Distribution and ecology of myxomycetes in the high-elevation oak forests of
Cerro Bellavista, Costa Rica

Carlos Rojas1

Steven L. Stephenson
Department of Biological Sciences, University of
Arkansas, Fayetteville, Arkansas 72701

Abstract: Myxomycetes associated with a high-eleva-
tion (.3000 m) oak forest in the Talamanca Range of
Costa Rica were studied for 7 mo. Field collections
were supplemented with collections obtained from
moist chamber cultures prepared with samples of
bark and ground litter of Quercus costaricensis.
Various microenvironmental parameters including
pH, substrate moisture and diameter, height above
the ground and canopy openness were recorded for
each field collection, whereas macroenvironmental
data for temperature and precipitation were obtained
from a meteorological station near the study area.
Niche breadth and niche overlap indices were
calculated to assess possible resource partitioning by
myxomycetes. Thirty-seven species were recorded,
including 11 new records for Costa Rica, eight for
Central America and one for the neotropics. Both
PCA and NMS multivariate analyses indicated that pH
and height above the ground explained most of the
observed variation, although substrate diameter also
seemed to be an important factor. Precipitation
showed an inverse correlation with the number of
fruitings, confirming its importance as a macroenvir-
onmental factor. Niche overlap values were not
higher for closely related species and values for niche
breadths were similar for most of the more common
species, suggesting that most members of the assem-
blage of myxomycetes present in the study site are
ecological generalists.

Key words: biodiversity, community ecology,
eumycetozoans, slime molds, Talamanca Range

INTRODUCTION

The myxomycetes (plasmodial slime molds or myx-
ogastrids) are a relatively small and homogeneous
group of eukaryotic, phagotrophic organisms (Ste-
phenson and Stempen 1994) phylogenetically related
to the amoeboid protists (Adl et al 2005) but
traditionally studied by mycologists. Only about 880
species are known worldwide (Hernández-Crespo and

Lado 2005). The myxomycete life cycle consists of two
vegetative stages, one a uninucleate amoeba, with or
without flagella, the other a multinucleate unicellular
structure known as a plasmodium (Martin and
Alexopoulos 1969). Under suitable conditions the
plasmodium gives rise to the reproductive stage,
a somewhat fungus-like fruiting body.

Most ecological studies of myxomycetes have been
carried out in temperate regions of the world (e.g.
Härkönen 1977, Stephenson 1989, Novozhilov et al
1999, Schnittler 2001b), and only recently have
certain regions of the tropics been investigated.
However most regions of the tropics remain under-
studied and thus have considerable potential for
future studies (Schnittler and Stephenson 2000,
Stephenson et al 2004b). Results obtained from
recent studies suggest that the relative abundance of
myxomycete fruiting bodies decreases with decreasing
latitude (Schnittler and Stephenson 2002) and in-
creasing elevation, apparently in response (at least in
part) to increasing environmental moisture (Stephen-
son et al 2004b). Stephenson et al (2000) suggested
that distribution ranges in myxomycetes can be large,
although factors related to microclimate and vegeta-
tion influence their presence and dispersal potential.
For example some species display strong preferences
for specific substrates or exhibit distribution patterns
that can be related to microenvironmental variables
such as pH and moisture (Novozhilov et al 2005).
However distribution patterns of myxomycetes have
yet to be studied in high-elevation forests of the
tropics and basic questions relating to the assem-
blages of myxomycetes present in these forests and
whether the abundance of fruitings is correlated with
the actual number of taxa present at a particular
locality warrant further investigation.

In Costa Rica many forests above 3000 m are
dominated by a single species of oak (Quercus
costaricensis Liebm.). These high-elevation oak forests
are characterized by an almost constant cloud
coverage, and for this reason they often are referred
to as ‘‘cloud forests’’ (Kappelle et al 1992). Other
major groups of organisms such as animals, plants
and fungi have distinctive assemblages of species
associated with these forests, and the assemblages
usually differ from those found at intermediate
elevations or in lowland moist sites in the same
region (Kappelle 1996).

The lack of ecological data on myxomycetes in the
Accepted for publication 7 May 2007.
1 Corresponding author. E-mail: crojas@uark.edu

Mycologia, 99(4), 2007, pp. 534–543.
# 2007 by The Mycological Society of America, Lawrence, KS 66044-8897

534



neotropics, especially in high-elevation forests, sug-
gested the study described herein, in which an effort
was made first to characterize the assemblage of
species associated with a high-elevation oak forest in
Costa Rica and then to investigate some of the
ecological patterns displayed by these species. Partic-
ular emphasis was placed on assessing patterns of
sporulation phenology, seasonal abundance and
microhabitat occurrence in relation to macro- and
microenvironmental factors and the range of avail-
able microhabitats.

MATERIALS AND METHODS

The study reported herein was carried out in 2004 in a high-
elevation oak-dominated forest community in the moun-
tains of south central Costa Rica. The forest examined
would be classified as a montane moist forest according to
the Holdridge life zone system (Beauvais and Matagne
1998). The actual study site is on the eastern slope of Cerro
Bellavista and within the boundaries of the Cerro de la
Muerte Biological Station (9u339420N, 83u449270W) in the
province of San Jose, Costa Rica. Elevations within the
study area are 3142–3230 m above sea level. The forest
is characterized by a canopy that is 20–25 m tall and
dominated by Quercus costaricensis. In the subcanopy (5–
15 m tall) the most common trees are members of the
genera Weinmannia, Comarostaphylis, Schefflera, Drymis,
Myrsine and Oreopanax. The understory is dominated by
ferns and the bamboo-like grass Chusquea. The mean
annual temperature in this region of Costa Rica is 10.9 C,
and precipitation averages around 3000 mm per year (data
from the National Meteorological Institute, Costa Rica).

Field sampling was carried out during the two periods of
the year characterized by a different precipitation regime.
The first period (referred to as the ‘‘dry season’’) occurs
between December and the beginning of the rains in May,
whereas as the second period (the ‘‘rainy season’’)
encompasses the rest of the year. However it should be
noted that recognizing these two periods as distinct is
somewhat arbitrary because the annual climate exhibits
a rather superficial seasonality and forests are mostly
evergreen throughout the year. Three collecting trips were
made in each of the two seasons, and these extended over
a total of 7 mo. The laboratory component of the study was
carried out during an additional 8 mo. Samples for
laboratory study were collected only in the first trip of each
season, but field collections were obtained on every trip.

Plots and collections.—Three 20 3 50 m (0.1 ha) plots were
established in portions of the study area clearly dominated
by oak trees. Although no analysis of tree dominance was
carried out, Jimenez et al (1988) reported a density of about
500 stems per hectare (DBH .10) for the oak forests in this
portion of the Talamanca Range. To minimize any marginal
effects the boundary for each plot was located at least 50 m
from the forest edge. The opportunistic sampling protocol
(Cannon and Sutton 2004) was used to search for fruitings
of myxomycetes in each plot. This method is effective for

studying myxomycetes, especially when the forest structure
is particularly complex as is the case in many areas of the
tropics because it lets the researcher make an a priori
selection of substrates. For this reason primary emphasis was
placed on examining dead leaves, decaying wood and twigs
for myxomycetes. No specimens still in the plasmodial stage
were collected. Nomenclature follows Hernández-Crespo
and Lado (2005) except for Arcyria leiocarpa and Stemonitis
smithii, where the treatment of Martin and Alexopoulos
(1969) is used.

All specimens obtained in the field were collected and
curated in the manner described by Stephenson and
Stempen (1994). To complement field collections, 120
moist chamber cultures were prepared with samples of bark
and ground litter collected from each plot. Sixty moist
chambers were prepared with samples collected during the
dry season and another 60 with samples collected during
the rainy season. In this part of the study samples were
collected in paper bags, transported to the laboratory and
placed in standard 9 mm diam Petri dishes lined with filter
paper. Distilled water was added to each dish and the
sample material in each culture remained soaked for 24 h,
after which excess water was poured off. Cultures were
examined every week for approximately 4 mo.

Environmental measurements.—General climatic data for
the total period during which the study was carried out were
obtained directly from the Villa Mills meteorological
station, 5 km south of the study site, through the National
Meteorological Institute (IMN) in San Jose, Costa Rica.
Microenvironmental variables were measured or deter-
mined directly in the field. For example the type of
substrate on which a fruiting occurred along with its
diameter (for woody substrates) and height above ground
were recorded for each specimen collected.

Canopy openness, measured with a spherical densi-
ometer, was used as an indicator of the quantity of light
reaching the forest floor. To determine this parameter,
each plot was divided in 10 subplots of 10 3 10 m, and these
were subdivided further into four 5 3 5 m areas. In each of
the latter, four measurements of canopy openness were
obtained and the mean value was calculated. Each specimen
collected in the plot was assigned the average value of the
specific area in which it occurred.

The stage of wood decay was recorded for discrete
categories as described by Stephenson et al (2004a), except
that in the present study categories 2 and 4 were not used.
As such only three categories (1, 3 and 5) were recorded for
wood decay and these were considered to represent early,
intermediate and late decay stages. Substrate moisture was
measured by collecting small samples of substrate from the
same microsites on which fruitings occurred. Within 24 h of
returning from the field these samples were weighed and
placed in a constant temperature oven for 72 h at 65 C.
Afterward it was assumed that most of the water content of
the substrates had been lost and samples were reweighed.
The difference between dry and original weights was used to
obtain the percentage moisture for each sample.

Data analysis.—Shannon-Wiener and taxonomic diversity
indices (Stephenson et al 1993) were calculated for the total

ROJAS AND STEPHENSON: MYXOMYCETES OF HIGH-ELEVATION OAK FORESTS 535



assemblage of species in an effort to quantify overall
myxomycete biodiversity in the forest studied. Sørensen’s
coefficient of community was calculated for the sets of
specimens obtained on different collecting dates to evaluate
temporal differences in species composition.

A multivariate analysis of microenvironmental measure-
ments was carried out to evaluate the possible effects of
micro- and macroenvironmental variables. To avoid the
‘‘noise’’ produced by the less common species, all species
were classified according to their abundance before
performing the analysis. In this classification, species
representing more than 3% of the total number of
collections were considered as abundant, those falling
between 1.5–3% as common, between 0.5 and up to 1.5%

as occasional and those less than 0.5% as rare (Stephenson
et al 1993). A nonmetric multidimensional scaling ordina-
tion (NMS) was performed with the numeric microenviron-
mental variables for only the species classified as abundant.
This ordination was carried out with the program PC-ORD
by exploring 50 runs of real data and 50 runs of randomized
data using the autopilot function and the scores generated
by weighted averaging. Sørensen distances and a Monte
Carlo test of significance were used. A principal component
analysis (PCA) based on correlations also was performed
with the same set of data to evaluate the relative importance
of the different variables to structure and composition of
the community and to evaluate similarities with the previous
ordination.

Values obtained for niche breadth and niche overlap
were calculated in the manner described by Stephenson
(1988). These were evaluated for the most abundant species
with Levin’s estimators, as recommended by Maurer (1982)
and Petraitis (1985). In this case substrate moisture and
diameter, pH, height above the ground and canopy
openness (and thus the level of light) were used as potential
indicators of resource partitioning. This type of analysis has
the potential to examine the evidence for potential
interactions that might occur within particular taxonomic
groups. Such interactions might not be revealed on a spatial
analysis, especially if they occur between closely related
species.

RESULTS

Thirty-seven species were collected (TABLE I), with 27
of these recorded during the dry season and 20
during the rainy season. Most species were represent-
ed by fruitings that occurred in the field under
natural conditions, and only four taxa were recovered
from moist chamber cultures. Eleven of the 37 species
are new records for Costa Rica, nine are new records
for the Central America and one (Diacheopsis sp.) is
a new record for the neotropics. The most abundant
species were Cribraria piriformis, Ceratiomyxa fruticu-
losa, Cribraria mirabilis and Cribraria vulgaris. Most
fruitings were recorded on logs, twigs and bryophytes,
although a few occurred on ground litter. For the
entire study the value calculated for Shannon’s index

of diversity was 3.27, whereas the values for the dry
season and the rainy season were respectively 1.09 and
1.30; the index of taxonomic diversity for the total
assemblage of species present was 2.11. Eighty-two
percent of all field collections consisted of stalked
fruiting bodies, and 75% of the species recorded were
examples that typically produce stalked fruiting
bodies.

The PCA analysis indicated that 60% of the total
variation was explained by pH and height above the
ground (not shown). However, according to the NMS
analysis, height above the ground, substrate diameter
and pH were the most important microenvironmental
variables accounting for the variation in the data
when a cutoff value of 0.01 is applied (FIG. 1). Of
note, pH values above 4.5 are absent when height
above the ground reaches 1 m (pH vs. height above
ground, Pearson’s product moment 5 20.43, P ,

0.0001) and values this high were rarely recorded for
substrates with a diameter greater than 20 cm (pH vs.
height above ground, Pearson’s product moment 5

20.36, P , 0.0001). Those relationships are ex-
plained by the fact that substrates with diameters
.25 cm are rare near the ground (substrate diameter
vs. height above ground, Pearson’s product moment
5 0.45, P , 0.0001). In fact a deeper analysis
indicated that Didymium squamulosum, Lycogala
epidendrum and Metatrichia floriformis seem to group
together both for high pH and lower heights. These
species also differ from other taxa in their relatively
high values for substrate pH (F 5 8.00, df 5 11, P ,

0.0001). Conversely examples of species showing
a preference for more acidic and higher substrates
include Cribraria mirabilis and Trichia botrytis.

Substrate moisture, pH and canopy openness
showed significant differences when a comparison
between the two seasons was made. Of interest, height
above the ground also showed a significant difference
from the dry to the rainy season. Temporal differ-
ences also appear to be apparent in the type of
substrate (x2 5 31.7, df 5 8, P , 0.0001), with logs
and twigs more commonly recorded during the dry
season and logs and bryophytes during the rainy
period.

Of interest, the only macroenvironmental factor
that appears to have an important influence on the
occurrence of myxomycetes is precipitation, which
had an inverse but significant correlation with the
number of fruitings recorded (FIG. 2, precipitation vs.
number of fruitings, Pearson’s product moment 5

20.77, p , 0.05). However a multiple regression
analysis showed that the combined effect of temper-
ature and precipitation seems to have an even greater
influence (precipitation and temperature vs. number
of fruitings, Pearson’s product moment 5 0.95, P ,

536 MYCOLOGIA



0.05). A significant relationship was not observed
when the number of species was considered (pre-
cipitation and temperature vs. number of species,
Pearson’s product moment 5 0.87, P , 0.05).

The coefficient of similarity value calculated for
species assemblages associated with the two seasons
was 0.48. Major differences in abundance were noted
for members of such genera as Lamproderma and
Lycogala, which invariably were absent during the

rainy period, or Didymium and Leocarpus, never
recorded for the dry season. When the values
calculated for niche breadth and overlap are exam-
ined it is clear that the values for most species are
similar and relatively high (TABLE II). If niche
breadth values represent the mathematical result of
a proportionally higher number of fruitings recorded
in the field, then some correlation between these two
variables should exist for the community as a whole.

TABLE I. Myxomycetes recorded at Cerro Bellavista and mean values for the most important associated microenvironmental
variables. Abbreviations are given only for the most abundant species. Note: NF/MC 5 Total number of field collections (NF)
and moist chamber records (MC) for each species, SD 5 Standard deviation, HAG 5 Height above ground in centimeters,
SubD 5 Substrate diameter in centimeters, O 5 Occasional, R 5 Rare, C 5 Common, A 5 Abundant and NA 5 Not available
(i.e. when a particular species was recorded only in moist chamber culture or when the number of specimens for that
particular species was only one)

Species Abundance NF/MC pH (SD) HAG SubD Abbreviation

Arcyria cinerea (Bull.) Pers. O 2/0 4.25 (0.20) 72.50 23.50
Arcyria denudata (L.) Wettst. R 1/0 4.30 (NA) 5.00 6.00
Arcyria leiocarpa (Cooke) Martin & Alexop. O 3/0 4.67 (0.35) 108.33 26.33
Ceratiomyxa fruticulosa (Mull.) Macbride A 19/0 4.67 (0.76) 34.47 6.89 CERfru
Clastoderma debaryanum Blytt C 5/0 4.72 (0.37) 10.00 18.40
Comatricha pulchella (Bab.) Rostafa C 4/0 4.23 (0.91) 5.00 6.75
Comatricha tenerrima (Curtis) G. Lister O 2/0 4.90 (0.84) 5.00 3.00
Cribraria intricata Schrad. C 6/0 4.30 (0.42) 57.50 25.67
Cribraria mirabilis (Rostaf.) Masseeb A 15/0 4.58 (0.58) 50.00 16.60 CRImir
Cribraria piriformis Schrad. A 20/0 3.90 (0.79) 31.20 20.40 CRIpyr
Cribraria vulgaris Schrad. A 14/0 3.83 (0.70) 68.57 18.64 CRIvul
Diacheopsis sp.c R 1/0 4.40 (NA) 150.00 20.00
Diderma chondrioderma (de Bary & Rostaf.) G. Lister R 1/0 6.90 (NA) 5.00 7.00
Diderma sp. O 2/0 5.75 (1.09) 5.00 4.00
Didymium dubium Rostaf.b R 0/1 NA NA NA
Didymium squamulosum (Alb. & Schwein.) Fr. A 9/0 5.80 (1.08) 5.22 1.67 DDYsqu
Hemitrichia calyculata (Speg.) Farr A 9/0 4.97 (0.58) 20.00 10.33 HEMcal
Hemitrichia serpula (Scop.) Rostaf. ex Lister A 0/7 NA NA NA HEMser
Lamproderma columbinum (Pers.) Rostaf.b A 12/0 4.21 (0.58) 27.50 20.50 LAMcol
Lamproderma cribrarioides (Fr.) R.E.Fr.b O 3/0 4.70 (0.69) 6.67 18.00
Lamproderma echinulatum (Berk.) Rostaf.b C 5/0 4.88 (0.79) 25.00 29.00
Lamproderma sauteri Rostaf.b R 1/0 4.60 (NA) 5.00 5.00
Leocarpus fragilis (Dicks.) Rostaf.b O 2/0 6.15 (0.63) 30.00 9.50
Lycogala epidendrum (L.) Fr. A 10/0 5.52 (0.87) 16.40 7.95 LYCepi
Metatrichia floriformis (Schwein.) Nann.-Bremek. A 13/0 5.52 (0.72) 7.62 14.00 METflo
Perichaena depressa Lib. R 0/1 NA NA NA
Physarum brunneolum (Phillips) Masseeb R 1/0 4.70 (NA) 100.00 10.00
Physarum contextum (Pers.) Pers.a R 1/0 5.80 (NA) 5.00 10.00
Physarum leucopus Link R 1/0 6.50 (NA) 80.00 4.00
Physarum melleum (Berk. & Broome) Massee R 0/1 NA NA NA
Stemonitis fusca Roth O 2/0 5.35 (0.07) 32.50 45.00
Stemonitis smithii Macbride O 4/0 5.28 (0.18) 11.25 10.25
Stemonitopsis hyperopta (Meyl.) Nann.-Bremek. O 3/1 4.10 (0.36) 48.33 6.67
Trichia botrytis (Gmel.) Pers. A 10/0 4.30 (0.51) 74.50 7.80 TRIbot
Trichia decipiens (Pers.) Macbride O 3/0 4.77 (0.32) 30.00 10.00
Trichia favoginea (Batsch.) Pers. A 7/1 4.41 (1.09) 32.14 22.00 TRIfav
Trichia verrucosa Berk. A 8/0 4.95 (1.32) 34.38 11.31 TRIver

a New record for Costa Rica.
b New record for Central America.
c New record for the neotropics.

ROJAS AND STEPHENSON: MYXOMYCETES OF HIGH-ELEVATION OAK FORESTS 537



However the actual correlation is weak (niche breadth
vs. number of fruitings, Pearson’s product moment 5

0.09, P . 0.05). Posterior analysis indicated that
Lamproderma columbinum has a broader niche and
Dydimium squamulosum a narrower niche than the
other species (F 5 14.5, df 5 7, P , 0.05). When
these two species are excluded from the analysis, no
appreciable differences in niche breadth values can
be noted among the rest of the taxa present (F 5

5.07, df 5 5, P . 0.05).
One might expect intrageneric overlap to be higher

than intergeneric overlap due to the theoretical
relatedness in resource use by closely related species.
However an analysis of the niche overlap values
calculated for Cribraria and Trichia, the two most
diverse genera, does not reveal significant differences
in the values between species within each genus or
among the species of the two genera (F 5 1.20, df 5

2, P . 0.05).

DISCUSSION

The number of species recorded in the present study
is comparable to the totals reported for studies
carried out elsewhere in the tropics (e.g. Schnittler
and Stephenson 2000, Stephenson et al 2004a).
However limiting the study to selected substrates
might have underestimated the actual number of taxa
present. Schnittler (2001a) and Schnittler and Ste-
phenson (2002) have described new microhabitats for
myxomycetes in the neotropics, and unpubl data
suggest that a number of others probably exist.
Preliminary surveys of the study area in 2001 and
2003 indicated that at least four plant species could
be serving as living microhabitats for some species of
myxomycetes. However these plants were not consid-
ered in the present study and some of the species of

myxomycetes typically associated with the microhab-
itats they provide were not recorded in the present
study either. In any case it is clear that surveys of
nontraditional substrates would yield new records for
any particular locality, especially in the neotropics as
discussed by Stephenson et al (2004b).

In a similar study carried out in the seasonal
tropical forests of Guanacaste in Costa Rica, Schnittler
and Stephenson (2000) reported a Shannon’s index
of 1.11. Such a low value can be obtained for a single
locality in temperate regions (e.g. Stephenson 1989),
but it seems more typical for the apparently less
diverse tropical regions of the world (Stephenson et
al 2004b). As such the value obtained in the present
study (3.27) seems more comparable to those
obtained for temperate rather than tropical regions,
and the values obtained for the different periods of
study suggest that diversity is similar throughout the
year. The taxonomic diversity index (2.11) obtained
in the present study contrasts dramatically with values
reported for other areas in the tropics. For example
Stephenson et al (1993) obtained a value of 3.93 for
the set of data contained in a published checklist of
the myxomycetes of Costa Rica (Alexopoulos and
Saenz 1976) and reported a value of 4.13 for southern
India.

When these values are considered together the
apparent conclusion is that tropical areas appear to
be characterized by lower overall numbers of species
but richer intrageneric diversity. However the set of
data from Cerro Bellavista shows a different pattern.
The assemblage of species contains a relatively higher
number of species, and this number is especially
impressive when one considers that for tropical
regions myxomycete diversity seems to decrease as
elevation increases (Stephenson et al 2004b). Of note,
the assemblage at Cerro Bellavista is dramatically poor
in terms of the number of species per genus, which

FIG. 1. Ordination of species and microenvironmental
variables with nonmetric multidimensional scaling (NMS).
Points define the abundant species. Lines indicate the
direction and strength of the most important microenvi-
ronmental variables. (For abbreviations used for species see
TABLE I.)

FIG. 2. Numbers of species and fruitings recorded
during 7 mo of field work in the current study. The line
indicates the average value of monthly precipitation for
Cerro Bellavista, based on data provided by the National
Weather Institute in Costa Rica.

538 MYCOLOGIA



resembles the apparent pattern for temperate areas.
This suggests that resource partitioning among
species should be lower than in typical tropical
assemblages, where requirements of particular species
might be expected to be more similar due to higher
numbers of closely related taxa.

Ceratiomyxa fruticulosa was the only species found
to be abundant both in the present study and
a previous study of cloud forests in Ecuador (Ste-
phenson et al 2004a). However Didymium squamulo-
sum, Lycogala epidendrum and Metatrichia floriformis,
reported as common in Ecuador, are three of the
species that form a cluster that seems to be related to
a basic pH in the ordination presented (FIG. 1).

When pH values are compared it is clear that most
substrates at Cerro Bellavista are only slightly more
acidic (values of 2.7–7.1) than those in Ecuador (3.3–
9.8). However none of the species associated with low
pH values in the present study, including the three
common species of Cribraria, were reported from
Ecuador (Stephenson et al 2004a). Moreover in their
study of a cloud forest in the northern part of Costa
Rica, Schnittler and Stephenson (2000) did not
report the same species recorded in the present
study. The forests investigated in these three studies
were different in terms of plant composition and
architecture. Consequently it seems obvious that the
differences in climatic conditions and plant compo-
sition that exist among these three areas are playing
an important role in determining the species compo-
sition of the assemblages of myxomycetes present.

The only common species shared between the
present study and cloud forests in Ecuador (Schnittler
and Stephenson 2000) occurred only on leaf litter,
which suggests that lignicolous substrates might have
a more important influence on the distribution of

myxomycetes, as has been suggested for plant
communities in other parts of the world (e.g.
Stephenson 1988, Schnittler 2001b). Of note, oak is
absent both in Ecuador (Ulloa and Møller 1993) and
in the cloud forest studied by Schnittler and
Stephenson (2000) in Costa Rica.

An interesting result of the present study is the high
proportion of stalked species (75%) and records of
those species (82%). The values we obtained are
similar to those reported for temperate deciduous
(but mostly oak) forests in the Mountain Lake region
of southwestern Virginia (Stephenson 1988), where
the same percentage (74%) was recorded for both
species that typically produce stalked fruiting bodies
and the total number of records represented by these
species. This would suggest that whatever ecological
factors (presumably those related to levels of substrate
and/or atmospheric moisture) are involved in de-
termining the relative proportions of sessile vs.
stalked forms in the assemblage of myxomycetes
present in a particular type of habitat are fairly
comparable for high-elevation oak forests in Costa
Rica and mid-latitude oak forests in eastern North
America.

Records of new taxa.—Studies of myxomycetes in
Costa Rica (e.g. Schnittler and Stephenson 2000)
have increased the number of species known from the
country to 126. However these studies did not
consider high-elevation communities. Although there
are few records from an elevation of approximately
2700 m near the El Empalme area (Alexopoulos and
Saenz 1976), the highest area studied in detail thus
far is the Monteverde Biological Reserve at around
1500 m (Schnittler and Stephenson 2000). However
both the structure of the forest and climatic condi-
tions in these two areas are very different from those

TABLE II. Niche breadth and niche overlap values for the 12 most abundant species recorded at Cerro Bellavista. Just one
value per species pair is shown. (For abbreviations refer to TABLE I.)

Species
Niche

breadth Species and Niche overlap values

CERfru CRImir CRIpir CRIvul DDYsqu HEMcal LAMcol LYCepi METflo TRIbot TRIfav TRIver
CERfru 2.54 *** 0.96 0.95 0.87 0.80 0.97 0.84 0.95 0.82 0.86 0.94 0.97
CRImir 2.98 *** 0.97 0.96 0.63 0.93 0.93 0.87 0.74 0.92 0.97 0.99
CRIpir 2.95 *** 0.90 0.70 0.97 0.92 0.92 0.86 0.82 0.99 0.97
CRIvul 3.03 *** 0.47 0.82 0.93 0.73 0.57 0.96 0.90 0.93
DDYsqu 2.00 *** 0.87 0.62 0.92 0.91 0.59 0.75 0.79
HEMcal 2.60 *** 0.85 0.98 0.92 0.79 0.96 0.96
LAMcol 3.86 *** 0.76 0.70 0.80 0.94 0.91
LYCepi 2.25 *** 0.95 0.74 0.92 0.92
METflo 2.16 *** 0.62 0.88 0.85
TRIbot 2.63 *** 0.82 0.89
TRIfav 3.00 *** 0.96
TRIver 3.04 ***

ROJAS AND STEPHENSON: MYXOMYCETES OF HIGH-ELEVATION OAK FORESTS 539



at Cerro Bellavista. Therefore the new records
generated in the present study are probably not
unexpected.

Four species of the genus Lamproderma were
recorded in the present study, and none of these
had been reported previously for the Central Amer-
ican region. Also three of the four most abundant
species recorded in Cerro Bellavista are members of
the order Liceales, which was the single most
abundant order. Stephenson and Stempen (1994)
indicated that both Lamproderma and Cribraria,
a member of the Liceales, are characteristic genera
of temperate forests. Cribraria mirabilis, one of the
new records, is well known in temperate areas,
especially in Europe (Lado and Pando 1997) but
appears to be rare in the tropics. Of interest, this
species was one of the most abundant myxomycetes in
the present study.

Leocarpus fragilis, another species typical of tem-
perate regions, also was recorded at Cerro Bellavista.
This species was collected previously (Martin Schnit-
tler, unpubl data) in the paramo of the Cerro
Chirripó (ca. 3700 m). Both mountains are in the
Talamanca region and represent two of the highest
peaks in Costa Rica, which suggests that L. fragilis
might be restricted to high elevations of the country.
This is apparently the case in Colombia (Uribe-
Meléndez 1995).

As a general observation it seems that the assem-
blage of myxomycetes at Cerro Bellavista more closely
resembles, both taxonomically and ecologically, the
assemblages associated with temperate forests rather
than those of tropical forests.

Microenvironmental factors.—The PCA analysis indi-
cates that pH and height above the ground account
for much of the variation associated with the more
common species. Other studies (e.g. Härkönen 1977,
Schnittler et al 2006) have found that pH is an
important ecological factor for myxomycetes and
height above the ground also seems to have some
influence on the distribution of the organisms,
especially on a macro scale when forest canopies are
studied (e.g. Black et al 2004). The results from the
PCA analysis are not surprising when the forest
dynamics of Cerro Bellavista, where there is a slight
seasonality affecting the phenology of oak trees, are
considered (Kappelle 1996). The variation that occurs
in the canopy coverage represents an important factor
because it largely determines the effective vertical
precipitation reaching the forest floor and the extent
to which leaching of nutrients from the canopy takes
place (e.g. Milla et al 2005), thus influencing the
substrate humidity and pH values in the lower strata
of the forest.

In the present study compositional differences in
the species assemblages recorded for the different
seasons might be expected to make this pattern even
more evident, especially when species recorded only
in one of the two seasons are considered. Lycogala
epidendrum and Didymium squamulosum for example
were recorded during only one season. Similarly it
seems that substrate moisture does not affect the
assemblage of species in Cerro Bellavista as has been
observed in other studies (e.g. Schnittler et al 2006).
However the major effect of horizontal precipitation
in the study area is that maintains relatively high water
content for most substrates throughout the entire
year.

Substrate diameter is another important variable
(FIG. 1). This is not surprising either, especially when
it is obvious that most of the twigs and lower diameter
logs are associated with the ground level. For example
Schnittler et al (2006) found these factors did not
explain the variation in their data when they studied
a community of canopy myxomycetes in Germany.
However the assemblages of species present in the
canopy and aerial strata of tropical forests are
different from those found at ground level (Black et
al 2004); therefore it would not be surprising if their
community ecology is different.

Cribraria vulgaris, C. mirabilis and Trichia botrytis
were recorded at the highest positions, whereas three
other species (Trichia favoginea, C. piriformis and
Lamproderma columbinum) were recorded for sub-
strates with the largest diameters. Of note, most of
those species were recorded in both seasons, which
suggests that these variables are not as macroenvi-
ronmentally dependent as is the case for pH.
Consequently the composition of the species assem-
blage observed in a particular season is a combination
of preferences for both dynamic and more static
microenvironmental conditions, depending on the
responses of particular taxa.

Associated with the microenvironmental variation
that exists between the seasons, there was a significant
difference in the height above the ground at which
fruitings were recorded, with higher values being
recorded during the rainy season. This pattern does
not seem to have been reported for myxomycetes in
previous studies.

In general the myxomycetes seemed to fruit
preferentially on logs and twigs on the ground.
However bryophytes were very important during the
rainy season. Of interest, although bryophytes have
been reported as apparently favorable substrates for
myxomycetes (Stephenson and Studlar 1985), they do
not seem to be more abundant above the ground in
the oak forests of Costa Rica (Holz et al 2002).
Consequently the apparent change in height above

540 MYCOLOGIA



the ground of the myxomycetes at Cerro Bellavista for
the two different seasons does not necessarily seem to
be the result of species that are able to grow on
bryophytes, as the data appear to show.

Macroenvironmental factors.—Data (FIG. 2) seem to
indicate that the sporulation patterns of myxomycetes
in the study area are linked intimately with pre-
cipitation. However, for this particular forest, it is
difficult to evaluate the effect of seasonal climate
variables on the community due to its tropical
evergreen character. Although patterns of seasonality
are known for some species of myxomycetes in
temperate regions, there is a lack of mid- and long-
term studies in the tropics (Stephenson et al 2004b).

In one of the few other studies of seasonal
differences in myxomycetes, Maimoni-Rodella and
Gottsberger (1980) examined the sporulation pattern
of the species present in a lowland tropical rainforest
in Brazil. Although these authors reported an overall
pattern similar to that of the present study, they also
considered temperature to be the more important
factor in the tropics because water often is not
a limiting factor. Of note, Ogata et al (1996) found
a positive correlation between both precipitation and
temperature and the overall abundance of fruitings in
a study carried out in eastern central Mexico. These
authors agreed with Maimoni-Rodella and Gottsber-
ger (1980) when suggesting that slight changes in
temperature during the period studied might be
responsible for changes observed in the community
composition. However both of these studies were
carried out in ecological situations totally different
from the present study, which is also reflected in the
very different species composition, as already dis-
cussed.

For the general area considered in the present
study Kappelle (1996) found that temperature is the
most important variable explaining the dynamics of
the overall ecosystem. It is interesting to note that our
data suggest that temperature has a positive additive
effect on the model when both factors are examined
together. More than one environmental factor appar-
ently affects the timing of sporulation in myxomy-
cetes. However the main effect of precipitation in the
present study contrasts dramatically with the negligi-
ble effect of substrate moisture in the microenviron-
mental analysis. This seems to indicate that, as has
been observed in other studies, myxomycetes have an
optimal range of substrate moisture conditions, which
is hypothesized to be constantly maintained by the
cloud coverage in the study area but exceeded when
rain occurs and temperature drops.

Climatic data for Cerro Bellavista indicate that
temperature varies only about 1.7 C from February to

October, whereas monthly precipitation increases
dramatically from around 21 mm to more than
450 mm over the same period. However the effect
of horizontal precipitation at high elevation in the
tropics is known to be correlated with temperature
because the latter affects the water content in the
atmosphere. For example Stephenson and Stempen
(1994) suggested that atmospheric humidity could be
the most important macroenvironmental variable
determining the timing of sporulation for myxomy-
cetes. This variable did not seem as important in the
present study; however it is known that high
environmental humidity is associated with high
horizontal precipitation and high precipitation recy-
cling (Dominguez et al 2006). Results obtained in the
present study seem to indicate that myxomycete
fruitings increase in number as precipitation drops
and decrease with increasing precipitation. However
similar studies carried out in lowland tropical forests
in other parts of the neotropics might reveal the
relative effect of each factor on the sporulation
pattern and abundance of myxomycetes.

Niche breadth and overlap.—The values calculated for
niche breadth are mathematical constructs that are
determined by the original values of resource
utilization used as input for the equations to generate
a value for a particular species. One assumes a normal
distribution in the resource utilization requirements;
consequently there is an increasing probability of
finding increasingly more narrow niches as more
input data are used. This is due to the fact that values
closer to the mean are more common than extreme
values. For this reason is important to test whether the
niche breadth values are the product of a probabilistic
artifact or external factors. The nonsignificant and
weak correlation between the number of fruitings
found in the field and values for niche breadth
suggests that these two variables are independent,
which makes niche breadth useful for making
inferences about biotic interactions.

The observed niche breadths include examples that
show significant differences for some species when
Lamproderma columbinum and Didymium squamulo-
sum are included in the analyses. Of note, these two
species are found at the extremes of the niche
breadth distribution for the study area, at least
suggesting that they are characterized by ecological
strategies different from those of the other taxa
present. Didymium squamulosum was the only abun-
dant species in which the fruiting bodies contained
calcium carbonate, which may influence its ecological
distribution. Lamproderma columbinum seems to
exhibit a strong substrate preference for bryophytes,
even though its niche breadth is the broadest overall.

ROJAS AND STEPHENSON: MYXOMYCETES OF HIGH-ELEVATION OAK FORESTS 541



When analyzing niche overlap it has been noted
that closely related species, for example those
belonging to the same genus, have more similar
resource requirements than less related species,
presumably as a result of their common evolutionary
path (Morin 1999). However, when the intrageneric
niche overlap values are evaluated with the two most
diverse genera in this study, no differences were
found. In fact the values obtained for species within
each genus differ only slightly from the values
obtained for species across genera. The implication
would seem to be that they reflect a common
ecological strategy. In fact it appears that the
assemblage of myxomycetes consists of a majority of
generalist species and few that are specialists.

In summary the results obtained in the present
study provide new and relevant data on the ecological
patterns displayed by myxomycetes in the neotropics.
The new records for the region also contribute to our
knowledge of the biogeographical patterns of myx-
omycetes. When the influence of both macro- and
microenvironmental variables on the sporulation of
myxomycetes at Cerro Bellavista was evaluated, in-
formation generated in the study indicates that
a combination of factors determines the timing of
this phenomenon, whereas analyses of niche breadth
suggest that the species of myxomycetes present are
mostly ecological generalists that are well adapted to
changing microenvironmental conditions.

ACKNOWLEDGMENTS

Appreciation is extended to Maria Julia Vargas and Randall
Valverde for their valuable support in the field, to Julieta
Carranza, Jose Francisco DiStéfano and Walter Marı́n at the
University of Costa Rica for their help in the research
process, and to Federico Valverde and the staff of the Cerro
de la Muerte Biological Station for their help with logistics.
Special thanks are extended to the Amistad Pacı́fico
Conservation Area authorities and to David Mitchell for
his comments on the some of the specimens of myxomy-
cetes collected during the study. The research reported
herein was supported in part by a grant (DEB-0316284)
from the National Science Foundation.

LITERATURE CITED

Adl S, Simpson A, Famer M, Andersen R, Anderson R, Barta
J, Bowser S, Brugerolle G, Fensome R, Fredericq S,
James T, Karpov S, Kugrens P, Krug J, Lane C, Lewis L,
Lodge J, Lynn D, Mann D, McCourt R, Mendoza L,
Moetru Ø, Mozley-Standridge S, Nerad T, Shearer C,
Smirnov A, Spiegel F, Taylor M. 2005. The new higher
level classification of eukaryotes with emphasis on the
taxonomy of protists. J Eukaryot Microbiol 52(5):399–
451.

Alexopoulos CJ, Sáenz JA. 1975. The myxomycetes of Costa
Rica. Mycotaxon 2(2):223–271.

Beauvais J, Matagne P. 1998. The Holdridge’s concept of
‘‘life zone’’: a tropical point of view in ecology.
Ecologie 29(4):557–564.

Black D, Stephenson S, Pearce C. 2004. Myxomycetes
associated with the aerial litter microhabitat in tropical
forests of northern Queensland, Australia. Syst Geog Pl
74:129–132.

Cannon P, Sutton B. 2004. Microfungi on wood and plant
debris. In: Mueller G, Bills G, Foster M, eds. Bio-
diversity of fungi: inventory and monitoring methods.
Burlington, Massachusetts: Elsevier Academic Press.
p 217–239.

Dominguez F, Kumar P, Xin-Zhong L, Mingfang T. 2006.
Impact of atmospheric moisture storage on precipita-
tion recycling. J Climate 19(8):1513–1530.

Härkönen M. 1977. Corticolous myxomycetes in three
different habitats in southern Finland. Karstenia 17:
19–32.

Hernandez-Crespo JC, Lado C. 2005. An on-line nomencla-
tural information system of Eumycetozoa. http://
www.nomen.eumycetozoa.com

Holz I, Gradstein S, Heinrichs J, Kappelle M. 2002.
Bryophyte diversity, microhabitat differentiation and
distribution of life forms in Costa Rican upper
montane Quercus forests. Bryologist 105:334–348.

Jimenez W, Chaverri A, Miranda R, Rojas I. 1988.
Aproximaciones silviculturales al manejo de un roble-
dal (Quercus spp.) en San Gerardo de Dota, Costa Rica.
Turrialba 38(3):208–214.

Kappelle M, Cleef A, Chaverri A. 1992. Phytogeography of
Talamanca montane Quercus forests. J Biogeo 19:299–
315.

———. 1996. Los Bosques de Roble (Quercus) de la
Cordillera de Talamanca, Costa Rica: biodiversidad,
ecologı́a, conservación y desarrollo. Costa Rica: Na-
tional Biodiversity Institute (INBio)-Amsterdam Uni-
versity. 336 p.

Lado C, Pando F. 1997. Flora Mycologica Iberica, Vol. 2.
Myxomycetes I. Ceratyomyxales, Echinosteliales, Li-
ceales, Trichiales. Madrid: J. Cramer. 323 p.

Maimoni-Rodella R, Gottsberger G. 1980. Myxomycetes
from the forest and the Cerrado vegetation in
Botucatu, Brazil: a comparative ecological study. Nov
Hedwig 34:207–245.

Martin G, Alexopoulos CJ. 1969. The myxomycetes. Iowa:
University of Iowa Press. 561 p.

Maurer B. 1982. Statistical inference for MacArthur-Levins
niche overlap. Ecology 63(6):1712–1719.

Milla R, Castro-Diez P, Maestro-Martinez M, Montserrath-
Marti G. 2005. Relationships between phenology and
the remobilization of nitrogen, phosphorus and
potassium in branches of eight Mediterranean ever-
greens. New Phytolog 168(1):167–178.

Morin P. 1999. Community Ecology. Malden, Massachusetts:
Blackwell Science Inc. 424 p.

Novozhilov Y, Schnittler M, Stephenson S. 1999. Myxomy-
cetes of the Taimyr Peninsula (north central Siberia).
Karstenia 39:77–97.

542 MYCOLOGIA



———, ———, Semliaskaia I. 2005. Synecology of myx-
omycetes in desert of the northwestern Caspian
lowland. Mikol Fitopatol 39(4):40–52.

Ogata N, Rico-Gray V, Nestel D. 1996. Abundance, richness
and diversity of myxomycetes in a neotropical forest
ravine. Biotropica 28(4b):627–635.

Petraitis P. 1985. The relationship between likelihood niche
measures and replicated tests for goodness of fit.
Ecology 66(6):1983–1985.

Schnittler M. 2001a. Foliicolous liverworts as a microhabitat
for neotropical Myxomycetes. Nov Hedwig 72:259–270.

———. 2001b. Ecology of myxomycetes of a winter-cold
desert in western Kazakhstan. Mycologia 93(4):653–
669.

———, Stephenson S. 2000. Myxomycete biodiversity in
four different forest types in Costa Rica. Mycologia
92(4):626–637.

———, ———. 2002. Inflorescences of neotropical herbs as
a newly discovered microhabitat for myxomycetes.
Mycologia 94(1):6–20.

———, Unterseher M, Tesmer J. 2006. Species richness and
ecological characterization of myxomycetes and myxo-
mycete-like organisms in the canopy of a temperate
deciduous forest. Mycologia 98(2):223–232.

Stephenson S. 1988. Distribution and ecology of myxomy-
cetes in temperate forests. I. Patterns of occurrence in
the upland forests of southwestern Virginia. Can J Bot
66:2187–2207.

———. 1989. Distribution and ecology of myxomycetes in
temperate forests II. Patterns of occurrence on bark
surface of living trees, leaf litter and dung. Mycologia
81(4):608–621.

———, Kalyanasundaram I, Lakhanpal T. 1993. A compar-
ative biogeographical study of Myxomycetes in the mid-
Appalachians of eastern North America and two
regions of India. J Biogeo 20:645–657.

———, Schnittler M, Lado C. 2004a. Ecological character-
ization of a tropical myxomycete assemblage—Maqui-
pucuna Cloud Forest Reserve, Ecuador. Mycologia
96(3):488–497.

———, ———, Novozhilov Y. 2000. Distribution and
ecology of myxomycetes in high latitude regions of
the northern hemisphere. J Biogeo 27:741–754.

———, ———, Lado C, Estrada-Torres A, Wrigley de
Basanta D, Landolt J, Novozhilov Y, Clark J, Moore D,
Spiegel F. 2004b. Studies of neotropical mycetozoans.
Syst Geo Pl 74:87–108.

———, Stempen H. 1994. Myxomycetes: a handbook of
slime molds. Oregon: Timber Press. 183 p.

———, Studlar S. 1985. Myxomycetes fruiting upon
bryophytes: coincidence or preference? J Bryolog 13:
537–548.

Ulloa C, Møller P. 1993. Arbustos y árboles de los Andes del
Ecuador. Denmark: Aarhus University Press. 264 p.

Uribe-Melendez J. 1995. Catalogo de los Myxomycetes
registrados para Colombia. Caldasia 18(86):23–36.

ROJAS AND STEPHENSON: MYXOMYCETES OF HIGH-ELEVATION OAK FORESTS 543