| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
CABI Bioscience, Bakeham Lane, Egham, Surrey TW20 9TY, UK
Coralie M. Simmons
Iwokrama International Centre for Rainforest Conservation & Development, 67 Bel Air, Georgetown, Guyana
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
Endophytic fungi were isolated from living symptomless leaves of 12 tree species from two locations in the Iwokrama Forest Reserve, Guyana. Sixty-four fungal morphotaxa were characterized from 2492 cultures, which were derived from a total of 2520 sample units. Species of Colletotrichum, Nodulisporium, Pestalotiopsis and Phomopsis were most frequently isolated. Colonization was greater in samples from the midrib than in those from laminar tissue, and slightly greater at the tip of the lamina compared with the base of the leaf. In contrast to studies in temperate ecosystems, no distinct fungal communities were identified for individual plant species, suggesting that the degree of host preference is low. The implications for estimation of fungal diversity in tropical systems are explored.
Key words: fungal diversity, host preference, leaf endophytes, neotropical mycology
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
). Numbers of species recovered from culture of surface-sterilized plant fragments may be substantial, and they have received much attention over the last ten years as potential sources of biologically active chemicals (e.g., Monaghan et al 1995
). Most studies have focused on endophytes of temperate plants, although the number of tropical studies is now increasing (Rodrigues and Petrini 1997
). In a few cases (e.g., the grass endophytes belonging to the Clavicipitaceae) fungi grow actively within host tissues in apparently mutualistic relationships (Lane et al 2000
). In other instances, fungal propagules appear to infect plant tissues laterally by transmission via air or water-splash from surrounding saprobic colonies (Wilson 2000
), and then become dormant. Active growth within the plant tissue is initiated only on its senescence, allowing the endophytes to become primary colonizers of the dead material.
Most research to date has investigated the endophytes of single plant species. Specificity of at least some fungus/plant interactions has been widely assumed at least at the genetic level, and it has been claimed that endophyte communities (or at least community profiles) are usually specific at the host species level (Petrini 1996
, Petrini and Fisher 1988, 1990
). However, most analyses have been carried out in temperate ecosystems where patterns of plant and probably also fungal communities differ from those observed in tropical ecosystems. Recent studies in the tropics (Arnold et al 2001
) have identified distinct host-related communities in tropical tree leaves, but on a quantitative rather than qualitative basis. Thus, few endophytic fungi were found to be entirely restricted to particular plant species, but significant differences were found in the frequency of infection of individual morphotaxa. This phenomenon has been termed host preference, following similar observations of decomposer fungal communities by Lodge (1997)
. Assumed host specificity at least of a proportion of tropical endophytes, coupled with the further assumption that fungi are significantly more diverse in tropical regions than in temperate systems, has provided one of the arguments supporting hypotheses of megadiversity in the fungal kingdom (Arnold et al 2000
, Dreyfuss 1986
, Dreyfuss and Chapela 1994
, Hawksworth 1991
).
The current study was designed to test the hypothesis that endophytic communities in the tropics are specific to their host plants. The research was carried out as a component of a combined fungal bioinventory and bioprospecting programme.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
|
UPGMA cluster analysis was performed on the species profile results using the Gower General Similarity Coefficient with the computer package MVSP for Windows (Kovach Computing, Anglesey, UK), and Sorensen's indices were computed using Colwell's program EstimateS version 6 (program and information on calculating the index available at http://viceroy.eeb.uconn.edu/estimates).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
|
|
|
|
|
|
), because not all cultures obtained from the leaf fragments were identified. However, Sorensen indices of similarity were calculated for all possible combinations of samples (Table IV
). The average for the subset of samples where two were analyzed from the same plant species (n = 10, µ = 0.496, 
= 0.206) was rather larger than the averages for all samples from each of the two collecting sites (n = 10, µ 0.315, = 0.114 for the 8 Mile site, and n = 7, µ = 0.342, = 0.133 for the 4 Mile site), but the data are insufficient to draw unequivocal conclusions. All plant species were present in both sites.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
). Comparison between studies is problematic because of differences in protocols, most obviously in the size of plant fragments employed. There were no obvious patterns in the extent of colonization, with wide variety within individual samples, and no clear correlation between multiple samples of the same plant species. These results suggest that the pattern of endophyte colonization is chaotic, and that the sample size was too small to identify host-related differences in colonization if these exist.
The greater colonization of the midrib compared with that of laminar tissues presumably at least partially reflects the slightly larger size of the leaf fragments, but perhaps also the more complex anatomical structure. The somewhat (but not usually statistically significant) greater colonization of the leaf tip region compared with the leaf base could possibly be explained by the leaf tip being more prone to infection, as rainwater draining off at the apex would tend to wash unestablished fungal propagules on the leaf surface towards the leaf tip (Wilson and Carroll 1994
).
Multiple colonization of individual leaf fragments may have been underestimated in this and other studies. Rapidly growing strains are likely to overgrow more slowly growing forms and reducing the number of colonies observed, so it is possible that morphologically similar young colonies will not be separated at the subculture stage. However, competitive interactions often result in sectoring, and therefore in different appearances. The high figures obtained for multiple colony growth from individual fragments underline the importance of culturing from minute leaf fragments (Lodge et al 1996a
). Our surface sterilization method (which took place after the leaves were cut into small fragments) might have reduced further the effective size of the plant subsamples due to penetration of the sterilant into the cut edges, but as both overall and multiple colonization measurements are comparable with those obtained elsewhere, we do not believe that significant undersampling occurred.
Identification and specificity –
Species concepts are currently uncertain on a worldwide basis for all of the dominant taxa isolated (Brasier 1977
, Cannon et al 2000
, Rehner and Uecker 1994
), and indeed for many of the remaining morphotaxa. Bearing in mind the current rudimentary state of knowledge of the Guyanese mycota (Nishida 1989
), difficulties in identification might be expected. At least one of the species detected (referred to as Ascomycota I) is sufficiently distinctive to be recognized provisionally as belonging to an undescribed genus, and will be described if appropriate after further studies to establish its relationships.
The number of fungal species recovered is probably a significant underestimate of the species richness actually present, although it has frequently been observed that a relatively small guild of species make up a dominant component of endophyte taxa in both temperate and tropical ecosystems (Bills 1996
). The recovery of between 7 and 25 fungal morphotaxa from sets of 150 leaf subsamples is comparable with results from other tropical studies. For example, Lodge et al (1996a)
found 22 species from 630 subsamples of Manilkara bidentata, and Rodrigues (1994)
found 57 fungal species from 13 320 samples of Euterpe oleracea, but only 13 species were recovered from 1512 leaf and rachis samples of Spondias mombin from Brazil (Rodrigues and Samuels 1999
). Rather higher figures have been obtained, e.g., 242 morphospecies derived from 984 samples of Heisteria concinna and 259 from 1008 samples of Ouratea lucens leaves (Arnold et al 2000
). In these cases, however, identifications were not attempted and morphospecies were defined using cultural characteristics alone. This may have resulted in significant overestimation of species numbers. Preliminary studies described in that paper comparing cultural morphology with ITS sequence variation suggested that the true species number was around 350 rather than the 418 cultural morphotaxa recovered from the two plant species, but within-species ITS sequence variation is frequently 3–4% and may be more than 11% (Cannon et al 2000
). Therefore, ITS sequences in isolation are not necessarily effective means of distinguishing species, although they have great potential as identification aids for non-sporing cultures (Guo et al 2000
).
These figures emphasize the extreme effort required to obtain anywhere near a complete survey of culturable endophytes from plant tissue. In our study, endophyte species diversity could not be estimated accurately due to the methods used: only a proportion of the cultures generated were grown up and identified due to the need to provide as wide a range of samples as possible for the bioprospecting programme. However, as the species richness results are comparable to those in studies where all cultures obtained were assessed, this suggests that the strategy of partial analysis did not result in the overlooking of a significant number of species.
The fungal species obtained as Manilkara bidentata endophytes from Iwokrama were contrasted with those reported from the same plant species in Puerto Rico by Lodge et al (1996a)
. The results are not exactly comparable, but we identified Colletotrichum (sexual morph Glomerella), Nodulisporium (probably asexual morphs of Xylaria spp.) and Pestalotiopsis species as frequent colonizers in common with the study from Puerto Rico. We also found Phomopsis species (single isolations of four different taxa), while Lodge et al found only a single "rarely isolated" Phomopsis species which they identified as P. manilkarae. The Phyllosticta species they isolated commonly was not recovered from the Iwokrama samples. The two data sets suggest that the endophyte communities of Manilkara bidentata in Guyana and Puerto Rico are not closely similar, suggesting that geographical rather than plant-linked factors may be more important in determining their composition.
Extensive research on endophytes of diverse plants in temperate regions has identified distinct communities for each plant species (Petrini 1996
), usually with a small number of dominant fungi which may be evident as saprobes following death of the host tissues. Host specificity is often studied at the morphological level, but has also been demonstrated in a few studies at a molecular level using RAPD markers and other electrophoretic techniques (Hämmerli et al 1992
, Leuchtmann et al 1992
). In this study, the morphotaxa were tightly defined using micromorphological characteristics and cultural appearance, and probably at least in some cases reflect infraspecific rather than specific variation.
In contrast to endophyte population profiles evident in temperate plants, this study has not identified distinct patterns of specificity or colonization preference. We found that 35 of the 64 fungal morphotaxa distinguished were recorded from more than one plant species, and 28 of these were isolated from at least three plant taxa. Although 29 morphotaxa were isolated only from one leaf sample, in most cases these involved single isolation events which could merely indicate rarity rather than host specificity. In the few cases where multiple isolations of taxa restricted to one collection did occur, the species involved were mostly recognized as widespread and plurivorous (e.g., Botryodiplodia theobromae, Clonostachys rosea, Fusarium decemcellulare, and Haematonectria haematococca). In only one case was a potentially host-specific species isolated on more than one occasion from the same plant collection, a possibly undescribed species tentatively assigned to Coniella, but even here the two isolates were derived from tip and base respectively of the same host leaf.
The statistical comparison of species profiles using the Gower General Similarity Coefficient (see Fig. 3 ) did not provide any clear evidence of host preference/specificity, but comparison of within-host and within-site Sorensen's indices (Table IV ) did suggest that the plant species plays a somewhat more significant role in defining endophyte communities than does the locality. However, the range of Sorensen indices for the paired plant samples is very large (0.27–0.82), and the most divergent pair of samples is actually the two Jacaranda species which were both made at the 8 Mile site, only a few yards from each other. These data do not therefore provide reliable evidence of host preference. There was no evidence that host identity is correlated with endophyte profile at family level.
Though it was not possible to identify distinct endophyte assemblages for the plants studied, specific fungal groups may be absent from particular plant species. For example, Colletotrichum species do not seem to be harbored by Jacaranda sp. and Cecropia sciadophylla leaves, and were rarely isolated from Carapa guianensis. The apparent absence of both Nodulisporium and Pestalotiopsis as endophytes of Chlorocardium rodiei leaves is of note, and largely explains the similarity of the two samples from this plant species as indicated using cluster analysis of the fungal species profiles (Fig. 3 ). It is possible that these apparent absences are due to undersampling; Xylaria stromata have been observed to develop from fallen Chlorocardium rodiei fruits in the Iwokrama Forest (Cannon unpubl), which are known to be sexually reproducing morphs of Nodulisporium species. There is no obvious correlation between physical structure of the leaf (which might affect both colonization and survival of the surface-sterilization process) either with endophyte diversity or the presence of individual morphotaxa.
The frequent isolation of apparently identical endophytic strains from plant tissues of unrelated species suggests that the extent of host preference/specificity in tropical leaf endophytes is small, and molecular research is being conducted on two key groups (Pestalotiopsis and Colletotrichum) in order to investigate these issues in more detail. The overall patterns of endophyte colonization observed in the samples from Iwokrama suggest that endophytic fungus/plant interactions may not be as strictly defined as those in temperate ecosystems. This may possibly be related to the more complex patterns of plant diversity encountered in tropical forests in comparison with those of temperate regions, where hectare plots sometimes contain hundreds of tree species in contrast with the temperate pattern of one or a few dominant taxa (Gentry 1988
, Valencia et al 1994
), although the forests of Guyana are notably less tree species-rich than the hyperdiverse ecosystems of Amazonian Ecuador (Lindeman and Mori 1989
, Polak 1992
). High tree species diversity in tropical forests has been suggested by Janzen (1970)
and Connell (1971)
to be associated with escape from natural enemies which cause disproportionately high mortality close to adult trees (Lodge et al 1996b
). Assuming that pathogenesis is linked to host specificity, one might predict that pathogen diversity would be low in such ecosystems. The observed lack of host specificity among the leaf endophytes studied here, many of which belong to well-known pathogen groups, provides tentative support for this hypothesis. Indications of extreme fungal diversity in the tropics have been discussed in a number of recent papers (e.g., Lodge et al 1995
, Hyde and Hawksworth 1997
) following a well-publicized global estimate of fungal diversity based largely on fungus/plant species ratios in temperate ecosystems (Hawksworth 1991
). Such estimates have been challenged, citing patterns of plant distribution as reason for caution (May 1991, 1994
), and it is debatable whether fungal diversity estimates can be extrapolated realistically to the tropics using these techniques. Tropical fungal diversity may not be so extensive as has been assumed.
|
ACKNOWLEDGMENTS |
|---|
Accepted for publication August 1, 2001.
|
LITERATURE CITED |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
———, ———, ———, Coley PD, Kursar TA., 2000 Are tropical fungal endophytes hyperdiverse? Ecology Letters 3:267-274
Bills GF., 1996 Isolation and analysis of endophytic fungal communities from woody plants In: Redlin SC, Carris LM, eds. Endophytic fungi in grasses and woody plants. St Paul, Minnesota: APS Press. p 31–65
Brasier CM., 1997 Fungal species in practice: identifying species units in fungi In: Claridge MF, Dawah HA, Wilson MR, eds. Species: the units of biodiversity. London: Chapman and Hall. p 135–170
Cannon PF, Bridge PD, Monte E., 2000 Linking the past, present and future of Colletotrichum systematics In: Prusky D, Freeman S, Dickman MB, eds. Colletotrichum: host specificity, pathology and host-pathogen interaction. St Paul, USA: APS Press. p 1–20
Connell JH., 1971 On the role of natural enemies in preventing competitive exclusion in some marine mammals and in rain forest trees In: Van der Boer PJ, Gradwell GR, eds. Dynamics of numbers in populations. Wageningen, Netherlands. p 298–312
Dreyfuss MM., 1986 Neue Erkenntnisse aus einem pharmakologischen Pilz-screening Sydowia 39:22-36
———, Chapela I., 1994 Potential of fungi in the discovery of novel, low-molecular weight pharmaceuticals In: Gullo VP, ed. The discovery of natural products with therapeutic potential. Boston: Butterworth-Heinemann. p 49–80
Gamboa MA, Bayman P., In press Communities of endophytic fungi in leaves of a tropical timber tree (Guarea guidonia: Meliaceae) Biotropica
Gentry AH., 1988 Tree species richness of open Amazonian forests Proc Nat Acad Sc 85:156-159
Guo LD, Hyde KD, Liew ECY., 2000 Identification of endophytic fungi from Livistona chinensis based on morphology and rDNA sequences New Phytologist 147:617-630
Hämmerli UA, Brändle UE, Petrini O, McDermott JM., 1992 Differentiation of isolates of Discula umbrinella (teleomorph: Apiognomonia errabunda) from beech, chestnut and oak using RAPD markers Molecular Plant-Microbe Interactions 5:479-483[Medline]
Hawksworth DL., 1991 The fungal dimension of biodiversity: magnitude, significance and conservation Mycol Res 95:641-655
Hyde KD, Hawksworth DL., 1997 Measuring and monitoring the biodiversity of microfungi In: Hyde KD, ed. Biodiversity of Tropical Microfungi. Hong Kong: University of Hong Kong Press. p 11–28
Janzen DH., 1970 Herbivores and the number of trees in tropical forests American Naturalist 104:501-528
Lane GA, Christensen MJ, Miles CO., 2000 Coevolution of fungal endophytes with grasses: the significance of secondary metabolites In: Bacon CW, White JF, eds. Microbial endophytes. New York & Basel: Marcel Dekker. p 341–388
Leuchtmann A, Petrini O, Petrini LE, Carroll GC., 1992 Isozyme polymorphism in six endophytic Phyllosticta species Mycol Res 96:287-294
Lindeman JC, Mori SA., 1989 The guianas In: Campbell DG, Hammond HD, eds. Floristic Inventory of Tropical Countries. New York: New York Botanical Garden. p 375–390
Lodge DJ., 1997 Factors related to diversity of decomposer fungi in tropical forests Biodiversity and Conservation 6:681-688
———, Chapela I, Samuels G, Uecker FA, Desjardin D, Horak E, Miller OK, Hennebert GL, Decock CA, Ammirati J, Burdsall HH, Kirk PM, Minter DW, Halling R, Læssoe T, Mueller G, Huhndorf S, Oberwinkler F, Pegler DN, Spooner B, Petersen RH, Rogers JD, Ryvarden L, Watling R, Turnbull E, Whalley AJS., 1995 A survey of patterns of diversity in non-lichenized fungi Mitteilungen aus dem Eidgenössische Forschunganstalt für Wald-, Schnee- und Landwirtschaft 70:157-173
———, Fisher PJ, Sutton BC., 1996a Endophytic fungi of Manilkara bidentata leaves in Puerto Rico Mycologia 88:733-738
———, Hawksworth DL, Ritchie BJ., 1996b Microbial diversity and tropical forest functioning In: Orians GH, Dirzo R, Cushman JH, eds. Biodiversity and Ecosystem Processes in Tropical Forests. Berlin & Heidelberg: Springer-Verlag. p 69–100
Magurran AE., 1988 Ecological Diversity and its Measurement Princeton, USA: Princeton University Press. 179 p
May RM., 1991 A fondness for fungi Nature 352:475-476
———. 1994 Conceptual aspects of the quantification of the extent of biodiversity Phil Trans Roy Soc London series B 345:13-20
Monaghan RL, Polishook JD, Pecore VJ, Bills GF, Nallin-Omstead M, Streicher SL., 1995 Discovery of novel secondary metabolites from fungi—is it really a random walk through a random forest? Can J Bot supplement 73:S925-S931
Nishida FH., 1989 Review of mycological studies in the Neotropics In: Campbell DG, Hammond HD, eds. Floristic Inventory of Tropical Countries. New York: New York Botanical Garden. p 494–522
Petrini O., 1996 Ecological and physiological aspects of host specificity in endophytic fungi In: Redlin SC, Carris LM, eds. Endophytic Fungi in Grasses and Woody Plants. St Paul, Minnesota: APS Press. p 87–100
———, Fisher PJ., 1988 A comparative study of fungal endophytes in xylem and whole stem of Pinus sylvestris and Fagus sylvatica Trans Br Mycol Soc 91:233-238
———, ———. 1990 Occurrence of fungal endophytes in twigs of Salix fragilis and Quercus robur Mycol Res 94:1077-1080
Polak AM., 1992 Major Timber Trees of Guyana. A Field Guide Wageningen, Netherlands: Tropenbos Foundation. 272 p
Rehner SA, Uecker FA., 1994 Nuclear ribosomal internal transcribed spacer phylogeny and host diversity in the coelomycete Phomopsis Can J Bot 72:1666-1674
Rodrigues KF., 1994 The foliar fungal endophytes of the Amazonian palm Euterpe oleracea Mycol 86:376-385
———, Petrini O., 1997 Biodiversity of endophytic fungi in tropical regions In: Hyde KD, ed. Biodiversity of Tropical Microfungi. Hong Kong: University of Hong Kong Press. p 57–69
———, Samuels GJ., 1999 Fungal endophytes of Spondias mombin leaves in Brazil Journal of Basic Microbiology 39:131-135
Valencia R, Balslev H, Paz y Miño G., 1994 High tree alpha-diversity in Amazonian Ecuador Biodiversity and Conservation 3:21-28
Wilson D., 2000 Ecology of woody plant endophytes In: Bacon CW, White JF, eds. Microbial Endophytes. New York & Basel: Marcel Dekker. p 389–420
———, Carroll GC., 1994 Infection studies of Discula quercina, an endophyte of Quercus garryana Mycologia 86:635-647
This article has been cited by other articles:
 |
|
 |
|
  A. E. Arnold and E. A. Herre Canopy cover and leaf age affect colonization by tropical fungal endophytes: Ecological pattern and process in Theobroma cacao (Malvaceae) Mycologia, May 1, 2003; 95(3): 388 - 398. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
