by M. S. PARRY
Forest Department, Tanganyika
FAO issuing a series of concise publications as parts of a World Forest Planting Manual. Some of these have already appeared and others are in course of preparation. The writer was commissioned to compile a short manual, the purpose of which would be to give an outline of the methods of tree planting commonly employed throughout the tropical zone of Africa, and also to summarize briefly the characteristics of the species most commonly planted.
This manual is now finished and will shortly be published. Information was obtained from published literature, unofficial notes and circulars, and also from private communications from a number of forest officers serving in the countries of both West and East Africa.1 In this article, the writer recapitulates the features of the environment with which he is dealing.
1The author makes special acknowledgement to A. Aubreville, Inspecteur général des eaux et forêts, Ministère de la France d'outre-mer. C. J. Taylor (Gold Coast); A. L. Griffith, East African Agriculture and Forestry Research Organization; H. C. Dawkins (Uganda); H. H. Pudden (Kenya): and to the Government of Tanganyika.
TROPICAL Africa is taken to include the whole geographical zone lying between the tropics of Cancer and Capricorn, and includes a large and rather heterogeneous assemblage of countries, embracing an immense diversity of planting conditions, from alpine grasslands to coastal mangroves, and from the wettest of rain forest to semi-desert. Until quite recently, there has been comparatively little interchange of ideas and information between the many separate countries within this great area. Consequently, it is not unusual to find that the techniques used even for the commonest species often vary considerably from place to place. There are several well-known nursery and planting techniques which have been adopted with local adaptations in most parts of tropical Africa, but there is surprisingly little uniformity about the conditions and species to which they are applied.
A certain degree of stability has been attained in the techniques of softwood planting but, with few exceptions, most of the indigenous hardwoods are planted by methods which vary not only from place to place, but often from year to year in one nursery - a sure sign that the foresters concerned are still feeling their way from one partial failure to the next. In most countries following a period of postwar expansion, there has been a phase of increased silvicultural research, the results of which will no doubt form the basis of standardized techniques in the future. It is inevitable, therefore, that any attempt to compile a handbook of planting methods for tropical Africa at the present time is bound to be considered somewhat premature but it is felt that a review of the methods now employed will serve a useful purpose. Research is a never-ending process and it is unlikely that silvicultural techniques in the tropics will ever become rigidly codified. This article, therefore, attempts to describe the available techniques in rather general terms with examples of their application. Planting methods must always be adapted to particular conditions on the basis of local experiment.
Any attempt to subdivide a large region into distinct climatic zones is bound to be artificial, particularly if the subdivision is done with a view to the selection of species suitable for each zone. Not only does climate grade imperceptibly from one place to another, but its effect on tree growth is modified by other conditions, particularly by soil fertility and by the amount of subsoil moisture available. Moreover trees vary a great deal in their adaptability, some being restricted within narrow limits and others growing successfully over a wide range of conditions. None the less, climate is the main factor determining the suitability of an area for tree growth, and therefore the most convenient basis for deciding which species will be suitable for a particular locality.
The climatic regimes of tropical Africa are governed primarily by the bi-annual swing of the sun and its attendant wind systems across the equator. In very general terms, the wet seasons follow the sun (lagging about a month behind), with the result that throughout most of the equatorial zone there are two wet seasons each year. At the equator itself (or, more accurately, at the "rainfall equator" which lies about 3 degrees north) the peaks of the two wet seasons occur about a month after each equinox. As one goes north or south from the equator, the two wet seasons move closer together until they coalesce into one wet season. In the northern hemisphere the rains coalesce in the northern summer (July) and in the south they coalesce in the southern summer (February). Owing to the northward shift of the "rainfall equator" the single wet season does not occur until about 8 degrees north, whereas in the south it is difficult to detect any marked double season more than 3 or 4 degrees from the equator, though a slight break in the rains may occur in February. The heaviest rains always occur with the northward movement of the sun.
These general influences are of course much modified by local topography and by distance from the sea or lakes. In general, local conditions have a great effect on the total amount of the rainfall, but very little effect on the march of the seasons, though there are a few exceptions to this.
In those parts of tropical Africa which have a cool season, such as the highlands of East Africa, the coldest part of the year is usually the dry season, and the climatic regime is of the kind known as the "summer rainfall" type, as distinct from the "Mediterranean" type which is characterized by a hot dry-season and a cool wet-season. The distinction affects the growth of certain species and is one of the reasons for the success of Caribbean species in the highlands of East Africa and in the summer rainfall areas of South Africa.
The main factors determining a climatic type are rainfall and temperature. As temperature is determined largely by altitude, it is common practice to describe a particular set of conditions by the basic criteria of:
(a) the average annual rainfall;
(b) the altitude.
But these two alone can be a dangerous over-simplification if used as the basis for predicting the suitability of an area for tree growth.
In tropical Africa, the distribution of rainfall is affected largely by whether it falls mainly in one or two wet seasons. The single short wet season followed by a long drought of six months or more imposes conditions which are more severe than those with the same total precipitation divided between two seasons. The single wet season is often preferable from the point of view of securing establishment of plantations, owing to the fact that planting conditions are often better and more reliable, but unless the total rainfall is high enough to eliminate any danger of drought the double season will be more satisfactory for sustained growth. The minimum rainfall for a given species may be 10 inches (250 mm.) higher in a single season area than with a double season.
The effectiveness of rainfall is affected also by humidity. As a rule rainfall and humidity are closely correlated, but in some places a high humidity may occur with a low rainfall, for example in mountains subject to prevalent mists, where the low rainfall gives a false impression that the area is dry. In fact, excellent growth of softwoods may occur with barely 30 inches (760 mm.) annual precipitation, under mist-bolt conditions. Even more important sometimes is the effect on microclimate of the vegetation itself. Many important forests in Africa, particularly at their drier limits, are of a relict nature, and are able to persist mainly by virtue of their own influence on the local environment. If the economic species in them have to be replaced after exploitation by planting, it may be necessary to devise an under-planting technique even though, theoretically, it should be feasible to clear-fell and replant in conjunction with crop cultivation.
Temperature in tropical Africa is influenced mainly by altitude above sea level, but what appears to be equally important is the relative altitude in relation to the surrounding country. Thus the climate at 4,000 feet (1,200 m.) on a mountain rising steeply from a low-lying plain will be much cooler than at the same altitude and with the same rainfall in the center of one of the immense plateaus which are a feature of the African hinterland. The best criterion of temperature is therefore not altitude but mean temperature itself, though within a limited region, altitude is much easier to refer to, because temperature data are frequently unknown. The best temperature data for defining a type are the mean summer maximum (i.e. mean daily maximum for the hottest months of the year) and the mean winter minimum. Absolute maxima and minima are of less interest except that in areas subject to frost it is desirable to know the extreme minimum likely to be experienced.
Sub-divisions into silvicultural zones
With the above considerations in mind the innumerable different climatic types of tropical Africa are here grouped into seven broad categories. It is emphasized that these are not distinct climatological units, which would be far too numerous for the present purpose, but are merely a convenient way of subdividing an immensely varied region into a manageable number of zones. Most species will grow throughout the whole of one zone, and many of them will grow in several. The zones selected for this purpose are as follows:
Zone I. - Wet tropical lowland
Such very high rainfall areas, with 70 inches (1,800 mm.) or more each year, have a high uniform temperature with little seasonal variation. The altitude is from sea level to about 1,500 feet (450 m.). Mean maximum temperature is about 85° to 90°F (29° to 32°C) in the hottest month, and mean minimum about 70°F (21°C) in the coolest. It includes the wetter rain-forest areas of West and Central Africa and limited foothill areas in the coastal plain of East Africa. Near the equator there is no marked dry season and there are two rainfall peaks, but at higher latitudes there is a well-defined single wet season with a marked dry period, but humidity is always high.
Zone II. - Moist tropical lowland
Moderately high rainfall between about 40 inches (1,000 mm.) and 70 inches (1,800 mm.) a year, combines with a high uniform temperature. Altitude from sea level is up to about 4,000 feet (1,200 m.) under plateau conditions, or only up to about 2,500 feet (750 m.) on steeply rising land. Temperature variations have a slightly wider range than in Zone I, the mean summer maximum being 85° to 90°F (29° to 32°C), and mean "winter" minimum about 60° to 65°F (15° to 18°C). Three rainfall regimes occur within the zone viz: equatorial (two peaks), north tropical (July peak), and south tropical (February peak). This type of climate occurs over an immense and diverse area and would be capable of almost limitless subdivision. It is taken here to include the drier high-forest areas of West Africa, much of the Congo basin, and the wetter parts of the East African coastal plain. It is also taken to include the rather distinct and important forests of medium altitude such as those of Uganda, as many of the important species in them grow equally well down to sea level, whereas few of them are found much above 4,000 feet (1,200 m.)
Zone III - Wet montane
Moderate to high rainfall, from 40 to 70 inches (1,000, to 1,800 mm.) a year, occasionally exceeds 100 inches (2,550 mm.) in very localized areas. If the rainfall occurs in only one season, more than 50 inches (1,250 mm.) would be necessary to give conditions equivalent to 40 inches (1,000 mm.) in two seasons. Temperatures arc cool, but rarely down to freezing point, the mean summer maximum being from 70° to 80°F (21° to 27°C), and the winter minimum about 50° to 60°F (10° to 15°C). The altitudinal range of this zone is mainly between 4,000 feet (1,200 m.) and 7,000 feet (2,100 m.), but certain freak localities rising abruptly from a plain, have a semimontane climate at altitudes as low as 2,500 feet (750 m.) though always with a high, well-distributed rainfall and much cloud. This climatic zone occurs mainly on the east and south-facing slopes of the mountain ranges of East Africa.
Zone I V. - Dry montane
Rainfall is low to moderate at high elevations; annual precipitation from 25 inches (600 mm.) to about 40 inches (1,000 mm.) in two seasons, or 30 to 50 inches (750 to 1,250 mm.) in a single season. Altitude is mostly between 5,000 feet and 9,000 feet (1,500 to 2,750 m.). At the lower altitudes a rainfall less than 30 inches (750 mm.) in two seasons or 40 inches (1,000 mm.) in one season produces conditions merging into those of Zone V.
Temperatures are cool throughout the year, becoming very cold in the dry season; frosts are not uncommon above 6,000 to 7,000 feet (1,800 to 2,100 m.), with considerable diurnal variations, e.g., from 35° to 75°F (2° to 24°C). The zone occurs mainly on the higher East African "Highlands", and on the dry sides of isolated mountains.
Zone V. - Dry plateau
Rainfall is low to moderate, mainly from 25 to 40 inches (600 to 1,000 mm.) a year, or up to 45 inches (230 mm.) with a single season. Altitude is mainly from 3,000 feet to 6,000 feet (900 to 1,500 m.). Temperatures are fairly high just before the wet seasons but become cool during the dry season, the mean summer maximum often approaching 90°F (32°C), while the dry-season mean minimum is usually well below 60°F (15°C). Frost may occur at the higher levels outside the equatorial belt. This type of climate occurs over a very wide area on the broad plateau which covers most of the East African hinterland from Uganda to the Rhodesias. Very similar conditions but with higher average temperatures occur in Zone VI. All three types of rainfall regime occur within this area, ranging from the southern single February peak, through the equatorial double peak, to the northern August peak.
Zone VI. - Dry lowland
Rainfall is low to moderate, mainly from 25 to 40 inches (600 to 1,000 mm.) a year, with altitude below 3,000 feet (900 m.). Temperatures are very high during hot seasons with considerable diurnal variation. The mean maximum temperature of the hottest month is usually more than 95°F (35°C), and the mean minimum of the coolest month about 60° to 70°F (15° to DISC). This type of climate occurs over an immense area stretching in a belt 200 to 300 miles (320 to 430 km.) wide right across the northern parts of the Guinea Coast territories from Senegal to the Sudan. In this area the rainfall regime is mainly of the north tropical type with a single peak in July-August, but a double season with a short gap in July-August may occur at lower latitudes. Similar conditions, but with the southern regime, occur in Angola and on the coast of East Africa. The Guinea coast from Gold Coast to Dahomey also has these conditions with the equatorial regime.
Zone VII. - Semi-desert
This is taken to include all areas with less than 25 inches (600 mm.) rainfall a year, regardless of altitude. Such conditions occur in a belt across Africa parallel to that of Zone VI but north of it, at about latitude ]12° to 15° N. Semi-desert also occurs in southwest Angola and in much of northern Kenya and Somaliland. There are also isolated pockets in central Tanganyika, and elsewhere in the rain shadows of the main mountains.
It would be quite impossible to attempt any sort of classification of the innumerable soil types of the tropical zone of Africa within an article of this nature. For such information reference should be made to published literature on the subject.
The following notes indicate the main soil characteristics which affect tree growth and suggest the points which should be examined when assessing a soil for the purposes of afforestation.
Steep topography should normally be no barrier to afforestation, unless planting has to be done with squatters. Even with squatters, slopes of up to 30 degrees can usually be planted quite safely as the land remains cleared for only three or four years, which is too short a period for the high permeability of the natural forest soil to be lost. Paradoxically, a loose friable soil is less liable to erode than a hard compacted soil. If planting is done directly into grassland or scrub, the limit of steepness is determined largely by the cost of road construction, which in turn depends mainly on the depth of soil and decomposed rock. On moderate slopes road construction may cost as little as £30 per mile ($62 per km.), but in steep country the cost will probably exceed 10 times this amount if there is undecomposed rock near the surface. Afforestation on land too steep for road building may sometimes be necessary for protective reasons or to "round off" a given acreage without leaving blanks, but as a general rule plantations without communications are to be avoided.
The important feature is rooting depth, which is not always the same as soil depth. In apparently shallow rocky soils, moisture often percolates into deep fissures where the roots can penetrate also, giving good tree growth on a soil unsuitable for agriculture. If this can be observed, for example on a road cutting, or deduced from the growth of existing trees, the presence of rock near the surface need be no deterrent to afforestation. On the other hand the rooting depth may be quite shallow in soils which appear deep, the usual reason being that water is poorly distributed throughout the profile.
Shallow rooting may occur because some soils have a seasonally high water table, indicated by gleying of the subsoil, while others are simply impermeable and, in dry country, rarely get wet in more than the top few feet so that trees have to live from one storm to the next solely on surface moisture. The surface run-off over impermeable soils is usually high and a large proportion of the rainfall may be lost unless the land is ridged on the contour. Trees nearly always prefer light permeable soils to compacted silts or heavy clays, though in coarse sands, if the profile is very deep, excess water from heavy stores may percolate rapidly down to a water table out of reach of tree roots. In this type of soil, trees which strike a taproot downward with vigor when in the seedling stage, like Eucalyptus camaldulensis and Casuarina are at an advantage over those like Cassia siemea which tend to root on the surface, whereas in a very impermeable soil the opposite might be the case. The maximum depth attained by the roots of certain dry country species is often over 50 feet (15 m.) and may exceed 100 feet (30 m.) but in many of the drier parts of Africa there is virtually no water table within 200 feet (60 m.) or more of the surface. All species are therefore to some extent likely to be dependent on surface water.
The effect of soil fertility is felt mainly during the early years of the crop. With long rotations and fairly high rainfall the quality of the site probably depends more on the nature of the parent rock than on the fertility of the soil at the time of planting. This is certainly the case where worn-out fields of local cultivators are being planted. Growth at first will be better on the more recently fallowed fields than on those which have been temporarily exhausted, but one would expect these differences to become evened-up during the course of the rotation.
This is not meant to imply that fertility can be overlooked; in fact, on certain soils the success or otherwise of establishment may depend very largely on making use of the initial fertility of a soil before it can become exhausted. Deep, permeable sandy soils in high rainfall areas carrying degrazed forest or scrub are particularly liable to lose fertility when they are cleared for re-planting. On such soils, if squatters are used to clear the land, it may be essential to plant trees on the first break to take advantage of the temporary increase in fertility following a burn. If this is not done, the soil may become virtually unplantable (with exacting species) until it has undergone a slow recovery under a long bush fallow.
In certain parts of Africa the basic fertility of the parent material itself is very low, and quality of growth is limited by this factor alone. Much of the land under dry deciduous woodland of the type known as miombo tends to be inherently sterile, particularly if the profile is free-draining and the rainfall high enough to exceed evaporation. In northern Rhodesia, for example, tree growth is limited by the general sterility of the soil in many places where the climate should theoretically support closed high forest. When assessing the value of a site for afforestation, therefore, it is necessary to distinguish between a soil which is temporarily exhausted, and one which is inherently of low fertility.
For practical purposes the initial fertility of a soil can be judged most easily from the quality of the crops or vegetation growing on it. In most places, impoverished land can be detected from species which may be well-known to the indigenous population as indicators of poor soil. It is always useful and interesting to have definite analyses made of the soils, but it is difficult to know how to interpret the results. A soil chemist's report is more likely to indicate the suitability of the soil for certain agricultural crops than for a long-term forestry crop. The quality of tree plantations may depend ultimately on the availability of nutrients from still decomposing rock, combined with soil depth, aeration and moisture retention, rather than on the initial nutrient status at the time of planting. Analyses are of value mainly where there are indications of a definite nutrient deficiency. A few important species (e.g. Chlorophora) often exhibit symptoms which may be due to nutritional disorders and any information which may lead to the diagnosis of this condition is always worth obtaining.
A test of the acidity of a soil is always worth making, as many species are sensitive in this respect and will grow only within a limited range. Most pines, for example, (except P. halepensis) require acid soils with acidity in the range pH. 4 to pH. 6 for optimum growth. Most soils in tropical Africa tend to be slightly acid unless derived directly from calcareous rock.
Indications of the presence of sub-soil moisture should always be looked for, as this is frequently one of the main factors affecting the growth of trees. Few species will stand waterlogging for more than a few days, and soils liable to temporary inundations or a seasonally high water table must be classed as special sites. They may be very suitable for afforestation but species specially adapted to these conditions must be selected (e.g. E. camaldulensis Lagerstroemia, Populus, etc.). In dry country the presence of a water table at depth but within reach of tree roots can completely alter the quality of a site, but its presence is difficult to deduce. The occasional large tree in a very low rainfall district probably indicates the presence of water, but this may be a local seepage and may not mean that uniformly good growth will occur over the whole area. Sandy river courses usually have water close to the surface for most of the year and their banks are ideal for afforestation, except that the alluvial deposits on either side are nearly always valuable agricultural soils.