Biodiversity and Conservation 7, 1563±1575 (1998)

Biodiversity and Conservation 7, 1563±1575 (1998)

Vascular plant species count in a wet forest in the Choco area on the Paci®c coast of Colombia

GLORIA GALEANO*

Instituto de Ciencias Naturales, Universidad Nacional de Colombia, Apartado 7495, BogotaÂ, Colombia

STELLA SUAÂ REZ

Apartado 265216, BogotaÂ, Colombia

HENRIK BALSLEV

Department of Systematic Botany, Institute of Biological Sciences, Building 137, University of Aarhus, DK-8000 Aarhus C, Denmark

Received 22 February 1998; revised and accepted 13 May 1998

The total number of vascular plant species was counted and growth form distribution was studied in the Choco area on the Paci®c coast of Colombia, in two transects 400 ́ 10 m and ten transects 2 ́ 50 m, for a total sampled area of 0.9 ha. The species count of the ten transects (442 species in 0.1 ha) appears to be the highest number of species recorded with this methodology. There were 970 species for the total area (0.9 ha). Ninety to ninety-®ve percent of the species were under 10 cm dbh and 70±86% under 2.5 cm dbh, epiphytes and small trees and treelets £10 cm dbh being the most diverse growth forms. The most species-rich families and genera were those represented by herba- ceous plants and treelets. Individuals were counted only in the ten 2 ́ 50 m transects (0.1 ha), where 4459 individuals were found. Palms and ferns were the most abundant growth forms. Arguments are presented against the way diversity is usually measured. Recommendations are made to include other growth forms besides trees when assessing alpha diversity for conservation purposes.

Keywords: species richness; growth forms; wet forest; Choco ; Colombia.Introduction

The bulk of information on structure, composition and diversity of neotropical forests is based upon the study of the woody component of the forest. Customarily, quantitative inventories include only trees 310cm dbh and less often they include all woody plants32.5 cm dbh (Gentry, 1982a, 1986, 1988, and see review in Gentry, 1995), and very few studies include plants 31 cm dbh (e.g. Franco-Rosselli et al., 1997). The few quantitative studies of neotropical forests that have considered all the species or have included other non-woody life-forms, have shown the importance of non-tree species for the structure, composition and species richness of tropical forests (Smith, 1970; Whitmore et al., 1985; Gentry and Dodson, 1987; Poulsen and Balslev, 1991; Duivenvoorden, 1994; Balsev et al., in press). The available information, however, is still very scarce. The total sampling of vascular plant alpha diversity faces the di􏰁culty of collecting in the forest canopy and of identifying sterile material (Whitmore et al., 1985). Thus, most inventories in tropical forests continue to include only individuals 310 cm dbh.

*To whom correspondence should be addressed. 0960-3115 Ó 1998 Chapman & Hall

1564 Galeano et al.

The Choco biogeographic region (Gentry, 1982b) is believed to hold about 8000 vas- cular plant species, including a high number of endemic species, and it is considered to include some of the most diverse forests on earth (Lellinger, 1975; Gentry, 1986, 1993; Forero and Gentry, 1989; McNeeley et al., 1990). Quantitative studies on the region have dealt mostly with trees 310 cm dbh or with individuals 32.5 cm dbh (Gentry, 1986; Faber- Langendoen and Gentry, 1991). The only published study dealing with a species count and including all growth forms and size classes in the Choco region is that of Gentry and Dodson (1987) at RõÂo Palenque, Ecuador, where they recorded a total of 365 vascular plant species in ten 0.01 transects. That locality, however, has a considerably drier climate (2980 mm of annual precipitation) than the site of the present study.

The structure, composition and richness of the forests at the Gulf of Tribuga , on the Paci®c coast of Choco , Colombia, were previously studied based on the individuals 35 cm and 310 cm dbh (Galeano et al., in press; Galeano, unpubl. data). In this paper we report the total number of vascular plant species counted in two 0.4 ha and ten 0.01 ha transects. This is the ®rst total species count ever recorded for a site with more than 5000 mm of rainfall, and it is the second one recorded for the Choco region. On the basis of these counts, we analyse the importance of di􏰀erent growth forms and taxonomic groups to forest structure, composition, and richness, and we make comparisons with similar studies in the Choco region and elsewhere in the Neotropics.

Methods

Study area

The study was carried out at the Gulf of Tribuga , on the Paci®c Coast of the Choco Department, Colombia (5° 29¢±6° 08¢N, 77° 10¢±77° 32¢W). The land around the gulf consists of hills lower than 300 m, with steep slopes of 10±50°, mostly covered by undis- turbed tropical wet forest according to Holdridge’s life-zone system (IGAC, 1977). Soil is derived from basaltic rocks, with a moderately developed structure, nutrient poor and acidic (Corte s, 1981). Annual precipitation ranges from 5100 mm (average for 20 years) as recorded in a weather station about 50 km to the north, and 7150 mm (average for 5 years) as recorded in E1 Amargal Biological Station, near the southmost point of the gulf. More information about the study area, the sites, and about the structure, composition and richness of the woody species can be found in Galeano et al. (in press).

Data

Two 0.4 ha transects (400 ́ 10 m) about 10 km from each other, were made at NuquõÂ and CoquõÂ. The third sampling unit was established at E1 Amargal Biological Station, ca 27 and 30km west of the NuquõÂ and CoquõÂ transects, respectively. Here, 10 transects of 2 ́ 50 m, mostly parallel among them and separated by 10±15 m were established. All transects were subjectively placed to include the range of topography common in the area (hillsides, ridges and some ̄at areas). In each transect, all vascular plant species were recorded and collected. Dbh and height were measured on all plant individuals. Trees whose stems straddled the transect border were included in the sample if the midpoint of their trunks fell within the range. Epiphytes and lianas were located with binoculars and then sampled. The sample was enhanced by bringing down one to two large tree branches in each of 20 subtransects of 10 ́ 20 m at NuquõÂ and CoquõÂ and in each of the ten 2 ́ 50 m transects at E1 Amargal, and searching them for epiphytes. At E1 Amargal the

Plant species count in ChocoÂ, Colombia 1565

number of individuals of each species was counted; epiphytes were counted with binoc- ulars; clones of rhizomatous plants were counted as single individuals only when we were certain that all ramets belonged to the same plant. Woody plants shorter than 1.3 m were not sampled at NuquõÂ and CoquõÂ , and saplings less than 50 cm tall were not sampled at E1 Amargal. Given the constrains of the last points mentioned, the samplings of both species and individuals were neither exact nor total. Vouchers were identi®ed at COL, or sent to specialists and duplicates were deposited at INDERENA, CHOCO, and AAU.

For analysis, each species (or morphospecies when identi®cation was not possible) was assigned to one of the following growth forms (some of them modi®ed from Gentry, 1985):

1. Trees (woody plants 35 cm dbh, excluding palms, lianas, and hemiepiphytes).
2. Treelets (woody plants <5 cm dbh, excluding palms, lianas and hemiepiphytes).
3. Palms (members of the family Palmae and Zamia obliqua, Zamiaceae).
4. Lianas (woody climbers; begining life as terrestrial seedlings).
5. Vines (herbaceous or subwoody climbers; begining life as terrestrial seedlings).
6. Hemiepiphytes (climbers; begining life as epiphytes; sometimes with woody stemsappressed to supporting trees).
7. Terrestrial herbs (including saprophytes).
8. Epiphytes (both erect and climbing epiphytes).

Results

Species richness and composition

The total number of vascular plant species found at the three sites is shown in Table 1. The species±area curve for the cumulative species in all sampling units (Fig. 1) showed no indication of levelling o􏰀, indicating that more species would have appeared if a larger area had been sampled. For the three sites pooled (an area of 0.9 ha) there were 970 species of vascular plants, distributed as follows: 890 species of ̄owering plants (dicots ac- counting for 70% of the species, and 80% of the families), 78 species of Pteridophyta, and two species of gymnosperms (Zamia obliqua A. Braun and Gnetum leyboldii Tul.). Family concept for the angiosperms follows Cronquist (1981), but the Fabaceae are treated as one large family including Caesalpiniaceae and Mimosaceae; the Pteridophyta are treated as a large group following Mabberley (1989). Only 38% of the species have been satisfactorily identi®ed so far. The ̄owering plants were distributed among 90 families and, at least, 295 genera. The most species-rich families are shown in Fig. 2. These 15 families (17% of all families found) included 63% of the total number of species. The species-rich families are roughly the same for the three transects, although ranked di􏰀erently. Figure 3 shows the species-richest genera. These 12 genera accounted for 23.4% of all the species; most of them are herbaceous*.

Density of individuals

The total number of individuals in ten 0.01 transects at E1 Amargal was 4459. Table 2 shows the number of individuals and their relative proportions in each growth form; Table 3 shows the most abundant species. Terrestrial herbs (most of them represented by ferns) and palms were the most dense groups.

*A complete list of species found can be obtained from the ®rst author.

Plant species count in ChocoÂ, Colombia 1567

Figure 1. Species±area curve for all vascular plant species in the cumulative area sampled at the Paci®c coast of Choco , Colombia.

Figure 2. Largest families of vascular plants in three plots totalling 0.9 ha, at the Paci®c coast of Choco , Colombia. Ferns are treated as one family. Figures at the end of the bars represent per- centages of the total.

Size and growth form distribution

Table 1 shows the growth form distribution at the three sites (including the ®rst 0.1 ha at NuquõÂ and CoquõÂ ). Epiphytes were the most diverse growth form in NuquõÂ and CoquõÂ and the second most diverse at E1 Amargal (Fig. 4). Epiphytes (including erect epiphytes, climbing epiphytes, and hemiepiphytes) accounted for 25±46% of the total species number, being more abundant at CoquõÂ . Small trees and treelets (<10 cm dbh, including their saplings), were also very species-rich, accounting for 26±46% of all species in all

1568 Galeano et al.

Figure 3. Largest genera (with 10 or more species) of vascular plants in three plots totalling 0.9 ha, at the Paci®c coast of Choco , Colombia. In parentheses growth form represented in order of occurrence (t, trees; tr, treelets; h, terrestrial herbs; e, epiphytes + hemiepiphytes; l, lianas).

counts (Table 1). At E1 Amargal, treelets <2.5 cm dbh were remarkably species-rich (Fig. 4). Ninety to ninety-®ve percent of the species in the transects were under 10 cm dbh and 70±86% were under 2.5 cm dbh (Fig. 5). The non-woody components (terrestrial herbs, epiphytes, vines, and herbaceous hemiepiphytes) accounted for about 50% of the total species number (Fig. 5).

Discussion

The total of 442 species of vascular plants found in the ten 2 ́ 50 m transects at E1 Amargal is, to our knowledge, the highest ®gure ever recorded for a total species count in 0.1 ha. The second highest number of species has been recorded for the RõÂ o Palenque area, Ecuador (365 species), employing the same methodology used at E1 Amargal but in- cluding saplings less than 50 cm high (Gentry and Dodson, 1987). If those size classes had been included at E1 Amargal, the number of the species would have been probably higher. Also, because we did not employ destructive sampling, it is possible that some species were missed, especially in the high canopy.

The species counts for NuquõÂ and CoquõÂ and for their respective 0.1 ha subtransects are similar to each other. However, the same area (0.1 ha) sampled at E1 Amargal had 1.6±2.0 times as many species as the 0.1 ha transects at NuquõÂ or CoquõÂ , and about as many species as either of the two 0.4 ha transects (Table 1). This remarkable di􏰀erence seems to be related to the di􏰀erent sampling methodology. First, the di􏰀erent size limit for sampling in NuquõÂ and CoquõÂ (1.3 m high) compared to that at E1 Amargal (50 cm high) accounts for a di􏰀erence of 69 species. Second, the exploded quadrat method (Gentry, 1982a) used at E1 Amargal seems to yield many more species than the contiguous areas used in NuquõÂ and CoquõÂ. Faber-Langendoen and Gentry (1991) using contiguous plots, recorded

Plant species count in ChocoÂ, Colombia 1569Table 2. Distribution of the total number of individuals among growth forms in 0.1 ha at E1

Amargal, Paci®c coast of Choco , Colombia

Growth forms No. individuals Percentage

Terrestrial herbs 1704 38.2 Epiphytes 492 11.0 Hemiepiphytes 144 3.2 Lianas 75 1.7 Vines 21 0.5 Palms 1177 26.4

Treelets <2.5 cm dbh (incl. Saplings)aSmall trees and treelets 2.5 £ dbh <10 cmaTrees 310 cm dbh

690 15.5 106 2.4 50 1.1

Total 4459 100a Treelets were divided into two size classes for easier comparison.

1.6±2.4 times fewer species as Gentry (1986) found with the exploded quadrat method; although both sites were extremely close to each other (Duivenvoorden, 1994). Finally, the exploded quadrat method involves a higher possibility of sampling error (Peters, 1996).

Compared to other forests from the Choco biogeographic region (Table 4), particularly in terms of species 310 cm and 32.5 cm dbh, the forest at the Gulf of Tribuga is less rich, and seems to be structurally more similar to the drier site at Rõ o Palenque, Ecuador than to other Colombian sites with similar or higher rainfall.

Most of the diversity at the study transects is due to the herbaceous component, es- pecially to epiphytes. This overall high importance of epiphytes is similar for some Andean forests as well as for some Central American forests, particularly La Selva (Hammel, 1990), and is one of the most notable di􏰀erences between these regions and the Amazon region (see Gentry, 1990; Duivenvoorden, 1994). The Choco sites studied here, compar- atively poor in tree species 310 cm dbh but with a high richness in vascular plant species, represent an extreme in the dichotomy of growth form distribution between Amazonian and Central American (and now Chocoan) forests (Gentry, 1990). Whereas Amazonia,

Table 3. The most abundant Choco , Colombia

Species
Wel®a regia H. Wendl.

Trichomanes diversifrons

(Bory) Mett.

Wettinia quinaria

(O.F. Cook and Doyle) Burret
Calathea elegans KennedyTrichomanes elegans Rich.Philodendron inaequilaterum Liebm.

species in ten 0.01 ha transects at

E1 Amargal Biological Station,

Family

Palmae Fern

Palmae

Marantaceae Fern Araceae

Growth form

palm terrestrial herb

palm

terrestrial herb terrestrial herb hemiepiphyte 93

No.

746 728

260

individuals Percentage

16.7 16.3

5.8

2.8 2.2 2.1

126 99

1570 Galeano et al.

Figure 4. Percentage of species per growth form in three 0.1 ha plots at the Paci®c coast of Choco , Colombia. At E1 Amargal, treelets £2.5 cm dbh include saplings higher than 50 cm. Some species may overlap between trees and treelets.

Figure 5. Contribution of size classes and growth forms groups to the total number of species in three 0.1 ha plots at the Paci®c coast of Choco , Colombia. Nonwoody species include terrestrial herbs, epiphytes, hemiepiphytes and vines.

Plant species count in ChocoÂ, Colombia 1571

Table 4. Number of species 310 cm and 32.5 cm recorded for 0.1 ha plots in the Choco biogeographic region, and at La Selva

Site

Choco sites
Bajo Calima 1 Bajo Calima 2 Tutunendo RõÂo Palenque NuquõÂ
CoquõÂ
E1 Amargal 1 E1 Amargal 2

La Selva, Costa Rica

310 cm dbh

77
42±48
55
32
24
31
37
21±40
43

32.5 cm dbh

265
108±166
258
119

65

77 102 90 130

References

Gentry (1986)
Faber-Langendoen and Gentry (1991) Gentry (1986)
Gentry (1986)
This study
This study
This study
Galeano et al. (in press)
Gentry (1995)

particularly the western part, seems to be the region of highest tree species richness (Gentry, 1990; Duivenvoorden, 1994; Valencia et al., 1994), the Choco region appears to be the one with the highest species richness of non-woody plants. If this generality proves true, one would expect that a total species count at any Choco site as rich in woody species as Bajo Calima or Tutunendo (Table 4) would result in a number of species higher than the one recorded here.

The low representation of lianas is noteworthy. The number of liana species 32.5 cm dbh per 0.1 ha at the three 0.1 ha transects in the Gulf of Tribuga (2±6) is more similar to that recorded in the temperate zone or in dry tropical forest (x 􏰇 12 species) than to the number recorded at moist (x 􏰇 25 species), or wet tropical forests (x 􏰇 32 species) (Gentry, 1982a). A low density of lianas has also been recorded for other forests in the Choco region (Gentry, 1986; Faber-Lagendoen and Gentry, 1991). To explain this unusual feature for a neotropical forest, Gentry (1986) suggested that in the Choco forest the lianas are, in an ecological sense, replaced by hemiepiphytes. Nevertheless, our data do not support this view because hemiepiphytes are also scarce and even less important than lianas at any scale. Gentry (1982a) discussed the existence of an apparent positive cor- relation between liana density and stand maturity. Nevertheless, the structure of the transects in the present study showed that the forest is a mature one (Galeano et al., in press). It may be possible, then, that epiphytes (instead of hemiepiphytes) replace (in an ecological sense) lianas at the study site, as was suggested by Hartshorn and Hammel (1994) for the forest at La Selva, which has also few lianas and abundant epiphytes. The site with fewest liana species, the 0.1 ha at Coquõ (2.6% of the species) was the same with the largest number of epiphytes (46% of the species).

Floristic composition, in terms of families and genera, re ̄ects growth form represen- tation. The species-richest families and genera at the Gulf of Tribuga were those of her- baceous plants and treelets (Fig. 2). A similar situation has been recorded at La Selva and, to a lesser extent, at Barro Colorado Island (Gentry, 1990). This is in strong contrast to the situation recorded in the Amazon, where families and genera of trees and treelets prevail (Foster, 1990; Gentry, 1990; Prance, 1990; Duivenvoorden, 1994; but see Balslev et al., in press). The Araceae, Piperaceae, ferns, and Cyclanthaceae are particularly important at our study sites. The substantial local diversity of these taxa is due to the coexistence of many sympatric species (Fig. 3); 74% of the species of Anthurium and 72% of the species

1572 Galeano et al.

of Philodendron are epiphytes; this supports Gentry’s (1982a) observation that infrageneric diversi®cation is most common among epiphytes. The diversity of Araceae recorded here (100 species in 0.9 ha) seems to be the highest ever recorded for such a small area. Croat (1992) stated that the aroid ̄ora of Colombia is the highest of any region in the world and that most of the species must be concentrated in western Colombia. Even La Selva, in Costa Rica, well known as rich in Araceae, has an aroid ̄ora as rich as the one recorded here (99 species), although the area of natural forest at La Selva is about 1000 times as large as our sample (Hartshorn and Hammel, 1994). The number of species of Cyclanthaceae found in our study (30 species in 0.9 ha) is probably also a record for this family; this supports the view that this family has its main center of diversi®cation in the Choco (Harling, 1958).

The number of individuals in 0.1 ha at E1 Amargal was lower than the existing records for wet and dry forests in western Ecuador, 7210 and 5428 individuals per 0.1 ha, respec- tively (Gentry and Dodson, 1987). The main di􏰀erences between these sites and E1 Amargal were the strikingly higher density of epiphytes in the wet forest at RõÂ o Palenque (although the species richness of this growth form was similar), and the higher density of herbs in the dry forest. Interestingly, the number of trees 310cm dbh and its relative density at E1 Amargal were almost the same as at RõÂ o Palenque (Gentry and Dodson, 1987). The density of terrestrial herbs at E1 Amargal (1704 individuals per 0.1 ha) was, however, higher than that recorded by Gentry and Dodson (1987) at RõÂ o Palenque (1220 individuals in 0.1 ha) and by Poulsen and Balslev (1991) in Amazonian Ecuador (813±1322 individuals per 0.1 ha). The contribution of palms to forest structure at the study area is apparent from their relative density (26.4% of all individuals). Palms not only characterize the woody com- ponent (Galeano et al., in press; Galeano, unpubl. data) but they are also dominant across the vertical gradient, especially in the understory.

The total number of species in the study area is higher than recorded in the small transects as shown by the species±area curve (Fig. 1). Considering the small area sampled (0.9 ha), the total number of species (970 species) is high compared with the number of vascular species recorded for larger areas, e.g. Gorgona island (in the same Choco region) with an area of 15.7 km2, has a known ̄ora of 512 species (Barbosa, 1986; Murillo and Lozano, 1989; Rangel and Aguirre, 1990); Barro Colorado Island, 15.6 km2, has 966 forest species (Croat, 1978; Foster and Hubbell, 1990); La Selva, Costa Rica, in about 14 km2has 1458 native species (Hartshorn and Hammel, 1994; McDade and Hartshorn, 1994). Most strikingly, the 970 species found in our sample represent about 25% of all species recorded in the checklist of the Choco Department (Forero and Gentry, 1989) which encompasses an area of more than 5 million times as large as our sample. This stresses the need of many more collections and ̄oras from the Choco region. Another count, in the Amazon, recorded a similar richness (1223 species) in ten 0.1 ha plots (Duivenvoorden, 1994). Those plots, however, were located farther from each other and on very di􏰀erent soil types, which adds to the possibility of ®nding di􏰀erent species.

Comparing the few data available, the forests at the Gulf of Tribuga seem to be more similar in terms of growth form distribution, tree species richness and density, and ̄oristic composition at the family level, to the forest at La Selva, Costa Rica (Hartshorn and Hammel, 1994), and at Rõ o Palenque, Ecuador (Gentry 1986; Gentry and Dodson, 1987), than to other sites so far studied in the Colombian Choco (Tutunendo and Bajo Calima: Gentry, 1986; Faber-Langendoen and Gentry, 1991). A similarity in ̄oristic composition between La Selva and RõÂo Palenque was pointed out by Gentry (1993).

Plant species count in ChocoÂ, Colombia 1573

The fact that most of the species were under 10 cm dbh, and even under 2.5 cm dbh (Fig. 5), supports the view that tree diversity and treelet or herb diversity are not always necessarily correlated (Gentry, 1982a), and that the earth’s richest plant community is the understory of the tropical rain forest (Gentry and Dodson, 1987; Hartshorn and Hammel, 1994). It is, then, clear that a classical forest inventory, including only individuals 310 cm dbh (and less often individuals 32.5 cm dbh) is not at all the appropriate means to study total composition and species richness. Gentry and Dodson (1987) stated that tree species310 cm dbh constitute only 15±20% of the complete ̄oras of many neotropical sites. Even for many Amazonian forests where tree species prevail, trees 310 cm dbh scarcely account for 10±30% of the total number of species (Duivenvoorden, 1994). Inventories attempting to assess alpha diversity for conservation goals should, of necessity, also consider growth forms other than large trees. If this is not done, by far the greatest part of the species richness will not be taken into consideration.

Acknowledgements

This paper is part of G. Galeano’s Ph.D. dissertation at the Department of Systematic Botany, Biological Institute, Aarhus University, directed by Henrik Balslev. Financial support was given by DANIDA (Grant no. 104.DAN.8.4/201), Fondo FEN-Colombia, The MacArthur Foundation, INDERENA, Universidad Nacional de Colombia, and FundacioÂn InguedeÂ. We thank L. GarcõÂa, J. Grueso, C. HernaÂndez, O. Hurtado, P. Hurtado, A. Moreno, G. Moreno, J. Moreno, N. Moreno, I. Perea, L. Perea, D. Tuberquia, F. Valencia and E. Ayazo for ®eld assistance. Logistic support was given by Instituto de Ciencias Naturales, Universidad Nacional de Colombia, the Department of Systematic Botany and the herbarium at Aarhus University, Denmark. L. Allorge, M. Amaya, L. Andersson, C.C. Berg, J. Betancur, R. Callejas, L. Clark, J.L. Ferna ndez, A. GoÂmez, P. Maas, R.T. Pennington, J. Sarmiento, and C. Taylor helped with identi®cations. We thank A. Barfod, R. Bernal, A. Cleef, J. Duivenvoorden, R. Foster, and P. Hall for reviewing the manuscript.

References

Barbosa, C. (1986) Contribucio n al conocimiento de la ̄ora y vegetacio n del Parque Nacional Natural Isla Gorgona y Gorgonilla. Perez-Arbelaezia 1, 311±35.

Balslev, H., Valencia, R., Paz y MinÄ o, G., Christensen, H. and Nielsen, I. (in press) Species count of vascular plants in 1-hectare of humid lowland forest in Amazonian Ecuador. In Forest Bio- diversity in North, Central and South America and the Caribbean: Research and Monitoring(F. Dallmeier and J. Comiskey, eds.) Man and the Biosphere Series, Vol 22, Unesco and the Parthenon Publishing Group, Carnforth, Lancashire, UK.

Corte s, A. (1981) Los suelos del AndeÂn Pacõ®co y su Aptitud de Uso. Bogota : Instituto Geogra ®co “AgustõÂn Codazzi”.

Croat, T. (1978) Flora of Barro Colorado Island. Stanford: Stanford University Press.
Croat, T.B. (1992) Species diversity of Araceae in Colombia: a preliminary survey. Ann. Missouri

Bot. Gard. 79, 17±28.
Cronquist, A. (1981) An Integrated System of Classi®cation of Flowering Plants. New York: Co-

lumbia University Press.
Duivenvoorden, J.F. (1994) Vascular plant species counts in the rain forests of the middle CaquetaÂ

area, Colombian Amazonia. Biodivers. Conserv. 3, 685±715.

1574 Galeano et al.

Faber-Langendoen, D. and Gentry, A.H. (1991) The structure and diversity of rain forests at Bajo Calima, Choco region, western Colombia. Biotropica 23, 2±11.

Forero, E. and Gentry, A.H. (1989) Lista anotada de las plantas del Departamento del ChocoÂ, Colombia. Biblioteca Jose Jero nimo Triana No. 10. Bogota : Instituto de Ciencias Naturales- Universidad Nacional de Colombia.

Foster, R.B. (1990) The ̄oristic composition of the Rõ o Manu ̄oodplain forest. In Four Neotropical Rainforest (A. Gentry, ed.) pp. 99±111. New Haven and London: Yale University Press. Foster, R.B. and Hubbell, S.P. (1990) The ̄oristic composition of the Barro Colorado island forest.

In Four Neotropical Rainforests (A. Gentry, ed.) pp. 85±98. New Haven and London: Yale

University Press.
Franco-Rosselli, P., Betancur, J. and Ferna ndez-Alonzo, J.L. (1997) Estudio comparativo de la

diversidad ̄orõÂstica en dos bosques subandinos del sur de Colombia. Caldasia 19, 205±34. Galeano, G., Cediel, J. and Pardo, M. (in press) Structure and ̄oristic composition of a one hectare plot of wet forest at the Paci®c Coast of Choco , Colombia. In Forest Biodiversity in North, Central and South America and the Caribbean: Research and Monitoring (F. Dallmeier and J. Comiskey, eds.) Man and the Biosphere Series Vol 22. Unesco and the Parthenon Publishing

Group. Carnforth, Lancashire, UK.
Gentry, A.H. (1982a) Patterns of neotropical plant species diversity. Evol. Biol. 15, 1±84.
Gentry, A.H. (1982b) Phytogeographic patterns as evidence for a Choco refuge. In Biological Di-

versi®cation in the Tropics (G.T. Prance, ed.) pp. 112±36. New York: Columbia University

Press.
Gentry, A.H. (1985) An ecotaxonomic survey of Panamanian lianas. In The Natural History of

Panama: La BotaÂnica e Historia Natural de Panama (W.D’Arcy and M. Correa, eds) pp. 29±42. Gentry, A.H. (1986) Species richness and ̄oristic composition of Choco region plant communities.

Caldasia 15, 71±9.
Gentry, A.H. (1988) Changes in plant community diversity and ̄oristic composition on environ-

mental and geographical gradients. Ann. Missouri Bot. Gard. 75, 1±34.
Gentry, A.H. (1990) Floristic similarities and di􏰀erences between Southern Central America and Upper and Central Amazonia. In Four Neotropical Rainforest (A.H. Gentry, ed.) pp. 141±57.

New Haven and London: Yale University.
Gentry, A.H. (1993) Riqueza de especies y composicio n ̄orõ stica de las comunidades de plantas de la

regio n del Choco : una actualizacio n. In Colombia Pacõ®co. Tomo I (P. Leyva, ed.) pp. 201±19.

Bogota : Fondo FEN-COLOMBIA.
Gentry, A.H. (1995) Patterns of diversity and ̄oristic composition in Neotropical Montane forests.

In Biodiversity and Conservation of Neotropical Montane forests (S.P. Churchill, H. Balslev, E.

Forero and J.L. Luteyn, eds) pp. 103±26. New York: The New York Botanical Garden. Gentry, A.H. and Dodson, C. (1987) Contribution of nontrees to species richness of a tropical rain

forest. Biotropica 19, 149±56.
Hammel, B. (1990) The distribution of diversity among families, genera, and habit types in the La

Selva ̄ora. In Four Neotropical Rainforests (A.H. Gentry, ed.) pp. 75±84. New Haven and

London: Yale University Press.
Harling, G. (1958) Monograph of the Cyclanthaceae. Acta Hort. Berg. 18, 1±428.
Hartshorn, G. and Hammel, B. (1994) Vegetation types and ̄oristic patterns. In La Selva. Ecology

and Natural History of a Neotropical Rain forest (L.K. McDade, K.S. Bawa, H. A, Hespenheide

and G.S. Hartshorn, eds) pp. 73±89. Chicago: The University of Chicago Press.
IGAC (Instituto Geogra ®co `Agustõ n Codazzi’) (1977) Zonas de Vida o Formaciones Vegetales de

Colombia. Memoria Explicativa sobre el Mapa EcoloÂgico. Bogota .
Lellinger, D.B. (1975) A phytogeographic analysis of Choco pteridophytes. Fern. Gaz. 11, 105±14. Mabberley, D.J. (1989) The Plant Book; A Portable Dictionary of the Higher Plants. Cambridge:

Cambridge University Press.

Plant species count in ChocoÂ, Colombia 1575

McDade, L.A. and Hartshorn, G.S. (1994) La Selva Biological Station. In La Selva. Ecology and Natural History of a Neotropical Rain Forest (L.K. McDade, K.S. Bawa, H.A, Hespenheide and G.S. Hartshorn, eds) pp. 6±14. Chicago: The University of Chicago Press.

McNeely, J.A., Miller, K.R., Reid, W.V., Mittermeier, R.A. and Werner, T.B. (1990) Conserving the World’s Biological Diversity. Gland: IUCN, WRI, CI, WWF-US, The World Bank.

Murillo, M.T. and Lozano, G. (1989) Hacia la realizacio n de una ̄o rula del Parque Nacional Natural Islas de Gorgona y Gorgonilla, Cauca-Colombia. Rev. Acad. Colomb. Cienc. Exactas, Fis. Nat 12, 277±304.

Peters, C.M. (1996) Beyond nomenclature and use: A review of ecological methods for ethnobot- anists. In Selected Guidelines for Ethnobotanical Research: A Field Manual (M.N. Alexiades, ed.) pp. 241±76. New York: The New York Botanical Garden.

Poulsen, A.D. and Balslev, H. (1991) Abundance and cover of ground herbs in an Amazonian rain forest, J. Veg. Sci. 2, 315±22.

Prance, G.T. (1990) The ̄oristic composition of the forests of Central Amazonian Brazil. In Four Neotropical Rainforests (A.H. Gentry, ed.) pp. 112±40. New York: Yale University Press. Rangel, O. and Aguirre, J. (1990) Caracterizacio n ecolo gica. Sõ ntesis ®nal. In Biota y Ecosistemas de Gorgona (J. Aguirre and O. Rangel, eds) pp. 277±93. Bogota : Fondo FEN-COLOMBIA. Smith, R.F. (1970) The vegetation structure of a Puerto Rican rain forest before and after short-term

radiation. In A Tropical Rain Forest (H. Odum, ed.) pp. D103±40. Oak Ridge: Atomic Energy

Commision.
Valencia, R., Balslev, H. and Paz y MinÄ o, G. (1994) High tree alpha-diversity in Amazonian Ec-

uador. Biodivers. Conserv. 3, 21±8.
Whitmore, T.C., Peralta, R. and Brown, K. (1985) Total species count in a Costa Rican tropical rain

forest. J. Trop. Ecol. 1, 375±8.