Año 10. Nº 2.


El potencial para la regeneración de un bosque de lenga en la Patagonia chilena después de su destrucción por un incendio


Rich Howorth 1,2* & Anne-Marie Truscott 3

 1 Raleigh International, 27 Parsons Green Lane, London SW6 4HZ, UK

2 Address for correspondence: Sussex Wildlife Trust, Woods Mill, Henfield, West Sussex BN5 9SD, UK. Tel/Fax: 44 1273 694801 / 01273 494500 richhow@gmail.com

3 Centre for Ecology and Hydrology Banchory, Hill of Brathens, Banchory, Aberdeenshire, UK. AB31 4BW amtru@ceh.ac.uk



 Los bosques de lenga (Nothofagus pumilio) son un hábitat importante para la biodiversidad y la provisión de productos forestales y servicios ambientales, sin embargo,  a menudo muestra una recuperación limitada después de una perturbación antrópica. En este trabajo se evalua su potencial de regeneración  en distintos hábitat secundarios, luego de su destrucción por un incendio en un área protegida de XI Región, Chile. Se establecieron pares de parcelas para establecer la composición y la estructura de la vegetación para árboles > 3 cm DAP, plántulas de especies leñosas y las demás plantas vasculares,además de las características ambientales y evidencias de pastoreo.

La regeneración natural de lenga es limitada en todos los hábitat secundarios, salvo en el renoval en una franja adyacente al bosque maduro, y no se observó ninguna  regeneración en la pradera con árboles. Aunque la franja de regeneración tenía pocos árboles, se observó una alta densidad de plántulas de lenga. El bosque maduro no quemado tuvo una alta densidad, sin embargo, tuvo, a la vez, un reclutamiento limitado de plántulas. La plantación de lenga tuvo más árboles que el matorral adyacente con su regeneración natural, sin embargo ambos hábitat secundarios tuvieron pocas plántulas. Hubo una presión leve de pastoreo por huemul nativo y liebres introducidas sobre lenga en todos los hábitat. Se concluye que la regeneración post-incendio de lenga es limitada en todos los hábitat salvo los alrededores inmediatos del bosque maduro, principalmente debido a la disponibilidad de fuentes de semillas, aunque las condiciones ambiéntales, la competencia y presión de pastoreo también pueden tener un impacto.

 Palabras clave: área protegida, Nothofagus pumilio, pastoreo, regeneración de bosque post-incendio, Aysén.



 Lenga (Nothofagus pumilio) forest is a valuable habitat for biodiversity and provision of forest products and ‘environmental services’ yet displays often limited recovery following anthropogenic disturbance, hence its potential for regeneration in different secondary habitats after destruction by fire (about 35 years previously) was assessed in a protected area of Region XI Aisén, Chile. Plot pairs were established to measure vegetation composition and structure of trees > 3 cm dbh, woody saplings and seedlings and all vascular plants. Environmental characteristics and evidence of grazing was recorded, and fencing was installed to enable future assessment of grazing impacts.

Diversity of vascular plant species declined from the open secondary habitats to closed forest habitats. Natural regeneration of lenga was limited in all of the secondary habitats except for the forest-edge regeneration area, with no lenga regeneration evident in pasture grassland. The forest-edge regeneration contained few trees but had a high abundance of lenga seedlings. The unburnt ‘control’ stand of mature forest had a high density of trees, with a large size range, but more limited recruitment of seedlings due to shade. The lenga plantation had more trees than the adjacent matorral’s natural regeneration, yet both had low numbers of lenga seedlings. Grazing of lenga by native huemul deer and introduced hares was light in all habitats. Post-fire regeneration of lenga is concluded to be limited or absent in all habitats except for the immediate forest surrounds, principally due to limited seed availability, although environmental conditions, competition and grazing pressure may have an impact.

Keywords: grazing, Nothofagus pumilio, post-fire forest regeneration, protected area, Aysén.



Large areas of Nothofagus pumilio (Poepp. & Endl.) Krasser) (Fagaceae) forests have been cleared using fire in southern Chile in the last one hundred years. Due to their value for wildlife, economic exploitation and provision of ‘environmental services’ such as watershed protection, it is important to understand their regeneration potential.
Forests dominated by N. pumilio, hereafter referred to by its common name of “lenga”, occur in southern temperate South America along the Andes mountain chain in Chile and Argentina between 36o 50’ and 56o S (Donoso 1981). In Region XI Aysén of Chilean Patagonia, pure lenga forests predominate towards the eastern continental interior in lower rainfall areas, typically in the altitudinal range 900-1500 m (Donoso 1981).
Lenga forests were subject to massive clearance in Aysén, particularly between 1920 and 1940, through uncontrolled fires set to claim territory and create grazing lands, reducing the estimated original 5 million hectares total forest area by half and resulting in widespread erosion (Veblen et al. 1995). Deciduous Nothofagus forests are still widespread in Aysén, many being exploited for timber and as sheltered grazing for livestock, but are often fragmented and few are designated as protected areas in Region XI (Lysenko & Bubb 2003).
Lenga is a long-lived ecological pioneer species whose growth and persistence in the forest landscape is favoured by disturbance, in common with other Nothofagus species (Veblen et al. 1996a). However, the occurrence of N. pumilio as monospecific forest stands at higher altitudes and latitudes, such as in Aysén where the climate is less favourable, is not dependent on such coarse-scale disturbance for regeneration (Pollmann and Veblen 2004). Lenga produces massive amounts of seed, 0.5-10 million seeds/ha. (Schlegel et al. 1979), especially in ‘mast’ years but does not maintain a soil seed bank, thus it relies on “seedling banks” which can survive for up to 25 years within forests awaiting regeneration opportunities (Cuevas & Arroyo 1999).
Lenga forest stand development corresponds to Oliver’s general model (Oliver 1981) of four broad stages: stand initiation, stem exclusion, understorey reinitiation, and old growth (Veblen 1985). Little regeneration occurs within lenga stands until the third developmental phase of "understorey reinitiation" (150-200 years of age), when treefall gaps become prevalent (Rebertus &Veblen 1993) enabling recruitment through development of the seedling bank (Veblen et al. 1996b). The patch size of even-aged regeneration cohorts varies from small treefall clearings to large areas greater than 1000 m2 (Veblen et al. 1995) resulting from coarse-scale disturbance such as wind blows. Stand turnover times are estimated to be in the order of 300-500 years (Veblen et al. 1995).
Lenga forest regeneration following destruction by fire is characterised by dense even-aged single-species stands of lenga where conditions are favourable. Regeneration can be variable and limited, however, especially on steep north-facing slopes, in exposed dry locations, where grazing pressure is high, and/or where soil nutrient availability is limiting (Veblen et al. 1996b).
Grazing is thought to be a key factor limiting lenga regeneration. In the study region, grazing is by both native vertebrate herbivores such as the huemul deer Hippocamelus bisulcus, guanaco Lama guanicoe, pudu deer Pudu pudu, and by introduced species including European hares Lepus capensis, red deer Cervus elaphus and domestic livestock. Domestic cattle Bos taurus have been present since the early 1900s in Aysén, hares are now widespread and red deer are progressively expanding from the north and east towards Coyhaique in Aysén (C. Saucedo pers. comm.). Deer and livestock are considered to be the principal causes of forest degradation throughout Patagonia (Pastur et al. 2004), both being able to impede tree regeneration especially during the first "stand initiation" phase (Veblen et al. 1992). Deer browse more selectively than cattle, however, actually promoting the growth of woody vegetation (Relva & Veblen 1998) in contrast to cattle grazing (Raffaele & Veblen 2001). In Tierra del Fuego, it was estimated that lenga seedlings take 4-10 years to reach 1.3 m height in the absence of domestic livestock grazing, but suffer time delays of 15-24% at equivalent sites with livestock grazing (Pastur et al. 2004), whereas a different study concluded that cattle browsing of woody forest vegetation was relatively unimportant (Cavieres & Fajardo 2005). Both livestock and deer eat lenga and can inhibit regeneration of saplings, as well as causing an increase in the cover of spiny shrubs (e.g. Berberis spp.) and introduced species of grasses and herbs (Veblen et al. 1996b).
Few studies have been carried out on the capacity of lenga to regenerate in secondary successions following forest clearance. An understanding of the regeneration potential of lenga forests that have been burnt and are subject to grazing is crucial to inform present-day management for conservation and sustainable use, given the widespread occurrence of these disturbances across the forest landscape. The objective of this study is thus to provide an assessment of regeneration (recruitment and growth) potential of lenga (N. pumilio) evident in different secondary habitats (including a plantation area established for conservation management) of former lenga forest destroyed by fire that is grazed by wild and domestic herbivores. The study was carried out at one site in a national reserve, with the potential for extrapolation to other similar areas, and represents one snapshot in time in the absence of posterior monitoring data being available.




 Study site

The Reserva Nacional (RN) Lago Cochrane (Tamango) protected area (47o 12’ S, 72o 30’ W) in Región XI Aysén was chosen as a study site as it has been subjected to partial forest destruction by fire and grazing by livestock (horses and cattle) and wild herbivores, including the endangered huemul deer for which the reserve is famous (Foto 1)

A plot series was established in March-April 2002 at the south-west fringe of the reserve, in an altitudinal range of 560-930 m on predominantly south/south-east facing slopes of variable steepness (5-27.5o) (Figure 1). The volcanic ‘Divisadero’ geological formation here gives rise to thin, unstable acidic soils with a high vulnerability to erosion. The area’s climate consists of a period of vegetation growth of just 2-4 months in duration and a dry season of 2-6 months each year. Mean annual temperature in the nearby town of Cochrane is 7.6oC, varying between 1.5oC in July and 14.4oC in January. Mean annual precipitation is 805 mm, with a maximum in May and minimum in February, rising to a mean of 1000 mm in the higher parts of the reserve (Alvear and Urrutia 1986). The reserve’s vegetation is predominantly Patagonian-Andean forest, (Foto 2) largely deciduous lenga (N. pumilio) forest at middle elevations (shaded in Figure 1), grading in to steppe to the east. Other more restricted forest types occur including ñirre (Nothofagus antarctica) and coigüe (Nothofagus dombeyi) dominated forests. “Matorral” (heathland/shrubland), pasture grassland and wetland habitats are also present, the former two vegetation types typically resulting from secondary succession following fire and deforestation. In addition, a plantation of lenga saplings covering 91 hectares in area was established around 1986.

Subsequent to the considerable use of fire by indigenous peoples prior to european settlement, the reserve’s forests were further impacted by human-caused fires around the time of establishment of the town of Cochrane in 1947. The most recent fire in the study area section of the reserve destroyed an area of c. 800 hectares around 1965-1968 (SAG 1970), making the secondary habitats present c. 35 years old at the time of plot establishment. The study area was subsequently occupied by settlers and used for intensive grazing. Deforested areas in the reserve have suffered from severe soil erosion (Alvear and Urrutia 1986).

The RN Lago Cochrane protected area, covering an area of 8,532 hectares, was declared in 1967 to conserve and enhance the soils, hydrology, flora and fauna, and facilitate educational study and biological research. A management plan and zoning system exist for the reserve, the study area occurring within the “temporary zone of recovery” of forest following destruction by fire.

Sample plots

Four plot pairs, each consisting of two 20 x 20 m adjoining plots, were established in different secondary habitats within an area of former lenga forest destroyed by fire approximately 35 years previously, with an identical fifth plot pair installed in an area of “mature” lenga forest which had not been burnt (Foto 3 y (Foto 4). ). The plot pairs were located within areas of:

·        Pasture Grassland (up to 40 cm height) with scattered shrubs ("pradera arborescente") (up to 1.5 m height) - Plots 1 and 2 (Foto 5).

·        Matorral (heathland/shrubland) (50-70 cm height) with scattered mainly lenga tree saplings (1.5-4 m height) - Plots 3 and 4 (Foto 6)

·        Plantation of native lenga (N. pumilio) saplings (c. 2.5 m height), established on matorral (c. 50 cm height) around 1986 (16 years old) - Plots 5 and 6 (Foto7)

·        Lenga (N. pumilio) forest-edge regeneration (‘renoval’) (three apparent height layers at 0.25, 0.65 and 3-4 m), forest stand initiation phase - Plots 7 and 8 (Foto 8)

·        Mature lenga (N. pumilio) Forest (‘bosque adulto’) (three apparent height layers at 0.3-0.6, 2.5-3 and 7-8 m), forest stand ‘stem exclusion’ phase - Plots 9 and 10 (Foto 3 , (4 y 9

Environmental characteristics were recorded, comprising location, altitude, slope gradient, aspect, soil type, drainage, vegetation height and evidence of disturbance (Howorth & Truscott 2003). An assessment of grazing impacts was made by recording the percentage of saplings of each species that were grazed, the height that grazing occurred at (an index of apparency) and occurrence of animal dung.

Plot establishment and enumeration largely followed the procedures of Allen (1993) for permanent plot monitoring studies in New Zealand, enabling comparison with other southern hemisphere Nothofagus forests. The corner points of each plot were marked by wooden posts, with one plot of each pair being fenced using tensioned wire strands to exclude large vertebrate herbivores.

All live tree stems > 3 cm diameter at breast height (dbh) were measured, with the measurement taken at exactly 1.35 m height on the uphill side of the tree. A protruding nail was hammered in 10 cm above the point of measurement of each measured stem, and a unique number marked on an aluminium tag on the largest stem of each individual. The following characteristics were recorded: individual number, species name, sub-quadrat number, stem heights (estimated to the nearest half-metre), and Dawkin’s Crown Position Class (CPC). Dead trees were measured but not marked. Species nomenclature follows Marticorena & Quezada (1985) for plants and Muñoz & Yañez (2000) for mammals.

"Saplings" (tree stems > 1.35 m height but less than 3 cm dbh) and the smaller class of "seedlings" were identified and measured in each of sixteen 5 x 5 m sub-quadrats. All discernable stems of woody seedlings and saplings, hereafter referred to jointly as "seedlings" unless specifically differentiated, were individually counted by height class, except for very dense and/or multi-stemmed clumps for which a single maximum height was recorded. For seedlings < 15 cm height, presence only of seedling species was noted except for lenga whose individuals were counted.

In addition, a subplot comprising a circular wire mesh with a maximum circumference of 2.5 m (maximum area c. 0.5 m2) was established in the centre of each of two sub-quadrats within each plot and the percentage cover of each plant species, or group in the case of grasses, mosses and lichens, was recorded. The cover of bare ground, rock and litter was also assessed. The wire mesh subplot of one sub-quadrat only was then staked out at the base and left in place permanently marked to monitor the impact of smaller vertebrate herbivores (hares and rabbits principally) over time (Foto 10).

 Data analysis

The Shannon diversity index (H’) was calculated for each plot pair. Tree diameters and tree and sapling heights were assigned to classes for frequency analysis following Allen (1993). Lenga tree and sapling numbers, and stand basal areas, were calculated and extrapolated to hectare values, including a correction being made to account for the effect of slope. Differences in tree characteristics between habitat types were analysed by fitting general linear models (GLM). Habitat and plot were included as fixed effects. All analyses were carried out in SAS v.8.02 (SAS 1999).




Vegetation composition

Diversity of vascular plants was found to decline from open non-forest habitats to the closed lenga forest, being highest in matorral and lowest in the mature forest
(Table 1). A total of 35 species of vascular plants were identified across the plot series, nine of which were non-native plants almost exclusively of european origin. Whilst the non-native species were largely confined to non-forest habitats, especially the pasture grassland, two species also occurred in the lenga forest-edge regeneration area.

A small number of species tended to dominate in each habitat, according to vegetation cover assessment of subplots and the woody seedling species composition of sub-quadrats. Dominant species in the pasture grassland included the grass Poa sp. with 16% average cover and the introduced herb Achillea millefolium with 21% cover within the circular "hare mesh" subplots. The small woody shrub Gaultheria mucronata dominated the matorral ground layer under the plantation in particular (64% average cover), and was very abundant in all habitats except in the pasture grassland. The dominant woody shrub species of the non-forest habitats were Ribes spp., averaging >35% of all seedling stems in plot sub-quadrats, and Berberis microphylla (average >25%). Prevalent woody species of the forest habitats, in addition to lenga seedlings, included the field layer shrub Escallonia rubra which averaged >30% of seedling stems in both forest habitats, whereas Myoschilos oblongum averaged c. 30% in mature forest only. The mature lenga forest ground layer was characterised by the shrub Maytenus disticha with 19% average cover, and there was also large amounts of fallen wood and leaf litter.


Vegetation Physiognomy - Lenga Trees

All of the trees recorded in the plot series were lenga, except for one and two individuals of Embothrium coccineum in the matorral and plantation respectively, as well as a single individual of Baccharis patagonica in the plantation. Lenga trees of the mature forest were single-stemmed, whereas all the other habitats almost exclusively contained multi-stemmed younger individuals. The pasture grassland contained no trees.

Natural regeneration of lenga was much sparser in the matorral than the planted individuals in the adjacent plantation, equating to just 89 individual trees (rather than stems) per hectare versus 530 trees per hectare respectively. Lenga trees of the matorral were also smaller on average, and in maximum stem diameter (maximum 5.4 cm, versus 7.3 cm in the plantation) and height (maximum 3.5 m versus 4 m), as well as basal area (Table 1). The number of trees present in the forest-edge regeneration area was intermediate between the two aforementioned open habitats, with 352 individuals equivalent per hectare, but they were generally a little larger on average.

The size of lenga trees in the mature forest varied considerably between its two constituent plots, Plot 9 containing significantly larger trees (Table 1) as well as the greatest range of size classes. In contrast to the secondary habitats, in which most live tree stems occurred in the smallest diameter and height size classes suggesting that natural regeneration of lenga has been relatively recent at Tamango, the mature forest had the greatest number of stems in the larger diameter classes of 5-9.99 cm and 10-19.99 cm dbh (Plots 10 and 9 respectively) (Figure 2) and taller height class of 3.1-5 m. The dead standing trees within the mature forest were smaller than the live individuals, with the dead trees of Plot 9 again larger on average than Plot 10. In contrast, the burnt dead standing trees remaining in the matorral and forest-edge regeneration had a greater average size than the live trees of the contemporary mature forest ‘control’ stand (Table 1).

There were significant differences between habitats in the diameter of trees (GLM procedure, F3, 309= 4.9, p= 0.0023).  There were significant differences between habitats in the height of trees (GLM procedure, F3, 227= 35.6, p<0.0001) with significant differences between all habitat types (Tukey HSD test p<0.001). There was also a significant difference between plots (GLM procedure, F3, 227= 62.9, p<0.0001), with large differences between plot 9 and 10 in the mature lenga forest.

The trees of all habitats were growing in conditions of full sunlight (Dawkin’s Crown Position Class 4), except for four trees in the forest-edge regeneration area which were only partially illuminated (CPC 3) and the mature forest. Most trees in the mature forest were subject to some shade (average CPC 2.9 ± 0.1, n = 158), especially in the less developed Plot 10 where the great majority of individuals received only indirect illumination (CPC 2).

Vegetation Physiognomy - Seedlings (including Saplings)

Densities of seedlings (< 3 cm dbh) of all species of woody plants were greatest in the forest-edge regeneration area (Table 1). All other habitats had about half this density, with the lowest abundance being in the pasture grassland, with the exception of matorral Plot 3. There was considerable variation in seedling density between plot pairings in all habitats. Most seedling stems occurred in the second smallest height class of 16-45 cm in all habitats, except for in the pasture grassland and the lenga plantation where more stems were in the taller height classes (Figure 3). Very few saplings and taller seedlings >76 cm in height occurred in mature forest and forest-edge regeneration, or the matorral.

Only the forest-edge regeneration habitat contained a high abundance of lenga seedlings (about one-third of all woody stems), and here only Plot 8 had high individual numbers and relative abundance (Table 1). The patchy regeneration evident here was probably due to the presence of a relict surviving mature tree from the previous fires in Plot 8, perhaps combined with its gentler gradient. Lenga seedlings were absent from the grassland and present at low abundance only in the matorral and adjacent plantation (average 1.7% and 3.0% respectively). The mature forest contained fewer lenga  seedlings on average than the forest-edge regeneration. A broad spread of individuals across all height classes was evident in the forest-edge regeneration, and to some extent in the matorral too (Figure 4). The majority of lenga in the plantation were planted saplings (>135 cm height). The mature forest’s seedling banks had most lenga seedlings in the smallest size class of <15 cm and very few individuals taller than 45 cm in Plot 9, whereas its counterpart Plot 10 had taller lenga seedlings with most individuals in the 16-45 cm height range but relatively few taller than 75 cm. The predominance of very small seedlings in the former plot suggests that there has been little active growth of lenga compared to the less abundant but more heterogeneously-sized seedlings of its counterpart, possibly a consequence of its larger trees casting greater shade.

Grazing Pressure

Grazing was variable both in intensity and herbivores present, with most grazing occurring in the pasture grassland where the domestic livestock (cattle and horses) grazed heavily (Table 2). Grazing pressure was light in the secondary and forest habitats, and was largely by native huemul deer and non-native hare. Ribes spp. shrubs and grasses were particularly selected in the grassland and plantation, with saplings and small trees also targeted in the plantation and adjacent matorral. The dominant heath species of matorral, Gaultheria mucronata, was not grazed. Lenga was browsed in all habitats where it occurred, although at a low intensity of less than 5% of shoots in all of the plots. Most browsing occurred at less than 1 m height, and within the forest habitats it was concentrated at 40-50 cm height. There was also evidence of bark-stripping.  




Comparative Assessment of Lenga Secondary Succession & Forest Dynamics


Secondary lenga regeneration following fire is variable but generally relatively limited at Tamango, with recruitment and growth of N. pumilio found to be restricted in all secondary habitats of the study area except for the dense band of seedlings in the forest-edge regeneration stand which were distributed across a broad range of size classes. This stand is broadly consistent with other areas of lenga forest, its extrapolated average density of almost 13,000 stems/ha being comparable to the 14,000-17,000 individuals/ha found in a young lenga regeneration stand at Reserva Coyhaique (Vera 1985). Limited regeneration of lenga has been found at a number of other but not all post-fire sites studied. Veblen et al. (1996b) remarked that lenga regeneration in southern Patagonia following burning is typically restricted to a narrow 5-20 m wide band at the forest margin where conditions are amenable, with adequate seed sources and shelter. Where conditions are limiting to lenga growth, in contrast, few tree seedlings are evident in fire sites over 40 years old (Veblen et al. 1996b), with matorral often replacing lenga forest as the dominant vegetation type in burnt areas with limited mature lenga seed sources and subsequent very slow regeneration (Veblen et al. 2004). However, Pisano & Dimitri (1973) found abundant regeneration of lenga in burnt areas of P. N. Los Glaciares in Argentina. Anecdotal observations in Aysén suggest that post-fire secondary regeneration is far better for evergreen Nothofagus species and other species of lower altitudes and milder climates. For example, a 70 year-old post-fire regeneration mixed forest in Region X contained a high average density of 3819 stems/ha composed principally of coigüe Nothofagus dombeyi with limited N. pumilio, although lenga increased in dominance with age (Pollmann & Veblen 2004).

The total absence of lenga regeneration in pasture grassland at Tamango suggests that forest reversion is unlikely here under current conditions. The management intervention of the lenga plantation appears to have had some positive effect in promoting forest reestablishment, however, given the slightly better growth of N. pumilio trees in the plantation relative to the adjacent matorral. Survival of plantation trees have been high furthermore, given the occurrence of an average 2,034/ha. lenga trees and seedlings versus an original planted density of c. 2,500/ha. (Quijada 1999). Their growth has fallen within the range of other studies furthermore, the plantation demonstrating an average annual growth increment over the 16 year growth period of c. 0.28 cm/yr diameter and 19.4 cm/yr height. These values are within the range of 0.21-0.36 cm/yr diameter found at a denser 27 year-old mixed lenga-coigüe natural regeneration area at the Reserva Coyhaique (Vera 1985), and at the low end of a typical annual height gain range of 15-45 cm/yr (Provincia de Santa Cruz 2000). Lenga typically follows a sigmoid growth curve, however, in which increases are greatest following an initially slow juvenile stage (Provincia de Santa Cruz 2000), and thus future annual growth of the young lenga trees may be greater.

The adjacent mature forest stand that was sampled for comparison is perhaps less representative of lenga forest dynamics in general, its limited and patchy regeneration and recruitment of N. pumilio seedlings at an extrapolated density of 5,122 seedlings/ha lying at the low end of the recorded densities of 4,000 and 10,000 individuals/ha at sites in Aysén quoted by Donoso (1981), and being much less than the 100,000-plus seedlings/ha reported by Veblen et al. (1996b). The trees of the mature forest stand are relatively poorly developed compared to other studies also, with a high density of smaller than average trees, due to its apparent relatively young age. The extrapolated average abundance here of trees > 5 cm dbh of 1782/ha exceeds the typical range of 100-1600 trees/ha, whereas its average height of 7.5 m and average diameter of 15.5 cm dbh are considerably smaller than the typical ranges of 14-20 m height and 21-60 cm dbh diameter quoted by Donoso (1981). The stand’s calculated average basal area for trees > 5 cm dbh of 64.8 m2/ha lay between the extremes of 25m2/ha. and 103 m2/ha found in other studies (Donoso 1981, Schmidt &Urzúa 1982) however, and was close to the 62 m2/ha of Reserva Coyhaique (Fajardo & de Graaf 2004).

The mature forest stand is estimated to be just c. 65 years old, based upon the annual diameter gain of 0.20-0.28 cm found at a larger mixed-age (40-300 years old) lenga forest of Alto Mañihuales in Aysén (Álvarez & Grosse 1978). It appears to be at the second ‘stem exclusion’ phase of forest development, which occurs between 60-140 years of age (Provincia de Santa Cruz 2000), in which competitive self-thinning is prevalent (Donoso 1993), the likely explanation for the occurrence of numerous small dead standing lenga trees. The adjacent burnt former forest stands, in contrast, would have been more than 100 years old at the time of their destruction by fire in the late 1960s, based upon the larger remnant dead standing trees in contemporary non-forest plots.

Despite the sometimes substantial physiognomic differences between replicate plots of a habitat type, and the relatively limited number of plots sampled overall, the plot series appeared to adequately represent the composition and structure of each habitat and differences between them, especially with regard to lenga regeneration patterns. Bearing in mind these limitations, we suggest that cautious extrapolation to and comparison with other areas of secondary lenga forest natural regeneration is thus possible and that the conservation management approach of the native plantation merits further application.

 Potential controlling factors of lenga forest regeneration.

This study focussed on the capacity of lenga to regenerate in different secondary post-fire habitats, although it is highly likely that other external factors contribute to the marked differences in lenga regeneration detected. Although such factors were not empirically tested in this study, it is believed that the following five variables are fundamental in influencing lenga establishment and growth.


1. Seed dispersal relative to distance from mature trees/forest.

The distance of post-fire secondary habitats from mature lenga forest seed sources is likely to be a critical factor in determining forest regeneration at Tamango. Lenga seeds are principally wind-dispersed and generally do not travel more than the equivalent distance of one to two tree heights (Veblen et al. 1996b) and not more than 50 m distance (Provincia de Santa Cruz 2000). Indeed, most seeds fall directly beneath the parent tree’s crown (Rusch 1987). The prolific regeneration of lenga evident at the forest edge at Tamango would thus be the result of “seed rain” from the mature forest and the single surviving mature tree there. In contrast, seed limitation would have restricted natural regeneration of lenga in the three non-forest secondary habitats sampled which lay c. 400-600 m from the core mature forest block. The 16 year-old plantation contained only young immature trees, lacking viable seeds since lenga does not start to flower and fruit until 20-40 years of age (Provincia de Santa Cruz 2000). The limited capacity of N. pumilio for long-distance seed dispersal has been found also in N. dombeyi in Argentina (Veblen et al. 1996b) and N. menziesii in New Zealand (Wardle 1984).


2. Light availability.

 In common with other Nothofagus species, N. pumilio is a shade-intolerant species (Veblen et al. 1995) and thus requires relatively open areas with plentiful light for its growth. Whereas the partially shaded conditions of the study site’s dense mature forest stand would have limited growth of smaller individuals, the lenga trees in the secondary habitats experienced full light conditions. However, lenga regeneration in these habitats could have been restricted through shading and competition by taller seedlings of other species, such as in the scrub patches of the pasture grassland.


3. Inter-specific competition of vegetation.

As a monospecific forest type in Patagonia, lenga is only vulnerable to competitive inhibition or exclusion when young. The woody shrub species Ribes spp. and Berberis microphylla are the most likely competitors in the open secondary habitats of the plot series, most individuals being up to 75 cm in height in contrast to the main woody shrub species of the matorral Gaultheria mucronata which was generally < 45 cm tall. Growth of dense introduced grass species in the pasture grassland might also be significant, since tussock grassland has been found to impede colonisation of Nothofagus species tree seedlings in New Zealand (Wardle 1984). The forest-edge regeneration had a greater number of seedlings of other species than the open secondary habitats, 80% of which were < 75 cm height, yet demonstrated abundant N. pumilio growth, therefore suggesting that inter-specific competition is not the most important factor limiting its regeneration in secondary habitats. The impact of the dense mat of field layer shrubs in the mature forest (90% of which were < 45 cm height) on lenga seedling growth is uncertain.


4. Grazing pressure from vertebrate herbivores.

Vertebrate herbivore pressure from huemul deer and hares was found to be generally low on lenga in all habitats studied in 2002, nor was there any evidence in the plots’ woody vegetation of a cumulative inhibition effect over time. In contrast, high levels of grazing damage has been detected previously in the lenga plantation, with 42% of saplings affected by huemul and 2% by hares in 1997, thereafter impacting 15% and 19% of saplings respectively (Quijada 1999). R. N. Tamango contains the highest recorded density of huemul deer, a flagship species for conservation, although this equates to a low density of c. 0.03 deer/ha in the south-east part of the reserve (C. Saucedo pers. comm.). Definitive studies of maximum deer thresholds that allow Nothofagus forest regeneration are lacking in the southern hemisphere (Husheer & Frampton 2005). At Tamango, domestic livestock were mainly restricted to the pasture grassland so may have prevented lenga establishment there but were unlikely to have been influential in the other habitats. Lenga stems were also browsed by hares, a species known to target lenga seedlings especially during snow-covered conditions when other food is scarce (Veblen et al. 1996b). The establishment of plot fencing and hare mesh exclosures by this study will allow future evaluation of grazing pressure by different species.


5. Environmental conditions. A range of physical environmental factors influence lenga establishment and growth in secondary habitats, including aspect, gradient, exposure, altitude, water availability, drainage, and soil type (Veblen et al. 1996b, Provincia de Santa Cruz 2000). Influential factors at Tamango include the severely eroded degraded soils of deforested areas (Alvear & Urrutia 1986), where eroded convex slopes with steep gradients (> 25%) and shallow soils (< 35 cm deep) over 650 m altitude were found to most limit planted lenga sapling growth (Quijada 1999). The non-forest plots in this study were generally located on lesser gradients and altitudes, but were on south to east facing slopes, reducing their solar insolation and potentially exposing them to strong cold dessicating winds, thus reducing moisture availability.




We conclude that the fire 35 years previously that impacted the former lenga forest under study has resulted in restricted natural lenga regeneration within the non-forest secondary habitats. Significant post-fire lenga regeneration has been restricted to a narrow forest fringe, the proximity of mature lenga seed sources apparently being a critical factor. Natural lenga regeneration over extensive degraded burnt areas in the short term is thus unlikely. The management intervention of the lenga plantation has had some limited success by promoting relatively better growth than in the adjacent matorral. The influence of grazing animals on lenga regeneration appeared to be limited during our survey, although monitoring of the plots is required to assess the relative impact over time of domestic livestock, native huemul deer and introduced hares.

We recommend that woody vegetation growth in the plots should be re-measured at 5-yearly intervals, and the ‘hare mesh’ sub-plots monitored more frequently every 1-2 years, to assess regeneration and the relative influence of grazing animals. Research on effective restoration techniques, such as planting seedlings, distributing seeds and fencing marginal forest areas from large herbivores, is advocated to assist the reinstatement of lenga (N. pumilio) throughout its former range to benefit biodiversity and ecosystem functions.




We acknowledge the European Commission for financing this research under the Biodiversity Aysén Project, through its “Environment in developing countries” programme, and thank everybody involved in this project for their support. Plot establishment and recording was carried out by the dedicated efforts of Raleigh International volunteers. Extensive support was also received in the field from Alvaro Tomé (Universidad Central de Chile) and the CONAF park rangers at R. N. Tamango, especially Hernán Velasquez and Javier Subiabre. Patricio Contreras at CONAF provided support with GIS and map images. We also acknowledge the advice of Dr. Thomas T. Veblen of the University of Colorado and Professor David M. Moore (University of Reading).




 Allen, R. B. 1993. A permanent plot method for monitoring changes in indigenous forests. Manaaki Whenua Landcare Research New Zealand Ltd.

Álvarez, S. & H. Grosse. 1978. Antecedentes generales y análisis para el manejo de lenga (Nothofagus pumilio Poepp. et Endl. Krasser) en Alto Mañihuales, Aysén. Tesis de la Facultad de Ciencias Forestales, Universidad de Chile, Santiago.  

Alvear, N. & J. Urrutia. 1986. Guía de manejo Reserva Nacional Tamango. Documento de Trabajo No. 15, CONAF XI Región.

Cavieres, L. A., A. Fajardo. 2005. Browsing by guanaco (Lama guanicoe) on Nothofagus pumilio in Tierra del Fuego, Chile. Forest Ecology & Management 204: 237-248. 

Cuevas, J. G., M. T. K. Arroyo. 1999. Ausencia de banco de semillas persistente en Nothofagus pumilio Fagaceae en Tierra del Fuego, Chile. Revista Chilena de Historia Natural 72:73-82.

Donoso, C. 1981. Investigación y desarrollo forestal – Tipos forestales de los bosques nativos de Chile. Documento de Trabajo No. 38, CONAF & FAO.

Donoso, C. 1993. Bosques templados de Chile y Argentina – Variación, Estructura y Dinámica. Editorial Universitaria, Universidad de Chile, Santiago.

Fajardo, A. &  R. De Graaf. 2004. Tree dynamics in canopy gaps in old-growth forests of Nothofagus pumilio in Southern Chile. Plant Ecology 173: 95-105.

Howorth, R. & A. M. Truscott. 2003. Regeneration ecology of lenga, Nothofagus pumilio, forests in a protected area of Aysén (R.N. Tamango). Research report for EC Biodiversity Aysén project, London.

Husheer, S. W. & C. M. Frampton. 2005. Fallow deer impacts on Wakatipu beech forest. New Zealand Journal of Ecology 29 (1): 83-94.

Lysenko, I.& P. Bubb. 2003. Analysis of deciduous forest spatial integrity in the Aysén Region of Chile. UNEP-WCMC report to the EC Biodiversity Aysén Project, Cambridge, UK.

Marticorena, C. & M. Quezada. 1985. Catálogo de la flora vascular de Chile. Gayana Botánica 42:1-157.

Muñoz, A & J.Yánez. 2000. Mamíferos de Chile. CEA Ediciones, Valdivia.

Oliver, C. D. 1981. Forest development in North America following major disturbances. Forest Ecology and Management 3:153-68.

Pastur, G. M, M.V. Lencinas, R. Vukasovic, P. Peri, B. Diaz & J.M. Cellini. 2004. Turno de corta y posibilidad de los bosque de lenga (Nothofagus pumilio) en Tierra del Fuego (Argentina). Bosque 25 (1): 29-42.

Pisano, E. & M. Dimitri. 1973. Estudio ecológico de la región continental sur del área andino-patagónico. I. Anales Instituto Patagónico IV (1-3).

Pollmann, W & T. T. Veblen. 2004. Nothofagus regeneration dynamics in south-central Chile: a test of a general model. Ecological Monographs 74 (4): 615-634.

Provincia De Santa Cruz. 2000. Dinámica del bosque de lenga. Consulted 19th October 2005. Available at  URL:  www.scruz.gov.ar/recursos/flora/dinamica.htm

Quijada, C. 1999. Evaluación de crecimiento en una plantación joven de lenga (Nothofagus pumilio) en la Reserva Nacional Tamango, XI Región. Tesis de Grado, Universidad Austral de Chile, Valdivia.

Raffaele, E. & T. T. Veblen. 2001. Effects of cattle grazing on early postfire regeneration of matorral in northwest Patagonia, Argentina. Natural Areas Journal 21: 243-49.

Rebertus, A. J. &  T.T  Veblen. 1993. Structure and tree-fall gap dynamics of old-growth Nothofagus forests in Tierra del Fuego, Argentina. Journal of Vegetation Science 4: 641-654.

Relva, M. A. & T. T. Veblen. 1998. Impacts of large introduced herbivores on Austrocedrus chilensis forests in northern Patagonia, Argentina. Forest Ecology and Management 108: 27-40.

Rusch, V. 1987. Estudio sobre la regeneración de la lenga en la cuenca del río Manso Superior, río Negro. Informe no publicado, Consejo Nacional de Ciencia y Tecnología, Buenos Aires.

SAG (Servicio Agrícola Ganadero). 1970. Antecedentes básicos de la Reserva Forestal Cochrane. Informe de trabajo, Servicio Agrícola Ganadero Patrimonio Forestal, Cochrane.

SAS. 1999. SAS software. SAS, Cary, NC, USA.

Schlegel, F.M. & .T. T. Veblen, R Escobar. 1979. Estudio ecológico de la estructura, composición y semillación del bosque de lenga (Nothofagus pumilio) XI Región. Serie Técnica, Facultad de Ingeniera Forestal, Universidad Austral de Chile.

Schmidt, H. & A. Urzúa. 1982. Transformación y manejo de los bosques de lenga en Magallanes. Ciencias Agrícolas (Universidad de Chile, Santiago) 11: 1-62.

Veblen, T. T. 1985. Stand dynamics in Chilean Nothofagus forests. In Pickett STA, PS White eds. The Ecology of Natural Disturbance and Patch Dynamics. Academic Press, New York. p. 35-51.

Veblen, T. T. &  M. Mermoz, C. Martin & T. Kitzberger. 1992. Ecological impacts of introduced animals in Nahuel Huapi national park, Argentina. Conservation Biology 6 (1): 71-83.

Veblen, T. T., T. Kitzberger, B. Burns & A.Rebertus. 1995. Perturbaciones y dinámica de regeneración en bosques andinos del sur de Chile y Argentina. En Armesto J, C Villagran, M Arroyo eds. Ecologia de los bosques nativos de Chile. Editorial Universitaria, Universidad de Chile, Santiago. p. 169-198.

Veblen T., C. Donoso, T. Kitzberger & A. Rebertus. 1996a. Epilogue: commonalities and needs for future research.  In Veblen TT, RS Hill, J Read eds. The Ecology and Biogeography of Nothofagus Forests. Yale University Press, London. p. 387-397.

Veblen, T., & C. Donoso, T. Kitzberger & A Rebertus. 1996b. Ecology of Southern Chilean and Argentinean Nothofagus Forests.  In Veblen TT, RS Hill, J Read Eds. The Ecology and Biogeography of  Nothofagus Forests. Yale University Press, London. p. 293-353.

Veblen, T., T. Kitzberger & R.Villalba. 2004. Nuevos paradigmas en ecología y su influencia sobre el conocimiento de la dinámica de los bosques del sur de Argentina y Chile. In Frangi J. L., A Brown Eds. Ecología y manejo de los bosques de Argentina. Editorial de la Universidad Nacional de La Plata, La Plata, Argentina. p. 1-48.

Vera O. J. 1985. Evaluación de intervenciones silvícolas en un renoval mixto de lenga (Nothofagus pumilio) y coigüe (N. dombeyi) ubicado en la Reserva Forestal Coyhaique, XI Región. Informe Técnico, Universidad Austral de Chile.

Wardle J. A. 1984. The New Zealand Beeches: Ecology, utilisation and management. New Zealand Forest Service.

 Cite este trabajo como:

Howorth, R. & Truscott, A.M. 2007. Potential for lenga forest regeneration in Chilean Patagonia following historic destruction by fire.
Chloris Chilensis, Año 10 N° 2. URL: http://www.chlorischile.cl

 Volver a la portada