Delay mowing or first grazing date on pasture or grassland
Overall effectiveness category Unknown effectiveness (limited evidence)
Number of studies: 11
Background information and definitions
This intervention involves delaying mowing or the onset of grazing on grasslands until later in the year. This may reduce damage to insect and spider natural enemy populations (or increase damage to pests) at sensitive points in their lifecycles, such as before overwintered individuals begin breeding. Ground-living invertebrates can be sampled by suction sampling, using a vacuum to suck-up and collect specimens for a given time or area of ground.
Supporting evidence from individual studies
A replicated, randomised, controlled trial in 1973-1975 on a tall oatgrass Arrhenatherum elatius-dominated grassland in Cambridgeshire, UK (Morris 1979) (same study as Morris 1981a, 1981b) found delaying the mowing date resulted in fewer bugs (Heteroptera). There were more bugs in plots cut in May only (averaging 176 individuals/plot) than plots cut in July only (52 individuals). Uncut plots had 288 individuals/plot. Four cutting treatments (uncut, May cut, July cut, May and July cuts) were replicated four times and randomly allocated to plots of 16 x 12 m. Invertebrates were sampled in October 1972-December 1975 using a D-Vac suction sampler.Study and other actions tested
A replicated, randomised, controlled trial in 1973-1975 on a tall oatgrass Arrhenatherum elatius-dominated grassland in Cambridgeshire, UK (Morris 1981a) (same study as Morris 1979, Morris 1981b) found that delaying mowing resulted in significantly fewer leafhoppers (Auchenorrhyncha). There was an average of 4,546 individuals/plot in plots cut in May compared to 1,906 individuals in plots cut in July. Uncut plots had similar numbers of leafhoppers (5,666 individuals) to plots cut in May. There was no difference in the number of species between cutting treatments. Four cutting treatments (uncut, May cut, July cut, May and July cuts) were replicated four times and randomly allocated to plots of 16 x 12 m. Invertebrates were sampled in October 1972-December 1975 using a D-Vac suction sampler.Study and other actions tested
A replicated, randomised, controlled trial in 1973-1975 on a tall oatgrass Arrhenatherum elatius-dominated grassland in Cambridgeshire, UK (Morris 1981b) (same study as Morris 1979, Morris 1981a) found that three of seven leafhopper (Auchenorrhyncha) species were more abundant on plots cut in July than in plots cut in May, in at least one of the three years. There were significantly more Macrosteles laevis and Neophilaenus campestris individuals on plots cut in July-only than May-only in all three years (M. laevis: 0.8-1.9 individuals in May-only vs. 0.1 individuals in July-only plots; N. campestris: 1.0-2.5 vs. 0.0-0.3). There were more Adarrus ocellaris leafhoppers on plots cut in July-only than May-only in 1975 (10.3 vs. 1.5 individuals). However, one leafhopper species, Recilia coronifera, was less abundant on plots cut in July than May-only in 1975 (0.0 vs. 0.4 individuals). Four cutting treatments (uncut, May cut, July cut, May and July cuts) were replicated four times and randomly allocated to plots of 16 x 12 m. Invertebrates were sampled in October 1972-December 1975 using a D-Vac suction sampler.Study and other actions tested
A replicated, randomised, controlled trial on plots of bird's-foot trefoil Lotus corniculatus pasture mixes in 1984-1985 at two sites in West Virginia, USA (Mackun & Baker 1990) found aphids (Aphididae) were significantly more abundant in plots cut first in July than those cut in June (4.5 vs. 3.7 aphids) at one site. Numbers of all other insects, including spittlebugs (Cercopidae), leafhoppers-planthoppers (Cicadellidae and Delphacidae) and mirids (Miridae), were not significantly different between plots cut in June and plots cut in July. Forage yields did not differ significantly between cutting treatments. Plots were cut on 15 June and 1 September, or 1 July and 1 September. There were four different pasture mixes of bird's-foot trefoil with one other pasture plant species (such as orchardgrass Dactylis glomerata) plus a bird's-foot trefoil monoculture treatment. Plots (11 x 5 m) were established in 1983. Insects were sampled seven times in 1984 and eight times in 1985, with five sweepnet samples/plot.Study and other actions tested
A replicated, randomised, controlled trial in 1985-1989 in Oxfordshire, UK (Gibson et al. 1992) found plots grazed in autumn-only had similar numbers of spider (Araneae) species and individuals (3.5-5.5 species, 69-197 individuals/m²) to those grazed in spring (3.3-5.3 species, 50-119 individuals) in 1989. An ungrazed control had 7.3-8.3 species and 111-207 individuals/m², while plots grazed in spring and autumn had lowest species richness and abundance (1.9-2.5 species, 16-51 individuals). Delaying mowing from spring to autumn did not have a clear effect on spider species richness or density in July over the three year period (autumn-only: 4.6-5.0 species, 69-99 individuals/m²; spring-only: 4.2-4.7 species, 50-99 individuals). The study took place in an ex-arable field (10 ha) and on old limestone grassland. In 1985, three treatments were applied (ungrazed, short-period spring or autumn sheep grazing) replicated six times in two square 3 x 3 grids of 30 x 30 m paddocks. Spring-and-autumn grazing was applied to larger areas outside the paddocks. Spiders were sampled by suction (using D-vac) and counting webs. Suction samples were taken in various months from May to October each year.Study and other actions tested
A trial in Dumfries, UK (Blake et al. 1996) found no detectable difference in the ground beetle (Carabidae) community between different cutting treatments on experimentally restored flower-rich grassland plots. A field was ploughed and sown with 17 plant species in August 1987 (five grasses, two clovers Trifolium spp. and 12 other flowering broadleaved species) and managed without fertilizers. Half the field was cut once each July. The other half was cut twice, in May and July. Both were grazed in autumn and winter. Ground beetles were sampled in 18 pitfall traps (laid out in two lines) in each treatment area, between April and September in 1989 and 1993.Study and other actions tested
A replicated, randomised, controlled study from 1987 to 1996 in Oxfordshire, UK (Bell et al. 2002) found that the predatory sheet web spider Lepthyphantes tenuis was approximately 2.5 and 6.5 times less abundant in cut versus uncut (control) field margins in May and July respectively. Following an early (April) cut, spider numbers in cut field margins recovered to match numbers in uncut control margins by July (around 4 spiders/m² in each). Recovery was less successful following a later (June) cut, with around 10 spiders/m² in cut field margins compared with around 15 spiders/m² in uncut control margins by September. Field margin treatment plots measured 2 x 50 m and were replicated around six arable fields. Spiders were counted in suction trap (D-vac) samples with data pooled from 1990, 1991, 1995 and 1996.Study and other actions tested
Referenced paperBell J.R., Johnson P.J., Hambler C., Haughton A.J., Smith H., Feber R.E., Tattersall F.H., Hart B.H., Manley W. & Macdonald D.W. (2002) Manipulating the abundance of Lepthyphantes tenuis (Araneae: Linyphiidae) by field margin management. Agriculture, Ecosystems & Environment, 93, 295-304.
A replicated, randomised, controlled trial in 2003-2005 on four farms in the southwest UK (Woodcock et al. 2007) (same study as Woodcock et al. 2009) found that 50 x 10 m plots of permanent pasture with delayed cutting (cut in July) had similar numbers of predatory beetles (Coleoptera) and slightly more seed- or flower-feeding beetles than plots cut in May. There were similar numbers of root or stem feeding beetles and foliage feeding beetles in plots cut in May and plots cut in July. Overall beetle numbers were similar between treatments, but there were slightly more beetle species in plots cut in July (30-38 species) than cut in May (27-34 species). The study also showed that reducing the management intensity on margins (by reducing or removing fertilizer, cutting and/or grazing) increased the fraction of seed- and flower-feeding beetles in the beetle community over the three years. The study tested seven treatments: cutting in May vs. July; cutting to 5 cm vs. 10 cm; grazing vs. no grazing; fertilizer vs. No fertilizer; and a treatment with no management. Treatments were replicated 12 times.Study and other actions tested
Referenced paperWoodcock B.A., Potts S.G., Pilgrim E., Ramsay A.J., Tscheulin T., Parkinson A., Smith R.E.N., Gundrey A.L., Brown V.K. & Tallowin J.R. (2007) The potential of grass field margin management for enhancing beetle diversity in intensive livestock farms. Journal of Applied Ecology, 44, 60-69.
A replicated, randomised, controlled trial in 2003-2005 on four farms in the southwest UK (Woodcock et al. 2009) (same study as Woodcock et al. 2007) found plots with delayed cutting (cut in July) had similar spider (Araneae), beetle (Coleoptera), true bug (Heteroptera), planthopper (Auchenorrhyncha), bumblebee (Bombus spp.) and butterfly (Lepidoptera) species richness to plots cut in May. Plots were 50 x 10 m on permanent pasture and were cut to 10 cm in either May or June. Each cutting treatment was replicated 12 times. Butterflies and bumblebees were monitored using transect walks, other invertebrates were monitored using a Vortis suction sampler.Study and other actions tested
Referenced paperWoodcock B.A., Potts S.G., Tscheulin T., Pilgrim E., Ramsey A.J., Harrison-Cripps J., Brown V.K. & Tallowin J.R. (2009) Responses of invertebrate trophic level, feeding guild and body size to the management of improved grassland field margins. Journal of Applied Ecology, 46, 920-929.
A replicated study in 1997-2005 at two pastureland sites at Pustnäs and Harpsund in southern Sweden (Lenoir & Lennartsson 2010) found that delaying the start of grazing had mixed effects on different groups of insects and spiders. Ground beetles (Carabidae) were found in higher numbers in late-grazed plots (2.0-5.4 beetles/trap) compared to continuously grazed plots (1.4-3.6 beetles/trap) at Pustnäs, while at Harpsund ground beetles were more abundant in continuously grazed pasture early in the season, but became more abundant in the late-grazed plot after grazing commenced. Spiders (Araneae) were more abundant in late-grazed plots at Pustnäs, but only until grazing started. At Harpsund, spider abundance was not affected by grazing, although some spider groups did show a response. Ant (Formicidae) numbers and diversity were higher in continuously grazed plots at Pustnäs until the start of grazing in late-grazing plots. At Harpsund there was no overall difference between treatments, although numbers of some individual species differed. The experiment used an enclosed 1 ha plot in a 2 ha pasture at Pustnäs and a 4 ha plot in a 12 ha pasture at Harpsund. The pastures were grazed from May to September with 1.2-1.8 cows/ha. Enclosed areas were ungrazed until late July.Study and other actions tested
A replicated, randomised, controlled study in 2003-2005 on four farms in the southwest UK (Blake et al. 2011) (part of the same study as Woodcock et al. 2007 and 2009) found similar combined numbers of planthoppers (Fulgoromorpha) and leafhoppers (Cicadomorpha) in pasture cut in July (averaging approximately 580 individuals/treatment) and cut in May (620 individuals). Planthopper and leafhopper species richness was also similar with 16.1 species in July-cut plots and 15.9 species in May-cut plots. More planthoppers and leafhoppers were found when pasture was cut only once (860 individuals/treatment), or not at all (595 individuals), compared with pasture cut twice (485 individuals). Cutting took place in July (a hay cut) or in May (a silage cut to 10 cm grass height) in permanent pastures. Plots were 50 x 10 m and treatments were replicated 12 times. Planthoppers and leafhoppers were collected in April, June, July and September in each year using a Vortis suction sampler, taking 75 ten-second suction samples/plot. June and September sampling occurred at least two weeks after cutting.Study and other actions tested