Fungal communities in wet tropical forests: variation in

Document technical information

Format pdf
Size 214.4 kB
First found May 22, 2018

Document content analysis

Category Also themed
not defined
no text concepts found


Alfredo Di Stéfano
Alfredo Di Stéfano

wikipedia, lookup




Fungal communities in wet tropical forests:
variation in time and space
D. Jean Lodge and Sharon Cantrell
Abstract: Understanding variation in tropical forest fungal populations and communities is important for
assessing fungal biodiversity, as well as for understanding the regulatory roles fungi play in tropical
forests. In wet tropical forests, the canopy is typically occupied by certain wood decomposers,
endophytes, epiphylls, and pathogens. Aphyllophoraceous canopy fungi are a subset of species found in
the understory. Marasmioid agarics in the understory often form extensive networks of rhizomorphs that
trap litter; these and other aerial species are rare on the forest floor. Decomposers are stratified within
the forest floor, with some species colonizing only fresh litter, others preferring decomposed litter, and
others restricted to soil organic matter. Specificity to particular host substrates is frequent among tropical
forest litter decomposers and contributes to spatial heterogeneity in fungal communities over the landscape.
Litter basidiomycetes and microfungal communities in patches of 1 m 2 or less do not significantly
resemble communities in similar patches located at distances greater than 100 m. Disturbances induce
changes in the environment and the abundance of different substrates, resulting in changes in fungal
communities through time, and variation over the landscape. Severe disturbances, as well as the slight
daily variations in rainfall, profoundly affect populations of fungal decomposers and their influence on
plant nutrient availability.
Key words: fungi, tropical forests, diversity, stratification, spatial variation, temporal variation.
Résumé: Il est important de comprendre la variation des populations fongiques et de leurs communautés
en forêts tropicales pour évaluer leur biodiversité, aussi bien que pour comprendre le rô1e régulateur
que jouent les champignons clans ces forêts tropicales. Dans les forêt tropicales ombrophiles, la canopée
est typiquement occupée par certains champignons décomposeurs du bois, endophytes, épiphylles et
pathogènes. Les champignons aphyllophoracé de la canopée constiment un sous-ensemble des
champignons qu’on trouve en sous-étage. Les agarics marasmioïdes du sous-étage forment souvent des
réseaux extensifs de rhizomorphes qui capturent la litiére; ces espéces et d’autres entités aériennes sont
rarement présentes sur le plancher forestier. La distribution des décomposeurs est stratifiée dans le sol
forestier, certaines espéces colonisant seulement la litiére fraîche, d’autres préférant la litière décomposée,
alors que d’autres se limitent à la matière organique du sol. La spécificité aux substrats provenant
d’hôtes particuliers est fréquente chez les décompaeurs de litières forestiéres tropicales, et contribue à
l’hétérogénéité spatiale des communautés fongiques sur l’ensembie du paysage. Les communautés de
basidiomycètes et de microfongi clans des plages de 1 m 2 ou moins, ne ressembient pas significativement
aux communautés occupant des plages similaires situées à des distances de plus de 100 m. Les perturbations
induisent des changements clans le milieu et clans 1‘abondance des différents substrats, ce qui conduit à
des changements des communautés fongiques avec le temps, et à leur variation clans 1’ensemble du
paysage. Les fortes perturbations, ainsi que les petites variations quotidiennes clans la precipitation
affectent profondément les populations des champignons décomposeurs et leur influence sur 1a
disponibilité des nutriments végétaux.
Mots clés : champignons, forêts tropicales, diversité, stratification. variation spatiale, variation temporelle.
[Traduit par la rédaction]
Received August 16, 1994.
D.J. Lodge and S. Cantrell.1 Center for Forest Mycology
Research, Forest Products Laboratory, USDA-Forest
Service, Palmer, PR 00721, U.S.A.
Current address: Department of Plant Pathology, The
University of Georgia. Athens, GA 30602, U.S.A.
The Forest Products Laboratory is maintained in
cooperation with the University of Wisconsin. This paper
was written and prepared by U.S. Government employees
on official time, and is therefore in the public domain and
not subject to copyright.
Understanding the variation in fungal populations in time and
space is important because of its relevance to questions of
biodiversity and the roles fungi play in regulating populations of other organisms and ecosystem processes. Hawksworth (1991) estimated global fungal biodiversity to be
1.5 million based on extrapolation from the ratio of the
number of native plant species to the number of described
fungal species in the British Isles (ca. 1:6). However, Hammond (1992) noted that such estimates of overall species
richness can only be tentative in the absence of good data on
Can. J. Bot. 73(Suppl. 1): S1391-S1398 (1995). Printed in Canada / lmprimé au Canada
tropical fungal communities. on latitudinal and other gradients in diversity, and on how rapidly the numbers of fungal
species increase at greater spatial scales. Of special concern
is that there may be less host specificity where plant communities are especially diverse, as in humid tropical forests
(Hammond 1992), because natural selection can act against
specificity that limits colonization of widely spaced hosts
(Janos 1983). Based on suggestive patterns in published data
and what is generally known about how higher abundance
and partitioning of resources allows for greater diversity,
Hammond (1992) went on to hypothesize that overall forest
architecture may be a better predictor of the number of fungi
and small animals present in a given area than plant species
richness. In addition to providing more resources and surfaces, forests with greater stature and structural complexity
can create more microhabitats and microclimates for insects
and fungi. For example, tall wet tropical forests have long
vertical gradients of temperature and relative humidity that
can accommodate fungal species that are adapted to different
strata (Hedger 1985).
Of equal importance to fungal biodiversity itself are variations in fungal populations that affect host plant diversity and
nutrient cycling. Fungal pathogens contribute to the high
local plant species richness (the total number of species in a
given area) that is typical of many tropical forests by causing
density-dependent mortality in seedlings and juveniles that
results in an unexpectedly greater spacing among individuals
of a population (Augspurger 1983a, 1983b; Gilbert et al.
1994). Burger (1992) hypothesized that plant pathogens
might also increase overall plant diversity on tropical mountains by causing more damage to hosts that are at the margins
of their range, leading to more rapid turnover of species
along elevation gradients. His hypothesis is based on the distributions on congeneric shrub species in Costa Rica that
replace each other more rapidly along elevational gradients
than can be explained by soils or the direct effects of climatic
differences. In tropical forests, temporal variations in fungal
and microbial biomass can potentially alter the fate of nutrients in the ecosystem (Lodge 1993; Lodge et al. 1994;
Zimmerman et al. 1995).
The primary objective of this paper is to review the literature on vertical stratification, small-scale spatial heterogeneity on the forest floor, and temporal variation in certain
fungal communities in wet and moist tropical forests. Some
original data on spatial patterns of basidiomycetes communities in the litter layer are included to complement the
published literature. The implications of these patterns are
discussed, but new estimates of global fungal diversity have
not been generated because the current knowledge base is
still too fragmentary. Most of the data presented here relate
to decompose fungi, although some references to plant
endophytes, pathogens, and epiphylls are included. Variation
in aquatic, mycorrhizal, lichenicolous, and entomopathogenic fungal communities and symbionts of invertebrates in
tropical forests is not addressed.
Vertical stratification
The canopy and understory
Fungi that grow asymptomatically inside live leaves and
Can. J. Bot. Vol. 73 (Suppl. 1), 1995
branches form one of the most pervasive but cryptic guilds
of fungi in rain forest canopies. Such fungi are distributed in
the aerial parts of trees (Rodrigues and Samuels 1990;
Rodrigues et al. 1993; Laessøe and Lodge 1994) and in
epiphytic plants such as members of the Araceae, Bromeliaceae, and Orchidaceae (Dreyfuss and Petrini 1984). Endophytic fungi are typically isolated from healthy foliage and
classified by their cultural and isozymic characteristics
(Rodrigues and Samuels 1990; Rodrigues et al. 1993:
Laessøe and Lodge 1994). Rodrigues and Samuels (1990)
isolated 11 endophytic fungi from unopened and newly
opened leaves of a fan palm, Licuala ramsayi (Muell.)
Domin., that was growing in a lowland tropical rain forest
in Queensland, Australia. Most of their isolates were anamorphs of xylariaceous fungi (Geniculosporium sepens
Chesters and Greenhalgh, Xylaria cubensis (Mont.) Fr.,
Xylaria sp., and Nodulosporium sp. ). but Fusarium solani
(Mart.) Sacc., Stagonospora sp., three species of Phomopsis, and a previously undescribed species, Idriella licuae
K.F. Rodrigues & Samuels, were also found. Rodrigues
et al. (1993) isolated 15 species of Xylaria from healthy
foliage of another palm, Euterpe oleracea Mart., that was
growing in periodically inundated rain forest in the Brazilian
Amazon. Laessøe and Lodge (1994) cultured anamorphs of
nine species of xylariaceous fungi (including X, cubensis.
X. arbuscula Sacc., Xylaria sp. c.f. X. allantoidea (Berk.)
Fr.; and Nodulosporium sp.) from healthy petioles of Schefflera morototoni (Aublet) Maguire et al. in subtropical wet
forest in Puerto Rico. Dreyfuss and Petrini (1984) identified
50 taxa of endophytic fungi growing in 33 species of vines
and epiphytes (Pteridophyta, Araceae, Bromeliaceae,
Orchidaceae, and Piperaceae), but their results were not
Except for X. palmicola Winter (Rodrigues et al. 1993),
the endophytic xylariaceous fungi that could be identified in
the studies cited above are known to have broad host ranges.
This gives weight to Hammond’s (1992) suggestion that host
specificity may be relatively uncommon among fungi of tropical forests with high diversity. However. several Xylaria
species with broad host ranges form species complexes, and
it is unknown whether physiological specialization for particular hosts occurs in these complexes. Isozyme patterns of
X. telfarii Berk. & Fr. and X. multiplex (Kunze) Fr. isolated
from stromata growing on wood clustered in an analysis of
similarity separately from but close to clusters of endophytic
isolates from live palm (Rodrigues et al. 1993). This suggests
that the wood decomposers and palm endophytes of these
Xylaria “species” came from separate populations. It is not
known, however, whether these population differences
resulted from geographic, host, or life history trait differences between the isolates derived from teleomorphs on
wood and those cultured from live palms. Isozyme analyses
of mostly temperate Xylaria isolates (Brunner and Petrini
1992) suggest that all these hypothesized sources of variation
are viable.
Although xylariaceous fungi are known to produce potent
secondary compounds (Whalley and Edwards 1987) that
might protect their host plants from herbivores, their potential beneficial role in tropical forests has not been studied. If
tropical forest endophytes serve to protect their hosts from
herbivores and pathogens, like endophytic clavicipitaceous
H5.1. Lodge and Cantrell
fungi of temperate grasses (Clay 1988), then we need to
know more about their host specificity and mode of dispersal. If they are beneficial, then broad host ranges or maternal
transmission of host-specific endophytes through seeds might
increase the resilience of tropical forests to disturbance and
fragmentation, but host specificity without maternal transmission might increase the susceptibility of these ecosystems
to perturbation.
Pathogens and parasites
Except for diseased leaves that fall to the forest floor, little
is known about plant parasitic fungi in rain forest canopies.
An exception is the Meliolaceae (black mildews), which
were monographed worldwide by Hansford (1961). The
hyperparasites and commensals of meliolaceous fungi have
also received significant attention (Stevenson 1975). Many
other types of leaf and branch parasites have been reported
from the aerial parts of trees and shrubs in Puerto Rico
(Stevenson 1975) and Venezuela (Dennis 1970). Records for
all but those affecting the most commercially valuable tree
crops are too uneven to deduce patterns in fungal populations
and communities. Mycena citricolor (Berk. & Curt.) Sacc. is
one of the best known agaric pathogens of Neotropical rain
forest understory because it is the causal agent of American
leaf spot on introduced coffee (Coffea spp., Rubiaceae).
Although it is a nonspecialized parasite reported on a wide
range of native and introduced hosts (32 spp. in Puerto Rico;
Stevenson 1975), it is especially devastating to coffee trees
(Pegler 1983). Coffee cultivation has largely disappeared
from the Luquillo Mountains of Puerto Rico since the 1940s,
and this fungus has become rare. Small outbreaks of
M. citricolor, however, were found recently (personal observation) on a native shrub in the Rubiaceae, Psychotria
berteriana DC, which had increased densities following
Hurricane Hugo. It may be significant that this plant is the
closest Neotropical relative of Coffea spp. (C. M. Taylor, personal communication).
Crinipellis perniciosa (Stahel) Sing. (Tricholomataceae)
causes Witches’ Broom of cacao (Hedger et al. 1987) and is
the agaric parasite with the most detailed population studies
in Neotropical forests. This fungus is endemic on Theobroma
and Herrania spp. (Sterculariaceae) in the Amazon and
Orinoco basins of South America, and is found as far as the
Lesser Antilles where Theobroma cacao L. is grown (Holliday 1970). Massive epiphytotics (disease outbreaks in plant
populations, analagous to epidemics in animals) occur in
cacao plantations, where the host density is high (Holliday
1970). Although Theobroma and Herrania spp. grow more
rapidly in full sun, they usually occur in the forest understory
because unshaded plants with rapidly growing meristems are
often killed by the disease (H .C. Evans, personal communication). Bastes and Evans (1985) described another pathotype of C. perniciosa that causes witches’-brooms and is
restricted to Solanum spp. in Amazonian Brazil. Hedger
et al. (1987) found that dikaryotic mycelia of the pathotypes
from Theobroma and Solanum were mutually antagonistic,
supporting the hypothesis that they belong to separate populations. Furthermore, genetic exchange among pathotypes is
improbable because C. perniciosa is monokaryotic in its
biotrophic stage, and has been shown to be homomictic
(inbreeding) when it converts to the saprotrophic phase
(Hedger et al. 1987). The production of dikaryotic mycelia
and basidiocarps from single basidiospores of C. perniciosa
(Purdy et al. 1983) is likely to cause reproductive isolation
among individuals if somatic incompatibility among dikary ons is also present. Such reproductive isolation might explain
the proliferation of morphological varieties of C. perniciosa
described by Pegler (1978). A reportedly saprophytic population of C. perniciosa on vines in Ecuador (Hedger et al.
1987) appears to be an undescribed species based on morphological and substratum differences, as well as on incompatibility with strains on cacao.
Other species of Crinipellis and Marasmius may be canopy
pathogens, although those that fruit on dead plant parts are
sometimes assumed to be saprophytic (Pegler 1983). These
fungi are not necessarily strict saprotrophs as shown by
Laessøe and Lodge (1994) when they obtained fruit bodies
of an undescribed Crinipellis sp. aff. C. stupparia and an
undescribed Marasmius sp. from live diseased petioles of
Schefflera morototoni (Aubl.) Maguire that were collected in
the canopy in Puerto Rico. Putative marasmioid parasites
causing thread and horsehair blights of leaves and twigs in
the tropical forests include C. stupparia, Marasmius pulcher
(Berk. & Br.) Petch, M. cyphella Dennis & Reid, and
M. scandens (Dennis and Reid 1957; Weber 1973; Hedger
1985), while M. cymatelloides Dennis & Reid was associated with leaf spots on a leguminous tree in Africa (Dennis
and Reid 1957). However, pathogenicity trials have not been
conducted for most of these aerial agarics (Hedger 1985).
There is no doubt, however, about the pathogenicity of certain aphyllophoraceous species, such as Koleroga noxia
Donk, ‘‘ Corticium” stevensii Burt. (a Koleroga sp. ), and
Corticium areolatum Stahel (Dennis 1970; Weber 1973;
Stevenson 1975). The horsehair fungus Marasmius crinisequi F. Muell. :Kalehbr. forms rhizomorph nets in the understory of many rain forests (Hedger 1990) but is not generally
considered to be a pathogen (Weber 1973).
Influence of pathogens on host diversity
The examples of C. perniciosa and M. citricolor above sug,gest a direct relationship between host density and epiphy totics in tropical forests. Gilbert et al. (1994) recently
showed that an unidentified canker disease of Ocotea whitei
(a common canopy tree in the Lauraceae) caused densitydependent mortality of juveniles close to the parent trees on
Barro Colorado Island in Panama. Augspurger (1983a.
1983b, 1984) and Augspurger and Kelley (1984) have shown
that higher proportions of tree seedlings in Panama survive
nonspecific damping off and other fungal diseases at distances further away from the parent plants. Such relationships may be one of the primary contributing factors to the
high dispersion of tree species and the consequent high
species richness that is characteristic of many wet tropical
forests. Burger (1992) has further suggested that plant pathogens may restrict the upper and lower elevation ranges of
plants, allowing for more rapid replacements of congeneric
species along elevational gradients. Furthermore, when disease resistant plant genotypes colonize an area at higher elevation with a climate more favorable to disease development,
pathogens may select against hybrids between the parent and
offspring populations resulting in reproductive isolation and
parapatric speciation (Burger 1992).
Hutton and Rasmussen (1970) compared fungi that were
cultured from surfaces of leaves of 20 tree species in moist
forest at Fort Clayton. Panama Canal Zone. Although their
methods were unclear as to the replication within plant
species and their presentation of data make certain analyses
difficult, one surprising pattern is suggested by their data. A
total of 36 types of fungi were isolated, of which 12 were
present only in the rainy season, 12 were present only in the
dry season, and 12 were found in both seasons. Three to four
fungal species were cultured from leaf surfaces on each tree
in the rainy season, although this is probably an underrepresentation of the total epiphyllic mycota. Such a large
seasonal turnover in fungal epiphylls was unexpected.
Combined foliage communities
Cowley (1970b) examined vertical stratification and host
specificity of microfungal communities of leaves of two tree
species in the Luquillo Mountains of Puerto Rico. Six samples of leaves were collected at heights of 3-6, 10, and
15-20 m above ground from two trees, Manilkara bidentata
(A. DC) Chev. and Dacryodes excelsa Vahl. The unwashed
leaves were ground and plated on modified Martin’s medium
(Holler and Cowley 1970), and 25 isolates were randomly
selected from the cultures obtained from each leaf sample.
The occurrence on both hosts of only 22 of the 101 fungal
species suggested some specificity, but most of these fungi
were rare. Of the 49 frequently isolated fungal species, however, only 21 were isolated from both tree species. Spores
lying on leaf surfaces may have contributed to the greater
number of fungal isolates that were recovered from the lower
levels of the canopies (Cowley 1970b). Comparison of only
the most frequent fungal isolates from the uppermost canopies, where spore contamination was less, showed that five
species were only on D . excelsa, eight were only on
M. bidentata, and just three were found on both tree species.
Wood decomposers
In a study comparing wood decomposer fungi that were fruiting on twigs in the canopy and the understory of a rain forest
in Cameroon, West Africa, the canopy fungi were mostly a
subset of those found between 1.5 and 2 m above the forest
floor (Nuiiez and Ryvarden 1992; Ryvarden and Nuiiez
1992). If data from the two reports on this study are combined, a total of 62 taxa were recognized at the species, morphospecies, or generic level. Of these, 14 were found in both
the canopy and the understory, and three (ca. 5%) were
found only in the canopy. The extreme moisture and temperature regimes of the canopy are apparently more selective,
resulting in a lower species richness (Ryvarden and Nuftez
1992). Species of Aphyllophorales that occurred in the
canopy generally had long-lived basidiocarps, such as perennial polypores, or a catahymenium that protected the fruiting
body during drought (Nuiiez and Ryvarden 1992).
Ephemeral fungi in the canopy might have been undersampled in the Cameroon study cited above, since canopy
exploration was conducted during the dry season using a
warm air balloon. Some species of agaric fungi are known
to live in canopies of rain forests, but fruiting may be rare
and confined to the wet season. For example, Armillaria sp.
is known to form mycorrhizal with epiphytic orchids but is
Can. J. Bot. Vol. 73 (Suppl. 1), 1995
rarely seen fruiting in the canopy (Hedger 1985). A rare
species, Cystoderma cf. austrophalax Sing.. was found on a
branch that fell out of a tree during a tropical storm in Puerto
Rico (personal observation). The only other specimen of
C. austrophalax was collected and described by the late
Dr. Rolf Singer from a branch that fell out of a tree during
a storm in Patagonia (personal communication).
Wood decompose fungi with ephemeral fruiting bodies
appear to be more common and diverse in the rain forest
understory than the canopy. Ephemeral Aphyllophorales in
the understory include species of Trechispora and L.eptosporomyces (Ryvarden and Nuiiez 1992) and Caripia.
Deflexula, Fistulina, Polyporus, Pleurotus. and Porodisculus (personal observation). Fruiting bodies of tremellaceous
fungi are also abundant in the understory (Ryvarden and
Nuiiez 1992).
The relatively high humidity in the understory of tropical
rain forests also allows the development of saprotrophic
agaric communities in litter and wood above the forest floor
(Hedger 1985). Agaric genera typically found in this forest
stratum are Anthracophyllum, Campanella, Chaetocalathus,
Collybia, Coprinus. Crepidotus, Crinipellis, Dictyopanus.
Gymnopilus, Hohenbuhelia, Marasmius, Marasmiellus~
Mycena, Gerronema, Gloeocephala, Favolaschia, Filoboletus, Lentinula, Melanotus, Micromphale, Nothopanus,
Psathyrella, Resupinatus, Rimbachia, Tetrapyrgos, and
Trogia. These aerial agaric communities generally have a
distinct species composition compared with those colonizing
the same substrata on the forest floor because they are
adapted to different moisture regimes and they differ in their
competitive abilities (Hedger 1985; Hedger et al. 1987). The
prevalence of species with gelatinized tissue in their basidiocarps among the ephemeral fungi of the understory,
including species of Auricularia, Campanella, Filoboletus,
Hohenbuhelia, Favolaschia, Marasmiellus, Mycena, Micromphale, Resupinatus, and Tremella, suggests that gelatinized
zones offer protection from desiccation. The ability of the
basidiocarps of marasmioid fungi to revive with moisture following desiccation is advantageous in an aerial environment,
where relative humidity fluctuates more rapidly than on the
forest floor (Hedger 1985). Marasmius crinis-equi, Marasmius nigrobrunneus (Pat.) Sacc., Micromphale brevipes
(Berk. & Rav.) Sing., and Crinipellis spp. form networks of
rhizomorphs in the understory (Pegler 1983; Hedger 1990)
that trap and decompose a significant percentage of the litter
fall (Hedger 1990).
The rain forest floor
Stratification in the litter layer
Hedger (1985) described in detail the stratification of basidiomycetes decomposer communities on the floor of a cacao
plantation with a moist forest overstory in Ecuador. He
found that freshly fallen leaf litter was dominated primarily
by species of Marasmius, Marasmiellus, and Mycena, wherein
the lower litter layer was dominated by species of Clitocybe.
Collybia, Geastrum, Lepiota, and Leucocoprinus. Hedger
further showed that colonization of autoclaved freshly. fallen
litter by Lepiota spp. (which normally grows in the lower litter layer; F1/F2 horizons) was all but impossible. but that
Lepiota pseudoroseola grew well in flasks of litter previously
H5,1. Lodge and Cantrell
decomposed by Marasmius pallescens Murr. and M. foliicola Sing. Thus, some of the vertical stratification in the litter layer appears to be related to successional substrate
preferences rather than competitive exclusion.
Litter communities in the subtropical wet forest floor at
350 m elevation in the Luquillo Mountains of Puerto Rico
were similar to those described by Hedger, but some differences were also apparent (personal observation). Most striking was the prevalence before Hurricane Hugo in 1989 of
Collybia johnstonii (Murr.) Dennis in both the middle and
upper litter layers (Lodge and Asbury 1988). These agarics
are important in rapidly trapping freshly fallen leaves and
binding them to the litter mat, thereby reducing the loss of
litter and soil organic matter from the ecosystem through
erosion (Lodge and Asbury 1988). Some of these litterbinding species were found to translocate significant quantities of 32P from partly decomposed litter into freshly fallen
leaves (Lodge 1993). The ability to translocate nutrients
probably increases the competitive abilities of these decomposers by increasing the speed by which they can colonize
newly fallen leaves, which are especially deficient in phosphorus (Lodge 1993). Furthermore, nutrient recycling by
these fungi may accelerate the rate of litter decomposition
(Lodge 1993).
attempt to match among samples, there remained 43.47, 56.
and 72 species per sample (total species richness within samples was 78–134; Bills and Polishook 1994). The number of
shared species in pairwise comparisons of samples was nine
to fourteen (15–28% overlap, or 72 – 85% complementarily). Complementarily in species composition among samples should approach or be less than 50% so as not to miss
an unknown portion of the total species present (Coddington
et al. 1991), suggesting that this fungal community was
undersampled despite the high number of species obtained by
Bills and Polishook. This result confirms the authors’ conclusions that the litter microfungal community was undersampled, which they based on the absence of asymptotes in
their species rarefaction curves (Bills and Polishook 1994).
In a study of microfungi cultured from leaf litter using comparable methods in a seasonal, moist tropical forest in
Panama (Cornejo et al. 1994), 500 morphospecies of fungi
belonging to 48 genera were identified. Morphospecies are
presumed to correspond with real species, since they are
recognized and separated by combinations of the same cultural, macro- and micro-morphological characters that are
used to separate recognized species, but collections are tallied regardless of whether they have been assigned to a
known species.
Fungi of soil organic matter
A few fungi on the rain forest floor apparently grow from
mineral soil, which may contain as much organic matter as
temperate loams. Some, but not all species of the Entolomataceae, Hygrocybe, and Neopaxillus belong to this
group. Most of the other macrofungal species that fruit on
mineral soil are connected to a large organic food base, such
as rotting wood (especially members of the Phallales), decaying roots (e.g., Laeriporus persicinus (Berk. & Curt. ) Gilbn.
and some Xylaria spp. ), and accumulations of humus from
wood or abandoned termite and leaf-cutting ant colonies
(e.g., ‘Tricholoma’ praegrande (Berk.) Sacc., and Trichoglossum hirsutum (Pers.) Baud.),
Agaric fungi in litter
J. Lodge and S. Cantrell (unpublished data) studied the
spatial relatedness of agaric litter decomposer communities
along two transects in Cuyabeño, Ecuador (Amazonian
forest, 00°0'N, 76°11'W; ca. 150 m elevation). Twelve
(1× 1 m) plots were established in Tierra Firme forest along
trails in each of two areas. Plots were located 50 m apart
along transect 1 and 50 paces (ca. 25 m) apart along transect 2. Except where noted, only plots sampled within the
same 2-day interval were compared since many of the species are ephemeral. Agarics were collected on the following
dates: August 3, plots 1 –6 on transect 1 and plots 1-3 on
transect 2; August 4, plots 7 – 12 in transect 1: August 7,
plots 4-9 in transect 2; August 10, plots 10– 12 on transect 2. Samples were identified to morphospecies in the field
and then compared microscopically to differentiate or match
similar collections.
Seventy species of litter agarics were found among the
24 plots. Total species richness for the sample period was
estimated to be 99.5 (88.3 – 110.7, 95% confidence interval)
using a Jacknife test (Krebs 1989). Sorensen’s similarity
coefficients (Krebs 1989) were calculated for each pair of
plots that were sampled within a 2-day interval. The average
similarity between agaric communities (on a scale off) to 1)
was 0.22 at ca. 25 m, 0. 15-0.17 at 50 m, and then fluctuated around 0. 13 at greater distances. For all plots within
transect 1, the species that occurred in more than one plot
were examined using a x2 analysis to determine if they were
more likely than expected from random to occur in an adjacent plot rather than a more distant plot. Agaric species were
significantly more likely to occur in an adjacent plot at 50 m
distance than expected (p = 0.02). Together, these data suggest that agaric communities in litter were spatially correlated only at distances of less than 100 m, and that 1 -m2
samples taken at greater distances were independent and
largely nonoverlapping in their species composition.
Variation in time and space
Understanding how fungal populations and communities are
spatially and temporally distributed in tropical forests is
fundamental to estimating their diversity. Such information
is also useful in determining how fungal populations affect
the abundance and distribution of other organisms and ecosystem processes at the landscape level.
Spatial variation
Microfungi in litter
Information on the patterns of dispersion in fungal populations and communities is needed to adequately sample their
diversity in tropical forests. To this end, data from a study
by Bills and Polishook (1994) in which microfungi were cultured from litter of a lowland tropical forest in Costa Rica
were reanalyzed to determine the degree of nonoverlap (i.e.,
complementarity; Coddington et al. 1991) in species composition among their four samples. Each litter sample was from
an area of approximately 0.2 m 2. located at least 0.5 km
apart. After eliminating from the data set single occurrences
of sterile mycelia and coelomycetes that the authors did not
Microfungi in roots and soil
There are few published studies of endophytic fungi in the
roots of tropical plants but work with European tree roots
suggests that endophytes may be present in considerable
numbers (Fisher et al. 199la, 1991b). In one of the few published studies from tropical forests, Holler and Cowley
(1970) compared microfungal communities cultured from
fine roots, litter, and soil in wet tropical forest at El Verde,
Puerto Rico. Microfungal community similarities were high
among samples within each of the three substratum types
(70-74%). In contrast, there was little similarity in the
microfungal communities among substratum types (10 –23 %
similarity). Thus. the species richness of microfungi growing
in roots and soil of wet tropical forests is largely in addition
to the species found in litter.
Can. J. Bot. Vol. 73 (Suppl. 1), 1995
nutrients in the ecosystem (Anderson and Swift 1983: Lodge
1993; Lodge et al. 1994). In subtropical wet forest in Puerto
Rico, fungal biomass accounted for a mean of one third of
the phosphorus in the litter layer, but this varied with litter
moisture from 3 to 85% (Lodge 1993). Such fluctuations in
fungal biomass may result in microbial immobilization and
conservation of nutrients against leaching during the rainy
periods (Behera et al. 1991, Yang and Insam 1991) and
pulses of mineralization in response to drying (Lodge et al.
1994). Trees may have a competitive advantage if proliferation and uptake of nutrients by roots is synchronous with
mineralization of nutrients from dead microbial biomass
(Lodge 1993, Lodge et al. 1994). In cases where woody substrates are especially abundant. however, fungi may compete
directly with trees for limiting nutrients. For example in
Puerto Rico, microbial nutrient immobilization resulting
from hurricane woodfall apparently reduced the availability
of nitrogen and possibly other limiting nutrients to the
canopy trees, thereby slowing forest recovery (Lodge et al.
1994; Zimmerman et al. 1995).
Disturbances from seasonal changes in rainfall and
treefalls to hurricanes can differentially affect fungal species.
Hutton and Rasmussen (1970) found in Panama that 8 of the
23 species they sampled had no epiphyllous fungi in common
between the rainy and dry seasons and 10 plant species had
only one epiphyllous fungus common to both seasons. Some
wood decomposer fungi fruit disproportionately in forest
gaps (e.g., Daldinia escholtzii (Ehrenb.) Starb., Pycnoporus
sanguineous (L.: Fr.) Murr., Schizophyllum commune Fr.,
and some Lentinus spp.), suggesting they may be adapted to
low moisture or high temperatures. This hypothesis is further
supported by growth data for P. sanguineous and S. commune (Castillo Cabello et al. 1994). In contrast, basidiomycetes litter decomposers that have superficial mycelia are
especially sensitive to drying (Hedger 1985). For example,
in the El Verde Research Area of Puerto Rico, only 3 of the
20 regularly surveyed mycelia of Collybia johnstonii could
be found in the litter layer during the 2 years following
Hurricane Hugo because of increased solar irradiation and
desiccation (J. Lodge, unpublished data). By 5 years posthurricane, nine additional mycelia of C. johnstonii had reappeared at previous localities and three mycelia appeared in
new locations. All the mycelia that were on ridges remained
uncolonized by mat-forming species or were replaced by
species of Marasmius, Marasmiellus, Tetrapyrgos, and
Micromphale that were more resistant to desiccation.
Cowley (1970a) found that microfungal communities in
decomposing leaf litter at El Verde were also sensitive to
stress and so affected by opening of the canopy and irradiation that they no longer resembled communities on the same
litter species in undisturbed forest. Thus, disturbances that
disrupt the canopy contribute to spatial as well as temporal
shifts in populations and communities of decompose fungi.
Substratum specificity
Host preferences among decompose fungi might contribute
to the high spatial heterogeneity in fungal communities noted
in the studies above. Several Xylaria spp. are known to be
host-specific, including one that was shown to be strictly
saprophytic (Laessøe and Lodge 1994). Some of the microfungi isolated from litter in Costa Rica were recovered as leaf
endophytes and pathogens in the same forest (Bills and
Polishook 1994), and distributions of those that are host
specific may correspond to the hyperdispersed patterns of
their hosts. Cowley (1970a) studied microfungi in leaf litter
of six different tree species at El Verde in the Luquillo
Mountains of Puerto Rico. Freshly fallen leaves of each
species were air dried and placed in litterbugs at two sites,
one of which was later subjected to gamma irradiation
(10-35 kr; 10 bags of each species per site). The decomposing litter samples were immediately frozen when brought
in from the field, which probably eliminated some fungal
species, but a total of 116 and 258 morphospecies were
recovered from the pre- and post irradiation sample periods,
respectively. Similarity indices of the microfungal communities based on frequencies of occurrences in samples pooled
across all the harvest times prior to irradiation showed a high
fungal community similarity within leaf litter species between
the two sites. Relatively larger differences in fungal communities were found between the six litter species (0.2 -0.48
similarity) as compared with within-leaf species (0.51 -0.68
similarity). In a study of microfungi in moist forest in
Panama, Cornejo et al. (1994) found that several species
occurred differentially on decomposing leaves of five tree
A few wet tropical forest agarics (personal observation)
and discomycete fungi (B. Spooner, personal communication) show distinct host preferences, but they are not the most
abundant nor dominant species. For example, Hedger (1985)
found in an Ecuadorian cacao plantation that agaric litter
decomposers were generally not host restricted, except at the
very gross level of monocotyledons versus dicotylendons. In
contrast, preferences for certain substratum types such as
twigs and woody petioles versus leaves is very common
(Hedger 1985; personal observation),
Implications of variation in time and
space-a summary
Temporal variation
The waxing and waning of fungal mycelia on the floor of a
wet tropical forest can significantly affect the fate of limiting
The great number of available environments and substrata in
wet tropical forests allows for niche partitioning among
agaric and xylariaceous fungi in vertical space. This leads to
high local diversity of these fungi when projected on an area
H5.1. Lodge and Cantrell
basis and also makes sampling such diversity technically
difficult. Although fungi that are endophytes of canopy
leaves may protect plants from herbivory, little is known of
their roles, host specificity. and population structures in tropical forests. Among the Aphyllophorales, the canopy species
are primarily a subset of those in the understory. There is
relatively little overlap of the agaric communities of the
understory and forest floor. Vertical stratification of fungal
communities within the rain forest floor results in efficient
recycling of nutrients, and may affect the rate of litter
decomposition and the temporal pattern of nutrient release.
Pathogenic fungi that affect stems and seedlings increase
in response to higher host densities, contributing to the
hyperdispersion of their hosts and high local species richness
typical of many tropical forests. Some endophytic and pathogenic fungi also play a role in the early stages of decomposition and nutrient cycling, and may contribute to the high
spatial heterogeneity of decomposer communities on the rain
forest floor. Disturbances that change the environment or
increase the availability of woody substrata can affect populations and communities of fungal decomposers on the rain
forest floor. Such changes in fungal populations and communities can directly affect populations of fungivorous animals,
as well as altering the availability and fate of limiting mineral
Can. J. Bot. Vol. 73 (Suppl. 1), 1995
on recycled paper

Report this document