Habitat requirements for the conservation of wood cricket (Nemobius sylvestris) (Orthoptera: Gryllidae) on the Isle of Wight, UK

TLDR: In this paper, a field investigation within three woodlands on the Isle of Wight was undertaken to identify its habitat preferences, including the presence of a well developed leaf litter layer, relatively low ground vegetation cover and height, low canopy cover and relatively short distances between individual populations.

Abstract: In the UK, wood cricket (Nemobius sylvestris) is a ‘Species of Conservation Concern’, being restricted to only three areas in southern England. Little information is available on the specific habitat requirements of this species. In 2006, a field investigation within three woodlands on the Isle of Wight was undertaken to identify its habitat preferences. Factors positively influencing wood cricket presence within woodlands included the presence of a well-developed leaf litter layer, relatively low ground vegetation cover and height, low canopy cover and relatively short distances between individual populations. Regression models identified the degree of isolation and variables describing vegetation structure as the main predictors for wood cricket presence within woodland fragments. The results of this study indicate the preference of wood cricket for open wooded edges. Conservation efforts for this species should focus on continuation of regular management activities aimed at providing permanent open edge habitat within woodlands, to maintain viable populations.

Figures

Fig. 2 Predicted probability of wood cricket being present related to the main explanatory variables. O indicates sites where wood cricket was present; x indicates sites where wood cricket was absent. Two outliers (167 cm for vegetation height and 273 m for distance) were omitted from Figure a and b. The curves were calculated with the following probability equations using the B values from Table 4:

Table 1: Variables that were computed from the field measurements and found to influence wood cricket presence/absence within woodland habitat.

Table 2: Mann-Whitney U test for the relationship between wood cricket presence/absence and fourteen habitat variables analysed through separate tests. n = 402; wood cricket present n1 = 201 and absent n2 = 201; Med Abs/Pres = median value for locations where wood cricket was absent or present, U = Mann-Whitney test statistic; z = test statistic given by SPSS when performing a Mann-Whitney U test and is used to test for a significant difference between two groups; P = probability or significance level; r = effect size.

Table 3: Spearman rank correlation between the individual habitat variables. n = 402, r = correlation coefficient, P = significance or probability value. Bold figures indicate correlation coefficients (r) values > 0.50.

Fig. 3 Predictive probability of the full model represented in 3-D. In each figure probability of wood cricket presence (y-axis) is plotted against occupied nearest neighbour distance (x-axis) and ground vegetation height (z-axis). Columns represent different levels of South orientated canopy cover and rows represent different values of cumulative ground vegetation cover.

Table 4: Summary of the logistic regression analyses. Model: variables included in the logistic regression model. Model performance: summary of model performance tests. Model fit test: tests if the model fits the data (P < 0.05 = good model fit). Hosmer & Lemeshow Model fit test: tests if the model fits the data (P > 0.05 = good model fit). Classification table (%): indicates percentage of cases correctly classified by the model. Effect size: indicates the amount of explained variation by the model (Nagelkerke R2; range 0 – 1). Variables in the equation: indicates the usefulness of the individual variables included in the model. Wald test: tests contribution to the model for the individual variables (P < 0.05 = significant). B (with Standard Error): indicates the direction of the relationship between the individual variables and wood cricket presence (- indicates a negative and + a positive relationship). Exp. (B) (with 95% Confidence Interval): indicates the odds ratio for wood cricket presence per unit increase of the individual variable (below 1 indicates a decrease above 1 an increase).

Full text

MURDOCH RESEARCH REPOSITORY
http://researchrepository.murdoch.edu.au
This is the author's final version of the work, as accepted for publication following peer review but without the
publisher's layout or pagination.
Brouwers, N.C. and Newton, A.C. (2009) Habitat requirements for the conservation of wood cricket
(Nemobius sylvestris) (Orthoptera: Gryllidae) on the Isle of Wight, UK. Journal of Insect Conservation, 13
(5). pp. 529-541.
http://researchrepository.murdoch.edu.au/4438
Copyright © Springer Science+Business Media B.V.
It is posted here for your personal use. No further distribution is permitted.
http://tweaket.com/CPGenerator/?id=4438
1 of 1 2/06/2011 4:34 P
M

1
Full reference:
Brouwers, N.C., Newton, A.C., 2009a. Habitat requirements for the conservation of wood
cricket (Nemobius sylvestris) (Orthoptera: Gryllidae) on the Isle of Wight, UK. Journal of
Insect Conservation 13, 529-541.
Title:
Habitat requirements for the conservation of wood cricket (Nemobius sylvestris) on the
Isle of Wight, UK
Niels C. Brouwers • Adrian C. Newton
N. C. Brouwers (Corresponding author) • A. C. Newton
School of Conservation Sciences
Bournemouth University
Talbot Campus, Fern Barrow, Poole, Dorset, BH12 5BB, UK
E-mail: ncbrouwers@hotmail.com
Phone: ++44 1202 965498 / Fax: ++44 1202 965255
Date of the manuscript draft: November 21, 2008

2
1 Abstract
In the UK, wood cricket (Nemobius sylvestris) is a ‘Species of Conservation Concern’, being
restricted to only three areas in southern England. Little information is available on the
specific habitat requirements of this species. In 2006, a field investigation within three
woodlands on the Isle of Wight was undertaken to identify its habitat preferences. Factors
positively influencing wood cricket presence within woodlands included the presence of a
well-developed leaf litter layer, relatively low ground vegetation cover and height, low canopy
cover and relatively short distances between individual populations. Regression models
identified the degree of isolation and variables describing vegetation structure as the main
predictors for wood cricket presence within woodland fragments. The results of this study
indicate the preference of wood cricket for open wooded edges. Conservation efforts for this
species should focus on continuation of regular management activities aimed at providing
permanent open edge habitat within woodlands, to maintain viable populations.
Keywords: woodland; forest; habitat requirements; conservation; wood cricket; Nemobius
sylvestris; Isle of Wight; United Kingdom

3
2 Introduction
Much interest has focused recently on the role of landscape-scale factors in maintaining
populations of species, particularly as a result of developments in metapopulation theory and
landscape ecology (Hanski and Gilpin 1997; Gutzwiller 2002; Crooks and Sanjayan 2006).
However, for the conservation of invertebrate species, factors acting at a local scale may
often be equally important for the persistence of individual populations as habitat availability
at the landscape scale, particularly for species with limited dispersal ability. Indications for
this are found in habitat fragmentation studies that have been undertaken at a range of
different spatial scales, revealing the relative importance of within-patch habitat compared to
spatial measures such as patch size and isolation between habitat fragments (e.g. Rukke
and Midtgaard 1998; Ranius 2000; Binzenhofer et al. 2005). For example, in a study on a
burnet moth species in an abandoned agricultural landscape in Germany, Binzenhofer et al.
(2005) found that presence of the species was mainly explained by total nectar plant cover
(i.e. habitat availability) within patches, whereas no patch size or isolation effect between
habitat patches was found. Two studies on beetles living in dead fungal fruiting bodies on
trees revealed similar results, where fragment area and isolation were found to be less
important explanatory variables for presence than the total amount of habitat (i.e. fungus
fruiting bodies) available within the individual woodland stands (Rukke and Midtgaard 1998;
Rukke 2000). Furthermore, in a study on an endangered longhorn beetle living on dead
trees, Buse et al. (2007) revealed that variables measured at the tree level were better
predictors of presence of the species than spatial measurements between trees. These
examples indicate the overall importance of within patch (i.e. local scale) habitat availability
in determining invertebrate presence and population persistence.
Detailed studies examining habitat factors influencing invertebrate populations are required
to be able to determine habitat suitability and species-specific requirements within individual
sites. Presence/absence studies are often used to analyse the responses of individual
species to habitat variables (e.g. Rukke 2000; Binzenhofer et al. 2005). Variables often
measured include habitat factors related to species-specific food availability, vegetation
structure (e.g. canopy cover), abiotic conditions (e.g. sunlight availability) and isolation
measures (e.g. nearest neighbour distance). Studies on grassland species have revealed
positive relationships with food availability and negative relationships with habitat distance,
but differing results for vegetation structure and related abiotic conditions (Binzenhofer et al.
2005; Strauss and Biedermann 2005; Heller and Gordon 2006). Studies specifically on
woodland species have found similar relationships. Most such studies to date have focused

4
on endangered ground or tree related beetle species (Rukke and Midtgaard 1998; Rukke
2000; Siitonen and Saaristo 2000; Sroka and Finch 2006; Buse et al. 2007; Matern et al.
2007) and butterfly species (Thomas et al. 1992; Konvicka et al. 2007). These studies have
similarly found positive relationships with measures of food availability (Rukke and Midtgaard
1998; Rukke 2000; Buse et al. 2007) and negative relationships with occupied nearest
neighbour distance between habitat patches (Thomas et al. 1992; Rukke and Midtgaard
1998; Rukke 2000; Siitonen and Saaristo 2000; Buse et al. 2007). For canopy cover in most
cases a negative relationship has been found (Rukke and Midtgaard 1998; Buse et al. 2007;
Matern et al. 2007), however the influence of vegetation structure differs widely between
species (Siitonen and Saaristo 2000; Sroka and Finch 2006; Buse et al. 2007; Konvicka et al.
2007; Matern et al. 2007; Sorvari and Hakkarainen 2007).
The research described here focused on wood cricket (Nemobius sylvestris) on the Isle of
Wight, United Kingdom. Although wood cricket is relatively widespread in Europe (Brown
1978), in the UK it has the national status of a ‘Species of Conservation Concern’ (NBN
Gateway 2007). In the UK, wood cricket reaches the northern limit of its European
distribution. Populations of the species are restricted to the South of England at three main
locations, the New Forest (Hampshire), South Devon and on the Isle of Wight (NBN
Gateway 2007). On the Isle of Wight, populations are largely restricted to relatively large
woodland fragments occurring in the northern half of the island (Brouwers and Newton 2008).
The specific habitat requirements of the species are poorly understood, and existing
knowledge is largely based on observational and anecdotal information (e.g. Richards 1952).
Wood cricket is a non-flying cricket species that is strongly associated with native
broadleaved woodland, often dominated by oak (Quercus spp.) (Richards 1952). It is
typically found in relatively open areas such as woodland clearings and edges of woodland
tracks, footpaths, railway lines and woodland peripheries (Richards 1952; Morvan and
Campan 1976; Beugnon 1980). Locally the species can reach high population densities
(Gabbutt 1959). The insects live on the ground and prefer a well-developed leaf litter layer,
which serves as shelter, a food source and as a breeding ground (Richards 1952; Brown
1978; Proess and Baden 2000). The species is considered to be omnivorous with the staple
diet mainly being composed of dead leaf litter material (Gabbutt 1959; Koehler and Samietz
2006). However, to date no detailed study has been undertaken of the specific habitat
requirements of the species in relation to its presence or absence within woodland stands.
To address this gap in knowledge, the distribution of wood cricket was investigated within
three separate woodlands in relation to a range of habitat characteristics. In order to define

Frequently asked questions

Q1. What are the contributions in this paper?

This paper found that within-patch habitat is more important for the persistence of individual populations as habitat availability at the landscape scale, particularly for species with limited dispersal ability. 

Q2. What are the future works in this paper?

This further highlights the need to obtain species-specific parameters relating to dispersal ability in order to improve and inform future modelling approaches. However, when clearings are created, these should preferably be adjacent to already inhabited locations ( e. g. permanent open ride edges ) in order to increase the potential of dispersal of the species into these newly created habitat areas. 

Q3. What should be the main focus of conservation efforts for this species?

Conservation efforts for this species should focus on continuation of regular management activities aimed at providing permanent open edge habitat within woodlands, to maintain viable populations. 

Q4. What factors positively influenced wood cricket presence within woodlands?

Factors positively influencing wood cricket presence within woodlands included the presence of a well-developed leaf litter layer, relatively low ground vegetation cover and height, low canopy cover and relatively short distances between individual populations. 

Q5. What are the main predictors of wood cricket presence within woodlands?

Regression models identified the degree of isolation and variables describing vegetation structure as the main predictors for wood cricket presence within woodland fragments. 

Q6. Why was extinction assumed to be more likely for B. reticulatus?

Because of the successional dynamics of the habitat locations (i.e. fungus fruiting bodies on dying trees), for B. reticulatus, extinction was assumed more likely to be a consequence of the environment becoming permanently unsuitable than stochastic population fluctuations within permanent stable habitat locations (see Thomas 1994). 

Q7. What could be done to improve the model performance?

Including more precise measures of, for example, humidity, light availability at ground level and wind exposure might improve the model performance. 

Q8. Why did Rieske and Buss 2001 find that wood crickets prefer sunlit conditions?

Ground-dwelling invertebrates generally favour sunlit conditions because of their thermophilic nature (e.g. Rieske and Buss 2001; Buse et al. 2007). 

Q9. What could be the effect of thinning of woodland stands on wood cricket?

Woodland restoration efforts (Defra 2005; Forestry Commission 2006) and thinning of woodland stands, which involve opening up the canopy, could also have a temporary positive effect on wood cricket populations by increasing habitat availability. 

Q10. How was the data used to distinguish the strata?

To reduce the impact of errors in precision (e.g. using hand-held GPS), these strata were distinguished using a combination of data sources. 

Q11. What is the main reason for the presence of wood cricket in woodlands?

Presence of permanent open edge habitat within fragments was found to be a strong indicator for wood cricket presence (Brouwers and Newton 2008), and generally occurs in locations that are actively managed. 

Q12. How do the authors prevent the natural overgrowth of wood cricket sites?

To prevent the natural overgrowing of these sites, these locations need to be actively managed through regular removal of the ground vegetation. 

Q13. What were the main hypotheses based on the findings of previous research?

Further hypotheses based on findings of previous research were that wood cricket would be more likely to be present when (1) ground vegetation cover was relatively sparse, and (2) canopy closure was relatively low. 

Q14. What was the important factor influencing wood cricket presence?

Locations where wood cricket was found tended to be relatively close to these permanent habitat locations, indicating that more isolated suitable habitat locations were less likely to be inhabited. 

Q15. Why was wood cricket not present in the study?

In such locations, either wood cricket was not present because of the lack of suitable habitat or because of their limited dispersal ability, or possiblybecause of the presence of internal barriers to dispersal within the woodlands. 

Q16. What is the probability of wood cricket presence in woodland?

With increasing values of ‘South orientated canopy closure’ and ‘Cumulative (i.e. structured) ground vegetation cover’, the probability of wood cricket presence decreased with increasing distance to the nearest occupied location and ground vegetation height (Fig. 3).# Fig 3 approx here #The results of this study confirmed earlier observations indicating the preference of wood cricket for open wooded edges. 

Q17. What is the main predictor of wood cricket presence in woodland?

Results of this study suggest that wood cricket is present only in early successional open woodland habitat with low vegetation cover. 

Q18. What is the main reason for the dispersal ability of wood cricket?

the dispersal ability of this species is a factor that needs to be considered in order to predict their presence with more accuracy than based on habitat suitability alone. 

Q19. What methods were used to examine the relative influence of the different habitat variables for explaining the presence or?

Several logistic regression methods were used to examine the relative influence of the different habitat variables for explaining the presence or absence of wood cricket within the woodlands.