myrillus did not differ significantly between ungrazed and grazed sites (data not presented). Cover of grass species was lower in sites where grazing had stopped than in sites that were grazed (data not presented). Rush and herbaceous cover did not differ significantly between ungrazed and grazed areas, and the same was true for the number of plant species in each site (data not presented). Sites were paired based on soil type and details of previous land-use. Sheep were removed from ungrazed sites in 1964–1987. In the summers of 1989 and 1990 four to eighteen 1 m2 quadrats were placed in each site and the cover of each plant species assessed.

14 

A site comparison study in 2006 in six heathland sites in France (Gachet et al. 2009) found that sites not grazed by cattle had a lower number of plant species, lower cover of grass and non-ericaceous shrubs, but higher cover of ericaceous shrubs. Ungrazed sites had a lower number of plant species and cover of grass (species: 7–8 species/plot, grass: 2–24% cover) than grazed sites (species: 13–14 species/plot, grass: 47–54% cover). Three of the four ungrazed sites also had lower cover of non-ericaceous shrubs (9–58%) than grazed sites (66-67%). However, cover of ericaceous shrubs was higher in ungrazed sites (86-95%) than in grazed sites (56-63%). Two moderately grazed sites and four ungrazed sites were selected for study. In 2006 four 1 m2 plots were placed at each site and plant cover estimated.

15 

A site comparison in 10 heathland sites that were subject to grazing in the UK (Pakeman & Nolan 2009) found that the cover of common heather Calluna vulgaris increased in sites with low sheep density and declined in areas with high sheep density. Common heather cover increased in sites where there were fewer than two sheep/ha and declined when there were greater than two sheep/ha (data reported as model results). Heather cover was estimated in seventy four plots across the 10 sites. Stocking density at the sites varied between zero and six sheep/ha.

16 

A site comparison study in 2011 in coastal heathland in the Netherlands (Damgaard et al. 2013) found that ungrazed areas had fewer plant species, lower sedge Carex spp. and grass cover, but higher cover of dwarf shrubs. Ungrazed areas has a lower number of plant species (5 species/plot) than grazed areas (6 species/plot). Ungrazed areas also had lower sedge and grass cover than areas that were grazed (data not reported). However, cover of dwarf shrubs was higher in ungrazed areas than in grazed areas (data not reported). Grazed and ungrazed areas were separated with a fence. Grazed areas were stocked with sheep at a density of 5.7 sheep/ha. Thirty-two plots were located in the ungrazed area and 33 in the grazed area. At each plot a point frame was used to estimate cover of different plant species.

Referenced papers

Please cite as:

Martin P.A., Rocha R., Smith R.K. & Sutherland W.J. (2019) Shrubland and Heathland Conservation. Pages 493-538 in: W.J. Sutherland, L.V. Dicks, N. Ockendon, S.O. Petrovan & R.K. Smith (eds) What Works in Conservation 2019. Open Book Publishers, Cambridge, UK.