The ecology of aquatic hyphomycetes in an Australian upland stream

Abstract

The principal aim of this study was to advance the understanding of aquatic hyphomycete fungi in streams of Australia, where previously very little work had been carried out. To achieve this aim, several projects were carried out in one selected stream ecosystem (Lees Creek; ACT, S.E. Australia). These projects included: an examination of the litter resource of aquatic hyphomycetes by measuring litter inputs to the stream and litter decomposition in the stream; determinations of the patterns of fungal occurrence on the litter; an attempt to manipulate patterns of fungal occurrence on a resource through chemical extraction treatments; investigations of temporal variation in the fungal stream spora; investigations of the spatial changes in fungal spora along headwaters and lower reaches of the stream. In addition a survey was made of stream spora in a range of streams in the Northern Territory, Northern New South Wales, Southern New South Wales and Victoria, for a comparison with the aquatic hyphomycete community of Lees Creek. Litter input into Lees Creek was 4.5t ha-l yr-1, with leafy materials accounting for 56% of the total. There was a six fold difference between maximum monthly litter input in summer and minimum input in winter. Leaves of Eucalyptus viminalis and phyllodes of Acacia melanoxylon both broke down faster in winter than in summer. Thus more litter resources are available to support the aquatic hyphomycete populations in Lees Creek in summer than in winter. The litter resource can be broadly classified into persistent materials (bark and twigs) and non-persistent materials (leaves and phyllodes). In summer the non-persistent resources dominate the fungal resource base; persistent litter is prevalent through the year. Over a three year period the fungal spore concentrations in Lees Creek averaged around 2,500 spores L-1, but there were strong seasonal fluctuations with high concentrations during late summer and early autumn (maximum 8,500 spores L-l) and low concentrations during winter and spring (minimum 600 spores L-1). The availability of litter in the stream was evidently the dominant factor determining these seasonal patterns. There was little change in species richness, but the relative spore concentration of individual fungal species changed markedly. Most major species had higher spore concentrations in summer/autumn and lower concentrations in winter/spring, e.g. Tetrachaetum elegans, Lunulospora cymbiformis, Flagellospora sp.33 and F. penicillioides. Other major species, e.g. Clavariopsis aquatica and Alatospora acuminata were more-or-less constant throughout the year. The relative abundance of species in the stream spora was strongly reflected in their relative abundance in naturally occurring litter taken from the stream, and also on samples of fresh litter inserted into the stream. On leaves and phyllodes inserted into Lees Creek aquatic hyphomycetes were first detected after one week in summer and after two weeks in winter, and thereafter until the litter had completely broken down. Chemical leaching treatments of litter before immersion in the stream sped up the colonization rate of most individual species, suggesting that both leaves and phyllodes contain minor amounts of inhibitory compounds. Evidence was found of successional change in patterns of fungal occurrence on several litter types. On leaves and phyllodes inserted into the stream in summer, the species most frequently detected in the early stage of decomposition were fungi such as T. elegans and L. cymbiformis; when these later declined, F. penicillioides and Flagellospora sp.33 tended to dominate the middle stages of succession; and A. acuminata consistently dominated the late stages. Species characteristic of the early phase appear to have either one of two colonization strategies: to saturate the stream with numerous small spores (L. cymbiformis); or to have relatively few, large, well resourced tetraradiate spores which impact very efficiently on litter (T. elegans). No evidence was found for absolute litter resource specificity from comparisons of fungi sporulating on eight different litter types sampled from the stream. However, many species of hyphomycetes showed a clear preference for some litter types, e.g.: L. cymbiformis and T. elegans were most abundant on leaves of E. viminalis and Pomaderris aspera; and A. acuminata on ferns, sedges and persistent litter types such as eucalypt bark. Differences in the physical structure of different types of litter were considered to be the main discriminating factor, with chemical and nutritional differences providing possible subsidiary explanations for litter preference amongst the aquatic hyphomycetes. Differing litter preferences, e.g. for persistent versus non­ persistent materials, and differing responses to water temperature, appear to be important factors determining patterns of seasonal occurrence of a species. Downstream from the sources of two headwater streams of Lees Creek, both the number of species and number of spores increased rapidly in the stream. Marked differences in dominant species between the two streams, and the labile nature of populations within each, suggested that there was a significant stochastic element to the composition of headwater communities, perhaps reflecting the importance there of chance inoculum arrivals from outside the stream. Further down in the main section of Lees Creek, the number of species and concentrations of spores continued to increase in a downstream direction, but in a more gradual and predictable manner. A simple computer model was developed to gain an insight into the dynamics of spore inputs and spore losses in a fungal spore population moving down such a stream. This gave rise to a method of graphical analysis which permitted real inputs and outputs of spores to be estimated for some fungal species in Lees Creek, and revealed unsuspected patterns of population change in others. The aquatic hyphomycete community in Lees Creek was found to be representative of streams of the local region, and to have more in common with other streams of SE Australia than with streams of N. NSW and the Northern Territory. Only one species (L. curvula) was found to be common in streams of these three widely separated regions. Aquatic hyphomycete species richness declined from south to north, which indicates that mean water temperature influences species composition of a stream. Spore abundance, however, was found to be related to litter availability, being highest in a NSW rainforest gully stream and lowest in a NT river running through predominantly grassland vegetation.