SEARCH

SEARCH BY CITATION

Fig. S1. Map of the study area; studied landscapes (4x4 Km squares) were distributed throughout the Lazio region and the Province of Siena.

Fig. S2. Values of percent forest cover (within the 4x4km squares) and PCA components interpreted as gradients in habitat amount (HA), habitat fragmentation per se (HF), and structural connectivity (C) for each of the 41 studied landscape in central Italy.

Fig. S3. Relative importance of variables in models for the dormouse (stippled) and red squirrel (clear), obtained by summing the Akaike weights for all models in the set including the target variable. HA = principal component interpreted as a gradient of habitat loss; HF = principal component interpreted as a gradient of habitat fragmentation per se; C = principal component interpreted as a gradient of structural connectivity (in the form of hedgerows); Luse 1 and Luse 2= principal components interpreted as north-south-east-west gradients in land use patterns. See materials and methods for a detailed description of the components.

Fig. S4. a) Results of the second ranked model for the red squirrel (Sciurus vulgaris): red squirrel probability of presence is expressed as a function of habitat fragmentation per se (principal component HF) controlling for habitat amount (principal component HA). Each line represents a fixed value of the HA component: e.g. a value of HA= -2.8 roughly corresponds to landscapes with less than 5% residual forest cover. An increase in habitat subdivision can increase the probability of presence only in landscape with relatively high amounts. b) Results of the third ranked model: red squirrel probability of presence is expressed as a function of structural connectivity (principal component C) controlling for habitat amount (principal component HA). Each line represents a fixed value of the HA component. An increase in structural connectivity can increase probability of presence in landscapes with moderate levels of habitat, but below a certain threshold of habitat even high values of structural connectivity will not increase probability of presence to relatively high values.

Table S1. Study design showing the combination of landscapes variables for each category; numbers correspond to landscapes in Fig. S1. See Fig. S2 for a graphical representation of the PCA axis values.

Table S2. Descriptive statistics of the landscape characteristics measured in 41 4x4Km landscapes in central Italy.

Table S3. Values of the principal components and presence (=1)/absence (=0) of the target species (Sciurus vulgaris and Muscardinus avellanarius) in each of the 41 sampled landscapes in central Italy.

As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials may be re-organized for online delivery, but are not copy-edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors.

FilenameFormatSizeDescription
JPE_1918_sm_figS1-S4b-tabS1-S3.doc647KSupporting info item

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.