What
Study of deforestation in the vicinity of the park, a UNESCO World Heritage Site.

Where
Taï National Park, 450,000 hectares, Côte d’Ivoire.

Keywords
Agroforestry, carbon credits, deforestation.

Context
Agroforestry is an integrated approach that increases total yields by combining food and annual crops with perennial trees and shrubs, and/or with livestock on the same land unit. These can be combined in a temporal sequence, in accordance with local ecological and socio-economic conditions. This approach helps to restore and sustain soil fertility.
Our work supported by Planet Action and UNESCO concerned an area of farming, agro-forestry and forest land in Côte d’Ivoire, in a buffer zone and a part of Taï National Park. A UNESCO World Heritage Site, this park is one of the last major remnants of the primary tropical forest of West Africa. Its rich natural flora and threatened mammal species, such as the pygmy hippopotamus and 11 species of monkeys, are of great scientific interest. Pro-Natura has been working with the Ivoirian non-profit association Vie et Forêt for over 6 years to develop practical training in agroforestry for peasant farmers living on the outskirts of the park, to reduce the pressure on its natural resources.

Partners/clients
Pro-Natura International, UNESCO, Planet Action.

Objectives
Study changes in the landscape and in particular deforestation in this complex zone from 2001-2007 using satellite imagery. As well as supporting analysis of deforestation, the results of this study will enable estimation of carbon stocks.

Methodology/Results
SPOT imagery was processed using to characterize the vegetation and identify water bodies.

The main information layers used to study vegetation are: green-leaf cover fraction, shade factor, brown/dry vegetation ratio, soil cover fraction and leaf moisture content.
A classification was established using a decision-tree tool provided in Overland. Results obtained show that:
• the dense forest zone inside the national park boundaries is well preserved, with little degradation at the edges
• shreds of forest in the farmland landscape are shrinking or disappearing
• a large reforested area is located in the south-west corner (in Liberia)
• there is a high degree of rotation between cropped fields and idle/fallow areas
• a large zone in the north-west corner of the park that appeared uniform in 2001 was more differentiated in 2007, with forest and cropped areas; this zone near the town of Zagné has seen an influx of population and a sawmill started operating there recently (Taï 2005)

Besides fulfilling their need for geoinformation to set up and monitor agroforestry activities, Pro-Natura and UNESCO decided to look at this pilot project in a carbon trading context. The world’s tropical forests are a huge reservoir of biodiversity and a key factor in maintaining the globe’s climate balances. More than half of them have disappeared in the last 30 years. Methodologies for inventorying carbon stocks are well defined, but surveying vast territories is a time-consuming and costly process and the approaches used sometimes generate significant uncertainties. Satellite data greatly reduce the degree of uncertainty when measuring forest biomass.

The land-cover classification maps produced can serve as a basis for stratification to ensure optimal positioning of sample plots to be inventoried in the field. These maps also make it possible to identify forest types and improve the spatial distribution of canopy biomass into different classes.

In conclusion, combining field and satellite data appears vital for measuring and monitoring a system designed to reduce carbon emissions due to deforestation and forest degradation.

Data used
Two SPOT 4 images (2001), one SPOT 5 image (2006) and one SPOT 4 image (2007).

Land-cover classification
Dense forest, forest vegetation (degraded forest, young trees, plantations, etc.), cropland (bare soil, developed fields, etc.), other types of vegetation (idle, fallow, savannah, wet grassland), bare ground (main villages, uncultivated areas), water, cloud (soil not visible), cloud shadow (soil unclear).