southwest fire, like its landscape, is syncretic, muted, balanced. No single fuel drives the regime; no single ignition source, no single wind, no single topographic feature, no singular climatic phenomenon imposes a domineering pattern onto the southwest ensemble. Instead a suite of biotas balances a suite of stresses, such that fire is one force among several, so routine as to be unexceptional. Fire is endemic, not demonic. It shapes and stirs the biotas rather than overturns them. If it is inexpungable, it is also less inclined to be catastrophic. It subjects the southwest to a low-intensity simmering, not a violent boiling; while there are seasons for fire, they are broad and accommodating. While there are few sites notorious for their fires, neither are there many refugia spared fire.7
The major terrain feature is a corrugated plateau, shallow yet sufficiently elevated to capture moisture enough to support scleroforest. The dry scleroforest is renowned for jarrah (E. marginata), a tree of extraordinary tenacity, almost impervious to fire. Its counterpart in the wet scleroforest is karri (E. diversicolor), a towering gum that resists fire and regenerates vigorously after even intense burns. Existing species tend to be self-perpetuating, immediately reclaiming their sites rather than emerging into dominance after a period of decades. The scleroforest suites appear secure, a stable compound of rain and fire; there is enough fire to ward off any resurgent rainforest, yet not so much (or so vicious) fire as to degrade scleroforest into grass or scrub. The regimes flourish amid abundant fire, yet can survive, within limits, in its absence.
Overall, the southwest represents an enclave of fire, a province small and a fire regime mild by Australian standards. Big fires do occur, of course; normally they accompany frontal passages as winds accelerate and shift, and as atmospheric instability encourages strong convection above burns. But storms approach from the Indian Ocean, with winds moderated by seas, not desiccated by deserts. Inland, on the lee side of the Darling Range, precipitation declines and desert winds stir. Here there is wind to drive fire but too little fuel to flame. Conversely, where ample fuel exists, winds are much less violent; only 5 percent of fires in the southwest, for example, have recorded winds that met or exceeded a strength of six on the Beaufort scale. (In Victoria and Tasmania, the percentage is 25; in South Australia, 30; in New South Wales, 35.) Moreover, during the summer fire season, high-pressure cells tend to be less intense. In the southwest enclave bushfires burn under more diverse conditions, few of which favor conflagrations.
There are two exceptions. One scenario requires that a stationary High draft desert air over the region, a kind of muted chinook splashing dry winds across the Darlings. The other involves infrequent hurricanes that churn out of the Indian Ocean. Monster storms not only hurl trees to the earth, piling abnormal fuels; they can, in the right circumstances, suck huge desert air masses into their vortex, like a black hole capturing streamers of stars from a passing galaxy. The fuel loads and fire winds of Cyclone Alby (1978) testified vividly to what, in Old Australia, must have been a rare but inevitable event.8
If so it was one the southwest enclave accepted and contained, even as it absorbed the cavalcade of Aboriginal firesticks that imprinted routine fire ever more firmly on the landscape. It is likely that the latter held the former in check. In recent years, lightning has accounted for approximately 12 percent of all fires in Western Australia, roughly equivalent to the frequency of lightning fire in the United States. In Old Australia the mountains would have experienced more starts; the open grasslands, the larger fires. All this only readied the southwest, however, for the profound reformation kindled by Aboriginal firesticks. Not every locale experienced those new fires with equal intensity, but where they burned more fires meant reduced fuels, which meant a smaller domain for natural burning. Anthropogenic fire branded the landscape with a seasonal regularity no less pronounced, and far more pervasive, than the meteorological minuet of lightning and cyclone. The firestick initiated the fissioning of Old Australia into New.9
FIRE FLUME: THE SOUTHEAST SUITE
Three points—Eyre Peninsula, Botany Bay, Port Phillip Bay—inscribe the great fire triangle of Australia. Here the centrifugal pump discharges all it has gathered up, as though every attribute that propels fire throughout the continent were hurled across this region, a running stream of fire. Everything within this colossal fire flume must accommodate big fires—not merely large in area or frequent in recurrence, but high-intensity holocaust, a fire possessed, vicious, unquenchable, final.10
Its geography prescribes an ideal formula for conflagrations. A Mediterranean climate reigns over a vast area, one of varied and elevated terrain that enriches the seasonal rains. Soils are newer overall, precipitation fuller, fuel loads abundant, ignition almost constant. Migrating storm tracks draft volumes of desiccating desert winds from the north, the desert interior, then shift abruptly to cold blasts from the south that drive fire flanks before them like a flaming blizzard. Those winds rush over the most rugged terrain in Australia, mountains that act less as baffles than as wings to accelerate and channel the most violent air on the continent. Over the course of the twentieth century, climatologists have noted that “potentially bad fire seasons” tend to occur in the southeast about once every three years, “bad fires” every six or seven years, and “very bad fires” every thirteen years. While these figures reflect ignition patterns (a largely human product), they do outline a rough proportion of fire severity relative to fire frequency.11
That severity is something with which every component of the southeast suite must live. Its fuels are varied—interior savannas, scleroforests, heaths, mallee, buttongrass moor, and rainforest. Each biota modifies the coarse geography set by mountains, climate, and winds into distinctive regimes with their own frequency and typology of fires. There are fires of all sorts, a kind of background count of chronic combustion, but each landscape must in some way also survive the holocaust that rushes down the fire flume, for that is the trying, the distinctively Australian fire.12
Fortunately, the Australian sirocco must interact with fuels readied by drought and kindled by a well-timed spark. Considered in this way, those varied regimes are out of sync. The ideal formula for grass fires calls for an exceptionally wet winter followed by an exceptionally dry summer—the one builds up fuel loads beyond what fauna can consume, the other cures and desiccates it. For scleroforest, however, a dry summer may not be sufficient. The larger fuels and deeper taproots drain only after true drought, or following several seasons of cumulative water deficits. Thus, the circumstances that favor grass-driven fires do not promote equally vigorous fires in scleroforest fuels, while, conversely, the prolonged drought necessary to prepare scleroforest strangles the heavy growth necessary to propel grassy fires. In any year large fires affect either the lowland savannas or the mountain forests but not both simultaneously. The intermediate scleroforest varies between those extremes in timing, fire frequency, and fire intensity, its fuels meshing with its biology in rough, sometimes uncanny synchronization.
In dry scleroforest eucalypt litter sets the fuel loads (65–70 percent of the total), with minor contributions from herbs and grasses. Fires are frequent, light, syncopated with fuel buildup. Surface fuels can reach levels of 15 tons/hectare within ten years, 22 tons/hectare after twenty years, and 27 tons/hectare after thirty. But fires typically burn between three and twelve years, between the time at which fuels are adequate to support sustained combustion and the time beyond which it is unlikely that a site can escape ignition. For several years after a burn, a dry scleroforest lacks the fuel to reburn, and this pause allows the woody scleromorphs to maintain their position. More rapid burning would degrade the biota to pyrophytic grasses, if any are present, or choke the forest with coppicing eucalypts and scleromorphic scrub.13
By contrast, wet scleroforest cultivates an understory rich in shrubs, a higher overall productivity, and a greater proportion of available fuels. Where dry scleroforest encourages easy ignition and spread, however, wet scleroforest discourages it with a microclimate that inhibits drying and a dense canopy that shelters the surface from wind. It claims niches protected from desert winds, sites often swathed in maritime moisture. Within their understory are often found the flora
