SD : People : Population dynamics and the assessment of land use changes and deforestation, Part 2

Updated July 1999

Population dynamics and the assessment of land use changes and deforestation, Part 2

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

< from Part 1

2. Population and land use changes in tropical countries

2.1 The assessment of land use changes through remote sensing ^[4]

Estimates of forest cover area and of the deforestation rate at country level merely indicate the magnitude of the deforestation issue through a modeling process. In order to undertake countermeasures, or simply to diagnose problems, a broader and more thorough understanding of the overall land use change process is essential. This entails answering the following questions:

How do tropical forest resources change?
How much of those are degraded? Or fragmented?
What are the causes of deforestation?
What becomes of deforested land?
What ecological and socio-economic conditions play a role in this transformation?

The only satisfactory way to provide reliable information on land use changes is to establish a global forest resources monitoring system, using a consistent method, providing reliable location-specific information on such change. Based on considerations of cost, precision and timeliness, an approach based on sampling and remote sensing was designed and applied to cover the entire tropical zone in a time frame of about one year. The specific objectives were:

To achieve the highest level of consistency and precision in the assessment of forest cover state and change at global and regional levels;
To develop and disseminate a simple and robust monitoring technique for producing estimates of forest cover state and change at global and regional levels with application also at national level;
To provide data for estimating changes by category of land cover and forest cover between two dates of interpretation at each location in the sample, and for producing change matrices at regional and global levels.

The method provides forest cover change data and maps for each sample location. A change matrix is produced for each location from the interdependent interpretation of two satellite images: the "historical" image acquired around 1980 and the recent one acquired around 1990. This enables estimation of class to class changes in land cover between the two dates of interpretation, which is best presented under the form of a matrix (see Figure 4). This information is essential to understand the complex processes taking place, such as deforestation, forest fragmentation and degradation, or afforestation.

Once established for a sample location, matrices can be aggregated, enabling assessments e.g. by geographic region, or by ecological zone. Prior to that, the matrices must be mathematically standardized for the exact period 1980-1990 on the basis of the annual class-to-class transition probabilities observed in each matrix.

Figure 4: Change matrix for an FRA 1990 sample area

An example is supplied by the following table (Figure 5), which synthesizes the changes assessed for the tropical areas as a whole, based on measurements done in 117 sample units - 47 in Africa, 30 in Asia and 40 in Latin America (FAO 1996).

The class-to-class changes reported in this table illustrate the complexity of the dynamics involved. As expected from the consideration of such a large and diversified area, almost all the possible class-to-class changes are represented. However, most of the area changes are located above the diagonal, implying a loss of forest area or cover, fragmentation etc.

This table also shows that thinking only in terms of deforestation is too simplistic. Deforestation is a multifarious phenomenon involving complex processes. It is important to consider all the aspects of those processes, i.e. the patterns of class-to-class changes. The changes observed in a given class are read along the respective rows, e.g. closed forest:

between 1980 and 1990, 6.4 percent of the original closed forest cover ([1367.96 - 1280.06] ( 1367.96 Mha) was transferred to other classes;
the most important change that the 'closed forest' class underwent was 34.79 Mha (39.6 percent of the lost area) that changed into other land cover, i.e. to permanent agriculture and to cattle ranching;
another 24.5 percent (21.57 ( 87.9 Mha) was converted to short fallow agriculture; etc.

Figure 5: Pan-tropical area transition matrix for the period 1980-1990

From the above table it appears that the most common class of destination of the changes - the one that gained most - is other land cover (mainly through a shift to cattle ranching and permanent agriculture), followed by short fallow and fragmented forest. The most important class-to-class transitions are: closed forest --> other land cover; closed forest --> short fallow; and shrubs --> other land cover.

The data of change matrices can also be presented in the graphic form of flux diagrams that afford a quick view of the relative importance of the various types of change: Figure 6 thus illustrates the pan-tropical transition matrix. The vertical ranking of land cover classes in the diagram corresponds to lesser and lesser woody biomass down the column (see scale on the left). Changes to a lower-ranking class (reported on the right side of the diagram) correspond to a loss of biomass and degradation of the cover: hence the label 'negative'. An examination of the said diagram shows that:

almost all the types of negative changes are represented, but negative changes are far more prevalent than positive ones;
the two largest transitions, namely closed forest --> other land cover and closed forest --> short fallow, also imply the largest losses of biomass;
for all other transitions, the estimated loss of biomass (represented by the area of the rectangle inside the line indicating the transition) depends more on the biomass gradient (vertical scale) than on the area involved in the change (horizontal scale);
most of the new plantation areas (both tree crops and forestry) have been created on previous closed forest area, implying a loss of biomass ^[5].

Figure 6: Pan-tropical woody biomass flux diagram
Note: Only class transitions involving more than 1.5 million hectares are shown in the diagram

2.2 Role of population factors in land use changes ^[6]

Human actions that cause these various categories of changes are influenced to different degrees by population dynamics (population growth, migration etc).

The change from closed forest to open forest is mostly the effect of fuelwood exploitation, grazing and related fire practices, and selective logging. Population dynamics (e.g. population growth, migration flows, urbanization) plays a role in these phenomena:

Fuelwood exploitation. To cover energy needs, most households in developing countries resort to freely gathered biomass fuels - crop residues, animal dung and, most of all, fuelwood. When the annual use of wood exceeds the sustainable yield of wooded areas, forests and woodlands are gradually destroyed. The impact of population growth on fuelwood consumption is direct, since energy needs are essentially proportional to population size. Urbanization also plays an important role. A first effect arises from the higher energy consumption per caput that occurs in cities. A second one arises from changes in habits: urban dwellers tend to prefer charcoal to wood; this increases the impact on wood resources per consumption unit. (On the other hand, where they are better off, urban populations may have access to alternative energy sources that do not affect local forest resources.) Finally, population concentration itself focuses the impact on wood resources on a peripheral zone of the city: instead of limited local damage spread over large forest tracts, one often observes deforested areas spreading circularly around the cities.
Grazing. Excessive pressure on the vegetal cover by animals can be a crucial problem. Examples such as the damage caused by goats in the Mediterranean area and elsewhere are well known. For the most part, however, the relationship between human population dynamics and overgrazing is not a matter of population growth driving the changes in livestock density. Such changes usually are driven by market forces or by animal health factors. Economic circumstances, or special goals such as the need for security, causing families to increase stocking densities, also play a part.
Logging. This component is weakly related to population growth; demand for logging products is more affected by income growth and export needs than by domestic population size or urbanization. In addition logging concerns smaller areas than forest clearing for agriculture, does not destroy the whole vegetation, and does not involve the destruction of organic matter (including seeds) that forest burning does. It does play an enabling role by opening access roads, but it does not create the need for land clearing.

The change from closed forest to long fallow, then to short fallow, is an effect of expanding small-scale subsistence farming (shifting cultivation), indicating the pressure of rural population. In effect, land clearing in shifting cultivation is essentially driven by population growth and related increases in requirements for food and other agricultural products. This pattern of change is dominant in Africa and significant in Asia, but less common in Latin America (see Annex 2).

The change to other land cover corresponds to the development of permanent agriculture and pastures, meaning that during the period observed the forest has been replaced by permanently cultivated fields or by ranches. As the period of observation (about 10 years) is relatively short as compared to the time frame of change in agricultural systems, the fact that there has not been a transition through the long and short fallow stages suggests either massive immigration or - more commonly - investment through planned operations, usually to develop commercial agriculture ^[7]. The change to plantations, of course, also indicates a similar pattern of planned operations. These two types of transition are evident in Latin America and in Asia, and to a much lesser extent in Africa (see Annex 2).

The change to fragmented forest results from partial deforestation by clearing of small patches of forest, which creates a mosaic of forest and non-forest cover; in turn, fragmented forest mainly changed to "other land cover", meaning that fragmentation is an intermediate stage towards permanent agriculture.

Overall, population pressure is an important factor in vegetation loss, especially in areas with scarce land reserves and limited sources of energy. To this extent, the close statistical association of population growth with deforestation easily leads to foresee further loss of forest cover in the wake of projected population growth. For instance the FRA 1990 report (FAO 1993) noted that the population of the developing regions was expected to grow from 4 to 7 billions during 1990-2020, concluding that "as a result, the risk of deforestation in the developing countries remains very high".

Such a statement probably holds at the broad geographical level to which it refers - i.e. developing countries as a whole - because local variations in the form and intensity of the population-deforestation linkage tend to offset one another, so that the overall correlation can be expected to change only slowly. But at country level - and a fortiori at the local level - a closer look at the linkages is required, because the mix of factors that impinge on forest futures varies considerably. We shall now look into this question.

2.3 Environmental impact of different population groups

As shown above, deforestation arises from a variety of human actions. At the small locality level, the nature and magnitude of the effect will depend on the types and mix of actions under way. It is important to note, in this respect, that the various types of actions may be undertaken by different population groups (loggers, shifting cultivators etc). At the regional or country level, the overall effect will therefore depend inter alia on the relative size of those groups.

Figure 7 provides a good illustration of this aspect of the population-deforestation relationship. That large tracts of forest have been cleared is evident, but the aspect of major interest here is that three different types of human impact can be distinguished:

The upper right part of the image corresponds to the location of an indigenous reserve. The population has a low density and a very low impact on the environment, with a negligible contribution to the deforestation rate.
The lower right part of the image corresponds to a resettlement area (another is partly visible in the upper left corner). The population density of settlers is much higher, and their contribution to deforestation is large.
The central part of the image corresponds to the establishment of large cattle ranches. Although population density is very low there, the impact on the forest is large since this type of land use has a high environmental impact per person involved.

Given minimal knowledge of the area - or the ability to interpret its deforestation patterns by analogy with known locations - the information supplied by this type of image enables formulating diagnoses with a timeliness that could hardly be matched if statistical data on population and areas had to be collected. When it comes to apply the information to policy making, however, it is necessary to quantify the relevant factors at work on a broad scale, hence to document those factors, other than mere total population size, which contribute to shape the relations between population and deforestation.

In a situation such as the one illustrated above, a view on the future of deforestation - magnitude of the risks, assessment of the pros and cons, means to counter it - requires more than sound data on population trends (natural growth and migration). It is important to take into account the nature of the activities undertaken by the population in question.

Figure 7: Land use in a deforested area

Policy-wise, seeking to minimize ecological damage would entail hard choices between land use changes that bring about very different economic outcomes (e.g. with regard to output value or potential foreign currency benefits) but benefit very different groups of population (in nature and size). Of course, economic policy measures proper are required. For instance shifting cultivation - the main factor in deforestation world-wide - uses several times more land per unit of produce than permanent agriculture does, but permanent agriculture requires inputs that some rural populations cannot afford. In this case, fiscal or price policies might be justified.

From the scientific viewpoint, the above information highlights the value of analysing population-deforestation linkages at several geographical levels, in order to combine complementary angles of analysis and avoid misconceptions and simplifications. The close statistical association at global and regional level revealed by the FRA 1990 report has its usefulness, but admittedly embodies more complex linkages. Consequently, geographically more detailed analyses are crucial when it comes to understanding and intervening on processes at the national or sub-national level. Besides population characteristics and dynamics, relevant elements in such analyses are land tenure, local political institutions, and the market relations in which local producers are engaged. This may have to include international markets which, as illustrated by the example above, can have an indirect but strong impact on the local environment.

Annex 1: Estimating the forest change model

Note: This model was developed by Walter A. Marzoli for the FRA 1990 exercise.

Very much like biological growth processes, deforestation has been observed to proceed relatively slowly at initial stages, much faster at intermediate stages, and slowly again at final stages. Using this analogy, forest area change can be represented by a differential equation of the form:

dy/dp = b₁ Y ^b2 - b₃Y      [1] where y is the percentage forested in a given area,
    dy/dp is the derivative of y with respect to population density,
    and b₁, b₂ and b₃ are parameters.

A high correlation between population density and forest cover was found using a logarithmic transformation. Measuring population density in persons by square kilometre, P is defined as:

P = log_n (1 + population density)

The change model structure can be briefly described as follows:

(i) dy/dp : Dependent variable. It is the ratio between population density change and forest area change, representing the response 'dy' to a given change of population pressure 'dp'. The higher the value of the ratio, the higher the change in forest cover per unit of population change. dy/dp is a function of the 'size' of the forest (Y in equation [1]) and of the following parameters:
(ii) b₁ : This parameter can be taken to represent productivity - or site quality, to follow on the analogy with yield modelling. In fact in the present model formulation it is a function of bio-climatic parameters. The meaning of b1 is that the same population growth rate has different effects in different ecological conditions. If time series of forest cover and population are available, the b1 parameter can be computed from the observed values.
(iii) b₂ : This parameter represents the culmination point of the derivative function. In the State Model curve (see below) it is the inflection point of the curve, where the deforestation rate culminates and start to decrease.
(iv) b₃ : This parameter is related to the maximum possible deforestation, represented by the asymptotic value of Y where an additional population increase has no effect on forest cover, which remains stable in time. This parameter can be related to accessibility (both physical and legal) of forest resources and to land suitability of forest areas for transfer to other land uses.

The state model

Integration of the differential equation [1] leads to a Chapman-Richards function of the form:

y = a₀ (1- a₁ e ^{-a 2 P}) ^a3 [2] State model

The cumulative growth curve defined by this equation has a sigmoid shape and an upper asymptote. The function gives the estimated forest area for a given population density level. The following system of equations defines the four model parameters:

a₀ = y_m

a₁ = 1 - (y₀ / y_m) ^{(1 - b2)}

a₂ = (1 - b₂) b₁ y_m ^{(b2 - 1)}

a₃ = 1 / (1 - b₂)

General model

where y_m = maximum possible non-forested area
y₀ = non-forested area for P = 0
b₁ and b₂ are parameters of the change model

It can be noticed that an additional variable y0 is needed to solve the state model equality [2]. Y₀ represents the initial condition of the site, i.e. the state of forest cover for a zero population density.

As new information and data become known, such models can be improved; FAO is in effect working to refine this model.

Annex 2: Regional biomass flux diagrams

The effects of rural population pressure are very clear in Africa where the single most dominant transition has been closed forest --> short fallow (small-scale subsistence farming) and where many transitions have been represented with similar frequency; the sequences closed forest --> open forest --> fragmented forest --> other land cover are typical and they represent clearly the various progressive stages of forest depletion.

In Asia this type of pressure is mainly represented by the sequences closed forest --> long fallow --> other land cover and closed forest --> short fallow --> other land cover, reflecting the expansion and intensification of shifting cultivation (classes Long Fallow and Short Fallow) in the hills of South East Asia. Abandoned shifting cultivation areas are represented by the positive transition long fallow --> closed forest. The difference between the forest area going into long fallow and the area of long fallow reverting to forest shows how unbalanced this originally sound practice has become.

In Latin America this type of change was less common but still visible in the transition closed forest --> short fallow, largely the effect of small-scale farming in Amazonia and/or Yucatan. The effects of centrally planned operations are evident in Latin America and in Asia but to a much lesser degree in Africa. The typical transitions have been closed forest --> other land cover or, relevant in Asia only, closed forest --> plantation. Typical land uses related to these processes were: large areas of deforestation owing mainly to cattle ranching in the Brazilian Amazon, large resettlement and plantation programmes in South East Asia and, to a lesser degree, in West Africa..

Notes

1. Forests are defined as ecosystems with at least 10 percent crown cover of trees or/and bamboos, generally associated with wild flora, fauna and natural soil conditions, and not subject to agricultural practices. Deforestation refers to change of land cover with depletion of tree crown cover to less than 10 percent.

2. This section is based on the results of collaborative activities jointly identified by the Population Programme Service and the Forest Resources Assessment Programme, funded by the former and conducted by Rudi Drigo (FAO 1998).

3. The 'Legal Amazon' comprises nine federated States, some of which (Amazonas, Para) are almost three times the size of France.

4. This section is largely based on FAO (1993).

5. Source for the diagram and comments: FAO (1996). Annex 2 presents the three respective diagrams, with comments, for the major developing regions, namely Africa, Asia and Latin America.

6. This section is based on FAO (1996) and Marcoux (1995).

7. Forest clearing for pastures is comparatively a minor factor on a global scale, but it is important in certain countries. It is driven by demand for animal products rather than by population growth, but its intrinsic impact (say by consumer concerned, or by calorie produced) appears to be much greater than that of other possible uses of cleared land. Section 2.3 illustrates this issue.

References

FAO. 1993. Forest resources assessment 1990 - Tropical countries. Forestry Paper 112. Rome.

FAO. 1995. Forest resources assessment 1990 - Global synthesis. Forestry Paper 124. Rome.

FAO. 1996. Forest resources assessment 1990 - Survey of tropical forest cover and study of change processes. Forestry Paper 130. Rome.

FAO. 1998. Population in deforestation assessment. Development of demographic data at sub-state level and study of relation between population and deforestation in Brazil. Final report.

Marcoux, Alain. 1995. Population and the environment: a review of issues and concepts for population programme staff - II. Population and land degradation. FAO.

United Nations. 1992. Agenda 21 - The United Nations Programme of Action from Rio. New York.