Providing evidence to improve practice

Action: Create beetle banks Farmland Conservation

Key messages

Supporting evidence from individual studies

1 

A replicated, randomized study in spring 1988-1990 on one beetle bank on a mixed farm in north Hampshire, UK (Thomas 1991) found weak evidence for a shift of predatory invertebrate activity from the beetle bank into the wheat crop over time. Individuals of the ground beetle (Carabidae) Demetrias atricapillus were more abundant on or very near the beetle bank in the first half of the study period (on average 12.2 individuals/m2 at 0-3 m distance from the beetle bank 14 April-3 May 1989) after which they were more evenly distributed (on average 0.4/m2 at 0-60 m from the beetle bank 8-22 May 1989) in 1989-1990. There was no consistent pattern in the distribution of the rove beetle (Staphylinidae) Tachyporus hypnorum in 1989-1990, although lower numbers were found on the beetle bank than in the crop by the end of the study in 1989. Money spiders (Linyphiidae) were more abundant on the beetle bank than in the crop, and significantly so for all but one sample date in 1989. There was evidence of some dispersal of money spiders and limited crop invasion by wolf spiders (Lycosidae) in spring 1990. The beetle bank was created in autumn 1986 through two-directional ploughing, it was 0.4 m high, 1.5 m wide and 290 m long. The bank crossed a 7 ha field with chalky-flint soil. The crop during the study was winter wheat in all years. The beetle bank was treated with a broad-spectrum herbicide in spring 1987 to remove broad-leaved herbs before the different treatments were hand-sown. Six replicates of each treatment (four single grass species, two mixes of three or four of the grass species, and bare ground) were created. Predation pressure was studied by placing dishes of prey at different distances along transects running from the beetle bank out into the crop (in 1988: 0, 1, 5 and 15 m, in 1989: 0, 3, 10, 30 and 60 m). The number of prey items remaining after 24 h was recorded. Dishes were active over one 24-h-period/week for seven weeks. Dispersal into the crop was studied by taking weekly vacuum-net samples along five of the same cock’s-foot Dactylis glomerata transects as above in 1989 (this part of the study is also reported in (Thomas et al. 1991)). In 1990, 10 perpendicular transects of barrier pitfall traps were placed at regular intervals along the beetle bank, avoiding non-grass treatments. Traps were placed 1, 4, 20 and 50 m into the crop, and set for one three-day-period each week and then emptied weekly throughout April and May. Vacuum-net samples were also collected in 1990 from five transects adjacent to five cock’s-foot plots but at the same distances as the barrier pitfall traps. This study was part of the same experimental set-up as (Thomas et al. 1991, Thomas et al. 1992, MacLeod 1994, MacLeod et al. 2004).

 

2 

A replicated, randomized and controlled study in the two winters of 1987-1989 at a mixed/arable farm in north Hampshire, UK (Thomas et al. 1991) found that two beetle banks sown with four different grass species (creeping bent Agrostis stolonifera, cock’s-foot Dactylis glomerata, Yorkshire fog Holcus lanatus and perennial rye grass Lolium perenne) produced densities of polyphagous invertebrate predators (invertebrates that feed on many different food sources) of up to 150 individuals/m2 in the first winter and over 1,500/m2 in the second winter. In the first winter, on bank 1 (in a 7 ha field) creeping bent held fewer predators (39/m2) than the other three grass species (66-102/m2), and similar numbers to bare ground (30/m2). In the second winter, Yorkshire fog held more predatory invertebrates (648-1,398/m2) than creeping bent (273-488/m2) and perennial ryegrass (276-394/m2) on both banks as well as cock’s-foot (218/m2) on bank 2 (in a 20 ha field) but not on bank 1 (cock’s-foot: 1,488/m2). In comparison, densities in the field were much lower (26-29/m2). In the second winter the two most abundant species were the ground beetle (Carabidae) Demetrias atriacapillus and the rove beetle (Staphylinidae) Tachyporus hypnorum. In spring 1989 D. atriacapillus occurred in higher numbers on or immediately adjacent to the banks up until 3 May (average density 12/m2 at 0-3m). After this date the distribution of this beetle throughout the field was more even (0.4/m2 at 0-60m). Significantly higher abundances of T. hypnorum occurred at 0 and 60 m into the field up until 18 April 1989 after which there were no consistent spatial patterns for this species, although there were lower numbers on the banks than in the field at the end of the study (22 May). The beetle banks were created in cereal fields on chalky-flint soil in autumn 1986 and treated with a broad-spectrum herbicide prior to hand-sowing in spring 1987. Six replicates of each treatment (four single grass species, two mixes of three or four of the grass species, and bare ground) were created. Predator communities were studied (November-February) through ground-zone searching in quadrats and destructive sampling (digging up turf samples) in the banks as well as mid-field. Crop penetration by emigrating predators was studied (once a week April-late May 1989) through transects of vacuum-net sampling at 0, 3 10, 30 and 60 m distance perpendicular to the cock’s-foot treatments on bank 1. This study was part of the same experimental set-up as (Thomas 1991, Thomas et al. 1992, MacLeod 1994, MacLeod et al. 2004).

 

3 

A replicated, randomized and controlled study in the three winters from 1987-1990 on two farms in Hampshire, UK (Thomas et al. 1992) (part of the same study as (Thomas et al. 1991) but extended with a third winter and a third beetle bank in a 51 ha field, also on chalky-flint soil, on a second farm) found that three years after beetle bank establishment, total predator densities and both ground beetle (Carabidae) and spider (Araneae) community compositions were not different to those in natural field boundaries. The tussock-forming grass, cock’s-foot Dactylis glomerata supported highest densities of ground beetles on all three beetle banks in the third winter. Community composition of both ground beetles and spiders changed significantly throughout the study in favour of species that prefer boundary or more permanent habitats. See (Thomas et al. 1991) for methods of beetle bank creation, experimental design and methods of predator sampling. This study was part of the same experimental set-up as (Thomas 1991, Thomas et al. 1991, MacLeod 1994, MacLeod et al. 2004).

 

4 

A replicated study on one beetle bank on a mixed farm in north Hampshire, UK (MacLeod 1994) found the densities of both ground beetles (Carabidae) and rove beetles (Staphylinidae) in four grass treatments showed two peaks in density over the study period (the seven winters of 1987-1988 to 1993-1994), in the second and sixth winter after establishment. The pattern was the same for spiders (Araneae) in cock’s-foot Dactylis glomerata, but in Yorkshire fog Holcus lanatus, creeping bent Agrostis stolonifera and perennial rye grass Lolium perenne the densities steadily increased and peaked in the fifth winter. Ground beetle densities over the seven year period in the different grass plots were as follows: cock’s-foot 11-110 individuals/m2, creeping bent 3-15/m2, perennial rye grass 2-11/m2 (only five winters), Yorkshire fog 1-76/m2. The respective rove beetle densities over the seven (or five) winters were: cock’s-foot 1-125 individuals/m2, creeping bent 0-67/m2, perennial rye grass 2-79/m2, Yorkshire fog 2-113/m2. Cock’s-foot and Yorkshire fog generally had the highest densities of predators but not always significantly so. The grass species composition in plots sown with cock’s-foot, Yorkshire fog and creeping bent remained relatively similar (min. 85% of original grass species left) during the study. Plots of false oat-grass Arrhenatherum elatius and red fescue Festuca rubra were created and added to the study in 1991. The 290 m-long beetle bank was created in spring 1987 and split into six blocks, each further sub-divided into eight plots with one treatment/plot. The eight treatments were sown cock’s-foot (3 g/m2), sown Yorkshire fog (4 g/m2), sown perennial rye grass (3 g/m2), sown creeping bent (8 g/m2), mix of three grass species (cock’s-foot, Yorkshire fog, perennial rye grass), mix of four grass species (previous three species plus creeping bent), bare ground, and sown flowering plants to provide pollen and nectar resources. Predatory invertebrates were sampled by taking two 20 x 20 x 10 cm turf samples/plot/winter. Percentage cover of grasses was measured in the four original grass treatments in October 1992 by placing six 25 x 25 cm quadrats in each grass plot, and in winter 1993-1994 it was measured in the collected turf samples. For methods in the first three winters see (Thomas 1991). This study was part of the same experimental set-up as (Thomas 1991, Thomas et al. 1991, Thomas et al. 1992, MacLeod et al. 2004).

 

5 

A replicated study in the winters of 1993-1994 to 1995-1996 on a lowland arable estate in Leicestershire, UK (Collins et al. 1996) (this study was continued in (Collins et al. 2003)) found that the average total predator, ground beetle (Carabidae) and rove beetle (Staphylinidae) (excluding aphid-specific species) density was higher in one hedge than one beetle bank over three winters. Out of five different grass species and areas of naturally regenerated vegetation, false oat-grass Arrhenatherum elatius, cock’s-foot Dactylis glomerata and timothy Phleum pratense held the highest densities of total predators, ground beetles and rove beetles on two other beetle banks. Beetle banks were 360-400 m long, 2-2.5 m wide, and sown in 1992-1993. Invertebrates were collected from soil samples using a cylindrical borer. This study was part of the same experimental set-up as (Moreby & Southway 2002, Murray et al. 2002, Bence et al. 2003, Collins et al. 2003).

 

6 

A replicated study in the summers of 1997-1998 in three regions (southern England, East Anglia and the Midlands) across the UK (Barker & Reynolds 1999) found no difference in the average catch of sawfly (Hymenoptera: Symphyta) larvae between beetle banks and grass strips planted along existing field margins. The total percentage cover of grass in planted grass strips affected the abundance of sawfly larvae positively. There were non-significant trends for sawfly larvae numbers to increase with strip age and to decrease with the amount of cock’s-foot Dactylis glomerata. Numbers of gamebird chick-food insects increased with strip age and area, but there was also a significant difference between farms. There was a non-significant trend for chick-food insect numbers to increase with the proportion of red fescue Festuca rubra. Cock’s-foot, red fescue and perennial rye grass Lolium perenne were the predominant grasses in most strips, being most common in 35, 25 and 17 strips respectively. A total of 116 grass strips (83 along pre-existing field margins and 33 beetle banks) on 32 farms were surveyed. For the overall analysis, the 11 strips on three of the farms were excluded. Grass strips had been established 0.5-12 years previously, both along pre-existing field margins and across cropped fields (beetle banks). Invertebrates were sampled by sweep-netting at the base of the vegetation in mid-June to mid-July. Percentage cover of all plant species and vegetation height was measured in 0.25m2 quadrats. Apart from where stated, this study does not distinguish between the effects of creating beetle banks and planting grass buffer strips/margins around arable or pasture fields.

 

7 

A 2000 literature review from the UK (Aebischer et al. 2000) found that populations of grey partridge Perdix perdix were 600% higher on farms with conservation measures aimed at partridges in place, compared to farms without these measures (Aebischer 1997). Measures included the provision of conservation headlands, planting cover crops, using set-aside and creating beetle banks.

Additional reference:

Aebischer N.J. (1997) Gamebirds: management of the Grey Partridge in Britain. Pages 131-151 in: M. Bolton (ed.) Conservation and the Use of Wildlife Resources. Chapman & Hall, London.

8 

A 2000 literature review (Holland & Luff 2000) looked at which agricultural practices can be altered to benefit ground beetles (Carabidae). It found three studies, two in the UK ((Thomas et al. 1991) (Collins et al. 1996)) and one in Denmark (a PhD thesis), showing higher ground beetle numbers in arable fields close to beetle banks.

 

9 

A replicated, paired, controlled study in the two winters of 1997-1999 and summer 1999 on five farm estates in the UK (Thomas et al. 2000) found different patterns of density and diversity for ground beetles (Carabidae), rove beetles (Staphylinidae) and spiders (Araneae) between five pairs of beetle banks and field margins in two consecutive winters. Rove beetle diversity was lower in beetle banks than in field margins in both winters, but density in beetle banks increased significantly between winters. There were no significant effects on ground beetles. The overall catch of chick-food invertebrates was lower in 22 beetle banks than in paired field margins on five farm estates, but the abundance of key prey groups was similar. There was no difference in grasshopper and bushcricket (Orthoptera) species richness between the two habitats (on average 1.4 species in beetle banks, 1.8 in field margins), but older beetle banks held higher abundances of grasshoppers and bushcrickets. Both abundance and species richness of butterflies and moths (Lepidoptera) was significantly lower in beetle banks than in field margins in June, July and August, but both habitats peaked in July. Destructive turf samples were collected randomly from the two habitats to assess predatory invertebrates. Chick-food invertebrates and grasshoppers and bushcrickets were sampled through sweep-netting and butterflies and moths through standard transect walks. This study was part of the same experimental set-up as (Thomas 2001, Thomas et al. 2001, Thomas 2002, Thomas et al. 2002).

 

10 

A replicated, controlled study in 1998 in two sites with autumn-sown crops on an estate in Hampshire, UK (Thomas 2001) found that boundary-overwintering ground beetle (Carabidae) species (species that migrate into fields in spring) were clustered near two beetle banks and a hedgerow in the early part of the season (March), after which activity-densities were more evenly spread until they clustered again later in the summer (July). The distribution of field-inhabiting species (species resident in fields year-round) was fairly uniform or more associated with the centre of the fields through the early part of the season. The two sites differed in the latter part of the season with one displaying a gappy distribution near the beetle bank, and the other clustering near the hedgerow and the beetle bank. The distribution of overwintering ground beetles in January was irregular within the beetle banks and the hedgerow, but there was no apparent pattern in distribution of active beetles from February to July. Two sets of ten transects (connected, paired pitfall traps at 5, 25, 50, 75, 100 and 150 m into the crop) were set up at each site. At site A, transects extended at 10 m intervals into the winter barley crop at right angles from both sides of a beetle bank sown with cock’s foot Dactylis glomerata. At site B, transects extended into the crop from one side of a beetle bank sown with cock’s-foot and from a hedgerow at the opposite side of the field, parallel to the bank, leaving a 50 m gap between traps at the furthest distance. Transects of pairs of unconnected pitfall traps were established within the beetle banks and the hedgerow. Pairs of traps were set at 10 m intervals and opened concurrently with the within-field traps for 72 h-periods March-July (A) or February-June (B). Fifteen 20 x 20 x 20 cm turf samples were removed from the beetle banks and the hedgerow in early January. This study was part of the same experimental set-up as (Thomas et al. 2000, Thomas et al. 2001, Thomas 2002, Thomas et al. 2002).

 

11 

A replicated, controlled study in 1998-1999 (winter-summer) on five farm estates in Hampshire and Wiltshire, UK (Thomas et al. 2001) found that diversity and average total abundance of chick-food invertebrates in sweep-net samples was higher in permanent field margins (65 individuals from 15 samples) than beetle banks (47 individuals from 15 samples) in 1999, and this was consistent between farms. In winter, the amount of plant litter, dead grass and tussocks that form important nesting material for game birds was higher in beetle banks (61%) than in field margins (27%), but overall vegetation cover in the two habitats was not different, and similar to that in summer (62-97%). Older beetle banks had higher diversity but not abundance of invertebrates. Invertebrate diversity also increased with plant diversity in both beetle banks and field margins. Invertebrate abundance and diversity was measured by vacuum suction-sampling and sweep-netting. Vegetation cover and composition was assessed with 0.71 m2 quadrats. Four to 22 banks/margins on one to five estates were included in the study in the two years. This study was part of the same experimental set-up as (Thomas et al. 2000, Thomas 2001, Thomas 2002, Thomas et al. 2002).

 

12 

A review (Evans et al. 2002) of two reports (Wilson et al. 2000, ADAS 2001) evaluating the effects of the Pilot Arable Stewardship Scheme in two regions (East Anglia and the West Midlands) from 1998 to 2001 found that grass margins benefited plants, bumblebees Bombus spp., bugs (Hemiptera) and sawflies (Symphyta), but not ground beetles (Coleoptera). The grass margins set of options included sown grass margins, naturally regenerated margins, beetle banks and uncropped cultivated wildlife strips, but the review does not distinguish between these different options. None of the beneficial effects were pronounced on beetle banks. The effects of the pilot scheme on plants, invertebrates (bumblebees, true bugs, ground beetles, sawflies) were monitored over three years, relative to control areas. Grass margins were implemented on total areas of 361 and 294 ha in East Anglia and West Midlands respectively.

Additional references:

Wilson S., Baylis M., Sherrott A. & Howe G. (2000) Arable Stewardship Project Officer Review. F. a. R. C. Agency report.

ADAS (2001) Ecological evaluation of the Arable Stewardship Pilot Scheme, 1998-2000. ADAS report.

13 

A replicated study in June 2000 in ten edge habitats on a lowland arable farm in Leicestershire, England (Moreby 2002) found that beetle banks contained the highest density of sawfly (Symphyta) larvae, significantly higher compared to hedge bottoms and winter wheat headlands, but not compared to grass/wire fence lines or edges of un-grazed pasture. Spider (Araneae) and rove beetle (Staphylinidae) densities were lower in beetle banks than in un-grazed pastures. Set-aside contained a higher density of weevils (Curculionidae) than beetle banks. There was no difference in ground beetle (Carabidae) or caterpillar (Lepidoptera) densities between habitats. Type of neighbouring crop did not affect invertebrate densities in the different habitats. Apart from the six habitats mentioned above, brood cover, one and two-year-old wild bird cover, and sheep-grazed pasture edges were included in the study. Invertebrates were sampled with a vacuum suction sampler in June 2000.

 

14 

A replicated study from 1995 to 1999 of arable habitats on a farm in Leicestershire, UK (Moreby & Southway 2002) found that the abundance of some invertebrate groups was higher in non-crop strips (grass beetle banks or wild bird cover), whereas other groups were more abundant in crops. Four invertebrate groups tended to have significantly higher densities in non-crop strips than crops in all years: spiders (Araneae) 7 vs 1-5 individuals/sample, true bugs (Homoptera) 29 vs 1-4, typical bugs (Heteroptera) 10-58 vs 0-9, and key ‘chick food insects’ 65 vs 2-10. In three of the years, true weevils (Curculionidae) were found at significantly higher densities in non-crop strips and beans (0-11) than other crops (0-2). In contrast, in three or four of the years, densities in crops were significantly higher than non-crops for: true flies (Diptera) 20-230 vs 25-100 individuals and aphids (Aphididae). Moth and butterfly larvae (Lepidoptera) and ground beetles (Carabidae) differed significantly in only one or two years, when density was higher in crops than non-crops. Total beetles (Coleoptera) varied between years and habitats. Sawfly larvae (Symphyta), leaf beetles (Chrysomelidae) and soldier beetles (Cantharidae) showed no significant differences. Grass strips (1 m-wide) planted as beetle banks were sown onto a raised bank along edges or across the centre of fields. Wild bird cover was sown as 2-5 m-wide strips along field boundaries and re-sown every few years with a cereal or kale-based mixture. Invertebrates were sampled each year in the centre of 5-11 grass/wild bird cover strips and 3 m into 3-4 pasture, 8-12 wheat, 6-8 barley, 3-6 oilseed rape and four field bean fields. Two samples of 0.5 m² were taken in each habitat using a D-Vac suction sampler in June 1995-1999. This study was part of the same experimental set-up as (Collins et al. 1996, Murray et al. 2002, Bence et al. 2003, Collins et al. 2003).

15 

A study of different set-aside crops on an arable farm in Leicestershire, UK (Murray et al. 2002) found that Eurasian skylark Alauda arvensis, but not yellowhammer Emberiza citrinella, used beetle banks more than expected compared to availability. Skylarks used beetle banks (planted tussocky perennial grasses) more than expected compared to availability and significantly more than unmanaged set-aside, broad-leaved crops and other habitats. Yellowhammer used beetle banks as expected compared to availability but significantly less than cereal and wild bird cover cereal set-aside. Field margin and midfield set-aside strips were sown with kale-based and cereal-based mixtures for beetle banks and wild bird cover. Other habitat types were: unmanaged set-aside, cereal (wheat, barley), broad-leaved crop (beans, rape) and other habitats. Thirteen skylark and 15 yellowhammer nests with chicks between 3-10 days old were observed. Foraging habitat used by the adults was recorded for 90 minutes during three periods of the day. This study was part of the same experimental set-up as (Collins et al. 1996, Moreby & Southway 2002, Bence et al. 2003, Collins et al. 2003).

16 

A small replicated controlled study from May-June 1992-1998 in Leicestershire, UK (Stoate 2002) found that the abundance of nationally declining songbirds and bird species of conservation concern significantly increased on a 3 km2 site where beetle banks were created (alongside several other interventions), although there was no overall difference in bird abundance, species richness or diversity between the experimental and three control sites. Numbers of nationally declining species rose by 102% (except for Eurasian skylark Alauda arvensis and yellowhammer Emberiza citrinella). Nationally stable species rose (insignificantly) by 47% (eight species increased, four decreased). The other interventions employed at the same site were managing hedges, wild bird cover strips, supplementary feeding, predator control and reducing chemical inputs generally.

 

17 

A replicated, paired, controlled study on five conventional arable estates in Hampshire and Wiltshire, UK (Thomas 2002) found that ground beetle (Carabidae) population patterns and vegetation composition in beetle banks and field margins changed across seasons. In winter there was no difference in ground beetle density (range: about 200-300/m2), species richness (range: 15-22 species in total) or diversity between beetle banks and field margins, but species richness increased with age in beetle banks. Ground beetle density and species diversity was higher in beetle banks than field margins in both spring and summer (beetle banks had on average about 75 individuals/m2 in spring and ca. 90/m2 in summer while field margins had about 45 and 60/m2 in each season respectively). Only eight sites were included in the spring analysis. Ground beetle species composition was similar in the two habitats during winter and summer. The winter catches contained especially large proportions of Bembidion lampros. In spring the species composition was different with far fewer B. lampros and more larvae (not identified to species). Total plant cover was high in both habitats in both seasons but significantly higher in field margins during summer. However cover of tussocky grasses was higher in beetle banks in both seasons and did not decline with bank age. Field margins had higher species diversity in summer and higher species richness in both seasons compared with beetle banks. Both measures however increased with beetle bank age so that older banks had a similar number of species to margins. A total of 22 beetle banks were included in this study, ranging from < 1 to 14 years old, each paired with a conventional permanent margin in the adjacent field. Ground beetle populations were sampled in four periods (winter (January-February), spring (May), summer (August) and winter (February)) through destructive sampling (vacuum suction-sampling and digging up turf samples). Vegetation composition was investigated in winter (January-February) and summer (July) through quadrats placed on the ground. This study was part of the same experimental set-up as (Thomas et al. 2000, Thomas 2001, Thomas et al. 2001, Thomas et al. 2002).

 

18 

A replicated, paired, controlled study in the summers of 1998-1999 and late winter 1998-1999 on five farm estates in southern UK (Thomas et al. 2002) found lower summer plant species richness and diversity in beetle banks compared with conventional arable field margins. Both measures increased with age of beetle banks in summer. Beetle banks had higher cover of tussock-forming and other grasses, but lower cover of herbaceous, woody and nectar-providing plants in the second summer. In winter there was no difference in overall plant cover between beetle banks and field margins but average species richness was lower in beetle banks. Species richness increased with age in beetle banks. There was no relationship between tussock cover and age of bank in winter. Beetle banks, aged <1-14 years, were sown mainly with cock’s-foot Dactylis glomerata and had received no, or little, active management since establishment. Percentage cover of all plant species was assessed in twenty 0.5-m2 quadrats along each bank or margin in July 1998, January-February and July-August 1999. This study was part of the same experimental set-up as (Thomas et al. 2000, Thomas 2001, Thomas et al. 2001, Thomas 2002).

 

19 

A controlled study in autumn 1998 on a predominantly arable farm in Leicestershire, UK (Bence et al. 2003) found overall more harvest mouse Micromys minutus nests in beetle banks (117 nests/ha) than field margins (14 nests/ha) although this difference was not statistically tested. Beetle banks were created and sown with grasses such as cock’s-foot Dactylis glomerata between September 1992 and 1994 and cut regularly in the year of establishment. Field margins were often adjacent to a hedgerow and normally left uncut. The two habitats were hand searched for harvest mouse nests in September to November, in a total of 1.8 km of beetle banks and 9.8 km of field margins. This study was part of the same experimental set-up as (Collins et al. 1996, Moreby & Southway 2002, Murray et al. 2002, Collins et al. 2003).

 

20 

A replicated study in 1994-1998 including two beetle banks on an arable estate in Leicestershire, UK (Collins et al. 2003) (a continuation of (Collins et al. 1996)) found higher densities of invertebrate predators in false oat grass Arrhenatherum elatius (2,045/m2) than in red fescue Festuca rubra (1,492/m2), crested dog’s-tail Cynosurus cristatus (1,380/m2) and naturally regenerated vegetation (1,060/m2). Rove beetles (Staphylinidae), were the dominant family in the predatory invertebrate catch, and showed the same significant pattern (1,716/m2 in false oat grass, 1,241/m2 in red fescue, 1,105/m2 in crested dog’s tail and 834/m2 in naturally regenerated vegetation). Spider (Araneae) density was higher in cock’s-foot (177/m2) compared with red fescue (119/m2) and naturally regenerated vegetation (107/m2). Ground beetle (Carabidae) density was 2.5 to 3.5 times higher (significant) in cock’s-foot than all other treatments. Boundary-type ground beetles dominated all treatments but were also higher in cock’s-foot (328/m2) compared with the other five treatments (69-126/m2). In the first year of the study (third summer after creation) all single grass treatments were dominated by their sown species. In the last year of the study false oat-grass had the highest cover (90%) followed by red fescue (75%), cock’s-foot and timothy (70%), and crested dog’s-tail (10%). Overall, cock’s-foot, false oat-grass and timothy were taller growing and formed denser grass coverage near ground level (0-30 cm) compared with the other treatments. Beetle banks were created in spring 1993, both situated in an 8.6 ha clay soil field. Six treatments (five species of grass and naturally regenerated vegetation) were established with two replicates/bank. Invertebrates were collected from soil samples gathered in January-February 1994-1997. Vegetation was examined visually as well as measured with a graduated board. This study was part of the same experimental set-up as (Collins et al. 1996, Moreby & Southway 2002, Murray et al. 2002, Bence et al. 2003).

21 

A randomized, replicated study over seven winters from 1987-1988 to 1993-1994 within one beetle bank on a mixed arable estate in Hampshire, UK (MacLeod et al. 2004), (an extension of (Thomas 1991)) found that ground beetle (Carabidae) and rove beetle (Staphylinidae) densities were in general highest in cock’s-foot Dactylis glomerata and Yorkshire fog Holcus lanatus respectively, although this was not always significantly higher in comparison with creeping bent Agrostis stolonifera or perennial rye grass Lolium perenne. Densities of money spiders (Linyphiidae) and wolf spiders (Lycosidae) were also higher, although not always significantly, in these two tussock-forming grasses. The ground beetle species composition changed from dominance by open field species to boundary species over the course of the study. In the last three winters, when sampled, field boundaries had lower densities of predatory invertebrates than the beetle bank, but this was not statistically tested. Percentage cover of  the grass species originally sown in plots remained high for all species, except perennial rye grass, plots of which were invaded by cock’s-foot by the sixth winter and excluded from sampling in the last two winters because perennial rye grass had become so rare. One beetle bank was created through two-directional ploughing and divided into six blocks in which eight treatments were sown (only the four single grass species treatments included in this study) in randomized order. Predator communities were sampled through ground-zone searching and destructive sampling November-February. Vegetation composition was examined in quadrats in October 1992. This study was part of the same experimental set-up as (Thomas 1991, Thomas et al. 1991, Thomas et al. 1992, MacLeod 1994).

 

22 

A replicated study in 1999 and 2003 on 256 arable and pastoral fields across 84 farms in East Anglia and the West Midlands, UK (Stevens & Bradbury 2006) found that out of 12 farmland bird species, none were strongly associated (either positively or negatively) with beetle banks. The species analysed were skylark Alauda arvensis, corn bunting Miliaria calandra, lapwing Vanellus vanellus, yellow wagtail Motacilla flava, chaffinch Fringilla coelebs, dunnock Prunella modularis, greenfinch Carduelis chloris, Eurasian linnet C. cannabina, reed bunting Emberiza schoeniclus, tree sparrow Passer montanus, whitethroat Sylvia communis and yellowhammer E. citrinella.

23 

A 2007 UK literature review (Stoate & Moorcroft 2007) describes a study which found that beetle banks held higher densities of harvest mouse Micromys minutus nests than field margins. Other studies found that grey partridge Perdix perdix and Eurasian skylark Alauda arvensis also nested in beetle banks. Skylarks were found to be more likely than yellowhammer Emberiza citrinella to forage in beetle banks. However, a study in Leicestershire, UK, found that lesser marsh grasshoppers Chorthippus albomarginatus did not use two species of plant commonly planted in beetle banks (cock’s-foot Dactylis glomerata and false oat grass Arrhenatherum elatius) as food plants.

 

24 

A replicated site comparison study from 2004 to 2008 in England (Ewald et al. 2010) found that grey partridge Perdix perdix overwinter survival was significantly and positively correlated with the presence of beetle banks in 2007-2008. Across all years there was a positive relationship with the ratio of young to old birds. There were no relationships between beetle banks and  brood size or year-on-year density changes. Spring and autumn counts of grey partridge were made at 1031 sites across England as part of the Partridge Count Scheme.

 

Referenced papers

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. (2017) Farmland Conservation Pages 245-284 in: W.J. Sutherland, L.V. Dicks, N. Ockendon & R.K. Smith (eds) What Works in Conservation 2017. Open Book Publishers, Cambridge, UK.