Introduction

Abstract

Tropical open woodlands account for about 40% of the total area of open and closed woodland and forest in the tropics but have so far not received the attention that they deserve in the monitoring and modelling of global environmental change. This Special Issue explores some of the challenges that need to be faced if this neglect is to be overcome. With the exception of Furley's paper on the cerrados of Brazil, the main geographical focus is Sub-Saharan Africa ( Fig. 1). The papers in this issue derive from a workshop held at the University of Leeds on 10 December 1997 with funding from the European Union's Framework 4 Co-operation With Developing Countries (INCO-DC) Programme. The editors would like to express their thanks to those involved in reviewing the papers featured in this issue. Open woodlands neare Linguère in northern Senegal (Photo: D. P. Bradley) Modelling changes in land cover in open woodlands and relating the changes to their underlying human causes is, as Grainger shows, more difficult than modelling deforestation in tropical closed forests, since only a proportion of the total change involves outright clearance (deforestation) and the same area of woodland can be exploited for multiple uses by multiple users. This includes the harvesting of non-timber forest products for subsistence uses, as described by Schreckenberg. Modelling is also limited by poor data availability on woodland area and rates of change, partly resulting from constraints on the use of remote sensing techniques to monitor change. Lambin examines these constraints and suggests how to overcome them. Several papers provide empirical insights into the processes of woodland degradation—the term now commonly employed to describe lesser changes in woody cover that do not involve complete clearance. Dougill and Trodd show how grazing and the use of fire may change the balance between different vegetation layers. Sullivan finds little evidence to support the notion that woodland degradation declines with distance from settlements and indeed questions the scientific basis of continuing fears of degradation driven by over-exploitation of resources by local herders in her study area (north-west Namibia). Although above-ground biomass and biodiversity may well decline when open woodland is converted to cultivated land, Mortimore, Harris and Turner report that total land productivity does not necessarily decrease. A number of papers draw attention to the importance of spatio-temporal variability. Fuller reports considerable intra-annual variation in tree canopy cover in open woodlands. Homewood and Brockington argue that, unlike deforestation, it would be wrong to expect degradation in open woodlands to maintain a consistent trajectory, owing to an inherent variability to which humans may make only a minor contribution. Uncertainty engendered by the absence of reliable data can easily lead to commonly held perceptions having little basis in reality. Ribot warns against perceptions of degradation biased by localised phenomena or colonial prejudice (a theme also evident in Sullivan's paper). Homewood and Brockington argue that spatial bias in biodiversity exploration can lead to problems in conservation assessment and planning. Furley reports that, according to recent research, the biodiversity of tropical open woodlands is not quite as inferior to that of tropical closed forests as previously thought. Our hope is that the papers in this special issue may be of use to those engaged in empirical research in the field by highlighting particular data needs on which advances in understanding may be based. Just as importantly, we hope that they succeed in highlighting some of the important unresolved theoretical debates with respect to these ecosystems. In particular, these seem to focus on the tension between: (i) received views of degradation and the ecological thinking on which these views are based, and; (ii) the ideas of continuing flux, and non-equilibrium, in the light of which the significance of human impacts is often argued to be of lesser importance.