SciELO - Scientific Electronic Library Online

vol.64 número1Novedades en Solanum y Salpichroa (Solanaceae) para Sudamérica australActividad biológica del suelo en dos bosques de Nothofagus del centro sur de Chile índice de autoresíndice de materiabúsqueda de artículos
Home Pagelista alfabética de revistas  

Servicios Personalizados




Links relacionados


Gayana. Botánica

versión impresa ISSN 0016-5301versión On-line ISSN 0717-6643

Gayana Bot. v.64 n.1 Concepción jun. 2007 


Gayana Bot. 64(1): 60-80, 2007


Seed germination of species of the Juan Fernández archipiélago under laboratory conditions

Germinación de semillas de especies del Archipiélago de Juan Fernández en condiciones de laboratorio

Jaime G. Cuevas1,2 & Javier A. Figueroa3

1Corresponding author: Instituto de Silvicultura, Universidad Austral de Chile, Casilla 567, Valdivia, Chile.

2Fundación Senda Darwin, Santiago, Chile. Present address: Instituto de Investigaciones Agropecuarias (INIA), Quilamapu, Chillán, Chile.

3Escuela de Arquitectura y Paisaje, Universidad Central de Chile, Santiago, Chile.


The Juan Fernández Islands contain one of the world's most interesting floras. However, their conservation status is fragile due to direct and indirect human activities. With the aim of contributing to the conservation of the native species of these islands, germination trials have been performed under laboratory conditions. We studied the germination patterns and characteristics of 25 taxa of the archipelago, at temperatures between 15°C and 25°C, with 12.5 h. of photoperiod. Thirty-five percent of the studied species germinated immediately (within the first month after sowing), while 50% had a delayed germination. Three species did not germinate. Regarding the germination pattern, nine taxa (53%) germinated synchronously (³ 90% of seeds germinated within a month after the first germination), while the remaining 47% germinated asynchronously. Most species showed a high germinative potential (³50% of germination). Generally, the germination results were good for most species assayed, with few species requiring special treatments as scarification. For each species, the pre-germinative treatments, germination characteristics and patterns are detailed and discussed in relation to other published antecedents.

Keywords: Conservation, Chile, endemic, island, propagation.


Las Islas Juan Fernández contienen una de las floras más interesantes del mundo. Sin embargo, su estado de conservación es frágil debido a las actividades humanas directas e indirectas. Se realizaron ensayos de germinación bajo condiciones de laboratorio con el propósito de contribuir a la conservación de las especies nativas de las islas. Se estudiaron las características y patrones de germinación de 25 taxa del archipiélago, a temperaturas de entre 15°C y 25°C, con 12,5 h. de fotoperíodo. El treinta y cinco por ciento de las especies germinaron inmediatamente (dentro del primer mes de la siembra), mientras que el 50% de ellas presentaron una germinación retrasada. Tres especies no germinaron. Con respecto al patrón de germinación, nueve especies (53%) germinaron sincrónicamente (³ 90% de las semillas germinaron dentro del primer mes después de la primera germinación), por el contrario el 47% germinó asincrónicamente. La mayoría de las especies ensayadas presentaron un alto potencial germinativo (³ 50% de germinación). En términos generales, los resultados de germinación fueron buenos para la mayoría de las especies ensayadas, con pocas especies requiriendo tratamientos especiales tales como la escarificación. En el presente estudio se discuten con detalle los tratamientos pre-germinativos, las características y los patrones de germinación para cada una de las especies ensayadas, en relación con otros antecedentes publicados.

Palabras claves: Conservación, Chile, endémico, isla, propagación.



Germination requirements and characteristics constitute key factors that determine the distribu-tion of plant species in space and time (Venable & Lawlor 1980). Germination attributes of the plant species are correlated with climate, season, and habitat (Angevine & Chabot 1979). For example, at one end of the spectrum, seeds of many desert plants remain dormant while buried in the soil, for many years after dispersal, until rainfall is high enough to trigger germination (Venable & Lawlor 1980). At the other end of the spectrum, seeds of mangrove trees in the moist non-seasonal tropics often germinate inside the ripe fruit, while they are still on the mother plants (Van der Pijl 1982). Between both extremes, in strongly seasonal environments, for example the Mediterranean-type climate, seeds tend to exhibit a considerable delay of germination until environmental conditions are suitable for seedling establishment (Figueroa et al. 2004). In summary, the period of time that a seed remains viable in the soil, without germinating, is a critical component of the germination strategy of plants in seasonal environments (Figueroa & Armesto 2001).

Besides the theoretical importance of germination, this process is very important to the con-servation and reproduction of endangered species. Many plant conservation and restoration pro-grams have needed the knowledge of germination requirements in order to help fragile species to persist (Porteous 1993; Akeroyd 1995; Clemente & Hernández 1995). The flora of the Juan Fernández Islands provides an excellent example of the importance of this knowledge. Juan Fernández Islands are widely recognized as one of the most important centers of plant endemism in the world (Hulm & Thorsell 1995) with 62% of their native vascular flora being endemic to the islands. They are inhabited by 131 species, 12 genera, and one family endemic to the archipelago (Marticorena et al. 1998). However, their important botanical uniqueness contrasts greatly with their known conservation status, as 73% of the endemic angiosperms are threatened with extinc-tion (Stuessy et al. 1992). Thirty species, including various ferns, are in a critical state (P. Danton & J. Lesouef, personal communication 19991; Ricci 2006). Four centuries of human influence have severely impacted and altered the native ecosystem through the introduction of exotic animal and plant pests, anthropogenic burning, and overexploitation of natural resources (Sanders et al. 1982). Consequently, the Juan Fernández National Park is considered by the International Union for the Conservation of Nature (IUCN) to be one of the most endangered parks in the world (Hulm & Thorsell 1995). These facts have motivated the development of conservation programs, which partially consist of the collection of seeds from the field, germination under laboratory or greenhouse conditions, initial growth in greenhouses, hardening-off in nurseries, and planting out in the field with protection against predators and abiotic stresses (Cuevas & van Leersum 2001).

In spite of the great importance of the Juan Fernández Islands flora, there are surprisingly few studies that address key processes to plant conservation, as germination characteristics, re-productive biology, and authoecology (e.g., Ricci 1998, 2006; Bernardello et al. 1999; Vargas 2004). In fact, most studies have been carried out about floristic, evolutive and biogeographic topics (e.g., Skottsberg 1922; Sanders et al. 1987; Valdebenito et al. 1990; Castro et al. 2007). Although efforts for propagating endemic species of Juan Fernández have been carried out since 19762, there are few studies on germination patterns of this flora (Elorza 1984; Ricci 1998; Leiva, personal communication 20023). These trials were done under greenhouse or nursery conditions. Moreover, Cuevas (2002c) has developed trials under controlled laboratory conditions, which can allow the achievement of better germination rates, as well as better care of seedlings. All of these studies are internal reports not readily accessible to a wide audience. Therefore, this is the first publication concerning the species germination of the Juan Fernández flora under laboratory con-ditions, and it is based on the Cuevas' (2002c) work. This paper systematizes relevant informa-tion dispersed in several publications, besides contributing original data.

The objective of this study was to investigate the germination patterns and requirements of 25 native or exotic species of the Juan Fernández archipelago. We also report the details of germination, which will be useful for plant's conservation programs carried out on the islands and on the mainland (Cuevas & Van Leersum 2001).


Study sites

Juan Fernández archipelago lies in the south Pacific Ocean, between 670 and 860 km from the Chilean coast (Valparaíso). It consists of three islands: Robinson Crusoe, 47.9 km², 33°37'S, 78°51'W; Alejandro Selkirk, 49.5 km², 33°45'S, 80°45'W; and a third smaller island (Santa Clara, 2.2 km²), located 1.2 km southwest from Robinson Crusoe.

The climate of the Juan Fernández archipelago, according to Köppen, is Csb, temperate-warm and humid, with a short and relatively dry summer (Marzol et al. 1996-1997). The maritime influence is intense. In San Juan Bautista village (Robinson Crusoe Island, RCI), the mean annual temperature recorded at 30 m a.s.l. is 15.3°C (Di Castri & Hajek 1976). The extreme maximum temperatures have been 28.9 °C in December, and 21.1 °C in August; the absolute minimum was 3 °C for August (Marzol et al. 1996-1997). The monthly fluctuation in mean tem-perature is only 6 °C (Fig. 1). The mean annual precipitation is around 1,150 mm (di Castri & Hajek 1976). Almost half of the rainfall (44%) occurs in winter (May to August), whereas 9% occurs in summer (November to February) (Marzol et al. 2001) (Fig. 1). Most precipitation falls at medium altitudes on RCI: 400 m a.s.l. according to Cereceda et al. (1996) and 550-600 m ac-cording to Zunino (1993). Mean air relative humidity is 74%, approx. (Marzol et al. 1996-1997).

Figure 1. Ombrothermic diagram for Robinson Crusoe Island, constructed with temperature data from CONAF (unpublished data; 1957-1971, dashed line), and precipitation data from Marzol et al. (2001) for the period 1902-1994 (solid line).

Figura 1. Diagrama ombrotérmico para la Isla Robinson Crusoe, construido con información de temperatura de CONAF (información no publicada; 1957-1971, línea segmentada), e información de precipitación de Marzol et al. (2001) para el período 1902-1994 (línea continua).

Species selection

Native flora (endemic and non-endemic) of Juan Fernández Islands is composed of 211 taxa (Marticorena et al. 1998). Cuevas (2002c) worked with 39 taxa; these were selected mainly from those endemic species with conservation problems (Danton & Lesouef, unpublished data; Ricci 2006), although 15 less-endangered or other non-endemic species were also chosen. The final determinant of species selection was, of course, the availability of an adequate number of seeds to carry out germination trials.

However, we restricted the species set as follows: from the initial 39 taxa, two species were discarded (Robinsonia evenia and Rhaphithamnus venustus) because they did not have a registered date of seed collection (i.e., we did not know if they were fresh or old seeds); seeds of 12 other taxa were rather old when sown (over 40 days), and as they had poor or null germination, we also left them out of the studies; otherwise we could confuse slow germination, dormancy, and a low germinative potential with poor accession viability. Therefore, we retained 25 species, 15 of which had been sown within 40 days of the date of their collection. Seeds of another five species (Acaena argentea, Dendroseris litoralis, Drimys confertifolia, Libertia chilensis, and Plantago fernandezia, see Table I) were older, but they were also considered in this group because they had good germinative potential, except for D. confertifolia. Finally, five other species were analyzed separately, because other factors apart from seed age (e.g., lack of ripening, inbreeding, lack of fecundation, laboratory conditions, pre-germinative treatments) may have lowered their germination rates. Overall, our main analyses were done with a sample size of 20 species.

All species tested, with the exception of Gavilea insularis (exclusively from Alejandro Selkirk Island), are found on RCI. Several species are also shared between the two larger islands.

The conservation status of Juan Fernández species was derived from Ricci (2006), with the exception of Lactoris fernandeziana (after Danton & Lesouef, unpublished data).

Seed harvest

Ripe fruits were collected directly from the mother plants during the months of maximum fruit production, which were inferred from the phenological patterns that CONAF (the National Forest Service) has developed for most endemic species (Cuevas & Díaz 2000). A variable number and weight of fruits per species were collected, varying from grams to kilograms, depending on the requirements for germination assays in the Park's greenhouse and on the availability of propagules in the field. Seeds were cleaned of pulp and/or any attached structures before germination assays were performed. Seeds showing signs of pathogens or insect damage were discarded. Seeds collected before January 2001 were stored in paper bags, under greenhouse conditions (12-30 °C and 75% of mean relative humidity) (11 species, see Appendix I). After that date, a home refrigerator was purchased, allowing seeds of 13 other species could be stored under cold conditions (~4 °C) in relatively hermetic plastic bags with silica gel. Finally, the seeds of one species were stored under mixed conditions (i.e., first in the greenhouse, then in the cold chamber).

Germination assays

From the numerous laboratory assays carried out on RCI, we selected only those that were initiated 2-40 days (mean: 17 days) after seed harvest in the field, except for five species (Acaena argentea, Dendroseris litoralis, Drimys confertifolia, Libertia chilensis, and Plantago fernandezia). For each species assayed, generally three replicates, each one with 50 seeds, were used. However, due to the unequal availability of seeds in the field, replicates varied between one and six, and seeds from 12 to 60 per replicate. Notwithstanding, within the same assay the seed number per replicate was always the same. Seeds were sown between moist paper towels in Petri dishes, taking care that they were kept constantly moist. With five species, pre-germinative treatments were applied (see Results and Discussion). Seeds were incubated in a germination chamber (BIOREF model VV-11, P + L Electrónica S.A., Santiago, Chile) with constant daily temperatures of 15, 20 or 25 °C, depending on the species (see Table I and Appendices I, II). Temperatures for assays were not identical for all species, because the objective of the assays in Cuevas (2002c) was to find the best conditions for germination.

Photoperiod was 12.5-h light / 11.5-h dark. Seeds were irrigated periodically with distilled water; in some cases a solution of commercial fungicide was applied (CAPTAN 80 WP, BASF, 0.18% m/v). A seed was considered to have germinated when its radicle became visible through a binocular stereoscope. Germination was monitored every week until no additional seeds germinated. If most seeds (>70%) had germinated, the trials were discontinued 0-86 days after last germination as the remaining seeds were rotten; for medium germination (31-70%), that period was 21-174 days. For even lower rates (1-30%), trials were discontinued after 57-216 days. Finally, for the null germinations, the period of waiting was 238-495 days.


Taxa were categorized regarding their: a) Germinative response: this was defined as immediate if germination began before or at 31 days after sowing, as delayed if it began after 31 days, or as no germination. b) Pattern of germination, where a synchronous behavior corresponds to ³ 90% of seeds germinated within a month after the first germination; asynchronous means > 10% of seeds germinated in different months. c) Germinative potential: this was arbitrarily considered as high when maximum germination was ³50%; germination of 15-49% was moderate, and < 15% germination was considered low (Ray & Brown 1994). These categorizations are only intended for descriptive purposes, and not for hypothesis testing as in Figueroa et al. (1996).

Comparison with other antecedents

When it was possible, we used as references the works about greenhouse and nursery procedures by Elorza (1984), Ricci (1998) and Leiva (unpublished data) for species in the Juan Fernández archipelago. The most complete study is the work done by Ricci (1998), who utilized different pre-germinative treatments. On the other hand, Leiva did not carry out any pre-germinative treatments. Elorza (1984), by contrast, only determined the time between sowing and the moment of seedling emergence, but he also contributes some worthwhile recommen-dations about pre-germinative treatments. When the trials were carried out with several replicates and treatments, we compared the results obtained in our study with the averages for the treatment that had the best results. Sample size for Ricci's and Leiva's studies is 100 seeds, except when otherwise stated; sample sizes were not specified in Elorza (1984).

The comparisons with the above studies are only gross approximations, as all investigations used seeds collected on different dates, from different sites, perhaps differing in viability, storage conditions, date of sowing, etc. Moreover, in a strict sense, greenhouse and nursery assays determine the rate of seedling emergence above the ground level, which is lower than the rate of germination.


Germination patterns

Thirty-five percent of the studied taxa (Table I) germinated during the first month after sowing (e.g., Acaena argentea, Dendroseris litoralis, Erigeron fernandezianus), while 50% had a de-layed germination (e.g., Azara serrata var. fernandeziana, Juania australis, Machaerina scirpoidea). The species with the shortest time to first germination was D. litoralis (seven days), while the one with the longest time to first germination was Selkirkia berteroi (mean 173 days). Three species did not germinate (Colletia spartioides, Drimys confertifolia, and Gavilea insularis).

Regarding the pattern of germination, nine taxa (53%) germinated synchronously (e.g., Lactoris fernandeziana, M. scirpoidea, Plantago fernandezia), while the remaining 47% germinated asynchronously (> 10% of seeds germinated in different months; e.g., A. serrata var. fernandeziana, Centaurodendron dracaenoides, Cuminia eriantha var. eriantha) (Table I). The germinative potential was high for 11 taxa out of 17 (65%), with the highest percentages belonging to E. fernandezianus and Herbertia lahue (above 90%). Two taxa had a moderate germinative potential (C. dracaenoides, Ochagavia elegans, both ca. 45%) and four species had a low germinative potential (e.g., Coprosma oliveri, M. scirpoidea, less than 4% germination) (Table I).

Germination characteristics

Here we make some comments about the results of the germination trials of the species assayed. Only for some species, we have data on the effect on germination of different seasons of seed collection. When we compare with antecedents of the literature, to simplify notation we refer to the works of Elorza (1984), Ricci (1998) and Leiva (unpublished data) as simply Elorza, Ricci and Leiva, respectively. We give the detailed information of these assays in Appendix I.

Table I. Percentages of germination (mean ± 1 SE) in species of the Juan Fernández archipelago.
TCS: time between seed collection and sowing, TFG: Time of first germination, TLG: time of last germination. £Germinative response: I = immediate, germination began before or at 31 days after sowing; D = delayed, after 31 days; N = null germination. ¥Pattern of germination: S = synchronous, ³90% of seeds germinated within a month after the first germination; A = asynchronous, > 10% of seeds germinated in different months. §Germinative potential: high, H ³50%; moderate, M 15-49%; low, L < 15%.

Tabla I. Porcentajes de germinación (media ± 1 EE) de especies del Archipiélago de Juan Fernández.
TCS: tiempo entre la colecta de las semillas y la siembra, TFG: tiempo de la primera germinación, TLG: tiempo de la última germinación. £Respuesta germinativa: I = inmediata, la germinación comenzó antes de o a los 31 días después de la siembra; D = retardada, después de 31 días; N = germinación nula. ¥Patrón de germinación: S = sincrónico, ³90% de las semillas germinaron dentro de un mes después de la primera germinación; A = asincrónico, > 10% de las semillas germinaron en diferentes meses. §Potencial germinativo: alto, H³50%; moderado, M 15-49%; bajo, L < 15%.

Acaena argentea (Rosaceae). "Trun". It is a perennial herb, native to South America, but it was introduced to the Juan Fernández Islands, becoming one of the most powerful invaders on the archipelago (Greimler et al. 2002). It covers large extensions on the deforested valleys and slopes of RCI (Cuevas 2001). It makes the hikes difficult, because Acaena grows like deep accumulations of biomass; their seeds also stick easily to the clothes. However, the species is considered as appropriate for plantations in eroded sites, and it does not impose a severe danger for forest vegetation, because Acaena is a shade-intolerant species (Cuevas & Le Quesne 2006). Moreover, in the open sites > 70% of the vascular flora is exotic (Cuevas 2001, 2002a). Germination was very high (mean 86%, Fig. 2), in spite seeds were several months stored, but in cool conditions. The maximum germination was achieved the day 20 after sowing. With these antecedents, Acaena fulfill one of the requirements for a "good" weed: good germination (Montaldo 1995). This phe-nomenon is compounded by its vigorous clonal reproduction (Greimler et al. 2002), explaining perfectly why is a serious invader of the natural prairies and denuded areas of the islands.

Azara serrata var. fernandeziana (Flacour-tiaceae). Endemic tree, which is in a critical conservation status. Few individuals are seen on RCI in natural conditions. The best date of seed collection seems to be in the middle of May, as other dates produced less optimum results (data not shown). Ricci found that maximum emergence (19% out of 500 seeds) occurred on day 80; thus, this emergence is lower and faster than occurs under laboratory conditions (60% in 262 days, Fig. 2). Therefore, if the aim is to execute the trials in a short time, greenhouse conditions are preferred. On the contrary, if the seed provision is limited and one does not want to lose an appreciable quantity due to null-germination, laboratory conditions are preferred, in spite of the delayed and asynchronous germination.

Centaurodendron dracaenoides (Asteraceae). Endemic arborescent rosette in a critical danger of extinction. It grows on cliffs of difficult access in the highest zones of RCI. Knowledge of this species biology is very scanty. Seed germination was relatively low after five months of collec-tion, when seeds are stored at ambient temperature and humidity (Cuevas 2002c). Germination is much better with fresh seeds (mean 45%) (Fig. 2). Latter trials were ongoing when this report was completed. Ricci documented a maximum emergence of 9% at 74 days, very low as compared with data from the present study.

Colletia spartioides (Rhamnaceae). Endemic shrub in critical danger of extinction. Few indi-viduals are seen in the field. The reason for non-germination of this species is unknown.

Coprosma oliveri (Rubiaceae). "Olivillo". Small tree, endemic to RCI. It is in endangered status of conservation. In fact, we have observed its bark gnawed possibly by rabbits, causing the death of some trees. Germination was very exiguous (1%, approximately). There are no antecedents of germination to compare with this species (but see C. pyrifolia in "Other species").

Cuminia eriantha var. eriantha (Labiatae). Endemic tree in a critical danger of extinction. It belongs to an endemic genus, which grows in the highest mountains of RCI, above 500 m of altitude. Seeds of this species had to be separated from the berry remnants, which are very gelatinous. It was necessary to wash them with water and/or soap/detergent. Pulp remnants can be a source of fungal infections, making it necessary to irrigate with Captan fungicide. As the results were not optimal, seeds were soaked in water again during a week after the germination trial had started (between days 34 and 41). Other alternatives have been suggested for similar kinds of fruits, such as acid or enzymatic treatments, but these can harm seeds or require expensive laboratory equipment, respectively, and thus they were not used (see Ellis et al. 1985). In spite of this species delayed germination (Table I), in which the maximum germination was achieved 8-13 months after sowing, the delay is justified considering the high percentage of germination (mean 87%, Fig. 2).

Figure 2. Germination dynamics for eight species found on the Juan Fernández Islands. Dendroseris' curve was truncated as no further germination was observed after day 78.

Figura 2. Dinámica de germinación para ocho especies encontradas en las Islas Juan Fernández. La curva de Dendroseris fue truncada debido a que no se observó germinación adicional después del día 78.

Dendroseris litoralis (Asteraceae). "Col". Endemic arborescent rosette, in critical status of conservation. In natural conditions, there are only scant individuals, while in the town area about 1,000 individuals live, mostly planted by CONAF. D. litoralis flowers are also a source of nectar for the endemic hummingbird Sephanoides fernandensis (Roy et al. 1999). Ricci obtained 93% emergence in only 19 days, with eight-month-old seeds. In the present study, seeds germinated very quickly with a high percentage of germination (63%) (Fig. 2), but the seeds were very old when sown (980 days). Most of the storage occurred under ambient conditions (two years); the remaining time of storage was under cold conditions (ca. 4°C). We believe that if fresh seeds would have been used, a germinative capacity similar to that achieved in greenhouse trials would have been obtained. Overall, seed longevity, as well as the quick and high germination, explains why D. litoralis propagation is so easy; this has helped to save this species from an imminent extinction in the past.

Drimys confertifolia (Winteraceae). "Juan Fernández Canelo". Endemic tree, which is less en-dangered in comparison to other Juan Fernández species. Seedling banks of this species are fre-quently observed in RCI forests. This suggests that the propagation of Drimys is very easy. Regarding germination, this is not true, because this species could not be germinated during this investigation. Cuevas (2002c) also took soil from the forests where Drimys grows, to test whether a specific factor in the soil (i.e., mycorrhizae) was required for germination. After 268 days, germination was still unsuccessful. Elorza could not achieve germination by soaking seeds in water at 60°C, or by using detergent. He suggested that the species could present dormancy for lack of embryonic development, a problem that could be resolved by seed stratification. This suggestion has not been evaluated in Juan Fernández, but the congeneric D. winteri of the Chilean mainland is dispersed with its immature embryo (Cabello & Botti 1987), and seeds do not germinate without stratification and low temperature (Donoso & Cabello 1978). Stratification would be an exception for the Juan Fernández flora's behavior, because Ricci demonstrated that in most species this treatment delays or decreases emergence. Lastly, there is some evidence of dormancy break due to the D. confertifolia seeds' passage through the digestive tract of a bird: seedlings were grown from seeds found in feces of the austral thrush (Turdus falcklandii magellanicus) after six months of trials4.

Erigeron fernandezianus (Asteraceae). Endemic shrub. This species is not considered to be in danger of extinction. However, as it grows well in eroded habitats, Erigeron is considered a good species for revegetaty eroded soil. The best results (mean 92%) were obtained with recently collected seeds (Fig. 2). After 13 days of the start of the trial, temperature was changed from 20 to 25 °C, and this action probably stimulated germination, which began the day 17. Ricci obtained 84% emergence after 56 days, which is slightly less than that obtained in the present study.

Gavilea insularis (Orchidaceae). Endemic geophyte, the only orchid on the archipelago, which is in a critical status of conservation. Germination was unsuccessful. In the middle of the trial, temperature conditions were changed, due to the consideration that Gavilea grows in the alpine zone of the Alejandro Selkirk Island. Thus, it could require a period of cold stratification. This cold treatment was applied for one month, and the result was also unfruitful. According to Ellis et al. (1985), some orchids require mycorrhizal associations to germinate. The next step would be take soil from the ground around adult individuals, to bring it to the laboratory, and to carry out trials in conditions similar to those of Gavilea's habitat.

Herbertia lahue (Iridaceae). "Lahue". Perennial herb, native in mainland Chile and Argentina, where is in an endangered status of conservation (Hoffmann 1995). The first citation of this species in Juan Fernández is very recent (Danton et al. 1999), however, according to the islanders, Herbertia had been seen earlier. In fact, we found a specimen (SGO7394A) in the Herbarium of the Museum of Natural History of Santiago, which was very similar to Herbertia- bulb included - collected on 12.4.1965 by Muñoz and Sierra. The lack of flowers may have made its identification difficult, as neither Muñoz and Sierra, nor those who examined samples a posteriori determined the species. Moreover, its geophytic life form may have made it difficult for this species to be detected earlier by botanists working in Juan Fernández. It remains to be solved whether the species arrived naturally to the archipelago, or it was introduced with the help of the human beings. This species had a very high germination (90%) in a relatively short time (four months) (Fig. 2).

Juania australis (Palmae). "Palma chonta". Endemic tree, vulnerable to extinction. It is the only palm on the archipelago and one of the most emblematic plants for the islanders. It can grow up to 15 m in height (Rodríguez et al. 1983). It was necessary to remove completely the fruit envelopes (exocarp, mesocarp, and endocarp) due to an expected physical dormancy, either using sandpaper (Elorza 1984) or manually. Most of the other species assayed showed the typical sigmoid curve of germination (Ellis et al. 1985), but this species showed a stepped curve of germination (Fig. 2): the number of germinated seeds stays stable over periods of variable duration, and thereafter the number increases by one unit. This is the consequence of an overly frequent examination of Petri dishes compared to the slow rate of germination, as is typical for most palms (Ellis et al. 1985). This caused the Juania trials to take more than two years, a record for the species studied. The mean germination was high (50%) by our standards. In the middle of the trial, temperature was increased from 15 to 20 °C in order to stimulate germination, but this action did not result in a consistent change on the germination rate for the two replicates. Non-germinated seeds decomposed. Fungal infections can be controlled applying the fungicide Captan. Ricci obtained 76% emergence in 165 days, beginning at 97 days (this lapse similar to Elorza). Leiva, on the other hand, obtained 26% emergence, which occurred between days 112 and 336. To accelerate germination, Ellis et al. (1985) have recommended pre-germinative procedures for other palms. For the other Chilean endemic palm, Jubaea chilensis, Cabello (1990, 1999) recommended the hydration of seeds for several days, taking care to adequately oxygenate the water (Infante 1989; Anonymous 1999). Thereafter, seeds are sown in a medium with good aeration and humidity retention.

This procedure could decrease the germination time of Jubaea from two years and up to only 2-3 months (Anonymous 1999). More recently, Camus (2003) cultivated Juania embryos in vitro, achieving germination after day seven.

Lactoris fernandeziana (Lactoridaceae). Small shrub, which is endemic to RCI and it is in extinc-tion danger. Moreover, this species is the only member of its genus and family. The nearest known relatives of Lactoris occur as fossil pollen grains found in South Africa, Australia, Argentina, between others sites (Zavada & Benson 1987; Barreda, personal communication, 20065). In spite of its classification as endangered, its abundance is estimated as approx. 1,000 individuals distributed over several relatively inaccesible sites, mainly above 600 m of altitude (Bernardello et al. 1999; Ricci 2001). Germination was immediate and with a high germinative capacity (69%) (Fig. 3). Ricci (1998) obtained 77% emergence in 90 days, and 95% in 39 days (Ricci 2001). However, seedling survival is problematic. When they grow in Petri dishes with forest soil mixed with sand, and reach 2 cm in height, they lose vigor, showing a progressive rotting that moves upwards from the roots. Neither Captan application nor a diminution in soil moisture reversed this process. Finally, all seedlings died. Up to now, nobody has been successful in culturing this species to the adult stage (Stuessy et al. 1998). This result can be the consequence of specific requirements for nutrients, mycorrhizae or conditions for growth (Bernardello et al. 1999). Alternatively, they point out the possibility of endogamic depression, as this species is self-compatible. Future assays should consider soil and microclimate analyses of the local conditions where Lactoris grows (Ricci 2001). In vitro techniques could also help to improve the null survival of this species in culture.

Libertia chilensis (Iridaceae). Perennial herb, this native species is not endangered in Juan Fernández nor on the mainland. The main reason for its propagation was its ornamental value (attractive white flowers), being an especially appropriate species for the botanic garden of the Park Administration. Eventually, it could be interesting for islanders to plant out L. chilensis in their public and private gardens, leaving aside the exotic ornamental plants (Cuevas et al. 2004). The high percentage of germination (84%) justifies the relatively long period of the trial (mean 203 days) (Fig. 3). Seeds were relatively old when sown (ca. 10 months), but the cold storage probably contributed to the conservation of their viability.

Machaerina scirpoidea (Cyperaceae). Endemic perennial herb, vulnerable to extinction. This species grows in the cascades and other sectors with lower slopes in the streams of RCI. It has a low frequency of occurrence. Germination was very poor (3%, approximately) (Fig. 3). Some seeds were probably immature; investigators carrying out future trials should therefore consider seed collection in January or afterwards. This will be easy, because the individual plant that produced these seeds grows in the Park Administration's botanic garden.

Nassella laevissima (Poaceae). "Coirón". Perennial herb not endangered. This native species was propagated because it is a useful species for revegetaty eroded sectors. When cattle are ex-cluded, it can reach 40 cm in height. Unfortunately, germination was very poor (mean 4%) after a mean period of 160 days. The species has a deep dormancy, as the first seed germinated only five months after sowing (Fig. 3). Other conditions for germination should be investigated, such as those reported by Ellis et al. (1985) for other members of the Poaceae.

Ochagavia elegans (Bromeliaceae). Endemic perennial herb, which is not menaced to extinction. This species usually grows on the rocky cliffs covering large areas. Seeds were soaked in potable water for two days in order to clean pulp remnants. Germinative capacity was moderate (47%), but germination was immediate (Fig. 3). Ricci obtained an emergence of 19% in 100 days; therefore, laboratory trials produced better results than were achieved under greenhouse conditions.

Plantago fernandezia (Plantaginaceae). Endemic arborescent rosette, which is in critical danger of extinction. This rare woody species grows in the highlands of RCI, in places of difficult access. The best results (80% germination) were obtained with seeds collected in February, although they were stored more than eight months under ambient conditions (Fig. 3). Recently collected seeds (1 month old) also germinated immediately (data not shown). Ricci found an emergence of 95% in 46 days, a little better than that found in the present study.

Selkirkia berteroi (Boraginaceae). Endemic shrub, which is in danger of extinction. It is a rare species found above 500 m of altitude on RCI. Germinative capacity was low (11%) (Fig. 3). The species shows a deep dormancy, as the first seed germinated after about six months. As Selkirkia does not possess a hard or impermeable seed coat, a physiologic requirement of the seed probably delays its germination. These trials were ongoing when this report was completed.

Sophora fernandeziana (Papilionaceae). "Leña dura". This endemic tree is in a critical danger of extinction. It lives in sparse populations (5-10; Crawford et al. 2001, Ricci 2006). Ricci (2006) counted only 156 adult individuals, with an almost complete absence of seedlings and saplings. Seedlings are especially vulnerable to predation by rabbits (Simonetti & Acevedo 1990). As a pre-germinative treatment to remove physical dormancy, seeds were soaked in hot water (90°C, initial temperature) for 24 h., the water being allowed to cool to ambient temperature. This procedure softens the hard seed coat (Ricci 1998). When seeds are swollen, they are ready for sowing. Seeds germinated since the first month with a high germinative capacity (80%) (Fig. 3). Non-germinated seeds were usually rotten; either these could have been dead at the beginning of the trial, or they died during the trial. Thus, a periodic irrigation with Captan is necessary. Ricci found that treated seeds emerged since day 15, reaching a maximum on day 60 (92%). On the other hand, Leiva obtained only 28% emergence in 267 days. This could be due to Leiva's non-application of a pre-germinative treatment. Finally, Elorza points out that seeds emerged after 26 days, but he had previously soaked seeds in cold water.

Figure 3. Germination dynamics for other eight species found on the Juan Fernández Islands. Lactoris and Ochagavia's curves were truncated after the stabilization of the germination.

Figura 3. Dinámica de germinación para otras ocho especies encontradas en las Islas Juan Fernández. Las curvas de Lactoris y Ochagavia fueron truncadas después de la estabilización de la germinación.


Although seeds of some of these species were rather old when sown, other factors distinct from seed age may have decreased their germination rates. For detailed conditions of germination, see Appendix II.

Coprosma pyrifolia (Rubiaceae). "Peralillo". Endemic tree vulnerable to extinction. It is distrib-uted in some populations with relatively abundant individuals. The fruit must first be dried, changing its color from green to black. Thereafter, it is broken longitudinally in two halves, each one containing a seed (Elorza 1984), by means of scissors or a scalpel. To facilitate the radicle emergence, an operculum that covers the flat surface of the endocarp is removed. However, this procedure did not result in high germination. Another alternative method, the complete removal of the seed from its envelope, is not recommended because it harms the seed. Moreover, the seeds are readily attacked by fungi. The best results were quite poor, 8% germinating in 115 days (Fig. 4). The seeds were apparently healthy, but they may have been not completely mature. However, a powerful trade-off makes later seed collection difficult: granivory by rats, especially during March-May (Cuevas 2002b). On the other hand, Ricci obtained 24% emergence in 96 days. Leiva obtained 1% emergence in 146 days. All studies agree that germination starts between two and three months after sowing, and we conclude that in either case germinative potential is not high. Similar causes can result in poor germination of the congeneric C. oliveri (Table I, Appendix I).

Chenopodium sanctae-clarae (Chenopodiaceae). Endemic shrub in a critical status of conservation. The species has been introduced from Santa Clara to RCI, where it is used as ornamental plant. Germination was very poor (2-4%) after 196 days of trials (Fig. 4). Germination began in mean on day 45 (mean). The seeds were about eight months old, but they were stored under cold conditions. This storage period does not explain the poor results obtained, as Ricci found that 96% of 10-month-old seeds emerged in only 23 days. Leiva obtained a mean of 15% emergence in 340 days, with three-month-old seeds. The probable cause for these discrepancies could be the site of collection and, thus, the initial quality of the propagules: Ricci's and Leiva's seeds were from Morro Spartan, Santa Clara Island, one of the few original localities for this species, where 113 adult individuals survive in only 1,700 m2 (Cuevas 2002a). On the contrary, Cuevas (2002c) used seeds from an isolated individual growing in the Park Administration's botanic garden. Possibly, gene flow in the Morro Spartan population produces seeds of better quality than those produced by the individual in the Administration's botanic garden, which is able to self-pollinate (Crawford et al. 1988).

Dendroseris micrantha (Asteraceae). Endemic arborescent rosette, in danger of extinction, be-longing to the genus with the greater species number on the archipelago (11). Germination was zero, in spite of the relative freshness of the seeds. Ricci also found scanty emergence, a maximum of 13% in 22 days. Some seeds used in the present study may have been empty. Crawford et al. (1987) suggested that this species is self-incompatible, although this possibility has not been experimentally tested. This phenomenon, if it is true, can make reproduction difficult at low population densities, as is the case with this species. Therefore, populations with more individuals in the appropriate phenological stage should be selected for seed collection.

Haloragis masatierrana (Haloragaceae). Endemic shrub. This species is in less danger of extinction than most others investigated in this study. As it grows frequently in open sites, it is very suitable for re-vegetating sites denuded and/or affected by erosion. Laboratory results were not very good: germination was in mean 8% after 67 days (Fig. 4). Ricci, on the other hand, obtained 95% emergence in 70 days. Leiva achieved 20% emergence in 77 days, with 14 monthold seeds. Ricci's seeds and those used in this study were of similar ages (4-5 months), and those of the pre-sent study were stored under cold conditions versus ambient temperature in Ricci's work. Thus, age cannot be a factor responsible for the great differences in germinative capacity. Apparently, laboratory conditions and/or seed quality were not favorable for the initial development of this species.

Figure 4. Germination dynamics for four species found on the Juan Fernández Islands, whose germinative potential was low for factors not attributable only to seed age (see text). All curves, with the exception of Rubus, were truncated after the germination stabilization.

Figura 4. Dinámica de germinación para cuatro especies encontradas en las Islas Juan Fernández, cuyo potencial germinativo fue bajo por factores no atribuibles sólo a la edad de las semillas (ver texto). Todas las curvas, con la excepción de Rubus, fueron truncadas después de la estabilización de la germinación.

Rubus ulmifolius (Rosaceae). "Zarzamora, blackberry". Shrubby species native from Europe. It is considered as one of the worst plagues on the islands, due to its vigorous clonal and sexual repro-duction, which excludes the native vegetation (Cuevas & Van Leersum 2001). The dispersal by the austral thrush ("zorzal") is suspected, but as far as we are aware, no study has proved this dispersal. As seeds without a pre-germinative treatment had germination about null, the acid-dependence of seeds to germinate was tested. Propagules were exposed to different lapses of treatment in pure sulphuric acid. Treatments of less than 45 minutes did not produce any positive result, but increasing the time up to 60 minutes the germination was possible (Fig. 4). In other assay, increasing the acid exposure up to 90 minutes, the germination also increased, but it never achieved more than 10%, in spite of the relatively long period of the trial (Fig. 4). Similarly, Montaldo (2000) achieved 12% emergence after eight months, using non-treated seeds sown in the soil. Possibly, a more prolonged treatment in acid would be necessary to make thin the Rubus seed coat, as it probably occurs in nature with the putative austral thrush dispersal. Moreover, Ellis et al. (1985) point out that the genus Rubus also requires other treatments to break its dormancy, as cold stratification, alternating cycles of temperature, chemicals, etc.


Generally, the germination results obtained with our main species set (20) were good, with few species requiring special treatments as physical scarification. In spite the 18 native species of that set make only a 9% of the native (endemic and non-endemic) taxa of the archipelago, the patterns were quite varied: we had species germinating immediately, others species germinated in a delayed lapse, and three species did not germinate. Half of species germinated synchronously, and most taxa had a high germinative potential. Finally, five other species analyzed separately had low percentages of germination. Besides the trivial possibility of seed age affecting the response of these seeds, other factors as seed quality may also influence the results obtained.

The topic of the germination of the endemic species of the islands, especially those threatened with extinction, does not end with the present study.

Many species remain to be tested under laboratory conditions; for some of them, previous investigations under greenhouse conditions already exist (Elorza 1984; Ricci 1998; Leiva, unpublished data). In either case, for many taxa it is desirable to develop better propagation techniques, which ideally permit fast germination, a high germinative capacity, and a high seedling survival rate (e.g., Juania and Lactoris). This includes the use of methods of in vitro culture for species with problematic germination and/or sur-vival. For conservation purposes, priority should be given to species in a critical state (Ricci 2006). With the aim of achieving the most successful germination, more effective procedures should be applied, considering the date and site of seed collection, storage period and conditions, sowing date, sowing substrate, and the conditions for germination (light, humidity, and temperature).


JC thanks the following people for their work in seed collection and revision of germination trials: Flavio Chamorro, Manuel Tobar, Vanessa Briones, María Díaz, Emma Elgueta, Olga González and Drina Pizarro. We also thank Mr. Iván Leiva, who kindly provided his unpublished data, Mr. Daniel Ekblaw, who helped revise the manuscript in English, and two anonymous referees for their valuable suggestions. This work was funded by the Proyecto de Cooperación Internacional Juan Fernández (N/ BHOL/ M.RR.EE/ 93/ 003) Chile- Gobierno de los Países Bajos, carried out by the Corporación Nacional Forestal (CONAF). The paper preparation was realized with the support of a postdoctoral fellowship for JC at the Millennium Nucleus P01-057-F of the Universidad Austral de Chile, and through the facilities received from the Instituto de Investigaciones Agropecuarias, Vicuña. Finally, Fundación Senda Darwin is acknowledged for funding of the publication costs.


1Philippe Danton, Botanist; 5 rue Galileé, 38000 Grenoble, France.

2Ramón Schiller, CONAF ranger, Robinson Crusoe Island, Chile (personal communication 2002).

3Iván Leiva, Administrator of Juan Fernández National Park, Robinson Crusoe Island, Chile.

4Christian Díaz, La Campana National Park, Administrator, personal communication 2002.

5Dra. Viviana Barreda, Museo Argentino de Ciencias Naturales Bernardino Rivadavia, Sección Paleopalinología. Av. Ángel Gallardo 470, Buenos Aires, Argentina.


Akeroyd, J. 1995. How to reintroduce a plant successfully. Plant Talk 2: 14-15.         [ Links ]

Angevine, M.V. & B.F. Chabot. 1979. Seed germination syndromes in higher plants. In: Topics in plant population biology (eds. O. Solbrig, S. Jain, G. Johnson & P. Raven), pp. 188-206. Columbia University Press, New York, USA.         [ Links ]

Anonymous. 1999. La ofensiva de la palma. Chile Forestal (Septiembre): 52-55.         [ Links ]

Bernardello, G., G.J. Anderson, P. López, M.A. Cleland, T.F. Stuessy & D.J. Crawford. 1999. Reproductive biology of Lactoris fernandeziana (Lactoridaceae). American Journal of Botany 86: 829-840.         [ Links ]

Cabello, A. 1990. Antecedentes sobre la germinación y el cultivo in vitro de la Palma chilena (Jubaea chilensis (Mol.) Baillón). Ciencias Forestales 6: 3-21.         [ Links ]

Cabello, A. 1999. Determinación del contenido de humedad y de la germinación de semillas de Jubaea chilensis en el vivero y en el laboratorio y de embriones extirpados cultivados in vitro. Notas del Centro Productor de Semillas de Árboles Forestales 11. facultades/cs_forestales/publicaciones/cesaf/n11/2.html         [ Links ]

Cabello, A. & C. Botti. 1987. Estudio de germinación en Drimys winteri Forst. (Canelo). Ciencias Forestales 5: 21-30.         [ Links ]

Camus, A.I. 2003. Estudio de la germinación de Juania australis, especie endémica de la isla Robinson Crusoe. Tesis carrera de Ecología y Paisaje, Escuela de Ecología y Paisaje, Facultad de Arquitectura, Urbanismo y Paisaje, Santiago, Chile. 93 pp.         [ Links ]

Castro, S.A., M. Muñoz & F.M. Jaksic. 2007. Transit towards floristic homogenization on oceanic islands in the south-eastern Pacific: comparing pre-European and current floras. Journal of Biogeography 34: 213-222.         [ Links ]

Cereceda, P., C. Castro, R.S. Schemenauer, R. Ponce & P. Osses. 1996. Comportamiento de las precipitaciones en cinco sectores de la isla Robinson Crusoe durante los inviernos de los años 1992 a 1995. Revista Geográfica de Chile Terra Australis 41: 37-48.         [ Links ]

Clemente, M. & E. Hernández. 1995. A conservation plan for the Andalucian flora. Plant Talk 2: 16-18.         [ Links ]

Crawford, D.J., T.F. Stuessy & M. Silva. 1987. Allozyme divergence and the evolution of Dendroseris (Compositae: Lactuceae) on the Juan Fernández Islands. Systematic Botany 12: 435-443.         [ Links ]

Crawford, D.J., T.F. STUESSY & M. SILVA. 1988. Allozyme variation in Chenopodium sanctae-clarae, an endemic species of the Juan Fernández Islands, Chile. Biochemical Systematics and Ecology 16: 279-284.         [ Links ]

Crawford, D., E. Ruiz, T. Stuessy, E. Tepe, P. Aqueveque, F. González, R. Jensen, G.J. Anderson, G. Bernardello, M. Baeza, U. Swenson & M. Silva. 2001. Allozyme diversity in endemic flowering plant species of the Juan Fernández Archipelago, Chile: Ecological and historical factors with implications for conservation. American Journal of Botany 88: 2195-2203.         [ Links ]

Cuevas, J.G. 2001. Distribución y abundancia de la flora vascular del Archipiélago de Juan Fernández. I Parte. Informe Técnico N° 23. Corporación Nacional Forestal V Región, Viña del Mar, Chile. 92 pp.         [ Links ]

Cuevas, J.G. 2002a. Distribución y abundancia de la flora vascular del Archipiélago de Juan Fernández. II Parte y Final. Informe Técnico N° 29. Corporación Nacional Forestal V Región, Viña del Mar, Chile. 120 pp.         [ Links ]

Cuevas, J.G. 2002b. Lluvia de semillas en los bosques de la isla Robinson Crusoe. Informe Técnico N° 32. Corporación Nacional Forestal V Región, Viña del Mar, Chile. 25 pp.         [ Links ]

Cuevas, J.G. 2002c. Germinación de semillas del Archipiélago de Juan Fernández en condiciones de laboratorio. Informe Técnico N° 33. Corporación Nacional Forestal V Región, Viña del Mar, Chile. 85 pp.         [ Links ]

Cuevas, J.G. & C. Díaz. 2000. Programa ``Conservación in situ de la flora amenazada del Archipiélago Juan Fernández", período 1999-2000. Informe Técnico N° 20. Corporación Nacional Forestal V Región, Viña del Mar, Chile. 65 pp.         [ Links ]

Cuevas, J. G. & C. Le Quesne. 2006. Low vegetation recovery after short-term cattle exclusion on Robinson Crusoe Island. Plant Ecology 83: 105-124.         [ Links ]

Cuevas, J.G. & G. Van Leersum. 2001. Project ``Conservation, Restoration and Development of the Juan Fernández Islands, Chile". Revista Chilena de Historia Natural 74: 899-910.         [ Links ]

Cuevas, J.G., A. Marticorena & L.A. Cavieres. 2004. New additions to the introduced flora of the Juan Fernández Islands: origin, distribution, life history traits and potential of invasion. Revista Chilena de Historia Natural 77: 523-538.         [ Links ]

Danton, P., E. Breteau & M. Baffray. 1999. Les îles de Robinson. Trésor vivant des mers du Sud. Entre légende et réalité. Yves Rocher, Nathan, France. 144 pp.

Di Castri, F. & E.R. Hajek. 1976. Bioclimatología de Chile. Vicerrectoría Académica de la Universidad Católica de Chile, Santiago, Chile. 129 pp.         [ Links ]

Donoso, C. & A. Cabello. 1978. Antecedentes Fenológicos y de germinación de semillas de especies leñosas chilenas. Ciencias Forestales 1: 31-41.         [ Links ]

Ellis, R.H., T.D. Hong & E.H. Roberts. 1985. Handbook of seed technology for genebanks. Volume I. Principles and methodology, Volume II. Compendium of specific germination infor-mation and test recommendations. International Board for Plant Genetic Resources, Rome, Italy. 667 pp.         [ Links ]

Elorza, F. 1984. Mejoramiento de las técnicas de propagación de especies leñosas en el vivero del Parque Nacional Juan Fernández. Informe Práctica Estival. Corporación Nacional Forestal V Región, Viña del Mar, Chile. 72 pp.         [ Links ]

Figueroa, J.A., J.J. Armesto & J.F. Hernández. 1996. Estrategias de germinación y latencia de semillas en especies del bosque templado de Chiloé, Chile. Revista Chilena de Historia Natural 69: 243-251.         [ Links ]

Figueroa, J.A. & J.J. Armesto. 2001. Community-wide germination strategies in a temperate rainforest of Southern Chile: ecological and evolutionary correlates. Australian Journal of Botany 49: 411-425.         [ Links ]

Figueroa, J.A., P. León-Lobos, L.A. Cavieres, H. Pritchard & M. Way. 2004. Ecofisiología de semillas en ambientes contrastantes de Chile: un gradiente desde ecosistemas desérticos a templado-húmedos. In: Fisiología ecológica en plantas. Mecanismos y respuesta a estrés en los ecosistemas (ed. H. Cabrera), pp. 81-98. Ediciones Universitarias de Valparaíso, Pontificia Uni-versidad Católica de Valparaíso, Chile.         [ Links ]

Greimler, J., T.F. Stuessy, U. Swenson, C.M. Baeza & O. Matthei. 2002. Plant invasions on an oceanic archipelago. Biological Invasions 4: 73-85.         [ Links ]

Hoffmann, A. 1995. Flora Silvestre de Chile. Zona Central. Tercera edición. Ediciones Fundación Claudio Gay, Santiago, Chile. 255 pp.         [ Links ]

Hulm, P. & J. Thorsell. 1995. Robinson Crusoe's islands face an uncertain future. Plant Talk 2: 19-21.         [ Links ]

Infante, L. 1989. Estudio de la germinación de la Palma chilena (Jubaea chilensis). Memoria Ingeniero Forestal, Facultad de Ciencias Forestales, Universidad de Chile. 132 pp.         [ Links ]

Marticorena, C., T.F. Stuessy & C.M. Baeza. 1998. Catalogue of the vascular flora of the Robinson Crusoe or Juan Fernández islands, Chile. Gayana Botánica 55: 187-211.         [ Links ]

Marzol, M.V., P. Cereceda, R. Schemenauer & C. Castro. 1996-1997. Caracterización de la pluviosidad de Bahía Cumberland (Isla de Robinson Crusoe, Chile). Cuadernos de Investigación Geográfica (Universidad de la Rioja) 22-23: 97-114.         [ Links ]

Marzol, M.V., P. Cereceda & J.M. Vide. 2001. The annual regime and persistence of precipitation in Robinson Crusoe Island. Insula 10: 53-59.         [ Links ]

Montaldo, P. 1995. Manejo ecológico de las malezas. Ediciones Universidad Austral de Chile, Valdivia, Chile. 156 pp.         [ Links ]

Montaldo, N. 2000. Éxito reproductivo de plantas ornitócoras en un relicto de selva subtropical en Argentina. Revista Chilena de Historia Natural 73: 511-524.         [ Links ]

Porteous, T. 1993. Native forest restoration. A practical guide to landowners. Queen Elizabeth II National Trust, Wellington, New Zealand. 184 pp.         [ Links ]

Ray, G.J. & B.J. Brown. 1994. Seed ecology of woody species in a Caribbean dry forest. Restoration Ecology 2: 156-163.         [ Links ]

Ricci, M. 1998. Técnicas de propagación y viverización de algunas especies de la flora vascular endémica del Archipiélago de Juan Fernández. Manual Técnico N° 1. CONAF V Región, Viña del Mar, Chile. 215 pp. + 2 appendices.         [ Links ]

Ricci, M. 2001. Evaluation of conservation status of Lactoris fernandeziana Philippi (Lactoridaceae) in Chile. Biodiversity and Conservation 10: 2129-2138.         [ Links ]

Ricci, M. 2006. Conservation status and ex situ cultivation efforts of endemic flora of the Juan Fernández Archipelago. Biodiversity and Conservation 15: 3111-3130.         [ Links ]

Rodríguez, R., O. Matthei & M. Quezada. 1983. Flora arbórea de Chile. Ediciones Universidad de Concepción, Concepción, Chile. 408 pp.         [ Links ]

Roy, M.S., J.C. Torres-Mura, F. Hertel, M. Lemus & R. Sponer. 1999. Conservation of the Juan Fernández firecrown and its island habitat. Oryx 33: 223-232.         [ Links ]

Sanders, R.W., T.F. Stuessy & C. Marticorena. 1982. Recent changes in the flora of the Juan Fernández islands, Chile. Taxon 31: 284-289.         [ Links ]

Sanders, R.W., T.F. Stuessy, C. Marticorena & M. Silva. 1987. Phytogeography and evolution of Dendroseris and Robinsonia, tree Compositae of the Juan Fernandez Islands. Opera Botanica 92: 195-215.         [ Links ]

Simonetti, J. & P. Acevedo. 1990. Efecto de Oryctolagus cuniculus sobre la regeneración de Chenopodium crusoeanum, Myrceugenia fernandeziana y Sophora fernandeziana. Informe Final. Corporación Nacional Forestal V Región, Viña del Mar, Chile. 44 pp.         [ Links ]

Skottsberg, C. 1922. The phanerogams of the Juan Fernández Islands. In: The Natural History of Juan Fernández and Easter Island, vol. 2. (ed. C. Skottsberg), pp. 95-240. Almqvist & Wiksells Boktryckeri AB, Uppsala, Sweden.         [ Links ]

Stuessy, T.F., C. Marticorena, R. Rodríguez, D.J. Crawford & M. Silva. 1992. Endemism in the vascular flora of the Juan Fernández Islands. Aliso 13: 297-307.         [ Links ]

Stuessy, T.F., U. Swenson, D.J. Crawford, G. Anderson & M. Silva. 1998. Plant conservation in the Juan Fernández Archipelago, Chile. Aliso 16: 89-101.         [ Links ]

Valdebenito, H.A., T.F. Stuessy & D. Crawford. 1990. A new biogeographic connection between islands in the Atlantic and Pacific Oceans. Nature 347: 549-550.         [ Links ]

Van der Pijl, L. 1982. Principles of dispersal in higher plants. Springer-Verlag, Berlin, Germany. 214 pp.         [ Links ]

Vargas, R. 2004. Caracterización de los bosques originales de isla Robinson Crusoe. Tesis Ingeniero Forestal, Facultad de Ciencias Forestales, Universidad Austral de Chile, Valdivia, Chile. 84 pp. + v pp.         [ Links ]

Venable, D.L. & L. Lawlor. 1980. Delayed germination and dispersal in desert annuals: escape in space and time. Oecologia 46: 272-282.         [ Links ]

Zavada, M.S. & J.M. Benson. 1987. First fossil evidence for the primitive angiosperm family Lactoridaceae. American Journal of Botany 74: 1590-1594.         [ Links ]

Zunino, H. 1993. Variabilidad espacial de las precipitaciones y potencial hidrológico de la niebla en la isla de Robinson Crusoe, archipiélago de Juan Fernández, Chile. Memoria para optar al título de Geógrafo. Pontificia Universidad Católica de Chile, Santiago, Chile. 100 pp.         [ Links ]

Appendix I. Summary of germination trials for the Juan Fernández flora. All provenances are from Robinson Crusoe Island, except when the contrary is indicated.

Apéndice I. Resumen de ensayos de germinación para la flora de Juan Fernández. Todas las procedencias son de la Isla Robinson Crusoe, excepto cuando se indica lo contrario.

Appendix II. Percentages of germination (mean ± 1 SE) for the Juan Fernández flora, trials which produced poor or null germination due to factors not attributable only to seed age. All provenances are from Robinson Crusoe Island.

Apéndice II. Porcentajes de germinación (media ± 1 EE) para la flora de Juan Fernández, ensayos que produjeron pobre o nula germinación debido a factores no atribuibles sólo a la edad de las semillas. Todas las procedencias son de la Isla Robinson Crusoe.

Recibido: 22.08.06,
Aceptado: 14.12.06 .


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