SEARCH

SEARCH BY CITATION

Keywords:

  • Conservation;
  • grassland;
  • habitat loss;
  • habitat selectivity;
  • management;
  • woodland

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgements
  9. References
  1. We studied the habitat requirements of a vulnerable butterfly, Erebia aethiops, in a grassland-forest mosaic within a nature reserve. This species inhabits seemingly abundant habitats such as forest edges, but it is declining in many parts of Europe.

  2. We analysed mark-recapture data, focusing on the effects of distinct vegetation structures, nectar sources and management regimes on population density and mobility.

  3. Adult E. aethiops preferred abandoned grasslands and small open enclaves surrounded by forest; i.e. highly heterogeneous habitats. Male densities were higher in sparse woodlots, female densities at grassland patches. These intersexual differences in habitat use emphasise the need for heterogeneous vegetation.

  4. Like other inhabitants of traditional woodlands, E. aethiops suffers from canopy closure, leading to its retreat to transitional structures such as forest edges or abandoned grasslands. Such preferences are in conflict with regular grassland management, necessary for conserving many other grassland organisms. Therefore, sparse woodlands containing forest free enclaves should be restored to protect this and other woodland organisms.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgements
  9. References

Ecologists have traditionally defined animal habitats by prevailing vegetation type (Rodwell et al., 1995; Crofts & Jefferson, 1999), and this view persists in conservation policies. The increasingly detailed knowledge of vital requirements of diverse animal species reveals that this approach cannot be applied universally (Webb & Lott, 2006; Vanreusel & Van Dyck, 2007; Konvicka et al., 2008a). Mobile animals utilise diverse resources, and their habitats may encompass various landscape components or vegetation units. This has been convincingly demonstrated in butterflies, whose larvae use different resources than adults, and these resources vary during the season or among broods (Larsen, 1991; Dennis et al., 2003; Dennis, 2004; Vanreusel & Van Dyck, 2007). The specialised requirements of individual species are linked with butterfly losses from human-dominated landscapes, such as Central Europe (Kadlec et al., 2010; Ibbe et al., 2011; Stefanescu et al., 2011). The fact that many of the currently declining species were common several decades ago documents the severity of human-induced landscape changes, with far-reaching implications for our understanding of the history of European landscapes.

The poor match between traditional ecologists' and the animal view of natural habitats is particularly apparent in woodland butterflies, which frequently depend on fluid mosaics of patches in certain successional stages. A majority of woodland butterflies utilise early successional stages within woodlands, such as gaps and clearings or edges separating woodlands from other land-cover types (Sutcliffe et al., 1997; Bergman, 2001; Benes et al., 2006; Konvicka et al., 2008b). They often require heterogeneous vegetation with close proximity of structures in different successional ages. Thus, larvae and ovipositioning females of Lopinga achine (Scopoli, 1763), Nymphalidae: Satyrinae, spend much time within small gaps containing larval host plants, whereas males establish perches in rather dense understorey shrubs (Bergman, 2001). Larvae of Apatura spp. Fabricius, 1807, Nymphalidae: Apaturinae, emperors feed on low and sun-exposed tree saplings, whereas adult males perch on prominent tall trees (Benes et al., 2002). Adults of Pararge aegeria (Linnaeus, 1758), Nymphalidae: Satyrinae, utilise sunny spots within closed forest, moving from these gaps to close canopy with changing ambient temperature (Shreeve, 1984, 1987). Fine-grained microhabitat mosaics were common under traditional woodland uses, such as coppicing (repetitive felling on the same stump followed by spontaneous shoots regeneration), selective harvest and grazing, but became scarce under modern high-forest management (Warren & Key, 1991; Benes et al., 2006). Consequently, an increasing number of woodland species became restricted to ecotone structures or escaped to such sites as abandoned pastures or orchards, which succumb to shrubs and trees following abandonment (Fartmann, 2006; Clarke et al., 2011; Ellis et al., 2011). These habitats, consisting of trees and open patches patchworks, resemble traditionally used sparse forests.

Erebia aethiops (Esper, 1777; Lepidoptera: Nymphalidae) is one of the species illustrating the effects of woodland transformations. It is considered to be a forest inhabitant (Van Swaay et al., 2006), but populations also exist at various semi-open habitats such as sparsely wooded heathlands, moorlands and forest margins (Loertscher, 1991; Kirkland, 1995; Benes et al., 2002; Tolman & Lewington, 2008). Its conservation status is Least Concern for Europe (Van Swaay et al., 2010), but declines have occurred in many countries. It is Extinct in Luxembourg and Lithuania, Endangered in Belgium and Vulnerable in Germany and the Czech Republic (Van Swaay & Warren, 1999); the number of localities has declined sharply in the latter country (Slamova et al., 2010). Given the apparently wide habitat range reported in the literature, the declines appear surprising. The key to understanding E. aethiops' situation may be in the details of its habitat requirements.

This article analyses the habitat requirements of E. aethiops adults in the Czech Republic, evaluating the effects of vegetation management on its population density. We studied a strong population of E. aethiops in a large grasslands-forest mosaic nature reserve. The grasslands are either mown or pastured; the formerly sparse forest is mostly unmanaged. Diverse approaches to grassland management applied in the reserve enabled us to test their effects on E. aethiops population density. Our major goals were to (i) describe the demographic structure of the studied population; (ii) quantify basic mobility parameters; (iii) evaluate the effect of vegetation structure and distribution of resources on the butterfly density and mobility; (iv) link the effects of vegetation management to E. aethiops population density; and finally (v) solve the conundrum of why E. aethiops is in decline in the Czech Republic, despite its occurrence at ubiquitous habitats such as forest margins.

Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgements
  9. References

Study system

The site was the Vysenske kopce Reserve, SW Czech Republic (48°49′N, 14°18′E, altitude range 505–610 m). It protects 55 ha of a highly heterogeneous mosaic of xeric calcareous grasslands and woodlots, subject to various forms of management, ranging from rotational sheep and goat grazing and mowing to temporary neglect (grasslands) and from shrub reduction and coppicing again to neglect (forest). Adjoining the reserve are abandoned grasslands, which cover slopes of an abandoned limestone quarry and a former army-training field, as well as intensively mown meadows.

The reserve hosts regionally exceptional butterfly diversity (69 species, Hanc, 2005), including one of the largest populations of E. aethiops in the Czech Republic (Slamova et al., 2010). The focal species inhabits a temperate Palaearctic range, from Scotland to eastern Siberia, in altitudes of 300–2000 m. Its life cycle is univoltine, with hibernating larvae and adult flight lasting from mid-July to mid-August. Larvae develop on coarse grasses: Bromus erectus, Brachypodium pinnatum and Calamagrostis epigejos (Benes et al., 2002). Males patrol above tussocks of these grasses in search of females. E. aethiops distribution in low altitudes is exceptional in mostly cold-adapted Erebia. Within the study area, E. aethiops occurs in three clusters of sites, separated by larger stretches (600–1000 m) of intensive grasslands.

Data collecting

Intensive mark-recapture (MR) covered the entire E. aethiops flight period, between 17 July and 23 August 2007 (33 days when marking was possible, ca 200 person-hours). A total of 27 subsites varying in area (mean: 1982 m2, range: 128–6072 m2) and encompassing all types of open vegetation in the reserve were distinguished. Subsites were visited daily in a random order, usually from 9 a.m. to 5 p.m. Impenetrable woodlots and thickets, almost impossible to enter with a net, were excluded from MR. The butterflies were marked using individual numeric codes, their sex and capture subsite were recorded.

Environmental variables

Subsites were characterised by their geography, geomorphology, vegetation structure, nectar supply, borders structure and management (Table 1). Subsite area and longitude and latitude of subsite centroids were obtained using GIS-compatible aerial photographs, processed using arcview 9.2 software (ESRI, 1996). The slope was classified categorically as low (0–3°), medium (3–12°) or steep (12–17°). Exposure was characterised using a ranked 1–6 scale roughly describing the intake of solar energy (1-North, 2-Northeast, 3-East and Northwest, 4-Southeast and West, 5-South, 6-Southwest). Vegetation structure was described by the percentage cover of trees, shrubs and litter per subsite area. Solitary/grouped trees and short (<1.2 m)/high (>1.2 m) shrubs were also distinguished. Nectar supply was expressed as an average from a weekly estimated cover (% area) of pooled Centaurea spp., Knautia spp. and Origanum spp., which were the main E. aethiops nectar sources at the locality (Slamova et al., 2011). Subsite borders were characterised by the proportion of intensively/occasionally mown meadows, forest, and shrubs on the subsite perimeter. Management was classified as a categorical variable with four levels – intensively mown, occasionally mown, grazing and neglect.

Table 1. The list of environmental variables used to explain the variation of population densities, female proportion and mobility of Erebia aethiops among subsites within the study system
Variable typeVariableUnits
GeographyLongitudem
Latitudem
Aream2
GeomorphologySlopeLow, medium, high
Exposure1–6, see 'Methods' section
Vegetation structureSolitary treesCover of subsite (%)
Grouped treesCover of subsite (%)
Trees totalCover of subsite (%)
Short shrubsCover of subsite (%)
Tall shrubsCover of subsite (%)
Shrubs totalCover of subsite (%)
LitterCover of subsite (%)
Nectar supply Centaurea, Knautia, Origanum Cover of subsite (%)
Borders structureForestCover of subsite perimeter (%)
ShrubsCover of subsite perimeter (%)
Intensively mownCover of subsite perimeter (%)
Occasionally mownCover of subsite perimeter (%)
Management4 Management typesCategorical, see 'Methods' section

Data analysis

Population size and structure

Constrained linear models in mark 5.1 software (White & Burnham, 1999) were used to estimate the size and sex ratio of the studied population. The Jolly-Seber method, implemented in MARK: POPAN was used, because it is suitable for repeatedly sampled open populations (Schwarz & Arnason, 1996). The method assumes the existence of a hypothetical superpopulation and estimates daily population sizes (N i) and total population size (N tot) (Schtickzelle et al., 2002). The method returns three primary parameters; apparent daily survival (φ i), capture probability given the animal is alive and stays in the study area (p) and the probability of a new animal entering the population (p ent). These parameters may be constant for sexes and time (.), sex-dependent (g), factorially (t), linearly (T) or quadratically (T 2) dependent on marking day, exhibiting additive (t) or interactive (gt) effects of sex and time.

For model selection, we relied on the AIC-values corrected for small sample sizes (AICc) (Akaike, 1981; White & Burnham, 1999) values, reflecting the strength of evidence of the best model relative to other models in the set of models considered. We ran 22 models to find the best one describing the most faithful population size and structure. Following many other butterfly studies (e.g. Schtickzelle et al., 2003; Zimmermann et al., 2011), we presumed that (i) capture probabilities p fluctuate factorially during flight period and differ between sexes in an additive manner p(g + t); (ii) the probability of a new animal entering the population p ent depends on time in a domed manner, either additively p ent(g + T 2 ) or multiplicatively p ent(g*T 2 ); and finally, that (iii) the total population sizes differ between sexes, N(g). Constrained by these assumptions, we computed all possible alternative models, which therefore differed in responses of apparent daily survival, φ i, and daily recruitment, p ent, to sex, time and their interactions. We first fitted the global model {φ(g*t) p(g + t) p ent(g*T 2 ) N(g)}, with link functions sin for φ and p, Mlogit for p ent and identity for N. We subsequently simplified the model by removing interactions and additive parameters, and checking individual simplified models for linear (T) and quadratic (T 2) responses to time.

The average male and female longevities were obtained as −(lnφ)−1 (Tabashnik, 1980; Zimmermann et al., 2005), using models containing the same parameter combinations as the selected final models, but having φ set constant or were sex-dependent. Likelihood-ratio test was used to compare the fit of φ(.) and φ(g) models.

Mobility

Range size of E. aethiops individuals was characterised by the direct distance crossed by individuals between the first and the last capture. Individual trajectory lengths were estimated as the sum of the distances between centroids of subsequent capture sites. Calculations of the mean and median trajectory lengths included zeroes for individuals staying at a single patch. The dependence of the trajectory lengths (log10-transformed) on the time between the first and the last capture (in days) was tested using a generalised linear model (GLM, Gamma distribution, log link function).

We estimated parameters of the inverse power function (IPF) describing the decay of the probability of migration over increasing distance (Fric & Konvicka, 2007): lnI = lnCmlnD where I is migration probability density, D is distance (in kilometres) and C and m are the estimated parameters. The proportions of marked males and females overcoming different trajectory lengths were used to estimate the IPF parameters. Both the explanatory variable and the predictor values were ln-transformed and ordinary least squares regressions were fitted separately for males and females in R 2.9.0. (R Core Development Team, 2009). Regression slopes m were then compared between sexes using the t-test. The distribution of the migration probabilities differed significantly from the normal distribution (Shapiro–Wilk normality test, = 0.939, = 0.0001). Hence, we used bootstrapped t-test (1000 replicates) to test the significance of differences of slopes from zero and to test the difference of slopes in males and females.

Effect of vegetation structure and subsite area

Per-subsite response variables were male and female density (log10-transformed), proportion of females, and fractions of emigrants, immigrants and residents. Male and female densities were defined as the numbers of individuals recorded during the whole study period at individual subsites (standardised by the number of visits of a subsite) per hectare. Female proportion was calculated as the number of females divided by the pooled number of both sexes recorded at a given subsite, i.e. F/(M + F). Raw data about numbers of subsite residents (R), immigrants (I) and emigrants (E) were transformed to the fraction of residents, R/(I), emigrants, E/(E), and immigrants, I/(I) (e.g. Sutcliffe et al., 1997; Bergman & Landin, 2001).

Generalised linear models (GLM) in R 2.9.0 (Faraway, 2006) were used to investigate the effects of vegetation structures on population densities (normal distribution for log10-transformed data), female proportion (quasi-binomial distribution) and fractions of emigrants, immigrants and residents (quasi-binomial distribution). First, forward selection of nuisance covariables, i.e. area (log10-transformed), longitude, latitude, slope and exposure was performed. Afterwards, the effect of vegetation structures was tested (all variables log10-transformed, significant covariables retained in the model).

The same response variables and GLM procedures were used to test the independent effect of area (log10-transformed). Forward selection of all potential covariables (longitude, latitude, slope, exposure and vegetation structures) was performed to control for confounding effect of possible correlation between subsite area and other explanatory variables.

Effect of population density on mobility

We tested whether or not the fractions of emigrants, immigrants and residents are affected by population density [log10(x + 1)-transformed; separately for sexes and sexes combined] using GLM. All environmental variables which significantly affected the fractions of residents, emigrants and immigrants (i.e. the cover of grouped trees for both sexes, cover of litter for females and nectar supply for males) were included in the models.

Effect of site management

The effect of site management on population densities (log10-transformed) and proportion of females was tested using GLM (normal distribution for log10-transformed population densities and quasi-binomial distribution for the proportion of females). The link between management and vegetation structure was investigated using the Redundancy analysis (RDA), a linear constrained multivariate ordination method, with subsites as samples and vegetation structures as dependent variables (centred, standardised and log10 (x + 1)-transformed) in canoco 4.5.0. (ter Braak & Smilauer, 1998).

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgements
  9. References

Population size and structure

We marked 541 males and 365 females; 309 males and 174 females were recaptured. Total numbers of recapture events were 736 for males and 369 for females (including eight dead individuals caught by crab spiders and two couples marked in copula).

The changes of population size during the season were most appropriately described by POPAN model {φ(g*T 2 ) p(g + t) p ent(g*T 2 ) N(g)}. All other candidate models differed by ΔAIC > 4.0 (Table 2). According to the best model, apparent daily survival depended on time quadratically, being higher on the beginning of flight period and subsequently decreasing more rapidly in males than in females. Daily values of capture probability were higher in males. Probability of entering the population was affected by a multiplicative effect of sex and time: for males, it was highest at the beginning of flight period, and declined gradually, whereas females started emerging at the beginning of the study period, peaking in the middle of flight season. This multiplicative effect corresponded with a protandric adult recruitment, with sex ratio balanced in the middle of the flight period (Fig. 1). The population contained 820 (±28 SE, 95% CI 773–885) males and 610 (±28 SE, 95% CI 557–667) females.

image

Figure 1. Daily estimates of Erebia aethiops population size (±SE) of males (black circles) and females (empty circles) estimated by the most appropriate Jolly-Seber model. See Table 2 for model specification.

Download figure to PowerPoint

Table 2. Overview of the mark-recapture models (MARK: POPAN) used to describe the population structure of Erebia aethiops in The Vysenske kopce Reserve and its environs. Models are ordered according to decreasing AICc values. Note that only nine of 22 models considered, plus a global model, are presented herein
ModelAICcΔAICcAICc weightModel likelihoodParametersMale N tot (SE)Female N tot (SE)
φ(g*T 2) p(t) p ent(g*T 2) N(g)6676.5100.861148829.11 (28.60)612.04 (28.28)
φ(g + T 2) p(t) p ent(g*T 2) N(g)6680.644.1290.1090.12746828.46 (27.98)598.12 (26.36)
φ(T 2) p(t) p ent(g*T 2) N(g)6683.486.9690.0260.03145819.85 (27.25)618.42 (26.34)
φ(T) p(t) p ent(g*T 2) N(g)6688.5312.0200.0020.00345823.87 (28.40)608.01 (27.17)
φ(g*T) p(t) p ent(g*T 2) N(g)6690.9214.4120.0000.00146812.65 (26.04)612.47 (28.24)
φ(t) p(t) p ent(g*T 2) N(g)6692.4115.9000.0000.00076815.29 (26.91)595.03 (25.80)
φ(T) p(t) p ent(g*T 2) N(g)6692.4515.9370.0000.00044807.98 (25.50)623.51 (26.68)
φ(.) p(t) p ent(g*T 2) N(g)6696.3319.8200.0000.00043828.15 (25.82)614.92 (25.65)
φ(g) p(t) p ent(g*T 2) N(g)6696.7420.2280.0000.00044833.20 (27.03)609.26 (27.04)
Global model
 φ(g*t) p(t) p ent(g*T 2) N(g)6738.1761.6630.0000.000107811.36 (27.29)611.28 (29.03)

Subsequent well-fitting models still retained the domed pattern in recruitment, either responding additively to sex and time (second model) or not differing between sexes (third model). Models with linear, factorial or time-independent survival performed considerably worse. However, the estimates of total population sizes from those subsequent models were almost identical to that from the best fitting model (Table 2).

Using the model {φ(g) p(g + t) p ent(g*T 2) N(g)} for estimating the average longevity returned φ values 0.851 (±0.0071 SE, 95% CI 0.837–0.865) for males and 0.863 (±0.0096 SE, 95% CI 0.843–0.881) corresponding to longevities 6.2 and 6.8 days, while the model {φ(.) p(t) p ent(g*T 2) N(g)}, i.e. a model with longevity identical for sexes, returned φ value 0.855 (±0.0058 SE, 95% CI 0.844–0.866), corresponding to a middle value of 6.4 days. The maximum intervals between the first and the last individual capture were 20 (a male) and 19 (a female) days, providing rough estimates of maximum longevities.

Mobility

Maximum range sizes were 1982 m for a male and 2101 m for a female. Mean (±SE)/median range sizes were 164 m (±305)/85 m in males and 137 m (±223)/85 m in females. In both sexes, a majority of individuals had small ranges, <250 m for more than 80% individuals (Fig. 2). The large stretches of unsuitable habitat among the three separate clusters of subsites were crossed by 13 males and 3 females.

image

Figure 2. Histograms of relative frequencies of Erebia aethiops (a) male and (b) female individual range sizes, i.e. the maximum effective distances, as observed during our mark-recapture study.

Download figure to PowerPoint

The longest detected trajectories were 2197 m (a male) and 2289 m (a female). Mean (±SE) trajectory was 210 m (±415.2) in males and 169 m (±327.5) in females. Median trajectories were 85 m for both sexes. The trajectory length increased with the length of the time interval between the first and the last handling occasion (males: F 1,306  = 18.50, < 0.001, b = 1.10; females: F 1,174  = 5.73, < 0.05, b = 0.84).

According to IPF, the rate of decrease of probability of migration with distance (Fig. 3) did not differ between sexes (bootstrapped t-test, P = 0.133). The radius of the area which can be colonised from this population is a few kilometres. For example, the probability of migration over 5 km is 0.0046 in males and 0.0024 in females, corresponding to 3.7 male and 1.5 female individuals, given the estimated population size.

image

Figure 3. Inverse power function describing the probability density of Erebia aethiops movements (I) to particular distances (D); separately for males (full circles and solid line) and females (empty circles and dashed line). The fitted function equations are, for males: ln(I) = −3.199 −1.356lnD (F 1,66 = 813.46, P < 0.001); and for females: ln(I) = −3.643 −1.483lnD (F 1,35 = 560.74, P < 0.001). The axes have logarithmic scale.

Download figure to PowerPoint

Effect of subsite area and vegetation

No spatial trends, neither for population densities nor for the fractions of migrants and residents, were identified. Therefore, only the effects of area and exposure entered the models for the effects of vegetation structure (Table 3). Male densities were positively related to the cover of trees and to forested borders of subsites. Female densities increased with higher cover of solitary trees and, as in males, with more forested borders. The proportion of females in the population was lower on subsites with higher cover of grouped trees, short shrubs and intensively mown borders on subsite perimeter (Table 3).

Table 3. Results of tests for the effects of vegetation structures on Erebia aethiops population densities and female proportion. Significant covariables were included to the models before testing the effects of individual vegetation structure variables. Nominally significant effects are printed in bold. d.f. = number of degrees of freedom. For regression slopes, b, the +/− signs denote the direction of the relationships
 Male densityFemale densityFemale proportion
d.f. F P bd.f. F P bd.f F P b
Covariables
Longitude1,230.000.955 1,230.630.437 1,250.020.886 
Latitude1,231.470.238 1,231.160.293 1,250.780.387 
Area 1,24 10.30 0.004 1,24 12.31 0.002 1,251.040.318 
Slope1,230.0190.892 1,230.4010.533 1,252.9990.096 
Exposure 1,24 7.73 0.010 + 1,24 5.92 0.023 +1,252.030.167 
Vegetation structure
Solitary trees1,211.780.197  1,22 6.10 0.022 +1,220.020.886 
Grouped trees1,210.360.557 1,210.050.826  1,23 14.71 0.001
Trees total 1,22 13.00 0.002 + 1,210.140.712 1,220.010.938 
Short shrubs1,210.430.520 1,211.780.197  1,23 17.22 0.001
Tall shrubs1,210.000.999 1,210.010.935 1,220.710.409 
Shrubs total1,210.050.821 1,212.390.137 1,220.170.682 
Litter1,210.110.745 1,210.910.350 1,220.250.621 
Nectar supply1,212.240.150 1,213.860.063 1,220.420.522 
Forested borders 1,22 15.68 0.001 + 1,22 14.57 0.001 +1,220.060.807 
Shrubby borders1,210.230.640 1,212.170.156 1,220.000.958 
Intensively mown borders1,212.230.150 1,212.830.107  1,23 8.50 0.008
Occasionally mown borders1,210.560.461 1,211.260.274 1,221.170.290 

Basic mobility parameters were unaffected by any potential covariables (Table 4). Fractions of emigrants and immigrants decreased and the fraction of residents increased with the cover of grouped trees in both sexes. In males, higher nectar supply led to an increase in the fraction of emigrants. In females, higher cover of litter led to a decrease in the fraction of emigrants and immigrants, while the fraction of residents increased (Table 4).

Table 4. Effects of vegetation structures on Erebia aethiops fractions of emigrants, immigrants and residents, separately for males and females. As no covariable was nominally significant, none was included to the models for individual vegetation structures. Nominally significant effects are printed in bold. d.f. = number of degrees of freedom. For regression slopes, b, the +/− signs denote the direction of the relationships
 SexFraction of emigrantsFraction of immigrantsFraction of residents
d.f. F P bd.f. F P bd.f. F P b
Covariables
LongitudeMale1,250.230.635 1,240.350.558 1,250.260.617 
Female1,230.780.387 1,230.530.473 1,230.600.445 
LatitudeMale1,251.490.233 1,242.160.155 1,251.640.213 
Female1,233.890.061 1,234.200.052 1,234.110.054 
AreaMale1,251.260.272 1,241.030.321 1,251.010.326 
Female1,230.180.673 1,230.430.519 1,230.260.613 
SlopeMale1,252.080.162 1,240.540.468 1,251.010.325 
Female1,231.690.207 1,230.740.400 1,231.060.313 
ExposureMale1,252.850.104 1,242.770.109 1,252.190.152 
Female1,230.080.779 1,231.380.252 1,230.480.494 
Vegetation structures
Solitary treesMale1,233.820.063 1,230.770.388 1,241.020.323 
Female1,210.620.441 1,212.650.118 1,211.220.281 
Grouped treesMale 1,24 8.55 0.007 1,24 7.04 0.014 1,25 6.15 0.020 +
Female 1,22 14.00 0.001 1,22 20.70 0.000 1,22 17.58 0.000 +
Trees totalMale1,231.960.175 1,230.320.576 1,240.420.521 
Female1,210.280.600 1,211.090.308 1,210.460.503 
Short shrubsMale1,230.000.986 1,230.040.835 1,240.010.922 
Female1,210.070.797 1,210.400.536 1,210.090.771 
Tall shrubsMale1,231.090.307 1,230.320.577 1,240.350.559 
Female1,210.000.968 1,210.670.422 1,210.170.686 
Shrubs totalMale1,230.500.487 1,230.140.713 1,240.310.586 
Female1,210.020.881 1,210.520.479 1,210.100.751 
LitterMale1,230.090.771 1,230.000.999 1,240.000.954 
Female 1,22 7.09 0.014 1,22 7.54 0.012 1,21 7.92 0.010 +
Nectar supplyMale 1,24 4.44 0.046 + 1,231.820.191 1,242.830.105 
Female1,211.380.253 1,210.680.418 1,210.930.345 
Forested bordersMale1,230.150.702 1,230.230.639 1,240.130.717 
Female1,212.300.144 1,210.460.506 1,211.070.313 
Shrubby bordersMale1,231.760.198 1,232.290.144 1,240.690.413 
Female1,210.570.458 1,210.210.648 1,210.510.484 
Intensively mown bordersMale1,230.300.587 1,231.440.242 1,240.410.528 
Female1,210.240.632 1,210.070.789 1,210.110.744 
Occasionally mown bordersMale1,230.310.583 1,231.620.216 1,241.130.298 
Female1,211.480.237 1,212.270.147 1,211.820.192 

The negative effect of increasing subsite area on population densities was significant in both sexes (males: = 38.08, P < 0.001, b = −0.85; females: F = 31.10, P < 0.001, b = −0.85); significant vegetation structures were included in the model as covariables. Still, there was no effect of subsite area on the proportion of females and fractions of emigrants, immigrants and residents (all > 0.05).

Effect of site management

Management affected local density of males (F 3,23  = 6.52, = 0.002) and females (F 3,23  = 6.86, = 0.002) by the same manner (Fig. 4), whereas the local proportion of females was not significantly affected (F 3,23  = 2.42, = 0.093). The highest densities of both sexes were recorded on neglected, occasionally mown and grazed subsites, in descending order. Intensively mown subsites supported the lowest densities.

image

Figure 4. Effects of subsite management on Erebia aethiops (a) male and (b) female local densities (mean ± SE); the y-axis has logarithmic scale.

Download figure to PowerPoint

Vegetation structure was linked with subsite management (RDA: F = 3.05, P < 0.001) (Fig. 5). Grazed and abandoned subsites tended to contain shrubs and trees. Solitary trees, litter and Origanum spp. were more abundant on pastures, whereas clumps of trees and Centaurea spp. were common rather on neglected meadows. Occasionally mown meadows resembled the neglected ones. Intensively mown meadows with occurrence of E. aethiops contained Knautia spp. nectar sources, but otherwise lacked other vegetation structures used by the butterfly, such as shrubs and trees.

image

Figure 5. Redundancy Analysis (RDA) ordination biplot showing the relationship between subsite management type (full squares) and vegetation structure (arrows) in Vysenske kopce Reserve, the locality of our Erebia aethiops population.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgements
  9. References

Population size and structure

In terms of adult demography, the studied E. aethiops population shared several features with mountain Erebia congeners occurring in the Czech Republic (Cizek et al., 2003; Kuras et al., 2003) as well as with a mountain E. aethiops population from the Alps (Loertscher, 1991). In all these cases, adult recruitment is protandrous, with the maximum abundance and balanced sex ratio in the middle of the flight period. A difference in relative timing of male and female emergence at different altitudes was reported in some butterflies with broad altitudinal range, such as in Euphydryas aurinia (Rottemburg, 1775), whose lowland populations are protandric (Zimmermann et al., 2011), whereas sexes appear synchronously at high altitudes (Junker et al., 2010). In contrast, Erebia butterflies seem to be protandrous across altitudes (Loertscher, 1991; Cizek et al., 2003; Kuras et al., 2003; Slamova & Klecka unpubl. data). The studied E. aethiops population contained approximately 1400 individuals, similar to a population inhabiting a network of mountain meadows in Switzerland (Loertscher, 1991), whereas populations containing tens of thousands of individuals may exist at large areas of suitable habitat in England (Kirkland, 1995). Apparently, the population sizes reflect the available habitat area.

Habitat requirements and mobility

E. aethiops adults utilise both open and wooded patches, and its sexes differ in habitat use. Adults reach the highest densities on small open enclaves within forests and the lowest densities at large meadows and pastures with no scattered trees. Female densities were higher than male densities at the open grasslands with solitary trees, and male densities were higher at shrubby sites (Fig. 6). At the open enclaves surrounded by forest, the resources essential for males (shade, mating substrates) and females (nectar and larval sources) overlapped spatially, in contrast to large pastures or meadows where especially male resources were scarce.

image

Figure 6. (a) Grassland patch with nectar supply of Origanum spp. and Centaurea spp. preferred by females and (b) abandoned pasture with solitary trees, habitat favoured by males of Erebia aethiops (photo I. Slamova).

Download figure to PowerPoint

The intersexual differences in habitat use reflect the associations of individual activities with vegetation structures, investigated by Slamova et al. (2011). Males stay most of the time at the partially shaded clearings to avoid overheating, whereas females nectar, bask and oviposit at grasslands with solitary trees and a high litter accumulation. Mating takes place on shrubs. Females tolerate warmer ambient temperatures, probably because they periodically cool down while laying eggs at bases of grass tussocks (Slamova et al., 2011); they also likely profit from lighter colouration and higher weight (cf. Heinrich, 1986; Konvicka et al., 2002). The female tolerance for warmer microclimate, and resulting association with warmer sites, is likely advantageous during cooler days of adult flight, enhancing their fertility by prolonging oviposition time (cf. Gossard & Jones, 1977; Karlsson & Wiklund, 2005). It also provides an advantage for larvae, which should profit from the warmer microclimate at such sites in autumn and spring. There is a trade-off involved, however, due to desiccation risk at the warmest spots of the calcareous locality. E. aethiops larvae require a humid and stable microclimate, best provided by a litter in older grass tussocks (Leopold, 2006).

Sex-related differences in habitat use were previously reported for at least two other woodland butterflies. In a cage experiment with artificial vegetation, Leimar et al. (2003) observed that females of L. achine and P. aegeria preferred open parts of the cage, whereas males preferred shaded ones. Such intersexual differences in habitat perception might represent a common but overlooked phenomenon in butterflies living in structurally diverse habitats and utilising different structures for different activities (Dennis et al., 2003; Vanreusel & Van Dyck, 2007). Different motivations affect female and male habitat choice. Females strive to maintain an optimum body temperature (e.g. Karlsson & Wiklund, 2005), maximise nectar intake to increase fecundity (e.g. Mevi-Schutz & Erhardt, 2003) and locate egg-laying sites. Males invest mainly in mating activities, such as locating virgin females and defending them against rivals (e.g. Rutowski, 1991).

Despite different microhabitat preferences of E. aethiops sexes, both sexes display similar range size and movement probabilities to longer distances. The majority of individuals stayed within a subsite or crossed only short distances (Fig. 2), similar to the mountain congeners E. epiphron (Knoch, 1783) and E. sudetica (Staudinger, 1861) (Kuras et al., 2003). The closest locality occupied by another large E. aethiops population is within the Boletice military area, 5 km westerly from our study system, well within the range of predicted movement distances. The ability to reach more remote locations, however, likely depends on permeability of the landscape matrix (Hanski, 1998; Ricketts, 2001; Baguette & Van Dyck, 2007; Lange et al., 2010). We did not detect, however, any effect of subsite area or borders on E. aethiops movement patterns. We admit, however, that we delimited habitat boundaries somewhat deliberately, which is often unavoidable in complex environments, particularly if the animals use diverse structures, change activity patterns in time (Cizek & Konvicka, 2005; Vanreusel & Van Dyck, 2007) or, as in E. aethiops, differ in space utilisation between sexes.

Relating mobility to habitat structure, both sexes tended to stay at subsites with sparsely growing trees, and females at subsites with accumulated litter. In addition, males more frequently emigrated from flowery patches, perhaps due to increased intrasexual interactions with other males, or while alternating between sun-exposed and shaded patches (cf. Slamova et al., 2011). While marking, we often noticed males crossing forest along narrow paths. Females left some subsites isolated by forest too, but they probably did so by direct flight, because we observed only a few female individuals within these shady areas.

Reserve management and conservation

Abandoned grassland patches, containing solitary trees, scattered shrubs and high accumulation of litter, supported the highest butterfly densities, in line with the requirements for oviposition (litter-rich tufts) mating (shrubby structures) and overheating prevention (shade). Grassland abandonment, however, is not a viable management option in the long term, as unmanaged grasslands eventually turn to scrub. Occasionally mown or lightly grazed patches hosted the second-highest E. aethiops densities, and alteration of these management options with temporary abandonment appears as a suitable compromise, keeping the habitat open while still supporting the study species. Intensively mown grasslands, on the other hand, supported the lowest densities, likely due to the absence of litter, temporary nectar shortages, and perhaps direct mortality caused by mowing (Dover et al., 2010; Cizek et al., 2012). All the above observations point to E. aethiops' requirements for a highly diversified vegetation structure with alteration of shaded, half-shaded and sun-exposed spots, with both managed grassland patches offering nectar and neglected patches with litter-rich grass tussocks for oviposition. Such habitats traditionally existed in open woodlands, which have virtually disappeared from many regions of Europe (Spitzer et al., 2008; Warren & Bourn, 2011).

At least two other open woodland butterflies, Hamearis lucina (Linnaeus, 1758) and Leptidea sinapis (Linnaeus, 1758), have locally retracted from shady woodlands to abandoned shrub-encroached meadows (Fartmann, 2006; Clarke et al., 2011). This also applies to the studied E. aethiops population, as open woodlots are scarce within the Vysenske kopce Reserve, and the species reaches the highest densities at shrubby grasslands. On the other hand, shrub-encroached grasslands are unsuitable for several other priority butterflies occurring in the reserve and inhabiting short-sward grasslands, such as Polyommatus coridon (Poda, 1761) and Pyrgus trebevicensis (Warren, 1926) (Hanc, 2005). Therefore, instead of allowing the grasslands to turn into scrub, E. aethiops population should be supported by managing the forests towards more open conditions (see Turner et al., 2009; for an analogous situation). Gradual formation of small coppiced panels within local forests would create a continuity of diverse successional stages not only for butterflies, but also for other invertebrates, birds, plants and fungi (Warren & Bourn, 2011). To increase the connectivity with other E. aethiops localities on a regional scale, stepping-stone sites could be formed by managing wider transitional zones between forests and grasslands in the largely cultivated landscape.

Conclusions

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgements
  9. References

Our study provides comprehensive information about population size and structure, habitat preferences and mobility of a lowland population of E. aethiops, a species vulnerable in many parts of Central and Western Europe. Although including data only from one season, and insect numbers fluctuate widely among years, the main conclusions of our study (adult habitat preferences, viable population size) should be robust against such variation. In the Czech Republic, E. aethiops represents a species of sparse woodlands, retracting to abandoned grasslands and forest edges. Its requirement for a high heterogeneity of habitat conditions is emphasised by different patterns of resource use in males and females. Although our studied population retracted to grassland patches, intensively managed grasslands do not satisfy some of its vital demands. In modern landscapes, E. aethiops thus faces double trouble: forests are too shady and grasslands are too open. To safeguard its further existence, local restoration of such traditional woodland use as coppicing or woodland pasture is desirable.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgements
  9. References

We are obliged to the Blansky Les Protected Landscape Area administration (R. Janak, T. Rejnkova, Z. Hanc) for permits to work in the nature reserve, for logistic help. J. Danko, T. Dudikova, D. Hisem, J. Kadrman, Z. Karova, V. Pouska, A.Vitova and P. Vlasanek helped in the field, K. Zimmermann helped with MARK modelling, and two anonymous reviewers contributed valuable comments. Funding was provided by the Grant Agency of the Czech Republic (P505/10/2248, P505/10/1630, 206/08/H044), Czech Ministry of Education (LC-06073, MSM 6007665801) and the Grant Agency of the University of South Bohemia (144/2010/P, 145/2010/P, 106/2010/P, 135/2010/P).

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgements
  9. References
  • Akaike, H. (1981) Likelihood of a model and information criteria. Journal of Econometrics, 16, 314.
  • Baguette, M. & Van Dyck, H. (2007) Landscape connectivity and animal behavior: functional grain as a key determinant for dispersal. Landscape Ecology, 22, 1171129.
  • Benes, J. , Cizek, O. , Dovala, J. & Konvicka, M. (2006) Intensive game keeping, coppicing and butterflies: the story of Milovicky Wood, Czech Republic. Forest Ecology and Management, 237, 353365.
  • Benes, J. , Konvicka, M. , Dvorak, J. , Fric, Z. , Havelda, Z. , Pavlicko, A., , Vrabec, V. & , Weidenhoffer, Z. (eds) (2002) Butterflies of the Czech Republic: Distribution and Conservation I., II. SOM, Prague, Czech Republic.
  • Bergman, K.O. (2001) Population dynamics and the importance of habitat management for conservation of the butterfly Lopinga achine . Journal of Applied Ecology, 38, 13031313.
  • Bergman, K.O. & Landin, J. (2001) Distribution of occupied and vacant sites and migration of Lopinga achine (Nymphalidae: Satyrinae) in a fragmented landscape. Biological Conservation, 102, 183190.
  • ter Braak, C.J.F. & Smilauer, P. (1998) CANOCO Reference Manual and User's Guide to Canoco for Windows: Software for Canonical Community Ordination (version 4). Microcomputer Power, Ithaca, New York.
  • Cizek, O. , Bakesova, A. , Kuras, T. , Benes, J. & Konvicka, M. (2003) Vacant niche in alpine habitat: the case of an introduced population of the butterfly Erebia epiphron in the Krkonoše Mountains. Acta Oecologica, 24, 1523.
  • Cizek, O. & Konvicka, M. (2005) What is a patch in a dynamic metapopulation? Mobility of an endangered woodland butterfly, Euphydryas maturna. Ecography, 28, 791800.
  • Cizek, O. , Zamecnik, J. , Tropek, R. , Kocarek, P. & Konvicka, M. (2012) Diversification of mowing regime increases arthropods diversity in species-poor cultural hay meadows. Journal of Insect Conservation, 16, 215226.
  • Clarke, S.A. , Wollen, K. , Green, D.G. , Butler, I. & Joy, J. (2011) Leptidea sinapis (Wood White butterfly) egg-laying habitat and adult dispersal studies in Herefordshire. Journal of Insect Conservation, 15, 2335.
  • Crofts, A. & Jefferson, R.G. (1999) The Lowland Management Handbook, 2nd edn. English Nature, The Wildlife Trusts, Peterborough, UK.
  • Dennis, R.L.H. (2004) Butterfly habitats, broad-scale biotope affiliations, and structural exploitation of vegetation at finer scales: the matrix revisited. Ecological Entomology, 29, 744752.
  • Dennis, R.L.H. , Shreeve, T.G. & Van Dyck, H. (2003) Towards a functional resource-based concept for habitat: a butterfly biology viewpoint. Oikos, 102, 417426.
  • Dover, J.W. , Rescia, A. , Fungarino, S. , Fairburn, J. , Carey, P. , Lunt, P. , Dennis, R.L.H. & Dover, C.J. (2010) Can hay harvesting detrimentally affect adult butterfly abundance? Journal of Insect Conservation, 14, 413418.
  • Ellis, S. , Wainwright, D. , Bulman, C. , Bourn, N. & Berney, F. (2011) Landscape-scale conservation in practice: lessons from northern England, UK. Journal of Insect Conservation, 15, 6981.
  • ESRI (1996) ArcView, Environmental Systems Research Institute, Redlands, California.
  • Faraway, J.J. (2006) Extending the linear model with R: generalized linear, mixed effects and nonparametric regression models.http://www.maths.bath.ac.uk/~jjf23/Chapman & Hall/CRC, Boca Raton, Florida, 301 pp.
  • Fartmann, T. (2006) Oviposition preferences, adjacency of old woodland and isolation explain distribution of the Duke of Burgundy butterfly (Hamearis lucina) in calcareous grasslands in central Germany. Annales Zoologici Fennici, 43, 335&#x2013;347.
  • Fric, Z. & Konvicka, M. (2007) Dispersal kernels of butterflies: power-law functions are invariant to marking frequency. Basic and Applied Ecology, 8, 377&#x2013;386.
  • Gossard, T.W. & Jones, R.E. (1977) The effects of age and weather on egg-laying in Pieris rapae L. Journal of Applied Ecology, 14, 65&#x2013;71.
  • Hanc, Z. (2005) Denni motyli (Lepidoptera: Papilionoidea, Hesperioidea), Narodni prirodni rezervace Vysenske kopce. Klapalekiana, 41, 33&#x2013;42. In Czech Republic.
  • Hanski, I. (1998) Metapopulation dynamics. Nature (London), 396, 41&#x2013;49.
  • Heinrich, B. (1986) Comparative thermoregulation of four montane butterflies of different mass. Physiological Zoology, 59, 616&#x2013;626.
  • Ibbe, M. , Milberg, P. , Tuner, A. & Bergman, K.O. (2011) History matters: impact of historical land use on butterfly diversity in clear-cuts in a boreal landscape. Forest Ecology and Management, 261, 1885&#x2013;1891.
  • Junker, M. , Wagner, S. , Gros, P. & Schmitt, T. (2010) Changing demography and dispersal behaviour: ecological adaptations in an alpine butterfly. Oecologia, 164, 971&#x2013;980.
  • Kadlec, T. , Vrba, P. , Kepka, P. , Schmitt, T. & Konvicka, M. (2010) Tracking the decline of the once-common butterfly: delayed oviposition, demography and population genetics in the Hermit, Chazara briseis . Animal Conservation, 13, 172&#x2013;183.
  • Karlsson, B. & Wiklund, C. (2005) Butterfly life-history and temperature adaptations: dry open habitats select for increased fecundity and longevity. Journal of Animal Ecology, 74, 99&#x2013;104.
  • Kirkland, K.L. (1995) A review of the distribution, ecology and behaviour of the Scotch Argus. The Bulletin of British Ecological Society, 26, 95&#x2013;102.
  • Konvicka, M. , Benes, J. , Cizek, O. , Kopecek, F. , Konvicka, O. & Vitaz, L. (2008a) How too much care kills species: Grassland reserves, agri-environmental schemes and extinction of Colias myrmidone (Lepidoptera: Pieridae) from its former stronghold. Journal of Insect Conservation, 12, 519&#x2013;525.
  • Konvicka, M. , Benes, J. & Kuras, T. (2002) Behaviour and microdistribution of two sympatric alpine butterflies (Erebia epiphron and E. euryale): relation to weather, vegetation and time of day. Biologia, 57, 223&#x2013;233.
  • Konvicka, M. , Novak, J. , Benes, J. , Fric, Z. , Bradley, J. , Keil, P. , Hrcek, J. , Chobot, J. & Marhoul, P. (2008b) The last population of the Woodland Brown butterfly (Lopinga achine) in the Czech Republic: habitat use, demography and site management. Journal of Insect Conservation, 12, 549&#x2013;560.
  • Kuras, T. , Benes, J. , Fric, Z. & Konvicka, M. (2003) Dispersal patterns of endemic alpine butterflies with contrasting population structures: Erebia epiphron and E. sudetica . Population Ecology, 45, 115&#x2013;123.
  • Lange, R. , Durka, W. , Holzhauer, S.I. , Wolters, V. & Diek&#x00F6;tter, T. (2010) Differential threshold effects of habitat fragmentation on gene flow in two widespread species of bush crickets. Molecular Ecology, 19, 4936&#x2013;4948.
  • Larsen, T.B. (1991) The Butterflies of Kenya and their Natural History. Oxford University Press, Oxford, UK.
  • Leimar, O. , Norberg, U. & Wiklund, C. (2003) Habitat preference and habitat exploration in two species of satyrine butterflies. Ecography, 26, 474&#x2013;480.
  • Leopold, P. (2006) Die Larval&#x00F6;kologie des Waldteufels (Erebia aethiops) in Nordrhein-Westfalen und deren Bedeutung f&#x00FC;r den Erhalt der Art. Larval&#x00F6;kologie von Tagfaltern und Widderchen in Mitteleuropa (ed. by T. Fartmann and G. Hermann ), Vol. 68pp. 61&#x2013;82. Westf&#x00E4;lisches Museum, Monster, Germany.
  • Loertscher, M. (1991) Population biology of two satyrine butterflies, Erebia meolans (De Prunner, 1798) and Erebia aethiops (Esper, 1777) (Lepidoptera: Satyridae). Nota lepidopterologica, 2, 22&#x2013;31.
  • Mevi-Schutz, J. & Erhardt, A. (2003) Effect of nectar amino-acids on fecundity of the wall brown butterfly (Lasiommata megera L.). Basic and Applied Ecology, 4, 413&#x2013;421.
  • R Core Development Team (2009) R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria. &#x003C;http://www.R-project.org> 1st October 2011.
  • Ricketts, T.H. (2001) The matrix matters: effective isolation fragmented landscapes. The American Naturalist, 158, 8799.
  • Rodwell, J.S. , Pignatti, S. , Mucina, L. & Schaminee, J.H.J. (1995) European vegetation survey – update on progress. Journal of Vegetation Science, 6, 759762.
  • Rutowski, R.L. (1991) The evolution of male mate-locating behavior in butterflies. American Naturalist, 138, 11211139.
  • Schtickzelle, N. , Baguette, M. & Le Boulenge, E. (2003) Modelling insect demography from capture-recapture data: comparison between the constrained linear models and the Jolly-Seber analytical method. Canadian Entomologist, 135, 313323.
  • Schtickzelle, N. , Le Boulenge, E. & Baguette, M. (2002) Metapopulation dynamics of the bog fritillary butterfly: demographic processes in a patchy population. Oikos, 97, 349360.
  • Schwarz, C.J. & Arnason, A.N. (1996) A general methodology for the analysis of capture-recapture experiments in open populations. Biometrics, 52, 860873.
  • Shreeve, T.G. (1984) Habitat selection, mate location, and microclimate constraints on the activity of the speckled wood butterfly Pararge aegeria . Oikos, 42, 371377.
  • Shreeve, T.G. (1987) The mate location behaviour of the speckled wood butterfly, Pararge aegeria, and the effect of phenotypic differences in hind-wing spotting. Animal Behaviour, 35, 682690.
  • Slamova, I. , Klecka, J. & Konvicka, M. (2011) Diurnal behavior and habitat preferences of Erebia aethiops, an aberrant lowland species of a mountain butterfly clade. Journal of Insect Behavior, 24, 230246.
  • Slamova, I. , Spitzer, L. & Konvicka, M. (2010) Kde unas preziva okac klubenkovy? Vyznam stanovistni mozaiky pro ustupujiciho motyla. Ziva, 2010, 3234. In Czech Republic.
  • Spitzer, L. , Konvicka, M. , Benes, J. , Tropek, R. , Tuf, I.H. & Tufova, J. (2008) Does closure of traditionally managed open woodlands threaten epigeic invertebrates? Effects of coppicing and high deer densities. Biological Conservation, 141, 827837.
  • Stefanescu, C. , Carnicer, J. & Penuelas, J. (2011) Determinants of species richness in generalist and specialist Mediterranean butterflies: the negative synergistic forces of climate and habitat change. Ecography, 34, 353363.
  • Sutcliffe, O.L. , Thomas, C.D. & Peggie, D. (1997) Area-dependent migration by ringlet butterflies generates a mixture of patchy population and metapopulation attributes. Oecologia, 109, 229234.
  • Tabashnik, B.E. (1980) Population structure of pierid butterflies, 3. Pest population of Colias philodice eriphyle. Oecologia, 47, 175183.
  • Tolman, T. & Lewington, R. (2008) Collins Butterfly Guide: The most Complete Field Guide to the Butterflies of Britain and Europe. HarperCollins Publishers, London, UK.
  • Turner, E.C. , Granroth, H.M.V. , Johnson, H.R. , Lucas, C.B.H. , Thompson, A.M. , Froy, H. , German, R.N. & Holdgate, R. (2009) Habitat preference and dispersal of the Duke of Burgundy butterfly (Hamearis lucina) on an abandoned chalk quarry in Bedfordshire, UK. Journal of Insect Conservation, 13, 475486.
  • Van Swaay, C.A.M. , Cuttelod, A. , Collins, S. , Maes, D. , Lopez Munguira, M. , Sasic, M. , Settele, J. , Verovnik, R. , Verstrael, T. , Warren, M.S. , Wiemers, M. & Wynhof, I. (2010) European Red List of Butterflies. IUCN and butterfly conservation Europe, Publications Office of the European Union, Luxembourg, UK.
  • Van Swaay, C.A.M. & Warren, M.S. (1999) Red Data book of European butterflies (Rhopalocera). Nature and Environment, No. 99, Council of Europe Publishing, Strasbourg, France.
  • Van Swaay, C.A.M. , Warren, M.S. & Lois, G. (2006) Biotope use and trends of European butterflies. Journal of Insect Conservation, 10, 189209.
  • Vanreusel, W. & Van Dyck, H. (2007) When functional habitat does not match vegetation types: a resource-based approach to map butterfly habitat. Biological Conservation, 135, 202211.
  • Warren, M.S. & Bourn, N.A.D. (2011) Ten challenges for 2010 and beyond to conserve Lepidoptera in Europe. Journal of Insect Conservation, 15, 321326.
  • Warren, M.S. & Key, R.S. (1991) Woodlands: past, present and potential for insects. The Conservation of Insects and their Habitats (ed. by N.M. Collins and J.A. Thomas ), pp. 155211. Academic Press, London, UK.
  • Webb, J.R. & Lott, D.A. (2006) The development of ISIS: a habitat-based invertebrate assemblage classification system for assessing conservation interest in England. Journal of Insect Conservation, 10, 179188.
  • White, G.C. & Burnham, K.P. (1999) Program MARK: survival estimation from populations of marked animals. Bird Study, 46(Suppl. 001), S120S139.
  • Zimmermann, K. , Blazkova, P. , Cizek, O. , Fric, Z. , Hula, V. , Kepka, P. , Novotny, D. , Slamova, I. & Konvicka, M. (2011) Demography of adults of the Marsh fritillary butterfly, Euphydras aurinia (Lepidoptera: Nymphalidae) in the Czech Republic: Patterns across sites and seasons. European Journal of Entomology, 108, 243254.
  • Zimmermann, K. , Fric, Z. , Filipova, L. & Konvicka, M. (2005) Adult demography, dispersal and behaviour of Brenthis ino (Lepidoptera: Nymphalidae): how to be a successful wetland butterfly. European Journal of Entomology, 102, 699706.