Towards a global view of ectomycorrhizal ecology


(*Author for correspondence: tel +64 3 321 9646; fax +64 3 321 9998; email

Research on ectomycorrhizal ecology has failed to reflect the diversity and distribution of ectomycorrhizal plants, with most (> 60%) published research concerning a single, northern-hemisphere plant family (the Pinaceae; Table 1) and a geographic range largely restricted to Europe and North America. In this issue of New Phytologist, Tedersoo et al. (pp. 479–490) make a large step towards addressing this imbalance in the literature by providing a detailed analysis of ectomycorrhizal communities and host preference for three plant hosts in Tasmania: Eucalyptus, Nothofagus and Pomaderris. Other than a passing reference by Warcup (1991) this is the first below-ground analysis of fungal associates in the Pomaderrideae (an Australasian subfamily of the otherwise arbuscular mycorrhizal Rhamnaceae), and one of only a limited number of studies of Nothofagus fungal associates.

Table 1.  Survey of ectomycorrhizal research in Web of Science shows a disproportionate focus on a single northern hemisphere plant family (the Pinaceae), with other ectomycorrhizal groups much less well represented
Ectomycorrhizal cladeEctomycorrhizal citations (n)Percentage of literature
  1. Citations were downloaded on August 18 2008 from Web of Science using the search term TS = (ectomycorr*). For each ectomycorrhizal plant clade all genera and plant families were searched, including known synonyms based on the Kew Gardens listing of plant families and genera ( Of 6675 citations, 2495 cited the name of a particular ectomycorrhizal plant genus in the abstract. Tedersoo et al. double the available information on Pomaderris ectomycorrhiza, and significantly increase our knowledge of Nothofagus and Eucalyptus associates.

Pinaceae1536 62
Fagales 553 22
(Nothofagus)  (28) (1.1)
Myrtaceae 288 12
(Eucalyptus) (272)(11)
Salicaceae (sensu stricto) 182  7.3
Ectomycorrhizal Leguminosae 130  5.2
Basal Malvales (Cistaceae, Sarcolenaceae, Dipterocarpaceae)  89  3.6
Tilia  36  1.4
ECM Rosaceae  25  1.0
Uapaca  15  0.60
Nyctaginaceae (Neea/Pisonia/Guapira)   8  0.32
Polygonum   7  0.28
Kobresia   7  0.28
Coccoloba   3  0.12
Gnetum   2  0.08
Asteropeiaceae   1  0.04
Pomaderridae   1  0.04

‘The cosmopolitan distribution of ectomycorrhizal fungi differs markedly from the distribution of ectomycorrhizal plants.’

A remarkable unremarkability

Despite representing a major expansion in the geographic and phylogenetic spread of ectomycorrhizal knowledge, one of the most remarkable outcomes of the community analysis carried out by Tedersoo et al. is the sheer unremarkablity of fungal community composition at the level of genera. Of the dominant genera (Lactarius, Laccaria, Descolea, Russula, Tomentella, Tulasnella, Cortinarius, Clavulina), only one (Descolea) is not found in the northern hemisphere. Many of these genera also occur as important components of Palaeotropical ectomycorrhizal communities, and the same pattern is observed in the Neotropics: Russula, Lactarius and Tomentella/Thelephora occur in Ecuador on Nyctaginaceae hosts (Haug et al., 2005); Clavulina, Cortinarius, Russula and Lactarius are among the most common species in Guyana on roots of Caesalpinaceous hosts (K. L. McGuire, pers. comm.); and Clavulina, Cortinarius, Tomentella and Russula are common associates of the dipterocarp endemic, Pakaraimaea, from Venezuelan Guayana (Moyersoen, 2006; B. Moyersoen, unpublished). The cosmopolitan distribution of ectomycorrhizal fungi differs markedly from the distribution of ectomycorrhizal plants: no ectomycorrhizal plant genera are shared among Tasmania, the tropics and the northern hemisphere.

Does the cosmopolitan distribution of fungal genera inform our understanding of fungal biogeography and ecology? At the most trivial level, this clearly indicates some combination of vicariance (Halling et al., 2008) and the ability to disperse over long distances (Moyersoen et al., 2003), although the importance and scale of dispersal and vicariance remain unclear for most genera (Lumbsch et al., 2008). However, there may also be more general lessons to be learned.

Outside the fungal kingdom, there are a number of other genera shared between the northern hemisphere and Tasmania. Examples include the genera Tyto (barn owls), Circus (harriers), Poa (grasses), Carex (sedges) and Gaultheria (snow berry). These nonfungal cosmopolitan genera all have distinct ecological niches that they fill across their distributions. That is to say, at least in a general sense, there is an ‘owl niche’ filled by the genus Tyto that is not filled by other avian genera, while the genus Tyto is not likely to occupy fundamentally different niches across its wide distribution.

Can we extend these concepts to fungi by analogy? If so, this implies that individual genera such as Cortinarius, Russula, Lactarius, or Laccaria have functional niches that are unique to these genera. We have only the most rudimentary knowledge of what these niches might be. Laccaria, for example, are frequently nitrophilic species that respond positively to disturbance. In some cases, clues as to the niches of species may come from morphology – the extensive hyphae and rhizomorphs common in Cortinarius imply a different ecological role from that of the nearly smooth mycorrhiza common in many Russula species (Agerer, 2001). Further elucidating the niches of these fungi will require deliberate efforts to achieve cross-continental comparisons in fungal community ecology. We also recognize that there will be considerable variation within genera and that genera may not be the best taxonomic level at which to describe niches. Nonetheless, to the extent that there are distinct fungal niches at the level of genus, elucidating these niches provides a conceptual framework for understanding ectomycorrhizal ecology despite the overwhelming diversity of ectomycorrhizal fungal communities.

Dissimilarities as well as similarities

While the cosmopolitan nature of ectomycorrhizal genera argues for broad-scale generic niches, Tedersoo et al. and others (e.g. Ishida et al., 2007) have shown clear evidence for host-preference in fungal communities, suggesting that host specificity may be more the rule than the exception (Bruns et al., 2002). Previously, the most widely recognized cases of host preference have been where preference occurs at the level of the fungal genus. For example, the strong host-preference of the Suillus–Rhizopogon group on Pinaceae and Alnicola on Alnus explains the absence of these fungal genera from Tasmania. The results of Tedersoo et al. suggest that host-preference is also common at the level of species within fungal genera with cosmopolitan distributions. This suggests that while there may be environmental ‘niches’ at the level of genus, host-preference may, in some cases, control community composition at the level of species. Host-preference has important environmental implications because the success of invasion by foreign ectomycorrhizal fungal species will depend on matching both niche and plant host in the native vegetation.

The occurrence of widespread host-preference among ectomycorrhizal fungi raises important questions as to the evolutionary pressures and physiological controls over specificity (Hoeksema & Bruna, 2000; Bruns et al., 2002; Egger & Hibbett, 2004). Tedersoo et al. find that specificity is more strongly driven by root identity than by tree proximity, which suggests that host-preference is not driven by litter chemistry but rather by interactions at the scale of individual roots. Further work is needed on the mechanisms driving this host-preference, particularly exploring both fungal (e.g. competition) and plant (e.g. active selection) pathways favouring host-preference. Mechanisms for reproductive isolation caused by host-preference also need to be further explored. We know that ectomycorrhizal fungi can influence host phenology (Dickie et al., 2007 and citations therein); if the reverse is true, could plant hosts be favouring reproductive isolation of fungi by shifting the phenology of sporocarp production?

From molecules to a global integration

The ‘black box’ of ectomycorrhizal communities is opening increasingly wider. Researchers are identifying individual species in communities using sporocarp matching via restriction fragment length polymorphism (RFLP) or terminal-restriction fragment length polymorphism (T-RFLP), or sequencing and blast comparisons, with upcoming techniques such as pyrosequencing likely to further this progress (Martin & Slater, 2007; Bruns et al., 2008). Combined with molecular phylogenies, these advances make the possibility of a global perspective on ectomycorrhizal community ecology possible. To achieve this, however, there needs to be a much greater emphasis on the under-studied ectomycorrhizal plant taxa and regions of the world. Without these cross-continental comparisons, patterns observed only in northern hemisphere pines have to be seen as merely local phenomena.


Some of the ideas in this Commentary were developed with, and tested on, colleagues attending the Fungal Environmental Sampling and Informatics Network (FESIN) meeting, 2–3 August 2008, Milwaukee, WI, USA; P. G. Kennedy and M. S. McGlone also provided helpful discussion. I. A. D. received support from the New Zealand Foundation for Research, Science, and Technology. B. Moyersoen was funded by FNRS for fieldwork in Venezuela.