SUPPLEMENTARY Material s129

Barton, Ikin, Smith, MacGregor, Lindenmayer. (2014).Vegetation structure moderates the effect of fire on bird assemblages in a heterogeneous landscape. Landscape Ecology.

SUPPLEMENTARY material

Figure S1. Ordination of principal components scores of sites grouped by vegetation type. Principal component 1 (PC 1) explained 44% of variance and had positive loadings for overstorey and leaf litter cover, representing a gradient of increasing canopy density, i.e. increasing vertical structure. Principal component 2 (PC2) explained 29% of variance and had positive loadings for understorey cover, representing a gradient of increasing shrub density, i.e. increasing horizontal structure.

Figure S2. Mean (± SE) PC1 (a, b) and PC2 (c, d) scores of sites in burnt and unburnt vegetation at one and six years after fire.

Table S1. Summary of the three vegetation types considered in this study, including dominant species, area and number of burnt and unburnt sites used in our analyses.

Table S2. Summary of principal components analysis performed on six vegetation structural attributes across three vegetation types and both years of surveys. Component loadings were rotated to assist interpretation. High loadings are shown in bold.

Table S3. Results of the linear mixed model analysis examining the effects of fire, vegetation type, and time on principal component scores representing gradients in vertical (PC1) and horizontal (PC2) vegetation structure.

Table S4. List of bird species recorded from our 2004 and 2009 surveys in forest, woodland and heath vegetation types in Booderee National Park, Jervis Bay Territory, Australia.

Table S5. Summary of the number of observed species and estimated ‘true’ species richness, in each vegetation type from two separate years of sampling.

Table S6. Results from the generalised linear mixed model analysis examining the effects of vegetation type, fire, and year on bird species richness in each site.

2