Identifying Key Drivers of Peatland Fires Across Kalimantan's Ex-Mega Rice Project Using Machine Learning.

Throughout Indonesia ecological degradation, agricultural expansion, and the digging of drainage canals has compromised the integrity and functioning of peatland forests. Fragmented landscapes of scrubland, cultivation, degraded forest, and newly established plantations are then susceptible to extensive fires that recur each year. However, a comprehensive understanding of all the drivers of fire distribution and the conditions of initiation is still absent. Here we show the first analysis in the region that encompasses a wide range of driving factors within a single model that captures the inter-annual variation, as well as the spatial distribution of peatland fires. We developed a fire susceptibility model using machine learning (XGBoost random forest) that characterizes the relationships between key predictor variables and the distribution of historic fire locations. We then determined the relative importance of each predictor variable in controlling the initiation and spread of fires. The model included land-cover classifications, a forest clearance index, vegetation indices, drought indices, distances to infrastructure, topography, and peat depth, as well as the Oceanic Niño Index (ONI). The model performance consistently scores highly in both accuracy and precision across all years (>75% and >67.5% respectively), though recall metrics are much lower (>25%). Our results confirm the anthropogenic dependence of extreme fires in the region, with distance to settlements and distance to canals consistently weighted the most important driving factors within the model structure. Our results may help target the root causes of fire initiation and propagation to better construct regulation and rehabilitation efforts to mitigate future fires.

Combined impacts of deforestation and wildlife trade on tropical biodiversity are severely underestimated.

Tropical forest diversity is simultaneously threatened by habitat loss and exploitation for wildlife trade. Quantitative conservation assessments have previously considered these threats separately, yet their impacts frequently act together. We integrate forest extent maps in 2000 and 2015 with a method of quantifying exploitation pressure based upon a species' commercial value and forest accessibility. We do so for 308 forest-dependent bird species, of which 77 are commercially traded, in the Southeast Asian biodiversity hotspot of Sundaland. We find 89% (274) of species experienced average habitat losses of 16% and estimate exploitation led to mean population declines of 37%. Assessing the combined impacts of deforestation and exploitation indicates the average losses of exploited species are much higher (54%), nearly doubling the regionally endemic species (from 27 to 51) threatened with extinction that should be IUCN Red Listed. Combined assessment of major threats is vital to accurately quantify biodiversity loss.

Global Demand for Natural Resources Eliminated More Than 100,000 Bornean Orangutans.

Unsustainable exploitation of natural resources is increasingly affecting the highly biodiverse tropics [1, 2]. Although rapid developments in remote sensing technology have permitted more precise estimates of land-cover change over large spatial scales [3-5], our knowledge about the effects of these changes on wildlife is much more sparse [6, 7]. Here we use field survey data, predictive density distribution modeling, and remote sensing to investigate the impact of resource use and land-use changes on the density distribution of Bornean orangutans (Pongo pygmaeus). Our models indicate that between 1999 and 2015, half of the orangutan population was affected by logging, deforestation, or industrialized plantations. Although land clearance caused the most dramatic rates of decline, it accounted for only a small proportion of the total loss. A much larger number of orangutans were lost in selectively logged and primary forests, where rates of decline were less precipitous, but where far more orangutans are found. This suggests that further drivers, independent of land-use change, contribute to orangutan loss. This finding is consistent with studies reporting hunting as a major cause in orangutan decline [8-10]. Our predictions of orangutan abundance loss across Borneo suggest that the population decreased by more than 100,000 individuals, corroborating recent estimates of decline [11]. Practical solutions to prevent future orangutan decline can only be realized by addressing its complex causes in a holistic manner across political and societal sectors, such as in land-use planning, resource exploitation, infrastructure development, and education, and by increasing long-term sustainability [12]. VIDEO ABSTRACT.

Fire Distribution in Peninsular Malaysia, Sumatra and Borneo in 2015 with Special Emphasis on Peatland Fires.

In this paper, we analyze the spatio-temporal distribution of vegetation fires in Peninsular Malaysia, Sumatra, and Borneo in the severe El Niño year of 2015, concentrating on the distribution of fires between mineral soils and peatland areas, and between land cover types in peatland areas. The results reveal that 53% of all Moderate Resolution Imaging Spectroradiometer (MODIS) fire detections were recorded in peatlands that cover only 12% of the study area. However, fire occurrence in the peatland areas was highly dependent on land cover type. Pristine peat swamp forests (PSF) experienced only marginal fire activity (30 fire detections per 1000 km2) compared to deforested undeveloped peatlands (831-915 fire detections per 1000 km2). Our results also highlight the extreme fire vulnerability of the southern Sumatran and Bornean peatlands under strong El Niño conditions: 71% of all peatland hotspots were detected in the provinces of South Sumatra and Central Kalimantan, which contain 29% of peatlands in the study area. Degraded PSF and all deforested peatland land cover types, including managed areas, in the two provinces were severely affected, demonstrating how difficult it is to protect even managed drained agricultural areas from unwanted fires during dry periods. Our results thereby advocate rewetting and rehabilitation as the primary management option for highly fire prone degraded undeveloped peatland areas, whenever feasible, as a means to reduce fire risk during future dry episodes.

Remotely sensed evidence of tropical peatland conversion to oil palm.

Rising global demands for food and biofuels are driving forest clearance in the tropics. Oil-palm expansion contributes to biodiversity declines and carbon emissions in Southeast Asia. However, the magnitudes of these impacts remain largely unquantified until now. We produce a 250-m spatial resolution map of closed canopy oil-palm plantations in the lowlands of Peninsular Malaysia (2 million ha), Borneo (2.4 million ha), and Sumatra (3.9 million ha). We demonstrate that 6% (or ≈880,000 ha) of tropical peatlands in the region had been converted to oil-palm plantations by the early 2000s. Conversion of peatswamp forests to oil palm led to biodiversity declines of 1% in Borneo (equivalent to four species of forest-dwelling birds), 3.4% in Sumatra (16 species), and 12.1% in Peninsular Malaysia (46 species). This land-use change also contributed to the loss of ≈140 million Mg of aboveground biomass carbon, and annual emissions of ≈4.6 million Mg of belowground carbon from peat oxidation. Additionally, the loss of peatswamp forests implies the loss of carbon sequestration service through peat accumulation, which amounts to ≈660,000 Mg of carbon annually. By 2010, 2.3 million ha of peatswamp forests were clear-felled, and currently occur as degraded lands. Reforestation of these clearings could enhance biodiversity by up to ≈20%, whereas oil-palm establishment would exacerbate species losses by up to ≈12%. To safeguard the region's biodiversity and carbon stocks, conservation and reforestation efforts should target Central Kalimantan, Riau, and West Kalimantan, which retain three-quarters (3.9 million ha) of the remaining peatswamp forests in Southeast Asia.

Status of peatland degradation and development in Sumatra and Kalimantan.

Peatlands cover around 13 Mha in Sumatra and Kalimantan, Indonesia. Human activities have rapidly increased in the peatland ecosystems during the last two decades, invariably degrading them and making them vulnerable to fires. This causes high carbon emissions that contribute to global climate change. For this article, we used 94 high resolution (10-20 m) satellite images to map the status of peatland degradation and development in Sumatra and Kalimantan using visual image interpretation. The results reveal that less than 4% of the peatland areas remain covered by pristine peatswamp forests (PSFs), while 37% are covered by PSFs with varying degree of degradation. Furthermore, over 20% is considered to be unmanaged degraded landscape, occupied by ferns, shrubs and secondary growth. This alarming extent of degradation makes peatlands vulnerable to accelerated peat decomposition and catastrophic fire episodes that will have global consequences. With on-going degradation and development the existence of the entire tropical peatland ecosystem in this region is in great danger.