16. Financial and economic analyses of conventional and reduced impact harvesting systems in Sarawak - Aaron Ago Dagang*, Frank Richter**, B. Hahn-Schilling*** and Penguang Manggil****

* Forester, Counterpart to FOMISS Project, Forest Department Sarawak, Tel: ++(60 82) 31 9171, E-mail: [email protected]

** Consulting Forester, Consultant to FOMISS Project, Germany, Tel: ++ (49 5544) 940221, E-mail: [email protected]

*** Chief Technical Advisors to FOMISS Project, GTZ, Germany, Tel: ++(60 82) 44 9446, Fax: ++(60 82) 44 9326, E-mail: [email protected]

**** FOMISS Project Director, Forest Department Sarawak, Tel: ++(60 82) 31 9188, Fax: ++(60 82) 44 5640, E-mail: [email protected]

INTRODUCTION

Timber harvesting with tractors forms the standard extraction method in the mixed hill dipterocarp forests of Sarawak. Due to the frequent occurrence of red-yellow podzolic soils, heavy rainfall and environmentally difficult terrain, conventional logging (CL) with tractors can cause substantial disturbance and damage to soils and forest stands (Nussbaum, 1995; Butaud, 1998). For this reason, during the last few years guidelines on reduced impact logging (RIL) have been developed and implemented in many trial areas of tropical natural forests (Jonathan et al., 1999).

Forest managers who, in principle, support environmentally sound harvesting methods are concerned about the financial implications. They claim that the costs will not be offset by savings, additional revenue or fiscal incentives to enable them to cope with the additional financial burden arising from implementing RIL.

This paper summarizes the results of comparative financial and economic analyses of CL and RIL with tractors for the FOMISS-Samling Pilot Area (FSPA) at Ulu Baram, Sarawak. Incentives to promote RIL are also discussed.

STUDY SETTING

Study area

Since January 1998, the FOMISS Project (Forest Department Sarawak and GTZ Cooperation Project) and the private sector (Samling Strategic Corporation Sdn. Bhd.) have implemented a sustainable forest management system (SFMS) in a pilot area. The FSPA is located in the Upper Baram of northeast Sarawak, covering 169 000 ha. The mean annual rainfall ranges from 4 600 mm to 5 000 mm. Sedimentary rocks and a moderately sloping to steep mountainous topography characterize the area. Fifty-six percent of the FSPA exceeds gradients of 25 degrees. The major soil groups are Regosols/Acrisols. Mixed hill dipterocarp forests are the dominant forest type, covering 80 percent of the area. Fifty percent percent of the FSPA comprises primary forests with an average commercial volume of 303 m³/ha. The average stocking of logged-over forests averages 170 m³/ha.

Logging systems

The RIL system tested in the FSPA consists of the following elements (Jonathan et al., 1999):

1. Alignment of skid trails.

2. The skid trail alignment is mapped; the harvestable trees are marked and climbers are cut.

3. A final harvesting map is produced indicating skid trail locations and areas to be excluded from logging.

4. The final harvesting map is verified and cross-checked.

5. The harvesting operation is carried out. A Caterpillar D6 DLS crawler tractor with a winch is used for skid trail and log landing preparation and for log extraction. The tree fellers use Stihl 07 chainsaws and are trained in directional felling. Tractors should not leave the skid trails when extracting logs. All trees felled and extracted are recorded. The complete harvesting operation is monitored.

6. Post-harvesting and damage assessments are carried out.

7. The RIL trial is analysed, including a RIL compliance assessment.

CL is also carried out with crawler tractors. In contrast to RIL, the winch is used only for short-distance log extraction. Breakouts from the main skid trails are the rule. Tree fellers choose the felling direction. Trees are felled in the direction of their natural lean. Under CL, no pre- or post-harvesting operations are conducted.

METHODS

General

The study analysed two different time frames.

1. The financial costs and benefits of CL and RIL were calculated for a one-year period of harvesting. Costs and benefits were not discounted.

2. A period of 40 years was taken into account for the financial and economic analyses. Harvesting operations are carried out after a 40-year rotation period in a logged-over forest. To compare costs and benefits streams occurring at different times, discounting was applied.

Cost-benefit analysis was used as follows:

1. The financial analysis was conducted from an enterprise perspective and dealt with actual cash flows. Only traded goods and services were considered and valued by applying market prices.

2. The economic analysis was conducted from society’s perspective. It considered traded and non-traded costs and benefits. Shadow pricing was used to adjust financial costs and benefits to reflect their economic values.

Data sources

The project established 13 RIL trial blocks in the FSPA. Additional data were obtained from RIL trials in the Model Forest Management Area (MFMA) in Sibu. General information on CL was obtained as average data from the entire FSPA. Harvesting costs for six RIL blocks and for five CL blocks were provided by Samling (Samling, 2001). The main source for the evaluation of non-timber forest values was a study by Sander (2000a).

Post-harvest damage assessments were conducted in six RIL blocks and in six CL blocks. Only severely damaged trees were considered for the calculation of tree damage to the residual stand.

For RIL, data on skid trail length was obtained from the harvesting map. A study by Richter (2000) focusing on forest rehabilitation in the FSPA provided information on the proportion of ground area compacted by CL.

The utilization factor was defined as the percentage of the wood volume extracted as commercial timber and measured in the stand in relation to the gross volume of the standing trees. During the harvesting operation, FOMISS field staff measured felled and trimmed logs.

Logging waste occurs when logs are left behind at the log landing, and through second trimmings at the log landing or royalty assessment point. Logging waste was estimated by comparing the commercial log volume measured after the first trimming at the felling site with the official commercial volume, as recorded at the royalty assessment point.

Figures on forest growth of the residual stand were predicted with the dipterocarp forest growth simulation model (DIPSIM). The model is based on permanent growth and yield plots from various locations in Sarawak. Input data was obtained from 275 sample plots distributed within the FSPA. The simulation was carried out for a net production area of 39 103 ha.

The tree volume used for the estimation of biomass (Brown, 1997) was obtained from DIPSIM. Based on figures provided by Brown (1997), an arithmetic mean wood density of 0.57t/m³ was applied. A conversion factor of 0.5 was used to express biomass in terms of carbon (Moura Costa, 1996). Carbon storage per hectare was calculated for five-year intervals. The average of these values was considered for the economic analysis of carbon storage.

Economic indicators and sensitivity analysis

The economic indicators selected were the net present value (NPV) and the benefit-cost ratio (BCR). A sensitivity analysis was carried out to estimate how changes in key technical and economic parameters would alter the economic performance of the two harvesting systems. The following parameters were considered: (1) discount rate, (2) timber price, (3) carbon trading, (4) harvesting costs and (5) harvesting volume.

DATA FOR THE COST-BENEFIT ANALYSIS

General data

A 10 percent discount rate ® was selected, reflecting real interest rates. To simplify the calculation of shadow prices, a standard conversion factor of 1.2 on traded financial prices was applied.

Forest zoning has been used to classify the FSPA area according to ecological, social and managerial criteria. The net production area is derived by deducting from the gross area the areas for protection, community use, permanent infrastructure, rivers and buffer zones (Kleine, 2000). The cost-benefit analysis was carried out for two theoretical blocks with a net production area of 100 ha, representing the average block size within the FSPA.

Experiences with RIL in the FSPA have demonstrated that due to small-scale terrain features, like rock outcrops, some sections of the block cannot be harvested. This results in a reduction of the net production area. The remaining net area where trees can be harvested is defined as net operable area (Kleine, 2000). For the first harvest, the net operable area amounts to 90 ha for CL and to 70 ha for RIL.

The net operable area is reduced further after the first harvest by highly compacted skid trails and log landings. The net operable area for the second harvest under CL was reduced by 13 ha (Butaud, 1998; Richter, 2000), resulting in a net operable area of 77 ha. For the RIL blocks it was calculated that 3.1 percent of the area was subject to compaction. Hence, the net operable area for the 40-year period under RIL amounts to 67 ha.

Based on DIPSIM simulation results, the rotation period for CL and RIL was set at 40 years. The results of the damage assessment demonstrated that the percentage of severely damaged trees was 54 percent for CL and 28 percent for RIL.

RIL reduces damage to the residual stand and leads to better quality of the future harvestable stand. In this regard, a quality factor was introduced. Under CL, a quality factor of 50 percent was defined; thus, 50 percent of the volume increment was allocated to trees that are defective or damaged. A quality factor of 65 percent was set for RIL. In the absence of sound scientific data, these values represent professional estimates.

Under RIL, a utilization factor of 80 percent was calculated. The utilization factor under CL amounted to 75 percent. Under CL, the volume of logging waste amounted to 20 percent of the total extracted volume. In RIL, logging waste was assumed to be zero.

Company records indicated that the timber volume extracted between 1994 to 1999 averaged 44.5 m³/ha in CL. A study by Richter (2000) revealed that an additional 20 m³/ha had been extracted, mainly for the construction of bridges and camps. Thus, the overall harvested volume under CL totalled 64.5 m³/ha. The average harvested volume under RIL amounted to 27.8 m³/ha.

The DIPSIM simulation indicated that in the second harvesting operation the timber volume available for extraction would be 23 m³/ha in CL and 85 m³/ha in RIL. The results considered the utilization factor, quality factor and logging waste. However, a maximum harvested volume of 40 m³/ha was defined.

Cost centres for the financial analysis

Harvesting operation costs

Only those aspects of the harvesting operations that are influenced by the introduction of RIL were analysed: (1) training; (2) harvest planning; (3) skid trail preparation, tree felling, log extraction; and (4) post-harvest operations.

The costs amount to RM28/m³ and RM43/m³ for CL and RIL, respectively (Table 1).^[21] Costs related to training and damage assessments are introductory costs. These cost centres were not taken into account for the 40-year period. Thus, the overall RIL costs are reduced to RM37/m³.

The average royalty cost, weighted by species share, was RM 37/m³.

Revenues for the financial analysis

Sander (2000b) provided prices for calculating timber revenues. An export quota of 40 percent was considered. A price difference of 17.5 percent between domestic prices and international prices was applied (Sander, 2000a). A 10 percent certification premium was added to the timber prices for RIL. In applying these figures the weighted price per cubic metre comes to RM314.5 (CL) and RM346.0 (RIL).

Based on a study by Sander (2000a) it was determined that 70 percent of the total timber revenues under CL (RM221/m³) is spent on forest operations, provided that costs for harvesting operations and royalties are not considered. It was assumed that, except for harvesting operations and royalties, the operational costs of the two options are identical. Thus, RM221/m³ was subtracted from the weighted timber price, resulting in a profit of RM95 m³ and RM126/m³ for CL and RIL, respectively.

Cost centres for the economic analysis

The economic analysis considers the same components of the harvesting operation as the financial analysis. The percentage of costs of tradable goods relative to the total harvesting costs was estimated for the individual subcentres: (1) training (0 percent); (2) harvest planning (2 percent); (3) harvesting operation (70 percent); and (4) post-harvest operation (2 percent). These cost components were multiplied with the standard conversion factor of 1.2 in order to estimate the economic values of the harvesting operations.

Table 1. Comparative costs for CL and RIL

Explanation:

Revenues for the economic analysis

The following aspects were analysed: (1) timber; (2) carbon storage; (3) non-timber forest products (NTFP); (4) soil values; (5) recreational values; and (6) biodiversity values.

Timber

The economic value of timber was calculated by multiplying the extracted volume with the average export price, weighted by species share. This results in an average timber price of RM351/m³ for CL and RM387/m³ for RIL (including 10 percent certification supplement). After deducting the overall operational costs, the economic value of timber revenues amounts to RM105/m³ (CL) and RM167/m³ (RIL).

Carbon storage

The current cost of tropical forestry carbon offsets ranges from US$2-10 per ton of carbon, averaging about US$8 (Stuart and Moura Costa, 1998). This average was used for the valuation of carbon storage. The total value of carbon storage per hectare net production area amounts to RM4 522 (CL) and RM4 962 (RIL). The values of carbon stocks were translated into flows by calculating annuities discounted at 10 percent.

NTFPs

The valuation of NTFPs considered a legal decision of the High Court of Johor Baru in 1996, which determined that compensation of RM 1 236/ha had to be paid to local communities for a total loss of land resulting from the construction of a reservoir. In order to account for different economic NTFP values under different forest management options, a percentage-based valuation was applied (Sander, 2000a): (1) non-production areas produce 100 percent of the value (RM1 236/ha); (2) production forest under CL produces 50 percent; and (3) production forest under RIL generates 70 percent of the total value. The annuity was RM6 798 and RM9 764 per block for CL and RIL, respectively.

Soil values

Sander (2000a) calculated an annual value of RM104/ha for total watershed protection. To estimate the different soil erosion prevention benefits of CL and RIL, a percentage-based valuation approach was applied. Glauner (2000) considered a management factor (C = conservation practice factor) for erosion modelling. A value of C = 0.001 has been proposed for undisturbed forests (Wischmeier and Smith, 1978). In contrast, a value of C = 1 indicates barren land. CL increases C by a factor of 20 relative to undisturbed forests. Within 10 years, C returns to the original value (Glauner, 2000). RIL increases C by a factor of five and cuts the time of recovery relative to CL by half (Marsh et al., 1996).

Based on these C-values it was assumed that non-production areas provide 100 percent of the maximum soil erosion prevention value (RM104). The percentage is reduced to 5 percent and 20 percent under CL and RIL, respectively. Compacted areas (skid trails and log landings) were rated with 0 percent. By multiplying the respective areas of the management options with these values and by considering the different time frames for recovery, the following total annual soil erosion prevention benefits per hectare of net production area was obtained: (1) CL: years 1-10 = RM14 and years 11-40 = RM90; and (2) RIL: years 1-5 = RM45 and years 6-40 = RM101.

Recreational values

It was assumed that the net operable areas managed under CL have no recreational value. The recreational value of non-production areas was set at RM19 ha/yr (Pearce et al., 1999). By multiplying these values with the respective areas the annual recreational benefits per hectare were RM19 for CL and RM19 for RIL.

Biodiversity values

In adopting the values given by Sander (2000a), an annual biodiversity benefit of RM11.40/ha was assumed for non-production forest areas. Furthermore, it was assumed that under CL, 50 percent of this biodiversity value is maintained in comparison to 70 percent under RIL. Accordingly, the annual economic value of biodiversity per hectare amounts to RM6 for CL and to RM9 for RIL.

RESULTS

Comparison of costs and benefits for CL and RIL - one-year period

The costs of harvesting and royalties per cubic meter are approximately 23 percent higher under RIL relative to CL (Table 2, Scenario 1). If the introductory costs for training and damage assessment are excluded, the difference is reduced to 14 percent (Scenario 2).

The profit was estimated at RM29/m³ and RM45/m³ under CL and RIL, respectively (Scenario 1). If the introductory costs are excluded the profit increases to RM52/m³. The profit is reduced to RM20/m³ (Scenario 3) if the concessionaire does not obtain a certification premium.

However, considering the extracted volume, the profit is higher under CL. The profit per hectare of the net production area totals RM1176 for CL and RM883 for RIL (Scenario 1).

Table 2. Cost and revenues for CL and RIL systems in a one-year period

Note:

Scenario 1 Harvesting costs under RIL include costs for training and damage assessment; revenues consider a certification premium under RIL and it was assumed that the costs for road construction and maintenance, debarking, scaling grading, transport, inventory, overheads, silvicultural treatment operations etc. were the same under CL and RIL.

Scenario 2 Harvesting costs under RIL do not include costs for training and damage assessment.

Scenario 3 Harvesting costs under RIL do not include costs for training and damage assessment; revenues do not consider a certification premium under RIL.

Financial analysis - 40-year period

The NPV and the BCR indicate that both management options are profitable (Table 3). CL was slightly more profitable and exceeds the NPV of RIL by a factor of 1.3. The BCR for RIL exceeds that of CL by a factor of 1.1. This means that RIL is more robust with regard to cost increases.

In terms of the financial cash flow (Figure 1 and 2), the first harvest was more profitable under CL than under RIL, whereas the financial benefits of the second harvest were higher under RIL. However, the effect of discounting reduces the revenues of the second harvest to a very small amount.

Table 3. Indicators for the financial analysis of CL and RIL systems; discount rate - 10 percent; calculation period - 40 years; production area = 1 ha

Note: The IRR cannot be calculated, because the aggregated flow of costs and revenues is never negative over the whole calculation period.

Figure 1. Financial cash flow under the CL system

Figure 2. Financial cash flow under the RIL system

Economic analysis - 40-year period

The economic analysis showed that the RIL system (NPV = RM9 905/ha) provided a higher level of overall benefits and welfare to the society as a whole as opposed to CL (NPV = RM9 100/ha).

Sensitivity analysis

The results of the sensitivity analysis demonstrated that RIL was more profitable than CL if:

1. a discount rate of less than 3 percent is applied; or

2. the harvesting costs are reduced by >30 percent; or

3. the timber price increases by >15 percent; or

4. a minimum of RM30/ha of annual carbon trading payments are generated; or

5. the extraction volume in the first harvesting operation under RIL is increased to >36 m³/ha.

To assess the effects of modifying several factors on the NPV, two realistic scenarios were tested (Table 4).

The results demonstrate that the RIL system is more profitable if the harvesting costs are reduced by 10 percent and that the logging intensity during the first cut is increased by 20 percent. Assuming that the concessionaire receives additional revenues of RM15/ha/yr through carbon trading payments, the NPV under RIL would increase to RM1 340, i.e. a 13 percent increase over the CL system.

Table 4. NPV of the CL and RIL systems, assuming that several elements under RIL are modified - 40-year period

Explanation:

INCENTIVE SYSTEMS TO PROMOTE THE APPLICATION OF RIL TECHNIQUES

Security of tenure

The development of long-term, legally binding land use planning has been stressed as the most important aspect to promote SFM (De Graaf, 2000). Increased tenure security for timber companies will facilitate contractual arrangements with outsiders and lower the private discount rates (Richards and Moura Costa, 1999). However, the conversion from short-term harvest licences to long-term agreements will not prevent land users from acting in a way that imposes social costs. Hence, the extension of the licence period is a basic prerequisite, but additional incentives are required to encourage the implementation of RIL.

Regulation of royalties

The study demonstrated that about 20 percent of the extracted volume is wasted. The logging waste is a consequence of the present royalty assessment procedures. The state levy on harvested timber is only collected far from the felling site in the forest. Any timber that does not arrive at the royalty assessment point is not accounted for. Logs are trimmed excessively and often only the best log grades will reach the assessment point, whereas lower quality timber is left to rot in the forest.

To achieve higher timber utilization efficiency the royalty rate should be determined as close as possible to the felling site. The taxation could be based on the gross standing volume. This method, however, appears to be impractical since the determination of standing stock volumes frequently is biased in measuring the clear bole height. Alternatively, the volume measured immediately after tree felling could serve as a reference for determining the royalty.

Domestic fiscal market-based instruments

Market-based instruments internalize social costs and benefits into private costs and returns. Performance bonds are a type of market-based instrument. Before harvesting is initiated, the entrepreneur deposits a refundable bond into a state government account. If harvesting is executed in accordance with RIL standards, the value of the bond is returned gradually to the concessionaire. Any fines for poor compliance with RIL standards are deducted.

Market instruments based on public goods benefits

The most important market instruments based on public goods benefits are timber certification and carbon trading. The main objective of certification is the establishment of an environmentally discriminating market, which should cause concessionaires to move from timber mining to SFM so they can sell products globally and without restrictions.

Carbon trading has a high potential with regard to the promotion of RIL. It imparts public goods values to the forest. International polluters compensate for these values through transfer payments, and these are internalized by timber concessionaires. The key advantage of trading certified carbon credits lies in the generation of a continuous revenue flow from the first year onwards.

Payment systems

In contrast to the above, innovative financing and internal mechanisms, which are based on national or international schemes, are ‘internal’ incentive systems of timber concessionaires. The payment system for forest workers is a crucial aspect within this context (Sist, 2000).

The implementation of RIL demands skilled, trained and responsible harvesting teams. In addition, a modified payment system is desirable to provide an adequate incentive for the achievement of the RIL objectives. A simple piecemeal rate is not suitable (Sist, 2000). Instead, a performance-based remuneration system is required that takes into account the quality of work and rewards forest workers for high-quality work. Experiences during RIL trials indicate that a pure time-based work rate does not provide an adequate incentive to efficient work.

Considering these aspects, a payment system (Figure 2) is proposed that is composed of three elements: (1) a fixed monthly salary, (2) a piecemeal bonus and (3) a quality dependent reward. The quality of work is controlled through a RIL compliance assessment that applies a simple point ranking system. In case points (2) and (3) are satisfactorily achieved, the total monthly salary will be higher than the average monthly income under CL.

Figure 2. Payment system for harvesting crews under a CL system and a modified payment system for harvesting crews under a RIL system

Explanation: Compliance classes: (a) full compliance, (b) good compliance, (c) satisfactory compliance, (d) poor compliance, (e) non-compliance

CONCLUSIONS AND RECOMMENDATIONS

The high extraction volume of the first harvest under CL strongly influences the results of the cost-benefit analysis for the 40-year period. Disregarding the negative environmental and social costs associated with CL, this system is financially more profitable than the RIL system, especially because of the effects of discounting, which reduce the higher RIL revenues of the second cut to a very small value.

The high logging intensity and careless extraction methods under CL result in an excessive number (54 percent) of severely damaged trees in the residual stand. In addition, the productive forest area is reduced by 13 percent due to unplanned skid trail and log landing construction.

Growth simulations indicate that both factors diminish the overall value of the forest resource. Forty years after the initial harvest the gross commercial volume of conventionally logged stands amounts to 196 m³/ha. This is in a sharp contrast to a commercial gross stocking of 303 m³/ha in primary forest stands.

The introduction of RIL reduces the damage to the residual stand to 28 percent due to improved logging practices and a lower harvesting intensity, which leads to an increased quality of the future harvestable stand. The area loss due to skid trail construction was lowered to 4 percent. Improved felling and trimming and, even more important, the reduction of logging waste accelerates the recovery of the logged-over forest. DIPSIM simulations support the sustainability of RIL. Forty years after the initial harvest, the gross commercial volume of stands that were logged with an environmentally sound harvesting method amounts to 317 m³/ha.

Post-harvesting assessments of stand conditions under RIL indicate that the benefits for silviculture and biodiversity conservation are potentially significant. This is supported by the economic analysis that considers timber and NTFP values, recreational values, soil conservation values, carbon stocks and biodiversity values. The analysis demonstrates that the introduction of RIL increases the overall value of the forest resource.

However, from the perspective of a timber concessionaire, CL is presently more profitable.

Nevertheless, reduced harvesting costs due to advanced experiences with RIL will contribute to a higher profitability of improved harvesting. However, as indicated by the sensitivity analysis, a reduction of 30 percent in harvesting costs would be required to make RIL more profitable than CL. Such a substantial decrease appears to be unrealistic.

The sensitivity analysis also indicates that a slight reduction of the harvesting costs and a simultaneous 20 percent higher harvesting intensity in the first cut is sufficient to make the RIL system more profitable than the CL system. Heavy cuts, however, would compromise the benefits of RIL seriously (Sist and Bertault, 1998), and harvesting intensity should be increased only after careful assessment of the consequences.

In addition, incentive systems, such as certification premiums or carbon trading payments can also help to facilitate the introduction of RIL.

To increase the profitability of RIL, it is recommended to increase the harvesting intensity of the initial harvest. Damage assessments should be carried out to control the effects of higher extraction volumes. CL and RIL still use the same piecemeal-dependent payment system. It is recommended to modify the piece-work-dependent salary system. A payment system is required that takes into account the quality of work and rewards workers for good practices.

REFERENCES

Brown, S. 1997. Estimating biomass and biomass change of tropical forests. FAO Forestry Paper 134. Food and Agriculture Organization of the United Nations. Rome, Italy.

Butaud, J.F. 1998. Damage assessment of hill dipterocarp forests subject to reduced impact logging guidelines. Junior professional report no. 10. Malaysian-German Technical Cooperation Project. Forest Department Sarawak and Forest Management Information System Sarawak (FOMISS): Kuching, Malaysia.

De Graaf, N.R. 2000. Reduced impact logging as part of the domestication of neotropical rainforest. International Forestry Review, 2 (1): 40-44.

Glauner, R. 2000. The role of soil-and-site science in management planning in managed natural forests. An example from Sabah, Malaysia. Mitteilungen der Bundesforschungsanstalt für Forst- und Holzwirtschaft no. 197. Komissionsverlag Max Wiedebusch: Hamburg, Germany.

Jonathan, R., Rani, A. & Hahn-Schilling, B. 1999. Reduced impact logging trial in the FOMISS-Samling Pilot Area. RIL - Phase II. Implementation guideline. Misc. report no. 22. Malaysian-German Technical Cooperation Project. Forest Department Sarawak and Forest Management Information System Sarawak (FOMISS): Kuching, Malaysia.

Kleine, M. 2000. Annual allowable cut (AAC) determination. Draft technical paper. Malaysian-German Technical Cooperation Project. Forest Department Sarawak and Forest Management Information System Sarawak (FOMISS): Kuching, Malaysia.

Marsh, C.W., Pinard, M.A., Putz, F.E., Tay, J. & Sullivan 1996. Reduced impact logging: a pilot project in Sabah, Malaysia. In: Schulte, A. & Schöne, D. (eds.): Dipterocarp forest ecosystems. World Scientific: Singapore: 293-307.

Moura Costa, P. 1996. Tropical forestry practices for carbon sequestration. In: Schulte, A. & Schöne, D. (eds.): Dipterocarp forest ecosystems. World Scientific: Singapore: 308-334.

Nussbaum, R.E. 1995. The effects of selective logging of tropical rainforest on soil properties, and implications for forest recovery in Sabah, Malaysia. Thesis. Faculty of Science. University of Exeter: Exeter, UK.

Pearce, D., Putz, F. & Vanclay, J.K. 1999. A sustainable forest future. Centre for Social and Economic Research of the Global Environment (CSERGE) Working Paper GEC 99-15. Center for Social and Economic Research on the Global Environment, University of East Anglia, Norwich.

Richards, M. & Moura Costa, P. 1999. Can tropical forestry be made profitable by “internalising the externalities”? Natural resource perspectives no. 46. Overseas Development Institute: London.

Richter, F. 2000. Silvicultural Decision Support System for the treatment of harvested natural forests based on the interpretation of aerial photos. Consultancy report no. 48 Malaysian-German Technical Cooperation Project. Forest Department Sarawak and Forest Management Information System Sarawak (FOMISS): Kuching, Malaysia.

Samling. 2001. Cost comparison between conventional logging and RIL. Unpublished data.

Sander, K. 2000a. Financial and economic analysis of forest management systems at FOMISS-Samling Pilot Area (FSPA). Consultancy report no. 52. Malaysian-German Technical Cooperation Project. Forest Department of Sarawak and Forest Management Information System Sarawak (FOMISS): Kuching, Malaysia.

Sander, K. 2000b Market analysis for exports of logs and selected timber products from Sarawak, Malaysia. Annex to: Financial and economic analysis of forest management systems at FOMISS-Samling Pilot Area (FSPA). Consultancy report. Draft 1. Malaysian-German Technical Cooperation Project. Forest Department Sarawak and Forest Management Information System Sarawak (FOMISS): Kuching, Malaysia.

Sist, P. 2000. Reduced impact logging in the tropics: objectives, principles and impact research. International Forestry Review, 2(1): 3-10.

Sist, P. & Bertault, J.-G. 1998. Reduced impact logging experiments: impact of harvesting intensities and logging techniques on stand damage. In: Bertault, J.-G. & Kadir, K. (eds.): Silviculture research in a lowland dipterocarp forest of East Kalimantan. The contribution of STREK project. CIRAD-forêt, FORDA & P.T. Inhutani I: Montpellier, France: pp. 139-160.

Stuart, M.D. & Moura Costa, P. 1998 Climate change mitigation by forestry: a review of international initiatives. Policy that works for forests and people series no. 8. Discussion paper. International Institute for Environment and Development: London, UK.

Wischmeier, W.H. & Smith, D.D. 1978 Predicting rainfall erosion losses - A guide to conservation planning. US Department of Agriculture. Agriculture Handbook no. 537. US Government Printing Office: Washington, DC.