The flat-headed cat is a biological oddity among felids. Its common name comes from its unusually flattened head, and its toes are semi-webbed while its claws do not fully retract, adaptations that make it an effective swimmer and a specialised hunter of slippery aquatic prey such as fish and amphibians. The species is strongly tied to lowland wetland ecosystems, including peat swamps, mangroves and riverine forests, habitats that are vanishing rapidly across Southeast Asia. The International Union for Conservation of Nature classifies it as endangered, primarily because of habitat loss and a chronic lack of ecological data. In Sarawak, only a handful of camera-trap detections had ever been published, and no long-term study had ever specifically targeted the species.

The study area, covering approximately 3582 hectares in southwestern Sarawak, is a challenging place to work. It lies far from human settlements and can only be reached by roughly 45 minutes of boat travel from the nearest village, across extensive swamp forests and tidally influenced wetlands. The landscape comprises a mosaic of ten distinct habitat types, ranging from near sea level in tidal zones to about 15 to 25 metres inland. The researchers focused on three habitat types: nipah swamp, dominated by the palm Nypa fruticans at low brackish elevations heavily influenced by tides; buta-buta habitat, dominated by Excoecaria agallocha in slightly raised and less inundated zones; and nibong habitat, dominated by the palm Oncosperma tigillarium on elevated ground between 10 and 20 metres above sea level, where freshwater tributaries support amphibians and small fishes.

Between August 2021 and beyond, the team deployed 148 camera trap units across three sampling zones separated by two river systems. Cameras were installed at 100-metre intervals along ten transect lines, between 500 and 1500 metres long, positioned roughly 300 metres inland from the tidal boundary and running parallel to the riverbanks. In dry, elevated areas the cameras sat about two feet above the ground, but in tidally inundated zones they had to be mounted as high as five feet and angled downward. Each unit captured three consecutive still images at 3-second intervals followed by a 10-second video per trigger, with no delay between triggers. A practical problem soon emerged: the automatic transition from still-image mode to video recording took 12 to 15 seconds, frequently producing incomplete footage showing only a tail or posterior region. From January 2022, all cameras were reconfigured to record video exclusively.

A pivotal moment came after 113 days of trapping, on 21 November 2021, when a single still image and video of a flat-headed cat were recorded at the ecotone between buta-buta and nibong habitat. That confirmed detection prompted a methodological shift from systematic transect placement to targeted strategy. Eighty-five camera units were repositioned along three new transects within the nibong habitat, primarily near puddles and comparatively drier microhabitats, while six additional cameras were deployed opportunistically outside the formal transects. The remaining units stayed in the lowland mangrove and nipah zones to enable habitat comparisons.

The camera network yielded 7426 independent wildlife records, but flat-headed cats accounted for just 0.8 percent of them, 59 detections with an encounter rate of 0.043. That rarity is consistent with the species’ naturally low population densities and its specialised wetland requirements. Among the detections were records of an adult female travelling with a kitten, a genuinely encouraging sign of breeding in the area. The broader community captured by the cameras was remarkable in its own right, including small-clawed otters, bearded pigs, Sunda pangolins, proboscis monkeys, mousedeer, treeshrews, monitor lizards and even estuarine crocodiles, confirming the site’s conservation significance.

The spatial pattern of detections was striking. Forty-nine of the 59 flat-headed cat records came from the nibong habitat, nine from buta-buta and only one from the nipah swamp, with no detections at all in the true mangrove areas. Habitat use differed significantly across the three types, and Bonferroni-adjusted confidence intervals showed that proportional use of nibong, between 0.712 and 0.944, was far greater than its availability of 0.333. Jacobs’ index, which ranges from minus one for strong avoidance to plus one for strong preference, delivered a value of 0.74 for nibong, confirming strong positive selection, while buta-buta showed moderate avoidance at minus 0.45 and nipah swamp near-complete avoidance at minus 0.95. Intriguingly, naive occupancy, the proportion of camera stations recording the species at least once, was highest in buta-buta at 0.364, followed by nibong at 0.231, suggesting that buta-buta functions as a transitional foraging area while nibong supports the most intensive, repeated use.

Within the buta-buta habitat, detections clustered at just two water puddle sites, with intervals between consecutive visits ranging from two days to nine months. In the nibong habitat, several stations recorded repeated detections, including events on up to three consecutive days. These patterns suggest that small water bodies act as key foraging microhabitats, and that nibong stands, which occur on slightly elevated, non-inundated ground close to water sources, offer an optimal configuration: aquatic prey accessible nearby, yet stable terrestrial substrates for efficient movement. The frequently inundated conditions of the nipah swamp, by contrast, appear to reduce suitability for this semi-aquatic but still terrestrial-moving felid.

Temporal analysis added another surprise. In the nipah swamp and buta-buta habitats, detections were largely restricted to night and early morning, matching the species’ traditional reputation as nocturnal. But in the nibong habitat, the cats were detected throughout the full 24-hour cycle, with repeated daytime activity recorded across the study. Kernel density estimation showed activity concentrated from evening through late morning, peaking around midnight, while the continued daytime detections in nibong suggest temporal flexibility rather than strict nocturnality. The researchers propose that reduced disturbance or higher prey availability in this habitat may allow individuals to forage throughout the day, a pattern occasionally documented in other small felids where environmental conditions favour opportunistic hunting.

The findings arrive at a consequential moment for Sarawak. The proposed Lundu Mangrove Protected Forest was first suggested for designation in 2000 and has remained in limbo ever since, but the state government’s initiative to expand the protected forest network to six million hectares, combined with the confirmation of the flat-headed cat and other wildlife of conservation importance, has put gazettement back on the table. The study’s authors argue that protecting nibong stands and adjacent wetland microhabitats, particularly areas containing small water bodies, should be a conservation priority, and that formal protected status would safeguard the hydrological conditions these specialised species depend on. Because the flat-headed cat is so tightly bound to coastal wetlands, it may serve as an indicator of ecosystem integrity, meaning conservation measures aimed at it would simultaneously benefit otters, pangolins, proboscis monkeys and countless other wetland-dependent species. With coastal mangroves across Borneo under pressure from land conversion, hydrological modification and human disturbance, this rare cat’s preference for a specific palm-dominated wetland may prove to be one of the most persuasive arguments yet for keeping this unique landscape intact.

Subject of Research: Habitat preference and spatiotemporal activity patterns of the endangered flat-headed cat in a proposed mangrove protected forest in Sarawak, Malaysia

Article Title: Habitat Preference and Spatiotemporal Habitat Use of the Endangered Flat‐Headed Cat in the Proposed Lundu Mangrove Protected Forest, Sarawak, Malaysia

Article References: Ampeng, A., Mohammad, H., Sait, H., Omar, A., Md‐Zain, B. M., Md‐Nor, S., Traeholt, C., Abang Abdillah, D. N., Osman, S., Ujeng, J., Cedric Jess, R. R., Abg Anoi, A. R., Ahmad, S. J., John Apiu, R. R., Deli, B., Yahya, H., & Mansor, M. S. (2026). Habitat Preference and Spatiotemporal Habitat Use of the Endangered Flat‐Headed Cat in the Proposed Lundu Mangrove Protected Forest, Sarawak, Malaysia. Ecology and Evolution, 16(10), Article e74248. https://doi.org/10.1002/ece3.74248

Image Credits: AI Generated

DOI: 10.1002/ece3.74248

Keywords: flat-headed cat, Prionailurus planiceps, camera trapping, mangrove wetlands, Sarawak, Borneo, habitat preference, Jacobs' index, kernel density estimation, nibong habitat, conservation, endangered species