A total of 266 studies and articles on reduced impact logging or conventional logging conducted in closed broad-leaved tropical forests and published since 1950 were reviewed and classified in this project. The majority of these studies were published in the last decade, but were not presented according to a standardized system. In the 1970s, the number of publications on conventional logging (CL) increased rapidly whereas RIL was covered in only few publications prior to 1980.
Minimizing the damage to the residual trees and advance regeneration during logging is essential for the success of all polycyclic silvicultural systems. In practice though, very little consideration is given to this. Referring to Appendix B it is quite apparent that damage to the residual stand in conventional logging operations is excessive and the percent of residual trees damaged ranges from 33-70% in areas with higher (e.g., >30 m3/ha) logging intensity (e.g., Nicholson 1958b, Fox 1968, Burgess 1971, Tinal and Palenewen 1978, Abdulhad et al. 1981, Masson 1983, Yeom 1984, Korsgaard 1985, Ayres and Johns 1987, Uhl and Viera 1989, Pinard et al. 1995, Dykstra et al. 1996, Elias 1996, Greiser-Johns 1996, Berthalt and Sist 1997) . In areas with lower logging intensity (e.g., in Africa with removal of 1-2 trees/ha) residual stand damage generally ranges from 10-20% (e.g., Ola-Adams 1987, White 1994, Cordero and Howard 1996, Scharpenberg 1997). However, tree damage does not increase in direct proportion to felling intensity (Verissimo et al. 1987). Levels of damage typical of conventional logging operations are unacceptably high.
Implementation of RIL has resulted in the reduction of residual tree damage from 50-200% (Mattsson-Marn and Jonkers 1981, Bote 1983, Reyes 1983, Malvas 1987b, Buenaflor 1989, Hendrison 1989, Johns et al. 1996, Berthault and Sist 1997, Moura-Costa 1997). With the implementation of RIL techniques, logging intensity can be significantly increased and still result in less damage to residuals. For example, Buenaflor (1989) found 67% of residuals damaged in uncontrolled logging with 23 m3/ha removed, while in a controlled logging area 22% of residuals were damaged with 32 m3/ha removed.
Some damage will always occur with the felling of trees and it can be expected that with careful felling approximately 200 m2 of forest area will be damaged with each felled tree (Weidelt 1996). Therefore, there is a maximum logging intensity threshold beyond which maintaining stand integrity is difficult in selection cutting. For example, Watanabe (1992) gives this threshold as 30% of stand basal area. Skid trails are also required, but skidding damage can be minimized by planning the trails, utilizing the optimum trail spacing, keeping the trails straight, directional felling of trees on an angle towards trails, keeping the skidders on the trails, utilizing the winch more, limiting skidding operations during wet periods, using the correct size of skidder (i.e., not too large and not too small), skidding log lengths, and utilizing bumper trees where required. The skill and work ethic of both fellers and skidder operators are also critical in minimizing damage.
As with residual stand damage, site impacts in conventional logging of NCTF are excessive. In high logging intensity and uncontrolled logging areas, 30-75% of the area can be serious impacted with roads, tractor trails, landings or just otherwise bulldozed (e.g., when gathering logs with the blade) (Chai 1975, Kartawinata 1978, Neil 1984, DeBonis 1986, Buenaflor 1989, Bruenig 1996, Dykstra et al. 1996). However, typically in higher logging intensity areas (e.g., 30-50 m3/ha), 10-25% of the area is impacted by roads, skid trails and landings (Nicholson 1958a, Borhan et al. 1987, FAO 1989b, Hendrison 1989, Uhl and Viera 1989, Malmer and Grip 1990, Sim and Nykvist 1991, Verissimo et al. 1992, Cannon et al. 1994, Ohn et al. 1996, Winkler 1997). In lower logging intensity areas the soil disturbance is from 6-13% of the area (Bullock 1980, Uhl et al. 1991, White 1994, Agyeman et al. 1995, Scharpenberg 1997). Bruenig (1996) states that with excessive roading and skidding, and thus excessive compaction and erosion, felling cycles of 25-50 years are not sustainable and 60-100 years is more realistic. Tropical soils are also highly susceptible to degradation when physically disturbed, and exposed to the sun and/or the direct impact of heavy tropical rains (Poore and Sayer 1987).
Better wood utilization efficiency in both harvesting and mill processing can greatly enhance the sustainability of the tropical timber industry (Noack 1995). The extent of logging waste reported in the literature generally ranges from 30% (Silitonga 1987, Bhargava and Kugan 1988, Gerwing et al. 1996, Muladi 1996, Scharpenberg 1997) to 50% (Virtucio and Torres 1978, Dykstra 1992, Noack 1995) of the extracted log volume.
Through a review of tropical countries Dykstra (1992) estimated felling recovery rates to be 54% in Africa, 46% in Asia/Pacific, 56% in Latin America and the Caribbean, and 50% on average for all tropical areas. A similar study by Noack (1995) in Ghana, Cameroon East Kalimantan and Sarawak found that on average 53.5% of the tree was extracted; of the remaining volume 4.6% was stump, 5.2% buttress, 10.4% stem off-cuts and 26.3% were parts of the crown with diameter >20 cm. Variations in felling recovery rates reported in the literature are due to operational efficiency and skill of workers, available markets for lower grade logs, and differences in the definition of merchantable wood.
One source of logging waste is felled and bucked trees which are not found during the skidding operation. For example, Mattsson-Marn and Jonkers (1981) found that 11 m3/ha (20% of extracted volume) of logs could not be found by the skidder in current operations. In a planned harvesting block the volume lost was reduced by 100% to 5.5 m3/ha. Gerwing et al. (1996) found that 6.6 m3/ha (22% of extracted volume) of usable timber was felled but never skidded. A similar result was found by Uhl et al. (1997) who reported 7 m3/ha (20% of extracted volume) felled and never recovered. Through RIL techniques, and mapping of felled trees and felling directions the loss of logs can more or less be eliminated.
Logging wastes also develop due to poor work methods, and felling and bucking techniques which result in the splitting and breaking of felled trees (Hendrison 1989). High stumps, felling above the buttress and topping at too large a diameter also result in excessive waste (Balachandra 1988, Gerwing et al. 1996). Brotoisworo (1991) attributes the low skill of workers to part of the 35-40% of the logging waste he found. The estimated volume of waste due to felling and bucking losses is about 6.5-8.5% of the utilizable stem volume (FAO 1989a, Winkler 1997). In addition to volume loss due to poor felling and bucking techniques there can be significant value losses.
A problem outlined by (Quirós et al. 1997) is that loggers are paid based on the volume removed. Therefore, they only take out the best and largest logs, resulting in 20-25% of the cut volume not being extracted due to felling damage or poor quality. In many cases the logging waste left is suitable for supplying local markets through small-scale sawmilling. Hendrison (1989) also found that serious wood damage and quality loss can occur during positioning and collecting (bunching) logs with the blade of a skidder.
Logger training is a key factor in reducing logging waste and value loss. Uhl et al. (1997) found that trained loggers were able to achieve a 300% reduction in waste associated with felling a bucking, while Winkler (1997) found a 120% reduction. DeBonis (1986) also found that a 15-30% increase in wood volume at the mill could be realized through proper felling and bucking techniques. Cross-cutting training programs have also shown that log values can be increased by 10-50% (Dykstra and Heinrich 1996).
Wood volume losses or waste also occur at roadside landings, export ports, millyards and in manufacturing itself. For example, Kilkki (1992) found in a study in Papua New Guinea that 10-35% of the export volume was left at the harbour as not fulfilling export grade rules. Bethel (1984) states that overall product yield from a tree can be as low as 10 to 20%, and typically averages no more than 30%. This is supported by Buenaflor and Karunatilleke (1992) who state that 70% of the wood being logged from natural forests is wasted owing to both poor harvesting and mill processing, and the non-availability of markets for all wood.
Mill process yields have been reported to be as low as 33% of delivered log volume (e.g., Barros and Uhl 1995, Gerwing et al. 1996, Uhl et al. 1997). Noack (1995) reported sawmill lumber recovery factors (LRF) ranging from 36% to 57%. In other reports reviewed, the LRF reported varied from 45-55% (Silitonga 1987, Kilkki 1992, Verissimo et al. 1995, Muladi 1996). When sawing large diameter tropical hardwood logs the LRF should be at least 50% (Uhl et al. 1997), and yields of 56-68% should generally be expected (Niedermaier 1984).