Population People

Updated July 1999

Population dynamics and the assessment of land use changes and deforestation, Part 1

by Rudi Drigo
FAO Consultant
and Alain Marcoux
Senior Officer (Population and Environment), FAO Population Programme Service
Ida-Eline Engh, Associate Professional Officer, SDWP, assisted in the preparation of this publication. This paper represents the views of the authors, and not necessarily those of the FAO

1. Population density and forest cover

Forests in general, and tropical forests in particular, have been receiving increasing attention since the first United Nations Conference on Human Environment in Stockholm in 1972. A high point was reached in 1992 at the United Nations Conference on Environment and Development held in Rio de Janeiro, which devoted a full chapter of its action plan (United Nations 1992) to forest conservation and development. The multiple roles of forests have been more and more widely recognized and stressed, together with the need to optimize the combination of benefits derived by society from the forest and minimizing conflicts around the exploitation of forest areas.

1.1 Population and deforestation: statistical correlation

The large degree of uncertainty associated with the information on trends in deforestation contributes to the concern for the current state of tropical forests [1]. Examples of related questions posed by policy-makers, the scientific community and the public at large are: What is the area of the remaining tropical forests? At what rate are they being depleted? Is that rate rising or declining? What are the causes of deforestation? What are its ecological, economic and social impacts? In what way do population factors make an impact on forest cover and deforestation rates? Indeed population factors are linked in important ways with the management and ultimately the conservation or degradation of forests, but those linkages are complex, because they are affected by a host of locally specific economic and social factors.

The association between human population dynamics and deforestation has been studied at various geographic levels and in various ecological and human contexts. We refer here to the Forest Resources Assessment (FRA) 1990 project conducted by FAO, the purposes of which were:

In the course of the global assessment a procedure had to be devised to adjust original forest cover observations - which had different reference dates from one country to another - to the reporting dates, 1980 and 1990. For this purpose, the statistical relationship between observed forest area change and ancillary variables, including population, was analysed. Analysis within the sample units demonstrated a significant relationship between forest cover and changes in human population density. The same analyses showed that the rate of change was influenced by ecological conditions which play an important role in the man-forest interaction. On the basis of these observations, the FRA 1990 programme developed an algorithm to project forest cover changes. Annex 1 describes the "pan-tropical deforestation model" developed for that project. The effect of population density growth on forest cover under different ecological conditions is illustrated in Figure 1.


Figure 1: Deforestation model curves for different ecological zones

At an elementary level, it is obvious that there must be an inverse correlation between population density and forest cover: as soon as humans build shelters and housing, they need to clear the wooded areas if these are present. If in addition they engage in agricultural activities, they need even more land per caput. In a dynamic perspective, this implies a relationship between population growth and deforestation. However, the intensity of that relationship can vary considerably, because it is mediated by a series of socio-cultural, economic and ecological factors. Hence the differences among populations as to the response to population growth in terms of extension (versus densification) of human settlements, extension (versus intensification) of agriculture, etc.

1.2 A case study of the Brazilian Amazon

In studying the relationship between population and forest cover changes, the spatial resolution of data sets - i.e. the size of the geographical units for which population and forest data are collected - is important. A few large units mean not only fewer cases on which to base correlation calculations, but also less information available on the variability of the said relationship from one ecological or human setting to another. Large unit size causes particular difficulties for analysis in regions where forest-people interfaces are highly dynamic, e.g. in the pioneer fronts of the Brazilian Amazon [2].

Better understanding the connection between population expansion and other factors in the process of deforestation in Brazil is important because that country holds a key position in the global picture of forest resources. In fact, the forest area lost annually in Brazil accounts for approximately one fourth of total tropical deforestation. However, population data available to FRA 1990 for matching with forest resources data were compiled at the state level, which is too broad to help describe in detail and better understand the population-forest interactions [3]. Therefore a separate effort was made in co-operation with Brazilian institutions to collect, and match, population data and forest cover data at the lowest administrative level (municipio). Building a detailed data set was of great relevance also in view of the rapid changes in forest cover that affect these regions, which may go virtually unnoticed in large-scale statistics.

New demographic time series and population maps were established for 138 spatial units. Inside Legal Amazon those units were municipios, the smallest administrative unit (groups of municipios, in high population density areas). Outside Legal Amazon the state level was maintained. At the same time an adequate data set and related maps of forest cover and forest change for the same units was undertaken; this required a considerable amount of GIS work using all available statistical and spatial information. The two sets of data were integrated into the Forest Resources Spatial and Statistical Information System (FORESSIS) of the FRA 1990 Project.

The deforestation model was then applied to the new data sets. The model in its general form is used to estimate forest area change at pan-tropical level. In this case, local observations of forest area and population change were used as to the inputs with the local value of the b1 parameter calculated for each unit. This enabled the interpolation of forest cover data for 1980 and 1990, and subsequently extrapolating them to 2000 using population projections.

Figure 2 shows the scattergram of the 138 units for forest cover and population density in 1991. There is a strong similarity between this distribution and the curve shown in Figure 1 for wet ecological zones.


Figure 2: Forest versus population density in Legal Amazon

The results consisted of time series of forest area change for each of the 138 units of the new data set. These results are illustrated by Figure 3. The maps show for each unit [a] the observed and projected deforestation rates for the periods 1980-1990 and 1990-2000, and [b] the deforestation trend (acceleration or deceleration) resulting from the comparison of the two decennial rates.


Figure 3: Deforestation rates and trends in Legal Amazon, 1980-2000

The information thus generated enables a more precise identification of the areas most at risk of deforestation - taking into account projected population growth - hence a better targeting of policy interventions. Without going into details here, the above maps show for instance that central Amazon - which was relatively unaffected during the 1980-1990 decade - is now undergoing an acceleration of the deforestation process.

The work done can serve also as a basis for adequately stratifying area samples in future studies where the pace of deforestation, or/and the rate of population growth, are relevant factors.

  • To Part 2, Population and land use changes in tropical countries

    Back to Top FAO Homepage