Providing evidence to improve practice

Action: Use grazing instead of cutting for pasture or grassland management Natural Pest Control

Key messages

Natural enemies: Two studies (one before-and-after and one replicated trial) from Australia and the UK found grazing instead of cutting had mixed effects on natural enemies, with some species and groups affected on some dates but not others. One replicated study from New Zealand found no effect.
Pests and diseases: One of five studies (including three replicated trials) from Australia, New Zealand, the UK and the USA found more pests, and two studies found effects varied between pest groups and sampling dates. Two studies found no effect on pests. One study found no effect on disease when grazing was used in addition to cutting.
Pasture damage and plant survival: One randomised study found more ryegrass shoots were attacked by pests. One study found lower survival of alfalfa plants but another found no effect.
Yield: One of four randomised, replicated studies (one also controlled) found lower yields and two found no effect. One study found lower ryegrass and higher clover yields, but no difference between clover varieties. Another randomised study found more ryegrass shoots.

Crops studied were alfalfa, cock’s-foot, perennial ryegrass, other grasses and white clover.

Supporting evidence from individual studies


A before-and-after trial in 1979 in lucerne Medicago sativa in New South Wales, Australia (Bishop et al. 1980) found a smaller reduction in predatory adult brown lacewings Micromus sp. after grazing (57% decline, from 8.6 to 3.7 adults/2 m²) than after cutting (91-94% decline, 5.1-7.0 to 0.3-0.6 adults). Numbers remained higher in grazed than cut lucerne seven days after treatment. Brown lacewing larvae declined by 82% in grazed compared to 98% in cut lucerne. Grazing and cutting caused similar declines for transverse ladybirds Coccinella transversalis (68% vs. 78-83%, respectively). Blue-green aphids Acyrthosiphon kondoi Shinji declined less under grazing (70% decline, from 93 to 28 aphids/3.1 m²) than cutting (89-90% decline, 66-113 to 6.6-12.1 aphids) but numbers were similar after seven days. Mowing, windrowing (piling cut vegetation in rows on the field) and baling lucerne before collection had little effect on pest or natural enemy numbers compared to harvesting directly into a trailer. Treatments included grazing (84 cattle for 1 day on 0.45 ha), cutting with a forage harvester and collecting the crop immediately (0.3 ha), and mowing and windrowing before baling and collection (0.3 ha). Aphids and predators were sampled with a suction sampler at 10 random quadrat sites/treatment.


A replicated trial in perennial ryegrass Lolium perenne and white clover Trifolium repens pasture in 1976-1977 in County Kildare, Ireland (Purvis & Curry 1981) found that effects of grazing vs. cutting varied between invertebrate groups and sampling dates. Fewer spiders (Araneae) occurred in continuous, lightly grazed (2-105 spiders) or intermittent, heavily grazed (2-121 spiders) plots than in cut plots (10-429 spiders/suction sample) for seven of eleven months. Wasps (Hymenoptera) showed mixed effects with fewer in continuous, lightly grazed than cut plots for three months, but the opposite for one month and no difference for seven months. In total, fewer small invertebrates occurred in grazed (13,120-17,750 invertebrates) than in cut (17,800-21,050 invertebrates/suction sample) plots during peak abundance in July-August 1977, but numbers were similar after cutting took place in September. The treatments included continuous grazing with 10-30 sheep/ha, intermittent grazing with 60-100 sheep/ha for 1-2 week periods, grass cut twice a year for silage, and two treatments combining cutting and grazing. Each treatment was tested in two 0.2 ha plots. Plant-dwelling invertebrates were sampled using a D-vac suction net in ten areas/plot (each measuring 0.09 m²). Most natural enemy and pest groups were not differentiated.


A randomised experiment in 1980-1982 in Berkshire, UK (Moore & Clements 1984) found more stem-boring fly (Oscinella spp. and Geomyza tripuncta) larvae in plots of grazed vs. cut perennial ryegrass Lolium perenne (reaching peaks of approximately 3,370-5,740 vs. 985-1,770 larvae/m²) during summer and winter. Numbers were similar during late spring when adults emerged and larvae were scarce. Peak numbers of adult female flies were also higher in grazed vs. cut (approximately 165-590 vs. 75-150 flies/treatment) plots in both years. The study reported that more perennial ryegrass shoots were attacked by fly larvae in grazed (11-13%) than cut (7%) plots but more grass shoots also occurred in the former than the latter (45,000 vs. 33,000 shoots/m² respectively, at peak numbers). Three plots were sheep-grazed at 28 day intervals (beginning March 1980) and each grazing event used 20 sheep for 24 hours. Three other plots were cut with a Mayfield autoscythe on the same dates as grazing events and cut material was removed. Plots were 10 x 10 m. Fly larvae were counted by dissecting grass shoots, sampled using 50 mm-diameter turf cores (five per plot) on 26 occasions over two years. Effects on natural processes of pest control were not presented.


A randomised, replicated experiment in 1977-1981 involving 22 lucerne Medicago sativa varieties in New South Wales, Australia (Lodge 1985) found lower yields when lucerne was grazed (average total 16,230 kg/ha over four years) rather than cut (30,893 kg/ha). Another experiment testing seven varieties found the same effect (9,740 vs. 19,122 kg/ha in grazed vs. cut over 3.5 years). The number of lucerne plants in grazed and cut plots declined by 89% and 51% (respectively) over four years in the first experiment and by 82% and 39% over 3.5 years in the second. Lucerne varieties that are active in winter performed better than dormant varieties when both types were grazed (9,887-11,611 vs. 7,638-9,991 kg/ha yields), but yields were similar for these varieties in cut plots (16,067-21,213 vs. 19,040-20,954 kg/ha) in the second experiment. All lucerne varieties were tested in four replicate plots (10 x 2 m) divided into grazing areas of 8 x 2 m (grazed with 85 Merino sheep/ha) and cutting areas of 2 x 2 m. Grazing occurred approximately every 6 weeks and for 4-29 days each time (202-287 days in total). Yields were measured as dry vegetation matter. Effects on natural processes of pest control were not presented.


A randomised, replicated, controlled experiment in 1990-1992 on a pasture in Essex, UK (Vickery et al. 1994) found similar grazing intensities of brent geese Branta bernicla (pests) on sheep-grazed plots (averaging 31.6-39.5 total goose droppings/m²/winter), cut and grazed plots (28.2-36.4 droppings), and cut-only plots (28.5-36.8 droppings). The amount of vegetation was similar between grazed (223-236 g dry weight/m²), cut and grazed (195-255 g/m²) and cut-only plots (188-232 g/m²). In another randomised, replicated, controlled experiment, grazing intensities of brent geese were similar in sheep-grazed (59.6 total droppings/m²) and cattle-grazed (60.2 droppings/m²) plots. In the first experiment, grazed plots contained sheep in April-May or June and July-September and grazing intensities varied from 13.5-92.2 livestock unit days. Cut and grazed plots were cut on 26 June then grazed for one or two one-month periods. Cut plots were cut in late June and late August. Each treatment was replicated six times in 100 x 75 m plots. In the second experiment six plots (of 50 x 50 m) were grazed by 14 cattle and six plots were grazed by 6-11 sheep in June-August. Goose droppings were monitored in sample areas (with 1.5 m-radiuses) at 5 and 10 random points/plot (first and second experiments, respectively).


A randomised, replicated experiment in 1986-1988 in Wyoming, USA (Gray & Koch 2004) found similar alfalfa Medicago sativa yields in plots cut twice and grazed (2.6-9.8 Mg/ha) compared to plots cut three times but not grazed (2.8-9.9 Mg/ha). Plant density at the end of the experiment (1988) was similar in plots cut twice and grazed (47.5% plants remaining) and plots cut three times (43.8%). Grazing reduced yields when used in addition to cutting, for example in 1988 plots cut twice and grazed yielded 2.55 Mg/ha compared with 2.96 Mg/ha in plots cut twice only, while plots cut three times and grazed yielded 2.31 Mg/ha compared with 2.78 Mg/ha in plots cut three times only. In another experiment, grazing in addition to cutting did not affect Verticilium wilt severity (caused by Verticillium albo-atrum), alfalfa yield or plant density in wilt-resistant and wilt-susceptible alfalfa varieties. The first experiment compared plots cut twice, cut twice and grazed in autumn, cut three times, and cut three times and grazed in autumn. Each treatment was replicated four times in 3.7 x 3.7 m plots. Plots were grazed after the first autumn frost (5 cows/ha). The second experiment tested the same treatments plus two alfalfa varieties.


A randomised, replicated experiment in 2001-2004 in a mixed perennial ryegrass Lolium perenne and white clover Trifolium repens pasture in Aberystwyth, UK (Williams et al. 2007) found higher perennial ryegrass yields under grazing (770-2,312 kg/ha in 2002-2003) compared to cutting (171-1,083 kg/ha) regimes on most dates in two experiments. White clover yields were lower under grazing (111-1,352 kg/ha) than cutting (247-1,430 kg/ha) regimes on most sampling dates, but total yields (ryegrass and clover) were higher with grazing. Grazing did not improve the performance of a nematode-resistant white clover variety compared to a conventional variety (yields of 94-1,412 vs. 98-1,266 kg/ha respectively when grazed; 512-1,442 vs. 264-1,334 kg/ha when cut). Two experiments tested the effects of sheep grazing (April-October) vs. cutting six times/year, as well as using two white clover varieties (both individually and mixed together). One experiment was conducted under natural conditions (results were not presented) while the other supplemented the pasture with plants artificially infested with pest stem nematodes Ditylenchus dipsaci. Twelve plots of 5 x 4 m were subdivided into grazing (3.5 x 4 m) and cutting (1.5 x 4 m) areas. Yield samples measuring dry vegetation matter were taken on all cutting dates for three years (2002-2004).


A replicated study in 2002-2007 in Taranaki, New Zealand (Schon et al. 2010) found similar numbers of predatory and omnivorous (plant and animal-eating) nematodes (Nematoda) in grazed (approximately 6,000-30,000 individuals/m²) and cut (10,000-50,000 individuals) pasture. Numbers of small predatory invertebrates, including mites (Acari), beetles (Coleoptera), spiders (Araneae) and other groups, were also similar in grazed vs. cut plots (4,000-20,000 vs. 7,000-24,000 individuals/m²). Numbers of plant-feeding or plant-parasitic nematodes were similar between grazed vs. cut plots, for example 7,700-53,900 vs. 9,000-23,400 Pratylenchus spp. individuals/m² and 0-99,200 vs. 4,700-13,000 Meloidogyne spp. juveniles/m² in up to 10 cm-deep soil samples. Numbers of plant-eating small invertebrates, including mites, springtails (Collembola), beetle larvae, moths and butterflies (Lepidoptera), were also similar in grazed vs. cut plots (600-5,500 vs.600-7,000 individuals/m²). Grazed plots were stocked at 3, 4 or 5 cows/ha. Pasture was mown and vegetation was removed in cut plots. Each treatment was applied to four 0.1 ha plots. Measurements were taken in 2007 after five years of treatment. Invertebrates were sampled using soil cores (up to 15.5 cm deep) in autumn and winter. Natural enemies and pests were not differentiated in many invertebrate groups.

Referenced papers

Please cite as:

Wright, H.L., Ashpole, J.E., Dicks, L.V., Hutchison, J., McCormack, C.G. & Sutherland, W.J. (2017) Some Aspects of Enhancing Natural Pest Control. Pages 359-381 in: W.J. Sutherland, L.V. Dicks, N. Ockendon & R.K. Smith (eds) What Works in Conservation 2017. Open Book Publishers, Cambridge, UK