Reduce chemical inputs in grassland management

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

Study locations

Key messages


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 replicated trial from 1972 to 1988 in the Drentsche A nature reserve, Drenthe, the Netherlands (Olff & Bakker 1991) found that stopping fertilizer inputs on grassland mown annually for hay led to a gradual change in plant species composition, and an increase in the number of species in two out of three experimental fields. All three fields had a maximum number of plant species recorded in the middle of the study (between 1980 and 1985), followed by a slight decrease in the number of species as species initially present were replaced. In all fields, previously dominant grass species Yorkshire fog Holcus lanatus and creeping bent Agrostis stolonifera were replaced by creeping buttercup Ranunculus repens and sweet vernal grass Anthoxanthum odoratum, amongst others. In 1988 there were 23 and 28 species on two peaty fields, which had risen from 19 and 20 species in 1974, and 30 plant species on a sandy field, the same number as in 1974. Plant species were monitored in six 4 m2 quadrats each June from 1974 to 1988, on two 50 x 10 m fields on peaty soil and one 20 x 10 m field on sandy soil. Fertilizer application stopped in 1972, and all fields were cut for hay either once or twice, but no earlier than July.

    Study and other actions tested
  2. Two long-term replicated trials near Wageningen in the Netherlands (Berendse et al. 1992) found that ceasing fertilizer inputs reduced the rate of plant species loss over 30 years, relative to conventional fertilizer application rate, but did not affect plant species loss over 17 years relative to a reduced rate of fertilizer application. The first experiment compared no fertilizer with 160 kg N/ha/year from 1958 to 1988, with two replicate 16 x 2.5 m plots of each treatment. Fertilized plots changed from 39 plant species to 10, while unfertilized plots had a slight but not significant drop in the number of species, from 33 to around 25. Plants were monitored annually, in fifty 0.25m2 quadrats in each plot. The second experiment compared no fertilizer with 50 kg N/ha/year, from 1972 to 1989, with four replicate 10 x 10 m plots for each treatment. The number of plant species steadily declined in both treatments from 20 in 1973 to 12 in 1989. All plots were mown twice a year, with hay removed

    Study and other actions tested
  3. A 1994 review of methods to restore grasslands in the Netherlands (Bakker 1994) found three experiments in which the plant community became more similar to species-rich, lower nutrient grassland over 20 years, after the cessation of chemical fertilizer input. The location of these experiments is not given, they may be those reported in (Olff & Bakker 1991). Plant species of fertilized, nutrient-rich grasslands (such as perennial rye grass Lolium perenne and broad-leaved dock Rumex obtusifolius) were gradually replaced by species of lower-nutrient grassland (such as red fescue Festuca rubra and sweet vernal grass Anthoxanthum odoratum on two sites). One site had dry sandy soil and two sites had wet peaty soils. Fertilizer applications were stopped in 1972 and hay was cut and removed in July each year. Plants were monitored from 1972 to 1992. The method was not considered as effective on the wet sites. The review briefly described the results of an extensive Dutch study on the effects of grassland management on birds (Dijkstra 1991). The study showed that reducing fertilizer to 200 kg N/ha favours common meadow birds such as northern lapwing Vanellus vanellus, Eurasian oystercatcher Haematopus ostralegus and black-tailed godwit Limosa limosa. Reducing fertilizer to 50 kg N/ha favours common redshank Tringa totanus, common snipe Gallinago gallinago and ruff Philomachus pugnax. Reducing to no fertilizer reduces meadow bird numbers. Data were not given.

    Additional references:

    Dijkstra H. (1991) Natuur- en landschapsbeheer door landbouwbedrijven : eindverslag van het COAL-onderzoek [Nature and landscape management by agricultural enterprises: final report of the COAL study]. University of Wageningen Monograph. COAL-publikatie report. 60.

    Study and other actions tested
  4. A trial from 1972 to 1989 on a grassland in the Netherlands (Neuteboom et al. 1994) found that the number of plant species hardly increased over a twelve-year period of reduced fertilization, and the abundance of herbaceous (non-grass) species decreased. From 1986 to 1988 no fertilizer was applied and there was no increase in the number of plant species. By 1990, only two herb species (cuckoo flower Cardamine pratensis and meadow buttercup Ranunculus acer now R. acris) were present in more than 5% of 100 vegetation samples. The authors suggest that plant diversity did not increase because of the dense growth of a small number of competitive grass species. Species that could have colonized were present on the field borders and ditches. The 6.6 ha grassland in the Netherlands (location unknown) had been fertilized at 200 kg N/ha, grazed and mown for silage for many years. From 1973-1985, fertilizer was reduced to 50 kg N/ha. The stocking rate was five steers/ha from April to July, then 3.5 steers/ha until October. From 1986-1988, no fertilizer was applied. Three paddocks of 2.2 ha had stocking rates of 2.3, 3.6 or 4.9 steers/ha from April to October.

    Study and other actions tested
  5. A replicated controlled trial from 1990 to 1995 on two sown agricultural grasslands in Scotland (Marriott et al. 1996) found that ceasing fertilizer input had only small effects on the vegetation over six years, if grazing continued. On plots with a grass height of 4 cm in summer (75% of the number of ewes relative to fertilized grassland), white clover Trifolium repens increased from 10% to over 20% by 1994. White clover did not increase on plots with a grass height of 8 cm in summer (42-57% of the number of ewes relative to fertilized grassland). Perennial rye grass Lolium perenne was more sensitive to autumn grazing pressure, and decreased on unfertilized plots grazed to 4 cm in the autumn (from around 60% to 33-35% by 1994). Both white clover and perennial rye grass were sown agricultural varieties. Unsown species only increased substantially in unfertilized plots left ungrazed. There were two replicates of each treatment at each of two upland sites: Hartwood Research Station in central Scotland and Sourhope Research Station in southeast Scotland. The percentage cover of different plant species in each plot was measured in 1990, 1992 and 1994, using a point quadrat.

    Study and other actions tested
  6. A replicated, controlled, before-and-after trial from 1991 to 1993 on a species-rich hay meadow at Tadham Moor in Somerset, UK (Mountford et al. 1996) found that ceasing fertilizer application for three years led to a gradual increase in the number of plant species. All plots where fertilizer was stopped in 1991, having been applied at four different levels from 1986 to 1989, had fewer plant species than unfertilized control plots throughout the experiment (they did not revert to the original condition in three years). However, in 1992 and 1993, Yorkshire fog Holcus lanatus and perennial rye grass Lolium perenne declined in the plots without fertilizer, and were being replaced by common bent Agrostis capillaris and crested dog’s tail Cynosurus cristatus. In 1993, plots without fertilizer had more plant species than plots with continued fertilizer, and species richness was increasing at an estimated 1 species/m2/year in all treatments. The average estimated time for the vegetation to revert to the original community following three years of fertilizer application was between four (at 25 kg N/ha/year) and eight years (for 50-200 kg N/ha/year). From 1986-1989, experimental plots (1.5 x 5 m) were fertilized at five different levels: 0 (control plots), 25, 50, 100 and 200 kg N/ha/year. There were three replicates of each treatment. From 1990 to 1993, half of each plot continued with the same fertilizer treatment as before, the other half stopped receiving any fertilizer. Plants were monitored on sixteen 1 m2 quadrats/plot, fifteen times between 1991 and 1993.

    Study and other actions tested
  7. A replicated study over one year of a neutral meadow grassland at a farm in England (Smith et al. 1996) found that fertilizer treatments significantly affected seed numbers for a small number of plant species. More downy oat-grass Avenula pubescens seed was recorded where fertilizer had been applied, conversely there was less ribwort plantain Plantago lanceolata, yellow rattle Rhinanthus minor and common daisy Bellis perennis. The meadow was divided into nine contiguous 20 x 30 m plots in 1990. Treatments were mineral fertilizer (June cut), no fertilizer (July cut, autumn cattle grazing, spring sheep grazing) and a September cut. The long-term management of the meadow prior to the experiment involved manure spreading in April-May each year. Vegetation was cut at 3 cm in three randomly placed 0.06 m² quadrats and seed collected.

    Study and other actions tested
  8. A review of experimental evidence (largely published in German language) in 1997 (Nösberger & Kessler 1997) described a 40 year experiment with different fertilizer application rates on the Eggenalp, in the Bernese Oberland, Switzerland (Baumberger et al. 1996) which found that the number of plant species was highest on unfertilized plots and decreased as fertilizer rates increased. There were fewer than 40 species in all fertilized plots (fertilizer application rates not given here).

    Additional references:

    Baumberger N., Koch B., Thomet P., Christ H. & Gex P. (1996) Entwicklung der artenvielfalt im langzeitversuch Eggenalp [Development of species diversity in the Eggenalp long-term experiment]. Agrarforschung, 3, 275-278.

    Study and other actions tested
  9. A 1998 review of how soil animals, especially nematodes (Nematoda) and microarthropods, change according to management of agricultural grasslands (Bardgett & Cook 1998) found two studies showing a higher density of nematodes or microarthropods in organically managed or low input grasslands, compared to intensively managed grasslands with conventional chemical inputs (Siepel 1996, Yeates et al. 1997). One also found higher numbers of microarthropod species on low input grasslands (Siepel 1996). This is in contrast to four studies that found that adding mineral fertilizer can increase numbers of nematodes in the short term (Edwards & Lofty 1969, Coulson & Butterfield 1978, King & Hutchinson 1980, Bardgett et al. 1993), although one study found reduced abundance and diversity of microarthropods after nitrogen fertilizer was added to grassland (Siepel & van de Bund 1988). The authors argue that soil communities are functionally different under low input systems.

    Additional references:

    Edwards C.A. & Lofty J.R. (1969) The influence of agricultural practices on soil micro-arthropod populations. Pages 237-246 in: J.R. Sheal (ed.) The Soil Ecosystem, Systematics Association, London.

    Coulson J.C. & Butterfield J.E.L. (1978) An investigation of the biotic factors determining the rates of plant decomposition on blanket bog. Journal of Ecology, 66, 631-650.

    King L.K. & Hutchinson K.J. (1980) The effects of superphosphate and stocking intensity on grassland microarthropods. Journal of Applied Ecology, 17, 581-591.

    Siepel H. & van de Bund C.F. (1988) The influence of management practices on the microarthropod community of grassland. Pedobiologia, 31, 339-354.

    Bardgett R.D., Frankland J.C. & Whittaker J.B. (1993) The effects of agricultural practices on the soil biota of some upland grasslands. Agriculture, Ecosystems and Environment, 45, 25-45.

    Siepel H. (1996) Biodiversity of soil microarthropods: the filtering of species. Biodiversity and Conservation, 5, 251-260.

    Yeates G.W., Bardgett R.D., Cook R., Hobbs P.J., Bowling P.J., & Potter J.F. (1997) Faunal and microbial diversity in three Welsh grassland soils under conventional and organic management regimes. Journal of Applied Ecology, 34, 453-471.

    Study and other actions tested
  10. A 1998 review of case studies in France, gathered from published and unpublished literature (Muller et al. 1998) found no monitoring results from France for the effects of reduced management intensity on agriculturally improved grasslands. Three Dutch studies were cited (including (Berendse et al. 1992)) which showed that stopping fertilization does not cause a rapid increase in plant species richness.

    Study and other actions tested
  11. A 2000 literature review of grassland management practices in the UK (Wakeham-Dawson & Smith 2000) found one study that reported that densities of invertebrates such as species of mites and ticks (Acari), springtails (Collembola), flies (Diptera), beetles (Coleoptera) and millipedes and centipedes (Myriapoda) were higher in unfertilized permanent pasture than pasture receiving over 140 kg nitrogen/ha/year (Curry 1994). One study found that although soil macro-invertebrate densities did not differ in fields with and without farmyard manure applications, bird usage was higher in those that had received moderate applications (Tucker 1992).

    Additional references:

    Tucker G.M. (1992). The effects of agricultural practice on field use by invertebrate-feeding birds in winter. Journal of Applied Ecology, 29, 779-790.

    Curry J.P. (1994) Grassland Invertebrates. London, Chapman & Hall.

    Study and other actions tested
  12. A randomized, replicated, controlled trial from 2003 to 2006 on four farms in southwest England (Defra 2007) (same study as (Woodcock et al. 2007, Potts et al. 2009)) found that 50 x 10 m plots of permanent pasture cut just once in May or July or not at all during the summer and left unfertilized supported greater numbers and more species of beetles (Coleoptera) in suction traps, true bugs (Hemiptera) and plant hoppers (Hemiptera: Fulgoroidea), as well as greater abundances of spiders (Araneae), crane and St Mark’s flies (Diptera) and more species of woodlice (Isopoda) than control fertilized plots cut in May and July (managed for silage). Small insectivorous birds (dunnock Prunella modularis, wren Troglodytes troglodytes and European robin Erithacus rubecula) and seed-eating finches (Fringillidae) and buntings (Emberizidae) preferred the less intensively managed treatments (particularly the plots uncut in summer) to control plots for foraging. There were twelve replicates of each management type, monitored over four years.

    Study and other actions tested
  13. A replicated, randomized study of three intensive cattle farms in Ireland over three to five years (Geijzendorffer 2007) found that the number of plant species per field was significantly higher with reduced fertilizer application on only one of three farms, although numbers per quadrat increased with decreased fertilizer application at all sites. At Johnstown Castle, the total number of species per field with no fertilizer application (10 species) was significantly higher than fields with 225 (5-7 species) or 390 kg N/ha (6-8 species). There was no significant difference at Solohead (80 kg N: 5-8 species, 175 kg N: 5-10 species, 225 kg N: 3-8 species, 350 kg N: 6-9 species) or Grange (88 kg N: 10-13 species, 225 kg N: 8-15 species). However, the average number of species per quadrat decreased with increasing levels of fertilizer (e.g. Johnstown Castle: 0 kg N: 4 species, 390 kg N: 3 species; Solohead: 80 kg N: 3 species, 350 kg N: 2 species; Grange: 88 kg N: 4 species, 225 kg N: 3 species). Fertilizer treatments were applied in a randomized block design with two to five replicates per treatment. Vegetation was sampled in 50 quadrats (3 dm³) in each field, three to five years after the treatments commenced.

    Study and other actions tested
  14. A randomized, replicated, controlled trial in 2003-2005 on four farms in southwest England (Woodcock et al. 2007), (same study as (Defra 2007, Potts et al. 2009)) found that 50 x 10 m plots of permanent pasture cut just once in July or not at all during the summer and left unfertilized attracted a greater abundance and more species of beetle (Coleoptera) than control fertilized plots cut in May and July (managed for silage), in the third year of monitoring. Plots without fertilizer added also had higher proportions of seed- and flower- feeding beetle species in the community. There were twelve replicates of each management type, monitored over three years.

    Study and other actions tested
  15. A randomized, replicated, controlled study in 1999-2005 on four hay meadow sites in Cumbria and Monmouthshire, UK (Kirkham et al. 2008) found that applying fertilizer reduced the number of plant species. The number of species declined at all sites when 24 t/ha/year of farmyard manure was applied. The maximum level of manure that could be applied without reducing species richness depended on past site management. The study compared two pairs of unimproved and semi-improved meadows. On the semi-improved Cumbrian meadow, which had previously been fertilized, species richness was unaffected when manure was applied at 12 t/ha/year. However, on the Monmouthshire meadows, which had no recent history of fertilizer use, species richness was reduced by even low levels of manure (≤ 6 tonnes/ha/year). The effects of liming also depended on past site management. Treatments were applied in March/April in 7 x 5 m plots from 1999 to 2005, with plants surveyed annually in May in three 1 m2 quadrats/plot. Treatments were replicated in three plots at each study site.

    Study and other actions tested
  16. A replicated, controlled trial in the Czech Republic (Hrevušová et al. 2009) found that 16 years after fertilizer applications were stopped, the effects of different rates of fertilizer applications on plant communities were still apparent. Cover of tall nitrogen-loving grasses (especially Yorkshire fog Holcus lanatus) in the treatment with 400 kg N/ha plus PK remained significantly higher than in control plots in 2007. Meadow buttercup Ranunculus acris was most negatively affected by former application rates. Although species richness was not statistically significantly affected by treatment, the number of species decreased from controls to treatments receiving 400 kg N/ha plus PK, after 16 years. The following treatments were applied to 5 x 6 m plots of alluvial meadow foxtail Alopecurus pratensis grassland: unfertilized, phosphorous and potassium (PK), 100 kg N/ha plus PK, 200 kg N/ha plus PK, 300 kg N/ha plus PK, 400 kg N/ha plus PK, from 1966 to 1990 or 1975 to 1990 for the final two treatments. Annual application rates of P and K were 40 and 100 kg/ha, respectively. There were four replicates of each treatment. Fertilizer application was stopped in 1991 in half of each plot and the responses of plant communities monitored until 2007.

    Study and other actions tested
  17. A randomized, replicated, controlled trial from 2003 to 2006 in southwest England (Potts et al. 2009) (same study as (Defra 2007, Woodcock et al. 2007)) found plots of unfertilized permanent pasture cut just once in July or not cut at all during the summer attracted more adult butterflies (Lepidoptera), but not more butterfly species or common bumblebees Bombus spp. than control fertilized plots cut in May and July (managed for silage). Plots cut just once in May, plots cut twice either unfertilized or ungrazed, and plots with a higher cutting height did not support more adult butterflies than control plots. Caterpillars were more abundant in unfertilized plots cut just once in May or July, or not at all in summer, than in other treatments. None of the grass treatments supported more common bumblebee species or individuals than control plots. Experimental plots 50 x 10 m were established on permanent pastures (more than five-years-old) on four farms. There were nine different management types, with three replicates/farm, monitored over four years. Seven management types involved different management options for grass-only plots, including conventional silage practices, no cutting in summer, early summer cut (May), late summer cut (July), raised mowing height. Bumblebees and butterflies were surveyed along a 50 m transect line in the centre of each experimental plot, once a month from June to September annually. Butterfly larvae were sampled on two 10 m transects using a sweep net in April and June-September annually.

    Study and other actions tested
  18. A replicated, controlled study from 2006 to 2009 on four permanent improved grassland fields in Herefordshire and North Yorkshire, UK (Defra 2010) found fertilized plots with moderate grazing and without early cattle exclusion had consistently higher invertebrate and bird food invertebrate abundance compared to plots with the same grazing treatments but with reduced fertilizer application (eg. 2009: approximately 45 average total invertebrate numbers in reduced fertilizer vs approximately 80 in normal fertilizer plots), however these differences were not significant. Plots without early cattle exclusion (grazing until October) received either reduced fertilizer input (a single 50 kg N/ha application) or normal fertilizer input (three 50 kg N/ha applications over the growing season). Grazing control was based upon weekly measures of grass height. Prior to the experiment, fields had received fertilizer inputs of approximately 150 kg N/ha/year. Vegetation composition was surveyed in four 1 x 1 m quadrats/plot. Invertebrates were surveyed using sweep nets, Vortis suction sampling and pitfall traps.

    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