Maintain species-rich, semi-natural grassland

How is the evidence assessed?
  • Effectiveness
  • Certainty
  • Harms
    not assessed

Study locations

Key messages

  • One site comparison from Finland and northwest Russia found that butterfly species richness, diversity and total abundance did not differ significantly between mown meadows and grazed pastures and that grassland age and origin had a greater impact on butterfly communities than present management.

About key messages

Key messages provide a descriptive index to studies we have found that test this intervention.

Studies are not directly comparable or of equal value. When making decisions based on this evidence, you should consider factors such as study size, study design, reported metrics and relevance of the study to your situation, rather than simply counting the number of studies that support a particular interpretation.

Supporting evidence from individual studies

  1. A 1994 review (Smith 1994) describes the results of a controlled trial in 1990-1992 on a species-rich upland meadow with a plant community characterized by sweet vernal grass Anthoxanthum odoratum and wood cranesbill Geranium sylvaticum (MG3 under the UK National Vegetation Classification) in the Pennines, northern England. The highest species richness (on average 17 species per 25 x 25 cm quadrat) was produced by a combination of a mid-July hay cut, spring and autumn grazing and no fertilizer. The lowest species richness was produced by no grazing, application of fertilizer and either a September (on average 10 species/quadrat) or June (11 species/quadrat) hay cut. Autumn-only grazing produced a vegetation layer with an intermediate species-richness (13-16 species/quadrat). Combinations of the following treatments were compared: cutting in mid-June, mid-July or early September; spring-sheep and autumn-cattle grazing, autumn-cattle grazing or no grazing; nitrogen, phosphorous, potassium (NPK) fertilizer (400 kg/ha) or no fertilizer. The number of replicates is not stated.

    Study and other actions tested
  2. The same trial is described in (Younger & Smith 1994), which reports the number of plant species found under single applications of the same treatments. The four treatments that produced the highest species richness were cutting in mid-July, grazing in autumn and spring, grazing in autumn only, and applying no fertilizer. Each of these treatments produced on average 15 plant species/quadrat (25 x 25 cm). Cutting earlier, in mid-June reduced species richness to 13 species/quadrat. Although grazing reduced the hay yield (from 6 t/ha to 4-5 t/ha), it was necessary to maintain species richness. Under grazed treatments, 15 plant species/quadrat were recorded, compared to 12 species/quadrat with no grazing. Use of NPK fertilizer increased hay yield (from 4 to 6 t/ha), but reduced species richness (from 15 to 13 species/quadrat). Plants were surveyed in June 1992 and dry matter yield was measured at the time of cutting (averaged between 1990-1992). The number of replicates is not stated.

    Study and other actions tested
  3. A long-term replicated controlled trial in the Jura Mountains near Schaufhaussen, Switzerland (Ryser et al. 1995) found plots that were cut annually (in July or October) or every second year (in July) retained a higher number of plant species than those burned, or cut less frequently than every two years. Frequently cut plots had 53 plant species/40 m2, and 37 species/m2 on average, compared to 45 species/40 m2 and 24 species/m2 on average for annually burned plots, those cut every fifth year and control unmanaged plots. There were three replicate 50 m2 plots for each treatment, and the experimental management regimes were carried out from 1978 to 1993. The percentage cover of plant species was estimated in 40 m2 and 1 m2 sample areas in each plot, in the last week of June 1991 and 1993.

    Study and other actions tested
  4. A 1997 review (Tilzey 1997) concluded that the Environmentally Sensitive Areas scheme made a significant contribution to halting the loss of semi-natural grasslands in England, but was less effective in enhancing and restoring grassland biodiversity. The paper made a broad assessment of the effectiveness of the scheme in protecting England’s lowland semi-natural grasslands, a decade after its introduction. Among Environmentally Sensitive Areas of greatest significance for their lowland grassland, six were of ‘outstanding’ significance (containing >40% of the English resource of a grassland type) and two were of ‘considerable’ significance (containing >10-40% of 1-2 grassland types or >5-<10% of three or more grassland types). Entry of land supporting semi-natural grassland was generally high (e.g. covering 80% of chalk grassland in the South Downs Environmentally Sensitive Area). However, there was evidence in some Environmentally Sensitive Areas that grassland habitats were declining in quality due to management being insufficiently tailored to biodiversity interest, such as permitting the use of inorganic fertilizers.

    Study and other actions tested
  5. A before-and-after trial in England (Defra 2002) concluded that management prescriptions in the Exmoor Environmentally Sensitive Area are maintaining the condition of unimproved grassland, based on trends in bird populations in parts of the Environmentally Sensitive Area under long-term management agreements. The same study found that five red/amber-listed species of conservation concern (Eurasian linnet Carduelis cannabina, Eurasian bullfinch Pyrrhula pyrrhula, grey partridge Perdix perdix, house sparrow Passer domesticus and garden warbler Sylvia borin) appeared to be increasing in density within the Cotswolds Environmentally Sensitive Area while declining nationally, suggesting that they benefit from some aspect of Environmentally Sensitive Area management. In each Environmentally Sensitive Area, breeding birds were surveyed in May-August 2002, and results were compared with baseline survey information from 1992 to 1993 (Exmoor) and 1997 (Cotswolds). In the Cotswolds Environmentally Sensitive Area, birds were surveyed in 96 randomly-selected 1 km squares, while the majority (153 km2) of the Exmoor Environmentally Sensitive Area was surveyed.

    Study and other actions tested
  6. A long-term replicated before-and-after trial in 1987-2002 on upland hay meadows in the Pennine Dales Environmentally Sensitive Area, County Durham, England (Defra 2004a) found that plant species richness improved overall between 1987 (when the Environmentally Sensitive Area was introduced) and 1995, then declined to close to its original level by 2002. Condition was broadly maintained and slightly enhanced on semi-improved and improved hay meadows, but there was deterioration among the best quality unimproved hay meadows, with an increase in grass species at the expense of herbs. There were clear relationships between vegetation change and management, with the following practices leading to reduced vegetation quality: early cutting (before 15 July), spring grazing (especially if prolonged after 15 May) and application of inorganic nitrogen. The first two practices led to a decline in herb richness, while the third led to a decline in characteristic hay meadow vegetation. The study built on monitoring carried out between 1987 and 1995, by resurveying a subset of 164 sites between June and August 2002. During the resurvey, vegetation was surveyed in three 1 m2 quadrats/site, and management information from 1995 to 2002 was collected by interviewing farmers.

    Study and other actions tested
  7. A randomized, replicated before-and-after trial in England (Defra 2004b) found that the conservation value of hay meadows in the Dartmoor Environmentally Sensitive Area increased during the nine years following its introduction in 1994. Eighteen randomly chosen Environmentally Sensitive Area hay meadows (with agreements aimed at habitat enhancement) were surveyed in June-July 1995 and 2003. Most hay meadows increased in conservation value, with the biggest improvement seen in poorer quality sites. There was an overall increase in plant species richness and the vegetation became closer to that of meadows characterized by crested dog’s tail Cynosurus cristatus and common knapweed Centaurea nigra (MG5 under the UK National Vegetation Classification Scheme). This was accompanied by a general trend of declining soil fertility and a narrowing of the difference in nutrient availability between sites (particularly potassium).

    Study and other actions tested
  8. A replicated before-and-after trial in England (Defra 2005a) found that the Environmentally Sensitive Area scheme maintained extensively-managed permanent pasture and wet grassland/hay meadows in the Avon Valley and Upper Thames Tributaries Environmentally Sensitive Areas, introduced in 1993 and 1995, respectively. The range of vegetation communities in these Environmentally Sensitive Areas was broadly similar between baseline surveys (following introduction) and resurveys in 2003. However, in the Somerset Levels and Moors Environmentally Sensitive Area (baseline surveys in 1993, 1995 and 1998), management under the ‘raised water tier’ of the scheme appeared to be encouraging the formation of less species-rich inundation and rush pasture communities, rather than maintaining species-rich grassland. There were 34 replicate sites in Avon Valley, 39 in Upper Thames Tributaries and 25 in Somerset Levels and Moors Environmentally Sensitive Area.

    Study and other actions tested
  9. A randomized, replicated before-and-after trial in England (Defra 2005b) found that management under the Environmentally Sensitive Area scheme broadly maintained species richness on enclosed rough grassland in the Blackdown Hills, Shropshire Hills and South West Peak Environmentally Sensitive Areas. In all three, the range of vegetation communities was similar between baseline surveys (in 1994-1995) and resurveys in May-August 2003. The number of species in survey plots was also similar between years, except in the South West Peak Environmentally Sensitive Area, where the average number of species per plot declined by 1.5 (to 23). The number of species at a finer within-plot scale declined in the Shropshire Hills and South West Peak Environmentally Sensitive Area, but not by an extent to indicate loss of ecological quality. While the results suggested that management prescriptions were broadly maintaining species richness, they appeared less effective in encouraging desirable species to colonize improved/semi-improved grasslands. There were 22 replicate sites in the Blackdown Hills, 29 in the Shropshire Hills and 38 in the South West Peak Environmentally Sensitive Area.

    Study and other actions tested
  10. A replicated site comparison study in 1998-2001 in Switzerland (Herzog et al. 2005) found that on average 86% of litter meadows in Ecological Compensation Areas on the Swiss plateau were of ‘good ecological quality’ and attracted wetland birds, which had significantly more territories (52) than expected (31) in these areas. However, only 20% of hay meadow Ecological Compensation Areas were of good ecological quality, and they did not appear to benefit birds of open cultivated land, containing fewer of these birds’ territories (68) than expected (151). Previous intensive management appeared to limit the effectiveness of hay meadow Ecological Compensation Areas, with significantly lower ecological condition in the more intensively farmed ‘lowland’ zone of the Swiss plateau, compared to ‘pre-alpine hills’ zone. Under the Ecological Compensation Area scheme farmers must carry out low intensity management on 7% of their land. For litter meadows, this includes traditional litter use, mowing and no fertilizer and for hay meadows, restrictions on fertilizer use and mowing (late cut). Plant species were recorded in 1,306 hay meadow Ecological Compensation Areas and 104 litter meadow Ecological Compensation Areas, in eleven study areas between 1998 and 2001. Breeding bird territories were mapped in 23 study areas, on three visits to each area between mid-April and mid-June.

    Study and other actions tested
  11. A 2005 review (Jefferson 2005) of seven studies exploring the role of cutting, grazing and fertilizer in maintaining species richness of upland UK hay meadows concluded that the best management involves spring and autumn grazing, a mid-July hay cut and no inorganic fertilizer. The review recommends using only low levels of farmyard manure as Edwards et al. 2002 and Tallowin 2005 found that it can lead to a shift towards improved grassland plant communities, and is unlikely to assist seed dispersal.

    Additional references:

    Edwards A.R., Younger A. & Chaudry A.S. (2002) The role of farmyard manure in the maintenance of botanical diversity in traditionally managed hay meadows: the effects of rumen digestion on seed viability. Pages 159-162 in: J. Frame (ed.) Conservation Pays? Reconciling Environmental Benefits with Profitable Grassland Systems. Occasional Symposium No. 36, British Grassland Society, Reading, UK.

    Tallowin J.R.B. (2005) The impact of organic fertilizers on semi-natural grasslands. Defra BD1415.

    Study and other actions tested
  12. A replicated study of different management regimes at two dry limestone grassland sites over 23 years in Switzerland (Köhler et al. 2005) found that compared to traditional management, alternative management regimes resulted in greater changes in plant species composition and abundance. Changes were most evident 12 years after the start and were continuing even after 22 years. By the end of the study at one site, the number of species had declined in the following order compared to plots under traditional management (annual mowing in July; 50 species): mowing annually in October (93% of species in traditionally managed plots), mowing every second year in July (90%), mowing every 5th year in July (79%), annual controlled burning February-March (79%) and abandoned (70%). At the other site, species number did not tend to differ with management (85-120% of traditionally managed plots). Cover of specific species tended to differ with management regime at both sites. Plots with mowing every second year in July, followed by mowing annually in October were the most similar in composition to traditional management. The six management regimes were investigated with three replicate plots (5 x 10 m) at each site from 1977 to 2001. Cuttings were removed immediately. Species composition was sampled 4-6 times at each site 1977-1996. The inner 4 x 9 m of each plot was also surveyed at both sites in May, June and September in 1997-1999.

    Study and other actions tested
  13. A site comparison study of 12 pastures and meadows (4 ha) in northwest Russia and four in Finland (Saarinen & Jantunen 2005) found that butterfly (Lepidoptera) species richness, diversity and total abundance did not differ significantly between mown meadows and grazed pastures, although meadows were preferred by more species (46 vs 42). A total of 3,660 individuals were recorded in the meadows and 2,082 in pastures. Butterfly communities were affected more by the origin and age of the grassland than the present management method. Landscape factors such as surrounding habitat, abundance of nectar plants and intensity of tilling were the most important factors differentiating older grasslands from the younger ones. Meadows were mown annually in late July or August and pastures were grazed by cattle, some with sheep or horses temporarily. Tilling and fertilization (manure) tended to occur at intervals of 3-10 years. Butterflies were sampled 11-13 times along transects (640-720 m) in June and July 1997-1999.

    Study and other actions tested
  14. A randomized, replicated, controlled trial on an abandoned meadow in northern Finland (Hellström et al. 2006) found that the number of plant species (excluding mosses and lichens) hardly changed over five years in response to regular annual mowing. However mosses and lichens were maintained by mowing and declined to almost nothing in unmown control plots. There were 14 plant species/plot on average in 1998 and 12 species/plot in 2003, with no differences between mowing treatments. Some species responded to treatments. For example, August mowing plus disturbance favoured harebell Campanula rotundifolia but reduced cover of common bent grass Agrostis capillaris. The cover of mosses and lichens fluctuated between years, but was not different between mowing treatments. The meadow was abandoned in 1985. In 1993, it was divided into forty 50 x 50 cm study plots, each at least 2 m apart, and annually mown in August, without grazing. From 1998, 10 plots were mown in June, 10 in August, and 10 mown in August with bare soil exposed in 25% of the plot area, using a spade. Ten control plots were not mown. The percentage cover of all plant species (including mosses and lichens) in the treatment plots was monitored in June every year from 1998 to 2003.

    Study and other actions tested
  15. A 2006 systematic review (Stewart & Pullin 2006) of the impacts of grazing management on one type of species-rich old meadow in the UK found that intermediate grazing intensities were generally most appropriate for plant conservation. Choice of stock type (sheep, cattle or horses) appeared less critical than intensity but there is some evidence that sheep grazing can result in lower diversity of herbaceous plants than cattle grazing at high stocking rates. Two studies showed that sheep grazing led to lower plant diversity and herbaceous plant cover than cattle grazing on old or restored pastures (one study for each type). Data from Welsh grasslands showed that with sheep grazing, a lower grazing intensity (grass height >10 cm) was needed to maximize the abundance and number of species of herbaceous plants than for horse or cattle grazed sites. There was little information regarding taxa other than higher plants. The review searched for studies from the UK or Ireland comparing the effects of sheep grazing with horse, cattle or no grazing, on pastures characterized by crested dog’s-tail Cynosurus cristatus and black knapweed Centaurea nigra, known as MG5 under the UK National Vegetation Classification. Studies on similar grassland types in northwest Europe were also considered. Forty-two studies were found, of which 22 were reviews.

    Study and other actions tested
  16. A before-and-after trial of 16 upland hay meadows in the Pennines, northern England (Critchley et al. 2007) found that 15 years of low intensity management under the Environmentally Sensitive Areas agri-environment scheme maintained the total number of plant species, but not the number of broadleaf (forb) species. The meadow plant communities were characterized by sweet vernal grass Anthoxanthum odoratum and wood cranesbill Geranium sylvaticum (MG3 under the UK National Vegetation Classification). The number of broadleaf species declined from 14 to 10 species/m2 on average between 1987 and 2002, while the average number of grass species increased from 8 to 12 species/ m2. The authors linked the loss of broadleaf species to inappropriate grazing intensity. Plant species were recorded in three 1 m2 permanent quadrats at each site, in 1987 (the year the scheme was introduced) and 2002. On 62 sites in the same study with areas of other types of species-rich grassland (not MG3), the average number of plant species increased slightly from 21.6 to 22.8 species/m2. Again, the increase was predominantly in the number of grass species, not broadleaf herbaceous species.

    Study and other actions tested
  17. A long-term unreplicated trial from 1975 to 2000 in Baden-Württemberg, Germany (Kahmen & Poschlod 2008) found that all tested treatments: low intensity grazing, mowing-and-mulching (in June or June and August) and winter burning, maintained a variety of types of semi-natural grassland. Plants showed similar functional responses to treatments: grazing encouraged species with small seeds and a persistent seed bank (and woody species, though this was probably due to very low grazing pressure) mulching led to an increase in ground-layer species and winter burning benefited species with storage organs. The authors suggest that this functional approach may help to predict vegetation changes following management. Treatments were applied in 20 x 40 m fields from 1975. Plants were surveyed in one 25 m2 plot/field until 2000 and graded according to 11 functional traits. The study was carried out over 14 sites, characterized by different grassland vegetation types. There were no replicates.

    Study and other actions tested
  18. A site comparison study in Switzerland (Kampmann et al. 2008) found that nine alpine meadows managed under the Ecological Compensation Areas agri-environment scheme for at least five years had more plant species (around 37 species/meadow on average) than nine conventionally managed meadows (around 27 species/meadow), but not more grasshopper (Orthoptera) species (3-3.5 grasshopper species/meadow on average). Conventionally managed meadows were cut two to six times annually, with 20-30 kg N/ha added after each cut as liquid or solid manure. Ecological Compensation Area meadows were first cut in July and had a maximum of 30 kg N/ha applied annually as solid manure. Sites in the study were randomly selected from target regions in the northern and east central Alps. Ecological Compensation Area sites were in higher, more remote locations and had steeper slopes than conventionally managed meadows. Since 79% of sites in the study had been managed in exactly the same way for 10 years, the results show that the Ecological Compensation Areas scheme can maintain higher plant species richness in alpine meadows.

    Study and other actions tested
  19. A randomized, replicated trial in 1995-2006 at a semi-natural grassland site near Maribor, Slovenia (Kramberger & Kaligaric 2008) found that occasional late cutting may allow farmers to maintain a stable grass yield, without stimulating the spread of broadleaved plants (forbs). The study applied different cutting frequencies (at 2, 4, 6, 8, 10 and 12 week intervals) on semi-natural grassland between 1995 and 2006, and measured the effect on the dry matter proportions of grasses, legumes and non-legume broadleaved plants in May 1995, 1999, 2002 and 2006. There were four replicates. In the final year (2006), the treatment with the least frequent cuts resulted in the highest proportion of grasses (77%) and lowest proportion of non-legume broadleaved plants (19%) in the dry matter of harvest. Although the digestibility of forage produced by infrequent cutting does not meet the needs of modern livestock production, occasional very late cuts could offer a compromise between maintaining grassland biodiversity, while allowing farmers to maintain a stable grass yield.

    Study and other actions tested
  20. A controlled replicated trial in the Harz mountains of northern Germany (Dierschke & Peppler-Lisbach 2009) found that annual mowing from the end of June onwards was the best method for conserving and restoring mountain meadows. Ideally, this management should occur within a mosaic of areas which are mown less frequently or not at all. Mulching was a less effective form of management, due to the development of a thick litter layer. Mowing every two or three years benefited species with more rigorous growth, and abandonment led to the development of stands of tall herb communities. The study was replicated over four meadow sites, with treatments (annual mowing, two or three yearly mowing, mulching and abandonment) commencing in 1987-1988, and plant surveys continuing until 2003.

    Study and other actions tested
  21. A replicated, controlled, randomized study of previously restored semi-natural grassland in central Finland (Ilmarinen & Mikola 2009) found that mowing affected woody plants, leguminous broadleaved plants (forbs) and bacterial-feeding nematodes (Nematoda), but not overall plant abundance or species diversity. Mowing significantly decreased cover of woody plants (mown: 2-10%; unmown: 10-45%) and increased cover of leguminous broadleaved plants (16-42% vs 5-18%), selfheal Prunella vulgaris, birdeye pearlwort Sagina procumbens and white clover Trifolium repens in 2004. Grasses and broadleaved plants and horsetails did not differ and there was no significant difference between annual and bi-annual mowing. Bacterial-feeding nematodes were more abundant in annually mown (9/g dry soil) compared to bi-annually mown (7/g) and unmown (4/g) plots in the upper, but not lower, part of the grassland. Fungal, root and omnivorous-feeding nematodes and pot worms (Enchytraeidae) were not affected by mowing. The three mowing treatments were allocated randomly to the 30 plots (1 x 1 m, 60 cm-wide buffer zones). Plants were sampled in June 2002-2004.

    Study and other actions tested
  22. An unreplicated trial in 1996-2007 on a species-rich mountain meadow in the Bohemian Forest Mountains, Czech Republic (Maskova et al. 2009) found that mulching once a year in July produced a greater number of plant species, and a greater proportion of broadleaved plants (forbs), than traditional mowing once a year in July (without mulching) or abandonment. Mulching promoted many short broadleaved plants, grasses and graminoids, which were suppressed in the fallow treatment by an increasing share of tall grasses. The study concluded that mulching may be a viable alternative for preventing succession in situations where regular mowing is not economically or technically feasible. Treatments were applied from 1996 in 50 x 100 m plots. Plant species composition was measured each year shortly before mowing or mulching (June/July) in five 1 m2 quadrats/plot. Aboveground plant biomass and litter (harvested from four 0.33 x 0.33 m subplots/plot) and belowground plant biomass (to 15 cm depth in four 0.15 x 0.15 m subplots/plot) were also measured.

    Study and other actions tested
Please cite as:

Dicks, L.V., Ashpole, J.E., Dänhardt, J., James, K., Jönsson, A., Randall, N., Showler, D.A., Smith, R.K., Turpie, S., Williams, D.R. & Sutherland, W.J. (2020) Farmland Conservation. Pages 283-321 in: W.J. Sutherland, L.V. Dicks, S.O. Petrovan & R.K. Smith (eds) What Works in Conservation 2020. Open Book Publishers, Cambridge, UK.


Where has this evidence come from?

List of journals searched by synopsis

All the journals searched for all synopses

Farmland Conservation

This Action forms part of the Action Synopsis:

Farmland Conservation
What Works 2021 cover

What Works in Conservation

What Works in Conservation provides expert assessments of the effectiveness of actions, based on summarised evidence, in synopses. Subjects covered so far include amphibians, birds, mammals, forests, peatland and control of freshwater invasive species. More are in progress.

More about What Works in Conservation

Download free PDF or purchase
The Conservation Evidence Journal

The Conservation Evidence Journal

An online, free to publish in, open-access journal publishing results from research and projects that test the effectiveness of conservation actions.

Read the latest volume: Volume 21

Go to the CE Journal

Discover more on our blog

Our blog contains the latest news and updates from the Conservation Evidence team, the Conservation Evidence Journal, and our global partners in evidence-based conservation.

Who uses Conservation Evidence?

Meet some of the evidence champions

Endangered Landscape ProgrammeRed List Champion - Arc Kent Wildlife Trust The Rufford Foundation Save the Frogs - Ghana Mauritian Wildlife Supporting Conservation Leaders
Sustainability Dashboard National Biodiversity Network Frog Life The international journey of Conservation - Oryx Cool Farm Alliance UNEP AWFA Bat Conservation InternationalPeople trust for endangered species Vincet Wildlife Trust