"Rural households need to reduce fuelwood use in favour of other energy alternatives," Konrad Wessels of the CSIR-Meraka Institute, South Africa, told environmentalresearchweb. "Currently the only viable alternative is electric stoves but the switch to electric stoves has been slow, apparently due to the cost of stoves and electricity."

The average household in the Lowveld region's communal areas – the former "homelands" established by forced resettlement during the apartheid era – burns three to four tonnes of fuelwood each year. Although electricity was introduced almost a decade ago, more than 90% of households still use fuelwood for cooking.

"Since unemployment is high, there is ample labour to collect free fuelwood, even as it becomes scarce," said Wessels. "However, as the fuelwood becomes scarcer and inaccessible on foot, more households are expected to purchase fuelwood from vendors. At some point the fuelwood might become so expensive that the switch to electric stoves becomes a cheaper option."

The team found that settlements were typically surrounded by an area of depleted wood, up to about 1.5 km away – equivalent to the maximum distance people will walk to collect fuelwood. In general, biomass increased at a rate of two tonnes per hectare for every 100 m beyond the 1000 m-point from a settlement.

"The use of alternative sources of energy for cooking should be promoted to balance out the current unsustainable rate of fuelwood extraction," said Wessels. "However, it will require interventions aimed at general poverty reduction and culturally acceptable energy alternatives. Various alternative energy stoves have been tested in these communal areas, but these remained engineering projects with little uptake by the community due to cultural preferences and practical considerations."

Along with colleagues, Wessels used light detection and ranging (LiDAR) from the Carnegie Airborne Observatory to estimate biomass over 25,000 hectares of sample sites in South Africa's north-east in 2008. The team, from the University of Pretoria, University of the Witwatersrand, CSIR-Natural Resource & Environment, CSIR-Meraka Institute and South African National Parks, all in South Africa, and Carnegie Institution for Science and Rochester Institute of Technology in the US, calibrated the data by measuring the diameter of woody stems in selected sites in the field.

"Airborne LiDAR data is the only way to effectively get tree height and tree cover over large areas to calculate woodland standing biomass," said Wessels. "The study area was strategically located within a long-term human surveillance site with a heritage of socioeconomic and environmental research. By combining these data with socio-economic data from household surveys, a supply–demand model could be applied to explore the future sustainability of fuelwood extraction under various scenarios."

Areas of communal landscape above granite rock typically contained less than half the biomass of the conservation areas of the Kruger National Park and SabiSand Private Game Reserve, both of which exclude people and domestic livestock but are subject to damage by fire and elephants. In communal savannahs above gabbro rock, areas more than 1200 m from a settlement had almost double the biomass of the protected parks. The team believes this is due to the absence of elephants, lack of accessibility for fuelwood extraction, and different fire management.

The village of Justicia proved ideal for the study as its location next to the SabiSand reserve and distance from other villages meant it was the only village taking fuelwood from its surrounding area. The team found that 67% of the 1279 households in Justicia use only fuelwood while 33% sometimes supplement fuelwood with electricity.

Now the team plans additional LiDAR campaigns to determine changes in woodland biomass and compare them to their supply–demand model predictions. "In order to scale-up these methods we are using the LiDAR data to "train" machine learning algorithms to map woodland biomass from optical and synthetic aperture radar satellite images across much larger areas to ultimately provide a national picture of woodland biomass," said Wessels. Wessel and colleagues published their study in Environmental Research Letters (ERL).