ONLINE RESOURCE European Journal of Wildlife Research

ONLINE RESOURCE – European Journal of Wildlife Research

Surviving on the edge: a conservation-oriented habitat analysis and forest-edge manipulation for the hazel dormouse in the Netherlands

Contents:

-Detailed description of the methods

-Table S1: recorded plant species in the 2011 habitat study

-Table S2: results (AIC model ranking) of the 2011 habitat study

-Figure S1: map of the study area

-References

Jip J. C. Ramakers1,2*, Martijn Dorenbosch2 and Ruud P. B. Foppen3

1Department of International Wildlife Management, Van Hall Larenstein University of Applied Sciences (part of Wageningen University and Research Centre), Agora 1, 8934 CJ Leeuwarden, The Netherlands

2Bureau Natuurbalans – Limes Divergens BV, Radboud University (Mercator III), Toernooiveld 1, 6525 ED Nijmegen, The Netherlands

3Dutch Mammal Society, Radboud University, Natuurplaza (Mercator III), Toernooiveld 1, 6525 ED Nijmegen, The Netherlands

4Department of Animal Ecology and Ecophysiology, Radboud University, Huygens building, Heyendaalseweg 135, 6525 AJ Nijmegen, The Netherlands

5Bureau Waardenburg, P.O. Box 365, 4100 AJ Culemborg, The Netherlands

*Correspondence author:

E-mail:

Phone: +31(0) 6 33 10 20 52

Address: Hagedis 20, 1275 BR, Huizen, The Netherlands

Methods

Study area

The study area is located in the most south-eastern part of the Netherlands (Fig.S1). The area typically encompasses hills and valleys of the South-Limburgian Loess Area, in which the typical forests are of the Luzulo-Fagetum, Fago-Quercetum and Carpinion betuli type, as well as Rhamno-Prunetea and Lonicero-Rubetea thorny shrubs (Schaminée and Janssen 1998). Mean temperatures in this area range from 15.5-16 °C in June and 3-4 °C in December, whilst mean precipitation can be up to 80 mm in June (measured from 1971-2000; Berendsen 2008). Within the study area, dormouse populations are found in three forest types: nutrient-poor beech-oak Fago-Quercetum forests, nutrient-poor wood-rush-beech Luzulo-Fagetum forests, and nutrient-rich hornbeam-oak Querco-Carpinetum(Foppen et al. 1999).

The present study was conducted at four forest locations that encompass a large part of the distribution area of the dormouse in the Netherlands: (1) Groote Bosch (further referred to as GB), (2) Schweiberger & Dunnenbosch (SDB), (3) Onderste & Bovenste Bosch (OBB), and (4) Vijlenerbossen (VB). GB, SDB and OBB are separated from VB by open agricultural fields, but are located within 7 km from one another. GD, SDN and OBB are connected to forest complexes stretching into Belgium, whereas VB is connected to forest complexes in Germany. In each forest, one study site was selected. Since VB forest covers an extensive area, this forest was subdivided into three sites (south, west, and north). Consequently, six study sites were distinguished in total (Fig. S1).

Annual dormouse census

As part of an annual dormouse monitoring programme (Verheggen and Boonman 2006), two survey rounds take place each year between the second half of September and the first half of November. Because the main reproductive period in the Netherlands is from July to September this reflects the time of the year when the population is at its peak (Foppen et al. 2002). During a survey, transects in forest edges or hedgerows are thoroughly searched for dormousenests, following Foppen et al. (2002). The number of observed nests in this period is expected to provide a relative estimate of dormouse densities. The width of surveyed forest transects depends on the edge vegetation structure, usually dense bramble shrubs, bracken, occasionally young trees or holly (Foppen et al. 2002). In general, the surveyed forest transects have a width of only a few meters;this could be more (up to 50 m) in cases with absent or sparsely scattered mature trees. Hedgerows are sampled across their whole width. The exact locations of dormouse nests are recorded in the field with a hand-held GPS and saved on a mobile GIS device. For each nest a descriptive protocol is followed, filling out a standard form with e.g., the plant species that supports the nest and the height of the nest from the forest floor.

Forest-edge manipulation

In the areas VB-W and VB-S, 75-100% of the mature trees were cut in the winter of 2009-2010in parts of the forest edges. These forest-edge measures were directed at stimulating the development of a well-developed forest edge, particularly bramble (Rubus fruticosus agg.), with the expectance to boost the dormouse population. In total, ten forest-edge segments were cleared in the VB-W and VB-S forests (mean ± SD length of each segment: 91.8 ± 43.8 m; width: 17.1 ± 5.4 m). Dormouse nest counts were performedfrom 2009 through to 2013 in these managed segments, as well as in ten unmanaged segments (length: 180.5 ± 21.8 m; width: 8.9 ± 3.0 m) within the same study sites. The length and width of both managed and unmanaged segments were a posteriori determined in an independent vegetation survey conducted in 2011 (see next section).

Forest-edge segments incorporated in the clearing management were arbitrarily chosen based on their vegetation succession; segments used as ‘unmanaged controls’ were a random subset from the remaining segments that were located next to managed segments. Hence, strictly speaking, the study is a quasi-experiment (i.e. not fully random). To avoid confusion, therefore, we shall refer to it as a forest-edge manipulation, and analysis and subsequent interpretation of the results will take this limitation into account.

Habitat description study

In 2011 a study was conducted on habitat selection by the dormouse in forest edges and hedgerows in all six study sites (totalling 19.7 km), using the census methods described above. One census round was conducted from mid-September to mid-October, and a second round from mid-October to mid-November. Independently of the dormouse census, a vegetation survey of forest-edge and hedgerow transects was conducted from late September to mid-October, by a qualified phytosociologist not a priori biased by knowledge of dormouse habitat preferences. Transects were divided into spatial segments based on homogeneity of both tree and shrub species (mean length ± SD: 118.9 ± 68.1; n = 166). The relative abundance of each tree and shrub species (Table S1) in these segments was determined using the Braun-Blanquet scale. The shape of each segment was recorded in the field with a GPS and saved on a mobile GIS device. Plant species abundance in each segment was automatically assigned a value of percentage cover using the programme TURBOVEG (Hennekens and Schaminée 2001). Besides species abundances, several measures relating to structural complexity were noted (Table 1 in main text).

Data analysis

For each individual segment as obtained from the procedure described in the previous section, the pooled number of dormouse nests from both survey rounds was determined by projecting nest data onto the shape data containing all segments using MapInfo Professional 9.0.2 GIS software (PB Mapinfo Corporation, Troy, New York).

We determined the effect of the 2009 clearing management on dormouse nest counts by means of a Generalised Linear Mixed Model (GLMM) with a Poisson error structure, using the lme4 package (Bates et al. 2014)in R 3.0.0 (R Development Core Team; Nest count per segment per year was the response variable, offset by the log of segment length. ‘Treatment’ (managed vs. unmanaged), ‘year’ (2009–2013) and their interaction were included as fixed factors; the approximate width of each segment was added as a covariate. To alleviate the potential role of spillover effects of clearing management in unmanaged segments as a result of the lack of a fully randomised design, an autocovariate (Dormann et al. 2007) was added to the model. This autocovariate was constructed within the spdep package (Bivand et al. 2014)using inverse-distance weighting of each segment’s nest count; the neighbour radius to incorporate segments in the weighting was set to 500 m, which should more or less cover dormouse dispersion distances (i.e. 250–500 m; Büchner 2008). ‘Segment ID’ was specified as random term, estimating an interceptfor each year to account for multi-year sampling. As direct comparison between managed and unmanaged segments was not meaningful (because of the lack of complete randomisation), we instead tested within-treatment changes in nest numbers, relative to their starting point. Significance levels resulting from five pairwise comparisons (i.e. four comparisons between successive years + the net change between 2009 and 2013) within each treatment were corrected using the Bonferroni method.

To analyse the effect habitat variables recorded in 2011 on dormouse nest counts in that year, we determined for each segment: overall diversity of tree and shrub species (Simpson’s reciprocal diversity index; Magurran 2004); diversity of hard-mast species (i.e. species producing nuts or winged seeds; Table S1: species 17-21)and soft-mast species (i.e. species producing fleshy fruits; Table S1: species 1-16); and the cumulative abundance of soft-mast species including hazel (Corylus avellana)(= food-species abundance) (Table 1 in main text). Hazel was included in the latter variable as this species is an important food source for dormice (e.g., Bright and Morris 1993; Richards et al. 1984). We tested the effect of the recorded soft- and hard-mast species only when they occurred in at least 40 segments (a natural boundary in the data), as well as Bracken (Pteridium aquilinum)and Old man’s beard (Climatis vitalba) (given their suitable nesting conditions). Honeysuckle (Lonicera spp.), a potential food source for dormice (e.g., Bright and Morris 1993; Juškaitis and Šiožinytė 2008; Richards et al. 1984), was not included as it was considered a rare species throughout the study sites. To reduce the number of variables to be tested, each variable was first tested in separate GLMMs; variables with an unadjusted P-value of ≥ 0.5 were disregarded in further analyses (see below).

In each model (GLMM with Poisson errors) nest count was the dependent variable, offset by the log of segment length. In addition to the independent variables listed above, an autocovariate was added to account for potential spillover effects (inverse-distance weighting with a 500-m neighbour radius; same procedures as outlined above). Each model contained ‘site’ and an observation-level factor (‘segment ID’) as random terms, to account for spatial clustering and overdispersion, respectively. We adopted an information-theoretic approach (Akaike’s Information Criterion or AICc; Burnham and Anderson 2002) to compare all possible subsets with the restriction of using a minimum of four and a maximum of eight independent variables in each model (based on prior, single-predictor analysis and to avoid overfitting) and excluding pairs of collinear (rSpearman > 0.5) variables.Model ranking based on AICc is based on the “smaller is better” principle; final (standardised) variable coefficients were generated using model averaging with the MuMIn package (Barton 2014) over all models that were within six AICc units from the top-ranked one, ensuring that the best model was chosen with high (> 95%) probability (Richards 2005).

Model fit in both sets of analyses was validated using residual plots, as well as observed-vs.-fitted plots to assess the predictive accuracy of the final models (r ≥ 0.97; in the habitat description study this was the model containing only significant model-averaged coefficients). GLMM procedures closely followed Bolker et al. (2008).

Table S1. Tree and shrub species identified in the field and used in analyses. Species under ‘Soft mast’ and ‘Hard mast’ include only those used as such in the data analysis; the rest is filed under ‘Other’.

Species / Common name / Species / Common name
Soft mast / Other
1 Cornus sanguinea L. / Dogwood / 22 Acer platanoides L. / Norway maple
2 Crataegus laevigata (Poir.) DC / Woodland hawthorn / 23 Acer pseudoplatanus L. / Sycamore maple
3 Crataegus monogyna Jacq. / Common hawthorn / 24 Alnus glutinosa L. / Black alder
4 Humulus lupulus L. / Hop / 25 Betula pendula Roth / Silver birch
5 Ilex aquifolium L. / Holly / 26 Betula pubescens Ehrh. / Downy birch
6 Mespilus germanica L. / Medlar / 27 Carpinus betulus L. / Hornbeam
7 Prunus serotina Ehrh. / Black cherry / 28 Clematis vitalba L. / Old man's beard
8 Prunus spinosa L. / Blackthorn / 29 Cytisus scoparius (L.) Link / Broom
9 Rhamnus frangula L. / Alder buckthorn / 30 Euonymus europaeus L. / Spindle
10 Rosa canina L. / Dog rose / 40 Fraxinus excelsior / Ash
11 Rubus fruticosus L. agg.* / Common blackberry / 41 Picea abies (L.) H. Karst / Norway spruce
12 Rubus idaeus L. / European raspberry / 42 Populus tremula L. / Aspen
13 Rubus spp.* / Blackberry spp. / 43 Populus x canadensis / Hybrid black poplar
14 Sambucus nigra L. / Elder / 44 Prunus avium L. / Wild cherry
15 Sorbus aucuparia L. / Rowan / 45 Pteridium aquilinum (L.) Kuhn / Bracken
16 Viburnum opulus L. / Guelder rose / 46 Quercus rubra L. / Red oak
Hard mast / 47 Ribes rubrum L. / Red currant
17 Acer campestre L. / Field maple / 48 Ribes uva-crispa L. / Gooseberry
18 Castanea sativa Mill. / Sweet chestnust / 49 Robinia pseudoacacia L. / Black locust
19 Corylus avellana L. / Hazel / 50 Salix alba L. / White willow
20 Fagus sylvatica L. / Beech / 51 Salix caprea L. / Goat willow
21 Quercus robur L. / Pendunculate oak / 52 Taxus baccata L. / European yew
53 Tilia cordata Mill. / Small-leaved lime
54 Ulmus minor Mill. / Field elm

*Grouping of blackberry (or bramble) was done based on morphological features, where Rubus spp. represents low-growing bushes, with very little or no fruit production, often found in the forest interiors, and R. fruticosus agg. represents higher (often > 0.5 m) bushes, generally richer in fruit production.

Table S2. Rank list of all the GLMMs (< 6 ΔAICc) for the 2011 habitat description study (explanations given below the table).

Rank / K / ΔAICc / Model formula / w
1 / 10 / 0.00 / + AUTO + Cave + Cmon + Fsyl + SHHT + Rfru + Rspp + Sauc / 0.04
2 / 10 / 0.12 / + HAB + AUTO + WID + Cmon + SHHT + Rfru + Rspp + Sauc / 0.03
3 / 10 / 0.15 / + AUTO + WID + Cmon + Fsyl + SHHT + Rfru + Rspp + Sauc / 0.03
4 / 9 / 0.53 / + AUTO + Cmon + Fsyl + SHHT + Rfru + Rspp + Sauc / 0.03
5 / 10 / 0.54 / + HAB + AUTO + Cmon + Fsyl + SHHT + Rfru + Rspp + Sauc / 0.03
6 / 9 / 0.64 / + HAB + AUTO + Cmon + SHHT + Rfru + Rspp + Sauc / 0.03
7 / 9 / 0.87 / + AUTO + WID + Cmon + SHHT + Rfru + Rspp + Sauc / 0.02
8 / 10 / 1.05 / + AUTO + WID + Cave + Cmon + SHHT + Rfru + Rspp + Sauc / 0.02
9 / 10 / 1.19 / + AUTO + WID + Cmon + SHHT + Rfru + Rspp + DSOF + Sauc / 0.02
10 / 10 / 1.62 / + HAB + AUTO + Cave + Cmon + SHHT + Rfru + Rspp + Sauc / 0.02
11 / 10 / 1.81 / + AUTO + WID + Cave + Cmon + Fsyl + Rfru + Rspp + Sauc / 0.01
12 / 9 / 1.84 / + AUTO + Fsyl + SHHT + Rfru + Rspp + Sauc + FDSP / 0.01
13 / 10 / 1.88 / + AUTO + Fsyl + SHHT – Pspi + Rfru + Rspp + Sauc + FDSP / 0.01
14 / 8 / 2.01 / + AUTO + WID + SHHT + Rfru + Rspp + Sauc / 0.01
15 / 10 / 2.02 / + AUTO + WID + SHHT – Pspi + Rfru + Rspp + Sauc + FDSP / 0.01
16 / 10 / 2.03 / + HAB + AUTO + Cmon + SHHT – Pspi + Rfru + Rspp + Sauc / 0.01
17 / 9 / 2.07 / + AUTO + WID + SHHT – Pspi + Rfru + Sauc + FDSP / 0.01
18 / 10 / 2.13 / + AUTO + WID + Cmon + SHHT – Pspi + Rfru + Rspp + Sauc / 0.01
19 / 9 / 2.13 / + AUTO + WID + SHHT + Rfru + Rspp + DSOF + Sauc / 0.01
20 / 10 / 2.13 / + AUTO + WID + Fsyl + SHHT + Rfru + Rspp + Sauc + FDSP / 0.01
21 / 10 / 2.16 / + AUTO + Cmon + Fsyl + SHHT – Pspi + Rfru + Rspp + Sauc / 0.01
22 / 10 / 2.32 / + HAB + AUTO + Cmon + SHHT + Rfru + Rspp + DSOF + Sauc / 0.01
23 / 9 / 2.33 / + AUTO + WID + SHHT + Rfru + Rspp + Sauc + FDSP / 0.01
24 / 8 / 2.35 / + AUTO + Fsyl + SHHT + Rfru + Rspp + Sauc / 0.01
25 / 9 / 2.39 / + AUTO + WID + Fsyl + SHHT + Rfru + Rspp + Sauc / 0.01
26 / 10 / 2.42 / + AUTO + WID + Cmon + SHHT + Rfru + Rspp + DTOT + Sauc / 0.01
27 / 10 / 2.43 / + AUTO + Cmon + Fsyl + SHHT + Rfru + Rspp + DTOT + Sauc / 0.01
28 / 10 / 2.43 / + AUTO + Cmon + Fsyl + SHHT + Rfru + Rspp + DSOF + Sauc / 0.01
29 / 9 / 2.57 / + AUTO + Cave + Cmon + Fsyl + Rfru + Rspp + Sauc / 0.01
30 / 10 / 2.74 / + HAB + AUTO + Cmon + SHHT + Rfru + Rspp + DTOT + Sauc / 0.01
31 / 10 / 2.76 / + AUTO + WID + Fsyl + SHHT – Pspi + Rfru + Sauc + FDSP / 0.01
32 / 9 / 2.86 / + HAB + AUTO + WID + SHHT + Rfru + Rspp + Sauc / 0.01
33 / 10 / 2.94 / + AUTO + WID + Cave + Cmon + Rfru + Rspp + DSOF + Sauc / 0.01
34 / 10 / 3.04 / + AUTO + WID + SHHT – Pspi + Rfru + Rspp + DSOF + Sauc / 0.01
35 / 9 / 3.06 / + AUTO + WID + Cmon + Rfru + Rspp + DSOF + Sauc / 0.01
36 / 8 / 3.07 / + HAB + AUTO + SHHT + Rfru + Rspp + Sauc / 0.01
37 / 9 / 3.09 / + AUTO + Cave + Fsyl + SHHT + Rfru + Rspp + Sauc / 0.01
38 / 9 / 3.11 / + AUTO + WID + Cave + SHHT + Rfru + Rspp + Sauc / 0.01
39 / 9 / 3.12 / + AUTO + Fsyl + SHHT – Pspi + Rfru + Sauc + FDSP / 0.01
40 / 9 / 3.31 / + AUTO + WID + SHHT – Pspi + Rfru + Rspp + Sauc / 0.01
41 / 10 / 3.34 / + AUTO + WID + Fsyl + SHHT + Rfru + Rspp + DSOF + Sauc / 0.01
42 / 10 / 3.35 / + AUTO + WID + Cmon + Fsyl + Rfru + Rspp + DSOF + Sauc / 0.01
43 / 10 / 3.38 / + HAB + AUTO + WID + SHHT – Pspi + Rfru + Sauc + FDSP / 0.01
44 / 10 / 3.41 / + AUTO + WID + SHHT + Rfru + Rspp + DSOF + Sauc + FDSP / 0.01
45 / 9 / 3.42 / + AUTO + WID + Cave + Cmon + Rfru + Rspp + Sauc / 0.01
46 / 10 / 3.45 / + AUTO + WID + Cave + Fsyl + SHHT + Rfru + Rspp + Sauc / 0.01
47 / 9 / 3.47 / + AUTO + WID + Cmon + Fsyl + Rfru + Rspp + Sauc / 0.01
48 / 10 / 3.51 / + HAB + AUTO + WID + SHHT + Rfru + Rspp + DSOF + Sauc / 0.01
49 / 10 / 3.53 / + HAB + AUTO + WID + SHHT + Rfru + Rspp + Sauc + FDSP / 0.01
50 / 9 / 3.58 / + HAB + AUTO + Fsyl + SHHT + Rfru + Rspp + Sauc / 0.01
51 / 9 / 3.59 / + AUTO + Cave + Cmon + SHHT + Rfru + Rspp + Sauc / 0.01
52 / 10 / 3.59 / + HAB + AUTO + Fsyl + SHHT + Rfru + Rspp + Sauc + FDSP / 0.01
53 / 10 / 3.62 / + AUTO + WID + Cave + SHHT + Rfru + Rspp + DSOF + Sauc / 0.01
54 / 9 / 3.65 / + AUTO + SHHT – Pspi + Rfru + Rspp + Sauc + FDSP / 0.01
55 / 10 / 3.77 / + HAB + AUTO + SHHT – Pspi + Rfru + Rspp + Sauc + FDSP / 0.01
56 / 8 / 3.83 / + AUTO + Cmon + SHHT + Rfru + Rspp + Sauc / 0.01
57 / 9 / 3.84 / + HAB + AUTO + SHHT + Rfru + Rspp + Sauc + FDSP / 0.01
58 / 8 / 3.85 / + AUTO + WID + SHHT + Rfru + Sauc + FDSP / 0.01
59 / 9 / 3.88 / + AUTO + WID + SHHT + Rfru + Rspp + DTOT + Sauc / 0.01
60 / 10 / 3.93 / + AUTO + WID + SHHT – Pspi + Rfru + DSOF + Sauc + FDSP / 0.01
61 / 9 / 3.95 / + AUTO + Fsyl + SHHT – Pspi + Rfru + Rspp + Sauc / 0
62 / 9 / 3.96 / + AUTO + Fsyl + SHHT + Rfru + Rspp + DSOF + Sauc / 0
63 / 7 / 4.00 / + AUTO + SHHT + Rfru + Rspp + Sauc / 0
64 / 9 / 4.03 / + HAB + AUTO + SHHT – Pspi + Rfru + Sauc + FDSP / 0
65 / 10 / 4.03 / + AUTO + WID + Fsyl + SHHT – Pspi + Rfru + Rspp + Sauc / 0
66 / 10 / 4.04 / + AUTO + Fsyl + SHHT + Rfru + Rspp + DSOF + Sauc + FDSP / 0
67 / 10 / 4.12 / + AUTO + Fsyl + SHHT + Rfru + Rspp + DTOT + Sauc + FDSP / 0
68 / 10 / 4.15 / + HAB + AUTO + WID + Fsyl + SHHT + Rfru + Rspp + Sauc / 0
69 / 9 / 4.18 / + AUTO + WID + Fsyl + SHHT + Rfru + Sauc + FDSP / 0
70 / 10 / 4.18 / + AUTO + WID + Cave + SHHT – Pspi + Rfru + Rspp + Sauc / 0
71 / 10 / 4.19 / + AUTO + Cave + Cmon + SHHT – Pspi + Rfru + Rspp + Sauc / 0
72 / 10 / 4.22 / + AUTO + Cave + Cmon + Fsyl + Rfru + Rspp + DSOF + Sauc / 0
73 / 9 / 4.23 / + HAB + AUTO + WID + Cmon + SHHT + Rfru + Sauc / 0
74 / 10 / 4.29 / + AUTO + WID + SHHT – Pspi + Rfru + DTOT + Sauc + FDSP / 0
75 / 8 / 4.30 / + AUTO + SHHT + Rfru + Rspp + Sauc + FDSP / 0
76 / 10 / 4.32 / + HAB + AUTO + WID + SHHT – Pspi + Rfru + Rspp + Sauc / 0
77 / 10 / 4.33 / + SHCO + AUTO + Cave + Cmon + Fsyl + Rfru + Rspp + Sauc / 0
78 / 9 / 4.39 / + HAB + AUTO + SHHT + Rfru + Rspp + DSOF + Sauc / 0
79 / 8 / 4.42 / + AUTO + SHHT – Pspi + Rfru + Sauc + FDSP / 0
80 / 9 / 4.44 / + AUTO + Fsyl + SHHT + Rfru + Rspp + DTOT + Sauc / 0
81 / 9 / 4.44 / + HAB + AUTO + SHHT – Pspi + Rfru + Rspp + Sauc / 0
82 / 10 / 4.44 / + AUTO + Cave + Cmon + Fsyl + Rfru + Rspp + DTOT + Sauc / 0
83 / 10 / 4.46 / + HAB + AUTO + WID + Cave + SHHT + Rfru + Rspp + Sauc / 0
84 / 10 / 4.47 / + AUTO + WID + Fsyl + SHHT + Rfru + Rspp + DTOT + Sauc / 0
85 / 10 / 4.50 / + AUTO + Cave + Fsyl + SHHT – Pspi + Rfru + Rspp + Sauc / 0
86 / 10 / 4.50 / + AUTO + WID + Cmon + Fsyl + Rfru + Rspp + DTOT + Sauc / 0
87 / 10 / 4.58 / + AUTO + WID + SHHT + Rfru + Rspp + DTOT + Sauc + FDSP / 0
88 / 9 / 4.63 / + HAB + AUTO + Cave + SHHT + Rfru + Rspp + Sauc / 0
89 / 10 / 4.64 / + HAB + AUTO + Fsyl + SHHT – Pspi + Rfru + Sauc + FDSP / 0
90 / 9 / 4.66 / + HAB + AUTO + WID + SHHT + Rfru + Sauc + FDSP / 0
91 / 10 / 4.68 / + AUTO + Cave + Cmon + Fsyl – Pspi + Rfru + Rspp + Sauc / 0
92 / 8 / 4.70 / + AUTO + Cave + SHHT + Rfru + Rspp + Sauc / 0
93 / 8 / 4.74 / + AUTO + WID + Cmon + Rfru + Rspp + Sauc / 0
94 / 9 / 4.74 / + AUTO + WID + Fsyl + Rfru + Rspp + Sauc + FDSP / 0
95 / 10 / 4.82 / + AUTO + WID + Cave + Cmon + Rfru + Rspp + DTOT + Sauc / 0
96 / 10 / 4.84 / + AUTO + WID + Cmon – Pspi + Rfru + Rspp + DSOF + Sauc / 0
97 / 9 / 4.85 / + AUTO + Cmon + SHHT – Pspi + Rfru + Rspp + Sauc / 0
98 / 10 / 4.87 / + HAB + AUTO + Cave + Fsyl + SHHT + Rfru + Rspp + Sauc / 0
99 / 10 / 4.87 / + HAB + AUTO + Cave + Cmon + Fsyl + Rfru + Rspp + Sauc / 0
100 / 8 / 4.87 / + AUTO + Fsyl + SHHT + Rfru + Sauc + FDSP / 0
101 / 8 / 4.88 / + AUTO + SHHT + Rfru + Rspp + DSOF + Sauc / 0
102 / 9 / 4.89 / + AUTO + Cmon + SHHT + Rfru + Rspp + DSOF + Sauc / 0
103 / 10 / 4.93 / + AUTO + Cave + Fsyl + SHHT + Rfru + Rspp + DSOF + Sauc / 0
104 / 8 / 5.00 / + AUTO + Cmon + Fsyl + Rfru + Rspp + Sauc / 0
105 / 10 / 5.01 / + HAB + AUTO + WID + SHHT + Rfru + Rspp + DTOT + Sauc / 0
106 / 10 / 5.02 / + AUTO + WID + SHHT – Pspi + Rfru + Rspp + DTOT + Sauc / 0
107 / 8 / 5.06 / + AUTO + Fsyl + Rfru + Rspp + Sauc + FDSP / 0
108 / 10 / 5.08 / + AUTO + Cave + Cmon + SHHT + Rfru + Rspp + DSOF + Sauc / 0
109 / 8 / 5.09 / + AUTO + SHHT – Pspi + Rfru + Rspp + Sauc / 0
110 / 10 / 5.13 / + HAB + AUTO + WID + Cmon + SHHT – Pspi + Rfru + Sauc / 0
111 / 10 / 5.16 / + HAB + AUTO + WID + Cmon + Rfru + Rspp + DSOF + Sauc / 0
112 / 10 / 5.20 / + SHCO + AUTO + WID + Cmon + Rfru + Rspp + DSOF + Sauc / 0
113 / 10 / 5.21 / + AUTO + SHHT – Pspi + Rfru + Rspp + DSOF + Sauc + FDSP / 0
114 / 10 / 5.23 / + AUTO + WID + Cave + SHHT + Rfru + Rspp + DTOT + Sauc / 0
115 / 9 / 5.25 / + HAB + AUTO + SHHT + Rfru + Rspp + DTOT + Sauc / 0
116 / 10 / 5.25 / + HAB + AUTO + Fsyl + SHHT – Pspi + Rfru + Rspp + Sauc / 0
117 / 10 / 5.27 / + HAB + AUTO + Fsyl + SHHT + Rfru + Rspp + DSOF + Sauc / 0
118 / 10 / 5.28 / + AUTO + WID + Cave + Cmon – Pspi + Rfru + Rspp + Sauc / 0
119 / 9 / 5.29 / + AUTO + WID + Cmon + Rfru + Rspp + DTOT + Sauc / 0
120 / 10 / 5.33 / + AUTO + Fsyl + SHHT – Pspi + Rfru + Rspp + DSOF + Sauc / 0
121 / 10 / 5.35 / + AUTO + Cave + Fsyl + SHHT + Rfru + Rspp + DTOT + Sauc / 0
122 / 8 / 5.37 / + HAB + AUTO + Cmon + SHHT + Rfru + Sauc / 0
123 / 9 / 5.38 / + AUTO + Cmon + SHHT + Rfru + Rspp + DTOT + Sauc / 0
124 / 10 / 5.38 / + AUTO + Fsyl + SHHT – Pspi + Rfru + DSOF + Sauc + FDSP / 0
125 / 10 / 5.41 / + AUTO + Fsyl + SHHT – Pspi + Rfru + DTOT + Sauc + FDSP / 0
126 / 8 / 5.43 / + AUTO + WID + Rfru + Rspp + Sauc + FDSP / 0
127 / 8 / 5.44 / + AUTO + WID – Pspi + Rfru + Sauc + FDSP / 0
128 / 10 / 5.49 / + SHCO + AUTO + WID + Cmon + Fsyl + Rfru + Rspp + Sauc / 0
129 / 10 / 5.49 / + HAB + AUTO + SHHT – Pspi + Rfru + Rspp + DSOF + Sauc / 0
130 / 10 / 5.49 / + SHCO + AUTO + WID + Cave + Cmon + Rfru + Rspp + Sauc / 0
131 / 9 / 5.52 / + AUTO + Cave + SHHT – Pspi + Rfru + Rspp + Sauc / 0
132 / 8 / 5.52 / + HAB + AUTO + SHHT + Rfru + Sauc + FDSP / 0
133 / 10 / 5.55 / + AUTO + Cave + Cmon + SHHT + Rfru + Rspp + DTOT + Sauc / 0
134 / 10 / 5.55 / + AUTO + Cmon + SHHT – Pspi + Rfru + Rspp + DSOF + Sauc / 0
135 / 10 / 5.57 / + HAB + AUTO + WID + Cave + Cmon + Rfru + Rspp + Sauc / 0
136 / 9 / 5.58 / + AUTO + WID + Fsyl – Pspi + Rfru + Sauc + FDSP / 0
137 / 9 / 5.60 / + AUTO + WID + Rfru + Rspp + DSOF + Sauc + FDSP / 0
138 / 9 / 5.60 / + AUTO + SHHT – Pspi + Rfru + Rspp + DSOF + Sauc / 0
139 / 7 / 5.62 / + AUTO + WID + SHHT + Rfru + Sauc / 0
140 / 10 / 5.65 / + AUTO + WID + Fsyl – Pspi + Rfru + Rspp + Sauc + FDSP / 0
141 / 10 / 5.70 / + HAB + AUTO + SHHT + Rfru + Rspp + DSOF + Sauc + FDSP / 0
142 / 8 / 5.71 / + HAB + AUTO + WID + SHHT + Rfru + Sauc / 0
143 / 10 / 5.73 / + HAB + AUTO + WID + Cave + Cmon + SHHT + Rfru + Sauc / 0
144 / 10 / 5.74 / + AUTO + WID + Cmon + Fsyl – Pspi + Rfru + Rspp + Sauc / 0
145 / 10 / 5.76 / + HAB + AUTO + WID + Cmon + Fsyl + Rfru + Rspp + Sauc / 0
146 / 8 / 5.78 / + AUTO + WID + Cmon + SHHT + Rfru + Sauc / 0
147 / 10 / 5.79 / + HAB + AUTO + Cave + SHHT – Pspi + Rfru + Rspp + Sauc / 0
148 / 10 / 5.81 / + HAB + AUTO + Fsyl + SHHT + Rfru + Rspp + DTOT + Sauc / 0
149 / 8 / 5.82 / + AUTO + SHHT + Rfru + Rspp + DTOT + Sauc / 0
150 / 10 / 5.87 / + AUTO + SHHT – Pspi + Rfru + Rspp + DTOT + Sauc + FDSP / 0
151 / 9 / 5.88 / + AUTO + Cmon + Fsyl + Rfru + Rspp + DSOF + Sauc / 0
152 / 9 / 5.88 / + AUTO + WID + SHHT + Rfru + DSOF + Sauc + FDSP / 0
153 / 9 / 5.90 / + AUTO + Cave + SHHT + Rfru + Rspp + DSOF + Sauc / 0
154 / 8 / 5.90 / + AUTO + WID + Fsyl + Rfru + Sauc + FDSP / 0
155 / 9 / 5.91 / + AUTO + Fsyl – Pspi + Rfru + Rspp + Sauc + FDSP / 0
156 / 9 / 5.92 / + AUTO + SHHT + Rfru + Rspp + DSOF + Sauc + FDSP / 0
157 / 10 / 5.92 / + HAB + AUTO + WID + Fsyl + SHHT + Rfru + Sauc + FDSP / 0
158 / 10 / 5.95 / + AUTO + Fsyl + SHHT – Pspi + Rfru + Rspp + DTOT + Sauc / 0
159 / 9 / 5.95 / + HAB + AUTO + Fsyl + SHHT + Rfru + Sauc + FDSP / 0
160 / 9 / 5.97 / + AUTO + WID – Pspi + Rfru + Rspp + Sauc + FDSP / 0

K = number of parameters; ΔAICc= difference in Akaike’s Information Criterion between current and top-ranked model; w = Akaike weight of the model. Model terms: AUTO = autocovariate (weighting variable); Rfru = Rubus frutiscosus agg.; Rspp = Rubus spp.; Sauc = Sorbus aucuparia; Pspi = Prunus spinosa; Fsyl = Fagus sylvatica; Cmon = Crataegus monogyna; Cave = Corylus avellana;HAB = habitat type; WID = edge width; SHHT = shrub height; DSOFT = Simpson’s D soft-mast species; DTOT = Simpson’s D total trees and shrubs; FDSP = food-species abundance.