<p>Concern about worldwide deforestation and exploitation of the tropical rain forests has led to friction between national governments, wood industries and timber trade on the one hand, and scientists and environmental organizations on the other. One way to safeguard the tropical rain forests is to avoid human interference and to use forests only as nature reserves and as buffer zones of environmental protection. Some vulnerable tropical rain forests and those with unique flora and fauna should, indeed, be treated in this way. Most forests, however, have the potential to produce timber and other products on a sustainable basis provided that they are managed wisely. This study is concerned with sustained timber production of tropical rain forests, particularly with damage control during timber harvesting.<p>The harvesting study is part of multidisciplinary research to develop a forest management system for sustained timber production in Suriname. The research started in the 1960s on an experimental scale in the northern Forestry Belt of Suriname. At first, a monocyclic silvicultural system was chosen with the aim of converting selectively logged rain forests by silvicultural measures into stands of valuable commercial timbers in a period of 60 to 80 years (Schulz, 1960; Boerboom, 1965). The long rotation period, intensive weed control and tending of future crop trees made this system economically unattractive.<p>In the 1970s, a polycyclic sivicultural system was tested, envisaging timber harvests in felling cycles of 20 years by treatments of logged-over forest. This system, later known as CELOS Silvicultural System (CSS), was developed after further research (de Graaf, 1986; Poels, 1987; Jonkers, 1987). In 1981 harvesting studies were included to extend the CSS into a forest management system. Field experiments were carried out in order to compare controlled (organized) logging with conventional (unorganized, haphazardous) logging. These experiments covered damage to the remaining stand caused by felling and skidding of trees, the impact of skidding machines on the forest soil, and organizational and administrative aspects of logging.<p>The research was carried out mainly in Mapane region, 100 km south-east of Paramaribo (Figs 2.1 and 3.1). Conventional, commercial logging operations and controlled logging experiments were studied in a forest area of 640 ha during the period 1982-1984. A number of observations were made in an area of 100 ha, 120 km south-east of Paramaribo, in the forest estate Patamacca, of the Bruynzeel Wood Company (BSH), the largest forest concession holder in Suriname.<p>Felling and skidding impacts to the remaining stand were determined by mapping and calculating the affected forest area. Felling damage was assessed in terms of gaps (chablis) in the forest resulting from a felled tree, and also by the number of damaged trees per ha. Controlled felling, which includes directional felling based on a tree location map in order to facilitate skidding, was shown to reduce felling damage significantly. At a felling intensity of 8-10 trees per ha (20 m <sup><font size="-2">3</font></SUP>), approximatly 14% of the forest area was damaged in conventional felling compared with 8% in controlled felling. In controlled skidding, damage was restricted to 5-8% of the remaining forest, whereas in conventional skidding the affected area exceeded 14%. Skidding damage was also substantial (13%) in the organized harvesting operations of BSH, amounting to 7% due to pre-sorting operations of crawlers in the stump area and 6% due to wheeled skidding on trails.<p>Soil damage may result from movement of skidders. As well as rutting and disturbance to the structure, the soil may also be compacted. The type and degree of compaction are determined by soil characteristics such as structure, texture, field moisture content and organic matter content, and by characteristics of the skidders such as gross vehicle weight, steering system, and tyre and track type. Soil compaction was measured on soil samples and with the aid of a penetrometer. The soil moisture was also assessed by determining the saturated conductivity (K factor) and pF value. Subjected to intensive machine traffic, primary skid trails (main trails from stump areas to log landings), were found to be highly compacted. Most secondary branch trails (log-collecting trails in the stump areas) were compacted to some degree, depending on traffic intensity.<p>Trail sections with moisture content near field capacity were maximally compacted after a few round trips of the loaded skidder. Moisture content appeared to be a decisive factor in the soil degrading process. Soil recovery was a slow process, and skid trails used eight years previously were still found to be maximally compacted.<p>Controlled logging was found to be more efficient than conventional logging. This was examined by measuring logging production in relation to effective crew and machine time, and also by studying the suboperations in a working cycle. Measures to control felling damage were not found to have a negative effect on production. Mean felling productivity was similar for both systems, being 0.07 man-day per m <sup><font size="-2">3</font></SUP>in the conventional and 0.08 man-days per m <sup><font size="-2">3</font></SUP>in controlled system. On an annual basis, skidding production under the controlled system was twice that of conventional logging. This difference can be explained in terms of planning and preparation of harvesting activities undertaken in the controlled system, and also the skidding method based on a pre-established skid trail network. The controlled skidding method was more efficient as 40% less machine time was required per unit of product transported.<p>The research findings provided the basis for the<TT>CELOS</TT>Harvesting System (CHS) aiming at combining logging efficiency and damage prevention. Timber harvesting is based on planning and pre-harvesting preparations, work organization and adapted logging techniques.<p>The cornerstone for planning is forest prospecting, that is a full (100%)<TT></TT>enumeration of harvestable commercial trees. Prospecting data, including terrain characteristics, are mapped and recorded for the use of felling and skidding crews in searching for trees. The maps are also used to design a skid trail network for efficient terrain transport of logs from the stump area to landings along the truck roads. In this way a plan of operations can be made annually for all harvesting work.<p>The skid trail system should be established prior to harvesting in order to facilitate tree felling according to the skidding pattern, thus combining damage prevention and skidding efficiency. The damaging effects of skidding in the stump area can also be reduced by winching logs to the trails. The additional costs of damage prevention are compensated by overall improvement in logging efficiency.<p>The logging organization of a management unit is composed of one prospecting crew (5 men), one felling crew (3 men), one crawler tractor crew (2 men) and one wheeled skidder crew (2 men). This unit is supervised by an assistant manager (ranger). All field staff need to be trained in basic techniques such as tree spotting, scaling, and power saw operation to enable job rotation and thus greater flexibility in operation. Training of machine operators especially should focus on damage prevention and logging efficiency. Such an organizational unit should be able to harvest annually an area of 1000 ha.<p>The CELOS Harvesting System and the CELOS Silvicultural System are the subsystems of the CELOS Management System (CMS) which has been designed for a forest unit of approximately 25 000 ha. Central to this polycyclic system are the management objectives and planning for the subsystems. A major aim is restriction of logging impacts to the soil and the remaining stand. Felling intensity is restricted, for instance for Suriname to 30 m <sup><font size="-2">3</font></SUP>/ha, in order to maintain the ecological, conservational, and protective functions of the forest. The present study gives guidelines for the layout of a standard management unit with a production target of at least 25 000 to 30 000 m <sup><font size="-2">3</font></SUP>per year and also indicates how harvesting and silvicultural operations can be integrated. The CELOS Management System is applicable in the northern Forestry Belt of Suriname and probably in similar forest types of the Amazon basin. Timber harvesting according to this system is applicable or adaptable to most lowland tropical rain forests.