Forest cover dynamics of shifting cultivation in the Democratic Republic of the Congo 2000–2010 (2015 Environ. Res. Lett. 10 094009)

An error in the unit conversion from pixels to hectares lead to all the areal quantities in the text being smaller than they should have been. Only the number of hectares were changed; none of the text nor tables were changed. The changes do not affect the overall results or conclusions.

Shifting cultivation has traditionally been practiced in the Democratic Republic of Congo (DRC) by carving agricultural fields out of primary and secondary forest, resulting in the rural complex: a characteristic land cover mosaic of roads, villages, active and fallow fields and secondary forest. Forest clearing has varying impacts depending on where it occurs relative to this area: whether inside it, along its primary forest interface, or in more isolated primary forest areas. The spatial contextualization of forest cover loss is therefore necessary to understand its impacts and plan its management. We characterized forest clearing using spatial models in a Geographical Information System (GIS), applying morphological image processing to the Forets d'Afrique Central Evaluee par Teledetection (FACET) product. This process allowed us to create forest fragmentation maps for 2000, 2005 and 2010, classifying previously homogenous primary forest into separate patch, edge, perforated, fragmented and core forest subtypes. Subsequently we used spatial rules to map the established rural complex separately from isolated forest perforations, tracking the growth of these areas in time. Results confirm that the expansion of the rural complex and forest perforations has high variance throughout the country, with consequent differences in local impacts on forest ecology and habitat fragmentation. Between 2000 and 2010 the rural complex grew by 10.2% (2 771 238 ha), increasing from 11.9% to 13.1% of the total land area (1.2% change) while perforated forest grew by 74.4% (1 431 355 ha), from 0.8% to 1.5%. Core forest decreased by 3.8% (3 291 142 ha), from 38% to 36.6% of the 2010 land area. Of particular concern is the nearly doubling of perforated forest, a land dynamic that represents greater spatial intrusion of forest clearing within core forest areas and a move away from the established rural complex.

Forest fragmentation results
In 2000 core forest accounted for 86 868 974 ha (38%) of DRC land area. In 2010, of that core forest, 3 291 142 ha (3.8%) became patch, edge, perforated and fragmented forest. Patch forest increased by 12.4%, edge forest by 2% and perforated forest by 53%. Of the 11 second-level administrative units, in 2010 Equateur had the most core forest, followed by Province Orientale, Nord-Kivu and Maniema. Maniema also had the most fragmented forest, followed by Province Orientale and Kasai-Occidental. The largest losses of core forest in the study period were in Nord-Kivu (−3.1%), Equateur (−2.5%) and Sud-Kivu (−2.4%).
In 2000 the 6 CBFP landscapes accounted for 32 754 657 ha (36.6%) of core forest. Of that core forest 912 234 ha (2.8%) had been lost by 2010, 0.7% less than the country average. The landscape with the highest percentage of core forest is Salonga-Lukenie-Sankuru (93%) followed by Ituri-Epulu-Aru (87%). Virunga had the highest percentage of core forest loss, while the lowest was in Salonga-Lukenie-Sankuru. Fragmented forest increased in all but one landscape, caused by the loss of viable forest corridors, eroded by expanding edge forest as well as increased forest perforation areas. Virunga and Ituri-Epulu-Aru had the highest rates of growth of perforated forest area, showing a higher-than-average move away from established rural complexes. Patch forest grew in all landscapes and by almost 50% in Ituri-Epulu-Aru, an important dynamic which highlights isolated forest patches completely enclosed by the rural complex but not reported as different from primary core forest in products such as FACET. These isolated forest patches can have fundamentally different ecological functions and habitat characteristics than core primary forest.
Of the protected areas (PA), 15 have no core or fragmented forest. Of the remaining 21 with core forest, all of them lost some during the study period, 10 of them losing more than 2%. PAs contained 13 240 495 ha of core forest in 2000 and lost 174 578 ha (1.3%) to fragmented classes by 2010. Perforated forest increased by over 10% in 10 PAs. Patch forest increased in 28 PAs, more than 50% in 6 of these, and doubling in 2 of them.

Rural complex footprint results
The rural complex grew by 10.2% between 2000 and 2010, a yearly average rate of 1%, from 11.9% to 13.1% of the DRC total land area (a 1.2% change in area). This change added 2 771 238 ha of rural complex to the country. The area of established agriculture is growing, although with high variability throughout the country. Nord-Kivu and Sud-Kivu had the highest rural complex growth, probably because of the growth in population densities, and seven other provinces had over 10% rural complex growth. Kinshasa, Bandundu and Bas-Congo, as expected, saw much lower rural complex expansion rates due to the nearly exhausted primary forest resources there. In Katanga the isolated forest perforation area grew greatly in proportion to the low baseline rural complex and isolated perforation areas in 2000. Sud-Kivu, Province Orientale, Nord-Kivu, Maniema and Equateur all had perforated forest growth between 83% and 110%.
The rural complex grew in all the CBFP landscapes. Lac Tele-Lac Tumba had the largest 2010 rural complex footprint area (17.8%), followed by Maiko-Tayna-Kahuzi-Biega and Maringa-Lopori-Wamba. The highest relative rural complex footprint growth occurred in Ituri-Epulu-Aru (34.5%). Lac Tele-Lac Tumba had the highest forest perforation area, covering 2.24% of the landscape in 2010 while the highest forest perforation area growth occurred in Ituri-Epulu-Aru.
Among PAs, the highest rural complex expansion occurred in the Bombolumene Hunting Reserve (45.4% of its areas in 2000) followed by Lomako-Yokala Natural Reserve, Kahuzi-Biega National Park and Tumba-Lediima Nature Reserve all in the 22%−25% range. If normalized by the amount of standing core and fragmented forest, the rural complex footprint growth seen in parks with a higher ratio of rural complex area to core forest area, such as Virunga National Park, is higher than it would seem otherwise, indicating larger impacts in those areas where rural complex expansion has more likelihood of eroding and fragmenting core and fragmented forest. Perforated forest more than doubled in at least 12 PAs.