SciELO - Scientific Electronic Library Online

vol.73 número4Dos especies nuevas dedicadas al Dr. Patricio Sánchez RBiodiversidad del complejo de artrópodos asociados al follaje de la vegetación del norte de Chile, II región índice de autoresíndice de materiabúsqueda de artículos
Home Pagelista alfabética de revistas  

Servicios Personalizados




Links relacionados


Revista chilena de historia natural

versión impresa ISSN 0716-078X

Rev. chil. hist. nat. v.73 n.4 Santiago dic. 2000 

Late Holocene vegetation dynamics and lake geochemistry at
Laguna Miranda, XI Region, Chile

Dinámica vegetaciónal y geoquímica lacustre del Holoceno tardío en
Laguna Miranda, XI Región, Chile


1School of Geography and Environmental Sciences, Monash University, Clayton, Victoria 3800
Australia, email:
2 Deceased (Department of Geography, Queen's University, Kingston, Ontario K7L 3N6, Canada).
3Quaternary Geology, Uppsala Universitet, Geocentrum, Villavägen 16, S-752 36 Uppsala, Sweden


Palynological and geochemical analysis of late Holocene lake sediments and dendrochronological analysis of Pilgerodendron in a volcanically active region of southern Chile reveal the long-term impact of a series of tephra fall events and tectonic activity on lake sedimentation and local vegetation. An upper 0,75 m core overlaps with a 4,35 m long Livingstone piston core to give a 4,60 m long sediment record, extending back to 4800 yr BP. Geochemical data shows the shift from allogenic dominance to authigenic and biogenic dominance as waterlogged soils developed within the catchment. This is presumed to have occurred under the influence of continued addition of nutrients to the catchment from tephra deposition and the associated high sedimentation rates. The palynological record from this site is dominated by Nothofagus dombeyi-type and Filicales. The most prominent changes in the pollen record, however, are the gradual decline in Podocarpus nubigena pollen throughout the last 4800 yr; the appearance and increase of Gramineae pollen during the past 2100 yr; and the rapid increase in the pollen of Pilgerodendron uviferum within the past 300-400 yr. Pilgerodendron tree-ring analysis and the pollen results over the last 400 years show that the most recent expansion of Pilgerodendron at the northern and eastern margins of this site is a response to either, periodic tectonic induced watertable changes, or is part of a long-term trend in gymnosperm growth around a shallowing lake margin. The decline of shade-intolerant trees such as Weinmannia and Podocarpus within a Nothofagus-rich forest community towards an increased presence of Pilgerodendron and Gramineae (likely Chusquea bamboo), that began over 2100 yr BP, may have been due as much to autogenic processes such as a change in the disturbance regime resulting in the development of waterlogged soils, rather than to an episode of climate change. It is important to recognise the potential effects of autogenic processes that may result from disturbance such as volcanic/tectonic activity in reconstructing past vegetation dynamics from pollen records.

Key words: Hudson Volcano, tephra, pollen, geochemistry, dendrochronology.


Análisis palinológicos y geoquímicos de sedimentos lacustres del holoceno tardío y análisis dendrocronológicos de Pilgerodendron uviferum en una zona volcánica activa del sur de Chile revelan el impacto a largo plazo de diversas caidas de tefra y actividad tectónica en la sedimentación lacustre y la vegetación de la zona. Una sección sedimentaria superior de 0,75 m traslapa una columna de pistón Livingstone de 4,35 m de largo para dar un registro de sedimentación de 4,60 m de longitud extendiéndose 4800 años AP. Datos geoquímicos muestran el cambio de dominancia alogénica a dominancia autígena y biogénica según se formaban los terrenos anegadizos dentro de la cuenca. Se presume que esto pudo producirse bajo la influencia de la llegada constante de nutrientes a la cuenca del lago procedentes del depósito de materias volcánicas conjuntamente con una alta sedimentación. El registro palinológico de este lugar está dominado por el tipo Nothofagus dombeyi y Filicales. Los cambios más notables que registra el polen son, no obstante, la declinación gradual del polen Podocarpus nubigena durante los últimos 4800 años; la aparición y aumento del polen Geamineae durante los pasados 2100 años; y el rápido aumento del polen Pilgerodendron uviferum en los últimos 300 a 400 años. El análisis de los anillos de leño de Pilgerodendron y los resultados que da el polen en los últimos 400 años muestran que la expansión más reciente de Pilgerodendron en las márgenes norte y este del lugar es una respuesta, bien a cambios periódicos de la capa freática por interacción tectónica, o es parte de un lento proceso de desarrollo de gimnospermas en las márgenes poco profundas del lago. La disminución de Weinmannia y Podocarpus, hace alrededor de 2100 años AP, se podría deber a procesos autogénicos resultando en el desarrollo de terrenos anegadizos, o a cambios en el régimen de perturbación, más bien que a un episodio de enfriamiento climático. En la reconstrucción de la dinámica de vegetaciones pasadas y partiendo de los registros del polen, es importante reconocer los efectos potenciales de los procesos autogénicos que pueden derivarse de perturbaciones tales como actividad volcánico/tectónica.

Palabras clave: volcán Hudson, tefra, polen, geoquímica, dendrocronológico.


The eruption of Hudson Volcano in the austral winter (August) of 1991 was the second most violent volcanic event in Chile this century. It produced large plumes of tephra, accompanied by SO2 gas emissions, that had a major environmental impact over an area of 150 000 km2 in Chile and Argentina (Naranjo et al. 1993). The impact of the tephra fall and associated tectonic disturbance on vegetation was widespread and led to mortality and severe damage to many species in the regional forest communities (Vogel 1996). While there have been a number of studies in Chile on the impact of volcanic eruptions on vegetation over time scales of years to decades (Kitzberger et al. 1995, Veblen et al. 1977, Vogel 1996), the consequences of volcanic and tectonic disturbance on vegetation dynamics over longer time scales (century to millennium) are not well understood.

The importance of large and small scale disturbance in the vegetation dynamics of southern Chilean forest communities is now well established (Veblen et al. 1996). Disturbance may vary from processes associated with large scale mass movement triggered by tectonic activity, volcanic eruptions, glacial processes, wind blowdown, and fire (natural or anthropogenic) to the finer-scale treefall gap processes. Climatic fluctuations also have an impact on vegetation, either directly through changes in parameters such as temperature and precipitation, or indirectly through the influence of these changes on the disturbance regime. That short-term fluctuations in climate are also important determinants of the disturbance regime has been demonstrated by recent ecological based studies that show population level dynamics being affected by annual to decadal scale climate change (Kitzberger et al. 1997, Szeicz 1997, Villalba & Veblen 1997). For example, forest dieback due to drought stress and fire, or the patchy destruction of canopy trees after blowdown events, create gaps and speed up regeneration dynamics. However, the relatively short time span involved in these ecological studies limits their insight into the actual long-term environmental variance that may affect many generations of trees.

Tephra layers preserved in lake or bog deposits provide clear evidence for past volcanic eruptions. This evidence can be used for stratigraphic correlation and chronological markers when reliably dated, and also provide the opportunity to investigate the effect of past eruptions on the environment and the pattern of ecosystem recovery. Surveys of lakes have shown many lake basins where tephras are preserved and that are suitable for palaeoenvironmental investigations (Haberle et al. in press). A multiproxy approach, using high-resolution palynology, dendrochronology and palaeolimnology records, has been employed to investigate the natural and human induced processes driving late Holocene vegetation change in southern South America at decadal to century time scales over the last millennium (Szeicz et al. 1998). Here, we extend this analysis of vegetation change to over the last 5 millennia, by focussing on the impact of successive volcanic eruptions on vegetation and lake sedimentation at a small lake basin in the vicinity of Hudson Volcano, as evidenced by pollen, geochemistry, and tephra analysis.

Study Area

Laguna Miranda (46°08'40"S, 73°26'40"W) is a small (0.5 ha), more or less ombrotrophic lake, lying at an altitude of 120 m above sea level in a remote forested region some 50 km southwest of Hudson Volcano and 65 km north of the San Rafael Glacier (Fig. 1). The site is about 0.5 km from the shoreline of a complex channel system, carved by extensive ice erosion from the Patagonian icefield that covered this region during the last glaciation (Clapperton & Sugden 1988). The bedrock of the region consists of granite of Cretaceous age (Niemeyer et al. 1984). Tectonic activity is centred around the Liquiñe-Ofqui fault system, with the main north-south aligned fault only 5 km to the west of Laguna Miranda. The site lies within a zone of high precipitation, produced by the coupled ocean-atmospheric influence of the Southern Polar Front, that migrates seasonally between 50°S (summer) and 40-45°S (winter). In general the climate is strongly oceanic, with annual precipitation in the region of 3000 mm and mean annual temperatures around 8-10°C. In the nearby Andes, rising to 4300 m altitude, precipitation is much higher with estimates for annual rainfall as much as 10,000 mm (Fujiyoshi et al. 1987).

Fig. 1. Location map of the Taitao-Coyhaique area, XI Region, Chile. Key to numbers representing sites referred to in the text: 1, Laguna Stibnite (Lumley & Switsur 1993); 2, Istmo de Ofqui I and II (Heusser 1964); 3, Río Témpanos (Heusser 1964); 4, Interfluctuational section (Heusser 1964); 5, Juncal Alto (Naranjo & Stern 1998). The Liquiñe-Ofqui fault system (LOFS) is shown as a dotted line.
Mapa del área de Taitao-Coyhaique, XI Región, Chile. La clave numérica de los lugares mencionados en el texto: 1, Laguna Stibnite (Lumley & Switsur 1993); 2, Istmo de Ofqui I y II (Heusser 1964); 3, Río Témpanos (Heusser 1964); 4, sección interfluctuante (Heusser 1964); 5, Juncal Alto (Naranjo & Stern 1998). El sistema de falla Liquiñe-Ofqui (LOFS) está representado con una línea de puntos.

The vegetation of the lake catchment is North Patagonian rain forest dominated by evergreen broadleaf and conifer taxa (Gajardo 1995, Veblen et al. 1983). The lowlands of this region support a dense forest of Nothofagus nitida, N. betuloides, Podocarpus nubigena, Weinmannia trichosperma, Drimys winteri, Caldcluvia paniculata, and Pseudopanax laetevirens (botanical nomenclature follows Marticorena & Quezada 1985). The understorey of these forests, particularly where light can easily penetrate, are frequently occupied by bamboo (Chusquea spp.). Poorly drained sites, including the waterlogged margins of Laguna Miranda, are dominated by Pilgerodendron uviferum and Tepualia stipularis. The branches of Nothofagus spp. are occasionally observed to be infested with the semiparasitic plant Misodendron spp. Potamogeton spp. was common around the margins of the lake.


Fieldwork was conducted in the southern hemisphere summer of 1995, with the logistic support of Raleigh International and CONAF. Parallel cores were taken from the centre of Laguna Miranda with a modified Livingstone corer. An undisturbed core of the uppermost sediment was obtained from the centre of the lake using a clear plastic piston corer. Each core was described in detail and scanned by a magnetic susceptibility core scanning loop at 10-20 mm intervals. This identifies changes in the abundance of ferrimagnetic minerals, which is used as an indication of changing erosional input to the lake, and to assist in the location of basaltic tephras. Three major tephra layers were located in this way and the glass shards extracted (H2O2 digestion) and mounted for electron microprobe analysis (EMPA, Department of Earth Sciences, University of Cambridge), following the guidelines set down by Froggatt (1992) and Hunt & Hill (1993). The amount of organic matter was determined by loss-on-ignition at 550°C. Chronological control is provided by using a combination of AMS/bulk radiometric 14C dating (Table 1) and the dates are reported throughout this paper as calibrated years before present (present is 1950 AD, Stuiver & Reimer 1993).

Geochemical analysis of lake sediments follows a fractional method introduced by Engstrom & Wright (1984) and modified by Lumley (1993), providing information on the chemical environment in the lake in relation to erosional inputs and redox conditions within the surrounding landscape (identifying the elements Mn, Fe, Al, Mg, Na, K, Ca, Ti, and Si). The wet-chemical extraction technique separates acid-soluble "authigenic fraction" and biogenic (diatom) silica from the clastic mineral "allogenic fraction". Element analysis of the separate fractions was carried out on an ICPES (Inductively Coupled Plasma Emission Spectrometer, University of London, Royal Holloway, UK).

Pollen analysis follows the standard acetolysis method described by Faegri & Iversen (1989). Pollen identification and nomenclature follows that set out in authored reference material (Heusser 1971, Villagran 1980, Zhou & Heusser 1996) and regional reference collections held at the Department of Plant Sciences, University of Cambridge (Appendix 1). Pollen counts are expressed as percentages of the total pollen sum (excluding pollen of aquatic vascular plants and all spores) which reaches a minimum of 300 in all samples. Charcoal particle concentrations were calculated following the point counting method outlined by Clark (1982). Charcoal analysis provides basic data on the abundance of charcoal in the sediment, from which the extent and intensity of fires around the site can be inferred. Numerical zonation and principal components analysis (PCA) were performed with only major taxa whose pollen or spore values exceeded 5% at least once. Numerical zonation employed optimal splitting by sum-of-squares analysis to partition the Laguna Miranda data into 3 significantly different pollen zones (Bennett 1996). PCA is used to reduce the pollen and spore data to a two-dimensional plot and the resulting data set is displayed as a biplot for samples and taxa (Birks and Gordon 1985). All numerical analyses, including zonation and principal component analysis (PCA), have been implemented within PSIMPOLL, a c program for plotting pollen data, developed by Bennett (1994).


Sediment, tephra analysis and chronology

The sediment stratigraphies for two cores from Laguna Miranda are shown in Fig. 2. The sediment is brown gyttja with thin green to grey inorganic inclusions. The upper 0.75 m core overlaps with the 4,35 m long Livingstone piston core to give a 4.60 m long sediment record at the site. Penetration of the core to greater depth was not possible due to increased sand (tephra) content at the base.

Fig. 2. a) Sediment characteristics, stratigraphic correlations between cores, and b) sediment chronology from Laguna Miranda.
Características sedimentarias, correlaciones estratigráficas entre columnas, y b) cronología sedimentaria de Laguna Miranda.

A total of 22 distinct inorganic layers, all containing volcanic glass shards, were identified as tephra deposited directly from atmospheric fallout, or from remobilised deposits within the lake catchment. Three tephra deposits, Mir-3 (olive silty sand, 15 mm), Mir-2 (olive silty sand, 8 mm) and Mir-1 (Olive grey silty sand, 80 mm) that gave strong magnetic susceptibility signals and were clearly visible in the core, were selected for tephra analysis. Geochemical compositions of these different tephra layers are shown in Appendix 2. The analysis show that the three tephras contain a suite of glass shards derived from both dacitic-andesite type eruptions (SiO2 content of 60-65%, and a high K2O content), and basaltic-andesite type eruptions (SiO2 of around 50-55 % and a high TiO2 content). Comparison with geochemical data from the nearest volcanic sources (Austral Volcanic Zone, 49°-55°S, and the Southern Southern Volcanic Zone, 41.5°-46°S, see Futa & Stern 1988, López-Escobar et al. 1993, Haberle & Lumley 1998, Stern 1990, 1991, Naranjo et al. 1993, Naranjo & Stern 1998) show that the distinct geochemical signature from Volcán Hudson eruptions (Fig 3a) is consistent with those obtained from tephras deposited in Laguna Miranda (Fig. 3b).

Fig. 3. a) SiO2 versus TiO2 and K2O for selected volcanic sources in Southern South America. Data for the Hudson, Maca, Cay volcanoes in the South Southern Volcanic Zone (SSVZ) and the Lautaro, Burney and Aguilera volcanoes in the Austral Volcanic Zone (AVZ) is from Futa & Stern (1988), Haberle & Lumley (1998), López-Escobar et al. (1993), Stern (1990, 1991) and Naranjo et al. (1993). b) SiO2 versus TiO2 and K2O for analysis on the three Laguna Miranda tephras compared to the tephra analysis envelope for Hudson Volcano (data given in Appendix 2).
a) SiO2 versus TiO2 y K2O por fuentes volcánicas seleccionadas de la parte meridional de Sudamérica. Los datos sobre los volcanes Hudson, Maca y Cay en el Zona Volcánica Sur Sur (ZVSS), y los volcanes Lautaro, Burney y Aguilera en la Zona Volcánica Austral (ZVA) son de Futa & Stern (1988), Haberle y Lumley (1998), López-Escobar et al. (1993), Stern (1990, 1991) y Naranjo et al. (1993). (b) SiO2 versus TiO2 y K2O por análisis de las tres tefras de Laguna Miranda comparadas con el análisis de tefra del Volcán Hudson (datos en el apéndice 2).

Radiocarbon dating indicates that average rates of sedimentation at Laguna Miranda were close to linear throughout the last 4800 years (Fig. 2). The temporal resolution of the cores varies from 9 to 14 year 10mm-1 depending on the depth. Table 2 lists the inferred age of inorganic (tephra) layers, showing that major phases with high frequency deposition events occur between 2800 - 3100 cal yr BP and 4400 - 4830 cal yr BP. The tephras identified at this site have not been dated directly, but there is a strong correspondence between the ages determined for two Hudson Volcano tephras recorded 70 km to the southwest in Laguna Stibnite (1690-1530 cal yr BP and 2790-2490 cal yr BP, Haberle and Lumley 1998), one recorded 80 km to the east of Hudson Volcano at Junco Alto (3885 cal yr BP, Naranjo & Stern 1998), and those recorded at Laguna Miranda (1695 cal yr BP, 2800 cal yr BP and 3820 cal yr BP). These results may provide additional chronological control when compared to regional tephrochronology being developed from lake sediment records (e.g., Haberle & Lumley 1998, Naranjo & Stern 1998).


Sediment chemistry (Fig. 4)

Authigenic fraction: Fe, Al, Ca and Si dominate the authigenic chemistry of the Laguna Miranda sediments (Fig. 4a). Ca and Mn concentrations are high between ca. 4800 and 4000 cal yr BP, though subsequently decreasing towards the top of the core. The Fe:Mn ratio shows a steady increase upwards through the core. i, Fe, Al and Si reach maximum concentrations between 4000 cal yr BP and 1200 cal yr BP, when there is a striking decrease in the concentrations of all these elements.This coincides with a decrease in the total authigenic content and an increase in sediment organic content, Mg and the Fe:Mn ratio.

Fig. 4. Sediment chemistry of Laguna Miranda samples against lake depth and inferred age (note independent scaling for each element). a) Authigenic fraction, b) Biogenic and Allogenic fraction.
Química sedimentaria de muestras de Laguna Miranda contra profundidad del lago y edad inferida obsérvese la escala independiente para cada elemento). a) Fracción autígena, b) Fracciones biogénicas y alogénicas.

Biogenic fraction: Biogenic Si shows a general trend of increasing concentrations towards the top of the core (Fig. 4b). Two periods in which biogenic Si increases markedly are ca. 2000 cal yr BP and after ca. 1200 cal yr BP.

Allogenic fraction:The allogenic fraction is dominated by Si, Al, Fe and Ca (Fig. 4b), which is similar to the elemental composition of the three tephra deposits. Total allogenic content is highest in the basal sediments, between ca. 4800 - 4000 cal yr BP, and higher in the core 1800 - 1200 cal yr BP. All elements follow a similar pattern of change as shown in the total allogenic curve.

Pollen, spores and microscopic charcoal (Fig. 5)

Zone 1 (ca 4800 - 2100 cal yr BP): Pollen spectra are dominated by Nothofagus dombeyi type (50-70%), Podocarpus nubigena (15-25%), and Filicales (30-40%). Podocarpus nubigena and Weinmannia trichosperma have relatively high percentages between ca. 4800 - 4000 cal yr BP, followed by a gradual decline towards the upper part of the core. Other tree taxa remain in consistently low abundances. The first appearance of Gramineae pollen occurs after ca. 2400 cal yr BP. The ratio between Misodendron and Nothofagus pollen shows only minor fluctuations. Microscopic charcoal is also low in abundance with only a slight rise between ca. 4000 - 3300 cal yr BP.

Fig. 5. Pollen percentage diagram of combined cores from Laguna Miranda against depth and inferred age. Pollen counts are expressed as percentages of the total pollen sum, excluding pollen of aquatic vascular plants and all spores. Selected taxa percentages are drawn with an exaggeration of x10 (single line). Nothofagus to Misodendron ratio as an indicator of relative Nothofagus infestation. Charcoal is expressed as a percentage of the total pollen sum. For explanation of sedimentary column see Fig. 2a.
El diagrama de porcentajes de polen de columnas combinados de Laguna Miranda contra profundidad y edad inferida. El conteo se expresa como porcentajes de la suma total de polen, excluyendo el polen de plantas acuáticas vasculares y todas las esporas. Los porcentajes de toxones seleccionados se sacan con una exageración de x10 (línea singular). La proporción de Nothofagus a Misodendron indica una infestación relativa de Nothofagus. El carbón vegetal se expresa como un porcentaje de la suma total de polen. Ver Fig. 2a para una explicación de la columna sedimentaria.

Zone 2 (ca 2100 - 200 cal yr BP): This zone is characterised by a rise in the percentage abundance of Gramineae. Cyperaceae and the aquatic, Potamogeton, increase after 1200 cal yr BP. Other taxa are relatively unchanged. The are slight increases in Pilgerodendron uviferum and Tepualia stipularis, and a decrease in Nothofagus dombeyi type towards the top of the zone. The ratio between Misodendron and Nothofagus continues to show only minor fluctuations. Microscopic charcoal remains very low.

Zone 3 (ca 200 cal yr BP to present): Pilgerodendron uviferum pollen becomes important in this zone. Tepualia stipularis, Maytenus and Lepidothamnus fonkii show slight increases in representation. Microscopic charcoal is very low.

Principal Components Analysis (PCA): The major patterns of variation in the pollen data were summarised by means of PCA, with pollen taxa accounting for significant sample variance displayed as a biplot (Fig. 6). The sample distribution shows a U-shaped shift through time showing the three phases of vegetation change: (i) early dominance of Nothofagus forest with Podocarpus, and to a lesser extent Weinmannia and Pilgerodendron important taxa, (ii) reduced influence from all forest taxa except Nothofagus under the influence of increased Gramineae, and (iii) the shift to Nothofagus forest including Pilgerodendron and Weinmannia as important components of the forest. High negative loadings for Nothofagus and Misodendron in axis 2 reflect changes in the forest dominant Nothofagus, and the close correspondence between abundance of the semiparasitic plant Misodendron and its host Nothofagus.

Fig. 6. Principal Components Analysis of pollen data from Laguna Miranda (taxa with relative abundance of <5% were excluded from the analysis). Eigenvalues are expressed as proportions of total variation with axis scale linear and plot centred on origin. The two dimensional representation accounts for around 82% of the original variability (Axis 1 = 62%, Axis 2 = 20%). The samples are identified according to their respective zone and the midpoint of samples from a single zone (mean of principal components) are joined up in stratigraphic order (thick stippled line). The directions of influence for each pollen taxa used in this analysis are overlain on the stratigraphic sample data in this diagram.
Análisis de los integrantes principales del polen de Laguna Miranda (taxones con una abundancia relativa de <5% no se incluyeron en el análisis). Los valores característicos se expresan como proporciones de variación total con escala axial lineal y plano bidimensional centrados en origen. La representación bidimensional supone alrededor del 82% de la variabilidad original (Eje 1 = 62%, Eje 2 = 20%). Las muestras se identifican según las zonas respectivas y el punto medio de muestras de una sola zona (media de componentes principales) se hallan unidos en correlación estratigráfica (línea de puntos gruesos). Las direcciones de influencia por cada taxón de polen usado en este análisis van sobrepuestas a los datos estratigráficos de la muestra en este diagrama.


The lake sediment record begins around 4800 cal yr BP at which time a North Patagonian rainforest, including a mixture of shade-tolerant and shade-intolerant species, was already established around the lake. Burning appears to have been minimal throughout the record. Between 4800 - 2100 cal yr BP very little change is evident in the forest composition, though there is a gradual decline in Podocarpus over this time and a reduction in the importance of Weinmannia trichosperma after 4000 cal yr BP. High allogenic inputs into the lake, particularly prior to 4000 cal yr BP, are probably derived from direct airfall of tephra and subsequent inwashing and reworking via overground water flow. Similarly, authigenic minerals formed in the waterlogged organic catchment soils, and mobilised by eluvial processes, are likely to have been derived from weathered products of older tephra deposits. The concentration of biogenic Si is at its lowest during this period, despite the high mineral inputs from tephra airfall events. Biogenic Si tends to be negatively correlated with total allogenic inputs suggesting the addition of tephra derived minerals resulted in an aquatic environment that was relatively poorly suited for the growth of diatoms and other aquatic organisms.

Between 2100 - 200 cal yr BP there are important changes in both the vegetation and geochemical record. Gramineae appears around 2400 cal yr BP and becomes increasingly important after 2100 yr BP. These changes may relate to the earlier rise in authigenic minerals, particularly Fe and the Fe:Mn ratio, indicative of reduced soil redox conditions during the development or expansion of waterlogged soils (Mackereth 1966, Pennington et al. 1972, Engstrom & Wright 1984). With continued addition of mineral nutrients to the environment from tephra fall and high levels of disturbance, there would have been an increase in the availability of sites for invasion. Nothofagus and Gramineae (likely the Chusquea bamboo) may have been best suited to take advantage of the disturbance regime and changing soil conditions.

The striking reduction in total authigenic elements in the sediment after 1200 cal yr BP coincides with a sharp reduction in the addition of tephra to the sediment. The increases in Cyperaceae and Potamogeton, around the edge of the lake, are indicative of progressively shallow lake margin and possibly marginal peat development. This is also supported by the rise in Fe:Mn ratio in response to a further reduction in soil redox conditions. The development of a lake margin peat, today occupied by Chusquea bamboo and the tree taxa Pilgerodendron and Tepualia, may also have contributed to the reduction in both total authigenic and allogenic elements by restricting mobility of inwashing allogenic and authigenic minerals. Dense growth of Chusquea bamboo may have acted as a filter to inwashing minerals and also utilised authigenic minerals in plant growth. Lake productivity increases during this period as biogenic Si concentrations reach a maximum. The last 200 cal years sees a rise in Pilgerodendron, presumably expanding around the lake margin as open sites become available for tree growth. There are no marked changes in sediment chemistry or microscopic charcoal to suggest a possible cause for this change. Szeicz (unpublished results) has examined this period in high resolution, comparing Pilgerodendron tree-rings analysis to determine stand age and pollen results over the last 400 years, and suggests that the most recent expansion of Pilgerodendron at the northern and eastern margins of this site may be a response to periodic tectonic induced watertable changes or is part of a long-term trend in gymnosperm growth around a shallowing lake margin. Fluctuations in pollen taxa percentages over this time may suggest that periodic disturbance still plays an important part in local vegetation dynamics, despite the lack of erosion or tephra evident in the lake sediment record.


This study has focussed on late Holocene environmental change recorded in a lake catchment with minimal human disturbance. Despite the long history of human occupation in southern Chile that reaches into at least the late glacial transition (Dillehay 1989) the influence of fire, which in this high rainfall environment is most likely associated with human activity, is negligible in the Laguna Miranda record (low charcoal levels throughout). The occupation of the channel islands by amerindians (Chono or Wayteca), whose subsistence was based primarily on fishing and coastal resource exploitation, appears to be relatively late, with sparse archaeological remains dating to no earlier than 5800 cal yr BP (Cardenas et al. 1993). Amerindian impact on this environment may have been focussed on the coastal margins and occasional excursions to the interior of the islands of the Taitao-Chonos region. Lake sediments under investigation from the northern Chonos islands show high levels of charcoal over at least the last 4000 years, suggesting much greater impact from human activity and possibly higher populations living in this environment relative to the Taitao Peninsula region (Haberle, unpublished results). Currently, human settlement is very sparse and clearance for grazing or agriculture is non-existent, though some areas have been burnt over the last 150 years for the purpose of timber extraction. So, what was the role of natural disturbance, in the absence of human activity, on the vegetation and sediment history of Laguna Miranda over time scales of several millennia?

Sources of instability in the catchment include volcanic and tectonic disturbance, changes due to disease, fire, and climate induced changes from drought and wind. The site is prone to earthquakes as it lies close to the main Liquiñe-Ofqui fault system, where evidence for uplift rates of around 10 m per 1000 years have been recorded around Chiloé Island for the late Holocene (Hervé & Ota 1993). Mass movement scars and drowned forests around Laguna San Rafael (Reed et al. 1988), and throughout south-central Chile (Veblen & Ashton 1978), attest to the capacity of tectonic movement to impact forests within the area. The absence of significant eroded soils in the Laguna Miranda sediment demonstrate that no widespread slope failures have taken place during the recorded history of the catchment.

The record of tephra layers found at Laguna Miranda represent only a partial record of the sequence of tephra eruptions at Hudson Volcano during the last 5 millennia. It is likely that there were additional eruptions of Hudson Volcano during this period that were not recorded in the Laguna Miranda sequence, either because the eruptions were small and deposition was not sufficient to be visible in this analysis, or because tephra dispersal was in a different direction during the eruption. The predominant winds in the region favour dispersal in an easterly direction, away from Laguna Miranda, so the tephras record at Laguna Miranda can only be considered to represent a minimum estimate of eruption frequency. Indeed, it is important to note that the lack of tephra layers after 1200 cal yr BP in Laguna Miranda may be a function of dominant dispersal direction rather than lower eruption frequency or magnitude.

What is the impact of tephra deposition on the vegetation around Laguna Miranda? Observations from the forested slopes of Hudson Volcano after the 1991 eruption showed that most of the plants in the Nothofagus forest community survived up to 100 mm thick tephra deposits, though with thicker deposits plant mortality increased (Vogel 1996). Complete destruction of the forest only appeared to occur when buried by more than 1,80 m of tephra. Most of the Laguna Miranda tephra layers are very thin (<80 mm), suggesting thickness of deposits in the catchment and impact from burial may have been minimal. The long-term shift to Nothofagus dominance from 4800 to 1200 cal yr BP is here considered to be a response to periodic volcanic disturbance, in which shade-intolerant Nothofagus species are best adapted. By contrast, if the absence of tephra layers after 1200 cal yr BP in the Laguna Miranda record is taken to indicate a period of relative volcanic, if not tectonic stability, then it is possible that the shift to a more mixed forest with a reduced Nothofagus component was the product of a long period of stability. This appears to be consistent with the response observed in other Chilean Nothofagus communities subjected to variable disturbance regimes (Veblen et al. 1996).

The addition of thin layers of tephra to the Laguna Miranda catchment may have caused damage to vegetation through mechanical and chemical processes, such as defoliation and subsequent prolonged absence of leaves (Rees 1979) or noxious gases and acid-bearing rains (Thorarinsson 1979). Fires associated with volcanic activity may have been a contributory factor to vegetation change, though there is no evidence to support this in the Laguna Miranda record. Disease and dieback are also known to affect Nothofagus forest structure in southern Chile, and in many cases large areas of canopy tree mortality have been observed (Veblen et al. 1996). The infestation of individual Nothofagus trees by the semi-parasitic plants Misodendron appears to have been a persistent feature in the Laguna Miranda catchment. Minor increases in Misodendron infestation may have been due to periodic disturbance events, such as tephra-tectonic activity, climatic variability, natural successional changes and human activity, that may have increased the susceptibility of individual Nothofagus trees to infestation for short periods of time.

Edaphic conditions may change with the addition of mineral nutrients and sand/silt sized particles after tephra deposition, altering the nutrient balance within the forest and influencing regeneration. Rapid rates of sedimentation of around 1 m per 1000 years and the development of waterlogged soils and peats around the margins of Laguna Miranda are likely to be a response to frequent renewal of nutrients from tephra deposition. Some support for this is evident in a comparison of sedimentation rates from a similar lake basin in a more remote region some 120 km from Hudson Volcano. The sedimentation rate at Laguna Stibnite on the Taitao Peninsula is around 28 cm per 1000 years, one quarter the rate recorded for Laguna Miranda (Lumley & Switsur 1993), which may reflect a relatively poor supply of nutrients to the catchments there.

Is there evidence that climatic fluctuations have influenced vegetation dynamics during the late Holocene? Previously investigated pollen profiles from the region have been interpreted to indicate vegetation responding to changes from warmer-drier to cool-moist conditions during the late Holocene (Heusser 1964, Markgraf 1989). Neoglacial advances have also been proposed for the southern Chilean region, culminating 5750-4450 cal yr BP, 3200-1950 cal yr BP, 1250-950 cal yr BP and 700-50 cal yr BP (Clapperton & Sugden 1988), and suggesting that climates where variable during the mid to late Holocene. The complicated dynamics of tide-water glaciers, such as the San Rafael Glacier some 50 km to the south of Laguna Miranda, make the interpretation of glacial fluctuations in the Laguna Miranda region highly problematic. In the Laguna San Rafael region, high abundance of Weinmannia in the lower levels of Ismo de Ofqui I site and the Interfluctuational section is considered to represent a warmer climate phase following neoglacial retreat from the sites (Heusser 1964). The sites are fragmentary and the dating is highly problematic (Clapperton & Sugden 1988); but, stratigraphic indications from each site point to disturbance, from both volcanic eruptions and nearby physical influences from glacial meltwater flooding, as possible significant factors in local vegetation change during the high Weinmannia phase. Weinmannia trichosperma is a relatively shade-intolerant species that is able to readily invade disturbed sites in much the same way as Nothofagus. Clearly caution must be exercised when inferring climate change from a single species such as Weinmannia, as it may be responding to changing disturbance regimes as much as climate variables.

Over shorter time scales of years to decades, climate variability may have been high during the late Holocene due to an increase in ENSO (El Niño-Southern Oscillation) climate events (Markgraf et al. 1992), though the southwestern region of Chile may not have been as strongly influenced by ENSO events as the subtropical-tropical regions (Ortlieb & Macharé 1993). The impact of these short-term fluctuations on vegetation are poorly understood, though over longer time scales the cumulative impact may be to increase the success of species adapted to frequent disturbance. There is no conclusive evidence of climatically induced forest change at Laguna Miranda. This raises the possibility that short-term climatic disturbances may produce similar responses in the pollen record as does volcanic or tectonic activity.


We would like to thank Raleigh International and CONAF for logistic support and permission to carry out this work in the XI Region, Chile. Many international venturers, including some 25 Chilean venturers, contributed outstandingly to the success of this work in difficult and remote terrain. he funding for this project was provided by the Leverhulme Foundation, UK, and the Queen's University Advisory Research Council, Canada.


BENNETT KD (1994) Confidence intervals for age estimates and deposition times in late-Quaternary sediment sequences. The Holocene 4: 337-348.         [ Links ]

BENNETT KD (1996) Determination of the number of zones in a biostratigraphical sequence. New Phytologist 132: 155-170.         [ Links ]

BIRKS HJB & AD GORDON (1985) Numerical methods in Quaternary pollen analysis. Academic Press, London. 317 pp.         [ Links ]

CARDENAS R, D VERA & CG HALL (1993) Los Chono y Los Veliche de Chiloe. Olimph, Santiago. 277 pp.         [ Links ]

CLAPPERTON CM & DE SUGDEN (1988) Holocene glacier fluctuations in South America and Antarctica. Quaternary Science Reviews 7: 185-198.         [ Links ]

CLARK RL (1982) Point count estimation of charcoal in pollen preparations and thin sections of sediments. Pollen et Spores 24: 523-535.         [ Links ]

DILLEHAY T (1989) Monte Verde: A Pleistocene Settlement in Chile. Smithsonian Institution, Washington D.C. xxiii + 306pp.         [ Links ]

ENGSTROM DR & HE WRIGHT Jr. (1984) Chemical stratigraphy of lake sediments as a record of environmental change. In: Haworth EY & JWG Lund (eds) Lake sediments and environmental history: 11-67. Leicester University Press, Leicester.         [ Links ]

FAEGRI K & J IVERSEN (1989) Textbook of Pollen Analysis, 4th Edition. John Wiley and Sons, Chichester. 328 pp.         [ Links ]

FROGGATT PC (1992) Standardisation of the chemical analysis of tephra deposits. Report on the ICCT working group. Quaternary International 13/14: 93-96.         [ Links ]

FUJIYOSHI Y, H KONDO, J INOUE & T YAMADA (1987) Characteristics of precipitation and vertical structure of air temperature in northern Patagonia. Bulletin of Glacial Research 4: 15-24.         [ Links ]

FUTA K & CR STERN (1988) Sr and Nd isotopic and trace element compositions of Quaternary volcanic centers of the southern Andes. Earth and Planetary Science Letters 88: 253-262.         [ Links ]

GAJARDO R (1995) La Vegetación Natural de Chile: Clasificación y Distribución Geográfica. Editorial Universitaria, Santiago de Chile. 165 pp.         [ Links ]

HABERLE SG & SH LUMLEY (1998) A postglacial eruption history of Hudson volcano, southwestern Chile. Journal of Volcanology and Geothermal Research 84: 239-256.         [ Links ]

HABERLE SG, SH LUMLEY, JM SZIECZ, R SWITSUR & KD BENNETT (in press) Late Quaternary environmental change in the Chonos Archipelago to Taitao Peninsula region of southern Chile: an overview of recent investigations. Revista Chilena de Historia Natural.         [ Links ]

HERVE F & Y OTA (1993) Fast Holocene uplift rates at the Andes of Chiloé, southern Chile. Revista Geológica de Chile 20: 15-23.         [ Links ]

HEUSSER CJ (1964) Some pollen profiles from Laguna de San Rafael area, Chile. In: Cranwell LM (ed) Ancient Pacific Floras: 95-114. University of Hawaii Press, Honolulu.         [ Links ]

HEUSSER CJ (1971) Pollen and Spores of Chile. Arizona Press, Tucson. xiv + 167pp.         [ Links ]

HUNT JB & PG HILL (1993) Tephra geochemistry: a discussion of some persistent analytical problems. The Holocene 3: 271-278.         [ Links ]

KITZBERGER T, TT VEBLEN & R VILLALBA (1997) Climatic influences on fire regimes along a rainforest-to-xeric woodland gradient in northern Patagonia, Argentina. Journal of Biogeography 24: 35-47.         [ Links ]

KITZBERGER T, TT VEBLEN & R VILLALBA (1995) Tectonic influences on tree growth in northern Patagonia, Argentina: the roles of substrate stability and climatic variation. Canadian Journal of Forest Research 25: 1684-1696.         [ Links ]

LOPEZ-ESCOBAR L, R KILIAN, PD KEMPTON & M TAGIRI (1993) Petrography and geochemistry of Quaternary rocks from the Southern Volcanic Zone of the Andes between 41°30' and 46°00'S, Chile. Revista Geológica de Chile 20: 33-55.         [ Links ]

LUMLEY SH (1993) Late Quaternary vegetational and environmental history of the Taitao Peninsula, Chile. Ph.D. Thesis, Cambridge University, Cambridge. 282 pp.         [ Links ]

LUMLEY SH & R SWITSUR (1993) Late Quaternary chronology of the Taitao Peninsula, southern Chile. Journal of Quaternary Science 8: 161-165.         [ Links ]

MACKERETH FJH (1966) Some chemical observations on post-glacial lake sediments. Philosophical Transactions of the Royal Society of London B 250: 165-213.         [ Links ]

MARKGRAF V (1989) Palaeoclimates in Central and South America since 18,000 BP based on pollen and lake-level records. Quaternary Science Reviews 8: 1-24.         [ Links ]

MARKGRAF V, JR DODSON, AP KERSHAW, MS MCGLONE & N NICHOLLS (1992) Evolution of late Pleistocene and Holocene climates in the circum-South Pacific land areas. Climate Dynamics 6: 193-211.         [ Links ]

MARTICORENA C & M QUEZADA (1985) Catálogo de la flora vascular de Chile. Gayana 42: 1-157.         [ Links ]

NARANJO JA, H MORENO & NG BANKS (1993) La erupción del volcan Hudson en 1991 (46oS), Región XI, Aisén, Chile. Servicio Nacional de Geología y Minería Boletín 44: 1-50.         [ Links ]

NARANJO JA & CR STERN (1998) Holocene explosive activity of Hudson Volcano, southern Andes. Bulletin of Volcanology 59: 291-306.         [ Links ]

NIEMEYER HR, JM SKARMETA, RP FUENZALIDA & WN ESPINOSA (1984) Hojas Península de Taitao y Puerto Aisén, Región Aisén del General Carlos Ibañez del Campo. Servicio Nacional de Geología y Minería, Carta Geólogica de Chile, No. 60-61, 80 pp.         [ Links ]

ORTLIEB L & J MACHARE (1993) Former El Niño events: records from western South America. Global and Planetary Change 7: 181-202.         [ Links ]

PENNINGTON W, EY HAWORTH, AP BONNY & JP LISHMAN (1972) Lake sediments in northern Scotland. Philosophical Transactions of the Royal Society of London B 264: 191-294.         [ Links ]

REED DJ, R MUIR-WOOD & J BEST (1988) Earthquakes, rivers and ice: Scientific research at Laguna San Rafael, southern Chile, 1986. The Geographical Journal 154: 392-405.         [ Links ]

REES JD (1979) Effects of the eruption of Parícutin volcano on landforms, vegetation, and human occupancy. In: Sheets PD & DK Grayson (eds) Volcanic activity and human ecology: 249-292, Academic Press, London.         [ Links ]

STERN CR (1990) Tephrochronology of southernmost Patagonia. National Geographic Research 6: 110-126.         [ Links ]

STERN CR (1991) Mid-Holocene tephra on Tierra del Fuego (54ºS) derived from the Hudson volcano (46ºS): Evidence for a large explosive eruption. Revista Geológica de Chile 18: 139-146.         [ Links ]

STUIVER M & PJ REIMER (1993) Extended 14C database and revised CALIB radiocarbon calibration program. Radiocarbon 35: 215-230.         [ Links ]

SZEICZ JM (1997) Growth trends and climatic sensitivity of trees in the North Patagonian rain forest of Chile. Canadian Journal of Forest Research 27: 1003-1014.         [ Links ]

SZEICZ JM, BA ZEEB, KD BENNETT & JP SMOL (1998) High-resolution palaeoecological analysis of recent disturbance in a southern Chilean Nothofagus forest. Journal of Paleolimnology 20: 235-252.         [ Links ]

THORARINSSON S (1979) On the damage caused by volcanic eruptions with special reference to tephra and gases. In: Sheets PD & DK Grayson (eds) Volcanic activity and human ecology: 125-160, Academic Press, London.         [ Links ]

VEBLEN TT & DH ASHTON (1978) Catastrophic influences on the vegetation of the Valdivian Andes, Chile. Vegetatio 36: 149-167.         [ Links ]

VEBLEN TT, DH ASHTON, FM SCHLEGEL & AT VEBLEN (1977) Plant succession in a timberline depressed by vulcanism in south-central Chile. Journal of Biogeography 4: 275-294.         [ Links ]

VEBLEN TT, C DONOSO, T KITZBERGER & AJ REBERTUS (1996) Ecology of southern Chilean and Argentinean Nothofagus forests. In: Veblen TT, RS Hill & J Read (eds) The ecology and biogeography of Nothofagus forests: 293-353. Yale University Press, New Haven.         [ Links ]

VEBLEN TT, FM SCHLEGEL & JV OLTREMARI (1983) Temperate broad-leaved evergreen forests of South America. In: Ovington JD (ed) Temperate Broad-leaved Evergreen Forests: 5-31. Elsevier, Amsterdam.         [ Links ]

VILLAGRAN C (1980) Vegetationsgeschichtliche und pflanzensoiziologische Untersuchengen im Vicente Pérez Rosales Nationalpark, Chile. Dissertationes Botanicae 54: 1-165.         [ Links ]

VILLALBA R & TT VEBLEN (1997) Regional patterns of tree population age structures in Northern Patagonia: climatic and disturbance influences. Journal of Ecology 85: 113-124.         [ Links ]

VOGEL A (1996) Beobachtungen zur Regeneration der Vegetation nach Ascheeruption am Hudson-Vulkan im südlichen Chile. In: Holtmeier F-K (ed) Beiträge aus den Arbeitsgebieten am Institut für Landschaftsökologie, Arbeiten aus dem Institut für Landschaftsökologie 1: 3-11. Münster.         [ Links ]

ZHOU M & CJ HEUSSER (1996) Late-glacial palynology of the Myrtaceae of southern Chile. Review of Palaeobotany and Palynology 91: 283-315.         [ Links ]

Associate Editor J. Armesto
Received November 28,1997; accepted September 8, 1998


Abundances of major elements in glass shards extracted from Laguna Miranda tephra samples. Tephra glass geochemistry is expressed as percentage weight, normalised to 100 wt%. Analysis was performed at the Department of Earth Sciences, University of Cambridge, using a Cameca SX50 spectrometer microprobe, an electron microscope with three wave dispersive spectrometers and a LINK AN10000 energy dispersive spectrometer running PAP matrix correction software. The probe operated at 20kV, using a 10nA beam current and a 10µm defocused beam to minimise loss of Na and K (50s count time). A mixture of minerals, natural oxides and pure metals were employed as standards that were periodically checked in order to verify internal consistency of the results

Abundancia de elementos principales en fragmentos de cristales extraídos de las muestras de tefra de Laguna Miranda. La geoquímica del cristal de tefra se expresa como peso porcentual, con normalidad de 100 wt%. El análisis se realizó en el Departamento de Ciencias de la Tierra (Department of Earth Sciences), Universidad de Cambridge, usando un espectrómetro microsonda Cameca SX50, un microscopio electrónico con tres espectrómetros dispersores de ondas y un espectrómetro LINK AN1000 dispersor de energía utilizando software PAP de matriz correctora. La sonda operaba a 20kV, usando una corriente de haz de 10nA y un haz desviado de 10µm para minimizar pérdida de Na y K (50s tiempo controlado). Se empleó una mezcla de minerales, óxidos naturales y metales puros como patrón, verificándolos periódicamente a fin de comprobar la consistencia interna de los resultados

Creative Commons License Todo el contenido de esta revista, excepto dónde está identificado, está bajo una Licencia Creative Commons