Vegetation and landscape connectivity control wildfire intervals in unmanaged semi‐arid shrublands and woodlands in Australia

• Published in: Journal of biogeography Vol. 38; no. 1; pp. 112 - 124
• Main Authors: O'Donnell, Alison J, Boer, Matthias M, McCaw, W. Lachlan, Grierson, Pauline F
• Format: Journal Article
• Language: English
• Published: Oxford, UK Blackwell Publishing Ltd 2011; Blackwell Publishing; Blackwell; Wiley
• Subjects: Animal and plant ecology; Animal, plant and microbial ecology; Australian environmental dynamics; Biological and medical sciences; Connectivity; Environmental Sciences; Fire ecology; fire fighting; Fire interval length; Fire regimes; fire spread; Forest and land fires; Forest fires; fuel age; fuel continuity; fuel loading; Fundamental and applied biological sciences. Psychology; General aspects; lakes; landscapes; Musical intervals; Phytopathology. Animal pests. Plant and forest protection; probability distribution; Shrublands; shrubs; south-western Australia; Synecology; topography; Vegetation; Vegetation structure; vegetation type; Weather damages. Fires; Weibull distribution; Wildfires; Woodlands; Australia; Western Australia; Oceania; South Australia; fuel continuity; Landscape; Woodland; Ecological connectivity; Vegetation type; Semi arid zone; Fires; Topography; Fuel; Vegetation; Fire interval length; fuel age; Vegetation fire; Weibull distribution; Age; Shrubbery; Interval; south-western Australia; SUD OUEST AUSTRALIE; AUSTRALIE; WEIBULL MODELE
• ISSN: 0305-0270; 1365-2699
• DOI: 10.1111/j.1365-2699.2010.02381.x

Aim The aim of this study was to determine how spatial variation in vegetation type and landscape connectivity influence fire intervals in a semi‐arid landscape with low relief and complex mosaics of woodland and shrubland vegetation. Location Our study focused on a 15,500‐km² area of relatively undisturbed and unmanaged land in south‐western Australia, referred to as the Lake Johnston region. Methods We modelled fire‐interval data from a 67‐year (1940-2006) digital fire history database using a two‐parameter Weibull function, and tested for the effects of vegetation type and landscape connectivity on estimates of the length of fire intervals (Weibull parameter b) and the dependence of fire intervals on fuel age (Weibull parameter c). Results Vegetation type and landscape connectivity significantly influenced fire interval probability distributions. Fire intervals in shrublands (dense low shrub assemblage) were typically shorter (b = 46 years) and more dependent on fuel age (c = 2.33) than most other vegetation types, while fire intervals in open eucalypt woodlands were much longer (b = 405 years) and were less dependent on fuel age (c = 1.36) than in shrub‐dominated vegetation types. Areas adjacent to or surrounded by salt lakes burnt less frequently (b = 319 years) and fire intervals were less dependent on fuel age (c = 1.48) compared with more exposed areas (e.g. b < 101 years, c > 1.68). Fire intervals in thickets (dense tall shrub assemblage) were longer (b = 101 years) than would be expected from fuel loads, most likely because they were protected from fire by surrounding fuel‐limited woodlands. Main conclusions Fire intervals in south‐western Australia are strongly influenced by spatial variation in vegetation (fuel structure) and landscape connectivity. The importance of fuel structure as a control of fire intervals in south‐western Australia contrasts with other landscapes, where topographical gradients or climatic influences may override the effects of underlying vegetation. We found that, regardless of low relief, topographical features such as large salt lake systems limited the connectivity and spread of fire among landscape units in an analogous manner to lakes or mountainous features elsewhere.