Bosque, Vol. 23 N° 2, 2002, pp. 19-28

ARTICULOS

Thinning schedules to reduce risk of windthrow in Nothofagus pumilio forests of Patagonia, Argentina

Propuesta de aplicación de raleos fuertes para reducir los volteos de viento en bosques de Nothofagus pumilio de Patagonia, Argentina

PABLO PERI ^{1, 2}, GUILLERMO MARTINEZ PASTUR ³, RICARDO VUKASOVIC ⁴, BORIS DIAZ ², MARIA VANESSA LENCINAS ³, JUAN MANUEL CELLINI ³

¹ Instituto Nacional de Tecnología Agropecuaria, cc 332 (9400) Río Gallegos, Santa Cruz, Argentina.
² Universidad Nacional de la Patagonia Austral, Río Gallegos, Santa Cruz, Argentina.
³ Centro Austral de Investigaciones Científicas, Ushuaia, Tierra del Fuego, Argentina.
⁴ Consultora "Servicios Forestales", Río Grande, Tierra del Fuego, Argentina.
E-mail: cadicforestal@arnet.com.ar; cadicforestal@gmx.net

SUMMARY

A thinning method is proposed to reduce windthrow risk in a high quality site (SI₆₀= 23.2 m) of lenga (Nothofagus pumilio) forests. Sixty seven-year-old (base age at 1.3 m) even-aged stands were thinned in 1996. The method involved thinning from above to maintain the stand stability and proposed a gradual reduction of crown density increasing the growth of selected trees (target trees). The system was carried out by five successive operations that incorporated the technique of girdling follower trees. Two thinning levels were defined: at 872 stems/ha (light thinning treatment) and 489 stems/ha (heavy thinning treatment). Each treatment had an area of 1 hectare with six long-term plots. Four permanent plots were established as controls and were not thinned. The diameter, basal area and volume growth were recorded for all treatments, and mortality rate and crown class dynamics were measured in the control. Total stem volume over bark and merchantable volume by log type were collected from trees felled in the first thinning intervention. Six year results from this study indicated that the proposed thinning method could achieve maximum growth and simultaneously avoid large gaps between tree crowns that may induce windthrow.

Key words: growth, stability, yield, wind damage, silviculture.

RESUMEN

En este trabajo se propone y analiza un método de raleo para reducir el riesgo de volteos por viento en bosques jóvenes de lenga (Nothofagus pumilio) en un sitio de alta calidad (IS₆₀= 23,2 m). Los raleos se llevaron a cabo en un bosque puro y coetáneo de 67 años (edad al DAP), efectuándose las primeras intervenciones en 1996. Este método propone raleos sucesivos por lo bajo para mantener la estabilidad del rodal, a través de reducciones sucesivas de la densidad de copas y el crecimiento paulatino de los árboles remanentes seleccionados (árboles meta). El sistema se lleva a cabo en cinco etapas sucesivas (anillado del 50% de los árboles no seleccionados al año cero; extracción de los árboles anillados y anillado del restante 50% al año 3; extracción de los árboles anillados y anillado de los árboles padres remanentes al año 6; y extracción de los árboles padres al año 10). En el ensayo se definieron dos niveles de raleo a lo largo de dos hectáreas: un raleo suave con 872 ind/ha y un raleo fuerte con 489 ind/ha. A lo largo de las mismas se establecieron 6 parcelas permanentes, mientras que en sectores no raleados se establecieron 4 parcelas testigo. Anualmente se registraron los crecimientos en diámetro, área basal y volumen total, así como la mortalidad y la dinámica de copas en los tratamientos testigo. Por otra parte, los rendimientos de los productos de la primera intervención (año 3) fueron medidos y discriminados según tipo de productos. Los resultados de los primeros seis años de intervención de esta propuesta de raleo presentan altos crecimientos y rendimientos, evitando la apertura brusca del dosel que puede poner en peligro al rodal frente a un volteo de viento.

Palabras claves: crecimiento, estabilidad, rendimiento, daño por viento, silvicultura.

INTRODUCTION

Nothofagus pumilio (Poepp. et Endl.) Krasser (lenga) is the native species with the greatest economic importance in the southern areas of Argentina and Chile (from 36°50' to 50°00' S). Its timber quality is widely accepted in local and international markets (Rechene and Gonda 1992).

The stocking rate and spatial distribution of stems affects growth, development and structure of the future stand. A level of competition between individuals is essential for good tree form and straightness (Schmidt et al. 1979). Therefore, it is desirable to maintain a high stocking rate in young stands. Early thinning (stands under 50 years of age) requires pruning operations to maintain stem quality (Martínez Pastur et al. 2001). According to Martínez Pastur et al. (1997a, 1997b), the starting point of thinning applications without pruning ranged from 60 to 70 years, when the stand had reached 80% of its final height. The systematic practice of intermediate cuts, during the forest growth cycle, characterizes intensive forestry management (Hawley and Smith 1982). However, the successful application of silvicultural systems depends upon understanding the limiting factors that affect trees throughout the forest rotation. Disturbance agents need to be identified and studied in relation to silvicultural operations. In some areas of lenga forest, windthrow could be the most important factor leading to failure of silvicultural planning (Ruel 1995). Damage by wind and windthrow are common events in lenga forests (Rebertus et al. 1997), and affect several hectares. For example, in summer 1998, nearly 2000 ha of lenga forest was blow down along the Tierra del Fuego (Argentina). Windthrow mainly occurs in young stands with a high height/diameter ratio (H/D). For example, trees of 20 cm quadratic mean diameter (QMD) growing in site class I have a total height of 26 m (H/D = 130) and consequently are more susceptible to windthrow than trees with the same QMD growing site class V trees (12 m of total height and H/D = 60). The severity of windthrow is also dependent on the quality of the site, increasing in wet or shallow soils, places exposed to wind and in overstocked stands.

Thinning can change the turbulence of airflow over the stand, increasing wind force on the trees. Usually, thinning from below is suggested to achieve an optimum final growth phase in the stand. However, to avoid windthrow in susceptible areas, the number of trees removed by thinning is reduced and consequently the full harvested yield potential of the site is not achieved. An alternative is to girdle (ring bark) the follower trees in several stages. In this way, the growth of the selected (target) trees is increased, by capturing the growth potential of the site, while the stability of the stand is maintained. While the girdled trees are dying the selected trees develop their crown, reaching larger diameters. Also, the girdled trees can add greatly to the structural diversity of bird fauna habitats. This modified silvicultural method may maintain the heterogeneity in the managed forest by simulating natural stand dynamics, and may create new interrelations between fauna and the environment. Many antecedents of intermediate treatment schedules (Rechene and Gonda 1992, Fernández et al. 1997) and thinning trials for Nothofagus pumilio were reported (Groose 1987, 1989, Donoso 1988, Espinosa et al. 1988, Nuñez and Vera 1992, Rubilar 1992, Donoso et al. 1993, Schmidt and Caldentey 1994, Schmidt et al. 1995, 1996, Peri et al. 2000, Martínez Pastur et al. 2001, Monelos et al. 2001, Donoso and Nirkos Gutiérrez 2001), but there are no antecedents of thinning schedules to avoid windthrow problems. Therefore, the objective of this work was to evaluate the proposal of two "high stability" thinning levels on the forest structure and growth of a pure second growth stand of Nothofagus pumilio at a site of high quality in Tierra del Fuego (Argentina), and to quantify the merchantable volume by log types of the first stage of the study.

MATERIALS AND METHODS

Study Area. The area is located in the San Justo ranch (54 05'S, 68 37'W) in Tierra del Fuego province, southern Patagonia, Argentina (figure 1). Treatments were applied in two hectares of a homogeneous, even-aged, 67-year-old second growth lenga stand. The site index in the study area was 23.2 m at base age of 60 year (Martínez Pastur et al. 1997b). The Bitterlich method (K=7.5) was used to describe the general characteristics of the stand from 20 plots homogeneously distributed (table1).

Figure 1: Location of the study stand in Tierra del Fuego (Argentina). Ubicación del rodal bajo ensayo en Tierra del Fuego (Argentina).

Alternative thinning schedules. The methodology proposed successive thinning from above to maintain stand stability. This schedule proposes a gradual reduction of crown densities facilitating growth of the trees selected to remain (target trees). The plan was carried out in five successive interventions (figure 2). The first operation was the girdling of half of all the non-target trees using a chainsaw (1-2 cm deep from the bark). The girdled trees died slowly over the three following years and consequently the target trees began to occupy the gaps that had been created. Secondly, in year 3, the girdled and dead trees were harvested as logs (diameter up to 25 cm and up to 3 m length), large poles or "postes" (18 to 25 cm diameter and 2.4 m length) and small poles or "varas" (up to 10 cm diameter and 4 m length). Thirdly, the second 50% of the follower trees were girdled at the same time that the first commercial thinning was carried out. Fourthly, the second commercial harvest will take place after another three years. Finally, seed trees will be girdled without harvesting (except for trees with merchantable value) to avoid damage to the remaining crop trees.

Treatments. Thinning levels were defined according to Fernández et al. (1997). There were two thinning treatments (figure 2): 800 stems/ha (light thinning) and 489 stems/ha (heavy thinning), each applied over a 1 hectare area. Within this hectare, six long-term plots of variable area (that included 15 trees) were established. Also, four permanent plots of 225 m² were installed in an area without thinning interventions (control). All trees were identified with numbered tags and two steel nails to mark the tape measurement position. Diameter and mortality rates were recorded for all treatments annually, and crown class dynamics were measured for the control plots twice. The basal area and total over bark volume growth were calculated using measurement of the diameter. The volume function used was defined by Peri et al. (1997). The merchantable volume by log types was measured from trees felled in the first thinning, as well as the tree mortality resulting from the logging. Statistical analysis was carried out by analysis of variance (F test), with means separation using Tukey's test at 5% probability.

Figure 2: Diagrams of the high stability thinning schedule. Diagramas de alternativas de raleos de alta estabilidad.

RESULTS AND DISCUSSION

The evolution of forest structure characteristics over six years is summarized in table 2. The lowest values of QMD (22-24 cm) were observed in the control treatment. The QMD increased after the selection of the target trees in the light thinning (24-26 cm) and heavy thinning (25-27 cm) treatments. The basal area of selected trees represented 67-70% of the original basal area in the light thinning treatment, while in the heavy thinning treatment they represented 38-41%. The control treatment had the highest stem volume (974-999 m³/ha). Because of the tree selection and increasing growth after six years from the first operation, trees in thinning treatments reached a greater diameter than the control. In the thinning treatments, mortality induced by competition was reduced to zero, while the mortality rate in the control treatment was 18-22 stems/ha.year. However, a mortality of 24-30 stems/ha was recorded for the thinning treatments due to logging operations. Windthrow damage may be reduced for the heavy thinning alternative where the H/D (height/diameter) ratio showed a lowest value (96-101) compared with the light thinning alternative (102) and the control (109-111), due principally to the diameter growth rate. However, although precise interpretation of an H/D ratio is difficult, because of interactions with soil and wind factors, there is general agreement that a value larger than 90 indicates a potentially unstable stand since the stems may not have sufficient strength to withstand the wind loading exerted on the crowns.

There were no significant differences between treatments for gross growth (table 3). The mean annual increment in QMD ranged from 0.2 cm/year (control and light thinning treatments) to 0.4 cm/year (heavy thinning treatment). The average over bark volume increment varied between 0.7 (heavy thinning treatment) and 4.7 m³/ha.year (light thinning treatment). Gross growth values showed high dispersion due to mortality. Therefore, net growth was analyzed. The control treatment had the lowest net QMD growth (0.18 cm/year), followed by the light (0.24 cm/year) and the heavy thinning (0.29 cm/year) treatments. The net total over bark volume growth was maximal in the control treatment (14 m³/ha.year), and decreased in both thinned treatments. This net volume growth represented 83% (light thinning) and 65% (heavy thinning) of the control. The total volume growth was similar to those found for other Nothofagus species, such as N. obliqua (Mirb.) Oerst., N. nervosa (Phil.) Dim. et Mil. and N. dombeyi (Mirb.) Oerst., with values up to 20 m³/ha.year in Valdivia (Chile) (Donoso et al. 1993) and up to 17 m³/ha.year in plantations in England (Tuley 1980). Similarly, N. pumilio stands for the initial growth phase (development phase classification proposed by Schmidt & Urzúa 1982) grown in site quality II-III had maximum growth of 12.7 m³/ha.year (Martínez Pastur et al. 2001). Schmidt et al. (1996) reported a growth rate of 8 m³/ha.year for thinned stands grown in a site quality of IV in the final growth phase in Chile (Magallanes). Also, these results were consistent with another Fagaceae species where Castanea sativa reached a growth of 13.5 m³/ha.year in the southern England and 16.0 m³/ha.year in the northern France (Everard and Christie 1995).

 

TABLE 2

Forest structure dynamics, H/D ratio and mortality rate for thinning treatments and control after six years. Dinámica de la estructura forestal, relación H/D y mortalidad para los tratamientos de raleo y el control a lo largo de 6 años.

QMD = quadratic mean diameter; BA = basal area; TOBV = total over bark volume; H/D = ratio between total height stand and QMD. Total data presented as average ± standard deviation. QMD = diámetro cuadrático medio; BA = área basal; TOBV = volumen total con corteza; H/D = relación entre la altura total del rodal y el QMD. Los datos totales son presentados como promedio ± desviación estándar.

Annual increments observed in QMD, basal area (BA) and total stem volume over bark (TOBV) increased over the studied growing periods (figure 3). This was mainly due to the time required for the thinning response. However, the annual increments for the light thinning treatment were similar between 2000-2001 and 2001-2002 periods indicating that potential growth may be achieved in this treatment. However under heavy thinning, annual increments continued to increase (figure3).

There was no significant difference between the two thinning alternatives in respect of merchantable volume by log types (table 4). The number of harvested trees averaged 667 trees/ha for the light thinning treatment and 740 trees/ha for the heavy thinning treatment. From these, the proportion of dead trees in light thinning was half (100 trees/ha) that of the heavy thinning treatment (200 trees/ha). The average QMD harvested was 19 cm for both treatments with a range of 10-35 cm. The total merchantable volume from the first thinning stage was 140 m³/ha and 128 m³/ha for the light and heavy thinning alternatives, respectively. The volume of small poles and number of small poles were slightly higher for light thinning treatment (55 m³/ha and 933 pieces/ha) than heavy thinning treatment (46 m³/ha and 820 pieces/ha). However, the opposite trend was manifested for large pole volume (56 m³/ha for light thinning treatment and 66 m³/ha for heavy thinning treatment). The pole volume from dead trees was nearly three times higher in light thinning treatment (14 m³/ha) than in heavy thinning treatment (5 m³/ha).

 

TABLE 3

Analysis of variance of annual increment (MAI) in diameter, basal area and total volume growth. Análisis de varianza de los crecimientos medios anuales (CMA) del diámetro, área basal y volumen total.

QMD = quadratic mean diameter; BA = basal area; TOBV = total over bark volume. Different letters in each column and for each factor show significant differences at P < 0.05 by Tukey's test. F values and significance: (A) Gross Growth: QMD = 2.51 (0.114); BA = 0.14 (0.868); TOVB = 0.13 (0.875); (B) Net Growth: QMD=3.55 (0.055); BA = 5.78 (0.014); TOBV = 6.03 (0.012). QMD = diámetro cuadrático medio; BA = área basal; TOBV = volumen total con corteza. Letras diferentes en cada columna para cada factor presenta diferencias significativas a P < 0,05 por el test de Tukey. Valores de F y significancia: (A) Crecimiento Bruto: QMD = 2,51 (0,114); BA=0,14 (0,868); TOVB = 0,13 (0,875); (B) Crecimiento Neto: QMD = 3,55 (0,055); BA=5,78 (0,014); TOBV = 6,03 (0,012).

Figure 3: Mean net annual increments for quadratic mean diameter (QMD), basal area (BA) and total over bark volume (TOBV). LT = light thinning; HT= heavy thinning. The bars represent the standard deviation for each sample. Incrementos medios anuales netos del diámetro cuadrático medio (QMD), área basal (BA) y volumen total con corteza (TOBV). LT = raleo suave; HT = raleo fuerte. Las barras representan la desviación estándar para cada muestra.

Merchantable volume by log types and growth rates obtained during the first stage of the proposed thinning schedule showed satisfactory results, indicating that the plan proposed in figure 1 can be continued over time. However, Thorsen and Helles (1998) reported that the optimal number of thinning operations increases and the optimal intensity of each thinning decreases when the risk of destruction by windthrow increases. This dynamic component indicates that adjustments may be needed in the future.

TABLE 4

Merchantable volume by log types for the first stage of the thinning considering the total production and the logged of live and dead trees. Rendimiento de la primera fase del raleo considerando la producción total y el aprovechamiento de árboles muertos y vivos.

Trees = harvested trees; QMD = quadratic mean diameter; BA = basal area; TOBV = total over bark volume; Vlogs = sawlog volume; Nlogs = sawlog number; Vpt = large pole volume; Npt = large pole number; Vp = small pole number; Np = small pole number. F values and significance for the analysis of variance considering the total production of the two treatments: Trees = 0.14 (0.723); QMD = 0.00 (0.957); BA = 0.08 (0.791); TOVB = 0.19 (0.681); Vlogs = 0.77 (0.414); Nlogs = 0.09 (0.769); Vpt = 0.27 (0.619); Npt = 1.07 (0.341); Vp =1.72 (0.238); Np = 0.46 (0.523). Trees = árboles cosechados; QMD = diámetro cuadrático medio; BA = área basal; TOBV = volumen total con corteza cosechado; Vlogs = volumen de trozas maderables; Nlogs = número de trozas maderables; Vpt = volumen de postes; Npt = número de postes; Vp = volumen de varas; Np=número de varas. Valores de F y significancia para el análisis de varianza considerando la producción total de los dos tratamientos: Trees=0,14 (0,723); QMD = 0,00 (0,957); BA = 0,08 (0,791); TOVB = 0,19 (0,681); Vlogs = 0,77 (0,414); Nlogs = 0,09 (0,769); Vpt = 0,27 (0,619); Npt =1,07 (0,341); Vp =1,72 (0,238); Np = 0,46 (0,523).

CONCLUSIONS

Results obtained over six-year period indicated that the proposed thinning method could achieve maximum growth for a high quality site, and simultaneously avoid large gaps between the crowns of neighbouring trees that may create windthrow problems. However, the implementation of this thinning system on a greater land scale needs to be analyzed over a longer time period incorporating the economic costs and benefits involved during its development.

ACKNOWLEDGEMENTS

To the sawmill "Los Castores", Universidad Nacional de la Patagonia Austral, Centro Austral de Investigaciones Científicas and Instituto Nacional de Tecnología Agropecuaria for institutional and financial support.

REFERENCES

DONOSO, P. 1988. "Caracterización y proposiciones silviculturales para renovales de roble (Nothofagus obliqua) y raulí (Nothofagus alpina) en el área de protección 'Radal - Siete Tazas'", Bosque 9 (2): 103-114.

DONOSO, P., T. MONFIL, L. OTERO, L. BARRALES. 1993. "Estudio de crecimiento de plantaciones y renovales manejados de especies nativas en el área andina de las provincias de Cautín y Valdivia", Ciencia e Investigación Forestal 7 (2): 253-287.

DONOSO, S., F. NIRKOS GUTIERREZ. 2001. Evaluación del crecimiento en ensayos de plantación de Nothofagus obliqua y Nothofagus alpina y rendimiento volumétrico al aplicar intervenciones silvícolas. Informe de Avance. Departamento de Silvicultura, Universidad de Chile, Santiago, Chile. 22 p.

ESPINOSA, M., J. GARCIA, E. PEÑA. 1988. "Evaluación del crecimiento de una plantación de raulí (Nothofagus alpina) a los 34 años de edad", Agro-Ciencia 4 (1): 68-74.

EVERARD J., J.M. CHRISTIE. 1995. "Sweet chestnut: silviculture, timber quality and yield in the forest of Dean", Forestry 68 (2): 133-144.

FERNANDEZ, C., G. MARTINEZ PASTUR, P. PERI, R. VUKASOVIC. 1997. Thinning schedule for Nothofagus pumilio forest in Patagonia, Argentina. Proceedings XI World Forest Congress. Vol. 6.H (Ecoregional Review), Topic 38.1 (Boreal Forest) Antalya (Turkey). October 13-22.

GROOSE, H. 1987. "Desarrollo de renovales de raulí raleados", Ciencia e Investigación Forestal 1 (2): 31-43.

GROSSE, H. 1989. "Renovales de raulí, roble, coigüe y tepa. Expectativas de rendimiento", Ciencia e Investigación Forestal 3 (6): 37-72.

HAWLEY, R., D. SMITH. 1982. Silvicultura Práctica. Ed. Omega S.A. 544 pp.

MARTINEZ PASTUR, G., P. PERI, V. PIRIZ CARRILLO, S. VACCARO. 1997a. Crecimiento en altura como indicador del sitio en bosques secundarios de lenga. Proceedings XVIII Reunión Argentina de Ecología. Pp. 80. Buenos Aires (Argentina). April 21-23.

MARTINEZ PASTUR, G., P. PERI, R. VUKASOVIC, S. VACCARO, V. PIRIZ CARRILLO. 1997b. Site index equation for Nothofagus pumilio Patagonian forest. Phyton, Vol. 61 (1/2): 55-60.

MARTINEZ PASTUR, G., J. M. CELLINI, M. V. LENCINAS, R. VUKASOVIC, R. VICENTE, F. BERTOLAMI, J. GIUNCHI. 2001. "Modificación del crecimiento y de la calidad de fustes en un raleo fuerte de un rodal en fase de crecimiento óptimo inicial de Nothofagus pumilio (Poepp. et Endl.) Krasser", Ecología Austral 11: 95-104.

MONELOS, L., P. PERI, G. MARTINEZ PASTUR, B. DÍAZ. 2001. Influencia del dosel y los raleos en el crecimiento de la regeneración avanzada de bosques de Nothofagus pumilio en Santa Cruz (Argentina). Proceedings XX Reunión de la Asociación Argentina de Ecología. Pp. 165. Bariloche (Argentina). 23-27. April.

NUÑEZ, P., O. VERA. 1992. "Evaluación de intervenciones silvícolas en un renoval mixto de lenga (Nothofagus pumilio) y coihue (Nothofagus dombeyi), ubicado en la reserva forestal Coyaique, XI Región". CIEFAP. Publicación Técnica 8: 111-125.

PERI, P., G. Martinez Pastur, B. Diaz, F. Fucaraccio. 1997. Uso del índice de sitio para la construcción de ecuaciones estándar de volumen total de fuste para lenga (Nothofagus pumilio) en Patagonia Austral. Proceedings II Congreso Forestal Argentino y Latinoamericano. Bosques Nativos y Protección Ambiental. Pp. 309-316. Posadas (Argentina). August 13-15.

PERI, P., G. MARTINEZ PASTUR, L. MONELOS, B. DIAZ. 2000. Thinning on growth of Nothofagus pumilio regeneration phase in south Patagonia, Argentina. Proceedings III Southern Connection Congress. Pp. 55. Canterbury (New Zealand). 17-22 January.

RECHENE, D., H. GONDA. 1992. "Propuesta metodológica y plan de manejo preliminar para lenga en sitios de calidad media de Chubut". CIEFAP. Publicación Técnica 8: 126-159.

REBERTUS, A., T. KITZBERGER, T. VEBLEN, L. ROOVERS. 1997. "Blowdown history and landscape patterns in the Andes of Tierra del Fuego, Argentina", Ecology 78 (3): 678-692.

RUBILAR, J. 1992. Respuesta de la lenga (Nothofagus pumilio (Poepp et Endl) Krasser) ante intervenciones de raleo en la provincia de Ultima Esperanza, XII Región. Undergraduate Thesis. Universidad de Chile. 63 p.

RUEL, J. C. 1995. "Understanding windthrow: Silvicultural implications", The Forestry Chronicle 71 (4): 434-445.

SCHMIDT, H., R. IPINZA, L. VIAL. 1979. Regeneración en bosque nativo de raulí. I. Estudio bibliográfico. Investigación y desarrollo forestal. FO: DP/CHI/76/003. Documento de Trabajo N24. Santiago, Chile.

SCHMIDT, H., A. URZUA. 1982. Transformación y manejo de los bosques de lenga en Magallanes. Universidad de Chile. Ciencias Agrícolas 11. 62 p.

SCHMIDT, H., J. CALDENTEY. 1994. Apuntes del tercer curso de silvicultura de los bosques de lenga. CONAF - CORMA - Universidad de Chile. Punta Arenas. 118 p.

SCHMIDT, H., J. CALDENTEY, S. DONOSO. 1995. Informe: Investigación sobre el manejo de la lenga - XII Región. Universidad de Chile - CONAF. 40 p.

SCHMIDT, H., J. CALDENTEY, S. DONOSO, K. PEÑA. 1996. Informe: Seguimiento forestal y ambiental del uso de los bosques de lenga - XII Región. Universidad de Chile - CONAF. 37 p.

THORSEN, B., F. HELLES. 1998. "Optimal stand management with endogenous risk of sudden destruction", Forest Ecology and Management 108: 287-299.

TULEY, G. 1980. Nothofagus in Britain. Forestry Commission. Forest Record 122. England. 26 p.

Recibido: 14.11.2000