Tree community patterns along a deciduous to evergreen forest
gradient in central Bolivia

Patrones de comunidades de árboles en un gradiente de bosque
semideciduo a bosque siempreverde en Bolivia central

R. Linares-Palomino, V. Cardona, D. soto, S.K. Herzog, M. Kessler

¹Department of Systematic Botany, Albrecht-von-Haller-Institute for Plant Sciences,
University of Göttingen, Untere Karspüle 2, 37073-Göttingen, Alemania

2 Herbario Nacional de Bolivia, Instituto de Ecología, Universidad Mayor de San Andrés, ]]> Casilla 10077, La Paz, Bolivia

³Herbario del Oriente Boliviano, Museo Noel Kempff Mercado, Santa Cruz, Bolivia

⁴Asociación Armonía - BirdLife International, Casilla 3566, Santa Cruz, Bolivia

⁵Present address: Institute of Systematic Botany, University of Zürich,
Zollikerstrasse 107, CH-8008 Zürich, Suiza
* Autor de correspondencia: e-mail: r.linaresp@yahoo.co.uk

Abstract

Keywords: Amboró National Park, seasonal forest, tree density, tree dominance, tree inventory.

Resumen

Estudiamos tres parcelas de una hectárea cada una incluyendo todos los árboles con diámetro a la altura del pecho de 10 cm o más en el ecosistema estacional del Refugio Los Volcanes (1.000-

1.400 m de altitud), situado en el centro de Bolivia. Las parcelas incluyeron variantes locales de bosques deciduos, semideciduos y siempreverdes. Encontramos un total de 115 especies y 43 familias. La parcela más diversa fue la semidecidua con 70 especies, 56 géneros y 31 familias; con bosque siempreverde tuvo valores de diversidad similares o ligeramente más bajos (63 especies, 49 géneros, 31 familias); mientras la que presenta bosque deciduo tuvo valores mucho más bajos (44 especies, 38 géneros y 21 familias) y compartía sólo 19 y 15 especies con las parcelas en el bosque semideciduo y siempreverde, respectivamente. Entre un 34% y 50% de las especies en las parcelas eran localmente raras con sólo uno o dos individuos por parcela. Por otro lado, un alto porcentaje del número total de individuos y del área basal era contribuido por unas pocas especies dominantes en cada parcela. Pachystroma longifolium (Nees) I.M. Johnst. (Euphorbiaceae) fue la única especie que estaba entre las tres especies más dominantes en cada una de las tres parcelas. Leguminosae fue la familia más importante en las parcelas de bosque deciduo y semideciduo. El bosque siempreverde contenía un conjunto de familias dominantes distintas a las otras parcelas. Los valores de riqueza de especies fueron similares a otros estudios de inventarios forestales en ecosistemas estacionales comparables realizados en Bolivia. La composición de especies y familias por otro lado, fue diferente, exceptuando el hecho bien conocido que Leguminosae es la familia dominante en bosques estacionalmente secos del Neotrópico. Fitogeográficamente, la flora arbórea de las parcelas de Los Volcanes es heterogénea, estando compuesta de elementos de bosques estacionalmente secos así como de los bosques tropicales húmedos.

Palabras clave: Amboró, Bosque estacional, Densidad de árboles, Dominancia de árboles, Inventario forestal.

Introduction

The Amboró National Park located in the Santa Cruz department (Bolivia) must be one of the most biodiverse protected areas in the world with vegetation types such as lowland humid evergreen forests, inter-Andean dry forests, high altitude cloud forests and puna vegetation from which more than 3,000 plant species have been reported to date (Nee 2004). The southeastern part of the national park is mainly composed of deciduous vegetation and previous classifications have tended to include it in single broad categories. Navarro et al. (1996) for example, included this area in the ‘subhumid to humid deciduous forest of southeastern Amboró’, a vegetation type usually found between 900-1,100 m and where the dominant tree species are Aspidosperma cylindrocarpum (Apocynaceae), Cariniana estrellensis (Lecythidaceae), Cedrela lilloi (Meliaceae),

Gallesia integrifolia (Phytolaccaceae), Pachystroma longifolium (Euphorbiaceae), Pogonopus tubulosus (Rubiaceae), and Tabebuia lapacho (Bignoniaceae). More recently, Navarro & Maldonado (2002) assigned the vegetation in this area to the ‘upper central subandean Cerrado semideciduous forest’, found between 1,000-1,900 m and with P. longifolium and C. estrellensis dominating the 1,000-1,300 m range. Based on visual inspections of the area, however, we observed substantial variations of the vegetation within these broad categories. During a multidisciplinary study on the ecology of animals and plants at the Refugio Los Volcanes (hereafter LV, an area adjacent to the south of Amboró National Park), we observed that the dominant zonal vegetation in the area was indeed semi-deciduous forest (about 30-50% deciduous trees), mainly found on the shaded south-facing slopes. Local differences in topography and aspect, however, seemed to lead to ecologically

relevant differences in water availability. Consequently, steep, sunny and north-facing slopes were occupied by deciduous forest (70-90% deciduous), whereas the fat, shaded valleys with groundwater supply supported evergreen forest (10-20% deciduous). In order to test these visual observations we established three permanent one-hectare plots along the deciduous to evergreen seasonal forests vegetation gradient of LV. Our aims were to characterise the forest types within the LV area, to understand the relationship between these distinct types using quantitative tree inventories and to investigate the wider biogeographic affinities (within Bolivia and beyond) of the LV forest.

The study was carried out at the Refugio Los Volcanes (1,000-1,400 m altitude) in Santa Cruz, Bolivia. LV is located at about 18°06’ S and 63°36’ W, adjacent to the southern border of Amboró National Park, on the transition from the humid inner tropics to the seasonally dry subtropics (Fig. 1). The geological substrate of the study area consists primarily of red sandstone and locally of loamy sedimentary rocks (lutite). Annual precipitation is about 1,200-1,400 mm, with a three month dry season from July to October. The deciduous forest plot (18° 06’ 48’’S, 63° 36’ 10’’W) was located at an average altitude of 1,150 m on a sunny north­facing slope. The semi-deciduous forest plot (18° 06’ 13’’S, 63° 35’ 44’’W, 1,300 m northeast of the deciduous forest plot) was located at an average altitude of 1,070 m on a south­facing slope shaded by a rock-wall toward the southeast. The evergreen forest plot (18° 06’ 03’’S, 63° 35’ 58’’W, 500 m northwest of the semi-deciduous forest plot) was located at an average altitude of 1,100 m in a ravine with loose soils and permanent groundwater supply provided by a small watercourse that divided the plot into two parts with steep slopes on both sides of the creek.

Methods

We established one permanent plot of 1 ha in each forest type in 2002-2003. We subdivided each plot into 25 adjacent 20 m x 20 m subplots, which where laid out in such a way as to include only a homogeneous representation of the forest type under study (i.e., avoiding secondary vegetation or other forest types). Thus, our plots are not the traditional square 100 m x 100 m inventory plots, but have rather irregular shapes (Fig. 1). We recorded diameter at breast height (dbh) for each tree with a diameter of 10 cm or more. Voucher specimens of all species were collected for later taxonomic determination and are deposited at USZ (Santa Cruz) and LPB (La Paz) (herbarium acronyms following Holmgren & Holmgren 1998). Identification of the most common species was done directly in the field whenever possible. Vouchered specimens were identified at USZ and LPB using local (e.g. Jardim et al. 2003) and national tree floras (Killeen et al. 1993), as well as by comparison with reference collections and expert identification from local and visiting botanists at the mentioned herbaria. Sterile specimens that could not be completely identified were sorted into morphospecies.

We calculated relative density, relative dominance (using basal area) and relative frequency for each species in order to estimate the importance value index (IVI) following Curtis & McIntosh (1951). To calculate the family value index (FVI) we followed Mori et al. (1983) and used relative density, relative dominance and relative diversity (number of species per family). We constructed smoothed individual based species accumulation curves by randomly adding individuals to the species accumulation curve. This procedure was repeated 100 times and we used the mean and 1.96 standard deviations to plot the curve and 95% confidence intervals (Magurran 2004).

Figure 1. Study area in Refugio Los Volcanes, Santa Cruz (inset shows location within Bolivia). Plot shape and approximate location are given by D (deciduous forest), SD (semi deciduous forest) and E (evergreen forest).

Results
Diversity and density

We recorded 115 tree species from 43 families on the three 1-ha plots (Appendix 1). The semi­deciduous and evergreen forest plots were the most species-rich (Table 1). In contrast, density values were much higher on the deciduous forest plot, where multi-stemmed trees were about twice as common as on the other two plots. A higher percentage of species represented by a single individual (singleton) and by two

We found no lianas or hemi-epiphytes with a dbh ≥10 cm on the deciduous forest plot. The semi­deciduous forest plot contained five individuals of hemi-epiphytes (Ficus sp., F. obtusifolia Kunth in Humb., Bonpl. & Kunth and F. trigona L. f., Moraceae) and two of lianas (Machaerium cf. punctatum (Poir.) Pers., Leguminosae and Celtis iguanaea (Jacq.) Sarg., Ulmaceae). The evergreen

forest plot contained eight individuals of hemi­epiphytes (Ficus cf. eximia Schott in Spreng., F. maxima Mill., F. obtusifolia Kunth in Humb., Bonpl. & Kunth, F. pertusa L. f. and F. trigona L. f., Moraceae) and no lianas.

Individual based species accumulation curves showed that when up to ca. 180 individuals are sampled randomly from each plot, confidence intervals of the estimatedspecies richness values overlap, indicating that the observed differences could have arisen by chance (Fig. 2). Increased random sampling of individuals, however, revealed clear differences between the deciduous forest plot and the two other forest plots. Nevertheless, irrespective of increased sampling these two latter forest types showed very similar estimated species richness values up to the total individuals sampled.

Figure 2. Individual based species accumulation curves of the deciduous (black), semi-deciduous (red) and evergreen (green) forest plots. Lines indicate estimated species richness values by given number of individuals; polygons indicate a 95% confdence interval.

Diameter distribution in size classes

The range of dbh values increased from the deciduous forest plot (10-83 cm) to the evergreen forest plot (10-162 cm), with intermediate values on the semi-deciduous forest plot (10-125 cm). Basal area was highest on the semi-deciduous and deciduous forest plots with 27.7 _m2 and 27.5 m², respectively. The evergreen forest plot had a somewhat lower basal area with 24.2 m². All three plots showed a left-skewed “inverted J” distribution of the number of individuals per dbh size class (Fig. 3). Nevertheless, there were differences in the dbh size class that contributed most to total basal area (Fig. 3). Dbh size classes from 10-50 cm contributed most on the deciduous forest plot (with 11.4 to 14.9% of total basal area per size class). On the semi-deciduous forest plot, the 30-40 cm dbh size class contributed with 22.6% of total basal area. On the evergreen forest plot low dbh size classes, as well as high dbh classes, were responsible for most of the total basal area (dbh size classes 10-15 cm and 15-20 cm contributed with 12.1% and 11.4% of total basal area, respectively. The dbh size class of 100 cm or more contributed with 13.1%).

Dominant species and families

contributed with 24% of the total number of individuals and 41% of total basal area. Together these six species represented the 54% of all individuals and 62% of total basal area. The most important family was Leguminosae with 32% of the individuals, 12 species and 39% of the basal area, followed by Meliaceae and Euphorbiaceae. These three families contributed with 58% of the number of individuals, 54% of basal area and 40% of species richness (Table 2).

The semi-deciduous forest plot (Table 3) was dominated by Pachystroma longifolium and Myrciantes cf. pseudomato (Legrand) Mc. Vough. (Myrtaceae). Pachystroma longifolium had the highest number of individuals (118, 21% of the total), and both species had the highest basal area, comprising 38% of total basal area. Other important species in terms of density were Chomelia cf. sessilis Mull. Arg. (Rubiaceae) and Chrysophyllum gonocarpum (Mart. & Eichl.) Engl. (Sapotaceae). The most important family was Euphorbiaceae, in which only two species contributed 25% of the individuals and 21% of basal area, followed by Sapotaceae and Myrtaceae. These three families comprised 47% of the number of individuals, 55% of basal area and 13% of species richness (Table 4).

The evergreen forest plot (Table 5) was dominated by Aniba sp. (Lauraceae), with 76 individuals (14% of the total) and 8% of total basal area. Drypetes amazonica Steyerm. (Euphorbiaceae), Pachystroma longifolium and Clarisia biflora Ruiz & Pav. (Moraceae) were also important. These four species contributed 40% of the total number of individuals and 26% of total basal area. The most important family was Moraceae with 74 individuals (14% of the total), 25% of total basal area and eight species, followed by Lauraceae, Sapotaceae and Euphorbiaceae. These four families represented 66% of the individuals, 70% of basal area and 35% of species richness (Table 6).

Figure 3. Number of individuals (bars) and percentage contribution of total basal area (lines) per diameter size class on three 1-hectare forest plots in central Bolivia. Note different scales of the ordinates for each graph.

Table 2. Ten most dominant woody plants with dbh ≥10cm in the deciduous forest plot according to importance value indices (species are ranked by descending importance value index). Legend: # Ind = no. of individuals, rDe = relative density, BA = basal area, rDo = relative dominance, F = frequency, rF = relative frequency, IVI = importance value indices.

Table 3. Ten most dominant families of woody plants with dbh ≥10cm in the deciduous forest plot according to family value indices (families are ranked by descending family importance value). Legend: # Ind = no. of individuals, rDe = relative density, BA = basal area, rDo = relative dominance, Div = species diversity, rDiv = relative diversity, FIV = family importance value.

Table 5. Ten most dominant families of woody plants with dbh ≥10cm in the semi-deciduous forest plot according to family value indices (families are ranked by descending family importance value). Legend: # Ind = no. of individuals, rDe = relative density, BA = basal area, rDo = relative dom inance, Div = species diversity, rDiv = relative diversity, FIV = family importance value.

Discussion

Effect of plot shape on species richness and community composition estimations

Spatial distribution patterns of plant species in tropical forests have been reported to show mostly clumped distributions (e.g., Condit et al. 2000). Within this scenario, longer and narrower rectangular plots are prone to capture more species than square plots of similar area, because the former could include a potentially more heterogeneous area (Condit et al. 1996, Laurance et al. 1998). Nevertheless, small and statistically non-significant differences were found in a study comparing tree diversity in square plots versus rectangular plots in Central Amazonia (Laurance et al. 1998). Given the patchy and fragmented nature of the deciduous

and evergreen forests in our study area, it was impossible to survey all vegetation types in regular square or rectangular plots. Rather, we tried to survey as environmentally homogeneous an area as possible, leading to our irregular plot shape design. We acknowledge that plot shape has influenced our results, especially those of the transect-like evergreen forest plot. The extent of this influence, however, seems to be variable, as can be seen from the species accumulation curves. It significantly decreased in the irregularly shaped deciduous forest plot when the hectare was completely surveyed, suggesting that sampling was representative. In contrast, the species accumulation curves did not level off in the more-or-less regularly shaped semi-deciduous plot and in the long and transect-like evergreen forest plot, suggesting, that sampling, irrespective of plot shape, was incomplete.

Table 6. Ten most dominant woody plants with dbh ≥10cm in the evergreen forest plot according to importance value indices (species are ranked by descending importance value index). Legend: # Ind = no. of individuals, rDe = relative density, BA = basal area, rDo = relative dominance, F = frequency, rF = relative frequency, IVI = importance value indices.

Local variation of tree diversity, dominance and structure at Los Volcanes >

There were obvious differences in the diversity and floristic composition between the deciduous forest plot and the two other plots, which where rather similar in diversity and structure. Not only was the deciduous forest plot poorer in terms of generic and familial diversity, it also shared fewer species with the semi-deciduous and the evergreen forest plot. The deciduous forest plot shared 19 species with the semi-deciduous and 15 with the evergreen forest plot, whereas the latter two shared 39 species. Eleven species were present in all three plots and represented a selection of the tree species characteristic of the Amboró region like Aspidosperma cylindrocarpum, Gallesia integrifolia and Pachystroma longifolium (Navarro et al. 1996). Few of these species were dominant and abundant on each of the plots, such as Chrysophyllum gonocarpum. On the contrary, most were locally rare species, at least at the hectare-scale, with abundances

of one individual per hectare (e.g., Cariniana estrellensis (Raddi) Kuntze, Swartzia jorori Harms, Eugenia sp.).

Differences in diversity and composition between the three plots, especially between the deciduous forest plot and the two other, was mirrored by several environmental factors. For example, soil humidity and pH increased from the deciduous forest plot to the evergreen forest plot (Lendzion 2003). Therefore, even though there is a dominant regional flora, small-scale environmental variations produce significant differences in the composition and dominance of the local tree flora. This fact has already been shown for terrestrial herbs at LV (Lendzion 2003) and for woody species in seasonal forests elsewhere (Oliveira-Filho et al. 1998, Balvanera et al. 2002).

Lianas have been considered important components of tropical forests (Medina 1995), constituting in some cases 20% or more of the local flora (Gentry 1995, Schnitzer & Bongers 2002). In this study, however, we were not able to record lianas with a dbh ≥ 10 cm in the deciduous and evergreen forest plots. Only

the semi-deciduous forest plot contained two individuals belonging to two tree species (Machaerium cf. punctatum and Celtis iguanaea), which occasionally also grow as lianas, with a dbh ≥10 cm. This agrees with another study in Bolivian deciduous forests that found similar low levels of liana diversity and density using the same sampling methodology (Cayola et al. 2005). Nonetheless, this does not mean that the LV forests are devoid of vines and climbers. In fact, most of the liana diversity and abundance in each of the studied forest plots is found in dbh classes below 10 cm (Hoffmann 2004), so that our results have to be interpreted as a sampling artefact.

In mountainous areas, with forest growing on steep slopes, one of the factors than can influence and produce low basal area values are landslides (Chazdon 2003). Since the humid forest plot was located on steeper slopes than the other two plots and additionally had rather loose soils, we suggest that an increased frequency of localized landslides could have produced the lower basal area as compared to the other two plots.

The Los Volcanes forests in a ]]> Bolivian context

Our inventory on the deciduous forest plot contained species richness and densities of trees with dbh ≥ 10 cm within the range of other inventories in Bolivian seasonal forests (Table 8). It was much richer and denser than that of the Santa Cruz Botanical Garden (inventoried by Uslar et al. 2004), which is located in a slightly drier lowland area (350 m). Our plots in the semi-deciduous and evergreen forest variants, although much more species-rich, were less dense than the San Rafael (1,500 m, Vargas 1995) and Río Tuichi (880 m, Cayola et al. 2005) plots.

The dominant and most diverse family in seasonal deciduous forests in the Neotropics is usually Leguminosae (Gentry 1995, Penningtonet al. 2006). There is, however, ample variation as to which are the next most diverse and dominant families. This is confirmed by several other inventories in Bolivian deciduous forests (Killeen et al. 1998, Kessler & Helme 1999, Fuentes Claros et al. 2004) and by our study, where we had six different families following Leguminosae in importance. In nearby Santa Cruz (Uslar et al. 2004) Leguminosae was the dominant family (seven species), all other families with only one or two species (Table 8). In the adjacent Amboró National Park (Vargas 1995), they were Lauraceae and Myrtaceae. Finally, the Tuichi valley dry forests inventory in La Paz department (Cayola et al. 2005) reported also Myrtaceae and Bombacaceae as important. The dominant families on the evergreen forest plot in LV showed contrasting results with Moraceae, Lauraceae and Sapotaceae as the most diverse families (Table 8). The diversity of Moraceae, whose members are prominent components of tropical rain forests, already suggests the prevalence of different local environmental conditions on this plot.

In contrast, species composition and densities of trees with dbh ≥ 10 cm on the LV plots differed from other tree inventories in Bolivia with similar annual rainfall and similar seasonal vegetation. For instance, the Santa Cruz plot in the deciduous Chiquitania forests (Uslar et al. 2004) reported Aspidospermum cylindrocarpon (Apocynaceae), Myrciaria cauliflora (Myrtaceae) and Phyllostylon rhamnoides (Ulmaceae) as the most abundant species (Table 8). The Rio San Rafael plot in Amboró contained Nectandra sp. (Lauraceae), Casearia sp. (Flacourtiaceae) and Chrysophyllum cf. gonocarpum (Sapotaceae) as the most frequent species (Vargas 1995). The three most common species in the Tuichi plot were Phyllostylon rhamnoides (Ulmaceae), Trichilia catigua (Meliaceae) and Anadenanthera colubrina (Leguminosae) (Cayola et al. 2005). At LV, the most abundant species were Trichilia claussenii, Parapiptadenia excelsa and Pachystroma longifolium on the deciduous forest plot, Pachystroma longifolium, Chomelia cf. sessilis and Myrciantes cf. pseudo-mato on the semi­deciduous forest plot, and Aniba sp., Drypetes amazonica and Pachystroma longifolium on the evergreen forest plot (Table 8).

In terms of basal area, our deciduous and semi-deciduous forest inventories reached similar values of around 27 m2, which is comparable to other seasonal forest inventories in Bolivia, except for the Tuichi plot in La Paz (20 m²) and our evergreen forest plot in LV (24.2 m²). Murphy & Lugo (1986) stated that seasonal forests were less complex structurally than wet tropical forests, with basal area values in the range from 17 _m2 to 40 m². Several other studies have reaffirmed this statement (e.g., Lopes et al. 2008). Basal area values of the LV plots, however, were also within the range of other tree inventories in several vegetation types in South America, using the same diameter cut-off, and particularly within Bolivia. For instance, basal area values of several 1-ha inventories in humid tropical forests ranged between 21.4 _m2 (Beni, Bolivia) and 39.2 m2 (Amboró, Bolivia) (Smith & Killeen 1998).

Phytogeography of the Los Volcanes
species

Based on the distribution of over 70 plant taxa, Prado & Gibbs (1993) proposed the existence of a Pleistocenic arc of seasonal forest formations. The extant fragments of this arc are mainly located in three nuclei: the Caatinga nucleus in northeastern Brazil; the Misiones nucleus comprising the area from the Bolivia-Brazil border in Corumbá southwards to the Paraguay-Paraná river confluence and eastward to Santa Catarina in Brazil; and the piedmont nucleus extending from Santa Cruz de la Sierra in Bolivia southward to Tucumán and the sierras of Catamarca in Argentina. Because our study area is located within the area of the piedmont nucleus (cf. Prado 2000), we examined the distribution patterns of the 115 species found at LV. We recorded

Anadenanthera colubrina, Myracrodruon urundeuva Allemao and Celtis iguanaea in our plots, all of which have been reported as characteristic to the wider Pleistocenic arc formation (Prado & Gibbs 1993). We also found Diatenopteryx sorbifolia Radlk., classifed as a restricted Arc species since it only occurs in the Misiones and Piedmont nuclei. Navarro & Maldonado (2002), who considered the area around LV as transitional between the Brasilian-Parana biogeographic region and the Bolivian-Tucuman biogeographic province, also noted this affinity with the seasonal vegetation in the Misiones and piedmont nuclei. In total, we identified 21 species as typical of Bolivian seasonal forest formations. More complex and widespread distribution patterns are attributed to Cordia alliodora (Ruiz & Pav.) Cham. and Ximenia americana L., the later occurring in diverse vegetation types throughout the Neotropics.

On a local scale, the forests at LV shared only 21% (24 species) of all recorded tree species with the inventories in Santa Cruz, San Rafael and Tuichi. As expected, the deciduous forest plot shared most species (10 species, 23%) with the nearby (65 km) deciduous forests in the Santa Cruz Botanical Garden and less than 12% with the other two inventories (43 km to San Rafael, 660 km to Tuichi). In contrast, the semi-deciduous forest plot shared only 13% of its species with the inventory in the Santa Cruz Botanical Garden, and less than 10% of its species were also present in the other two inventories. The evergreen forest shared even fewer species with any of the other seasonal forest inventories (less than 11%). Low percentages of shared species between seasonal forest inventories, irrespective of geographical distance, have also been reported for the seasonally dry forest in Mexico and attributed to the existence of processes of high local diversification, possibly due to particular historical processes (Trejo & Dirzo 2002). Inde ed, the complex vegetational arrangement

Acknowledgements

We thank the reviewers and the handling editor for suggestions and critique that improved the original manuscript. We thank the owner of Los Volcanes, A. Schwiening, for allowing us to work on his land. SKH and MK are indebted to the Colección Boliviana de Fauna and the Dirección General de Biodiversidad for research permits. We thank the curators of the Herbario Nacional de Bolivia (LPB) and Herbario del Oriente Boliviano (USZ) for providing us with working facilities and allowing access to their collections. We are grateful to C. Hamel and M. Valverde for help in field and to J. Lendzion for providing both unpublished soil and environmental data from the plots. Financial support was provided by the Deutsche Forschungsgemeinschaft (DFG) and the German Academic Exchange Service (DAAD).

References

1.- Balvanera, P., E. Lott, G. Segura, C. Siebe & A. Islas. 2002. Patterns of -diversity in a Mexican tropical dry forest. Journal of Vegetation Science 13: 145-158.         [ Links ]

2.- Cayola, L., A. Fuentes & P. M. Jørgensen. 2005. Estructura y composición forística de un bosque seco subandino yungueño en el valle del Tuichi, Área Natural de Manejo Integrado Madidi, La Paz (Bolivia). Ecología en Bolivia 40: 396-417.         [ Links ]

3.- Chazdon, R. L. 2003. Tropical forest recovery: legacies of human impact and natural disturbances. Perspectives in Plant Ecology, Evolution and Systematics 6: 51-71.         [ Links ]

4.- Condit, R., S. P. Hubbell, J. V. LaFrankie, R. Sukumar, N. Manokaran, R. B. Foster & P. S.  Ashton. 1996. Species-area and species­individual relationships for tropical trees: A comparison of three 50-ha plots. Journal of Ecology 84: 549-562.        [ Links ]

5.- Condit, R., P. S. Ashton, P. Baker, S. Bunyavejchewin, S. Gunatilleke, N. Gunatilleke, S. P. Hubbell, R. B. Foster, A. Itoh, J. V. LaFrankie, H. S. Lee, E. Losos, N. Manokaran, R. Sukumar & T. Yamakura. 2000. Spatial patterns in the distribution of tropical tree species. Science 288: 1414-1418.        [ Links ]

6.- Curtis, J. T. & R. P. Mcintosh. 1951. An upland forest continuum in the prairie-forest border region of Wisconsin. Ecology 32: 476-496.         [ Links ]

7.- Fuentes Claros, A., A. Araujo Murakami, H. Cabrera Condarco, F. Canqui, L. Cayola, C. Maldonado & N. Paniagua. 2004. Estructura composición y variabilidad del bosque subandino xérico en un sector del valle del río Tuichi, ANMI Madidi, La Paz (Bolivia). Revista Boliviana de Ecología y Conservación Ambiental 15: 41-62.         [ Links ]

8.- Gentry, A. H. 1995. Diversity and foristic composition of neotropical dry forests. pp. 146-194. In: S. H. Bullock, H. A. Mooney & E. Medina (eds.). Seasonally Dry Tropical Forests. Cambridge University Press, Cambridge.         [ Links ]

9.- Holmgren, P. K. & N. H. Holmgren. 1998. [continuously updated]. Index herbariorum: a global directory of public herbaria and associated staff. New York Botanical Garden‘s Virtual Herbarium. http://sweetgum.nybg.org/ih/. Cited 5 February 2008.        [ Links ]

10.- Hoffmann, D. 2004. Lianen dreier Waldtypen im Übergang der Tropen zu den Subtropen in Zentralbolivien. Diploma thesis, Halle-Wittenberg University, Halle. 107 p.        [ Links ]

11.- Ibisch, P. L., S. G. Beck, B. Gerkmann & A. Carretero. 2003. Ecoregiones y ecosistemas. pp. 47-88. In: P. L. Ibisch & G. Mérida (eds.). Biodiversidad: La Riqueza de Bolivia. Estado de Conocimiento y Conservación. Ministerio de Desarrollo Sostenible y Planifcación, Editorial FAN, Santa Cruz.        [ Links ]

12.- Jardim, A., T. J. Killeen & A. Fuentes. 2003. Guía de los árboles y arbustos del bosque seco Chiquitano, Bolivia. Editorial FAN, Santa Cruz. 324 p.        [ Links ]

13.- Kessler, M. & N. Helme. 1999. Floristic diversity and phytogeography of the central Tuichi Valley, an isolated dry forest locality in the Bolivian Andes. Candollea 54: 341-366.        [ Links ]

14.- Killeen, T. J., A. Jardim, F. Mamani, N. Rojas, & P. Saravia. 1998. Diversity, composition and structure of a tropical semideciduous forest in the Chiquitanía region of Santa Cruz, Bolivia. Journal of Tropical Ecology 14: 803-827.        [ Links ]

15.- Killeen, T. J., E. García & S. G. Beck. 1993. Guía de árboles de Bolivia. Herbario Nacional de Bolivia/Missouri Botanical Garden, Edit. Quipus srl, La Paz. 958 p.        [ Links ]

16.- Laurance, W. F., L. V. Ferreira, J. M Rankin de Merona & R. W. Hutchings. 1998. Infuence of plot shape on estimates of tree diversity and community composition in Central Amazonia. Biotropica 30: 662-665.        [ Links ]

17.- Lendzion, J. 2003. Krautfora dreier Waldtypen im Übergang der Tropen zu den Subtropen in Zentralbolivien. Diploma thesis, Göttingen University, Göttingen. 101 p.        [ Links ]

18.- Lopes, C. G. R., E. M. N. Ferraz & E. L. Araujo. 2008. Physiognomic-structural characterization of dry- and humid­forest fragments (Atlantic Coastal Forest) in Pernambuco State, NE Brazil. Plant Ecology 198: 1-18.        [ Links ]

19.- Magurran, A. E. 2004. Measuring biological diversity. Blackwell Publishing, Oxford. 256 p.        [ Links ]

20- Medina, E. 1995. Diversity of life forms of higher plants in Neotropical dry forests. pp. 221- 242. In: S. H. Bullock, H. A. Mooney & E. Medina (eds.). Seasonally Dry Tropical Forests. Cambridge University Press, Cambridge.        [ Links ]

21.- Mori, S. A., B. M. Boom, A. M. de Carvalho & T. S. dos Santos. 1983. Southern Bahian moist forests. Botanical Review 49: 155-232.        [ Links ]

22.- Murphy, P. G. & A. E. Lugo. 1986. Ecology of tropical dry forest. Annual Review of Ecology and Systematics 17: 67-68.        [ Links ]

23.- Navarro G., I. Vargas, A. Jardim, M. Toledo & N. de la Barra. 1996. Clasifcación y diagnóstico para la conservación de la vegetación de la región del Parque Nacional Amboró. Santa Cruz, Bolivia. Universidad Complutense, Fundación Amigos de la Naturaleza, Museo de Historia Natural Noel Kempff Mercado, Universidad Mayor de San Simón, New York Botanical Garden. Santa Cruz. 78 p.        [ Links ]

24.- Navarro G. & M. Maldonado. 2002. Geografía ecológica de Bolivia: Vegetación y ambientes acuáticos. Fundación Simón I. Patiño, Cochabamba. 719 p.        [ Links ]

25.- Nee, M. 2004. Flora de la región del Parque Nacional Amboró. Bolivia. Vol. 2: Magnoliidae, Hamamelidae y Caryophyllidae. FAN. Santa Cruz. 208 p.        [ Links ]

26.- Oliveira-Filho, A. T., N. Curi, E. A. Vilela & D. A. Carvalho. 1998. Effects of canopy gaps, topography and soils on the distribution of woody species in a central Brazilian deciduous dry forest. Biotropica 30: 362-375.         [ Links ]

27.- Pennington, R. T., G. P. Lewis & J. A. Ratter. 2006. An overview of the plant diversity, biogeography and conservation of neotropical savannas and seasonally dry forests. In: R. T. Pennington, G. P. Lewis & J. A. Ratter (eds.). pp. 1-29. Neotropical Savannas and Seasonally Dry Forests: Plant Diversity, Biogeography and Conservation. CRC Press, Miami.        [ Links ]

28.- Prado, D. E. 2000. Seasonally dry forests of tropical South America: from forgotten ecosystems to a new phytogeographic unit. Edinburgh Journal of Botany 57: 437-461.        [ Links ]

29.- Prado, D. E. & P. E. Gibbs. 1993. Patterns of species distributions in the dry seasonal forests of South America. Annals of the Missouri Botanical Garden 80: 902-927.
        [ Links ]

30.- Schnitzer, S. A. & F. Bongers. 2002. The ecology of lianas and their role in forests. Trends in Ecology and Evolution 17: 223-230.         [ Links ]

31.- Smith, D. N. & T. J. Killeen. 1998. A comparison of the structure and composition of montane and lowland tropical forest in the Serranía Pilón Lajas, Beni, Bolivia. pp. 681-700. In F. Dallmeier & J. A. Comiskey (eds.). Forest Biodiversity Research, Monitoring and Modeling, Vol. 21. The Parthenon Publishing Group, New York.        [ Links ]

32.- Trejo, I. & R. Dirzo. 2002. Floristic diversity of Mexican seasonally dry tropical forests. Biodiversity and Conservation 11: 2063- 2048.        [ Links ]

33.- Uslar, Y. V., B. Mostacedo & M. Saldias. 2004. Composición, estructura y dinámica de un bosque seco semideciduo en Santa Cruz, Bolivia. Ecología en Bolivia 39: 25-43.        [ Links ]

34.- Vargas, I. 1995. Estructura y composición de cuatro sitios boscosos en el Parque Nacional Amboró. Tesis de grado, Carrera de Agronomía, Universidad Autónoma Gabriel René Moreno, Santa Cruz. 78 p.        [ Links ]

Artículo recibido en: Junio de 2008.

Aceptado en: Septiembre de 2008.