Role of fungi in freshwater ecosystems

Transcription

1 Biodiversity and Conservation 7, 1187±1206 (1998) Role of fungi in freshwater ecosystems MICHELLE K.M. WONG, TEIK-KHIANG GOH, I. JOHN HODGKISS, KEVIN D. HYDE, V. MALA RANGHOO, CLEMENT K.M. TSUI, WAI-HONG HO, WILSON S.W. WONG and TSZ-KIT YUEN Department of Ecology and Biodiversity, The University of Hong Kong, Pokfulam Road, Hong Kong Received 15 December 1997; accepted 28 January 1998 There are more than 600 species of freshwater fungi with a greater number known from temperate, as compared to tropical, regions. Three main groups can be considered which include Ingoldian fungi, aquatic ascomycetes and non-ingoldian hyphomycetes, chytrids and, oomycetes. The fungi occurring in lentic habitats mostly di er from those occurring in lotic habitats. Although there is no comprehensive work dealing with the biogeography of all groups of freshwater fungi, their distribution probably follows that of Ingoldian fungi, which are either cosmopolitan, restricted to pantemperate or pantropical regions, or in a few cases, have a restricted distribution. Freshwater fungi are thought to have evolved from terrestrial ancestors. Many species are clearly adapted to life in freshwater as their propagules have specialised aquatic dispersal abilities. Freshwater fungi are involved in the decay of wood and leafy material and also cause diseases of plants and animals. These areas are brie y reviewed. Gaps in our knowledge of freshwater fungi are discussed and areas in need of research are suggested. Keywords: biodiversity; biogeography; ecology; freshwater; fungi; nutrient cycling Introduction This paper reviews the biology of fungi in freshwater sediments. We use the term sediments in a broad sense to mean freshwater sand, gravel, silt, mud (Anon., 1989), wood, leaves and other organic matter that accumulates on the oor of freshwater habitats. We have not treated lichen-forming fungi on rocks in or by lake or stream margins, which merit a modern separate review. Biodiversity of freshwater fungi There are more than 600 species of freshwater fungi and a greater number are known from temperate, as compared to tropical, regions. These include ca 300 ascomycetes, 300 mitosporic fungi and a number of chytrids and oomycetes (Goh and Hyde, 1996). Three main groups can be considered: 1. The Ingoldian fungi which occur on decaying leaves in streams and lakes and which are probably the most well studied. They have been documented in many countries around the world, although the tropics have received less attention. 2. The aquatic ascomycetes and hyphomycetes occurring on submerged woody material have received less attention. Studies on these fungi in temperate regions are mainly based in North America, around Chesapeake Bay (Shearer, 1993a) and Ó 1998 Chapman & Hall

2 1188 Wong et al. Hong Kong (Hyde et al., unpubl.). Less intensive collections have been made in Australia, Brunei, England, Philippines, Seychelles and South Africa. 3. The chytrids and oomycetes, including those that cause diseases, are well-documented (Laidlaw, 1985; Fuller and Jaworski, 1987; Barr, 1988; Bruning, 1991; Powell, 1993). These fungi generally lack the ability to degrade cellulose, and are probably important in degrading noncellulosic entities entering the freshwater ecosystem (e.g. dead insects, keratin and pollen grains). Ingoldian fungi are relatively well studied, although inventories in numerous tropical countries are lacking. Studies in tropical countries have resulted in the discovery of many new species, e.g. the work carried out by Kuthubutheen and Nawawi in Malaysia. In the case of ascomycetes and hyphomycetes on woody material, other than the work of Hyde and co-workers in Asia and that of Shearer and co-workers in North America, there is little work being carried out. The task of documenting biodiversity is therefore problematical. Hyde et al. (1997) concluded that in any new freshwater system studied in new country, more than 50% of the ascomycetes identi ed may be new species. Although as more species are being described, this is no longer likely, we can conclude that the freshwater fungi are still relatively poorly studied. During studies of the fungi occurring on submerged woody material in the tropics over the last seven years, we have identi ed numerous ascomycetes and hyphomycetes that are new to science and have a large backlog of new species awaiting formal treatment. In one small river system, the Lam Tsuen River in Hong Kong, that we have now studied extensively, we have identi ed more than 200 fungi, which is more than the total number of non-ingoldian freshwater fungi presently known from the tropics. There is a great need for such studies to be implemented on a global scale. Habitats for freshwater fungi Freshwater habitats that support fungi can be divided into: (1) lentic (lakes, ponds, swamps, pools); and (2) lotic (rivers, streams, creeks, brooks) (Thomas, 1996). In addition, many freshwater fungi have been reported from arti cial habitats, such as water-cooling towers (Jones and Eaton, 1969; Eaton and Jones, 1970, 1971a,b; Udaiyan, 1989; Udaiyan and Hosagoudar, 1991). Lentic habitats Lentic habitats comprise any natural aquatic environment lacking a continuous ow of water. Lakes and ponds are typical lentic habitats (irrespective of their climatic zones) from which over 100 freshwater ascomycetes and their anamorphs are known (Shearer, 1993a). Lakes and ponds may provide calm environments in which fungi can develop (Sparrow, 1968). Low wave action may allow undisturbed fungal growth. However, the mild wave action may result in oxygen de ciency and reduce the number of fungi present (BaÈ rlocher, 1992). Almost one-third of freshwater ascomycetes are reported from lakes and/or ponds. They colonize both wood and leaves of various plant species and are widely distributed. One of the rst freshwater fungi from lentic habitats was described in the pioneering work of Weston (1929). He described an interesting discomycete, Loramyces juncicola occurring on fallen culms of Juncus spp. from a small pond in America. Other collections of

3 Role of fungi in freshwater ecosystems 1189 freshwater ascomycetes in lakes and ponds include those of Ingold (1951, 1954, 1955), Ingold and Chapman (1952), Tubaki (1966), Cavaliere (1975), Minoura and Muroi (1978), Magnes and Hafellner (1991) and Hyde and Goh (1998) and these articles should be consulted for further information. Discomycetes occur more commonly in temperate regions, especially in lentic habitats. Lotic habitats Lotic habitats comprise continuous owing water bodies, such as rivers, streams, creeks and brooks. Numerous species of fungi have been reported from this dynamic system (Shearer, 1993a; Thomas, 1996; Hyde et al., 1997). Almost one-third of the freshwater ascomycetes have been reported from rivers and/or streams. Temperate species were reported by Shearer and co-workers (Kjùller, 1960; Shearer, 1972, 1984, 1989a, b, 1993a, b; Willoughby and Archer 1973; Shearer and Crane, 1978a, b, 1980a, b; Webster and Descals, 1979; Shearer and von Bodman, 1983; Shearer and Zare-Maivan, 1988; Re vaâ y and GoÈ nczoè l, 1990) while tropical species were reported on by Hyde and co-workers (Goh, 1997; Hyde et al., 1997). Many of the tropical species found were new to science. Biogeography of freshwater fungi There is no comprehensive review dealing with the biogeography of all groups of freshwater fungi. The Ingoldian fungi are probably better known than the other freshwater fungi and have been reviewed by Wood-Eggenschwiler and BaÈ rlocher (1985). They conclude that: (1) Many species are cosmopolitan, although any single species may be more common in tropical or temperate regions. (2) Some species are restricted to temperate and cold, others to tropical and warm, regions. (3) Some species are restricted to very small geographical areas. Other freshwater fungi appear to follow similar distribution patterns. Origin of freshwater fungi Freshwater fungi are a diverse and heterogeneous group comprising species from di erent orders. The dominant groups are the ascomycetes and hyphomycetes, depending on geographical location and substrate. Shearer (1993a) stated that ``the presence of fungi in aquatic habitats alone may not be appropriate to de ne an ascomycete as a freshwater ascomycete''. This is because the occurrence of a species may simply be fortuitous and presence is therefore not conclusive evidence in assigning a particular fungus as ``freshwater''. The fungus could have its origin in terrestrial habitats and may have entered the freshwater system as spores. There are, however, numerous ascomycete (and hyphomycete) species which commonly occur in freshwater and have not been found in terrestrial habitats. These can con dently be categorized as freshwater fungi. Certain genera, e.g. Jahnula and Proboscispora, are con ned to freshwater habitats, while others have representatives in both terrestrial and marine habitats. Annulatascus has terrestrial (mainly on bamboo and palms) as well as freshwater representatives. Savoryella

4 1190 Wong et al. and Aniptodera has representatives in marine and freshwater habitats, while Ascotaiwania which was rst reported as a freshwater genus, is now known from terrestrial palms (Hyde, 1995). However, individual species are generally restricted to freshwater, marine or terrestrial habitats. The main di erences between species in a genus are found in the ascospores, with sheaths or appendages often occurring in freshwater and marine representatives, while other morphological characters vary little. It has been suggested that some marine fungi have a fungal-algal ancestor, which gave rise to ancestral pyrenomycetes, and which were mainly parasites of algae (Kohlmeyer and Kohlmeyer, 1979). Terrestrial loculoascomycetes and pyrenomycetes are thought to have originated from these parasitic pyrenomycetes. Subsequently these terrestrial ascomycetes moved back into the marine environment and are known as secondary marine fungi. These secondary marine fungi include the intertidal loculoascomycetes Halotthia, Leptosphaeria, Mycosphaerella, Pontoporeia, and pyrenomycetes such as Chaetosphaeria and Kallichroma. Kohlmeyer and Kohlmeyer (1979) organised marine fungi in two groups: (1) primary marine fungi (e.g. Ceriosporopsis, Corollospora, Halosphaeria) thought to have been derived from marine ancestors, that have not left their original marine environment; and (2) marine fungi which were thought to have evolved from terrestrial ancestors which have migrated back into the sea. An analogy can be made between freshwater and marine fungi. The genus Aniptodera (Shearer and Miller, 1977) may be classi ed as a primary freshwater fungus as species occur in both freshwater and marine environments and no terrestrial representatives have been found. Genera such as Annulatascus and Ascotaiwania may be classi ed as secondary freshwater fungi since they originate most probably from terrestrial habitats, since they have terrestrial representatives. Recent phylogenetic studies (Spatafora et al., 1995) have shown that many marine ascomycetes are likely to have evolved from terrestrial ancestors (e.g. Microascales), which have lost many of their characters. Features such as active ascospore ejection are thought to be unnecessary in the sea. It is probably also true, that most, if not all, freshwater ascomycetes evolved from terrestrial ancestors. Ecological adaptations of freshwater fungi Adaptations of freshwater Ascomycetes Many freshwater ascomycetes have ascospores with various sheaths, appendages or wall ornamentations, which probably function in ascospore dispersal and/or attachment. For example the base of the ascospores of Loramyces species have a long, tapering liform, caudate and sinuous appendage or `tail' which may be involved in the entrapment and attachment. In addition, the head of the ascospore is surrounded by a gel-like sticky mucilaginous sheath which may aid in adhesion to substrata (Weston, 1929; Ingold, 1954; Ingold and Chapman, 1952; Digby and Goos, 1987). ``Spores with sticky mucilaginous sheaths probably represent the most common type encountered in fungi'' (Jones, 1994). This appears to be true in freshwater ascomycetes. Shearer (1993a) listed ascomycete species possessing ascospores surrounded by mucilaginous sheaths which included: Nimbomollisia sp., and Obtectodiscus aquaticus (discomycetes); Annulatascus velatispora, and Fluviatispora spp. (pyrenomycetes); and Caryospora callicarpa, Kirschsteiniothelia elaterascus, Massarina spp., Phaeosphaeria spp., Pleospora scirpicola and Rebentischia sp. (loculoascomycetes). Some loculoascomycetes possess

5 Role of fungi in freshwater ecosystems 1191 mucilaginous appendages at their poles, e.g. Jahnula bipolaris, Lophiostoma spp., Rebentischia sp. and Wettesteinina niesslii. In addition, ascospores with unfurling appendages which uncoil in water to form long viscous threads, are also common in freshwater ascomycetes, e.g. Aniptodera spp., Annulutascus bipolaris and Halosarpheia spp. (Shearer and Crane, 1980b; Hyde, 1992). Species of Aniptodera and Halosarpheia also occur in the sea, and it appears that the morphological adaptations in marine and freshwater species in these genera are similar. Shearer (1993a) concluded that the appendage ontogeny in freshwater ascomycetes is simply mucilaginous rather than involving fragmentation and/or extension of a spore wall as in many marine fungi. However, Jones (1994) predicted that as our knowledge of a ascomycetes from freshwater habitat increases, ascospore appendages in marine fungi may not prove to be unique. Several unique appendage types have now been shown to exist in tropical freshwater ascomycetes (Wong and Hyde, 1998). Asci of aquatic ascomycetes also appear to be adapted for dispersal in these habitats. Deliquescent asci are common in the marine Halosphaeriales and in some freshwater species in the genera Halosarpheia and Nais. The nature of deliquescing asci allows liberation of ascospores without forcible discharge. The ascospores then accumulate around the break or tip of the neck, and are dispersed by water movement. Shearer (1993a) suggested that the disappearance of the apical apparatus in the asci might be an adaptation to life in water. In several freshwater ascomycetes, however, large and refractive apical rings (to 7±8 lm long and 4±5 lm wide) have been reported in several common tropical genera, e.g. Annulatascus, Ascotaiwania and Submersisphaeria. Similarly, several freshwater loculoascomycete genera have species with well-developed apical discharge mechanisms in their asci (e.g. Jahnula spp., Massarina spp.). In Kirschsteiniothelia elaterascus (Shearer, 1993b) the asci are ssitunicate, but a long and narrow posterior end of the endoascus becomes coiled within the ectoascus. The ectoascus ruptures in water and the endoascus elongates to a length about 1.7 times of its original length (Shearer, 1993b) by the uncoiling of the basal structure. The posterior end of the endoascus remains connected to the endoascus or may be completely liberated. The need to eject ascospores therefore appears to be important in freshwater. In Ophioceras and Pseudohalonectria species (Shearer, 1989a) asci are liberated from the ascomata into the surrounding water before the ascospores are probably ejected some distance from the ascomata, but the advantages of this are unknown. Adaptations of the anamorphs Spore attachment and entrapment mechanisms in Ingoldian hyphomycetes, which commonly have sigmoid or tetraradiate spores, are well-documented (Ingold, 1953, 1956, 1966, 1975, 1984; Webster, 1981; Read, 1990). Sigmoid conidia often become attached at their sticky poles, and then straighten in the direction of the water current so they are less likely to be washed away. Tetraradiate and branched conidia act as an anchor and allow their entrapment to the substrata or in surface foam (Ingold, 1942, 1953). Tetraradiate conidia can also attach to the substratum, with three ``legs'' forming a strongly adhesive tripod (Webster and Descals, 1981). Adhesive mucilaginous material is also produced at each arm of the conidium in contact with a surface (Read, 1990; Read et al., 1992) and attaches them rmly to the substratum. Read (1990) indicated that substrate colonization by Ingoldian hyphomycete may involve four factors: conidial entrapment and attachment; rapid germination; mucilage production on the spore surface and germ tubes; and the

6 1192 Wong et al. formation of appressoria. These characters may confer an advantage for the Ingoldian hyphomycetes over other fungi. Functional role of freshwater fungi The main role of the freshwater ascomycete, basidiomycetes and mitosporic fungi in freshwater ecosystems is in the degradation of dead plant material (e.g. Juncus, leaves and wood) that nds its way into the water (Goh and Hyde, 1996). They may also be involved in the degradation of animal parts such as insect exoskeletons, sh scales, and hair. Other ecological groups present are the plant pathogens and endophytes that may colonise living plant tissues. The decay of dead plant tissues is a result of the fungi's ability to degrade celluloses and lignocelluloses (Zare-Maivan and Shearer, 1988a, b). Their success in degrading woody tissues lies in an ability to form soft-rot cavities (Shearer, 1993a; Yuen et al., pers. obs.). Basidiomycetes are rare and mainly absent in freshwater as they are not soft-rotters, although they can degrade cellulose. It appears that the ability to degrade the lignocellulose from within the S 2 layer of the cell wall is important in submerged waterlogged wood. Several species have now been tested for their ability to cause soft-rot decay and although we have information for only a small proportion of known species, it is probably representative. Fungi on submerged wood Knowledge of tropical fungi on submerged wood The role of fungi, including freshwater fungi, in wood decay has been extensively examined in cultural studies (Shigo, 1965, 1972; Shearer and von Bodman, 1983; Benner et al., 1986; Zare-Maivan and Shearer, 1988a). However, no similar work has been carried out using tropical freshwater fungi. The growth parameters and woody plant species that these tropical fungi colonise may be di erent to similar fungi in temperate regions. Although, many new fungi have been found on submerged wood in tropical regions (Nawawi, 1985; Sridhar et al., 1992; Goh, 1997; Hyde et al., 1997), there has been little work carried out on the roles of these fungi in nutrient cycling or in the degradation of lignocellulose. Principal factors of the decay process Woody tissues are distinguished from other plant tissues by their high lignocellulose and low nitrogen content. The principal components of wood are cellulose, hemicellulose, and lignin. These components are degraded by di erent organisms to various extents. Only a limited group of fungi possess enzymatic capabilities to digest wood (Tubaki, 1958; Singh, 1982; Zare-Maivan and Shearer, 1988a; Abdullah and Taj-Aldeen, 1989), such as cellulase and lignin degrading enzymes. In the development of the decay process, colonisation, growth and survival of the organism in wood are important. This depends on substrate conditions, which involve the presence of a suitable substrate, a suitable temperature and also interaction between microorganisms. The growth rates are a ected by temperature and substrate, and re ect the di erences in colonization ability among the component species (Ogawa et al., 1996). It is interesting to note that the optimum temperature for growth of most tropical freshwater fungi is between 20±25 C (Yuen et al., 1998), although several isolates exhibit optimal growth at temperatures as low as 15 C or as high as 30 C. These results are similar

7 Role of fungi in freshwater ecosystems 1193 to those reported by Zare-Maivan and Shearer (1988a) and Koske and Duncan (1974) for temperate species. They also showed that the optimal temperature for growth of most temperate aquatic fungi was 25 C, but that they could also grow relatively well at temperatures as low as 10 C. Tropical freshwater fungi do not grow well in low temperatures, and so are absent in streams in temperate regions. Although temperate species grow best at 25 C, they are not able to grow as rapidly as tropical species and this probably accounts for their absence in tropical streams (Yuen et al., 1998). Interaction is an important factor in determining the organization, composition, and pattern of fungal colonization within freshwater ecosystems (Shearer, 1993a). The antagonistic activities of freshwater fungi on agar media have been investigated and it was found that some fungi can inhibit the growth of others by producing antibiotic substances (Khan, 1987; Zare-Maivan and Shearer, 1988a, b; Asthana and Shearer, 1990). Zare- Maivan and Shearer (1988a, b) showed that persistent and late colonizing fungi on longlasting substrata (e.g. wood) are more likely to produce antagonistic substances than those on less persistent substrata (e.g. leaves). Extent of decay The extent of decay can be expressed by loss in weight, number of soft-rot cavities and reduction in crushing strength. Investigations so far carried out indicate that freshwater fungi, with the exception of early succession species, cause signi cant losses in weight (Jones, 1981; Zare-Maivan and Shearer, 1988b). Signi cant weight loss and reduction in strength result from the formation of soft-rot cavities within the S 2 layer of the cell wall (Wilcox, 1978). Fungi on submerged wood in streams and water-saturated wood in cooling towers are also able to degrade wood and produce soft-rot cavities on wood test blocks in the laboratory (Eaton, 1976; Leightley and Eaton, 1977). Formation of cavities is closely related to hyphal growth, and is explained by the limited di usion of enzymes away from hyphal surfaces (Hale and Eaton, 1985). However, some of the fungi cannot form soft-rot cavities in ligni ed cell walls and it is thought that these fungi can only use dead cell contents and the walls of parenchymatous ray cells (Mouzouras, 1986). Fungi on leaves Relative importance of fungi in leaf decay Aquatic hyphomycetes are regarded as the dominant mycobiota associated with decaying leaves in streams (BaÈ rlocher and Kendrick, 1974; Butler and Suberkropp, 1986), although other fungal taxa have also been isolated from submerged decaying leaves (Kaushik and Hynes, 1968; Godfrey, 1983). Zoosporic fungi and terrestrial fungi also occur on leaves that are recovered from streams in addition to aquatic hyphomycetes. Oomycetes are generally early colonizers, but they decline rapidly (BaÈ rlocher and Kendrick, 1974; BaÈ rlocher, 1990). Terrestrial fungi (genera such as Alternaria, Aureobasidium, Cladosporium) that colonize the phyllosphere of senescent leaves often persist, but the aquatic hyphomycetes usually outgrow these terrestrial fungi (BaÈ rlocher and Kendrick, 1974; Suberkropp and Krug, 1980). It has also been suggested that the predominance of aquatic hyphomycetes on submerged leaves is based on their ability to remain active at low temperatures (BaÈ rlocher and Kendrick, 1974; Godfrey, 1983). There are some experimental data that suggest that di erences in temperature alone do not explain such

8 1194 Wong et al. di erences in the ability to grow and degrade submerged leaves (Graca and Ferreira, 1995). Such predominance is due to the ability of the aquatic hyphomycetes to degrade submerged organic matter through a wide range of climatic conditions, and that terrestrial fungi are unable to macerate leaf material when submerged (Graca and Ferreira, 1995). Moreover, the tetraradiate or sigmoid conidia produced on submerged substrata provide aquatic hyphomycetes with an additional colonisation advantage over terrestrial fungi, which possess spores more suited to air dispersal (Webster and Descals, 1981). Production of degrading enzymes in freshwater ecosystem Aquatic fungi have been shown to produce a rich array of enzymes able to degrade the major leaf polysaccharides (Suberkropp and Krug, 1980). These enzymes are able to degrade simple sugars, cellulose, and other plant polymers (Tubaki, 1957; Thornton, 1963, 1965; Nilsson, 1964; Singh, 1982; Chandrashekar and Kaveriappa, 1988), and lead to skeletonization of leaves through maceration. Experimental evidence shows that aquatic hyphomycetes have a ph-dependent degrading activity toward cellulose, xylan, and pectin (Chamier and Dixon, 1982; Suberkropp et al., 1983; Chandrashekar and Kaveriappa, 1988). There is at least circumstantial evidence that some species can attack lignin and cause soft rot (Jones, 1981; Fisher et al., 1983; Zemek et al., 1985). Nitrogen content as a percentage of remaining leaf mass typically increases during decay (BaÈ rlocher, 1985; Webster and Ben eld,1986). This is partly due to complex formation between leaf phenolics/lignins and proteins or other nitrogenous compounds, and partly due to the accumulation of microbial cells (Raviraja et al., 1996). Generally, it is assumed that pronounced nitrogen increases indicate higher fungal activity. It has been shown that the decay rates of leaves and fungal biomass are correlated (Maharning and BaÈ rlocher, 1996). A possible explanation is that increasing fungal colonization not only results in higher exoenzymatic activity, which contributes directly to mass loss and but eventually higher invertebrate feeding, but also increases susceptibility to mechanical fragmentation (i.e. indirect e ects) of the leaf (Raviraja et al., 1996). Fungi on grasses We are unaware of any work that has been carried out on the role of fungi on grasses in freshwater habitats. There have, however, been several studies on the role of fungi in the nutrient cycling of standing grasses, in salt marshes and estuaries with brackish water, e.g. Spartina alterni ora, Juncus sp. or Phragmites australis. Most litter and wood-decomposing fungi, which have been tested, cannot grow at a water potential (w) lower than )6 Mpa (Newell et al., 1991). Thus, freshwater habitats undoubtedly provide suitable physical environments for an ecologically specialized group of fungi. Gallagher and Pfei er (1977) discovered that there is a metabolically active `dead-plant community' of bacteria and/or fungi on submerged dead standing Spartina alterni ora leaves and leaf sheathing bases. Padgett et al. (1985) showed that if both bacterial and fungal carbonconversion e ciencies were equal to 50%, then P70% of the carbon released as CO 2 would be due to fungal respiratory activity. Newell et al. (1989) also provided evidence that nearly all dead-leaf nitrogen was scavenged into fungal mass after the rst sampling interval. Flux estimates for dead-leaf carbon indicated a ow of 11±15% of the original to fungal biomass, 2% to bacterial biomass, 15±21% to carbondioxide, 10±12% to dissolved leachate, and 34±36% to small particles; while 32±39% remained attached as shreds at the

9 Role of fungi in freshwater ecosystems 1195 end of the study period. Fungi therefore play an important role in the degradation of standing dead grass culms in brackish habitats. A variety of aquatic macrophytes including Juncus, Phragmites, Scirpus and Typha, serve as substrata for freshwater fungi. Magnes and Hafellner (1991) collected ascomycetes on emergent plants from alpine lakes to determine relationships. Among the 52 fungal species collected, one large group was considered unspecialized, occurring on a variety of plants hosts. Seventeen species were found to be substrate speci c, but were not considered to be parasitic. There are several studies on the freshwater fungi occurring on living macrophytes, which provide further information (Ingold and Chapman, 1952; Ingold, 1955; Webster and Lucas, 1961; Apinis et al., 1972a, b; Nannfeldt, 1985). The succession of fungi on emergent macrophytes has also been investigated (Pugh and Mulder, 1971; Apinis et al., 1972a, b; Taligoola et al., 1972). During the early stages of leaf emergence in Typha latifolia, the mycobiota was dominated by species of yeasts and dematiaceous hyphomycetes. After the leaves died, species with ssitunicate asci (e.g. Leptosphaeria, Nodulosphaeria, Paraphaeosphaeria and Phaeosphaeria) became prominent (Pugh and Mulder, 1971). Ascomycetes were present in early (Massarina and Wettsteinina) and late (Lasiosphaeria, Ophiobolus and Passeriniella) stages of succession on submerged culms of Phragmites australis (Apinis et al., 1972b). Fungi in muddy sediments As far as the role and occurrence of fungi in muddy sediments is concerned we know very little. If sediments are extracted and plated onto agar, then a range of fungal species are isolated. However, we have no knowledge if these fungi are functional in muddy sediments or if the isolates are from dormant fungal spores. Pathogens The majority of freshwater fungi have been reported as saprotrophs on dead plant substrata (Shearer, 1993a; Goh and Hyde, 1996; Thomas, 1996), but whether this re ects a real dominance of saprotrophs over parasites, or is due to collector bias, is unknown. Certainly, in contrast to the well-studied Ingoldian fungi or less well documented aquatic ascomycetes, there have been few studies on the pathogens of aquatic organisms, and there is no comprehensive account of them. The roles of these fungi have been brie y discussed in review by Shearer (1993a), Goh and Hyde (1996) and Thomas (1996). The extent of mycoparasitism in freshwater ecosystems is unknown. Several hyphomycete genera, chytrids and oomycetes associated with other fungi, have been reported from freshwater ecosystems (e.g. Crucella, Janetia, Nectria), but interactions between fungi are poorly documented (Thomas, 1996). Janetia curviapicis was described as ``associated with or growing on other hyphomycetes on submerged wood'' from a stream in north Queensland, Australia (Goh and Hyde, 1996), while the Crucella subtilis anamorph of Camptobasidium hydrophilum is apparently a mycoparasite of several aquatic hyphomycetes (Marvanova and Suberkropp, 1990). Nectria species are also regularly observed developing on old fruiting bodies of various ascomycetes on submerged wood, and these may be mycoparasites (Hyde, pers. obs.). Parasitism of plants and animals in water have been documented in Australia (Thomas, 1996). Aquatic plant pathogens have representatives in the oomycete, chytridiomycete,

10 1196 Wong et al. ascomycete, basidiomycete, and mitosporic fungi, and are commonplace. Most of these fungi cause diseases of the aerial parts of aquatic plants. Rusts (basidiomycetes) are often associated with the leaves of aquatic macrophytes, while smuts often develop in aerial in orescences (Thomas, 1996). Species of Phyllachora (ascomycetes) are often present as tar spots on the aerial leaves of aquatic grasses (Hyde, pers. obs), and the popular Asian leafy water spinach (Kang Kong, Ipomea aquatica) is often parasitised by the oomycete Albugo ipomoeae-aquaticae (Shivas et al., 1996). Infection of submerged plant parts is less well documented, however, the staple Taro (Colocasia esculenta) which grows along river banks, may be severely a ected by Taro Blight. This blight is caused by Phytophthora colocasiae (oomycete) which may cause devastating losses, as the underground rhizomes become soft and are attacked by a watery rot (Hyde et al., 1991). Semisubmerged leaves of water lilies may also be infected by the chytridiomycete Physoderma limnanthemi (Thomas, 1996). Infection of aquatic animals by fungi is better documented (Thomas, 1996). Achlya, Aphanomyces, Pythium and Saprolegnia species (oomycetes) are regularly reported growing in association with diseased sh (Thomas, 1996), and can also parasitise captive aquatic fauna such as turtles, tadpoles and sh. In many cases these infections result as the organisms are under stress from environmental conditions. The role of oomycetes as possible biocontrol agents of mosquitoes has also received attention. Several oomycetes, trichomycetes and mitosporic fungi are pathogens of mosquitoes or other aquatic arthropods and their potential for biocontrol is discussed by Thomas (1996). Mucor amphibiorum is an interesting fungal/animal association noted in Australia, as it is parasitic on Platypus spp. in Tasmania, and Cane Toads in Queensland and the Northern Territory (Thomas, 1996). Linkages to other organisms Plant material produced in the riparian zone has been regarded as a major energy source for low order streams (Vannote et al., 1980). Such material consists mainly of leaves and, to a lesser extent twigs, barks, seeds and owers. Senescent leaves are rich in high-energy structural compounds, but this energy is not easily accessible to aquatic detritivores. There is evidence that aquatic fungi can macerate the leaf matrix and make the energy available to feeding animals in freshwater habitats (Kaushik and Hynes, 1971; Suberkropp and Krug, 1980; Singh, 1982; Suberkropp et al., 1983). There is considerable experimental evidence that detritivores selectively feed on conditioned leaves, i.e. those previously colonized by fungi (Suberkropp, 1992; Graca, 1993). Fungi can alter the food quality and palatability of leaf detritus, a ecting shredder growth rates. Animals that feed on a diet rich in fungi have higher growth rates and fecundity than those fed on poorly colonized leaves (Graca et al., 1993). Interactions between invertebrates and aquatic hyphomycetes have been demonstrated (Rossi, 1985; Graca et al., 1993). Some shredders prefer to feed on leaves that are colonized by fungi, whereas others consume fungal mycelium selectively (Graca et al., 1993). At the same time, the palatability of detritus is a ected by a number of factors, including leaf softness, nitrogen content and fungal biomass (Suberkropp, 1992). The situation with wood is less well known, although mycophagy of lignicolous fungi has been observed (Tsui et al., pers. obs.). Woody substrata are more bulky and hard to decompose and consume, and therefore little data are available on the relationships between other wood inhabitants and fungi.

11 Role of fungi in freshwater ecosystems 1197 Human disturbance to aquatic fungal communities Most freshwater habitats are vulnerable to disturbance. Any perturbation within the drainage basin will a ect in-stream communities through wash-o or run-o processes and, owing to the downhill ow of water, any change in headstream areas will alter the downstream reaches (Dudgeon, 1992). Human in uence, such as the discharge of industrial and domestic e uents, indiscriminate use of pesticides and fertilizers, and clearance of riparian vegetation, all have deleterious e ects on the fungal community and are discussed by BaÈ rlocher (1992). Pollution With the exception of Ingoldian fungi, there is no information on the e ect of disturbance on freshwater fungi. Bermingham (1996) has reviewed the e ects of pollution on Ingoldian fungi. Because macroinvertebrates (shredders) have feeding preferences for leaf material colonized by a particular species of fungi, any perturbation of the fungal community could directly a ect the rate of incorporation of leaf material into the detrital food web. The e ects of organic pollution on aquatic fungi are apparent, and this subject will not be treated in detail as they are reviewed elsewhere (Cooke, 1976; van der Merve and Jooste, 1988; Au et al., 1992a, b). In general, the diversity of organisms, usually aquatic hyphomycetes and occasionally sediment fungi, in polluted and unpolluted freshwater systems are compared. The results indicate that species richness and conidial production is higher in unpolluted streams, whereas fungi that are the inhabitants of organic substrata (e.g. Cercophora spp.) are more common in polluted streams. The fungal diversity and rates of the decomposition of leaf baits was also suppressed in polluted streams (Au et al., 1992b). Human activities often increase the concentration of toxic metals in streams, and studies have reviewed the sensitivity of fungi to certain chemicals (Duddridge and Wainwright, 1980; Abel and BaÈ rlocher, 1984). Both Pb 2+ and Cd 2+ have been shown to inhibit the growth of the fungi studied, with Cd 2+ a ecting aquatic hyphomycetes (Abel and BaÈ rlocher, 1984). Apart from industrial and domestic wastes, agrochemicals and fertilizers from farms may leach into streams during heavy monsoonal rains and may a ect the fungal community. Pesticides and herbicides, such as PCP, paraquat and DDT, inhibit the growth and sporulation of fungi at low concentrations (Dalton et al., 1970; BaÈ rlocher and Premdas, 1988; Chandrashekar and Kaveriappa, 1989). However, most of these studies were carried out in temperate regions, and whether this impact of pollution is paralleled in tropical communities is unknown. The e ect of acid mine drainage and ph on temperate Ingoldian fungi has been studied. Low ph (<5.5) results in lower biodiversity, which corresponds to a lower rate of leaf decomposition (Chamier, 1987). Mine drainage, acidic and with heavy metals, has also been shown to reduce biodiversity, with the aquatic hyphomycete community being the most sensitive of the groups of fungi investigated (Maltby and Booth, 1991). It therefore seems that Ingoldian fungi are sensitive to low ph and metals, but few studies have been carried out. Deforestation and stream regulation As mentioned before, allochthonous organic matter accounts for a major energy input in stream ecosystems, as they are vital for providing substrates and in sustaining the aquatic

12 1198 Wong et al. fungal community (Shearer, 1993a). Removal of the vegetation along the streams due to road building or stream regulation will a ect the fungal community. Metwalli and Shearer (1989) found that the mean number of conidia per litre and the degree of colonization of leaf discs found in streams in wooded areas was greater than those in streams with clearcut riparian vegetation. The richness of riparian vegetation was also positively correlated with fungal species richness (Fabre, 1996). Major gaps in knowledge in the functional role of fungi in freshwater ecosystems There are numerous gaps in our knowledge of the biodiversity, ecology and role of freshwater fungi. Although the Ingoldian fungi are relatively well documented, other groups, such as ascomycetes on submerged wood and fungi in muddy sediments are poorly studied. This is particularly true of tropical regions where few studies have been carried out. Dematiaceous hyphomycetes are regularly collected on submerged wood (e.g. Acrogenospora spp., Beltrania rhombica, Sporoschisma spp.). However, we are not sure whether these fungi originate from freshwater or terrestrial habitats. Studies need to be implemented to establish the requirements or adaptations for a species to be truly aquatic. Very little work has also been carried out on interactions between freshwater fungi. Our knowledge of whether fungi can compete for substrata by inhibiting the growth of competitors is rudimentary. We also lack knowledge of the importance of fungi relative to other freshwater organisms in nutrient cycling. Fungi appear to be the dominant organisms involved in the decay of leaves, wood and probably dead animals and animal parts, but we are unclear of their role in the decay of particulate detritus in sediments. It is also unclear if single species can degrade tissues or if a suite of fungi complement each other in the decay of organic matter. Several studies have shown that fungi alter dead submerged leaves in such a way that they become palatable to small invertebrates, and also that some of these invertebrates can feed directly on fungal mycelium. However, the extent to which this occurs in nature is speculative, and no such studies have been carried out in the tropics. Although we have some idea of which fungi occur in freshwater ecosystems, we have little idea how biodiversity is linked to ecosystem functioning. Apart from the e ects of pollution on Ingoldian fungi and its e ect on leaf palatability, we are unaware of any other experiments attempting to link biodiversity with ecosystem functioning. Some metals are toxic to fungi (Abel and BaÈ rlocher, 1984). However, the causal relationship between the speci c metal and their e ects to natural communities are not well established (Bermingham, 1996). In addition to the fact that a species may be present in both polluted and unpolluted habitats, the question as to whether the fungus is a ected by pollutants is unknown. Ergosterol has been recognized as a marker to quantify fungal biomass (Newell, 1992), but the method cannot be used to determine the biomass of individual fungal species. The e ect of elevated carbon dioxide, rainfall and land use has not been investigated. These various disturbances may a ect species composition, and losses of keystone species (Box 1) may even occur. Methods available to answer these problems Most work on freshwater fungi has been of either the inventory type, or physiological studies that have been carried out on arti cial media or wood blocks and have largely concentrated on the early stages of decay. It is important that inventory is carried out

13 Role of fungi in freshwater ecosystems 1199 globally and in as many types of freshwater habitats as possible, as information is lacking for most countries and aquatic ecosystems. Although in vitro studies are important, the ndings of studies on arti cial media, need to be tested in vivo, in order to establish how fungi interact on submerged material. Less is known of the extent to which competitive interactions occur in fungal communities and of the extent of degradation of wood that occurs in nature. An attempt must be made to ascertain what events may occur under natural conditions. Wood blocks can be inoculated with fungi and then placed in the natural environment in order to establish competitiveness against the inhabitant mycobiota. Preliminary eld studies placing uncolonized and precolonized wood blocks in streams have been carried out (Yuen et al., pers. obs.). Precolonized wood blocks were inoculated with one or two isolates of tropical fungi, while no fungi were inoculated onto the uncolonized controls. After three months, wood blocks were collected and the number of fungi, the rate of decay and interaction activities between di erent fungal colonies were established. More fungi were found on uncolonized wood blocks, presumably as there is less competition compared to precolonized wood blocks. Other in vitro experiments can also be devised in order to study these competitive interactions. By developing an experimental system in which individual fungi and mixtures of fungi are inoculated onto sterile wood blocks or leaf samples, the functional role of individual and groups of fungi can be established. It is thought that some fungi may pre-condition the wood so as to allow settlement and colonization of the wood by other organisms. The e ect of prior infection of wood by a fungus, or in combination with di erent types of microorganisms, should be investigated. As in the case of leaves, the conditioned leaves are consumed in preference to nonconditioned leaves by invertebrates (Arsu and Suberkropp, 1985), and it is known that leaves colonized by certain fungal species are more attractive than others (Suberkropp, 1992). This also needs to be investigated in the tropics and with woody material. As wood is very resistant to decay, it takes a long time for the decay process to be completed and for the nutrients to be released and recycled. The later stages of decay can be traced by placing wood blocks in streams for a longer time before collection. Although we can isolate fungi from sediments, we cannot be sure that these fungi are involved in the nutrient cycling of particulate matter. Methods are needed in which we can measure fungal and not other organism activity. One possible method could be to measure ergosterol activity, but this method has been criticized. Alternatively, it may be possible to develop a way to measure fungal biomass in a sediment or water column. The relative amounts of fungi and bacteria could also be investigated in this way. Further experiments with feeding detritivores should also be carried out. This should include mixtures of fungi, and the role of tropical fungi particularly is in need of investigation. Acknowledgements We thank The University of Hong Kong for the provision of University Studentships, Postdoctoral Fellowships, and Part-Time Demonstratorships to allow this work to be possible. References Abdullah, S.K. and Taj-Aldeen, S.J. (1989) Extracellular enzymatic activity of aquatic and aeroaquatic conidial fungi. Hydrobiologia 174, 217±23.

14 1200 Wong et al. Abel, T.H. and BaÈ rlocher, F. (1984) E ects of cadmium on aquatic hyphomycetes. Appl.. Environ. Microbiol. 48, 245±51. Anon (1989) McGraw-Hill Dictionary of Scienti c and Technical Terms. 4 th Edition. USA: Mcgraw Hill. Apinis, A.E., Chesters, C.G.C. and Taligoola, H.K. (1972a) Colonisation of Phragmites communis leaves by fungi. Nova Hedwigia 23, 113±24. Apinis, A.E., Chesters, C.G.C. and Taligoola, H.K. (1972b) Microfungi colonizing submerged standing culms of Phragmites communis Trin. Nova Hedwigia 23, 473±80. Arsu, T.L. and Suberkropp, K. (1985) Selective feeding by stream caddis y (Trichoptera) detritivores on leaves with fungal colonized patches. Oikos 45, 50±8. Asthana, A. and Shearer, C.A. (1990) Antagonistic activity of Pseudohalonectria and Ophioceras. Mycologia 82, 554±61. Au, D.W.T., Vrijmoed, L.L.P. and Hodgkiss, I.J. (1992a) Fungi and cellulolytic activity associated with decomposition of Bauhinia purpurea leaf litter in a polluted and unpolluted Hong Kong waterway. Can. J. Bot. 70, 1071±9. Au, D.W.T., Vrijmoed, L.L.P. and Hodgkiss, I.J. (1992b) Decomposition of Bauhinia purpurea leaf litter in a polluted and unpolluted Hong Kong waterway. Can J. Bot. 70, 1061±70. BaÈ rlocher, F. (1985) The role of fungi in the nutrition of stream invertebrates. Bot. J. Linn. Soc. 91, 83±94. BaÈ rlocher F. (1990) Factors that delay colonization of freshwater alder leaves by aquatic hyphomycetes. Arch. Hydrobiol. 119, 249±55. BaÈ rlocher, F. (1992) The Ecology of Aquatic Hyphomycetes. [Ecological Studies No. 94] Berlin: Springer-Verlag. BaÈ rlocher, F. and Kendrick, B. (1974) Dynamics of the fungal populations on leaves in a stream. J. Ecol. 62, 761±91. BaÈ rlocher, F. and Premdas, P.D. (1988) E ects of pentachlorophenol on aquatic hyphomycetes. Mycologia 80, 135±7. Barr, D.J.S. (1988) Zoosporic plant parasites as fungal vectors of viruses: taxonomy and life cycles of species involved. In Viruses with Fungal Vectors (J.I. Copper and M.J.C. Asher, eds) pp United Kingdom: Lavenham Press. Benner, R., Moran, M.A. and Hodson, R.E. (1986) Biogeochemical cycling of lignocellulosic carbon in marine and freshwater ecosystems: relative contributions of prokaryotes and eukaryotes. Limn. Oceanol. 31, 89±100. Bermingham, S. (1996) E ects of pollutants on aquatic hyphomycetes colonizing leaf material in freshwaters. In Fungi and Environmental Change (J.C. Frankland, N. Magan, and G.M. Gadd, eds) pp Cambridge: Cambridge University Press. Bruning, K. (1991) E ects of phosphorus limitation on the epidemiology of a chytrid phytoplankton parasite. Freshw. Biol. 25, 409±17. Butler, S. and Suberkropp, K. (1986) Aquatic hyphomycetes on oak leaves: comparison of growth, degradation and palatability. Mycologia 78, 922±8. Cavaliere, A.R. (1975) Aquatic ascomycetes from Lake Itasca, Minnesota. J. Minnesota Acad. Sci. 44, 32±5. Chamier, A.C. (1987) E ect of ph on microbial degradation of leaf litter in seven streams of the English Lake district. Oecologia 71, 491±500. Chamier, A.C. and Dixon, P.A. (1982) Pectinases in leaf degradation by aquatic hyphomycetes: the enzymes and leaf maceration. J. Gen. Microbiol. 128, 2469±83. Chandrashekar, K.R. and Kaveriappa, K.M. (1988) Production of extracellular enzymes by aquatic hyphomycetes. Folia Microbiol. 33, 55±8. Chandrashekar, K.R. and Kaveriappa, K.M. (1989) E ects of pesticides on the growth of aquatic hyphomycetes. Toxicology letters 48, 311±5.

15 Role of fungi in freshwater ecosystems 1201 Cooke, W.B. (1976) Fungi in sewage. In Recent Advances in Aquatic Mycology (E.B.G. Jones, ed), pp. 389±434. London: Elek Science. Dalton S.A., Hodkinson, M. and Smith, K.A. (1970) Interaction between DDT and river fungi. 1. E ects of p,p -DDT on the growth of aquatic hyphomycetes. Appl. Microbiol. 20, 662±6. Digby, S. and Goos, R.D. (1987) Morphology, development and taxonomy of Loramyces. Mycologia 79, 821±31. Duddridge, J.E. and Wainwright, M. (1980) Heavy metal accumulation by some aquatic fungi and reduction in viability of Gammarus pulex fed Cd 2+ contaminated mycelium. Water Research 14, 1605±11. Dudgeon, D. (1992) Patterns and Processes in Stream Ecology: A Synoptic Review of Hong Kong Running Waters. Stuttgart: Schweizerbart Verlagsbuchhandlung. Eaton, R.A. (1976) Cooling tower fungi. In Recent Advances in Aquatic Mycology (E.B.G. Jones, ed), pp London: Elek Science. Eaton, R.A. and Jones, E.B.G. (1970) New fungi on timber from water-cooling towers. Nova Hedwigia 19, 779±86. Eaton, R.A. and Jones, E.B.G. (1971a) The biodeterioration of timber in water cooling towers I. Fungal ecology and decay of wood at Connah's Quay and Ince. Material und Organismen 6, 51± 80. Eaton, R.A. and Jones, E.B.G. (1971b) The biodeterioration of timber in water cooling towers II. Fungi growing on wood in di erent positions in a water cooling system. Material und Organismen 6, 81±92. Fabre, E. (1996) Relationships between aquatic hyphomycetes communities and riparian vegetation in 3 Pyrenean streams. C.R. Acak. Sci. Paris, Sciences de la vie/life sciences 319, 107±11. Fisher, P.J., Davey, R.A. and Webster, J. (1983) Degradation of lignin by aquatic and aero-aquatic hyphomycetes. Trans. Brit. Mycol. Soc. 80, 166±8. Fuller, M.S. and Jaworski, A. (1987) Zoosporic Fungi in Teaching and Research. Athens: Southeastern Publishing Corporation. Gallagher, J.L. and Pfei er, W.J. (1977) Aquatic metabolism of the communities associated with attached dead shoots of salt marsh plants. Limnol. Oceanogr. 22, 562±4. Godfrey, B.E.S. (1983) Growth of two terrestrial microfungi on submerged alder leaves. Trans. Brit. Mycol. Soc. 79, 418±21. Goh, T.K. (1997) Tropical freshwater hyphomycetes. In Biodiversity of Tropical Microfungi (K.D. Hyde, ed.), pp.189±228. Hong Kong: Hong Kong University Press. Goh, T.K. and Hyde, K.D. (1996) Biodiversity of freshwater fungi. Indust. Microbiol. 17, 328±45. Graca, M.A.S. (1993) Patterns and processes in detritus-based stream systems. Limnologica 23, 107± 14. Graca, M.A.S. and Ferreira, C.F. (1995) The ability of selected aquatic hyphomycetes and terrestrial fungi to decompose leaves in freshwater. Sydowia 47, 167±79. Graca, M.A.S., Maltby, L. and Calow, P. (1993) Importance of fungi in the diet of Gammarus pulex and Asellus aquaticus II. E ects on growth, reproduction and physiology. Oecologia 96, 304±9. Hale, M.D. and Eaton, R.A. (1985) Oscillatory growth of fungal hyphae in wood cell walls. Trans. Brit. Mycol. Soc. 84, 277±88. Hyde, K.D. (1992) Tropical Australian freshwater fungi II. Annulatascus velatispora gen. et sp. nov., Annulatascus bipolaris sp. nov. and Nais aquatica sp. nov. Aust. Syst. Bot. 5, 117±24. Hyde, K.D. (1995) Fungi from palms. XXII. A new species of Ascotaiwania. Sydowia 47, 199±212. Hyde, K.D. and Goh, T.K. (1997) Fungi on submerged wood in a small stream on Mt Lewis, North Queensland, Australia. Muelleria 10: 145±57. Hyde, K.D. and Goh, T.K. (1998) Fungi on submerged wood in Lake Barrine, North Queensland. Mycol. Res. 102, 739±749.