Abstract:
Mycorrhizas of vascular plants and mycorrhiza-like associations of liverworts are integral parts of terrestrial ecosystems, but have rarely been studied in tropical mountain rain forests. Our lightand electron microscopical studies at the RBSF revealed that the roots of nearly all tree species were well colonized by structurally diverse arbuscular mycorrhizal Glomeromycota (Haug et al. 2004, Beck et al. 2005), a few tree species by ectomycorrhizal Ascoand Basidiomycota (Haug et al. 2005), and that epiphytic, pleurothallid orchids formed mycorrhizas with Tulasnella species and members of the Sebacinales (Basidiomycota) (Suárez et al. 2006, Kottke et al. 2007). Species of Sebacinales also occurred in mycorrhizas of hemiepiphytic ericads (Setaro et al. 2006) and Tulasnella species were found in liverworts belonging to the Aneuraceae (Kottke et al. 2007). Traditionally, studies on biodiversity and host specificity have been based on morphologically defined species. No such approach was feasible in the case of the mycobionts in our study as the fungi did not display sufficient structural differences in the mycorrhizas for delimitation of morphospecies. The identification of the mycorrhiza-forming fungi in the forest could be done neither by spore nor by fruiting-body sampling, as both methods would have yielded only a very narrow spectrum of the fungal communities (Husband et al. 2002, Sanders 2004a). Instead, direct sequencing of the associated fungi from the plant material was carried out (Kottke et al. 2007). Given our present stage of knowledge, the sequence types (ribosomal genotypes) can rarely be precisely related to either morphological or biological species. However, the amount of information derived from the sequences of the ribosomal genes appeared to be meaningful in previous ecological studies on arbuscular mycorrhizas (Helgason et al. 2002, Husband et al. 2002, Sanders 2004b) as well as on mycorrhiza-forming Basidiomycota (Bidartondo et al. 2003, Bidartondo et al. 2004). However, analysis of biodiversity and specificity of the mycobionts from field samples using DNA sequences also poses problems. Firstly, results are limited by the available primers. Secondly, problems resulted from the observation that the ribosomal genes can show intraspecific variation, especially in the case of the multinucleate Glomeromycota (Sanders et al. 1995, Lloyd-Macgilp et al. 1996, Clapp et al. 2001, Sanders 2004b). These facts pose general, unresolved challenges to a species concept based on meaningful levels of genetic diversity. Thus the list of fungal sequences presented here is far from being complete and far from being a species list.
