H.F. Maitre
H.F. Maitre is Chief, Inventories and Management Division, Centre technique forestier tropical (Department of Cirad), France. This article is adapted from a paper presented at the 1986 World Congress of the International Union of Forestry Research Organizations in Ljubljana, Yugoslavia.
The reaction of the tropical rain forest to different silvicultural operations logging and thinning was studied in three forest reserves in Côte d'Ivoire. Measurements taken over a span of four years made it possible to follow the diameter growth of more than 50 commercially valuable species, providing information hitherto unavailable about the requirements and behaviour of these species. Taking into account the growth, mortality and regeneration of an entire stand, the results clearly indicated the value of silvicultural operations, which frequently resulted in a doubling of production, and also fostered the establishment of the most valuable species.
The species studied (known as "principal") were those destined for timber production, actively marketed or little known but technologically valuable.
· Over the past two decades, the future of tropical forests has become a matter of worldwide concern. Repeated warnings have been issued regarding the social, economic and environmental consequences of unchecked tropical forest destruction.
This shrinking of areas covered by tropical forest, especially the dense rain forest of West Africa, is usually the direct consequence of the rush for agricultural land, where forests are seen as reservoirs of soil that can quickly be cleared for agriculture, without regard to their fragility or to the forest-based values involved.
As Schmidt has recently noted (Schmidt, 1987), managing tropical forests for economic production is a key element in their conservation. This fact presents the forest manager and researcher with a two-pronged question: "What are the possibilities of reconstituting the potential of stands creamed by the first loggers? And what simple and inexpensive measures can be taken to bring about the full development of valuable species and thus ensure adequate timber production?"
These questions have particular relevance in Côte d'Ivoire, where there has been considerable deforestation in the past 30 years. In 1983, for example, the estimated total area under forest was 3.5 million ha, compared with an estimated 14.5 million ha 40 years earlier. In the decade from 1973 to 1982, approximately 3 million ha of forest were cleared, which represents a destruction rate of 300000 ha a year.
Faced with this situation, the Government of Côte d'Ivoire established a "permanent forest reserve" covering 2500000 ha of forests classified as relatively untouchable. This permanent forest reserve will receive from the Ivorian Forest Service the attention needed not only to protect it against clearing but also to ensure sustained and improved production through continuous and realistic management based on simple rules.
In pursuit of these objectives, three sets of silvicultural trials were started in 1976 by SO.DE.FOR (Société pour le développement des plantations forestières) with technical assistance from CTFT (Centre technique forestier tropical).
This article describes the principles and the first results of this research in dense natural forest.
Early research by various organizations in the management of natural forest was conducted without any coordination and without sufficient resources to achieve its goals.
Many trial plots (almost always too small) were studied in tropical forest regions, and there is plenty of literature quoting a host of incomplete and often uninterpreted data. This research, however, led to very few concrete ideas or forest management policies.
In these circumstances, some hard thinking in the early 1970s produced the following three basic principles for planning the natural forest research:
· utilize only large plots (several hectares) with the maximum number of repetitions in space;· measure simple parameters (diameter and localization of trees);
· provide facilities for data storage and interpretation (computers).
In 1976, on the basis of the above recommendations with regard to forestry research, a system was planned and put into practice by SO.DE.FOR in Côte d'Ivoire.
Definition of me study's objectives was based mainly on the above principles but also on previous successes and failures more or less everywhere in Africa (Ghana, Nigeria, Gabon, Côte d'Ivoire) in improving natural stands, mostly between 1945 and 1965. The idea is to standardize me forest, modifying natural regeneration by opening up the cover to varying degrees. The problem here was the proliferation of sun-seeking creepers, and also the high cost of operations that were excessive in number and took too long, making them difficult justify technically and economically.
The SO.DE.FOR study focuses on the higher stand constituted by stems and trees more than 10 cm in diameter (at 1.30 m, or above the buttress) to establish their reaction to simple and inexpensive treatments that could be carried out on a large scale.
Since regeneration at ground level proved difficult to master, no special treatment was envisaged to encourage it. It is inevitably influenced by the intensity and method of silvicultural treatment of the higher storey, and is therefore something to observe rather than induce.
The main lines of the study were as follows:
· to test and develop simple silvicultural techniques of thinning and logging;· to study the behaviour and growth of trees (by species) in response to these silvicultural treatments;
· to establish the trends in stands as a whole (induced mortality, natural recruitment of young stems, effect on creepers and coppice-wood), as a result of these treatments;
· to check the favourable or unfavourable response of young stems and seedlings (regeneration);
· to quantify the effect of the different operations on production, to reflect the most suitable treatments for different terrain and conditions of production and, lastly, to judge the results by comparison with non-treated stands.
The same type of trial was repeated three times, once for each type of forest in Côte d'Ivoire:
· evergreen forests (Irobo);· semi-deciduous forests (La Téné);
· transition forests (Mopri).
Research therefore covered three separate but identical "perimeters", of the same area and the same nature. Each perimeter was 2 km square, or 400 ha, divided into 25 16-ha plots; the total amount of land involved was therefore 1200 ha (see Fig. 1).
FIGURE 1. A trial perimeter (with identical numbering system for all three perimeters). Perimeter area: 400 ha for the trials and 900 ha, including the buffer zone
The species studied (known as "principal") were those destined for timber production, actively marketed or little known but technologically valuable. There are more than 70 such species in Côte d'Ivoire, and about 50 were well represented within the perimeters. Other tree species were also studied and monitored, but without botanical identification and all under the heading of secondary species.
Depending on the abundance of species and previous logging in the three forests, treatments took place as follows for the 75 plots:
· 30 control plots remained untouched (ten in each perimeter);· 35 plots were thinned by elimination of standing secondary species. This was done by girdling, using the rough notching technique, and spraying with arboricide. It was conducted at two intensities by suppression of either 40 percent or 30 percent of the basal area, systematically, and starting with the biggest trunks of secondary species until the desired percentage of elimination had been obtained;
· ten plots (only at La Téné) were commercially exploited by felling, skidding and sale of all principal species with a diameter of 80 cm or more (average extraction: 53 m³/ha).
The treatments coincided with the first set of general measurements marking the start of the experiment.
The pragmatic aspects of silvicultural treatments, namely thinning and logging and the increased production they induce, are now well established.
Although each 16-ha plot was treated over the whole of its area, only the four central hectares were observed and monitored, to avoid the edge effects. All trunks of more than 10 cm diameter were numbered and all trees belonging to the principal species were accurately located within each plot, following the standard system of rectangular coordinates (axes formed by the edges of the plot measured).
The circumferences of all trees (principal and secondary species) at 1.30 m, or above the buttresses, are measured manually every two years: at the beginning of the treatment system (year 0), two years later (year 2), then after four years (year 4) and so on. More than 48000 trees are involved. Also, a subsample of principal species, including 3750 individuals, is monitored and precisely measured every six months by dendrographs (permanently fixed metal diameter tapes).
These measurements therefore made it possible to monitor not only individual and general growth of diameter, basal area and volume (measurement rates having been established in advance) but also:
· the growth of young stems that have reached 10 cm diameter between two measurements; and· the mortality or natural disappearance of trees (for reasons other than logging or thinning) between two measurements.
Analysis and breakdown of stands with trees from 2 to 10 cm in diameter, corresponding to the different stages of regeneration reached before or since the measuring, were made on 100-m² patches systematically marked out within the 4-ha plots (40 patches per plot).
This mass of information has of course made it necessary to store and process data by computer; anything else would be unworkable.
The results set out below are from the third set of measurements (four years after silvicultural treatment) for the higher stand, whose stems are at least 10 cm in diameter, and from sampling by patches seven years after treatment of plantlets and young stems between 2 and 10 cm in diameter.
A few months after the silvicultural treatment (poisoning of secondary species or logging of principal species) the appearance of the stands was striking, because of windfallen wood, dead or broken trees and the gaps created either by extracting 50 m³/ha of merchantable timber, or by the elimination of 30 to 40 percent of the basal area.
Large gaps were unevenly distributed in stands on the plots logged, where clearings from felling and tree extraction were found side by side with absolutely untouched clumps. The impact of logging is haphazard compared with that of thinning by poisoning, which is more uniformly spread over the whole stand.
In fact, commercial logging leaves some trees almost isolated and groups of individuals untouched near the control plots.
Four years after treatment, however, the forest has a more normal appearance: the "candles" (dead trees) have mostly disappeared and the stand consists of sound trees without creepers or drooping crowns. The upper storey simply become more open, with an abundance of valuable species, particularly after thinning.
The treatment combining logging and poisoning of standing trees is shown in Figure 2.
FIGURE 2. Treatment combining commercial logging and thinning (Crowns of valuable species are shown in black)
Many small-diameter stems, brought into the light by this partial opening of the upper storey, spring up quickly from the undergrowth; in the gaps there is often considerable regeneration.
These contrasts are easy to see, since control and treated plots are side by side within the trial perimeters.
Early fears of irreversible destruction of a dense stand therefore quickly give way to the certainty of having added impetus to forest growth.
The responses of eight well-known species to treatment are summarized in the Table. Results obtained after four years of study are shown in the form of annual averages, both for diameter growth and for percentage of recruitment or mortality, depending on the type of treatment (the two intensities of thinning being grouped together) and for two categories of trees by size.
This table clearly shows what has been generally observed for most valuable species.
|
|
Plots |
Number of trees measured |
Average annual diameter growth (in cm) of stems between: |
Percentage of stems having reached 10 cm (recruitment) (%) |
Percentage of mortality of stems with diameters |
||
|
10 cm and 25 cm |
25 cm and 65 cm |
10 to 25 cm (%) |
Over 25 cm (%) |
||||
|
Khaya anthotheca |
Control |
198 |
0.20 |
0.61 |
+ 6.0 |
- 4.1 |
- 2.3 |
|
Thinning |
308 |
0.37 |
0.97 |
+ 10.4 |
- 5.4 |
- 1.4 |
|
|
Logging |
104 |
0.22 |
- |
+ 2.3 |
- 2.8 |
- |
|
|
Gambeya delevoyi |
Control |
741 |
0.29 |
0.32 |
+ 3.5 |
- 4.4 |
- 12.2 |
|
Thinning |
637 |
0.59 |
0.64 |
+ 10.4 |
- 5.1 |
- 8.3 |
|
|
Logging |
253 |
0.21 |
0.38 |
+ 2.4 |
- 4.8 |
- 11.2 |
|
|
Scotellia sp. pl. |
Control |
812 |
0.12 |
0.28 |
+ 3.1 |
- 4.1 |
- 4.4 |
|
Thinning |
910 |
0.29 |
0.42 |
+ 3.2 |
- 4.1 |
- 6.0 |
|
|
Logging |
240 |
0.16 |
0.42 |
+ 3.5 |
- 4.5 |
- 3.7 |
|
|
Aningeria robusta |
Control |
341 |
0.20 |
0.41 |
+ 4.5 |
- 3.9 |
- 3.1 |
|
Thinning |
371 |
0.39 |
0.58 |
+ 9.9 |
- 4.2 |
- 5.3 |
|
|
Guarea cedrata |
Control |
325 |
0.25 |
0.31 |
+ 7.0 |
- 5.5 |
- 3.1 |
|
Thinning |
566 |
0.47 |
0.53 |
+ 16.5 |
- 8.3 |
- 6.7 |
|
|
Nesogordonia papaverifera |
Control |
1082 |
0.27 |
0.23 |
+ 5.5 |
- 2.0 |
- 1.8 |
|
Thinning |
790 |
0.51 |
0.32 |
+ 12.5 |
- 3.2 |
- 2.2 |
|
|
Logging |
644 |
0.25 |
0.19 |
+ 2.6 |
- 3.7 |
- 1.9 |
|
|
Tarrietia utilis |
Control |
1280 |
0.28 |
0.57 |
+ 2.9 |
- 2.8 |
- 2.6 |
|
Thinning |
1942 |
0.58 |
0.91 |
+ 6.0 |
- 3.9 |
- 3.3 |
|
|
Triplochiton scleroxylon |
Control |
640 |
0.60 |
0.79 |
+ 5.7 |
- 2.8 |
- 1.4 |
|
Thinning |
359 |
1.53 |
1.48 |
+ 15.6 |
- 2.5 |
- 1.4 |
|
|
Logging |
511 |
0.87 |
1.13 |
+ 7.2 |
- 5.7 |
- 2.8 |
|
Treatment by thinning (elimination of secondary species) provokes a very favourable reaction on the part of the principal species. It leads to a strong increase in diameter growth (50 to 100 percent stronger), especially for medium-sized, stems. It also considerably accelerates the appearance of young stems of valuable species in the category over 10 cm in diameter, demonstrating that the dynamics of these small valuable stems are greatly encouraged by the partial and uniform opening of the forest canopy, which acts as a "suction pump" for natural regeneration.
Some hard thinking in the early 1970s produced three basic principles for planning natural forest research.
Under the treatment by straight commercial logging, gains in growth and recruitment are systematically below those with thinning, but often considerably more than those of the untouched control plots. Treatment by logging involves taking out a large amount of basal area, of the same order of magnitude as thinning, but it does not specifically promote growth of principal species as against secondary ones. This explains its relatively modest impact on the growth of valuable species.
As regards the phenomenon of mortality (down timber and natural die-back of principal species), this is not easy to interpret: no clear and direct link between the opening of the stand and increased mortality has been established. The main point to remember is the scale of volume losses that could be caused by natural mortality: for one hectare of forest, the quick disappearance of one or two stems of a normal size represents a decrease of standing volume often more important than the annual increment in volume resulting from the growth of the whole stand. The balance of production could be nil or even negative. Finally, it is recognized that the result of silvicultural treatments should be continued over a period of more than ten years, since precise measurements taken very regularly (half-yearly) by permanent dendrographs have shown that the accelerated diameter-growth caused by these treatments become progressively more pronounced. The rates of growth become higher and higher each time the measurement takes place during the four years of observation.
The study of different stages of regeneration started late and is still too recent to be conclusive. However, some important ideas have already appeared:
· compared with the untouched plots, silvicultural treatment hardly changes the floristic composition of the undergrowth of seedlings, saplings and small stems with diameters between 2 and 10 cm at 1.30 m. This undergrowth, which corresponds to the successive stages of regeneration, was also lightly helped in its activity, namely by the frequent appearance of young individuals. Neither creepers nor coppice-wood were able to hinder in any way the "normal" process of regeneration;· the representation of species with any future at this "lower" level of the population proved to be very modest, and only. 16 percent of the trees counted belonged, in an evergreen forest (Irobo), to species which could likely reach or exceed a size of 40 cm diameter at 1.30 m.
An estimate of the volume expansion of standing trees by species or by group of mixed species (whether principal or secondary) showed that forest stands left to themselves without silvicultural treatment tend to maintain a fairly constant volume of standing timber, since natural mortality largely compensates for the growth of the whole stand, and overall production is modest in untouched plots.
More precisely, each species presents its own behaviour and well-defined dynamics, and some species show a very strong tendency to decline or die back while standing, despite silvicultural treatment. This phenomenon is all the more serious since it can involve big trees and the consequent losses in standing volume can cancel out the overall production of the surrounding stand. Controlled commercial logging, taking into account the behaviour of each species, is necessary to reduce timber losses.
Early fears of irreversible destruction of a dense stand quickly give way to the certainty of having added impetus to forest growth.
It was also found that simple thinning is favourable to the productivity of the rest of the stand, and therefore of timber species, which is the intended purpose.
Treatment by ordinary logging also means increased production in volume of trees compared with those of untouched plots. This increase is, however, lower than that caused by thinning (despite extraction on the same scale), because of the very uneven impact of logging.
Logging leaves large gaps badly distributed over the terrain, and only well-positioned trees that have not been injured by the felling of others can benefit from the operation. The random nature of the treatment does not lead to more vigorous activity by the young saplings from recent regenerations.
Opening the forest cover has most effect on small- and medium-sized stems, which compete fiercely for light.
The production of secondary species was negligible, both in the control plots and in the treated plots. It was, however, stable and relatively large for the principal species, whose annual productivity rate was from 0.5 to 2 percent of standing volume in untouched plots, 1.5 percent in logged plots, and between 2 and 3.5 percent in thinned plots. For merchantable species alone, increases in annual volume were as follows (from 10 cm diameter):
· 0.7-1.8 m³/ha/year for untouched stands
· about 2.5 m³/ha/year for logged stands
· 2.2 to 3.6 m³/ha/year for thinned stands.
These amounts, showing a doubling of production, correspond approximately to a standing timber production of about 270 m³/ha, with valuable principal species accounting for 100-150 m³ of this.
The production of natural stands improved by silvicultural treatment compares very favourably with that of artificial stands of timber species in Côte d'Ivoire:
· 4-5 m³/ha/year for teak (Tectona grandis);
· 7-8 m³/ha/year for framiré (Terminalia ivorensis).
Paradoxically, the heterogeneity of dense forests can also constitute an asset, since the multiplicity of timber species provides a wide range of marketable products.
The above constitutes a first assessment, providing for the natural forest data hitherto little known or deduced from observations on too limited a scale.
The figures obtained during this first phase of the study have made it possible to compare the performance of stands as a result of simple silvicultural treatments, whose beneficial effect on the growth of valuable species has proved very encouraging.
These data, however precise they may be, nevertheless remain provisional, and only after a lapse of time and after new measurements and observations will they acquire the degree of reliability indispensable for a thorough knowledge of dense forest stands.
Most of the research objectives set during the planning of the trials have been achieved, particularly the main objective, which was to quantify the production of tree stands for different valuable species or for the whole stand.
The pragmatic aspects of silvicultural treatments, namely thinning and logging and the increased production they induce, are now well established. This has enabled SO.DE.FOR to start a practical programme, called Pilot Forestry Management, in Yapo, covering an area of 10000 ha of evergreen forest, where there are "classic" operations combining the two types of treatment (tested and recommended by research) illustrated by the design below.
It remains to establish accurately the duration of the effect of silvicultural treatments and to decide on the right time for further treatments, and in addition to gain a better idea of the direct effect of treatments on regeneration. The estimated cost of development operations, their economic profitability and the possibilities of applying them on a large scale are fields in which the Yapo Forestry Service may be able to contribute fresh information.
This research has also been rich in non-technical lessons which are nevertheless crucial to the success of future action to preserve the forest ecosystem.
The Ivorian authorities, on the advice of their foresters, decided in the mid-1970s to mark out a permanent forest reserve and manage it with a view to the maintenance of wood production and also effective protection, on the principle that the only forests respected are those that are "worked".
This government policy took concrete form in the allocation of adequate resources from national finances in order to conduct a wide-ranging study on the reaction of dense forest to various pragmatically envisageable silvicultural operations.
This experiment, meticulously conducted by SO.DE.FOR with technical assistance from CTFT, has already provided promising results: simple and inexpensive silvicultural treatments in natural forests maintain or even accelerate valuable timber production, with a higher financial profit than that obtained from plantations.
On the basis of these first results, a pilot management programme for the Yapo forest has been undertaken. Other operations are foreseen, covering increasingly extensive areas. The determination of Côte d'Ivoire, together with the first results of research, have impressed international donors, which are already helping to finance these actions.
Generally speaking, experience has shown that the success of dense forest management depends, most of all, on commitment at national level, implemented by a stable institution with the resources and the necessary trained and motivated forest workers. Finally come the technical aspects to be solved, and on this level progress must be gradual: very thorough practical research, demonstration pilot operations, and then action on a larger scale.
BERTAULT, J.G. 1986 Etude de l'effet d'interventions sylvicoles sur la régénération naturelle au sein d'un périmètre expérimental d'aménagement en forêt dense humide de Côte d'Ivoire (Study of the effect of silvicultural treatments on natural regeneration in an experimental management perimeter in dense rain forests in Côte d'Ivoire). Univ. Nancy. CTFT/SO.DE.FOR.
MAITRE, H.G. & HERMELINE, M. 1985 Dispositifs d'étude de l'évolution de la forêt dense ivoirienne suivant différentes modalités d'intervention sylvicole (Study of the development of Ivorian dense forests, using different methods of silvicultural treatment). CTFT/SO. DE. FOR .
SCHMIDT, R. 1987 Tropical rain forest management - a status report. Unasylva, 156(39): 2-17.
