There
have been many localized species-level plant radiations in South
America, the most famous of which is probably the Scalesia
radiation discovered by Charles Darwin in the Galapagos Islands of
Ecuador. That radiation consists of fifteen species of Scalesia,
spread out over about thirty thousand square kilometers, with a
maximum of six sympatric species per island (Valencia et al 2000). I
here report the discovery of a completely unsuspected species
radiation, far more dramatic than the Scalesia radiation in
terms of number and density of species, in the high mountains of the
Upper Pastaza Watershed in the eastern Andes of Ecuador. This
radiation consists of about 26 new closely related species of the
genus Teagueia (Orchidaceae), growing on four adjacent
mountains just 10-12 km apart, with up to 15 species sympatric on a
single mountain.
The
Upper Pastaza Watershed (here defined as the watershed of the Rio
Pastaza from 900 m upward) is the deepest and straightest valley in
the eastern Andes of Ecuador, and it serves as a major conduit for
airflow between Amazonia and the cold dry Interandean Valley. The
interaction of this airflow with the irregular topography creates
many unusual microclimates, and this has led to the evolution of many
endemic species of plants. The watershed is seldom appreciated as a
center of endemism, but in fact the number of vascular plant species
strictly endemic to the Upper Pastaza Watershed exceeds the number
endemic to the Galapagos Islands. About half of these endemic species
are orchids. Though the area has a long history of botanical
exploration, beginning with Richard Spruce in 1857, the higher
mountains of the watershed (above 3000 m) are nearly inaccessible and
are virtually unknown botanically.
Methods
This
work is part of a larger study to map the distributions of Teagueia
Luer, Lepanthes Sw., and certain other pleurothallid orchid
genera in the Upper Pastaza Watershed, concentrating on a long strip
8 km wide centered on the Rio Pastaza, beginning in Puyo and ending
above Baños. Sampling has been
ongoing for six years. Sampling is limited by the steepness of the
terrain, but all major ridgelines and most minor ones have now been
sampled to about 2400 m. Six of the seven highest ridge systems have
now been sampled to 3100m or beyond. Locations since 2001 have been
georeferenced using a GPS, and voucher collections for each species
of important genera (Lepanthes, Teagueia, Masdevallia,
Dracula, Trichosalpinx, Scaphosepalum,
Porroglossum) are stored in alcohol. At the end of this study
these collections will be distributed to QCNE and QCA. The new orchid
species discovered during the course of this project have been
described in collaboration with Dr C. A. Luer (Luer 1998, 1999, 2000,
2002 and in press) and Dr. Calaway Dodson (in press).
The
discovery of the Teagueia radiation in the Upper Pastaza
Watershed was a fortuitous byproduct of this mapping work.
Results
Twenty-six
new species of Teagueia Luer were discovered on four adjacent
mountains in the Upper Pastaza Watershed, more than quadrupling the
size of the genus. All are long-repent, unlike the caespitose or
short-repent species previously known, and all share several other
features (discussed below) that are unusual or absent in the
previously known species.
These
new long-repent species show considerable variation both within and
between species (Figure 1). They vary significantly in leaf size,
ranging from 5-8 mm in the smallest species to 25-35mm in the largest
species. The leaf texture varies greatly between species, and also
varies somewhat within species; some species have smooth leaves but
others have extremely pusticulate upper leaf surfaces. One species
has erose leaf margins. Inflorescences are very similarly shaped in
most species. The flowers vary greatly in size and shape, though the
basic structure is quite similar in all these long-repent species.
Colors range from white to yellow to red to black. Many of
these long-repent species are polymorphic, having a dark purple form
and a yellow form; these colour forms are often found intermixed, as
in several other pleurothallid genera (notably Brachionidium,
pers. obs).
Figure 1. Examples
of variation in Pleurothallidinae. Figure 1.
Ejemplos de variacion en los Pleurothallidinae.
The
new long-repent species are found on four adjacent mountains in the
Upper Pastaza Watershed: Cerros Mayordomo, Añangu,
Negro, and Candelaria (Figure 2). Cerro
Mayordomo, north of the Rio Pastaza , hosts seven of the long-repent
Teagueia species. Ten kilometers to the west, Cerro Añangu
hosts two of the long-repent species found on Cerro Mayordomo and
five additional long-repent species. Opposite Cerro Añangu,
on the south side of the Rio Pastaza, Cerro Negro hosts 9 long-repent
Teagueia species, none of which are found on the north side of
the Pastaza. Also on the south side of the Rio Pastaza and directly
across from Cerro Mayordomo, Cerro Candelaria hosts fifteen of these
new long-repent Teagueia species, including all of the Cerro
Negro species, and six additional species unique to Cerro Candelaria.
All of these sites are within 18 km of each other. In addition, one
population of long-repent Teagueia species has been found 90
km to the south of the Rio Pastaza.
Figure 2. Study
area. Figura
2. Area de estudio.
These long-repent
Teagueia species are found above 2800-3100 m on each of these
four mountains. Their upper elevation limit appears to be 3650 m;
above this elevation they were apparently absent on Cerro Candelaria.
Most species grow terrestrially in thick moss in deep shade in cloud
forests. Most species can be found less commonly growing
epiphytically within a meter or two of the ground. The epiphytic
habit is more common at higher elevations. Above 3400 m on Cerro
Candelaria there are patches of paramo, which often freeze at night,
and many of the Teagueia species of that mountain (including
several that grow in deep shade at 3100 m) grow in full sun in these
paramo patches, often rooting in sphagnum moss.
On all the
mountains where long-repent Teagueia species have been found,
they are among the most common higher plants, and by far the most
common orchids. Their lower altitudinal limit is usually extremely
well defined; below a certain point there are no Teagueia
plants whatsoever, and above that point they suddenly become
abundant. At elevations of 3100 m to 3400 m, densities on the order
of five individuals per square meter are common. In a circle of
diameter 5m one can normally find three to five species of Teagueia
growing together. Some hybridization appears to take place.
The first four long-repent species that were discovered
in the course of this work were published in Luer (2000).
Descriptions of the more recently discovered species are in
preparation.
Discussion
The
genus Teagueia Luer (Orchidaceae) was established to contain
six unusual species of the subtribe Pleurothallidinae (Luer 1991).
Some of these species had been originally placed in the genus
Platystele (Luer 1990), where they did not fit very well.
Three of these original Teagueia species are Ecuadorian, and
three are Colombian. All six species are caespitose or nearly so, and
all six have long-tailed sepals and disproportionately small petals.
The lip of all six species has an orifice formed by the folded edges
of the middle lobe of the lip, and also has a swelling or callus just
below the column tip. Most species bear inflorescences with all
flowers opening simultaneously.
The
26 new species of Teagueia discovered during the present study
all share certain characters not found in the six previously known
species. All are long-repent instead of caespitose or short-repent,
and none have a swelling or callus on the lip beneath the column.
Most have broad sepals without long tails, and all have successively
flowered inflorescences rather than the simultaneously flowered
inflorescences of most of the previously known species.
The sharing of
features that are unusual or absent in the six previously known
Teagueia species strongly suggests that all 26 new species
evolved from a recent common ancestor. Their extremely limited
geographical distribution also suggests a recent origin for this
clade. The limited distribution may be an artefact of the
inaccessibility of much of the eastern Andes, but it seems likely
that they are genuinely absent from the well-studied
Papallacta-Baeza-Tena corridor 110 km to the north of the Rio
Pastaza, and they are also probably genuinely absent from the
well-known Cuenca-Gualaceo-Limon corridor, 180 km to the south of the
Rio Pastaza.
One may speculate that the development of the
long-repent habit was the key innovation that opened up new niches
for this clade. In their extremely wet habitat, competition from
fast-growing mosses and liverworts is extreme, and the long-repent
habit may allow these species to keep their growing points above the
moss while maintaining contact with a humid substrate. It is likely
that this innovation was accompanied by a pollinator switch, as
indicated by the structural differences between the lips of the
long-repent species and those of the "normal" species.
The order of these developments is not clear; the pollinator switch
may have provided the reproductive isolation needed to permit the
fixation of the long-repent habit, or the long-repent habit may have
opened up new habitats with a new pollinator regime, leading to the
switch.
In any case, the
radiation of this clade does not seem to be driven by adaptations to
a variety of specific habitats. Many long-repent species typically
grow together in the same patch of moss, and most species seem
tolerant of a broad range of environmental factors. The radiation of
the clade is more likely driven by specialization onto different
pollinator niches. One species, T. cymbisepala, has tubular
orange flowers and may be hummingbird-pollinated. Small dipterans
have been seen visiting some of the other species, though actual
pollinization has not been observed. The variety of colours, sizes,
and shapes among sympatric species strongly suggests that they each
use a different set of pollinators. However, the presence of
occasional hybrids suggests that there is some slight overlap in
pollinator sets.
It is difficult to
explain certain aspects of the distributions of the long-repent
Teagueia species. Most unusual is the abrupt transition in the
cloud forest between the Teagueia zone and the non-Teagueia
zone. Teagueia species change from absent to abundant in the
space of a meter or two. This is not a by-product of simple colonial
growth, because the boundary line is formed by many different species
of Teagueia growing together. All individuals seem to respect
the same invisible line in the forest. This suggests that the
Teagueia species are all responding to some specific factor in
the soil. The obvious candidate for such a factor is a mycorrhizal
fungus, since it is known that orchid seed germination depends on the
presence of a suitable fungus. There is evidence that species of some
other pleurothallid genera, such as Lepanthes, require a
genus-specific symbiont fungus (pers. obs.). Perhaps the same is true
of Teagueia species, and the invisible line in the forest may
be the growth front of a very large mycorrhizal fungus.
Another difficult
aspect of Teagueia distribution is the striking difference
between the Teagueia floras of each of the four adjacent
mountains in the study area. Orchid seeds are the smallest of all
flowering plant seeds and are easily dispersed by wind. There can be
no doubt that seeds of all Teagueia species cross the valleys
that separate these four mountains. Most other orchid species freely
cross the Rio Pastaza, and most pleurothallid orchid distributions in
the study area form north-south bands related to precipitation (which
decreases from east to west). These bands almost always cross the
easterly-flowing Rio Pastaza. Yet the Rio Pastaza appears to be an
impenetrable barrier for Teagueia species. It is worth noting
that the long-repent Teagueia species are the only
pleurothallid orchids in the area that are primarily terrestrial.
Perhaps some biotic or abiotic soil factor differs from one side of
the Rio Pastaza to the other. We suspect that mycorrhizal fungi may
be that factor. Some terrestrial mycorrhizal fungi in the temperate
rain forests of northwest US are rodent-dispersed. If the mycorrhizal
fungus needed by the Teagueia species is rodent dispersed,
then the deep canyon of the Rio Pastaza could well form a barrier to
its dispersal, and there may be significant differences between the
terrestrial mycorrhizal fungi north and south of the Pastaza. It may
be that the Teagueia species adapted to the northern fungus
cannot germinate on the southern fungus, and vice versa. As a test of
this possibility, DNA studies of Teagueia mycorrhizal fungi
are planned.
Similar though much
smaller local radiations are known in Lepanthes (5 sympatric
closely related species on Cerro Abitagua, four of them endemic to
Cerro Abitagua; pers. obs.) and Trichosalpinx subgenus
Pseudolepanthes (six closely related species on Cerro del Torra in
Depto. de Choco, Colombia, four of them endemic to Cerro del Torra;
Luer 1997). The Teagueia radiation described here is
apparently the most diverse local radiation yet known among South
American orchids, and even so, we are probably underestimating its
size. Three of these 26 species are known only from single
individuals. These three species must have their principal
populations elsewhere, hidden in some as-yet-unexplored mountains in
the Upper Pastaza Watershed. It is likely that many more species will
turn up when these mountains are finally reached.
Acknowledgements
My students Andy Shepard and Pailin Wedel discovered the
Cerro Negro species discussed here. This work has been supported by a
series of generous grants from John and Ruth Moore to the Population
Biology Foundation, and by several grants from the San Diego County
Orchid Society, a grant from the Oregon Orchid Society, and grants
from the Andean Study Programs, Glenn and Marsha Staats, Alyssa
Roberts and Richard Bozek, and Cherise Miller and Kent Udell.
References
Luer, C. A. 1990.
Icones Pleurothallidinarum VII. Systematics of Platystele
(Orchidaceae). Monographs in Systematic Botany from the Missouri
Botanical Garden, 38:1-133.
Luer, C. A. 1991.
Icones Pleurothallidinarum VII. Systematics of Lepaanthopsis,
Octomeria Subgenus Pleurothallopsis, Restrepiella, Restrepiopsis,
Salpistele, and Teagueia (Orchidaceae).Monographs in
Systematic Botany from the Missouri Botanical Garden, 64:105-113.
Luer, C. A. 1997.
Icones Pleurothallidinarum XV. Systematics of Trichosalpinx
(Orchidaceae). Monographs in Systematic Botany from the Missouri
Botanical Garden, 64:105-113.
Luer, C. A. 1998.
Icones Pleurothallidinarum XVII. Systematics of Subgenus
Pleurothallis Section Abortivae, Section Truncatae, Section
Pleurothallis Subsection Acroniae, Subsection Pleurothallis, Subgenus
Dracontia, Subgenus Unciferia (Orchidaceae). Monographs in
Systematic Botany from the Missouri Botanical Garden, 72:103-120.
Luer, C. A. 1999.Icones Pleurothallidinarum XVIII. Systematics of Pleurothallis
Subgenus Pleurothallis Section Pleurothallis, Subsection
Antenniferae, Subsection Longiracemosae, Subsection
Macrophyllae-Racemosae, Subsection Perplexae,, Subgenus Pseudostelis,
Subgenus Acuminatia (Orchidaceae). Monographs in Systematic Botany
from the Missouri Botanical Garden, 76:139-147.
Luer, C. A. 2000.
Icones Pleurothallidinarum XX. Systematics of Jostia, Andinia,
Barbosella, Barbodria, and Pleurothallis Subgenus Antilla, Subgenus
Effusia, Subgenus Restrepioidia (Orchidaceae). Monographs in
Systematic Botany from the Missouri Botanical Garden, 79:1-4,
117-140.
Luer, C. A. 2002.
Icones Pleurothallidinarum XXIV. A First Century of New Species of
Stelis of Ecuador Part One. Monographs in Systematic Botany
from the Missouri Botanical Garden, 88:89-94.
Valencia,
R.; N. Pitman; S. Leon-Yanez & P. Jorgensen, eds. 2000. Libro
Rojo de las Plantas Endemicas del Ecuador 2000. Herbario QCA,
Pontificia Universidad Catolica del Ecuador, Quito.
