Previous Page Table of Contents Next Page

Using farm trees for fuelwood

Gunnar Poulsen

GUNNAR POULSEN is a Danish forestry and land-use consultant with extensive experience in Africa.

One of the foundation stones of forest management is forest mensuration - comprising primarily stocktaking of the standing volume of wood and assessment of the growth rate. The basic data for calculating volume are: diameter at breast-height (dbh), the height of the marketable part of the trunk or measurement to the top of tree, and the number of trees per unit area. The current annual increment is found by combining the volume figures with additional parameters for diameter and height growth.

Following this procedure, it is possible to obtain a fairly exact evaluation of the stock and growth rate of a forest stand. However, it is rarely possible to apply such a detailed and time-consuming method to larger areas, let alone to the entire forest estate of a region - or a country. Any large-scale evaluation must necessarily be based on sampling in some way, or in other words, by limiting actual mensuration to a small but representative part of the whole forest and from there arriving at figures for the total through extrapolation. A frequent variation of the method consists in cutting out the time-consuming mensuration of individual diameters and basing volume calculation on direct measurement of the compound basal area by means of a relascope. Nevertheless, whether one or the other approach is used, the basic principle remains the same. The results arrived at are function of diameters in breast - height, tree heights, form factor, density of stand (number of stems per unit area), and growth parameters for diameter and height. Additionally, it is often assumed that variations in stand density, within quite wide margins, do not influence the rate of volume growth. In this context, it should be pointed out that this is an assumption which does not hold true for all trees and environmental conditions.

The mensuration system as outlined here in its broad features is almost certainly well conceived for surveying volumes and growth rates for a wide range of forest types, both natural and artificial. However, a far from negligible and almost certainly rapidly expanding proportion of wood supplies in the tropics is not obtained from forest stands but from hundreds of millions of "farm trees" and other ligneous vegetation spread across crop and pasture land.

The motive for writing this article and for starting by describing the commonly used method for surveying forest resources is a nagging doubt concerning the suitability of this method for assessing the productive potential (in terms of wood) of farm trees and drawing conclusions on this basis concerning the wood supply situation, the need for afforestation, etc.

Farm trees, whether they be Grevillea robusta. Croton megalocarpus and Eucalyptus saligna of some highland areas in East Africa, or Anogeissus leiocarpus, Daniella oliveri and Prosopis africana in the north of the United Republic of Cameroon, are submitted to a silvicultural treatment which has little in common with forest practices as they are currently applied within reserves. This does not mean that farm trees are usually mismanaged. On the contrary, many farmers apply perfectly rational methods when dealing with their tree resources, methods that make sense within the context of the needs and possibilities of small farms.

What gives farm tree management its distinctive character is the widespread use of pruning and pollarding. To give an example, let us try to follow the life cycle of a row of Grevillea trees on a typical small farm on the slopes of Mt Kenya.

Grevillea robusta is the dominant farm tree in this region, with Croton megalocarpus and Eucalyptus saligna occupying second and third positions. The species we are choosing to observe is favoured because, if well managed, it does not compete severely with neighbouring crops. Its root system grows mainly in depth with only a few shallow, lateral roots. The foliage is feathery, throwing a relatively light shade. Furthermore, Grevillea robusta is a species that can be established simply and cheaply by direct sowing, or by using naturally occurring seedlings. Last but not least, the wood is appreciated as both fuel and timber.

On a typical farm, Grevillea trees may be seen forming single lines with 2 to 4 metres between stems, or they may be evenly spread across fields at a spacing of perhaps 10 x 10 metres.

Quite often, a combination of the two spatial arrangements is used. The density on a well-wooded farm may be of the order of 100-200 trees per hectare, apart from the possible existence of more densely stocked wood-lots.

Let us now assume that the farmer planted the row of trees we are going to observe five years ago and that they have grown vigorously ever since. Now the 8 to 10-metre-tall trees are beginning to cause problems. They compete too much with the neighbouring coffee, maize and cowpeas. It is mainly a matter of competition for water resources within a 5-metre-wide belt on each side of the tree line, but the shade of the trees is also beginning to become a problem.

The farmer has a simple solution to this which enables him, in fact, to kill three birds with one stone. He will prune the trees severely, removing all branches until nothing is left of each tree but the naked stem, looking rather desolate like a flag-pole without a flag. In this way, he will reduce competition for water and light almost to zero and he will at the same time obtain an always welcome supply of small wood.

Despite the excessive nature of the intervention (from a forestry point of view), the trees start sprouting fresh foliage surprisingly fast, from bottom to top of the denuded stem, making them look like giant bottle-brushes. With time, a full, vigorous crown reappears. Unfortunately and unavoidably, this development of the crown is accompanied by a corresponding expansion of water consumption and shading. Presently the farmer will arrive at the conclusion that the level of competition is passing that threshold beyond which the adverse effect on the crops exceeds the advantages produced by the trees. He will then carry out a second pruning - likely to be no less violent than the first. In this way, a pruning rotation is started, involving the removal of all branches at two- or three-year intervals, and sometimes even including a shortening of the terminal shoot. The system may be continued in this way for as long as 30, 40 or 50 years and may involve as many as 15 or 20 prunings. Despite all this interference, the trees will continue increasing in diameter and also in height, unless this is prevented by top-pruning. However, growth will be at an uneven pace. Just after each intervention, it will practically come to a standstill. Then it will gradually pick up again and reach a maximum just before the following pruning again deprives the trees of their photosynthesizing structures. Anyhow, at some stage the farmer will almost certainly become aware that the diameters have reached a size which will make it economically attractive to fell the trees and dispose of the logs for saw timber. He will either pit-saw them on the farm or sell them to a mill. This, then, will be the end of the life cycle of the tree row. Over the years the amount of wood obtained from the many prunings and the final exploitation involving entire trees, trunks and all, will have been considerable and possibly of fundamental importance for the proper functioning of the farmer's household.

RECENTLY EXPLOITED Grevillea robusta IN KENYA branch regrowth accounts for 60 to 80 percent of production

However, most likely little or nothing will be known about the quantity of wood actually produced. For good reason it will not be known either how much of the total was branch wood and how much stem wood. At a guess, the branches may have provided as much as 60 to 80 percent of the total.

Markhamia platycalyx, A GOOD FARMLAND FUELWOOD SPECIES production cannot be calculated from dbh and height

Data from sample plots, if available, will be of little guidance for arriving at a more precise evaluation. They apply to trees grown under fundamentally different environmental conditions, at least in some significant respects. The sample plot trees are exposed to competition, below and above ground and on all four sides, whereas the farm trees enjoy practically maximum available radiation on two sides and are also exposed to relatively little competition where their roots extend below those of the usually more shallow-rooted crops. No less important, the silvicultural treatments applied in the two cases have scarcely anything in common. Even an eventual application of "suitably adapted" trial plot data runs up against a fundamental difficulty. The trees on the farm do not occupy a well-defined area that can be used for calculation purposes. Strictly speaking, only the basal area of the stems constitutes a full-time, measurable occupancy. The ground coverage of the crowns, on the other hand, fluctuates continuously between nearly zero and a score or two of square metres per tree. All the same, the foliage is never allowed to reach such a density and extension that it would impede the satisfactory development of crops on more than a fraction of the land. Just as it is not possible to apply trial plot data for volume/hectare to trees which do not occupy a well-defined area, it does not make sense to use growth data of a similar origin for trees growing on a single line or widely scattered across farm land and exposed to conditions that have little in common with those inside a closed stand. In the case of the farm trees, both height and diameter growth can be expected to fluctuate widely and rhythmically together with the periodic changes in crown cover, reaching their maximum just before each pruning and then dropping practically to zero before the start of a new expansion.

A FREQUENTLY POLLARDED Grevillea robusta a one-tree wood-lot

During periods within the rotational cycle when the crown-cover is wide, growth rates can be expected to exceed those measured inside closed stands. Conversely, within a time interval after each pruning, they will necessarily be lower than those of the closed stands. Whether the average growth-rate of the individual trees will be superior or inferior to that of trees within a stand must depend on the length of the pruning intervals. It will probably not be wrong to assume that the mean annual increment will be a direct function of the length of the intervals.

Sample plot data are based on the mensuration of stem and branch parameters inside stands where silvicultural interference mainly takes the form of periodical thinning and, possibly, pruning to a height of two-thirds of the total - at the most - and with no expectation of regrowth. Consequently, no appropriate figures will be available from this source for assessment of the most important aspect of wood production on the farm, which is branch regrowth after pruning.

The inappropriateness of conventional forest mensuration techniques is not confined to the farm trees. In their elementary form these methods are equally ill-suited to recording the productive potential of hedgerows and other small shrub vegetation from which women and children quite often pick a large proportion of the total supply of domestic fuel, an activity for which they, it should be remembered, are often solely responsible.

Complete stem-pruning, as applied to Grevillea robusta on farm land in Kenya and used here to illustrate certain mensuration constraints, may be considered an extreme variant of pollarding, and not even the most widely used. Three other exploitation techniques which may be appropriately covered by the same denomination are seen more frequently. However, when it comes to mensuration, they all confront us with very similar problems. Basically, and in addition to complete stem-pruning, one may distinguish among top-pollarding (or classic pollarding), branch-pollarding, and selective pollarding.

When applying top-pollarding, the stem and all branches are cut back periodically at a height of about two metres (above goat reach) and then left to regenerate by means of coppicing. In the case of branch-pollarding, the stem and most or all of the principal branches are left more or less intact with the exploitation of these farm trees being confined to the periodic removal of all secondary branches. This method is widely used in parts of India, principally for harvesting tree fodder. In Africa it is seen more rarely in its purest form. Finally, selective pollarding consists in cutting a few branches now and then and here and there in the tree crowns, usually for the double purpose of reducing competition and obtaining produce This method offers the advantage of maintaining the tree crowns almost constantly in a condition of nearly full productive capacity.

In a previous article by Mr Poulsen entitled "The non-wood products of African forests" (Unasylva, Vol. 34, 137:15-21), it should have been mentioned that the article was based upon an FAO project, carried out for, and funded by, the UN Economic Commission for Africa.

When adapted to the conditions on farm and pasture land, these methods have one important feature in common: they conserve the production base. The roots and an essential part of the aboveground structure of the trees are left behind at the time of exploitation in such a way that the continuous renewal of the photosynthesizing structure is ensured even under conditions of heavy livestock pressure.

It is an undeniable fact that the methods described are not always properly applied. For many reasons, many farmers over-exploit tree resources in a destructive way, but this does not reflect on the silvicultural and economic characteristics of the techniques as such. In reality, there can be little doubt that these farmers' practical techniques, largely based on common sense, are very well-suited to many environmental, social and economic conditions or needs. Their more widespread use should be encouraged and initiatives should be taken for improving them.

However, this is not the point here. The purpose of this article is to draw attention to the need for designing appropriate mensuration techniques for obtaining reliable data about the stock and growth rate of a type of tree vegetation that covers vast and rapidly expanding areas throughout the tropics, and in particular in Africa. There should also be more scepticism exercised over mensuration data of farm-land tree cover which have been obtained using conventional mensuration methods.

FAO: Training series 2

Chainsaws in tropical forests

A manual prepared by FAO and ILO, this complete illustrated guide to the use of chainsaws in the tropics was published in 1980. Available in English, French and Spanish.

FAO, Distribution and Sales Section
Via delle Terme di Caracalla
00100 Rome, Italy

Previous Page Top of Page Next Page