Bushfire opinions in Madagascar: actions speak louder than words?

August 6, 2026

A team of outsiders comes to grassland villages in rural highland Madagascar and asks about bushfire (doro tanety). Fire is bad, respond the villagers. It has no advantages and degrades the environment. Asked the same questions, city dwellers in Antananarivo are more equivocal: it’s bad for the environment, but might be a necessary evil for cattle raising and farming. Meanwhile, fires continue to be lit prolifically in this grassy landscape with a long dry season, a little less than in the past, but still omnipresent in the landscape with smoke plumes, black scars, and night-time flickering lights.

What’s going on here? To understand these contradictions, one needs to understand fire ecology, agrarian livelihoods, the power of discourses, as well as the psychology and ethics of research methods. Fire ecology, because fire is inevitable in wet-dry tropical grassy landscapes. Agrarian livelihoods, because fire can be a simple and effective tool for pasture renewal and field preparation (at the same time as a threat to thatched roofs and tree plantations). The power of discourses, because a century of anti-fire propaganda by governments, foresters, and conservationists – serving their interests, reflecting older temperate zone biases – shapes the frame in which people talk about fire. And research methods, because how do you ask people their honest opinions about a criminal activity when you arrive in their village in a sparkling four-wheel drive vehicle? These questions are at the center of a paper “Anti-fire statements coexist with millions of anthropogenic fires in Madagascar” that we recently published, under the leadership of Víctor Fernández García and funded by a SNIS research grant.

Youths clowning in front of small grass fire at night in middle-west of Madagascar
Fire play in Ambarinanahary, 1994. The foreign researcher (me, a masters student at the time) arrives in a village, not yet conversant in the local language. My guide, interpreter, and friend, a geography student from Antananarivo, translates my questions about farming, land use, and village history. After sunset, people note my excitement at the bushfires lighting up the horizon in different directions across the grassy plains. An hour later, we’re called outside, there is laughter and urgency in the air, it seems like half the village is there. I take out my camera, struggle with the exposure for night fire photos, and the kids start mugging it up in front of the flames.

Read the rest of this entry »

Madagascar’s fire regimes compared to the rest of the tropics

June 15, 2022

Bushfire is often seen as symbolic of environmental catastrophe on Madagascar. But is it? A global comparison of fire regimes based on satellite image data suggests care in jumping to such conclusions. A recent article in Global Change Biology, led by Leanne Phelps and to which I contributed, finds that the island’s fire regimes have analogues to 88% of fire regimes in the global tropics with similar climate and vegetation. Madagascar’s fires, while exceptionally vilified, are not exceptional. It also demonstrates that the large, landscape-scale grassland fires common across highland and western Madagascar have no relationship to forest loss; indeed forest loss occurs in places without large-scale fires.

Figure 3 from Phelps et al. (2022). Colours represent areas with similar fire regimes (clustered based on burnt area, fire size, seasonality, and numbers of fires). Black pixels represent fire regimes not found on Madagascar. Gray pixels are places without landscape-scale fire regimes. Photos: (a) tapia branch and chameleon at Ibity, (b) uncontrolled, peri-urban landscape fire in Ambositra [photo C. Kull, 2019], (c) a forest-savanna boundary in Ambohitantely, (d) ancient biodiverse grasslands on Ibity, (e) landscape fire in an agricultural region near Ambositra, likely for grassland renewal [photo by C. Kull, 1998/9], (f) tree cover on a forest-savanna boundary in Ambohitantely, (g) smallholder land use on Ibity.
Read the rest of this entry »