Analysis of the contribution and efficiency of the Santuario de la
Naturaleza Yerba Loca, 33º S in protecting the regional vascular plant
flora (Metropolitan and Fifth regions of Chile)

Análisis de la contribución y eficiencia del Santuario de la Naturaleza Yerba Loca,
33º S, en la protección de la flora vascular regional (regiones Metropolitana y Quinta
de Chile)

MARY T. K. ARROYO¹, CLODOMIRO MARTICORENA², OSCAR MATTHEI²,
MÉLICA MUÑOZ³ & PATRICIO PLISCOFF¹

¹Center for Advanced Studies in Ecology and Research on Biodiversity (CMEB), Facultad de Ciencias,
Universidad de Chile, Casilla 653, Santiago, Chile;
e-mail: southern@abello.dic.uchile.cl
²Departamento de Botánica, Facultad de Ciencias Naturales y Oceanográficas, Universidad de
Concepción, Casilla 160-C, Concepción, Chile
³Sección Bótanica, Museo Nacional de Historia Natural, Casilla 787, Santiago, Chile

ABSTRACT

Santuario de la Naturaleza Yerba Loca (SN Yerba Loca), Metropolitan Region (MR), 33º S, Chile is analyzed for its conservation value and efficiency in protecting native vascular plants in a regional context. The reserve's flora of 500 species and subtaxa was evaluated for species richness, endemism, range size and marginally distributed taxa, using species-area analysis, and tendencies in the floras of the MR (1.434 species and subtaxa) and MR-Fifth regions (1,841 species and subtaxa) to set the regional pattern. The reserve (0.7 % of MR land area and 0.3 % MR-Fifth land area) contains 34 % of the MR and 27% of the MR-Fifth floras, and around 16-17 % of the mediterranean-climate area (regions IV-VIII) flora of central Chile. Veech's Relative Richness Index (RRI) revealed that SN Yerba Loca houses exaggerated richness in relation to its land area (28 % more species than expected from the regional model). However, endemism rates (35 % Continental Chile endemics, 22 % Mediterranean endemics, 3% MR-Vth endemics) are statistically lower than in the MR (44 %, 29 %, 9 %) and the MR-Vth (48 %, 31 %, 11 %) floras, and SN Yerba Loca houses proportionately fewer MR endemics (2 %) than the MR (6 %). Compared with the regional floras, the reserve contains statistically fewer marginally distributed species, and range size (median = five administrative regions) is significantly larger. The reserve's outstanding species richness compensates for its low endemism rates bringing the absolute number of endemics to 92 % of the regional expectation. Corresponding values for marginally distributed species are 81 % (northern limits), 63% (southern limits) and for median and shorter range taxa, 100 %. It is concluded that SN Yerba Loca is a highly efficient reserve from the point of view of vascular plant conservation, and represents an excellent conservation choice. SN Yerba Loca and MN El Morado (a second state protected area in the MR), conservatively, house 39 % of the native vascular plant flora of the MR (30 % of that of the MR-Fifth) on 0.9 % of MR land area. Our study emphasizes that relatively small land areas in central Chile can house significant amounts of biodiversity, and that moreover, the RM and RM-Fifth are areas of high species richness within the central Chilean biodiversity hotspot. The outstanding conservation value of SN Yerba Loca calls for a management plan designed to assure the integrity of the reserve under increasing pressures from the Metropolitan Region with six million inhabitants.

Key words: conservation efficiency, endemism, hotspot, protected area, mediterranean flora, species richness, SN Yerba Loca.

RESUMEN

Santuario de la Naturaleza Yerba Loca (SN Yerba Loca), 33ºS, un área protegida en la Región Metropolitana (RM) de Chile, es analizado en términos de su aporte a la conservación de la flora vascular y su eficiencia en un contexto regional. Se analizó la flora vascular nativa, compuesta por 500 taxa en terminos de riqueza de especies, endemismos, rango de distribución, y presencia de taxa con límites de distribución geográfica en la zona de estudio, empleando análisis de especie-área y en base de tendencias en las floras de la RM (1-434 especies y subtaxa) y RM-Quinta Región (1.841 especies y subtaxa). La reserva (de 0,7 % de la superficie de RM y 0,3 % de RM-Quinta Región) contiene el 34 % de la flora nativa de RM y el 27 % de RM-Quinta Región, y alrededor de 16-17 % de la flora mediterránea (regiones IV-VIII) de Chile central. El Índice de Riqueza Relativa (IRR) de Veech indica que SN Yerba Loca alberga una riqueza extraordinaria en relación a su superficie (28 % más especies de lo esperado del modelo regional). Sin embargo, los niveles de endemismo para SN Yerba Loca (35 % endémicos de Chile continental, 22 % endémicos de la zona mediterránea y 3 % endémicos a RM-Quinta Región) son estadísticamente más bajos que los niveles regionales (RM: 44, 29 y 9 %; RM-Quinta Región: 48, 31, y 11 %), y SN Yerba Loca alberga proporcionalmente menos taxas endemicos a RM (2 %) que RM (6 %). En comparasión con las floras regionales, SN Yerba Loca contiene menos especies marginalmente distribuidas, y los rangos geográficos para especies presentes en SN Yerba Loca (mediana = cinco regiones administrativas) son significativamente mayores que en las floras regionales. La riqueza destacada de la reserva compensa su bajo nivel de endemismos, con el número absoluto de taxas endemicos llegando al 92 % de la expectación regional. Valores correspondientes para especies con distribuciones marginales son 81 % (límites norte), 63 % (límites sur) y para taxas con rangos cortas, 100 %. Se concluye que SN Yerba Loca es una reserva muy eficiente desde el punto de vista de conservación de la flora vascular. SN Yerba Loca y MN El Morado (un segundo área protegida en RM) constituyen el 0,9 % de la superficie de RM. Albergan alrededor del 39 % de la flora de la RM (y 30 % de la flora de RM y RM-Quinta Región). Nuestro estudio indica que áreas pequeñas en Chile central pueden albergar mucha biodiversidad, y que RM y RM-Quinta Región son áreas de alta riqueza de especies dentro del "hotspot" Chile central de biodiversidad. El sobresaliente valor de SN Yerba Loca en lo respecto a la conservación exige un plan de manejo para asegurar la integridad de la reserva frente de presiones de la Región Metropolitana con seis millones de inhabitantes.

Palabras clave: eficiencia de conservación, endemismos, área protegida,"hotspot" flora , mediterránea, riqueza de especies, SN Yerba Loca.

INTRODUCTION

The conservation of biodiversity in a modern context (Heywood & Watson 1995) spans the gamut from the sustainable use of the components of biodiversity to strict preservation. Globally, protected areas are considered to constitute an important means for protecting biodiversity (Miller et al. 1995). Most national (and some private) protected areas combine the sustainable use of the components of biodiversity and biodiversity preservation. The protection and maintenance of biodiversity in the natural environment on the one hand, underpins recreation, while the use of protected areas for recreation permits biodiversity per se to be preserved through non-extractive sustainable use. At the landscape level, protected areas also can perform important ecological services, such as watershed protection, CO₂ sequestration (e.g. as in forests) and maintenance of carbon sinks (e.g., as in peat bogs).

Although the protection of biodiversity is now explicitly recognized as an objective of protected areas, many protected areas were set up prior to world concern about biodiversity, with scenic or wilderness value, or the protection of a select number of useful or charismatic species being the overriding criteria for their establishment (Armesto et al. 1998). For example, in India, protected areas of forest species have existed since the IV century (Groombridge 1992). Using data given in Groombridge (1992), around 60 % of all existing protected areas were established before the 1980s, which corresponds with the decade of accelarated global concern for biodiversity. It was in the late 80's that Wilson & Peters (1988) brought the notion of biodiversity to a wide field of scientists and others. The contracted form of biological diversity (biodiversity) was apparently coined by Walter G. Rosen in 1985 for the first planning meeting of the "National Forum on Biodiversity" held in Washington DC in September, 1986 (Heywood & Watson 1995). Gathering concern throughout the late 80's and early 90's eventually lead to the signing of the Convention on Biological Diversity which came into force in 1993. In Chile the establishment of protected areas dates to 1907, with the creation of the Reserva Nacional Malleco, followed by the Villarrica, Alto Bío-Bío and Llanquihue forest reserves created in 1912 (Benoit 1996). The first national park (still in existence), Parque Nacional Vicente Pérez Rosales, was created in 1926, with 10 new parks established between 1935-1945. Over the period 1945-1969, 52 additional national parks and forest reserves came into being. Of the 90-odd current protected areas in Chile, thus a very large number were created before 1980.

As biodiversity became a global concern, the notion of biodiversity conservation began to make its way into the hands of the decision makers (e.g., Muñoz et al. 1996). However, even today, new protected areas tend to be created more on the basis of a select group of species, or dominant species, rather than concern for adequate and efficient coverage of all species in a region. The early establishment of many protected areas on the basis of wild, scenic or utilitarian value, signifies that there can be no guarantee that a set of existing national parks and reserves protects all of a region's biodiversity. Nor can it be assumed that existing protected areas are efficient in terms of the amounts of biodiversity they protect. Given the latter, as a working hypothesis, individual protected areas should span the natural range of variation in the landscape, there being reserves with relatively low to relatively high levels of biodiversity: that is, some protected areas, by chance could house high levels of biodiversity, while others could contain low amounts.

In order to consolidate any national protected area system a first objective should be that of learning how much biodiversity is contained in existing protected areas. This information is fundamental for determining which species in the region of interest are protected and their level of protection. Secondly, individual protected areas should be evaluated for their relative conservation efficiency. This kind of information is essential to guide the right decisions when changes in the boundaries (reductions, amplifications) of existing protecting areas are contemplated. In this paper we outline and analyze the vascular plant flora of the Santuario de la Naturaleza Yerba Loca, Metropolitan Region, 33º S, Chile. The Santuario de la Naturaleza Yerba Loca is one of four protected areas managed by CONAF-Chile in the densely populated Metropolitan Region of Chile. Some 25 % of the land area of the Metropolitan Region is intensively used for agriculture and urban development; many semi-natural areas are today heavily subject to grazing (Arroyo et al. 2000). The Metropolitan Region occurs in the mediterranean-type climate area of central Chile (di Castri & Hajek 1976), known for its high levels of endemism and species richness (Arroyo & Cavieres 1997, Arroyo et al. 1999). Reflecting the latter, in a recent paper in Nature, central Chile (based on the original data provided by Arroyo et al. 1999 in Mittermeier et al. 1999) has been declared a "biodiversity hotspot for conservation priority" (Myers et al. 2000). Based on comprehensive floristic lists of the native flora for the Metropolitan Region and Metropolitan and Vth regions combined, we first determine the proportion of the regional floras contained in the Santuario de la Naturaleza Yerba Loca. We then employ a series of criteria designed to evaluate the conservation value of the reserve. Specifically we analyze the reserve's flora from the perspective of: (i) total number of vascular plants protected in relation to the physical size of the reserve; (ii) concentration and absolute numbers of endemic taxa; (iii) average range size of taxa; (iv) presence of marginally distributed species. For the purposes of our analysis, a highly efficient reserve is one in which: (i) a higher number of species in relation to the expected regional average for the reserve size are protected; (ii) the density of endemic species is high; (iii) the average range size of species is low, and (iv) marginally distributed species are well represented. Finally, we combine published floristic information for Monumento Nacional El Morado (a second protected area in the Metropolitan Region) with that for Santuario de la Naturaleza Yerba Loca and determine the proportion of the Metropolitan Region (and Metropolitan-Quinta Region combined) flora contained in the two protected areas.

MATERIAL AND METHODS

Study site

Santuario de la Naturaleza Yerba Loca (hereafter SN Yerba Loca) (Fig. 1) is located in the Municipality of Lo Barnechea to the NE of the city of Santiago. The area was decreed a "Nature Sanctuary" on 24 July, 1973 by the Chilean Council of National Monuments (Consejo Nacional de Monumentos Nacionales). As originally designated, the reserve covers 39,029 ha. However, only the 10,500 ha. of the reserve managed by the Corporación Nacional Forestal (CONAF) is considered here. The reserve is centered on the deeply-cut, northeast-southwest trending valleys of the Estero Yerba Loca running from the base of Cerro La Paloma, and Estero La Leonera, situated to the east and running into Estero Yerba Loca. It comprises steep mountainous territory located to the east of the Río San Francisco, from its junction with Estero Yerba Loca to the level of La Ermita. Estero Yerba Loca shows strong signs of heavily glaciation in the Pleistocene. Permanent glaciers are found today on Cerro La Paloma and Cerro Plomo. The reserve spans an altitudinal range of approximately 1,300 to 5,340 m. However, above 3,600 m vascular plants are essentially absent. Vegetation types include typical low-elevation mediterranean sclerophyllous scrub and succulent scrub (900-1,500 m), montane sclerophyllous woodland dominated by Kageneckia angustifolia (1,600-2,000 m), and high alpine vegetation (2,000-3,600 m). A significant extension of Andean wetlands (locally referred to as "vegas") is found on the eastern side of the reserve below Cerro La Parva. The upper reaches of Estero Yerba Loca below Cerro La Paloma contain one of the largest and best conserved extensions of high Andean cushion bog in the general area. The structure and ecology of the alpine vegetation of the Farellones_La Parva-Valle Nevado area immediately to the east, as described in Arroyo et al. (1981, 1982) and Cavieres et al. (2000), is typical of SN Yerba Loca. Details on the composition and ecology of montane sclerophyllous forest in the reserve can be found in Arroyo & Uslar (1993) and Peñaloza et al. (2001).

Fig. 1: Location of the Santuario de la Naturaleza Yerba Loca, Chile.

Ubicación del Santuario de la Naturaleza Yerba Loca, Chile.

Floristic data

Knowledge of the flora of SN Yerba Loca was generated from an intensive field collecting program supplemented with a small number of previous literature and herbarium records (3 % taxa). Field work incompassed 92 man/woman days of collecting time spread across 31 separate collecting trips in the early spring to late autumn of 1998-2001 and covered the entire elevation range of vascular plants, all major watersheds, ranges and expositions as well as vegetation types in the reserve. All plant species occurring along the explored gradients and trajectories were collected and herbarium specimens prepared. Each plant collection was formally recorded in relation to georeferenced points (Garnmin GPS). So as to document the flora of the reserve, herbarium collections have been deposited at the University of Concepción (CONC) and Museo National de Historia Natural (SGO). The field effort resulted in more than 1,700 herbarium collections and their respective duplicates. The floristic lists for the Metropolitan Region and Metropolitan-Fifth regions combined were generated from the Chilean flora data base, University of Concepción (version 23 August, 2001). This constantly up-dated electronic data base contains distributional data at the level of Chile administrative regions and life-form data for all currently accepted species, subspecies and varieties in the Chilean flora. The only major modification to the data base affected was replacement of the existing set of names for the genus Calceolaria with a much reduced set of taxa recognized in the recent taxonomic revision of the genus by Ehrhart (2000). A few new taxa for the Metropolitan Region discovered in SN Yerba Loca were added to the regional lists. The regional lists can be considered fairly complete, since the Metropolitan and Vth regions are among the best explored in Chile.

Conservation criteria: species richness

In order to evaluate the species packing efficiency of a reserve, a measure of relative species richness is required that takes into account that species richness does not increase linearly with size of area. The efficiency of SN Yerba Loca in concentrating species richness was evaluated by comparing the number of species contained in the reserve in relation to the expected number of species for the reserve land area as predicted from a regional species-area curve. Previous work for a wider area of central Chile (mediterranean area in general) showed that species richness is strongly linearly correlated with area under the power function (Arroyo et al. 1995, Cowling et al. 1996). In order to evaluate the species richness capacity of SN Yerba Loca, we constructed a new species-area curve for the area comprising the Fifth and Metropolitan regions. The data used to construct the new curve comprised: (a) current estimates of species richness for the Fifth and Metropolitan Regions obtained from the Chilean species data base (see above); (b) species numbers for the Valparaíso area, Marga Marga Valley, Santiago Valley and Quebrada de La Plata as in Arroyo et al. (1995); (c) a published species list for Monumento Nacional El Morado (hereafter MN El Morado) (Teillier et al. 1994) and the new data for SN Yerba Loca, as presented herein. Inclusion of MN El Morado also allowed comparison of the species packing efficiency of SN Yerba Loca with that of a second protected area in the Metropolitan Region. For the species richness analysis, subtaxa were eliminated from the data sets for the Vth, Metropolitan Region, SN Yerba Loca and MN El Morado, so as to enable use of the floristic lists previously published in Arroyo et al. (1995) which correspond to species-level counts. Following Veech (2000), we calculated the Relative Richness Index (RRI) as per Bowers & McLaughlin (1982) for all data points used in constructing the regression line. RRI is obtained from the regression residuals expressed as a percentage of the expected number of species for the size of the land area of a reserve (c.f., Bowers & McLaughlin 1982, Veech 2000):

where S_obs is observed species richness S_exp is expected species richness.

Conservation criteria: endemism

The efficiency of SN Yerba Loca in concentrating endemic taxa was evaluated by comparing levels of endemism in the reserve with those in the floras of the Metropolitan-Fifth regions combined and the Metropolitan Region. Four endemism categories were considered: (a) taxa restricted in distribution to continental Chile; (b) taxa restricted in distribution to administrative regions Fourth through Eigth; (b) taxa restricted in distribution to the Metropolitan and Vth regions combined; (c) taxa restricted in distribution to the Metropolitan Region. Taxa (species, or subtaxa) restricted in distribution to continental Chile will be referred to as "Continental Chile endemics". Taxa restricted in distribution to Chile and the area described by administrative regions IV-VIII will be referred to as "Mediterranean endemics". These last administrative regions were chosen as a surrogate for the mediterranean-type climate region based on their close fit to the latitudinal limits of the mediterranean-climate area as defined by Van Husen (1967) (30-38º S) and the availability of accurate species lists at the level of Chile's administrative regions in the Chilean flora data base (University of Concepción). It should be noted that taxa endemic to the Metropolitan Region constitute a subset of the Metropolitan-Vth regions combined endemics, which in turn are a subset of the Mediterranean endemics. Mediterranean endemics comprise a subset of the Continental Chile endemics. Endemism status at the level of continental Chile was determined by consulting the geographical distributions of species in monographic treatments, and checking the recently published checklists for neighboring or geographically-close Argentina (Zuloaga et al. 1994, Zuloaga & Morrone 1996, 1999), Peru (Brako & Zarucchi 1993) and Ecuador (Jorgensen & León-Yáñez 1999). Subsets of mediterranean and regional endemics were obtained by filtering distributional information contained in the Chilean flora data base at the level of Chile's administrative regions. Statistical comparisons of endemism levels for SN Yerba Loca with the regional floras were made using the G-test of independence (Sokal & Rohlf 1995).

Conservation criteria: range size

In order to determine whether the constituent taxa in the SN Yerba Loca reserve are more narrowly distributed on average than species in the regional floras, we compared the range sizes of species and their subtaxa in the SN Yerba Loca, the Metropolitan-Fifth regions combined and the Metropolitan Region. The number of administrative regions occupied by each taxa in Chile was used as an index of range size. Following Matthei (1995) and Arroyo et al. (2000), the Vth and Metropolitan Regions were amalgamated into one composite region. The latter was considered appropriate so as to define a geographical unit that is more equivalent to the other 11 administrative regions in Chile, all of which span the breath of Chile from the Pacific Ocean to the Andean crest. Range size was calculated separately for the subsets of Continental Chile endemics, Mediterranean endemics and non-endemic native taxa. The subset of Metropolitan-Fifth regions combined and Metropolitan Region endemics were not considered in this analysis, in that all species on the criteria used here, have the same distributional ranges (one administrative region). Use of the number of administrative regions gives an approximate measure of range size, in that there is considerable variation in the size of the individual administrative regions in Chile. However, when comparisons are made for subsets of flora from the same general geographical area, as is the case here, this variation is of little relevance, in that species in the areas compared are drawn from the same regional species pool, and thus will be subject to the same level of error. The number of administrative regions occupied by a species in Chile is useful as a comparative measure of geographical range in the absence of more accurate distributional data, in that it provides a rough measure of the latitudinal range. Statistical comparisons of range size among the floras were made using the non-parametric Mann-Whitney U-test (Statistica Version 6.0, 1998 version).

Conservation criteria: marginally distributed
species

We evaluated SN Yerba Loca's contribution to conserving marginally distributed species by considering presence in the reserve of species with northern and southern distributional limits found within the Metropolitan-Fifth regions combined, following the same criterion as used for range size for determining distributional limits. The percentages of marginally-distributed species in SN Yerba Loca were then compared with corresponding sets of data generated for the regional floras using the G-test of Independence (Sokal & Rohlf 1995). Subsets of species of different origins (endemic, non-endemic, etc.), as well as the entire native flora, were considered so as to detect differences between floristic components.

Conservation criteria: combined criteria

Proportional values provide a measure of the propensity of a local flora to contain species of a given category (for example endemic species). However, low proportional values for any given category of species, could be compensated by higher than average species richness. Absolute expected numbers of endemic species, numbers of species with marginal distributions and numbers of species with shorter than average distributions were estimated for an area the size of SN Yerba Loca, and compared with observed numbers in the reserve. For endemism, the expected number of species in the different endemism classes were obtained by multiplying the corresponding percentages found in the Metropolitan Region-Fifth regions combined by the expected number of species for SN Yerba Loca as obtained in the regression equation obtained under the procedures described in Material and Methods, section (a). A parallel procedure was employed for northern and southern distributional limits. For range size we first determined the proportion of species in the regional flora with median to lower range sizes, and then multiplied this number by the expected number of species for an area the size of SN Yerba Loca obtained from the regression. This number was compared for the observed number of taxa in SN Yerba Loca, having median and lower range sizes, as per in the regional flora. Because our range size and endemism data correspond to the species and subtaxa level, it was necessary to adjust the predicted species number of SN Yerba Loca as obtained from the regression, so as to account for subtaxa. This was achieved by multiplying the expected species number obtained in regression by the ratio (number of species + subtaxa) / number of species as in the observed data for SN Yerba Loca.

RESULTS

Species richness and life-form

The Appendix gives the full listing of native species found in SN Yerba Loca, along with life-form, endemism status and distribution according to administrative region. The identifications for eight species and one putative hybrid in the Appendix 1 (nine taxa in total) are tentative. Including the nine taxa, the native flora of SN Yerba Loca is comprised of 500 species and their subtaxa. The total number of species in the reserve is 488, excluding 11 varieties or subspecies in cases where species are represented by one or more varieties or subspecies (as in Adesmia papposa, Adiantum chilense, Bromus setifolia, Calceolaria ascendens, Calceolaria corymbosa, Hypochaeris tenuifolia, Mustisia ilicifolia, M. subulata, Oxalis compacta, Sisyrinchium arenarium), and one putative hybrid (Cistanthe arenaria x) (total de subtaxa = 12).

Table 1 compares species richness and the life-form composition for the native floras of SN Yerba Loca, the Metropolitan-Fifth regions combined and the Metropolitan Region. The native flora of the Metropolitan-Fifth regions combined stands at 1,841 species and subtaxa, with the Metropolitan Region housing 1,434 species and subtaxa. Disregarding subtaxa reduces these last numbers to 1,723 and 1,355 species, respectively. Eliminating the nine taxa of tentative identity in SN Yerba Loca (n = 491), it can be seen that the nature reserve conservatively contains an outstanding 34 % of all native vascular plant taxa reported for the Metropolitan Region and 27 % of all vascular plants known for the Metropolitan-Fifth regions combined. Present knowledge suggests that the entire mediterranean climate area as described by regions IV-VIII contains around 3,160 native species and subtaxa. Again, using a conservative estimate for SN Yerba Loca, it can be seen that SN Yerba Loca contains around 16 % of the entire flora of the mediterranean area of central Chile. The mediterranean region as defined here, contains 2,864 species when subtaxa are disregarded. Using a conservative estimate of 480 species for SN Yerba Loca (i.e., eliminating the tentatively identified species from the total species number of 488), SN Yerba Loca can be seen to house 17 % of the mediterranean region flora.

Life-form composition in both regional floras is diverse (Table 1) and not significantly different (G = 0.532, NS). The life-form spectra of the SN Yerba Loca flora does not differ statistically from that of either regional flora (Metropolitan-Fifth regions: G = 1.140, P < 0.05; Metropolitan Region: G = 0.968, P < 0.05). This indicates that the reserve's flora is highly representative of the regional life-form spectrum.

Figure 2 depicts the regional species-area relationship. The log-log regression is highly significant (F_{1, 7} = 99.781, P < 0.001, R² = 0.934), indicating that the curve can be used reliably as an indicator of the relationship between species richness and area. Table 2 gives values of the Relative Richness Index (RRI) for areas used in the analysis, ranked from high positive to low negative values. It can be seen that RRI for SN Yerba Loca ranks highest among all the areas compared, the reserve having an excess of 28 % species in relation to the model. On a regional scale, thus SN Yerba Loca supports outstanding species richness in relation to its land area. Of special note is the RRI value for MN Nacional El Morado. This reserve situated in the Andes in the Cajón de Maipo to the south of SN Yerba Loca, in contrast to SN Yerba Loca, is associated with a large negative RRI value, and thus is an area of low species richness for its particular land area. Other interesting asides are the higher intrinsic richness of the Metropolitan Region in comparison with the Fifth Region, and the very low intrinsic richness of the Santiago valley.

Fig. 2: Regional species-area relationship (power function) based on data for the Fifth and Metropolitan regions of Chile. Sources of floristic data are indicated in the Methods section. Regression equation: Log S = 2.109 + 0.234 Log A, F_1,7 = 99.718, P < 0.001, R² = 0.934. Data points for the two protected areas with available floristic lists are indicated by the arrows.

Relación regional entre riqueza de especies y área (función de potencia) basada en datos para las regiones Quinta y Metropolitana de Chile. Se indican las fuentes de los datos florísticos en la sección de Métodos. Ecuación de regresión: Log S = 2,109 + 0,234 Log A; F_1,7 = 99,718; P < 0,001; R² = 0,934. Los datos correspondientes a las dos áreas protegidas en la Región Metropolitana con listas florísticas disponibles están indicados con flechas.

Endemism

Table 3 provides endemism levels for SN Yerba Loca and the regional floras. So as not to overestimate endemism levels, the 9 taxa of tentative identity have been placed in the non-endemic category (see Appendix 1). Endemism levels in the regional floras are high (44-48 %, continental Chile endemics; 29_31 %, Mediterranean endemics). Moreover significant numbers of species are strictly endemic to the regions under consideration (see Table 3). Continental Chile endemism levels are higher in the Metropolitan-Fifth regions combined than in the Metropolitan Region (G = 5.832, P < 0.05), however this pattern did not characterize the other endemism categories.

Of the 883 continental Chile endemics in the Metropolitan-Fifth regions combined, 173 (19.6 %) are found in SN Yerba Loca. Of the 575 Mediterranean endemics found there, 110 (19.1 %) occur in SN Yerba Loca. Sixteen of the 193 species endemic to the Metropolitan-Fifth regions combined (8.3 %) are found in the reserve. Corresponding figures in relation to the Metropolitan Region flora are 27.6 %, 26.1 %, and 12.1 % respectively. The reserve contains 11 of the 91 (12.1 %) strictly Metropolitan Region endemics. Although high proportions of the endemic taxa found in the regional floras occur in SN Yerba Loca, the reserve's endemism levels are significantly lower than in the two regional floras: (a) Metropolitan-Fifth regions: continental Chile endemics: G = 28.800, P < 0.001; Mediterranean endemics: G = 16.883, P < 0.001; Metropolitan-Fifth endemics: G = 31.425, P < 0.001; (b) Metropolitan Region: continental Chile endemics: G =12.879, P < 0.001; Mediterranean endemics: G = 10.565, P < 0.001; Metropolitan-Fifth region endemics: G = 22.191, P < 0.001; Metropolitan Region endemics: G= 15.001, P < 0.001. Inclusion of the nine taxa with tentative identifications did not alter these tendencies. Overall, thus, SN Yerba Loca, is not an area outstanding for its concentration of endemic taxa.

Range size

Taxa occurring in the two regional floras may be distributed in one to 12 political regions with a median value of four regions in each case. In both regional floras, continental Chile endemics (range = 1-10, median = 3) and mediterranean endemics (range = 1-5, median = 2), as expected, have smaller geographical ranges than the overall flora, while non-endemic natives tend to have the largest distributions (range = 1-12; median = 5). Comparing SN Yerba Loca and the two regional floras, in all but one case (continental Chile endemics) the median values for range size were higher in SN Yerba Loca (total flora = 5, mediterranean endemics = 3, non-endemic natives = 6). Range size distribution was not normal, forcing the use of non-parametric statistics. Significant differences between SN Yerba Loca and the two regional floras were revealed, with range size always being larger in the SN Yerba Loca flora: (a) Metropolitan-Fifth regions combined: total native flora: Z = 6.534, P < 0.001; Continental Chile endemics: Z = 3.613, P < 0.001; Mediterranean endemics: Z = 9.138, P < 0.001; non-endemic natives: Z = 3.611, P = 0.005; (b) Metropolitan Region: total native flora: Z = 4.765, P < 0.001; continental Chile endemics: Z = 2.730, P < 0.01; Mediterranean endemics Z = 8.042, P < 0.001; and non-endemic natives: Z = 2.834, P = 0.005. Summarizing, in general the reserve contains proportionately more species with larger range sizes than the regional floras. The distribution of range sizes for the SN Yerba Loca and Metropolitan Region floras are shown in Fig. 3.

Fig. 3: Frequency distribution of range size for the flora of SN Yerba Loca, 33º S, Chile and the Metropolitan Region, Chile. Range size is represented as the number of administrative regions in which a species occurs: black bars = Metropolitan Region; white bars = SN Yerba Loca. See Table 4 for n values.

Distribución de frecuencias del tamaño de rango de las floras de SN Yerba Loca, 33º S, Chile y la Región Metropolitana, Chile. El tamaño del rango es representado por el número de regiones en la cual existe una especie: barras negras = Región Metropolitana; barras blancas = SN Yerba Loca. Ver Tabla 4 para los valores de n.

Marginally distributed species

Table 4 shows numbers of taxa in the Metropolitan-Fifth regions combined and the Metropolitan Region with northern and southern distributional limits, respectively, together with corresponding data for SN Yerba Loca. Marginal populations of a large number of species in both the Metropolitan Region and the Metropolitan-Fifth regions combined are found at the latitude of those regions (northern limits: 39-41 % - southern limits: 26-31 %; Table 4, Total flora), there being nevertheless significantly more northern than southern distributional limits in both cases (Metropolitan-Fifth: G = 37.003, P < 001; Metropolitan: G = 60.252, P < 0.001). Between 17-23 % of northern limit and 13-21 % of southern limit taxa in the two regional floras are found in SN Yerba Loca. These percentages tend to be slightly lower when only endemic taxa are considered, and somewhat higher when native non-endemic taxa are considered.

Comparing the two regional floras, for all floristic components, significant differences in relation to the proportion of northern limits are lacking. In contrast, for southern limits, the Metropolitan-Fifth regions combined exceed the Metropolitan Region, except in the case of the non-endemic natives (Total flora: G = 12.351, P < 001 - continental Chilean endemics: G = 8.709, P < 0.005 _ Mediterranean endemics: G = 6.847, P < 0.01). The two regional floras thus show interesting structural differences, a feature that probably relates to the presence of proportionately more high elevation species in the Metropolitan Region flora.

Comparing the total native flora of SN Yerba Loca with the Metropolitan Region and the Metropolitan-Fifth regions combined, the proportions of marginally distributed taxa occurring in the reserve are significantly lower (Metropolitan-Fifth, total flora: northern limit: G = 37.209, P < 0.001 - southern limit: G = 50.441, P < 0.001; Metropolitan, total flora: northern limit: G = 28.017, P < 0.001 - southern limit: G = 21.462, P < 0.001 ). This situation is repeated for all categories of endemic taxa in both regional floras: (a): Metropolitan-Fifth: continental Chile endemics: northern limit: G = 15.144, P < 0.001 - southern limit: G = 19.769, P < 0.001; Mediterranean endemics: northern limit: G = 7.835, P < 0.01- southern limit: G = 20.627, P < 0.001; (b) Metropolitan: continental Chile endemics: northern limit: G = 13.790, P < 0.001- southern limit: G = 6.492, P < 0.05; Mediterranean endemics: northern limit: G = 8.371, P < 0.005 - southern limit: G = 8.238, P < 0.005. It also characterizes non-endemic native taxa: (a) Metropolitan-Vth: northern limit: G = 16.711, P < 0.001- southern limit: G = 15.439, P < 0.001; (b) Metropolitan: northern limit: G = 11.073, P < 0.001 - southern limit: G = 9.488, P < 0.001. Thus, although a considerable number of marginally distributed taxa are found in SN Yerba Loca, it may be concluded that the reserve concentrates proportionately fewer marginally distributed species than the regional floras.

Combined criteria

As explained earlier, relatively low proportions of endemics and marginally distributed species could be compensated for in absolute terms in a species-rich area such as SN Yerba Loca, such that the reserve becomes efficient from a conservation perspective on these criterion in addition to species richness. Table 5 compares expected and observed numbers of taxa combining the three criteria previously considered separately. Because our species richness predictions were based on a curve for the general geographical circumscription of the Fifth and Metropolitan regions, the endemism levels and other parameters used in the projections in Table 5 are necessarily those for the Metropolitan-Fifth regions combined. The expected number of species and subtaxa for SN Yerba Loca (extrapolated from the expected number of species) was 391. It can be seen (Table 5) that the high species richness of SN Yerba Loca brings the absolute number of Continental Chilean endemics quite close to the regional expectation (92 %), although the deficit tends to be larger in the more restricted endemism categories. For species with median or smaller ranges (Table 5) the observed number is equal to the regional expectation (100%). High species richness least compensates the number of species with northern or southern limits (81 and 63%, respectively). Overall, the high species richness in SN Yerba Loca goes along way to compensate for the lower contributions of categories of species considered here to have high conservation priority.

DISCUSSION

The native vascular flora of SN Yerba Loca of 500 native taxa, conservatively containing 27 and 34 % of the floras of the Metropolitan-Fifth regions combined and Metropolitan Region, respectively, and an estimate of 16-17 % of all species distributed in the Mediterranean-type climate area (Fourth to Eigth regions) is clearly very rich by any standard. The regional floras themselves, in the context of the central Chile hotspot, are also very rich. Squeo et al. (2001), for a total land area of 40,462 km², report a total of 1,478 species and subtaxa for the Fourth Region of which 53.5 % are continental Chile endemics. While the endemism levels (43.7-48.0 %) are slightly lower in the regional floras considered here, intrinsic species richness (1,434 species and subtaxa on a land area of 15,480 km² (RM) and 1841 species and subtaxa on 31,794 km² (RM-Fifth) is clearly much higher in the north-central part of the Mediterranean area than in the arid Fourth Region to the north. What is more relevant in the present context, however, is the efficiency of the SN Yerba Loca in concentrating species richness in relation to its land area. Reserves that contain a high number of species in relation to the amount of land area protected can be considered to have high conservation value (Margules et al. 1988, Araujo 1999). The use of regression residuals expressed as a percentage of expected species richness (standardized residuals - Veech 2000) (RRI) constitutes the only legitimate way to establish whether a given area is richer than expected, given that richness will be always be a function of size of area. Based on these principles, SN Yerba Loca was revealed to concentrate an outstanding number of species for its land area. Clearly, on the basis of the richness criterion and from a utilitarian point of view the reserve constitutes an excellent conservation choice with regard to vascular plants. In contrast to SN Yerba Loca, the RRI value for MN El Morado was strongly negative, indicating that this reserve has low efficiency in terms of concentrating species. These two contrasting situations support the hypothesis that protected areas established on the basis of wild and scenic value will span the gamut of situations from containing very high to low levels of species richness. Of course, we recognize that our sample size is very limited at this stage and that it would be convenient to perform this analysis on a broader set of protected areas _ however, this is the only data available for the study area at this stage.

On face value, SN Yerba Loca could be considered a "hotspot" and MN El Morado a "coldspot" within the Metropolitan Region (c.f., Veech 2000). However it is important to separate a utilitarian view of the efficiency of a piece of land in concentrating richness, as has been the main focus of this paper and that of Veech (2000), from the issue of underlying intrinsic richness. The wide elevational range, presence of steep, opposite and mainly north-south facing valleys, and two large areas of wetlands undoubtedly contribute strongly to the high species richness in SN Yerba Loca. The low efficiency of MN El Morado, on the other hand, reflects its higher average elevation, and the much lower elevational range covered by the reserve (Teillier et al. 1994). Bowers & McLaughlan (1982) studied the influence of area, elevational range (and collecting effort) for twenty local floristic lists in the state of Arizona, USA. These authors came to the conclusion that elevational range was a much better predictor of species richness than area. In their study, elevational range and collecting effort accounted for 77 % of the variation in species richness. If SN Yerba Loca were a fairly flat piece of land, undoubtedly it richness would descend considerably. While SN is clearly an excellent conservation choice in relation to the regional average for its land area, it would be unadvisable at this stage to assume that is a true hotspot within the Metropolitan Region. Likewise, MN El Morado might not be a cold spot in relation to other areas in the high Andes in the Metropolitan Region. Before any conclusion of exaggerated or lower than average intrinsic richness can be accepted, additional studies of areas covering similar elevational ranges in the Andes in the Metropolitan Region will be necessary. This brings home the related point that the detection of relative richness will always be conditioned by the data available. Each time a new area is added to the species-area regression, the RRI values of individual areas can expect to change somewhat. Clearly, access to a stable regression equation is essential for reaching solid conclusions. It is presently unknown how much raw floristic data will be necessary to produce a stable species-area relationship in central Chile.

We argued that a well placed reserve will be one that concentrates a high number of endemic taxa, high numbers of narrowly distributed species, and high numbers of species with marginal populations. Endemic taxa can be considered a region's unique biodiversity. Narrowly distributed species, whether endemic or not, in general will be more vulnerable to severe anthropogenic disturbances outside reserves, and thus will be more prone to extinction (Holsinger & Gottlieb 1991). Particularly, this will be the case if rarity is combined with limited range size (Gaston & Kunin 1997). The conservation of representative populations from a species entire geographic range is an ideal conservation target. In particular, it is probably true to say that marginal populations tend to be overlooked. Marginal populations can contain unusual genotypes, adapted to extreme conditions. Moreover, increasing attention is being given to marginal populations on account of the role such populations could play under global climatic change (Arroyo et al. 1993). While SN Yerba Loca excels in terms of efficiency in housing species, the reserve was found to be less notable in terms of concentrating outstanding numbers of endemic species, species with marginal distributions and species with smaller geographical ranges. Nevertheless, it was seen that the outstanding species richness in the reserve compensated the absolute numbers of species in these categories to a large extent, thus warranting the conclusion that SN Yerba Loca is also a fairly efficient reserve on the basis of these criteria. The generally lower percentages of endemic taxa in the reserve suggest that greater concentrations of endemic taxa will be found at lower elevations in central Chile. Undoubtedly, the large component of high alpine species in the reserve has a strong influence on all categories of endemism. In general, a large proportion of the high Andean flora is shared with eastern slopes of the Andes in Argentina, thus lowering the percentage of endemism rate of reserves situated in the high Andes (on the continental Chile criterion used here). For conservation assessments it would be preferable to eliminate the traditional endemism concept and replace it for some quantitative measure of the size of a species distribution (c.f., Gaston 1994). In this study a first step in that direction was taken by employing a crude measure of range size (number of administrative regions occupied). Range size measured this way was generally larger in the SN Yerba Loca flora than in the regional floras. Our results provide the first quantitative data suggesting that components of the high Andean flora have larger geographical ranges than low elevation species in central Chile. Such a tendency is not unexpected, given that the high alpine area of the Andes forms a fairly continuous corridor along which species can migrate with ease from north to south. To advance in this general area of southern South American biogeography, floristic data for Argentina organized in electronic form is badly needed.

Our study of SN Yerba Loca invites additional comments. That over a third of the native flora of the Metropolitan Region and over a quarter of that of the Metropolitan-Fifth regions combined were found in SN Yerba Loca on a land area equivalent to 0.7 and 0.3 % of the size of the regional areas suggests that relatively small land areas can house considerable plant biodiversity. These same data suggest relatively low turnover in the mediterranean flora of central Chile. In general mediterranean floras tend to be characterized by much local endemism (Cowling et al. 1996). The fairly large distributional ranges of the species in the two regional floras, added to what appears to be low floristic turnover, suggests that local endemism might be less evident in the Mediterranean-type climate flora of central Chile in comparison with most other Mediterranean-type climate areas. Indeed, the proportion of regional endemics with more local distributions, is not particularly high. Moreover, several genera that have not been revised taxonomically in recent times figure strongly among the more local endemics (e.g., Viola). Chilean botany has been characterized by an excess of synonyms as a result of the earlier naturalists sampling disjunct portions of continuous ranges and giving them specific epithets. Thus a number of the more local endemics could eventually turn out to be artifacts of taxonomy. This suggests the hypothesis that endemism in the Mediterranean area of central Chile might be more strongly conditioned by the strength of the Andean barrier than rapid local evolution or adaptation to particular edaphic conditions. Arroyo et al. (1995) previously showed that the oceanic climate of central Chile is associated with a strongly woody Mediterranean flora containing many signs of evolutionary conservatism. The Mediterranean-type climate area also has significantly fewer annual species compared with climatically-similar California (Arroyo et al. 1995). Additionally, in relative terms, it has already been shown that intrinsic species-richness in central Chile is lower than in other Mediterranean-type climate areas (Arroyo et al. 1995, Cowling et al. 1996). Demonstration of lower levels of local endemism would constitute another expression of evolutionary conservatism in the Chilean Mediterranean flora. Low turnover and the presence of fairly large geographic ranges in many species, of course, is positive from a conservation viewpoint, since many individual species found in a reserve like SN Yerba Loca will have a high chance of falling into existing protected areas beyond the Metropolitan Region. The latter is highly relevant for the conservation of the mediterranean-type climate flora of central Chile in general, where less than 5 % of the land area is contained in the state protected area system in comparison with 19 % for all of continental Chile (Arroyo & Cavieres 1997).

A long range objective of our present research efforts concerns determining what proportion of the Mediterranean flora of central Chile occurs in the national protected area system, and to what extent individual species' geographical ranges are covered. We are also engaged in detecting areas of high biodiversity content on the landscape outside existing reserves so as to allow objective decisions in central Chile insofar as the establishment of new protected areas is concerned (c.f., Castro-Parga et al. 1996, Williams 1996, Araujo 1999). These goals are being pursued at different spatial scales: individual political regions, climatic divisions with the Mediterranean area and Mediterranean area as a whole. In this context, some preliminary comments concerning the conservation status of the vascular plant flora of the Metropolitan Region can now be made. The Metropolitan Region contains four national protected areas: SN Yerba Loca, MN Nacional El Morado, Reserva Nacional Río Clarillo (all in the Andes) and Reserva Nacional Robleria de Loncha (in the Coast Range). These reserves comprise a mere 1.9 % of the area of the Metropolitan Region. Thus far (including the present results) reliable floristic surveys are available for two of the four protected areas. Table 6 amalgamates floristic information for SN Yerba Loca and MN El Morado, which together comprise 0.9 % of Metropolitan Region land area. Conservatively, 39 % of the known native vascular plant flora of the Metropolitan Region (and 30% of the flora of the Metropolitan-Fifth regions combined) is contained in the two reserves. The remaining two Metropolitan Region reserves, which extend to lower elevations, undoubtedly will contain many lowland species not found in RN Yerba Loca and MN El Morado. However, many of the species in RN Río Clarillo found above 1300 m can be expected to be shared with one or more of the two inventoried reserved. All things being equal, it is likely that a good number of species found in the Metropolitan Region are not contained in any national protected area in the region.

Finally, a review of the species richness data (Arroyo et al. 2001) suggests that SN Yerba Loca contains many more native species than any other reserve (Bellotos de Melado, Fray Jorge and Talinay, Contulmo area) in central Chile studied to date. Outside of the mediterranean-type climate area, this trend also holds up. Parque Nacional Puyehue in the Tenth Region contains less than 300 native species (Muñoz 1980). Northern parks for which comprehensive species lists are available (Parque Nacional Pan de Azúcar, Rundel et al. 1966; Parque Nacional Llullaillaco, Arroyo et al. 1998) also exhibit lower richness. Parque Nacional Pan de Azúcar situated in the coastal desert, considered to be a floristically-rich area, contains 206 native vascular plants species in an area of 43,754 ha. Parque Nacional Llullaillaco in the hyper-arid sector of the northern Andes (24-25º S) contains less than 100 species in 262,000 ha. In making these comparisons, it must be recognized that the sample size is still very small: as mentioned earlier continental Chile has some 90-odd protected areas (Muñoz et al. 1996). Moreover, the protected areas for which species richness data is presently available vary enormously in size and include units that are much smaller than SN Yerba Loca (e.g., RN Bellotos del Melado). Thus it cannot be concluded at this stage that SN Yerba Loca ranks highest among all protected areas in Chile in terms of species richness in relation to area. Notwithstanding the latter, the reserve does contain a very large flora, and is very rich in relation to the regional average. It thus deserves special attention from a management perspective. Yearly, SN Yerba Loca is receiving increasing numbers of visitors from the Metropolitan Region, with a population of close to six million (2002 Census). Transhumance cattle herding was practiced in the area of the reserve before it was decreed a Nature Sanctuary; today cows and horses can still be found in the reserve which is only partially fenced off from surrounding lands. In our opinion, a concerted effort is needed to assure the integrity of SN Yerba Loca.

ACKNOWLEDGMENTS

Work supported by FONDECYT grant 1980705 (MKA-OM) and an Endowed Presidencial Science Chair (Cátedra Presidencial en Ciencias (MKA). This paper forms part of the research and outreach activities of the "Millennium Center for Advanced Studies in Ecology and Research on Biodiversity", supported by grant No. P99-103-F ICM, and of an IBOY (International Biodiversity Observation Year) satellite project. Many assistants participated in the field work. Foremost, we acknowledge Ana Maria Humaña for her outstanding dedication, organizational talent, and tireless effort. We also thank Alejandro Muñoz, Maritza Mihoj, Carlos Valdivia, Fernanda Pérez, Francisca Diaz, Peter McPherson, Antonio Maldonado, Carmen Castor, and Denise Rougier. We apologize if we have overlooked any volunteer helper. Park guides at SN Yerba Loca and Sr. Carlos Weber, Executive Director of Corporación Nacional Forestal are thanked for their support.

LITERATURE CITED

ARAUJO MB (1999) Distribution patterns of biodiversity and the assessment of a representative reserve network in Portugal. Diversity and Distribution 5: 151-163.

ARMESTO JJ, R ROZZI, C SMITH-RAMÍREZ & MTK ARROYO (1988) Conservation targets in South American temperate forest. Science 282: 1271-1272.

ARROYO MTK & P USLAR (1993) Breeding systems in a temperate mediterranean-type climate montane sclerophyllous forest in central Chile. Botanical Journal of the Linnean Society 111: 83-102.

ARROYO MTK & L CAVIERES (1997) The Mediterranean-type climate flora of central Chile - What do we know and how can we assure its protection? Noticiero de Biología (Chile) 5: 48-56.

ARROYO MTK, J ARMESTO & C VILLAGRÁN (1981) Plant phenological patterns in the high Andean cordillera of central Chile. Journal of Ecology 69: 205-233.

ARROYO MTK, R PRIMACK & JUAN ARMESTO (1982) Community studies in pollination ecology in the high temperate Andes of central Chile. I. Pollination mechanisms and altitudinal variation. American Journal of Botany 69: 82-97.

ARROYO MTK, JJ ARMESTO, F SQUEO & J GUTIÉRREZ (1993) Global change: The flora and vegetation of Chile. In: Mooney HA, R Fuentes & BI Kronberg (eds) Earth system response to global change: contrasts between North and South America: 239-263. Academic Press, San Diego, California.

ARROYO MTK, L CAVIERES, C MARTICORENA & M MUÑOZ (1995) Convergence in the mediterranean floras of central Chile and California: insights from comparative biogeography. In: Arroyo MTK, M Fox & P Zedler (eds) Ecology and biogeography of Mediterranean ecosystems in Chile, California and Australia: 43-88. SpringerVerlag, New York, New York.

ARROYO MTK, C CASTOR, C MARTICORENA, M MUÑOZ, L CAVIERES, O MATTHEI, F SQUEO, M GROSJEAN & R RODRÍGUEZ (1998) The flora of Llullaillaco National Park located in the transitional winter-summer rainfall area of the northern Chilean Andes. Gayana Botánica (Chile) 55: 93-110.

ARROYO MTK, R ROZZI, J SIMONETTI, P MARQUET & M SALABERRY (1999) Central Chile. In: MITTERMEIER RA, MYERS N & CG MITTERMEIER (eds) Hotspots: Earth's biologically richest and most endangered terrestrial ecosystems: 161-171. CEMEX, México, Distrito Federal.

ARROYO MTK, C MARTICORENA, O MATTHEI & L CAVIERES L (2000) Plant invasions in Chile: present patterns and future predictions. In: Mooney HA & R Hobbs (eds) Invasive species in a changing world: 385-421. Island Press, Washington, District of Columbia.

ARROYO MTK, O MATTHEI, C MARTICORENA, M MUÑOZ, F PÉREZ & AM HUMAÑA (2000) The vascular plant flora of the Bellotos del Melado National Reserve, VII Region, Chile: a documented checklist. Gayana Botánica (Chile) 57: 117-139.

BENOIT C (1996) Representatividad ecológica del Sistema Nacional de áreas Protegidas del Estado. In: Muñoz M, H Nuñez, J Yánez (eds) Libro rojo de los sitios prioritarios para la conservación de la diversidad biológica en Chile: 149-153. Ministerio de Agricultura, Corporación Nacional Forestal, Santiago, Chile.

BOWERS JE & SP MCLAUGHLIN (1982) Plant species diversity in Arizona. Madroño 29: 227-233.

BRAKO L & JL ZARRUCHI (1993) Catalogue of the flowering plants and gymnosperms of Peru. Missouri Botanical Garden, Saint Louis, Missouri. 1986 pp.

CASTRO-PARGA I, JC MORENO-SAIZ CJ HUMPHRIES & PH WILLIAMS (1996) Strengthening the natural and national park system of Iberia to conserve vascular plants. Botanical Journal of the Linnean Society 121: 189-206.

CAVIERES LA, A PEÑALOZA & MTK ARROYO (2000) Altitudinal vegetation belts in the high Andes of central Chile (33ºS). Revista Chilena de Historia Natural 73: 331-344.

CONAF-CONAMA-BIRF (1999) Catastro y evaluación de recursos vegetacionales nativos de Chile. Informe Nacional con Variables Ambientales. 11 Volumes. Comisión Nacional del Medio Ambiente, Santiago, Chile.

COWLING RM, PW RUNDEL, BB LAMONT, MTK ARROYO & M ARIANOUTSOU (1996) Plant diversity in mediterranean-climate regions. Trends in Ecology and Evolution 11: 362-366.

DI CASTRI F & ER HAJEK (1976) Bioclimatología de Chile. Dirección de Investigación, Vicerectoría Académica, Ediciones Universidad Católica de Chile, Santiago, Chile. 128 pp.

EHRHART C (2000) Die Gattung Calceolaria (Scrophulariaceae) in Chile. Bibliotheca Botanica 153: 1-283.

GAJARDO R (1995) La vegetación natural de Chile. Clasificación y distribución geográfica. Second edition. Editorial Universitaria, Santiago, Chile. 165 pp.

GASTON KJ (1994) Rarity. Chapman & Hall, London, United Kingdom. 205 pp.

GASTON J & WE KUNIN (1997) Concluding comments. In: Kunin WE & KJ Gaston (eds) The biology of rarity: 262-272. Chapman & Hall, London, United Kingdom. 280 pp.

GROOMBRIDGE B (ed) (1992) Global biodiversity. Status of the earth's living resources. Chapman & Hall, London, United Kingdom. 585 pp.

HEYWOOD VH & RT WATSON (eds) (1995) Global biodiversity assessment. Cambridge University Press, Cambridge, United Kingdom. 1140 pp.

HOLSINGER KE & LD GOTTLIEB (1991) Conservation of rate and endangered plants: principles and prospects. In: Falk DA & KE Holsinger (eds) Genetics and conservation of rare plants. Oxford University Press, New York, New York. 283 pp.

JORGENSEN PM & S LEÓN-YÁNEZ (eds) (1999) Catalogue of the vascular plants of Ecuador. Missouri Botanical Garden, Saint Louis, Missouri. 1181 pp.

MARGULES GR, AO NICHOLS & RL PRESSEY (1988) Selecting networks of reserves to maximise biological diversity. Biological Conservation 43: 63-76.

MATTHEI O (1995) Manual de las malezas que crecen en Chile. Alfabeta Impresores, Santiago, Chile. 545 pp.

MILLER K, MH ALLEGRETTI, N JOHNSON & B JONSSON (1995) Measures for conserving biodiversity and sustainable use of its components. In: Heywood V & RT Watson (eds) Global biodiversity assessment: 915-1061. Cambridge University Press, Cambridge, United Kingdom.

MITTERMEIER RA, MYERS N & CG MITTERMEIER (eds) Hotspots: Earth's biologically richest and most endangered terrestrial ecosystems. CEMEX, México, Distrito Federal. 429 pp.

MUÑOZ M (1980) Flora del Parque Nacional Puyehue. Editorial Universitaria, Santiago, Chile. 557 pp.

MUÑOZ M, H NÚÑEZ & J YÁÑEZ (eds) (1996) Libro rojo de los sitios prioritarios para la conservación de la diversidad biológica en Chile. Ministerio de Agricultura, Corporación Nacional Forestal, Santiago, Chile. 203 pp.

MYERS N, RA MITTERMEIER, CG MITTERMEIER, G DA FONSECA & J KENT (2000) Biodiversity hotspots for conservation priority. Nature 403: 853-858.

PEÑALOZA A, LA CAVIERES, MTK ARROYO & C. TORRES (2001) Efecto nodriza intra-específico de Kageneckia angustifolia D. Don (Rosaceae) sobre la germinación de semillas y sobrevivencia de plántulas en el bosque esclerófilo montano de Chile central. Revista Chilena de Historia Natural 74: 538-548.

RUNDEL PW, MO DILLON & B PALMA (1996). Flora y vegetación del Parque Nacional Pan de Azúcar en el desierto de Atacama. Gayana Botánica (Chile) 53: 295-315.

SOKAL RR & FJ ROHLF (1981) Biometry. Third edition. W. H. Freeman and Company, New York, New York. 887 pp.

SQUEO FA, G ARANCIO, JR GUTIÉRREZ (eds) (2001) Libro rojo de la flora nativa y de los sitios de prioritarios para su conservación: Región de Coquimbo. Ediciones Universidad de la Serena, La Serena, Chile. 372 pp.

TEILLIER S, A HOFFMANN, F SAAVEDRA & L PAUCHARD (1994) Flora del Parque Nacional El Morado (Región Metropolitana, Chile). Gayana Botánica (Chile) 51: 13-47.

VAN HUSEN C (1967) Klimagliederung in Chile auf der Basis von Häufigkeitsverteilungen der Niederschlagssummen. Freiburger Geographische Hefte 4: 1-13.

VEECH JA (2000) Choice of species-area function affects identification of hotspots. Conservation Biology 14: 140-147.

WILLIAMS PH (1996) WORLDMAP 4 WINDOWS. Software and help document 4.1. Privately Distributed. London, United Kingdom. 521 pp.

WILSON EO & FM PETERS (eds) (1988) Biodiversity. National Academy of Sciences Press, Washington, District of Columbia. 521 pp.

ZULOAGA FO, EG NICORA, ZE RÚGOLO DE AGRASAR, O MORRONE, J PENSIERO & AM CIALDELLA (1994) Catálogo de la familia Poaceae en la República Argentina. Missouri Botanical Garden, Saint Louis, Missouri. 178 pp.

ZULOAGA FO & O MORRONE (eds) (1996) Catálogo de las plantas vasculares de la República Argentina I. Missouri Botanical Garden, Saint Louis, Missouri. 323 pp.

ZULOAGA FO & O MORRONE (eds) (1999) Catálogo de las plantas vasculares de la República Argentina II. Missouri Botanical Garden, Saint Louis, Missouri. 1269 pp.

Associate Editor: P. Marquet
Received June 7, 2001; accepted September 4, 2002