Fire usually sounds like a force that breaks relationships in nature. But in this Australian savanna, frequent burning helped plants and ants connect more often—and brought more of the most protective ants into the picture. Fire is a major disturbance affecting ecosystems globally, but its impact on mutualisms has received minimal attention.

Here, a long-term field experiment examines how different fire patterns affect ant partnerships with honeydew and plant nectar in an Australian tropical savanna. These interactions provide ants with a key energy source, while their plant and insect hosts receive protection services.

It is a little like a neighborhood exchange: one side supplies food, and the other stands guard. Australian savannas are among the world’s most fire-prone biomes. Frequent fire promotes ant abundance and diversity because it produces an open habitat favored by mostly dry-country animals.

But the fire does not affect every dominant ant in the same way. One group nests in the ground and prefers open ground, while another nests in foliage and favors shady habitats. The study used a long-term burning experiment with plots burned at different times and intervals, alongside plots left unburnt for more than twenty-five years.

Woody cover decreased as fire became more frequent, and late fires were more intense than early fires. This gave the comparison a real-world setting: the same savanna could be viewed under repeated burning or long fire exclusion, while researchers examined ant interactions on plants.

Honeydew interactions occurred on thirty-six point three percent of eucalyptus plants and forty-four point six percent of acacia plants, while plant-nectar interactions occurred on only eighteen point nine percent of acacia plants. So the two food sources did not play equal roles.

Ants were found with honeydew more often than with the nectar offered directly by acacia plants. Interactions involving both honeydew and plant nectar were always lowest in unburnt plots. Across the burned plots, interactions with both food sources were more common than in unburnt plots.

The clearest difference involved plant nectar: it appeared on only two point eight percent of acacia plants in unburnt plots, compared with twenty-three point seven percent in burned plots. Fire markedly increased honeydew interactions on eucalyptus and nectar interactions on acacia.

The effect was especially clear for nectar, while fire did not affect acacia interactions involving honeydew. Several explanations are possible, including more ants, changes in how attractive nectar is after fire, or changes in the amount or quality of nectar.

The strongest effect suggests that burned plots may have offered more, or better, nectar. The expected change in ant partners did not occur. Interactions with the ground-nesting group were lower in unburnt plots, but the foliage-nesting group did not increase on individual plants.

The result suggests that excluding fire may favor the foliage-nesting ants by increasing the density of woody plants, and therefore the size or number of colonies, rather than increasing their presence on each plant. The study measured how often partnerships occurred and which partners were present, rather than directly measuring the benefits received by each partner.

Even so, more interactions and more highly protective ants suggest stronger protection for plants and insects in this savanna. That conclusion belongs to a landscape adapted to frequent fire. It cannot be readily generalized to places where fire is less common, so changing fire patterns elsewhere will require separate study.

In this fire-prone savanna, frequent fire increased ant partnerships, especially around plant nectar, and favored highly protective ants. That may help plants and insects, but the result may not apply where fire is rare.