Shorter fire-return intervals rearrange fuel arrays in dry forests of southeastern Australia

Fire is a fundamental force in the Australian landscape, shaping the forests just as the forests shape the fire. In many ecosystems, plants have evolved specific traits to survive and even depend on regular burns, such as seeds that only germinate after exposure to heat or trees that can regrow from their trunks after a blaze. This relationship usually keeps the forest in a stable condition, returning to a familiar state after a fire. However, when fires happen too frequently, they can push these resilient systems past their breaking point. Instead of recovering as a tall forest, the land can shift into a completely different, persistent state dominated by low, dense shrubs and small trees. This phenomenon, known as an alternative ecosystem state, is becoming more common as fire seasons grow longer and more intense. The critical question for land managers and scientists is whether these new, shrub-filled landscapes behave differently when they burn. Do they ignite more easily? Do they spread faster? Or do they burn with less intensity? Understanding these differences is vital for predicting how future fires will move across the countryside and for managing the safety of communities living near these forests.

Researchers set out to answer these questions by studying a specific area in southeastern Australia, within Wilsons Promontory National Park. Here, a series of intense fires in the mid-twentieth century destroyed the tall eucalyptus trees in certain patches, preventing them from growing back. Over the decades, these areas transformed into dense thickets of smaller shrubs and trees, while nearby forests with similar soil and climate retained their tall eucalyptus canopy. The team compared these two distinct environments: the "reference" forests that still held their tall trees, and the "alternative" states where the tall trees were gone. They did not just guess at how these forests would burn; they measured the physical fuel on the ground, tested how easily that fuel caught fire in a laboratory, and used computer models to simulate how a fire would behave across the landscape under various weather conditions, from mild prescribed burns to extreme wildfire days.

The physical structure of the two forest types turned out to be quite different. In the alternative states, the ground was covered with a thick layer of fine leaves and twigs, similar to the reference forests, but the vegetation above the ground was a dense, tall wall of shrubs reaching up to nearly five meters. In contrast, the reference forests had a more open understory with a tall canopy of eucalyptus trees overhead. Crucially, the reference forests were covered in loose, fibrous bark from the eucalyptus trees, which can peel off and fly through the air as embers, while the alternative states had almost no bark at all. When the researchers measured the danger posed by these fuels using standard field protocols, they found a complex picture. The alternative states scored higher for the height and density of the shrubs, but the reference forests scored higher for the amount of loose bark and the density of the low-lying grasses.

To see how these differences played out in a real fire, the team took samples of the dry leaves and twigs from the ground in both types of forests and tested them in a controlled laboratory setting. They found that the fuel from the alternative states was surprisingly harder to ignite. When a small, burning piece of cotton was dropped onto the samples to mimic a wind-blown ember, the fuel from the alternative forests was less likely to catch fire, burned more slowly, and consumed less of its own mass than the fuel from the reference forests. This suggests that the dense, tightly packed leaves of the shrubs in the alternative states do not allow enough air to flow through them to support a quick, easy ignition. However, once a fire does manage to start in these dense shrubs, the story changes. Because the vegetation is tall and packed so tightly, the flames can grow much taller. In the computer simulations, fires in the alternative states produced significantly taller flames across all weather conditions, reaching heights that could easily reach the tops of nearby tall trees.

The simulations also revealed a stark difference in how far a fire could jump. In the reference forests, the loose bark from the eucalyptus trees acted as a source of flying embers, allowing the fire to spot ahead and start new blazes hundreds of meters away. In the alternative states, the lack of bark meant that these flying embers were virtually non-existent, keeping the fire contained to the ground it was burning on. The speed at which the fire moved across the ground was similar in both forests under normal conditions, but under the most extreme, catastrophic weather scenarios, the fire in the alternative states spread faster. This suggests that while the alternative forests are harder to start, once they are burning, they can become more intense and spread more rapidly in severe weather, largely because the dense, tall fuel allows the fire to climb and move with great energy.

These findings paint a nuanced picture of fire risk in these changing landscapes. The alternative states are not necessarily more dangerous in every way; they are actually less likely to be ignited by flying embers and their ground fuel is less flammable. However, if a fire does get started, the unique structure of these forests allows it to burn with greater intensity and taller flames, which could then ignite the tall trees in the surrounding forests. This creates a feedback loop where the altered vegetation makes it harder for the forest to return to its original state, as the intense fires it produces prevent the tall trees from recovering. The research highlights that simply looking at how much fuel is on the ground is not enough to understand fire danger. The shape and type of that fuel matter just as much, creating a distinct fire signature for these alternative landscapes that requires a different approach to management and safety.