Rewilding aims to restore ‘self-willed’ ecosystems involving the creation of habitats subject to stochastic disturbance connected by favorable corridors for dispersal of animals, including insects. Reversion of arable land to grassland and scrub habitats adjacent to Arger Fen nature reserve in Suffolk (southeast England) through non-intervention allowed succession to occur largely unmanaged on fields with differing topography, from flat terrain to slopes. Monitoring of Orthoptera revealed statistical evidence that species diversity and richness was greater on the steeper slopes (gradient > 10%), while species varied in their topographic preferences from flat terrain (e.g., long-winged conehead Conocephalus fuscus Fabricius, 1793) to slopes (e.g., field grasshopper Chorthippus brunneus Thunberg, 1815). Lagomorph grazing by the wild brown hare Lepus europaeus (Pallas, 1778) and the rabbit Oryctoloagus cuniculus (Linnaeus, 1758) appeared to be critical in maintaining exposed soil for hillside species such as C. brunneus, which may require the egg-laying and basking habitat. A mosaic of scrub and grassland on a wooded hillside affected by ash dieback Hymenoscyphus fraxineus (Baral et al. 2014) was also important for Orthoptera. We postulate that rewilding schemes on arable land may be particularly effective when there are topographic undulations incorporating flat and hillside areas to promote the greatest diversity of Orthoptera.

Acrididae, bush-cricket, grasshopper, landscape, Tettigoniidae, topography, wilding

The term rewilding was first used in the 1980s (Noss 1985). Soulé and Noss (1998) proposed three key components of rewilding: large core protected areas, ecological connectivity, and keystone species that translated to the 3Cs of cores, corridors, and carnivores. Over time, Soulé and Noss’s original concept has shifted into local interpretations but still incorporates self-regulatory ecosystems with minimal or no anthropogenic influence where wild grazers have a critical role (Dempsey 2021).

In recent times, the aim of rewilding in Europe has focused on restoring natural processes by creating large areas of habitat subject to stochastic disturbances connected by favorable corridors for species to disperse along (van Klink et al. 2020, Carver and Convery 2021, Gordon et al. 2021a, b). Hart et al. (2023) provide an overview of the published studies relating to rewilding schemes in Europe, noting that it can take decades if not centuries to have the desired benefits to target species (e.g., Chernobyl Exclusion Zone rewilding site; Dombrovski et al. 2022).

The European Green Belt initiative, first discussed in 1989 after the fall of the Berlin Wall, led to the establishment of a 12,500 km long and 50 km wide green corridor from Norway to Greece following the former Iron Curtain (Fraser 2009). The development of naturally vegetated corridors benefits not only apex predators (e.g., the Balkan lynx Lynx lynx balcanicus Bures, 1941 and the brown bear Ursus arctos Linnaeus, 1758 (Sterr et al. 2012)), but also beetles (Coleoptera), flies (Diptera), and some moths (Lepidoptera) of woodland habitats (Thomas et al. 1994, Merckx 2015). Pollinator activity and abundance in horse-grazed areas increased in a Swedish rewilding scheme (Garrido et al. 2019, 2021, 2022), suggesting that introduced herbivores play a role in maintaining rewilded habitats for early successional invertebrates. In the Czech Republic, a 26-year experiment revealed that bison, pony, and cattle grazing led to a higher diversity and abundance of butterflies (Konvička et al. 2021). In the UK, reintroduced beavers (Castor fiber Linnaeus, 1758) act as keystone herbivores in rewilding schemes by modifying riparian habitats on farmland, creating ‘beaver meadows’ (Stringer and Gaywood 2016). These structurally diverse wet meadows can form quickly in response to dammed rivers and an absence of active grassland management (e.g., at Spains Hall in Finchingfield, south-east England (Essex County Council 2024)).

Ecological restoration that allows habitats to regenerate with a lack of active agricultural (e.g., fertilizer application) or conservation management, such as controlled livestock grazing, is known as rewilding max, a passive strategy (Gordon et al. 2021a, b). Domestic livestock (cattle, sheep, and ponies) are often used to graze rewilded farmland sites (e.g., Knepp Wildland in West Sussex, UK (Dempsey 2021)) after the initial establishment phase and grassland re-establishment (Casey et al. 2020). This form of conservation intervention without sole reliance on natural grazers is known as rewilding lite (Gordon et al. 2021b). However, introduced livestock can have detrimental impacts on Orthoptera when stocking density (number of animals) is too high and the resultant sward height is too short (Gardiner and Haines 2008). The consequences of livestock grazing are largely influenced by the intensity of grazing, type of grazer, and rotational or seasonal aspects of the regime, which in turn have an impact on the characteristics of grasslands, such as leaf litter development, plant species presence, sward height, and vegetation biomass (Marini et al. 2008, Fonderflick et al. 2014, Rada et al. 2014).

Rewilding and Orthoptera.—To reverse the decline of insects such as grasshoppers and bush-crickets, the rewilding of arable land may be highly beneficial (Tree 2017, van Klink and WallisDeVries 2018, Garrido et al. 2022). The ideal aim of rewilding is to restore natural processes, often involving the creation of large areas of habitat with stochastic disturbance connected by favorable corridors for species to disperse along (Carver and Convery 2021, Gordon et al. 2021a, b). Rewilding is well established at some UK sites, such as Knepp in West Sussex (Greenaway 2006, Tree 2017, Wallace 2018, Dempsey 2021), Spains Hall in Essex (Essex County Council 2024), and two sites in Suffolk: Black Bourn Valley and the Somerleyton Estate (Gardiner and Casey 2022a, b). Research at Black Bourn Valley revealed that Orthoptera were found in greatest abundance and diversity in fields > 8 years post-arable cropping cessation (Gardiner and Casey 2022a, b, Broad 2023). The Gardiner and Casey studies noted the importance of wild lagomorph grazing, anthills, and pond edge habitat, particularly for species such as the field grasshopper Chorthippus brunneus (Thunberg, 1815), which require exposed soil (Waloff 1950).

Key aspects of topographic heterogeneity.—Given their importance, various studies have addressed the effects of the topography elements of aspect, elevation, and slope gradient on Orthoptera. Landscape heterogeneity is an important factor in insect ecology (Li et al. 2022), such as for butterflies (Kumar et al. 2009, Beirão et al. 2021) and Orthoptera (Batáry et al. 2007, Marini et al. 2008, Rada et al. 2014, Joubert-Van der Merwe and Pryke 2018). Topography of landscapes is characterized by aspect, elevation, and slope gradient (Li et al. 2011) and constitutes a critical influence on environmental diversity affecting biotic (e.g., taxonomic richness and diversity, population structure, and animal dispersal) and abiotic (e.g., soil structure, micro/macroclimate, and hydrological processes) factors (Zhang et al. 2017, Amatulli et al. 2018). For example, plant-level scale microtopography (1 cm to 1 m) incorporates the same elements as broader landscape topography, including aspect and slope gradient (Diefenderfer et al. 2018). When occurring in any number, anthills at rewilding sites provide significant microtopographic heterogeneity due to the basking and oviposition opportunities provided by the slopes (King 2006, 2020, 2021).

i) Aspect.—Orthoptera are affected by significant changes in elevation and microclimatic variations due to north- and south-facing slopes (Voisin 1990, Weiss et al. 2013). South-facing slopes are particularly species rich and important for grasshoppers (Gardiner 2011). In these scenarios, a hot ‘microclimate’ with high orthopteran abundance can develop (Marshall and Haes 1988). A study of Furze Hills in southeast England found significantly more grasshopper nymphs and C. brunneus adults on the south-facing slope (Gardiner 2022), while a southern aspect appeared to promote higher orthopteran abundance and species richness on Hungry Hill at Lound Lakes (Gardiner 2021).

ii) Elevation (altitude).—Elevation, measured in meters above mean sea level (masl), is a common aspect of Orthoptera studies, particularly for species at high altitudes. Studies often record a decline in orthopteran abundance and species richness with increasing elevation (e.g., Gebeyehu and Samways 2006, Pitteloud et al. 2020), although not exclusively so (see Azil and Benzehra 2020). High-altitude species have adaptations for survival in the colder climate. For example, it appears that egg hatching and adult maturity of the common green grasshopper Omocestus viridulus (Linnaeus, 1758) in the Alps (>2000 masl) are reached much later at high elevations than at low elevations (Berner 2005). Hatching and maturation can also be severely affected by cloudy weather (low amount of sunshine) at high elevations, with cold weather affecting reproductive success (Berner et al. 2004). However, high altitude O. viridulus nymphs may have a much shorter period from egg hatching to adult (quicker nymphal maturation) than their lower-altitude counterparts, thereby maximizing their chances of survival (in essence, quicker development is necessary due to a more unfavorable climate) (Berner et al. 2004). Smaller hills at low elevations (summits of less than 100 masl) are also likely to have diverse environmental conditions depending on aspect and slope gradient.

iii) Slope gradient.—Typically, slope steepness is classified by gradient: 0–5% very weak slope, 6–10% weak slope, 10–15% moderate slope and > 16% moderately steep and greater (Sikdar et al. 2004). Studies have indicated a declining abundance and diversity of Orthoptera with increasing slope steepness (e.g., Gebeyehu and Samways 2006, Wersebeckmann et al. 2023). Steep slopes are often difficult to actively manage and become abandoned, leading to scrub encroachment and reduced Orthoptera abundance and diversity of open habitats (Marini et al. 2009, Wersebeckmann et al. 2023). In contrast, on small hills in the UK, grassy summits and weak-moderately steep slopes may be particularly susceptible to the burrowing activities of rabbits. This lagomorph digging activity creates an open mosaic of bare soil and fine-leaved grasses favorable for grasshoppers of shorter vegetation (e.g., C. brunneus) but not bush-crickets (e.g., Conocephalus fuscus Fabricius, 1793), which inhabit taller swards (Gardiner 2022).

Interaction between grazing and topography.—On low elevation hillsides (<100 m), the grazing of grassland by wild lagomorphs (e.g., hare Lepus europaeus Pallas, 1778 and rabbit Oryctoloagus cuniculus Linnaeus, 1758) or targeted livestock (e.g., cattle or sheep) appears to determine the abundance and diversity of Orthoptera in combination with aspect (Fonderflick et al. 2014). Two hillside studies suggested lower species richness on rabbit-grazed hill summits. Rabbit grazing leads to an absence of species such as Roesel’s bush-cricket Roeselii roeseliana (Hagenbach, 1822) and dark bush-cricket Pholidoptera griseoaptera (De Geer, 1773) and an abundance of C. brunneus and Pseudochorthippus parallelus (Zetterstedt, 1821) (Gardiner 2022). There is uncertainty over whether grazing merely expands favorable habitat on hills for species such as C. brunneus or if bare soil lures adults away from nymphal habitats. Soil slippage on hills also appears to provide the bare soil habitat required for adult oviposition and basking for species such as C. brunneus in addition to the trampling hooves of grazing livestock (cattle or sheep) and the burrowing activities of lagomorphs (Gardiner 2022).

Topography and rewilding schemes.—Variations in landscape topographic features (e.g., slope gradient) can create microclimate heterogeneity (e.g., soil drainage and solar radiation), which influences plant and Orthoptera distribution at differing altitudes (Unsicker et al. 2010) and may be important for the selection of rewilding sites. Plant communities are also altered by topographic heterogeneity that influences the height and structure of vegetation as well as the feeding areas of Orthoptera. For example, particularly dense patches of coarse grasses (e.g., cock’s-foot Dactylis glomerata Linnaeus, 1753) often form in the moist soil at the base of slopes, which can be important feeding areas for nymphs and adults of species such as the meadow grasshopper P. parallelus (Gardiner and Hill 2004). Rewilding on former cropland without significant heterogeneity (i.e., mostly flat terrain) can cause the development of an open layer of fine-leaved grasses (e.g., fescue Festuca and bent Agrostis spp.) and a warm microclimate favorable for nymphal and adult development (Gardiner 2022) but limited coarse grasses for feeding (Gardiner and Casey 2022a, b). A mosaic of fine and coarse-leaved grasses is required for all life stages. Movements of grasshopper nymphs and adults between different grass patches may therefore be frequent on post-arable rewilding sites. Topographic heterogeneity, particularly of aspect, elevation, and slope gradient, influences the distribution of these grassland patches (Unsicker et al. 2010).

One factor that was not investigated in a recent study of Orthoptera colonization at Black Bourn Valley was the influence of topographic heterogeneity in rewilding fields (Broad 2023). This was mainly due to the mostly flat topography of the landscape (c. fields 39–52 m elevation) and the absence of any notable slopes (very weak slopes: mean gradient 0.9%, min/max 0.2–2.2%) or small hills. The effect of aspect and slope gradient on weak–moderate slopes (5–15%) has not been thoroughly investigated, while its importance for Orthoptera in rewilding schemes is unknown.

The aim of this study was to compare the rewilded-grassland Orthoptera on flat and sloping former-arable fields at Arger Fen in Suffolk, UK. The results are discussed in relation to the topographic heterogeneity of the fields, specifically the abundance and diversity of Orthoptera in rewilded fields with differing slope gradients.

Study site.—Arger Fen nature reserve (eastern England, 51°59'14.9856"N, 0°49'9.3504"E) is owned and managed by Suffolk Wildlife Trust (SWT) and is a mosaic of lowland woodland, dry acid grassland, and fen wetland (110 ha total area) alongside farmland with slopes varying from very weak to moderately steep (mean gradient 3%, min/max 0.2–10.9%). Each of the four rewilding fields were taken out of arable production between 2004 and 2014 and had been treated with nitrogen (N) fertilizer for a range of annual crops, including winter wheat. Soil types at the reserve vary from clay loams with impeded drainage to freely drained sands/gravels. Much of the rewilding area was once part of Leavenheath, an extensive (69 ha) lowland heath that was enclosed and converted into arable farmland after 1817 (Chatters 1985). The intention of rewilding the arable fields was to establish a mosaic of rough grassland and scrub similar to that present on Leavenheath before enclosure.

A total of four fields were selected for this study due to their differing topographic heterogeneity: two ‘flat’ and two ‘sloping’ (Table 1, Figs 1, 2). The topography of the flat fields was almost completely even with a few very weak slopes (mean gradient 0.2–0.8%), while the sloping fields had slopes varying from very weak to moderately steep (gradient 1.5–10.9%). Vertical relief was greatest in the sloping fields (27 m) compared to the flat fields (13 m). All fields were last plowed and cropped before 2014, resulting in maturing grassland swards similar to other ex-arable rewilding sites in Suffolk, such as Black Bourn Valley (Gardiner and Casey 2022a, b). Once cropping ceased, all fields were allowed to naturally revert to grassland and scrub through succession with minimal intervention apart from mowing of footpaths and occasional light grazing (not during the study period).

Figure 1. 

Flat (A) and sloping (B) rewilding fields at Arger Fen. © Tim Gardiner.

Figure 2. 

The four study fields demarcated by flat or sloping status and the location of Arger Fen.

In the flat and sloping fields in this study, the grasses creeping bent (Agrostis stolonifera Linnaeus, 1753), crested dog’s tail (Cynosurus cristatus Linnaeus, 1753), D. glomerata, and Yorkshire fog (Holcus lanatus Linnaeus, 1753) were frequent, while tufted-hair grass (Deschampsia caespitosa Linnaeus, 1753, Palisot de Beauvois 1812) and red fescue (Festuca rubra Linnaeus, 1753) were often locally abundant. Marsh thistle (Cirsium palustre (Linnaeus 1753), Scopoli 1772) and compact rush (Juncus conglomeratus Linnaeus, 1753) were found in wetter patches in the flat fields. Herbaceous species were scattered throughout all four rewilding fields, including pyramidal orchid (Anacamptis pyramidalis (Linnaeus) Richard 1817), knapweed (Centaura nigra Linnaeus, 1753), wild carrot (Daucus carota Linnaeus, 1753), perforate St. John’s wort (Hypericum perforatum Linnaeus, 1753), cat’s-ear (Hypochaeris radicata Linnaeus, 1753), ox-eye daisy (Leucanthemum vulgare Lamarck, 1779), and ribwort plantain (Plantago lanceolata Linnaeus, 1753). In addition, sheep’s sorrel (Rumex acetosella Linnaeus, 1753) and bracken (Pteridium aquilinum Linnaeus, 1753, Kuhn, 1879) were found to be abundant in a sandpit on the edge of the sloping Ford’s Heath field.

Arable weeds persisting from former cultivation included the annuals field madder (Sherardia arvensis Linnaeus, 1753; Kingsland Lane) and scarlet pimpernel (Anagallis arvensis Linnaeus, 1753; Ford’s Heath). The near threatened (NT; Vascular Plant Red Data List for Great Britain, Cheffings et al. 2005) annual common cudweed Filago vulgaris (Lamarck, 1779) was recorded in the sloping Ford’s Heath and Hullback’s Grove fields, indicating a possible initial return of a heathy grassland flora through rewilding on the well-drained soil of the slopes. Filago vulgaris has only been found to be frequent in Suffolk in the acidic, freely draining soils of the Sandlings and Breckland heathlands (Sanford 2005).

Post-cropping cessation, scrubland communities have established in the rewilding fields. Common scrub species observed include hazel (Corylus avellana Linnaeus, 1753), hawthorn (Crataegus monogyna, Jacquin, 1775), broom (Cytisus scoparius Linnaeus, 1753, Link, 1822), (only in Ford’s Heath near sandpit), blackthorn (Prunus spinosa Linnaeus, 1753), oak (Quercus robur Linnaeus, 1753), field rose (Rosa arvensis Hudson, 1762), bramble (Rubus fruticosus Linnaeus, 1753), and goat willow (Salix caprea Linnaeus, 1753). A small area (c. 0.2 ha) of ash (Fraxinus excelsior Linnaeus, 1753) saplings affected by Hymenoscyphus fraxineus dieback occurred in Hullback’s Grove. Saplings were interspersed with tall grasses.

There was no active conservation management of any of the fields throughout the 2021 summer study period. All fields were surrounded by dense hedgerows and were adjacent to woodland (Fig. 2).

Transect surveys.—A 1 m wide × 400 m long transect was established in each of the four fields. Transects were arranged in a W shape (each arm 100 m) to ensure even coverage of each field and to avoid any edge habitat effects. The transect method followed the methodology of Gardiner et al. (2005), Gardiner and Hill (2006), and Gardiner (2021). Adult individuals of all Orthoptera species along all transects were recorded acoustically and visually to determine assemblage composition, species diversity, and species richness.

Each transect was walked at a slow, strolling pace (2 km/h) on three occasions from May to August 2021. Nymphs detected from a 1-m wide band in front of the observer were recorded along all transects. The observer walked through the grass and recorded any orthopteran that was disturbed and jumped. As it is difficult to distinguish between species in the early instars (though not impossible; see Thommen 2021), nymphs of all species were lumped together for recording purposes. The surveys were undertaken in vegetation sufficiently short (<50 cm) to minimize the possibility of overlooking nymphs in tall grass or non-stridulating species such as groundhoppers (Tetrigidae) (Gardiner et al. 2005). For an observer (TG) with 25 years’ experience identifying Orthoptera in the UK, it was relatively easy to ascertain the species of adults without capture (Gardiner and Hill 2006), although some species, such as C. fuscus and R. roeselii, are significantly under-recorded using visual transects (Gardiner and Hill 2006). It should be noted that inexperienced surveyors (<5 years identifying Orthoptera in the field) may produce much lower counts of Orthoptera (Gardiner 2009b).

A dual visual and acoustic monitoring method has been used by Weiss et al. (2013) to ensure complete coverage of the orthopteran fauna of sites. In the current study, a stridulation monitoring technique was used to record adult males of species that stridulated along the transects at the same time as visual monitoring. Stridulation monitoring has been used to record cryptic species in the county of Essex and has been found to be effective compared to visual sighting transects and pitfall traps (Harvey and Gardiner 2006, Gardiner et al. 2010, Schirmel et al. 2010). Acoustic signaling was used to locate Orthoptera to assess species richness/diversity and preferences between flat and sloping rewilding fields, but acoustic signals were not recorded or analyzed. Bat detectors were not utilized in the current study as the first author (TG) was able to detect stridulating males with non-ultrasonic calls up to 20 m away either side of the transect. The weather conditions on survey days were favorable for insect activity, being largely sunny and warm (>17°C).

Acoustic detection systems (Diwaker et al. 2007, Penone et al. 2013, Lehman 2014, Jeliazkov et al. 2016, Newson et al. 2017, Tomar et al. 2017, Walcher et al. 2022) are commonly used to obtain data on the presence and abundance of Orthoptera. Acoustic detectors allow the recording of orthopterans with ultrasonic stridulations not detectable by the human ear. They also provide a strong measure of repeatability between observers and surveys (Diwaker et al. 2007). For the Orthoptera recorded in the Arger Fen study area, the conehead C. fuscus has a peak stridulation frequency of 30 kHz (Iorgu and Iorgu 2011), beyond the upper range of human hearing (c. 20 kHz; Diwaker et al. 2007). Other local species, such as C. brunneus (peak frequency 12 kHz) and R. roeselii (peak frequency 17 kHz) (Iorgu and Iorgu 2011), are known by the surveyor (TG) to be easily detected without an acoustic detector based on his 25 years of experience, as their peak frequencies are within the range of human hearing. The use of auditory detection methods by trained surveyors without detection equipment has been found to be reliable for the detection of Orthoptera in tropical regions (Diwaker et al. 2007, Tomar and Diwakar 2020).

When identifying species, the results using combined visual-acoustic surveying techniques, such as that used in the current study, can be comparable to those of automated detection systems, particularly in regard to low-frequency calls, and are considerably less time intensive (Walcher et al. 2022). However, the human listener is not effective at detecting species with high frequency calls (>20 kHz, Walcher et al. 2022), such as C. fuscus, which, in the current study, required visual identification.

Environmental surveys.—The mean slope gradient for each 100-m arm of all transects was calculated using onthegomap.com. In late May 2021, a total of 40 grass heights were recorded at random positions (selected while walking) along the Orthoptera transects (10 on every 100-m arm) using a 1 m ruler for each of the four fields. In each field, anthills were counted along the 400 m long Orthoptera transects in a 1 m wide band, and the number of wild lagomorph (brown hare and rabbit combined) droppings (dung balls) were recorded on every 100-m arm (positions randomly selected while walking) to ascertain the level of grazing pressure in the fields (Wood 1988, Gibb and Fitzgerald 1998, Millett and Edmondson 2013). To provide further evidence of wild rabbit grazing, the number of burrow excavations was also recorded (positions randomly selected while walking) on the transects for every 100-m arm.

Statistical analysis.—All data were square-root transformed to correct for non-normality before analysis (Heath 1995). Significance for all tests was accepted as evidence on the following scale in accordance with Muff et al. (2022): p-value > 0.1, little or no evidence; 0.05–0.1, weak evidence; < 0.05, moderate evidence; < 0.01, strong evidence; or < 0.001, very strong evidence.

Slope gradient.—The mean slope gradient for each field was compared between flat and sloping fields using a Student’s t-test. Correction was made for unequal variance where necessary using Satterthwaite’s approximate t test, a method in the Behrens-Welch family (Armitage and Berry 1994, Heath 1995).

Orthoptera.—To allow comparison of grasshopper abundance between Arger Fen and other sites (e.g., Black Bourn Valley rewilding site; Gardiner and Casey 2022a), the overall number of grasshopper adults (of all species combined) per hectare was calculated by dividing the pooled visual detections for the 3 surveys by the transect area searched (e.g., 1600 m² searched for each survey × 3/number of visual detections).

All detections of Orthoptera (visual or acoustic) were summed for each field and survey period (3 surveys) to determine species preferences between flat and sloping rewilding fields. The independence of field transects was assumed, and data were pooled for each transect for analysis in a way similar to other monitoring studies (Nur et al. 1999).

Species richness was calculated for each field and 100-m transect arm. Assemblage diversity estimates were calculated using Species Diversity and Richness software, Version 4.1.2. (Pisces Conservation Ltd., IRC House, The Square, Pennington, Lymington, Hampshire). The Shannon-Wiener Diversity Index (H’, Kent and Coker 1992) was calculated using the total number of individuals recorded for each Orthoptera species in each field and for each 100-m transect arm.

Student’s t-tests were used to determine whether species richness/diversity, abundance of adults (all species) and nymphs, height of grass, and number of anthills and lagomorph droppings differed between flat and sloping fields. Where necessary, corrections for unequal variance were performed using Satterthwaite’s approximate t test, a method of the Behrens-Welch family (Armitage and Berry 1994, Heath 1995).

To test the relationship between different species and slope gradient, all detections of Orthoptera (visual or acoustic) were summed for each 100-m transect arm (the arms had differing gradients). Linear regression models were run to determine whether the number of adults and nymphs of each species, species richness/diversity, grass height, anthills, lagomorph droppings, and rabbit excavations had significant relationships with slope gradient, which varied between the different 100-m transect arms in each field.

Through the visual surveys, a total density of 1221 adult grasshoppers/ha was recorded at Arger Fen. Six species of Orthoptera, all widespread and abundant in Suffolk, were recorded in the Arger Fen rewilding fields (Table 2, Appendix 1). The most recorded species (visual and acoustic detections combined) were P. parallelus (n = 409, 34% of adult detections) and R. roeselii (n = 323, 27%), followed by C. fuscus (n = 261, 22%) and C. brunneus (n = 154, 13%). Both the lesser marsh grasshopper Chorthippus albomarginatus (De Geer, 1773, n = 23) and the dark bush-cricket P. griseoaptera (n = 20) were found in much lower abundance (2%).

Strong evidence (p < 0.01) was found that the abundance of C. brunneus was significantly higher in sloping fields compared to C. fuscus, which preferred flat fields (Table 2). For nymphs (all species) and P. parallelus, there was statistical evidence (moderate and weak, respectively) of higher abundance in the flat fields. Contrastingly, species diversity was significantly higher (p < 0.05) in the sloping fields.

In the flat fields, P. parallelus was the most abundant orthopteran, comprising c. 41% of the total number, while C. brunneus represented only 1% of detections. However, in the sloping fields, C. brunneus was the most abundant species and represented 27% of the total detections. In these sloping fields, P. parallelus accounted for 26% of adult detections. The two bush-crickets, R. roeselii (flat: 31%, sloping 23%) and C. fuscus (flat: 26%, sloping 17%), were in similar abundance in the sloping fields compared to the flat grasslands. All six species were detected in the scrubby grassland of the hillside ash dieback area in Hullback’s Grove field.

Mean slope gradient was significantly higher in the sloping fields compared to the flat grasslands (t value -31, moderate evidence p = 0.02). There was moderate evidence (p < 0.05) that lagomorph droppings and excavations were in higher density in the sloping fields (Table 3). However, there was no evidence that anthill density and mean grass height were different between flat and sloping fields (Table 3).

At a more localized level, slope gradient influenced the abundance and diversity of Orthoptera. Linear regression models revealed distinct flat and slope species (Table 4). Moderately significant negative relationships were detected for P. parallelus (Fig. 3) and R. roeselii, and weak ones were detected for C. fuscus and nymphs, indicating a preference for flatter ground in all cases. However, for C. brunneus (Fig. 3), C. albomarginatus, and species richness and diversity, statistical evidence revealed significant positive relationships, suggesting a preference for steeper gradient slopes in all cases. There was strong evidence of a significant positive relationship between slope gradient and lagomorph droppings/excavations but not between slope gradient and anthills or grass height (Table 4).

Figure 3. 

The relationship between slope gradient and the abundance of two grasshopper species with contrasting topographic preferences, Chorthippus brunneus (slope species) and Pseudochorthippus parallelus (flat species); line of best fit shown.

Rewilding can lead to the return of biodiversity to farmland. Leaving arable fields to revert naturally to grassland, scrub, and woodland without active herbivore introduction is a relatively unstudied aspect of rewilding, with little data available to determine the success of schemes despite theoretical discussion of the benefits (Hart et al. 2023). The impact of rewilding on insects in Europe is largely unknown, despite studies conducted in Black Bourn Valley in southeast England (Gardiner and Casey 2022a) and in Sweden (Garrido et al. 2022).

Colonisation of rewilded fields.—Rewilding of arable land can lead to rapid colonization by Orthoptera, including species such as C. albomarginatus, C. fuscus, and R. roeselii, which have expanded their ranges in the UK due to climate change (Gardiner and Casey 2022a, b). The total of 1221 adult grasshoppers/ha was much higher at the topographically varied Arger Fen than in the flatter fields of Black Bourn Valley (234 adults/ha), another Suffolk rewilding site 30 km to the north (Gardiner and Casey 2022a). The six species recorded at Arger Fen were all recorded at Black Bourn Valley, which had an overall higher species richness (9 species). Orthoptera inhabiting Black Bourn Valley’s post-arable rewilded fields (Gardiner and Casey 2022a, b) but absent from Arger Fen were the common green grasshopper O. viridulus, the slender (Tetrix subulata Linnaeus, 1758), and the common groundhopper Tetrix undulata (Sowerby, 1806). In contrast to the distant population of O. viridulus (nearest known site > 4 km distant from fields at Arger Fen), the two groundhoppers have been recorded < 1 km from Arger Fen, making it possible that they could colonize the regenerating grassland and scrub mosaic if the dry, heathy grassland of Leavenheath continues as part of the rewilding strategy (Chatters 1985).

Habitat preferences in flat and sloping fields.—It is important that the mosaic of habitats established at rewilding sites is suitable for a wide range of invertebrate species. The habitat preferences of Orthoptera may relate to the choice of oviposition site, food preferences, and vegetation height (Clarke 1948, Richards and Waloff 1954, Gardiner 2006, 2009a). Early colonists of flat post-arable grassland include the range-expanding bush-crickets C. fuscus and R. roeselii (Tables 2, 4). In flat fields, D. glomerata provides the tall grassland (c. 37 cm grass height) both species require as habitat. P. parallelus had a higher abundance in flat fields, perhaps preferring the lush D. glomerata vegetation for feeding, shelter, and oviposition in grass-covered soil (Waloff 1950, Gardiner and Hill 2004), even though the grass height was well above the optimal level (10–20 cm) for this grasshopper (Gardiner et al. 2002).

While nymphs were also in higher abundance in the flat fields, C. brunneus preferred the sloping fields (Tables 2, 4, Fig. 3) where lagomorph grazing created patches of bare soil (Tables 3, 4), which may be suitable for oviposition and basking (Clarke 1948, Waloff 1950). Weak evidence was found suggesting that the range-expanding C. albomarginatus was a species of the sloping fields (Table 4), perhaps preferring to oviposit at the base of grass blades exposed by rabbit excavations (Waloff 1950). The preference of nymphs for flat fields, with adults of C. brunneus inhabiting sloping fields, suggests that movements of nymphs downhill from egg-hatching (eclosion) sites occurred before adults moved uphill to find the lagomorph-grazed areas to oviposit in the bare soil established by rabbit excavations. Research has shown that early instar grasshopper nymphs of C. brunneus are often found in short grassland near oviposition sites before moving to taller swards (10–20 cm grass height) as they mature (Gardiner et al. 2002).

Exposed soil may offer other benefits for grasshoppers by providing sites where they can bask, as it is often much warmer than surrounding vegetation (Key 2000). However, in this study, the amount of bare soil did not vary between field type, and the number of anthills in flat and sloping fields was similar. Unlike the study at Black Bourn Valley (Gardiner and Casey 2022a, b), microenvironments provided by anthills were evenly distributed between fields, perhaps due to the relatively well-established post-arable grassland (>7 years old) and rapid colonization by ants.

Wild lagomorph grazing and interaction with topography.—Wild grazing animals play a significant part in reducing vegetation height and cover (Fargeaud and Gardiner 2018, Gardiner 2018). Rabbits grazed the closed grassland of the sloping fields, creating numerous patches of exposed soil due to their burrowing activities, which were favorable for adult C. brunneus and to a lesser extent C. albomarginatus (Gardiner et al. 2002). Rewilding fields on hillsides may therefore provide a range of topographic conditions and niches that can be utilized by C. brunneus, particularly where there is grazing by wild lagomorphs (Grayson and Hassall 1985).

Short sward patches established by lagomorph grazing may have excessively hot temperatures (>40°C) similar to hay meadows after cutting (Gardiner and Hassall 2009), which are unlikely to be favorable for grasshoppers in the absence of ‘cool’ tussocks in close proximity. Both the flat and sloping field types had a mean grass height > 30 cm, which may provide grasshoppers with numerous sheltered ‘cool’ areas of tall vegetation when temperatures are excessively hot as the climate warms. Such behavioral thermoregulation may account for the persistence of species such as C. brunneus on slopes where lagomorph excavations (basking and egg-laying sites) were frequently in close proximity to cooler tall vegetation for shade-seeking orthopterans (Fig. 1).

Rewilding max: scrub and bare soil provision.—The absence of domestic livestock grazing during the study period is akin to rewilding max (i.e., more than Rewilding Lite) where active conservation is absent (Gordon et al. 2021a, b) and there is a lack of ‘control’ by site managers (Dempsey 2021). In these situations, large and diverse populations of Orthoptera can build up in the initial phase of grassland regeneration after arable cropping has ceased. As observed at Knepp with satellite remote sensing over two decades, a heterogeneous patchwork of rewilding habitats with scrub and woodland can develop on post-arable land (Schulte to Bühne et al. 2022), removing early successional habitats important for Orthoptera.

Wild grazers such as lagomorphs may create the micro-heterogeneity in habitat necessary for egg-laying grasshoppers in a rewilding max scenario with sloping farmland, although other forms of soil disturbance, such as disc harrowing, may be necessary where bare soil is lost as succession progresses (Gardiner and Casey 2022a, b). Flower-rich grassland swards can also develop with soil disturbance; for example, in the areas of Hullback’s Grove grazed by lagomorphs, an open sward with patches of bare soil has developed, which is favorable for C. brunneus (Fig. 4). It has not been possible to study the effects of domestic livestock grazing (e.g., sheep or cattle) in either flat or sloping post-arable fields, so future studies should investigate the influence of managed grazing on Orthoptera abundance and diversity.

Figure 4. 

Flower-rich, open sward grazed by lagomorphs in Hullback’s Grove hillside field; © Tim Gardiner.

All six study species were detected in the hillside ash dieback area in Hullback’s Grove. The mortality of F. excelsior saplings affected by the fungus H. fraxineus led to the maintenance of an open grass and scrub mosaic that would otherwise have been shaded out as the canopy matured (Fig. 5). Such local-scale stochastic variations in habitats outside the control of conservation management may be important for Orthoptera on rewilded hillsides.

Figure 5. 

Scrub and grassland mosaic in an ash Fraxinus excelsior dieback (Hymenoscyphus fraxineus) area, Hullback’s Grove hillside field, inhabited by six species of Orthoptera; © Tim Gardiner.

Survey limitations.—The visual and acoustic surveying technique used at Arger Fen did not utilize acoustic detectors (e.g., those applied to bat detection) to record species in a standardized way (see Newson et al. 2017 and Walcher et al. 2022), which would ensure a strong measure of repeatability between observers and surveys (Diwaker et al. 2007). However, all six species were detected using both visual and acoustic survey techniques at Arger Fen (Appendix 1). The conehead C. fuscus generally has a peak stridulation frequency (30 kHz) beyond the upper range of human hearing (c. 20 kHz, Diwaker et al. 2007) and was therefore virtually undetected by ear without acoustic detection equipment in this study (only three acoustic detections, presumably at low frequency stridulation, were made). Estimates of C. fuscus abundance are also impacted by visual detection methods where the cryptic nature of the species leads to under-recording and lower species richness estimates (Gardiner and Hill 2006). Other local species, such as R. roeselii (peak frequency c. 17 kHz), were commonly detected by the human listener on the surveys (284 acoustic detections of R. roeselii, Appendix 1), suggesting that the method was useful for the other four study species with call frequencies close to or within the range of human hearing (the three grasshopper species in this study have a peak frequency limit ≤ 21 kHz; Meyer and Elsner 1996). Therefore, visual-acoustic surveying techniques as used in the current study may be comparable to automated detection systems where species have low frequency calls (≤20 kHz) in addition to being considerably less time intensive in the field (Walcher et al. 2022). For larger studies involving multiple observers and numerous species with ultrasonic stridulation, we suggest the use of an acoustic detection system (e.g., bat detector) for the acoustic component of the surveying technique to ensure standardization and repeatability.

Another source of error in this study may be the accuracy of the lagomorph dropping counts. Compared to the taller and moister vegetation present on the lower slopes, droppings may have been easier to locate in shorter, lagomorph-grazed vegetation and would also have dried and been less likely to decay. Thus, the lagomorph dropping counts must be viewed with some caution, and further investigation is required. The choice of random locations on site may also have led to surveyor bias in the sampling of environmental variables in the transects, and it would have been better to locate survey points via random number tables. The presence of only two replicates each for the flat and sloping fields is also a limitation of the current research, and future studies should incorporate more fields where possible.

Outlook for Arger Fen.—There is evidence that the heathy grassland and scrub vegetation of Leavenheath is beginning to re-establish in the sloping fields, following a trajectory similar to nearby Tiger Hill (<1 km distant). At Tiger Hill, the acid grassland is dominated by Agrostis and Festuca grasses with occasional R. acetosella, moss, and anthills (Kirby et al. 2002). On Ford’s Heath, Agrostis was frequent while R. acetosella was found in a sandpit connecting to the edge of the rewilding field. Along with occasional moss and anthills (Table 3), the presence of P. aquilinum, patches of C. scoparius scrub, and bare ground with the threatened annual F. vulgaris indicates the development of heathy vegetation similar to that found on dry, acid soil locally at Tiger Hill. The developing scrub and grassland mosaic on hillside slopes could be important for species of woody vegetation, such as P. griseoaptera, and those of sparse vegetation, including C. brunneus. The latter species can build up large populations on long-established heathland (1500–3100 adults/ha; Gardiner et al. 2002), which suggests that C. brunneus density in sloping fields (542–700 adults/ha) is on an initial trajectory toward the high abundance of heathy, acid grassland, whereas the flat fields are not (16–25 adults/ha).

Local species of Orthoptera quickly colonize new habitats created on former arable land, particularly species expanding their range due to climate change. In these areas, species diversity and community heterogeneity are improved by differing local topography. This study provides initial evidence that topographic heterogeneity may be important for the diversity of Orthoptera in rewilding schemes on former arable land. Our results suggest that Orthoptera can profit from rewilding schemes on sloping farmland.

The authors would like to thank Suffolk Wildlife Trust for supporting the project and giving permission to survey the fields. We would also like to extend our gratitude to the Suffolk Naturalists’ Society for funding the research through their Morley Grant Scheme. Professor Rob Fuller kindly helped with the discussion and review of the draft manuscript, while Anna Saltmarsh assisted with the fieldwork.