ARTICLE IN PRESS
Environmental Hazards 6 (2005) 109–114
www.elsevier.com/locate/hazards
Communities and bushfire hazard in Australia:
More questions than answers
Alison Cottrell
Understanding Communities Project, Bushfire Cooperative Research Centre and Centre for Disaster Studies, James Cook University,
Townsville 4811, Ql., Australia
Abstract
The issue of communities and their exposure to bushfire hazard is highly topical internationally. There is a perceived trend of greater
exposure to bushfire risk which is exacerbated by increased levels of building in fire-prone areas or peri-urban regions. There is a need to
clarify what we understand to be peri-urban regions, and how we conceptualise and describe the communities that reside in them, in
order that efficient and effective services are provided. However, more questions arise for us. For example: Where are these communities
located? What do we know about the people who live there? What are the implications for bushfire mitigation? Despite being
problematic, locality remains important to the understanding of communities, bushfire hazard and delivery of services.
r 2005 Elsevier Ltd. All rights reserved.
Keywords: Understanding communities; Bushfire risk; Wildfire risk
1. Introduction ing the diversity of the communities exposed to bushfire
hazard it is necessary to move on from the definitional issue
The issue of communities and their exposure to bushfire that tends to plague us. Bell and Newby (1971), Nisbet
hazard is highly topical internationally. Although the costs (1966, 1974), Sullivan (2003), and Wellman (1988) all
of bushfires in Australia are relatively lower than those for remind us of the long history of intellectual traditions
other natural hazards, they lead to the highest rates of which inform our debates about what is a community. For
death (BTE, 2001). There is also a trend for increased levels detailed discussion on the variety of what constitutes a
of building in bushfire prone areas (AFAC, 2001). This community see Cahnman and Heberle (1971), Hillery
paper addresses some basic questions about communities (1955), and Bell and Newby (1971). In the context of
that are vulnerable to bushfire hazard in Australia: where natural hazards see Buckle (1999), Cottrell (forthcoming
are these communities located, what do they look like, 2004), Marsh and Buckle (2001), and Sullivan (2003). In
what do we know about the people who live there and what essence, we need to be mindful that just as there are a
are the implications for bushfire mitigation? The conclu- variety of ways that researchers define communities, a
sion is that we do not know enough, and that local service similar variety is held by lay members of the community
providers are integral to expanding our knowledge of these including fire service providers. Understanding these
communities with respect to bushfire hazard. differences is essential to understanding where to start to
bridge gaps between communities and service providers.
2. Location A number of labels have been used for locations at the
urban periphery. In the context of bushfire hazard, we have
Describing and categorising the communities at greatest i-Zone (Queensland Fire and Rescue Service, 2004),
risk to bushfires is somewhat problematic. In demonstrat- wildland–urban interface (Canadian Forest Service, 2004;
Ewert, 1993), Hanson (1962) and forest–residential inter-
Tel.: +07 47 814653; fax: +07 47 814020. mix (Carroll et al., 2005). There is also rural–urban
E-mail address: alison.cottrell@jcu.edu.au. interface (Pahl, 1966), rural–urban fringe (Meligrana,
1464-2867/$ - see front matter r 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.hazards.2005.10.002
, ARTICLE IN PRESS
110 A. Cottrell / Environmental Hazards 6 (2005) 109–114
2003), exurban (Nelson and Dueker, 1990; Nelson 1992; Low-density housing communities or subdivisions lo-
Hugo et al., 2003; Davis, 1990; McKenzie, 1996; Hugo cated next to or in the forest.
et al., 2003) Burnley and Murphy (1995); Bunker Cottages, cabins and recreational and industrial facil-
and Houston (2003); Crump (2003); Exurban Change ities located in the forest.
Program (2004) and suburbanisation (cf. American web- Small-to-medium communities (often aboriginal settle-
sites, Crump, 2003; Edwards, 2004). Hugo (2002) also ments or resource-based communities with populations
identifies the use of ‘pen-metropolitan’ (Burnley and ranging from 500 to 50,000) surrounded by vast
Murphy, 1995 in Hugo, 2002) and ‘technoburbs’ (Fishman, expanses of forest.
1990 in Hugo, 2002). In general, wildland–urban interface
tends to be a term used by Canadians, exurbanisation in The physical layout of peri-urban zones is, therefore, a
the United States of America and in urban planning complex one with potential for high levels of diversity
(Gurran, 2005) and peri-urbanism by researchers of issues between and within locations. The Queensland Fire and
in developing countries (e.g., Browder et al., 1995; Rescue Service (2004) use the term i-Zone and provides a
Monaghan, 2004; Rawlings, 1999; Sadiki and Ramutsin- pictorial schema of two basic types which represent higher
dela, 2002). and lower settlement densities. The first is land parcels of
Although we need to be mindful of the relevance of these less than 0.25 ha and the second is land parcels greater than
terms in a contemporary world (Hugo et al., 2003) for the 0.25 ha but less than 10 ha (see diagram below). The dots
purposes of this discussion it is useful to recognise the represent trees on the landscape indicating suburban
variety of terms that are used. Peri-urban areas can be development adjacent to fire hazards and larger lot
taken to include those areas around urban centres which developments with fire hazards over the allotment.1
can be an extension of the suburbanisation process where
the suburbs encroach on the rural hinterland. It can also
include regional urban centres. In the case of bushfires, the
areas which catch the public eye are those on the periphery
of large cities. In Australia, the fire hazard to peripheral
areas of Sydney, Melbourne, Hobart, Brisbane, Perth,
Adelaide and, more recently, Canberra, are those which
gain the most attention because of the numbers of people
affected, the numbers of deaths, the costs of property
damage, the problems associated with mitigation and
response and the rapid changes in population at these
locations. Queensland Fire and Rescue Service (2004).
These typologies are highly schematic, but do provide a
3. What do these communities look like? bigger picture view of these zones. Current estimates by the
Queensland Rural Fire Service are that there are 1,774,565
The picture we have of the physical layout of these freehold land parcels in Queensland on the urban periphery
locations needs some clarification. Peri-urban zones can be which are vulnerable to bushfires. Chen and McAneney
totally new suburbs with small lot housing, shopping (2004, 2005) estimate that 6% of addresses in and around
centres, service facilities, but essentially a suburb next to major urban centres in Australia are vulnerable to bush-
bushland. Peri-urban zones can also be large lot develop- fires. For bushfire risk in terms of adjacency to national
ments. In many areas, they may be a mix of suburban and parks, Gurran (2005, p. 31) suggests that
industrial zones as housing has spread into peripheral
industrial estates. In other areas they may have the form of In Australia, more than 257,000 people neighbour the
conurbations gradually infilling unsettled areas on the boundaries of our 533 national parksyIn new South
periphery. In yet other areas they may be encroaching on Wales alone, this includes in excess of 8000 rural
rural production. The Canadian Forest Service (2004) landholders and over a hundred suburbs, towns and
defines the situation as: villages adjoining 162 national parks.
The wildland–urban interface is any area where struc- These approaches raise the issue of the utility of
tures (whether residential, industrial, recreational, or geographical information system (GIS) data for informed
agricultural) are located adjacent to or among combustible service delivery. There is no doubt that at the larger scale
wildland fuels. There are many different types of wild- these approaches are already useful, and as technologies
land–urban interface areas in Canada: improve, will become less expensive and more readily
available (Chen et al., 2003; Carniello, 2005; Cova et al.,
Large cities containing forest or grassland areas. 2004; and Goetz et al., 2003) and will remain useful for
Structures (such as homes) situated in lightly populated
agricultural areas which may also contain forests or 1
It should be noted that a number of parameters are used to define risk
rangelands. exposure by QFRS, not only the ones mentioned here.
