southern tip of Madagascar

The study sites were located within the area covered by the Landsat scene shown at the southern tip of Madagascar. The photograph shown within the outline of Madagascar features a stable forest study site; the picture is from Elmqvist’s paper.

May 9, 2007 • A recently released study provides evidence that social dictums influence the stability of tropical dry forests more than proximate population density. Often, it is assumed that a rising local population means rising pressure on forest resources. However, a team of Swedish and Malagasy researchers found that social dictums can trump the influence of population density on deforestation. This conclusion was reached by analyzing the social norms regulating local forest usage in the semi-arid Androy region of south Madagascar in conjunction with Landsat-derived forest loss, gain, and stability data.

After a land cover classification of three dry-season Landsat scenes (1984, 1993, 2000), a 7% net loss of forest cover between 1984 and 2000 was found; but over the seven years between 1993 and 2000 a 4% increase in forest cover was observed.  Information on local forest use was obtained through interviews with local forest officials, village counselors, and clan elders. Interestingly, the largest forest loss occurred in an area with a small local population and a long distance to market––but with markedly weak and insecure property rights. At the same time, forest stability and reforestation was observed in more densely populated areas that had well-defined property rights (largely in the form of inherited land claims and trespassing taboos).

This research indicates that local dictums and community enforcement can have profound effects on forest stability and can lead to sustainable resource management.

Reporter: Laura Rocchio

+ Read the full journal article:
Elmqvist T, Pyykönen M, Tengö M, Rakotondrasoa F, Rabakonandrianina E, et al. (2007) Patterns of Loss and Regeneration of Tropical Dry Forest in Madagascar: The Social Institutional Context. PLoS ONE 2(5): e402.