Professional Documents
Culture Documents
Abstract
Aim We aim to assess regional patterns in the distribution and species richness of
vascular epiphytes with an emphasis on forests that differ in altitude and the amount of
rainfall.
Location Tropical America, in particularly the 75,000 km2 large state of Chiapas in
southern Mexico at 14.5–18.0N. Chiapas is diverse in habitats with forests from sea-
level to the tree-line at c. 3800 m altitude and with annual amounts of rainfall ranging
from 800 to over 5000 mm. It is also one of the botanical best-explored regions in the
tropics.
Methods First we give an overview of epiphyte inventories to date. Such epiphyte
surveys were mostly carried out on the basis of surface area or individual trees and we
discuss their problematic comparison. Applying a different methodological approach, we
then used 12,276 unique vascular epiphyte plant collections from Chiapas that are
deposited in various botanical collections. The locality data were georeferenced and
compiled in a relational data base that was analysed using a geographical information
system. To compare the number of species between inventories that differed in the
numbers of records, we estimated the total richness, SChao, at each.
Results We recorded 1173 vascular epiphyte species in thirty-nine families (twenty-
three angiosperms), comprising c. 14% of all confirmed plant species in the state. About
half of all species were orchids (568). Ferns and bromeliads were the next species-rich
groups with 244 and 101 species, respectively. Most species were found in the Montane
Rain Forest and in the Central Plateau. Trees of different forest formations, rainfall
regimes, altitudes and physiographical regions supported a characteristic epiphyte flora.
Main conclusions We were able to confirm the presumed presence of a belt of high
diversity at mid-elevations (500–2000 m) in neotropical mountains. In contrast to pre-
dictions, however, we observed a decrease in diversity when the annual amount of
rainfall exceeded 2500 mm. The decrease is attributed to wind-dispersed orchids, bro-
meliads and Pteridophyta that may find establishment problematical under frequent
downpours. In the wet but seasonal forests in Chiapas, this decrease is not compensated
by plants in the animal-dispersed Araceae that are abundant elsewhere. We presume that
in addition to the annual amount of rainfall, its distribution in time determines the
composition of the epiphyte community.
Keywords
Botanical collections, canopy biology, elevation gradient, epiphyte quotient, geograph-
ical information system, rainfall gradient, SChao estimate of diversity, tropical forests.
*Correspondence and present address: Jan H. D. Wolf, Universiteit van Amsterdam, Institute for Biodiversity and Ecosystem Dynamics (IBED), P.O. Box 94062,
1090 GB Amsterdam, The Netherlands. E-mail: wolf@science.uva.nl
Table 1 The plant collections that were examined for vascular epiphytes from Chiapas
the richest in species of Mexico, despite its small territory of barium labels and entered in a relational data base (Micro-
slightly <4%. For example, the number of vascular plant soft Access). Specific information may be provided on
species recorded for Chiapas is 8248 species of an estimated request from the first author. Most labels contained no
total of 22,800 for the whole country (Breedlove, 1981; information about the latitude and longitude of the collec-
Breedlove, 1986; Rzedowski, 1992). tion site. With the help of topographical maps (Instituto
Nacional de Estadı́stica, Geografı́a e Informática, INEGI,
1 : 50.000) such data was estimated to a precision of
Methods
seconds. In case the identification of duplicate collections
We compiled label data of epiphytes in several herbaria that differed, the name given by the experienced taxonomist or
are known to have relatively large collections from the state group specialist was adopted. When no clear differentiation
of Chiapas (Table 1). Hemi-epiphytes were included, but between taxonomists could be made, both names were
facultative epiphytes were not if their presence on trees was maintained (197 cases).
regarded as highly unusual (e.g. Agave sp.). The ecologically The spatial distribution of species was analysed in a GIS
different heterotrophic Loranthaceae were also excluded. (ArcInfo, Redlands, California, USA) where the position of
We considered only specimens identified to the level of species was superimposed on digitized topographical,
species or below. Varieties and subspecies were treated as physiographical, rainfall and vegetation maps that have been
individual species in the diversity estimates. All information prepared at Ecosur. The topographical overlay was derived
concerning the collector, collector’s number, collection date, from maps (1 : 250.000) published in print by INEGI
taxonomy, locality and habitat were copied from the her- between 1985 and 1989 (locality references E15-07, E15-11,
E15-12, E15-05, E15-02, E15-10 and D15-01) and from and the size of regional floras is mostly politically based. As a
maps by the Secretarı́a de Programación y Presupuesto (SPP), general rule, epiphyte inventories where the sample effort
published in 1983 (locality references E15-08 and E15-09). was different can only be compared if the sampling was
The vegetation overlay was derived from maps (1 : 250.000) adequate, i.e. containing a large portion of all species. The
by INEGI (1985–1988; locality references E15-10, D15-01, fact that local florulas invariably contain many more species
E15-02, E15-07, E15-08, and E15-12) and by SPP (1984; than plot or tree-based inventories suggests otherwise. To
locality references E15-09, E15-11 and E15-05). The veget- estimate the total species richness of an inventory through
ation maps are based on aerial photographs taken between extrapolation, small samples also do not perform well
1972 and 1981. We used the physiographical overlay elab- (Colwell & Coddington, 1994). The aggregated distribution
orated by D. Navarrete (Ecosur) on the basis of Müllerried’s of epiphytes in the forest, moreover, calls for relatively large
(1957) geological map. As to rainfall, four climatic maps samples.
(1 : 500.000) were used, published in 1970 by the Comisión Sample size also influences the quantification of epiphyte
para el estudio del territorio nacional (Cetenal), Universidad ÔsuccessÕ, if the relative contribution of epiphyte diversity to
Nacional Autónoma de México (UNAM); locality references the entire flora is used: the ÔEpiphyte QuotientÕ (EQ; Hos-
15-PI, 15-PII, 15-QVII and 15-QVIII. okawa, 1950). Epiphytes contribute more in smaller plots,
We use Chao’s nonparametric diversity estimator to esti- because their accumulation curves per ground surface area
mate the overall diversity of the samples which typically will are steeper than those of forest trees, as pointed out by
have continuously rising species–accumulation curves (Chao, Nieder et al. (1999, 2001). Single trees may support up to
1984). Her estimator (SChao) provides an estimate of the seventy-seven epiphyte species (Freiberg, 1999). Small plots
completeness of the sampling, enables a comparison between of <1 ha often have EQs of over 40%. In local florulas, the
unequal-sized samples, has a relatively low sensitivity to next larger spatial scale, epiphytes contribute less but reg-
varying sample intensity and species richness, and performs ularly still over 20%, as for example at Rio Palenque (22%),
especially well in data with a preponderance of relatively rare La Selva (23%) and at Maquipucuna (27%). In large regions
species (Colwell & Coddington, 1994; Walther & Morand, such as Peru or the Guianas c. 10% of all vascular plant
1998). SChao is, moreover, easy to compute: SChao ¼ Sobs: þ species are epiphytes, a proportion comparable with the
F12 =2F2 , where Sobs. is the number of observed species; F1, the epiphyte contribution world-wide (Madison, 1977). In
number of species with one record, the singletons and F2, the addition to scale, the EQ is prejudiced by an edge effect if the
number of doubletons. The estimator variance may also be epiphytes in the crown fraction outside the plot boundary
computed: var(SChao) ¼ F2(G4/4 þ G3 þ G2/2), where are included. Transect studies in particular would be subject
G ¼ F1/F2. Computations were made using Excel and the to this source of error, possibly explaining the high EQ
statistical program EstimateS (Colwell, 1997). (35%) in Rio Palenque (Gentry & Dodson, 1987a).
Species nomenclature follows regional checklists and The spatial scale dependence of diversity patterns per
floras (Smith, 1981; Breedlove, 1986; Soto Arenas, 1988; surface area applies also to the three-dimensional space that
Utley, 1994). epiphytes inhabit. Johansson (1974) already pointed out that
comparisons of epiphytes of various regions must be per-
formed on host trees of the same size (and species) because of
RESULTS AND DISCUSSION the strong correlation between host tree and epiphyte. For
example, the significance of the higher diversity in the Sierra
Epiphyte inventories and their assessment
Nevada de Santa Marta plot at 2450 m in comparison with
The number of epiphyte inventories has recently increased the 3100-m plot is hard to appreciate because of the smaller
considerably (Appendix 1). Epiphyte diversity patterns on height of the forest (Sugden & Robins, 1979).
environmental gradients, however, remain elusive because In conclusion, much care is needed with the comparison of
several restrictions hinder a comparison between inventories. the currently available epiphyte inventories. Increased com-
First, it is not always clear whether next to true epiphytes, munication among canopy researchers is essential in the
the parasitic, accidental, facultative, and hemi-epiphytes that development and implementation of standardized protocols
spend part of their life cycle rooted in the soil, were also for comparative studies (Barker & Pinard, 2001). One way
included. Secondly, there is no agreement on the sampling to compare epiphytes between inventories is to plot epi-
unit of inventories. Epiphytes are either sampled per tree or phytes against trees of different sizes (Hietz & Hietz-Seifert,
parts thereof, per ground surface area or included in local 1995b; Hietz-Seifert et al., 1996; Wolf & Konings, 2001).
florulas or regional floras. Epiphyte diversity and abundance The drawing of species accumulation curves, preferably
on trees cannot be compared with epiphytes in surface area against some 3D-sampling unit, facilitates a comparison
plots in the absence of additional data about the structure of between inventories. Such curves visualize the sampling
the forest. An ecologically meaningful comparison between effort and may be used to estimate local species richness
floras requires information about the diversity in habitats (Colwell & Coddington, 1994). In this study we avoid the
within the area, the beta diversity. Thirdly, the sample effort difficult comparison of defined samples by compiling the
may vary considerably between inventories, ranging from plotless data from botanical collections in a data base. That
one to more than 100 trees or from 0.01 to 1.5 ha. The size approach also has its drawbacks that are discussed in the
of florulas often determined by the size of nature reserves section on data quality.
two-thirds (63%) of all epiphytes are orchids (Ibisch et al., because of orchids, but other epiphyte-rich groups such as
1996). The most species-rich orchid genera are Epidendrum Pteridophyta and bromeliads show a similar distribution. In
(seventy-nine), Pleurothallis (fifty-two), Encyclia (thirty- Costa Rica, epiphytic bromeliads are also most common in
seven), Maxillaria (thirty-three), Oncidium (thirty-one), mountain areas (Rossi et al., 1997). Above 2000 m, the
Spiranthes (twenty-one), Lepanthes (twenty-one), and Stelis number of epiphytic aroids and orchids decreases rapidly.
(twenty-one). In the Bromeliaceae, most species are tilland- The Pteridophyta decline in richness in a lower rate; but
sias (sixty-four); of the remaining genera, Catopsis (fourteen) similar to the rate reported for Bolivia and Borneo (Kessler,
and Vriesea (eight) are the most species rich. In contrast to 2001; Kessler et al., 2001). Hence, the relative contribution
South American florulas, Guzmania (two) has only few of ferns to total epiphyte diversity is higher in the temperate
species. Peperomia (Piperaceae) contributes similar (forty- mountain climates (Fig. 4). The common ferns Campylo-
nine species, 4.2%) to the total flora as in Peru (seventy- neuron amphostenon (Kunze ex Klotzsch) Fée and Poly-
seven species, 3.6%). Cactaceae, however, do contribute podium fissidens Maxon are characteristic for high
more in Chiapas than in Peru with 2.3 and 0.6%, respect- elevations.
ively. The contribution of species in the Pteridophyta is with Moreover, our compilation of epiphyte inventories world-
21% also larger than in Peru (16.6%). Possibly this reflects wide points towards an unimodal richness pattern on trop-
biased sampling. Ferns in Chiapas have been the subject of a ical mountains (Appendix 1). In Central America, Barro
regional flora (Smith, 1981) and their relatively low col- Colorado Island (<100 m, 204 species) and La Selva
lecting efficiency indicates a high collecting effort. Remark- (<150 m, 380 species) have fewer epiphytes than mountain
able in the Pteridophyta flora of Chiapas is the high number forests at Monteverde (700–1800 m, 878 species). In South
of Asplenium species (forty-nine). From altitudinal transects America the peak in richness appears to lie somewhat higher.
on Mt Kinabalu, Borneo and Carrasco, Bolivia, respectively, In plots of <1 ha, mid-elevation forests such as Sehuencas
only thirty-seven and twenty-six Asplenium species have (2100–2300 m, 204 species), La Carbonera (2200–2700 m,
been reported, including terrestrial species (Kessler et al., 191 species), Merida (2600 m, 128 species) and Cajanuma
2001). The number of Polypodiaceae (sixty-six) in Chiapas, (2900 m, 138 species), are richer in species than lowland
the largest fern family, was similar to that in those two forests at Rı́o Palenque, (<220 m, 127 species) and Sur-
mountain regions (sixty and fifty-five), while the number of omoni (100 m, fifty-three species), and forests near the upper
Elaphoglossum species (thirty-eight) is higher than in Borneo tree limit at La Caña (3300 m, thirty-nine species) and Santa
(eight), but lower than in Bolivia (eighty-seven). Marta (3000–3200 m, thirty species). The number of epi-
phytes in the florulas of Otonga (2000 m, 193 species), Rı́o
Guajalito (2000 m, 230 species), Maquipucuna (1100–
Epiphyte distribution patterns
2800 m, 441 species) and San Francisco (1800–3150 m, 627
Altitude species) is also higher than in the vast, 10,000 ha, Mabura
We recorded epiphytes up to an altitude of 4100 m on Hill area in Guyana (<100 m, 191 species).
isolated trees above the timberline. Highest species richness In the discussion of the observed altitudinal pattern, it is
in Chiapas is found at mid-elevations between 500 and useful to make a comparison with the non-vascular com-
2000 m (Fig. 3) corroborating Gentry and Dodson’s hypo- ponent of the epiphyte community as bryophytes and lichens
thesis (Gentry & Dodson, 1987a). The pattern is largely provide much information on the altitudinal patterns of
Table 2 Representation of vascular epiphyte families in the data attention to the vegetation in the canopy and we use that
base and their collection efficiency, i.e. the number of species study to summarize the three main altitudinal patterns that
encountered per 100 collections (– ¼ few data) have been observed. First, there was a continuous increase
with elevation in the richness of fruticose lichens. Also in
No. of No. of Contribution Collection
collections species (%) efficiency
terms of abundance these dependent outer canopy specialists
thrive on the combination of high light intensities and
Angiosperms re-occurring (fog) precipitation. This pattern exhibited by
Araceae 838 67 5.7 8.0 the specialist fruticose lichens, however, was exceptional.
Araliaceae 273 14 1.2 5.1 Secondly, there was a continuous decline in species num-
Asteraceae 39 3 0.2 7.7 bers with elevation for epiphytic mosses and for crustose
Begoniaceae 210 23 2.0 11.0 and foliose lichens. This decline was most significant in the
Bignoniaceae 3 2 0.2 –
cool and humid forests at higher elevations where branches
Bromeliaceae 1087 101 8.6 9.3
were enveloped in a heavy cloak of bryophytes, mainly
Burmanniaceae 3 2 0.2 –
Cactaceae 90 27 2.3 30.0 liverworts, suggesting increased competition for space. In
Crassulaceae 39 12 1.0 30.8 addition, mass effect, i.e. the influx of propagules from an
Cyclanthaceae 8 4 0.3 – adjacent core area to an area where the species cannot be
Dioscoreaceae 3 1 0.1 – self-maintaining (Shmida & Wilson, 1985), has been pos-
Ericaceae 147 14 1.2 9.5 tulated to produce a decline of richness with elevation,
Gesneriaceae 63 10 0.8 15.9 in combination with Rapoport’s rule (Stevens, 1992).
Guttiferae 162 14 1.2 8.6 According to Rapoport’s rule, the elevational (and latitu-
Lentibulariaceae 2 1 0.1 – dinal) range of species increases with increasing elevation.
Liliaceae 1 1 0.1 –
According to Stevens’s Rapoport rescue hypothesis, low-
Marcgraviaceae 51 6 0.5 11.8
lands may thus be richer in species because they are
Moraceae 31 4 0.3 12.9
Onagraceae 38 1 0.1 2.6 potentially sink habitats for a larger number of species. We
Orchidaceae 5350 568 48.4 10.6 tested Rapoport’s rule for our data and indeed found a
Piperaceae 629 52 4.4 8.3 good correlation (Pearson) between the mid-elevation of
Rubiaceae 3 1 0.1 – the distribution of the species and their altitudinal range
Solanaceae 17 1 0.1 5.9 (only species with more than ten occurrences; n ¼ 378,
Subtotal 9087 929 79.2 10.2 r ¼ 0.36, P < 0.001). According to Stevens’s hypothesis we
Pteridophyta, i.e. ferns and allies expect a monotonically decrease of richness with elevation,
Adiantaceae 56 6 0.5 10.7 but observed a unimodal pattern. This pattern, the third,
Aspleniaceae 697 49 4.2 7.0 was in Wolf’s study shown by the liverworts, the group
Blechnaceae 146 12 1.0 8.2 with by far the largest number of species in his transect.
Dennstaedtiaceae 8 2 0.2 – Interestingly, the mass effect was also associated with the
Dryopteridaceae 6 2 0.2 – presence of a mid-elevation zone of high richness as it
Grammitidaceae 70 10 0.9 14.3 coincided with a zone of overlap between a low- and a
Hymenophyllaceae 319 32 2.7 10.0
high-elevation flora (Wolf, 1993a). The presence of a zone
Lomariopsidaceae 370 38 3.2 10.3
where the altitudinal distributions of many species overlap
Lycopodiaceae 42 8 0.7 19.1
Nephrolepidaceae 71 7 0.6 9.9 may also have induced the high species richness at mid-
Polypodiaceae 1310 66 5.6 5.0 elevations in Chiapas. Most species-rich families have their
Psilotaceae 1 1 0.1 – highest diversity in this belt, which suggests a common
Schizaeaceae 4 2 0.2 – factor (Fig. 3). In addition, forests at both low and high
Tectariaceae 19 2 0.2 10.5 elevations have a distinct epiphyte flora (Table 3). In this
Vittariaceae 69 6 0.5 8.7 view, between 500 and 2000 m, typical tropical lowland
Woodsiaceae 1 1 0.1 – species coincide with the temperate species from the high-
Subtotal 3189 244 20.8 7.7 lands. The temperate origin of the mountain vegetation of
Total 12,276 1173 100.0 9.6 Chiapas is evidenced by the presence of holarctic tree
genera such as Abies, Acer, Alnus, Juniperus, Pinus,
Prunus, Quercus, Sambucus, Ulmus and Viburnum.
epiphytes on tropical mountains to date (Reenen & A separate factor that may contribute to the mid-eleva-
Gradstein, 1983; Kürschner, 1990; Frahm & Gradstein, tional hump results from the geometric constraint on species
1991; Wolf, 1993a; Kessler, 2000). The study of non-vas- ranges (e.g. Colwell & Lees, 2000). When species are dis-
cular epiphytes is often easier than that of their vascular tributed stochastically within a bounded domain (from sea
co-inhabitants because the classification of sterile specimen level to mountaintop), null models predict that there is a
is mostly possible, aided by an increasing number of taxo- mid-domain peak in richness. Unfortunately, the high
nomic reference books, and because of relatively small number of rare species in the data base render a detailed
minimal areas (Gradstein et al., 1996). Wolf (1993a), analysis of the altitudinal ranges of species impossible and
working in the northern Andes of Colombia, paid full the presence of zones of overlap or a mid-domain effect
could thus not be confirmed. The occurrence of overlap, a diversity is especially high on mid-elevations at wet moun-
mid-domain effect and Stevens’s Rapoport rescue hypothesis tains: finer niche partitioning and evolutionary explosion.
are not mutually exclusive. In general, we agree with Lawton The latter is associated with the dynamic character of the
(1996) and Rahbek (1997) that the quest for single expla- young Andean mountains. Our data show that the species
nations of patterns is unhelpful. richness patterns are comparable also on more stable
Perhaps the high diversity between 500 and 2000 m mountains in Chiapas.
results from a higher diversity of habitats, because next to Finally, in Bolivia a mid-elevational belt of high epiphyte
the wet mountain forests, the dry Central Depression also pteridophyte diversity is correlated with high amounts of
falls in this altitudinal range (Fig. 1). A separate analysis of precipitation in that zone (Kessler, 2001). In Chiapas,
the diversity pattern exclusively in the wet Sierra Madre however, there is no such relationship. The collection sites in
mountain range, however, also showed a unimodal distri- the lowlands (<500 m) receive more rainfall annually than
bution (Table 4). In addition to habitat diversity, Gentry and the sites between 500 and 2000 m, respectively, 2580 and
Dodson offer two more hypotheses as to why species 1750 mm.
Table 3 Relative contribution of species in the data base per altitudinal interval (number of records/total number of records in a particular
interval, times 10,000). Only species with a preference for a certain interval are shown, arbitrarily defined as being collected there at least ten
times more often than in any of the other intervals. Rare species having a relative contribution <0.50% in those intervals are not considered
Table 4 Number of epiphyte species in the Sierra Madre region per region is not poorer than many mountain forests. In the
altitudinal interval. Given are the number of records (n), the number Colombian part of that same region, 0.4 and 0.1 ha plots
of observed species (Sobs.), and the estimated number of species contain 140 species on average which is also remarkably rich
(SChao) with the 95% confidence interval (Galeano et al., 1998). On the other side of the Andean
Altitudinal interval n Sobs. SChao SChao 95% CI
mountains, the Amazonian Rı́o Caquetá region is with 212
species in a 0.75 ha inventory also rich in epiphytes. Pre-
0–500 m 279 108 166.78 (162.6, 170.9) sumably, the high diversity in the Chocó and Caquetá
500–1000 m 172 106 211.09 (204.3, 217.9) regions is related to the high amounts of annual rainfall
1000–1500 m 210 154 394.67 (384.1, 405.2) (Gentry & Dodson, 1987a) of c. 3000, 7000 and 3000 mm,
1500–2000 m 454 228 408.88 (403.2, 414.6) respectively. Next, the rainfall–diversity relationship is
2000–2500 m 437 193 306.64 (302.3, 311.0) analysed.
2500–3000 m 159 93 156.03 (151.4, 160.7)
>3000 m 37 28 83.13 (69.2, 97.0)
Rainfall
0–4100 1748 605 1302.69 (1294.2, 1311.1)
Recent epiphyte inventories in neotropical lowland forests
corroborate Gentry & Dodson’s (1987a) hypothesis that
species richness increases with the amount of rainfall
Possibly, a mid-elevational zone of higher species richness (Fig. 5). For the entire state of Chiapas that in contrast also
is not encountered everywhere. Ibisch et al. (1996) report a includes mountain forests, a different pattern emerges
decrease of epiphyte richness with elevation for the whole of (Fig. 6). After an initial rise in species richness, the number
Peru and the lowland florula at Rı́o Palenque (<220 m, 227 of epiphytes decreases again when rainfall exceeds 2500 mm
species) in the Ecuadorian part of the Chocó biogeographical annually. All species-rich groups exhibit this pattern, except
400
La Selva, Costa Rica
350
300
Number of species
250
R o Palenque, Ecuador
BCI, Panama R o CaquetÆ, Colombia
200
Mabura Hill, Guyana
Coqu , Colombia
the aroids that remain comparatively stable. The positive 2500 mm, while aroids remain stable (Fig. 6). In general,
relationship between rainfall and epiphyte diversity thus wind-dispersed epiphytes and lianas are better represented in
breaks down when mountain sites at high elevations are forests that are relatively dry (Gentry & Dodson, 1987a;
included. The cool mountains are effectively more humid Gentry, 1991). In contrast, the extremely wet forests in the
than their amounts of rainfall suggest because of the high Chocó have an unusual high number of animal-dispersed
average relative humidity of the air and low evapotranspi- species (Gentry, 1986). As Gentry & Dodson (1987a) pos-
ration (Wolf, 1993a). Cloud precipitation, not measured in tulate, in wet forests wind-dispersed propagules are ham-
rainfall gauges, may further enhance the moisture availab- pered in their establishment, i.e. dispersal and attachment, in
ility of the epiphyte habitat (e.g. Clark et al., 1998). the face of abundant rainfall. As to bromeliads, rainfall may
Even when only records from low elevations (<1000 m) fuse the coma hairs of seeds to an inert mass (pers. observ.).
are considered, a positive relationship with rainfall in Moreover many orchids, bromeliads and ferns are well
Chiapas could not be found. Again in important plant adapted to survive periods of drought (Benzing, 1990).
groups such as ferns, the bromeliads and orchids, the num- In agreement with the establishment hypothesis, we
ber of epiphytes decreases when annual rainfall exceeds observe that with increased rainfall the characteristic species
100
Other epiphytes
Piperaceae
90
Bromeliaceae
Ferns and allies
80
Orchidaceae
70 Araceae
60
Percentage
50
Figure 7 Familial composition of epiphyte
floras in neotropical lowland rain forests with 40
high amounts of rainfall. La Selva, Costa
30
Rica (Hartshorn & Hammel, 1994). Rı́o
Palenque, Ecuador (Gentry & Dodson, 20
1987a). Barro Colorado Island, Panama
(Croat, 1978). Rı́o Caquetá, Colombia 10
(Benavides, 2002). Horquetas, Costa Rica;
0
the ÔothersÕ category includes Piperaceae Horquetas, La Selva, Río Caquetá, Río Palenque, BCI, Suromoni, This study,
(Whitmore & R. Peralta, 1985). Suromoni, Costa Rica Costa Rica Colombia Ecuador Panama Venezuela Mexico
Venezuela (Engwald, 1999). (4000 mm) (4000 mm) (3060 mm) (2980 mm) (2750 mm) (2700 mm) (>2500 mm)
in the vegetation shift from wind-dispersed orchids and ferns fifty-three orchid species were found. Most aroids had a tree-
to Araceae (Table 5). The decrease in anemochoric species is, dependent growth form, Anthurium and Philodendron being
however, not fully compensated by zoochoric species. This the largest genera, with forty-eight epiphyte species in total.
appears contrary to other neotropical wet lowland forests, Our data base of Chiapas (75,000 km2) contains only sixty-
where aroids contribute more to total diversity (Fig. 7). In a seven epiphytic Araceae. Rı́o Palenque, as well in the Chocó
0.9-ha inventory in the extremely wet forests along the Pacific biogeographic region, is also rich in Araceae (Gentry &
Coast in Colombia (annual rainfall of 5100–7150 mm), Dodson, 1987a) and this pattern repeats itself on trees of wet
aroids were with 100 species even the largest plant family, forest in the Amazon where aroids have fifty-two species
comprising over 10% of the total vascular flora (Galeano compared with thirty-seven orchids and thirty-seven ferns
et al., 1998). Alongside less than fifteen bromeliads and only (Benavides, 2002). Aroid preponderance in wet climates is
1377.5
(column) totals is significantly different between all regions, except for the two coastal plains (unpaired t-test, P < 0.001). Ô–Õ no sufficient data (n < 20) to calculate SChao. *Note that the
87.2
55.2
30.0
14.3
14.5
82.3
114.9
648.0
126.5
275.5
SChao
Table 6 The number of observed epiphyte species, Sobs. and the number of estimated species, SChao, per physiographical region in epiphyte rich families. The SChao diversity of the
Costa Rica also contains many Araceae (ninety-nine species)
(n ¼ 12,276*)
and from wet Barro Colorado Island (BCI) in Panama
of Chiapas
The whole
twenty-four epiphytic aroids are reported compared with
1173
67
27
12
14
10
52
90
relatively few (eighty-two) tree-dwelling orchids (Croat,
101
568
244
Sobs.
total number of collections (12,276) is larger than that of the summed regions (11,724), because in this column 552 difficult to assign borderline cases are also included
1978; Hartshorn & Hammel, 1994). Also on trees in the
Mexican forest at Los Tuxtlas, the northernmost tropical
lowland rain forest, aroids are more species rich than orchids,
SChao
Plain (n ¼ 26)
89.7
bromeliads and Pteridophyta (Hietz-Seifert et al., 1996).
Gulf Coastal
–
–
–
–
–
–
–
–
–
–
The low manifestation of Araceae in Chiapas may be
related to the historical biogeography of the family. As a
Sobs.
separate hypothesis, we propose that the relatively high
25
4
3
0
0
0
1
9
2
4
2
contribution of anemochorous epiphytes is related to the dry
season that gives these plants a good opportunity to disperse.
338.5
210.3
725.0
44.3
62.5
35.3
41.3
SChao
In contrast to the aforementioned wet lowland forests, the
forests in southern Mexico are subject to a distinct dry
(n ¼ 2879)
–
–
–
–
Highlands
period that lasts several months. As the rainfall regime is an
Eastern
important element on which the internal division of Chiapas
Sobs.
36
40
23
31
277
128
545
4
0
2
4
in physiographical regions and associated vegetation types is
based, we expect that regions and forest formations have
distinctive epiphytes (Breedlove, 1978).
38.3
82.3
74.5
18.0
58.6
49.6
Central Plateau
385.0
191.8
864.1
SChao
(n ¼ 3529)
–
–
Physiographical region
For the analysis of the distribution of species over the seven
Sobs.
physiographical regions in the state (Fig. 1), 11,724 records
28
62
14
10
39
39
293
155
650
9
1
were used that could be attributed to a particular region,
comprising 1153 species. Of all regions, the Central Plateau
is the richest in epiphytes, and richer than the Eastern and
46.1
68.4
35.1
53.1
5.5
5.0
345.2
196.6
748.4
SChao
Northern Highlands that have lower elevations and higher (n ¼ 1482)
–
–
Highlands
Northern
183
143
492
7
1
5
5
32
50
27
39
species. The similarity in diversity suggests that these regions
share many species, but in reality the floristic similarity
between the regions is always <60% (Table 7). With con-
216.1
150.0
548.3
37.1
70.1
19.1
23.6
32.4
SChao
158
114
417
1
4
2
30
46
13
23
26
orchids. Next to differences in the historical biogeography of
that region, this may be because of the favourable combi-
nation of high temperatures at lower elevations and high
58.1
66.4
31.2
49.0
374.1
186.0
790.1
SChao
7
Sierra Madre
–
–
48
50
10
28
39
263
137
588
1
7
5
Vegetation type
In total, 8485 collections could be ascribed to one of the
forest formations of Chiapas, following Breedlove’s classifi-
115.8
10
12
51
7
1
0
1
2
6
8
4
Gesneriaceae
Crassulaceae
Pteridophyta
Orchidaceae
Ericaceae
Others
Total
Table 7 Floristic similarity (%) between the physiographical regions, using the qualitative Sørensen index
Table 8 Relative contribution of species in the data base per region (number of records/total number of records in a particular region,
times 10,000). Only species with a preference for a certain region are shown, arbitrarily defined as being collected there at least ten times
more often than in any of the other regions. Rare species having a relative contribution <0.50% in those regions are not considered as are
the coastal plains because there the sampling effort was low (<100 records)
collections) from sites that currently are devoid of any forest well adapted to the semi-deciduous habit of these trees. A
cover, however, it must be feared that their local populations thick water-absorbing bark may facilitate bromeliad survi-
have become extinct. The large number of rare species in the val, in particular that of the seedlings (Castro-Hernandez
data base raises concerns that recently many species may et al., 1999). High diversity is not associated with a unique
have gone extinct in Chiapas, possibly before they were ever flora, Tillandsia carlsoniae L.B. Smith being the only
sampled. The current situation is likely even gloomier exclusive species.
because the assessment of forest cover is based on aerial
photographs taken over 20 years ago. In the Central Plateau
CONCLUSIONS
of Chiapas, annual deforestation rates for 1974–1984 and
1984–1990 were 1.58 and 2.13%, respectively (Ochoa- For a region at 15N, the flora of Chiapas with 1173 species
Gaona & González-Espinosa, 2000). is remarkable rich in vascular epiphytes. Presumably this
Most epiphytes are found in the montane rain forest relates to the high diversity in habitats of the state. Each
(Table 9). Both in the tropical and lower montane rain physiographical region and forest formation supports char-
forest and in the two other forest types from lower eleva- acteristic epiphytes, indicating a low expansion rate and/or
tions that are all subject to a more prolonged dry season, habitat specialization.
the important Orchidaceae and Pteridophyta are less We were able to corroborate Gentry & Dodson’s (1987a)
diverse. However, some of these formations are not well hypothesis that on neotropical mountains most species are
collected, and additional species are likely to be found. present in a mid-elevational belt, in Chiapas between 500
Remarkably, the bromeliads are most diverse in the pine– and 2000 m altitude. A predicted increase of epiphyte
oak forest of the highlands. Of the two main tree genera, diversity when rainfall increases, however, was not
bromeliads are mostly found on oaks and are apparently observed. After an initial rise, the number of species
Table 9 The number of observed epiphyte species, Sobs., and the number of estimated species, SChao, per vegetation type in epiphyte rich
families. Forest formations from Breedlove (1978). The SChao diversity of the column totals is significantly different between all formations
(unpaired t-test, P < 0.001). The montane rain forest includes the evergreen cloud forest and the pine–oak Liquidamber Forest. Ô–Õ, no sufficient
data (n < 20) to calculate SChao. For family totals for the entire state of Chiapas, see Table 6
decreases when the annual amount of rainfall exceeds Nacional para el uso y Conocimiento de la Biodiversidad,
2500 mm. In southern Mexico, even the wettest forests are CONABIO, grants B060 and L050.
subject to a dry period. Wind-dispersed orchids, bromeliads
and Pteridophyta are the most species-rich groups. They
REFERENCES
appear well-adapted to the dry season when they disperse
their seeds and decline in the wetter forest. In the wettest Aguirre-Leon, E. (1992) Vascular epiphytes of Mexico: a
forests, animal-dispersed plants in the Araceae gain in preliminary inventory. Selbyana, 13, 72–76.
importance, but these are less diverse as in rain forests Annaselvam, J. & Parthasarathy, N. (2001) Diversity and
elsewhere. As a consequence, in Chiapas the rainfall–diver- distribution of herbaceous vascular epiphytes in a tropical
sity relationship is unusual. We conclude that next to the evergreen forest at Varagalaiar, Western Ghats, India.
Biodiversity and Conservation, 10, 317–329.
annual amount of rainfall, its distribution in time determines
Bader, M., Van Dunne, H.J.F. & Stuiver, H.J. (2000) Epiphyte
the composition of the epiphyte community.
distribution in a secondary cloud forest vegetation; a case
As we were able to interpret the diversity patterns, we also
study of the application of GIS in epiphyte ecology. Ecotro-
conclude that herbarium collections may be used to obtain a
pica, 6, 181–195.
good impression of those patterns when large numbers of Barker, M.G. & Pinard, M.A. (2001) Forest canopy research:
collections are available and despite the non-systematic sampling problems, and some solutions. Plant Ecology, 153,
sampling of the botanist. 23–28.
Benavides, A.M. (2002) Diversidad y distribución de epı´fitas
ACKNOWLEDGMENTS vasculares en el medio Caquetá, Amazonı´a noroccidental
Colombiana. Thesis, Universidad de Antioquia, Medellin,
We thank Juan Castillo, José L. Godónez-A., Teresa Santi- Colombia.
ago-V. and in particular Guadalupe Olalde who between Benzing, D.H. (1990) Vascular epiphytes. Cambridge University
them compiled most of the information in the data base. We Press, Cambridge, UK.
are grateful for the hospitality and co-operation encountered Biedinger, N. & Fischer, E. (1996) Epiphytic vegetation and
in all the institutions where we examined the botanical ecology in Central African forests (Rwanda, Zaı̈re). Ecotro-
collection. The information of orchid enthusiast Cisco Dietz pica, 2, 121–142.
yielded many new records for the state. Miguel Angel Boggon, J., Funk, V.A., Kelloff, C., Hoff, M., Cremers, G. &
Castillo and Dario Navarrete at the Ecosur GIS laboratory Feuillet, C. (1997) Checklist of the plants of the Guianas.
are acknowledged for providing digitized information on the Centre for the study of biological diversity, University of
distribution of rainfall, forest formations and physiographi- Guyana, Georgetown, Guyana.
cal regions. Gerard Oostermeijer and two anonymous Bøgh, A. (1992) The composition and distribution of the
vascular epiphyte flora of an Ecuadorean montane rain forest.
reviewers provided valuable suggestions to improve the
Selbyana, 13, 25–34.
manuscript. Financial support was provided by Comisión
Breedlove, D.E. (1978) The phytogeography and vegetation of Freiberg, M. (1996) Spatial distribution of vascular epiphytes on
Chiapas (Mexico). Vegetation and vegetational history of three emergent canopy trees in French Guiana. Biotropica,
northern Latin America (ed. by A. Graham), pp. 149–165. 28, 345–355.
California Academy of Sciences, San Francisco, USA. Freiberg, M. (1999) The vascular epiphytes on a Virola michelii
Breedlove, D.E. (1981) Introduction to the flora of Chiapas. tree (Myristicaceae) in French Guiana. Ecotropica, 5, 75–81.
Flora of Chiapas (ed. by D.E. Breedlove), pp. 1–35. California Freiberg, M. & Freiberg, E. (2000) Epiphyte diversity and
Academy of Sciences, San Francisco, California, USA. biomass in the canopy of lowland and montane forests in
Breedlove, D.E. (1986) Listados floristicos de Mexico IV. Flora Ecuador. Journal of Tropical Ecology, 16, 673–688.
de Chiapas. UNAM, Mexico City, Mexico. Galeano, G., Suarez, S. & Balslev, H. (1998) Vascular plant
Bussmann, R.W. (2001) Epiphyte diversity in a tropical Andean species count in a wet forest in the Choco area on the Pacific
forest–Reserva Biológica San Francisco, Zamora-Chinchipe, coast of Colombia. Biodiversity and Conservation, 7, 1563–1575.
Ecuador. Ecotropica, 7, 43–59. Gentry, A.H. (1986) Species richness and floristic composition
Castro-Hernandez, J.C., Wolf, J.H.D., Garcia-Franco, J.G. & of Chocó region plant communities. Caldasia, 15, 71–91.
Gonzalez-Espinosa, M. (1999) The influence of humidity, Gentry, A.H. (1991) Breeding and dispersal systems of lianas.
nutrients and light on the establishment of the epiphytic The biology of vines (ed. by F.E. Putz and H.A. Mooney),
bromeliad Tillandsia guatemalensis in the highlands of pp. 393–423. Cambridge University Press, Cambridge.
Chiapas, Mexico. Revista de Biologia Tropical, 47, 763– Gentry, A.H. & Dodson, C.H. (1987a) Diversity and biogeo-
773. graphy of neotropical vascular epiphytes. Annals of the
Catling, P.M. & Lefkovitch, L.P. (1989) Associations of Missouri Botanical Garden, 74, 205–233.
vascular epiphytes in a Guatemalan cloud forest. Biotropica, Gentry, A.H. & Dodson, C.H. (1987b) Contribution of
21, 35–40. nontrees to species richness of a tropical rain forest.
Chao, A. (1984) Nonparametric estimation of the number of Biotropica, 19, 149–156.
classes in a population. Scandinavian Journal of Statistics, 11, Gradstein, S.R., Hietz, P., Lücking, R., Lücking, A., Sipman,
265–270. H.J.M., Vester, H.F.M., Wolf, J.H.D. & Gardet, E. (1996)
Clark, K.L., Nadkarni, N.M., Schaefer, D. & Gholz, H.L. How to sample the epiphytic diversity of tropical rain forests.
(1998) Atmospheric deposition and net retention of ions by Ecotropica, 2, 59–72.
the canopy in a tropical montane forest, Monteverde, Costa Greeney, H.F. (2001) The insects of plant-held waters: a review
Rica. Journal of Tropical Ecology, 14, 27–45. and bibliography. Journal of Tropical Ecology, 17, 241–260.
Colwell, R.K. (1997) EstimateS: statistical estimation of species Haber, W.A. (2001) Number of species with different plant
richness and shared species from samples. Version 5. User’s growth forms. Monteverde: ecology and conservation of a
Guide and application published at: http://viceroy.eeb. tropical cloud forest (ed. by N.M. Nadkarni and N.T.
uconn.edu/estimates. Wheelwright), pp. 519–522. Oxford University Press, New
Colwell, R.K. & Coddington, J.K. (1994) Estimating terrestrial York, Oxford.
biodiversity through extrapolation. Philosophical Transac- Hartshorn, G.S. & Hammel, B.E. (1994) Vegetation types and
tions of the Royal Society (Series B), 345, 101–118. floristic patterns. La Selva: ecology and natural history of a
Colwell, R.K. & Lees, D.C. (2000) The mid-domain effect: neotropical rain forest (ed. by L.A. McDade, K.S. Bawa, H.A.
geometric constraints on the geography of species richness. Hespenheide and G.S. Hartshorn). The University of Chicago
TRENDS in Ecology & Evolution, 15, 70–76. Press, Chicago.
Croat, T. (1978) Flora of Barro Colorado Island. Stanford Hietz, P. (1997) Population dynamics of epiphytes in a Mexican
University Press, Stanford, USA. humid montane forest. Journal of Ecology, 85, 767–775.
Delgadillo, M.C. & Cárdenas-S, A. (1989) Phytogeography of Hietz, P. & Hietz-Seifert, U. (1995a) Structure and ecology of
high-elevation mosses from Chiapas, Mexico. The Bryologist, epiphyte communities of a cloud forest in central Veracruz,
92, 461–466. Mexico. Journal of Vegetation Science, 6, 719–728.
Ek, R.C. (1997) Botanical diversity in the tropical rain forest of Hietz, P. & Hietz-Seifert, U. (1995b) Composition and ecology
Guyana. PhD Thesis, Universiteit Utrecht, Utrecht, The of vascular epiphyte communities along an altitudinal
Netherlands. gradient in Central Veracruz, Mexico. Journal of Vegetation
Engwald, S. (1999) Diversität und Ökologie der vaskulären Science, 6, 487–498.
Epiphyten in einem Berg- und einem Tieflandregenwald in Hietz-Seifert, U., Hietz, P. & Guevara, S. (1996) Epiphyte
Venezuela. PhD Thesis, Universität Bonn, Bonn, Germany. vegetation and diversity on remnant trees after forest
Flores-Palacios, A. & Garcia-Franco, J.G. (2001) Sampling clearance in southern Veracruz, Mexico. Biological Conser-
methods for vascular epiphytes: their effectiveness in vation, 75, 103–111.
recording species richness and frequency. Selbyana, 22, Hofstede, R.G.M., Dickinson, K.J.M. & Mark, A.F. (2001)
181–191. Distribution, abundance and biomass of epiphyte-lianoid
Foster, R.B. (1990) The floristic composition in the Rio Manu communities in a New Zealand lowland Nothofagus-podo-
floodplain forest. Four neotropical rainforests (ed. by A.H. carp temperate rain forest: tropical comparisons. Journal of
Gentry), pp. 99–111. Vail-Ballou Press, Binghamton, New Biogeography, 28, 1033–1049.
York, USA. Hosokawa, T. (1950) Epiphyte-quotient. Botanical Magazine
Frahm, J.-P. & Gradstein, S.R. (1991) An altitudinal zonation of Tokyo, 63, 739–740.
tropical rain forests using bryophytes. Journal of Biogeogra- Hubbell, S.P. (2001) The unified neutral theory of biodiversity
phy, 18, 669–678. and biogeography. Princeton University Press, Princeton.
Ibisch, P.L. (1996) Neotropische Epiphytendiversität – das Moffett, M.W. (1993) The high frontier: exploring the tropical
Beispiel Bolivien. Martina Galunder-Verlag, Wiehl, rain forest canopy. Harvard University Press, Cambridge,
Germany. Massachusetts, USA.
Ibisch, P.L., Boegner, A., Nieder, J. & Barthlott, W. (1996) Müllerried, F.K.G. (1957) Geologı´a de Chiapas. Gobierno del
How diverse are neotropical epiphytes? An analysis based on estado de Chiapas, Tuxtla Gutiérrez, Mexico.
the ÔCatalogue of the flowering plants and gymnosperms of Nadkarni, N.M. (1994) Diversity of species and interactions in
PeruÕ. Ecotropica, 2, 13–28. the upper tree canopy of forest ecosystems. American
Ingram, S.W. & Nadkarni, N.M. (1993) Composition and Zoologist, 34, 70–78.
distribution of epiphytic organic-matter in a neotropical Nieder, J., Engwald, S. & Barthlott, W. (1999) Patterns of
cloud forest, Costa-Rica. Biotropica, 25, 370–383. neotropical epiphyte diversity. Selbyana, 20, 66–75.
Janzen, D.H. & Liesner, R. (1980) Annotated check-list of Nieder, J., Prosperi, J. & Michaloud, G. (2001) Epiphytes and
plants of lowland Guanacaste Province, Costa Rica, exclusive their contribution to canopy diversity. Plant Ecology, 153,
of grasses and non-vascular cryptogams. Brenesia, 18, 15–90. 51–63.
Jaramillo, J. (2001) Flora de Rı́o Guajalito. The flora of the Rı´o Nowicki, C. (2001). Epifitas vasculares de la Reserva Otonga.
Guajalito mountain rain forest (Ecuador). Results of the Epiphytes and canopy fauna of the Otonga rain forest
Bonn – Quito epiphyte project, funded by the Volkswagen (Ecuador). Results of the Bonn – Quito epiphyte project,
foundation. (ed. by J. Nieder and W. Barthlott), pp. 47–322. funded by the Volkswagen foundation. (ed. by J. Nieder
Books on Demand GmbH, Bonn. and W. Barthlott), pp. 115–160. Books on Demand GmbH,
Jarman, S.J. & Kantvilas, G. (1995) Epiphytes on an Old Huon Bonn.
Pine Tree (Lagarostrobos-Franklinii) in Tasmanian Rain- Ochoa-Gaona, S. & González-Espinosa, M. (2000) Land use
Forest. New Zealand Journal of Botany, 33, 65–78. and deforestation in the highlands of Chiapas, Mexico.
Johansson, D. (1974) Ecology of vascular epiphytes in west Applied Geography, 20, 17–42.
African rain forest. Acta Phytogeographica Suecica, 59, 1–123. Olmsted, I. & Gómez-Juarez, M. (1996) Distribution and
Kelly, D.L., Tanner, E.V.J., Lughadha, E.M.N. & Kapos, V. conservation of epiphytes on the Yucatán peninsula. Selbya-
(1994) Floristics and biogeography of a rain-forest in the na, 17, 58–70.
Venezuelan Andes. Journal of Biogeography, 21, 421–440. Perry, D.R. (1978) A method of access into the crowns of
Kessler, M. (2000) Altitudinal zonation of Andean cryptogam emergent and canopy trees. Biotropica, 10, 155–157.
communities. Journal of Biogeography, 27, 275–282. Pianka, E.R. (1966) Latitudinal gradients in species diversity: a
Kessler, M. (2001) Pteridophyte species richness in Andean review of concepts. American Naturalist, 100, 33–46.
forests in Bolivia. Biodiversity and Conservation, 10, 1473– Prance, G.T. (1994) A comparison of the efficacy of higher taxa
1495. and species numbers in the assessment of the biodiversity in
Kessler, M., Parris, B.S. & Kessler, E. (2001) A comparison of the neotropics. Philosophical Transactions of the Royal
the tropical montane pteridophyte floras of Mount Kinabalu, Society (Series B), 345, 89–99.
Borneo, and Parque Nacional Carrasco, Bolivia. Journal of Rahbek, C. (1997) The relationship among area, elevation, and
Biogeography, 28, 611–622. regional species richness in neotropical birds. American
Köster, N., Kreft, H., Schmit-Neuerburg, V., Küper, W., Nieder, Naturalist, 149, 875–902.
J. & Barthlott, W. (2003) High variability in diversity and Rauer, G. & Rudolph, D. (2001) Vaskuläre Epiphyten eines
abundance of vascular epiphytes within Amazonia. Abstract westandinen Bergregenwaldes in Ecuador. The flora of the
16. Jahrestagung der Gesellschaft für Tropenökologie – Rı´o Guajalito mountain rain forest (Ecuador). Results of the
Tropische Biodiversität im globalen Wandel. Universität Bonn – Quito epiphyte project, funded by the Volkswagen
Rostock, Rostock, Germany. foundation. (ed. by J. Nieder and W. Barthlott), pp. 323–470.
Kress, W.J. (1986) The systematic distribution of vascular Books on Demand GmbH, Bonn.
epiphytes: an update. Selbyana, 9, 2–22. Reenen, G.B.A. & Gradstein, S.R. (1983) A transect analysis of
Kürschner, H. (1990) Höhengliederung (ordination) von the bryophyte vegetation along an altitudinal gradient on the
epiphytischen Laub-und Lebermoosen in Nord-Borneo (Mt. Sierra Nevada de Santa Marta, Colombia. Acta Botanica
Kinabalu). Nova Hedwigia, 51, 77–86. Neerlandica, 32, 163–165.
Lawton, J.H. (1996) Patterns in ecology. Oikos, 75, 145–147. Ribeiro, J.E.L., Nelson, B.W., Silva, M.F., Martins, L.S.S. &
Lowman, M.D. (2001) Plants in the forest canopy: some Hopkins, M. (1994) Reserva florestal Ducke: Diversi-
reflections on current research and future direction. Plant dade e composição da flora vascular. Acta Amazonica, 24,
Ecology, 153, 39–50. 19–30.
Madison, M. (1977) Vascular epiphytes: their systematic Rossi, M.R., Mendez, V.H. & Monge-Najera, J. (1997)
occurrence and salient features. Selbyana, 2, 1–13. Distribution of Costa Rican epiphytic bromeliads and the
Madison, M. (1979) Distribution of epiphytes in a rubber Holdridge Life Zone System. Revista De Biologia Tropical,
plantation in Sarawak. Selbyana, 5, 107–115. 45, 1021–1031.
Malcolm, J.R. (1991) Comparitive abundances of neotropical Rzedowski, J. (1992) Diversidad y origen de la flora fanerogá-
small mammals by trap height. Journal of Mammalogy, 72, mica de México. La diversidad biológica de Iberoamérica (ed.
188–192. by G. Halffter), pp. 315–335. Acta Zoologica Mexicana
Mitchell, A.W., Secoy, K. & Jackson, T. (ed.) (2002) The global volumen especial.
canopy handbook. Techniques of access and study in the Schimper, A.F.W. (1888) Die epiphytische Vegetation Ameri-
forest roof. Global Canopy Programme, Oxford, UK. kas. Fischer, Jena.
Schneider, J.V. (2001) Diversity, structure, and biogeography of Werneck, M.d.S. & Espı́rito-Santo, M.M.d. (2002) Species
a successional and mature upper montane rain forest of the diversity and abundance of vascular epiphytes on Vellozia
Venezuelan Andes. PhD Thesis, Johann Wolfgang Goethe- piresina in Brazil. Biotropica, 34, 51–57.
Universität, Frankfurt am Main, Germany. Whitmore, T.C. & R. Peralta, K.B. (1985) Total species count in
Shmida, A. & Wilson, M.V. (1985) Biological determinants of a Costa Rican tropical rain forest. Journal of Tropical
species diversity. Journal of Biogeography, 12, 1–20. Ecology, 1, 375–378.
Sipman, H.J.M. & Wolf, J.H.D. (1998) Provisional checklist Winkler, H. & Preleuthner, M. (2001) Behaviour and ecology of
for the lichens of Chiapas. Acta Botánica Mexicana, 45, birds in tropical rain forest canopies. Plant Ecology, 153,
1–29. 193–202.
Smith, R.F. (1970) The vegetation structure of a Puerto Rican Wolf, J.H.D. (1993a) Diversity patterns and biomass of
rain forest before and after short-term gamma radiation. A epiphytic bryophytes and lichens along an altitudinal gradient
tropical rain forest – a study of irradiation and ecology at El in the northern Andes. Annals of the Missouri Botanical
Verde, Puerto Rico (ed. by H.T. Odum and R.F. Pigeon), pp. Garden, 80, 928–960.
D103–D140. US Atomic Energy Commission, Oak Ridge, Wolf, J.H.D. (1993b) Ecology of epiphytes and epiphyte
Tennessee. communities in montane rain forests, Colombia. PhD Thesis,
Smith, A.R. (1981) Pteridophytes. California Academy of Universiteit van Amsterdam, Amsterdam.
Sciences, San Francisco. Wolf, J.H.D. & Konings, C.J.F. (2001) Toward the sustainable
Soto Arenas, M.A. (1988) Listado actualizado de las orquı́deas harvesting of epiphytic bromeliads: a pilot study from the
de México. Orquı´dea (Mex.), 11, 233–272. highlands of Chiapas, Mexico. Biological Conservation, 101,
Stevens, G.C. (1992) The Elevational Gradient in Altitudinal 23–31.
Range – an Extension of Rapoport Latitudinal Rule to Zotz, G. (1999) The epiphyte vegetation of Annona glabra on
Altitude. American Naturalist, 140, 893–911. Barro Colorado. Journal of Biogeography, 26, 761–776.
Stork, N.E. (1988) Insect diversity: facts, fiction and specula-
tion. Biological Journal of the Linnean Society, 35, 321–
337.
Sugden, A.M. & Robins, R.J. (1979) Aspects of the ecology
of vascular epiphytes in Colombian cloud forests. I. The BIOSKETCHES
distribution of the epiphytic flora. Biotropica, 11, 173–188.
Ter Steege, H. & Cornelissen, J.H.C. (1989) Distribution and
Jan Wolf is a former senior scientist at Ecosur. Currently
ecology of vascular epiphytes in lowland rain forest of
he is professor holding the Dr Jakoba Ruinen Chair,
Guyana. Biotropica, 21, 331–339.
Phyllosphere Sciences at the Institute for Biodiversity and
Utley, J.F. (1994) Bromeliaceae. Flora Mesoamericana (ed. by
Ecosystem Dynamics, Universiteit van Amsterdam. His
G. Davidse), pp. 89–156. Universidad Nacional Autonoma de
general interest in canopy biology centres on the con-
Mexico, Mexico DF.
servation and ecology of epiphytes and epiphyte com-
Walther, B.A. & Morand, S. (1998) Comparative performance
munities, particularly in tropical America.
of species richness estimation methods. Parasitology, 116,
395–405.
Alejandro Flamenco-S, left his academic position at
Webster, G.L. & Rhode, R.M. (2001) Plant diversity of an
Ecosur to enrol in the PhD programme at the Instituto de
Andean cloud forest – checklist of the vascular flora of
Ecologı́a, Universidad Nacional Autónoma de México
Maquipucuna, Ecuador. University of California Press,
(UNAM). He is interested in the application of geo-
Berkeley-Los Angeles-London.
graphic information systems for biodiversity conserva-
Went, F.W. (1940) Soziologie der Epiphyten eines tropischen
tion issues.
Urwaldes. Annales du Jardin Botanique de Buitenzorg, 50,
1–98.
Appendix 1 Inventories of epiphytic vascular plants, arranged by sampling unit. EQ is Epiphyte Quotient (Hosokawa, 1950).
Appendix 1 continued