15. Financial indicators of reduced impact logging performance in Brazil: case study comparisons - Thomas P. Holmes*, Frederick Boltz** and Douglas R. Carter***

* Research Forester, Southern Research Station, USDA Forest Service, PO Box 12254, Research Triangle Park, NC, USA 27709, Tel: ++(1 919) 549 4031, Fax: ++(1 919) 549 4047, E-mail: [email protected]

** E.T. York Presidential Fellow, School of Forest Resources and Conservation, Institute of Food and Agricultural Sciences, 357 Newins-Ziegler Hall, University of Florida, Gainesville, FL, USA 32611-0410, Tel: ++(1 352) 846 0904, Fax: ++(1 352) 846 1277, E-mail: [email protected]

*** Associate Professor, School of Forest Resources and Conservation, Institute of Food and Agricultural Sciences, 357 Newins-Ziegler Hall, University of Florida, Gainesville, FL, USA 32611-0410, Tel: ++(1 352) 846 0893, Fax: ++(1 352) 846 1277, E-mail: [email protected]

INTRODUCTION

The neo-classical theory of the firm is built on the presumption that businesses attempt to maximize profits, where financial profits are simply the difference between the revenue received by a firm and the costs it incurs. Economic theory says that, for a given technology, the firm evaluates the various ways it can utilize labour, land and capital inputs to produce outputs. Maximum profits are gained by choosing input levels so that the value of the marginal product produced by each input is equal to its cost (Varian, 1984). If inputs to the production process are non-priced (such as environmental quality), or under-priced (such as standing timber) then they will be over-utilized from a social perspective. In the case of forestry, the historical result has been timber mining, degradation of environmental quality and industrial migration.

The cycle of timber depletion, environmental degradation and industrial migration is not a new story. It occurred in the primary forests of the USA (Williams, 1989) and has proceeded to such tropical countries as Brazil (Nepstad et al., 1999). The demand for tropical timbers suggests that this process will continue unless significant changes in technology and/or policy are implemented. Tropical forests, formerly under little pressure for timber production, are now increasingly the focus of logging industry development. Growth in the Latin American and African share of total tropical timber production will likely continue, as few Asian countries have the potential to substantially increase sustainable log production (ITTO, 1996). Recent trends in tropical timber production show a decrease in the Asia-Pacific region’s share of global production by 29.6 percent from 1992 through 1999. Production in the Latin America-Caribbean region increased 15.8 percent over this same period (ITTO, 1999).

Conventional logging (CL) practices are recognized as a principal contributor to degradation and ultimate conversion of tropical forest ecosystems to non-forest land uses (Johnson and Cabarle, 1993; Bryant et al., 1997). The decreased productivity of forests following damaging CL entries may translate into higher opportunity costs for long-term forest management and greater incentive for the conversion of forestland to alternative uses.

Reduced impact logging (RIL) practices comprise harvest planning, infrastructure development and operational techniques that aim to reduce the damaging impacts of logging while improving the production efficiency of logging operations. The FAO model code of forest harvesting (Dykstra and Heinrich, 1996) provides the basis for the RIL system design. RIL techniques and guidelines are not fixed prescriptions, but adapt the best harvesting techniques to existing biophysical and economic conditions. Throughout the tropics, RIL has proven more ecologically benign than conventional logging activities (Boxman et al., 1985; Johns et al., 1996; Pinard and Putz, 1996; Uhl et al., 1997). Furthermore, RIL reduces operational costs (Boxman et al., 1985) and, in some cases, generates higher initial financial returns than conventional operations (Barreto et al., 1998; Holmes et al., 2000). RIL systems may provide a low-cost method of maintaining the carbon sequestration functions (Putz and Pinard, 1993; Boscolo et al., 1997) and the structural diversity of tropical forests (Frumhoff and Losos, 1998). However, it has not been demonstrated that RIL operations alone are sufficient for the sustained production of merchantable timber or for the maintenance of the environmental service flows provided by tropical forests in their natural, unaltered state.

Financial self-interest is a strong motivating force. Understanding the financial aspects of RIL under different ecological, industrial and market conditions is imperative if sustainable forest management is ever to become a reality in tropical forests. If a “feasible financial set” of conditions is identified where RIL is more profitable than CL practices, then self-interest may help to protect ecological services after initial harvest entries in some logged tropical forests.

In this paper, we compare indicators of financial performance for three case studies in the Brazilian Amazon. To conduct the analysis, we disaggregate case study results into common measures of productivity, cost and profitability. Direct comparisons are complicated by the fact that standard protocols were not used across studies. To provide a tractable analysis, we utilize incremental measures, where increments measure proportional changes between CL and RIL systems.

CASE STUDIES

International donors and the private sector funded the following studies conducted during the mid-1990s:

1. Agrosete (Barreto et al., 1998; Johns et al., 1996): The RIL-CL comparison was conducted on private forestland of Fazenda Agrosete, approximately 20 km southeast of Paragominas, Pará, Brazil. RIL was conducted on a 105-ha plot and CL on an adjacent 75-ha plot. Trained operators worked with the research team on the CL and RIL plots.^[15] RIL extracted 4.5 and CL 5.6 trees per hectare. Productivity, cost and logging waste measures were drawn from observed operations and plot impacts. Lowland, closed-canopy terra firme forests of Paragominas are humid, evergreen with a canopy height of 25 to 40 m and emergents extending to 50 m. The terrain is moderately undulating and soils are kaolinitic red-yellow Oxisols. Annual rainfall averages 1 750 mm with a distinct dry season from June to November. The mean annual temperature is 28° C.

2. Cauaxi (Holmes et al., 2000): Research was conducted on private forestland of the CIKEL timber company of Fazenda Cauaxi, some 120 km southwest of Paragominas, Brazil. RIL was conducted by trained operators of Fundação Floresta Tropical (FFT) on 100 ha of undisturbed forest, while CL was implemented by local contractors hired by CIKEL on an adjacent 100-ha plot. CL harvested 39 and RIL 41 timber species, at intensities of 4.25 and 3.31 trees per hectare, respectively. Logging intensity data and wastewood measures were collected in the Cauaxi plots. Average productivity and cost measures for the study were calculated from a sample of FFT RIL operations and CL operations in the Paragominas region. The study was conducted in lowland, closed-canopy terra firme forests of the Paragominas timbershed (see Agrosete above for ecosystem description).

3. Itacoatiara (Winkler, 1997): The study examined private forestland of Mil Madeireira Itacoatiara S.A., a Brazilian subsidiary of Precious Woods, Ltd., which is located 227 km east of Manaus. Efficiency and environmental impact studies were conducted in two adjacent 10-ha cutting blocks. Production costs per component were estimated as a proportion of total logging costs, while specific per unit costs were not reported. Both RIL and CL operations were implemented by the Precious Woods, Ltd. logging team. CL removed 16 and RIL 6 trees per hectare, or 78.9 percent and 26.9 percent of the available merchantable volume per plot. Sixty-five tree species were of commercial interest, of which 24 were harvested under RIL and 32 under CL. The lowland, moist terra-firme forests lie upon inclined plateaus of tertiary origin. Steep ravines dissect the plateaus at slopes of 10° to 40° (Precious Woods, 1997). Soils are Oxisols. Canopy height is 30 to 37 m with emergents extending to 55 m. Annual rainfall is around 2 200 mm with a dry season from June to October. The mean annual temperature is 26° C.

Table 1. Logging characteristics and financial cost estimates from three studies in the Brazilian Amazon

Table 1 shows the general logging characteristics and provides a summary of relevant cost data at the three study sites. Variation in harvest intensity, particularly at Itacoatiara, is observed. Because logging costs generally decrease, up to some point, as harvest intensity increases, large differences in harvest intensity may obfuscate meaningful comparisons. In addition, the reader is warned that comparisons of cost data can be misleading because identical activities may or may not be included in each cost category, and different protocols may have been used to collect data. Further, cost data from Itacoatiara are not presented in monetary units, but only as percentages of total cost. However, the authors were careful in constructing Table 1, and it presents a summarization of the best comparative data available regarding RIL and CL parameters in the Brazilian Amazon.

PRODUCTIVITY COMPARISONS

Directional felling is more time consuming and thus less productive under RIL when RIL and CL extract similar volumes and target stems (Figure 1). CL sawyers were 10 to 22 percent more productive in volume produced per hour (m³/hr) than comparable RIL-felling teams. Importantly, at Fazenda Agrosete, Barreto et al. (1998) found that gains in directional-felling productivity by a 3-person team rendered RIL more efficient than CL 2-person felling. On average, the 3-person RIL team felled 34 trees per day relative to the 22 trees felled by 2 CL sawyers. Productivity gains exceeded the increased cost of labour and equipment for the 3-person team. At Itacoatiara, Winkler (1997) found that felling time per stem was higher under RIL. However, because the RIL operation focused on “only the most mature trees of commercial interest”, greater volume per stem was recovered in RIL felling operations, resulting in its greater efficiency relative to CL.

Figure 1. Incremental productivity of CL felling (m³/hr), computed as (CL productivity - RIL productivity)/RIL productivity

Skidding operations are more productive under RIL due to efficient planning and infrastructure development (Figure 2). RIL utilizing rubber-tyre skidders on moderately undulating sites in the Brazilian Amazon increased productivity by 41 to 49 percent over CL bulldozer operations. Unplanned, conventional skidding is less efficient and thus more costly due to delays and damage caused in “roaming”, or searching for felled stems in an uncharted forest. Using a bulldozer for skidding in a planned RIL operation increased skidding productivity by 5 percent over CL-skidding productivity (Barreto et al., 1998).

Winkler (1997) found lower RIL-skidding productivity at Itacoatiara. However, the components of RIL skidding comprised pre-skidding (skid trail opening, winching to skid trail) and skidding to the log deck. The CL operation only comprised traditional skidding activities, and used a rubber-tyre skidder.

Figure 2. Incremental productivity of RIL skidding (m³/h), computed as (RIL productivity - CL productivity)/CL productivity

COST COMPARISONS

RIL operations incur costs associated with pre-harvest activities (block layout and line cutting, inventory, vine cutting, data processing and mapmaking) and harvest planning activities (tree marking, road planning, log deck planning and skid trail layout) that are not incurred by CL operations. In addition, RIL requires special training of personnel that incurs costs beyond the on-the-job training received by CL operators. The crux of the matter is whether or not gains in efficiency attributable to planning operations equal or exceed the incremental RIL costs.

RIL costs

RIL investments in inventory, planning, vine cutting and infrastructure development up to a year before logging increases the proportional cost of pre-harvest operations (Figure 3). The incremental pre-harvest costs of RIL are expected to be an important disincentive to RIL adoption by the logging industry (Barreto et al., 1998; Hammond et al., 2000; Holmes et al., 2000). Inventory, vine cutting and road, log deck and skid trail layout generate the highest incremental costs to RIL. These costs are compounded forward from the time they are incurred to the time of harvest. In general, CL operations do not incur these “advance” costs. However, infrastructure costs associated with the construction of roads and log decks are decreased as a consequence of pre-harvest and harvest planning activities (Barreto et al., 1998; Winkler, 1997; Holmes et al., 2000).

Figure 3. Planning and infrastructure costs as a proportion of direct costs for RIL and CL

RIL training costs comprise 1 to 18 percent of total harvest cost for CL and RIL operations. Training costs vary considerably among the studies, though methods of calculation are not uniform, nor are these data reported by all studies. In Barreto et al. (1998), training costs were estimated as an immediate wage increase for RIL-trained personnel. In contrast, Holmes et al. (2000) amortized training costs over five years of logging operations.

Direct costs

RIL direct costs^[16] (in US$/m³) ranged from 3 percent lower to 34 percent higher than CL direct costs (Figure 4). Of the three case studies examined, RIL direct costs are lower than CL direct costs only at Fazenda Cauaxi in Paragominas (Holmes et al., 2000). At Cauaxi, pre-harvest and harvest planning activities, and the reduction in felling efficiency due to directional felling, resulted in a RIL incremental cost of $1.33/m³. However, efficiency gains in road construction, log deck construction, skidding and log deck operations resulted in a cost saving of $1.48/m³. Thus, the gain in efficiency more than offset the RIL incremental costs.

In contrast, Winkler (1997) found that direct costs of RIL were 9 percent higher than CL costs. This may be due, in part, to the fact that harvest intensity on the CL plot was more than 2.5 times greater than on the RIL plot, which would likely decrease the per unit direct cost of CL. Also, Winkler (1997), noted that planned changes to the RIL operation would result in RIL costs that are only 1.5 percent higher than CL costs.^[17]

Barreto et al. (1998) found that pre-harvest and harvest planning activities increased RIL costs by $1.87/m³. Gains in operational efficiency, particularly the skidding operation, resulted in a cost reduction of $0.24/m³, or a cost recovery of 13 percent of the incremental RIL expenditures. Overall, RIL direct costs were 34 percent higher than CL at Fazenda Agrosete.

Figure 4. Incremental change in direct costs attributable to RIL (US$/m³), computed as (RIL cost - CL cost)/CL cost

Wastewood accounting

At Fazenda Cauaxi, CL operations wasted 4.08m3/ha due to: (1) high stumps, (2) poor felling techniques resulting in split logs, (3) wood wasted in improper bucking, and (4) logs not found by skidding crews. RIL operations wasted 1.32m3/ha for these reasons. In addition, logs left unutilized on the log deck amounted to 1.97m3/ha (0.60m3/ha) for CL (RIL) operations (Holmes et al., 2000). Overall, wasted wood represented 24 percent (8 percent) of the recovered volume at Fazenda Cauaxi by CL (RIL) operations (Holmes et al., 2000).

At Fazenda Agrosete, CL (RIL) operations wasted 8.83m3/ha (0.40m3/ha) in the forest^[18]. These amounts represented 26 percent and 1 percent of volumes felled by CL and RIL crews, respectively (Barreto et al., 1998).

At Mil Madeireira Itacoatiara, CL (RIL) operations wasted 2.99m3/ha (1.31m3/ha). These amounts represented 9 percent (4 percent) of volumes extracted by CL and RIL crews, respectively.

When woodwaste is not accounted for, direct costs appear deceptively lower for CL. The Paragominas case studies reported accounting adjustments for costs associated with woodwaste, while Itacoatiara did not. Woodwaste incurs direct costs associated with felling, bucking, skidding and log deck activities and indirect costs^[19] by increasing the effective stumpage price (Holmes et al., 2000). It may be expected that waste costs are commonly not accounted for in CL operations, given that inventory and monitoring activities necessary for such accounting are not conducted. Although this asymmetric information effectively biases estimates of returns to logging, the exceptional profitability of logging provides conventional firms the luxury to function inefficiently and to ignore such losses.

PROFITABILITY COMPARISONS

When direct and indirect waste costs are accounted for, RIL net revenues are 18 percent to 35 percent greater than CL net revenues (Figure 5). Relative gains in profitability at Agrosete were due to two factors: (1) a greater amount of wood wasted per hectare was reported, and (2) waste adjustments were computed using stumpage and revenue impacts (Barreto et al., 1998). At Cauaxi, waste adjustments were computed using impacts on direct costs and on effective stumpage price. We can only speculate on the impact that waste accounting would have at Itacoatiara. However, accurate waste accounting clearly increases the competitiveness of RIL relative to CL.

Figure 5. Incremental change in net revenues attributable to RIL (US$/m³) after accounting for waste related losses, computed as (RIL net revenue - CL net revenue)/CL net revenue

Tropical timber harvesting of primary forests is highly profitable. CL in Paragominas demonstrates profit margins of 39 percent to 52 percent and RIL demonstrates profit margins of 46 percent to 63 percent^[20]. Although profit margins for RIL exceed CL for the studies reporting such measures, highly profitable CL firms face few incentives to alter their operations unless they face dramatic changes in market signals such as increases in stumpage prices or decreases in product prices. In this sense, logging firms may seek a “satisfactory”, rather than “maximal”, level of profit in initial harvest entries in primary forests.

DISCUSSION AND CONCLUSIONS

RIL appears competitive with or superior to CL in financial returns to initial harvest entries if wood wasted in the harvesting operation is accounted for fully. If stumpage is treated as a “free good”, or if it is under-priced, economic theory states that it will be over-utilized from a social perspective. This appears to be occurring in the areas of the Brazilian Amazon currently experiencing intensive commercial exploitation. Wasted wood incurs direct costs associated with wasted labour and equipment use. However, the major financial impact is related to the increase in effective stumpage price. Stumpage and timber prices are market signals that reflect economic scarcity. Current market signals (or the lack thereof) do not seem to provide incentives to adopt practices that appear immediately more costly. We recommend that a stumpage and timber price reporting series be instituted in the Brazilian Amazon. Such a series would benefit landowners and mill-owners by providing publicly shared information about resource values and trends. We expect that this would facilitate better resource planning and provide incentives for more informed and conservative use of timber resources.

Of major importance is the lack of standardized data that would permit an understanding of functional relationships between forest types, input and output prices, industrial scale and costs and returns. Development of data to facilitate the estimation of cost and profit functions, for the logging and milling industries in the Brazilian Amazon, would facilitate effective planning for industrial development by identifying conditions where the application of RIL operations in new markets would be efficient and competitive with more destructive logging practices.

Although this study reviewed only three case studies, the comparative analysis was hampered by the lack of standard protocols used by each study. We recommend that consideration be given to creation of standard cost accounting categories and methods for data collection. Only if a standard cost accounting system is developed and applied will meaningful broad-scale comparative analyses be possible.

RIL prescriptions define the pattern and intensity of harvesting and the resulting opportunity costs of RIL relative to CL. When RIL is designed to mimic CL harvesting in terms of the harvest level, species, size classes, and spatial distribution, gains in operational efficiency and waste reduction render RIL environmentally and economically superior to CL for initial harvest entries (Barreto et al., 1998; Holmes et al., 2000). However, when RIL is implemented as part of a forest management prescription, in which areas and stems are set aside to maintain productivity and ecosystem integrity, the opportunity costs relative to conventional liquidation harvest of all merchantable stems may be too great for RIL to be competitive. For instance, van der Hout (1999) found the cost and damage savings in spatially restricted harvesting of “clumped” species under CL were superior to those under a RIL prescription requiring spatially distributed, selective harvesting under RIL. Winkler (1997) notes that one-third of the RIL forest area was set aside as preservation forest, while no such measures were applied under CL. The opportunity costs of foregone merchantable timber in reservations likely leads to inferior RIL financial competitiveness relative to unconstrained, liquidation harvest of merchantable stems under CL, despite gains in operational and resource-use efficiency. These were the conclusions of RIL-CL studies in Sabah, Malaysia (Tay, 1999; Pinard et al., 2000), in which RIL was found financially inferior due largely to foregone timber excluded from RIL due to environmental harvesting restrictions. Economic incentives appear to be necessary to promote the adoption of RIL as part of a long-term forest management system.

Tenure security and forestland scarcity determine estimates of efficiency and profitability for loggers and the relative importance of incremental gains in resource-use efficiency that may be derived from RIL implementation. It is expected that important damage mitigation benefits will be derived in future harvest entries, given greater conservation of future crop trees and reduced environmental disturbance under RIL. Tenure security is critical to the inclusion of future harvest returns in management profitability analyses and to expectations of financial benefit for careful management and conservation relative to more destructive practices. Moreover, resource scarcity and harvests constrained to a fixed resource base provide greater relevance for issues of resource use and timber-recovery efficiency. Without land-availability constraints and clear market signals of scarcity, it is unlikely that loggers will be drawn to the marginal increments in resource-use efficiency that may be gained under RIL. In a broader landscape without resource constraints, the opportunity costs of more careful RIL management relative to maximizing turnover and throughput of timber may be too high for conventional firms to change their logging behaviour.

REFERENCES

Barreto, P., Amaral, P., Vidal, E. & Uhl, C. 1998. Costs and benefits of forest management for timber production in eastern Amazônia. Forest Ecology and Management,108: 9-26.

Boscolo, M., Buongiorno, J. & Panayotou, T. 1997. Simulating options for carbon sequestration through improved management of a lowland tropical rainforest. Harvard Institute for International Development, Cambridge, MA.

Boxman, O., de Graaf, N.R., Hendrison, J., Jonkers, W.B.J., Poels, R.L.H., Schmidt, P. & Sang, R.T.L. 1985. Towards sustained timber production from tropical rain forests in Suriname. Netherlands Journal of Agricultural Science, 33:125-132.

Bryant, D., Nielsen, D. & Tangley, L. 1997. The last frontier forests: ecosystems and economies on the edge. World Resources Institute, Washington, DC.

Dykstra, D.P. & Heinrich, R. 1996. FAO Model Code of Forest Harvesting Practice, Food and Agriculture Organization of the United Nations, Rome.

Frumhoff, P.C. & Losos. E.C. 1998. Setting priorities for conserving biological diversity in tropical timber production forests. Union of Concerned Scientists. Center for Tropical Forest Science, Smithsonian Institution, Washington, DC.

Hammond, D.S., van der Hout, P., Zagt, R.J., Marshall, G., Evans J. & Cassells, D.S. 2000. Benefits, bottlenecks and uncertainties in the pantropical implementation of reduced impact logging techniques. International Forestry Review, 2: 45-53.

Holmes, T.P., Blate, G.M., Zweede, J.C., Perreira, R. Jr., Barreto, P., Boltz, F. & Bauch, R. 2000. Financial costs and benefits of reduced-impact logging relative to conventional logging in the eastern Amazon. Tropical Forest Foundation, Washington, DC.

ITTO. 1996. Annual review and assessment of the world tropical timber situation. International Tropical Timber Organization, Yokohama, Japan. (May 5, 1998).

ITTO. 1999. Annual review and assessment of the world timber situation 1999. Document GI-7/99. International Tropical Timber Organization, Yokohama, Japan. (August 25, 2000).

Jenkins, M.B. & Smith, E.T. 1999. The business of sustainable forestry: strategies for an industry in transition. Island Press, Washington, DC.

Johns, J.S., Barreto, P. & Uhl, C. 1996. Logging damage during planned and unplanned logging operations in the eastern Amazon. Forest Ecology and Management, 89: 59-77.

Johnson, N. & Cabarle. B. 1993. Surviving the cut: natural forest management in the humid tropics. World Resources Institute, Washington, DC.

Nepstad, D.C., Veríssimo, A., Alencar, A., Nobre, C., Lima, E., Lefebvre, P., Schlesinger, P., Potter, C., Moutinho, P., Mendoza, E., Cochrane, M. & Brooks, V. 1999. Large-scale impoverishment of Amazonian forests by logging and fire. Nature, 398: 505-508.

Pinard, M.A., Putz, F.E. & Tay, J. 2000. Lessons learned from the implementation of reduced-impact logging in hilly terrain in Sabah, Malaysia. International Forestry Review, 2: 33-39.

Pinard, M.A. & Putz, F.E. 1996. Retaining forest biomass by reducing logging damage. Biotropica, 28: 278-295.

Precious Woods, Ltd. 1997. Management plan for sustained use of the forests of Mil Madeireira Itacoatiara Ltda. Draft version, February 18, 1997. Itacoatiara, Amazonas, Brazil.

Putz, F.E. & Pinard, M.A. 1993. Reduced-impact logging as a carbon-offset method. Conservation Biology, 7: 755-759.

Tay, J. 1999. Economic assessment of reduced impact logging in Sabah, Malaysia. Ph.D. dissertation. University of Wales, Bangor, UK.

Uhl, C., Barreto, P., Veríssimo, A., Vidal, E., Amaral, P., Barros, A.C., Souja, C. Jr., Johns, J. & Gerwing, J. 1997. Natural resource management in the Brazilian Amazon: an integrated research approach. BioScience, 47: 160-168.

Van der Hout, P. 1999. Reduced impact logging in the tropical rain forest of Guyana: ecological, economic, and silvicultural consequences. Tropenbos-Guyana series 6. Tropenbos-Guyana Programme, Georgetown, Guyana.

Varian, H. 1984. Microeconomic analysis. W.W. Norton & Company, New York.

Williams, M. 1989. Americans and their forests: a historical geography. Cambridge University Press, New York.

Winkler, N. 1997. Environmentally sound forest harvesting: testing the applicability of the FAO model code in the Amazon in Brazil. Forest Harvesting Case Study 8. Food and Agriculture Organization of the United Nations, Rome.