Nitrogen and ectomycorrhizal fungal communities: what we know, what we need to know

Understanding the effect of changing nitrogen availability on biodiversity is critical, not only to address basic questions about the factors that structure communities, but also because atmospheric nitrogen deposition has been increasing in recent decades (Galloway et al., 1995). One hypothesized effect of excess N deposition is the loss of diversity of ectomycorrhizal fungi (Arnolds, 1991). Sporocarp production is one way to assess ectomycorrhizal fungal diversity, but it is not reliable – for that, it is necessary to look belowground, now a realistic proposition with the development of PCR-based molecular identification methods. The study by Peter and co-workers (2001) in this issue (see pp. 311–325), is the first in-depth molecular study to track the changes in belowground ectomycorrhizal communities from the initiation of fertilization onwards. Given that N is often a limiting plant nutrient, one might think that its addition would be a welcome form of free fertilization. However, there is always the possibility of too much of a good thing, and excess N availability is known to have negative effects on plant biodiversity and ecosystem function (Vitousek et al., 1997). Analysing temporal trends in sporocarp production, Arnolds (1991) summarized the evidence for a dramatic decline of diversity in ectomycorrhizal fungi in Europe. He proposed that N deposition was a likely contributor to this decline. Subsequent fertilization experiments and deposition gradient studies demonstrated that additions of N can lead to changes in sporocarp production that paralleled those seen over time in Europe (Wallenda & Kottke, 1998; Lilleskov et al., 2001b). However, sporocarp production is unlikely to reflect below-ground communities: first, fungi may reduce allocation to sporocarps in response to N fertilization, without any change in community structure; second, even in the absence of fertilization, ectomycorrhizal fungal species are not equally represented as sporocarps and on roots. Not all ectomycorrhizal fungi produce conspicuous epigeous sporocarps. Some produce thin crusts on logs or leaf litter (e.g. Thelephoraceae and Corticiaceae), and others have no known sexual stage (e.g. Cenococcum geophilum). Of those fungi that do produce conspicuous sporocarps, a species’ sporocarp production does not necessarily reflect its below-ground abundance (Gardes & Bruns, 1996). Thus, it is necessary to look below-ground to know what is really happening. Unfortunately, this is a much more difficult proposition, because of the methodological challenges involved in fungal identification on roots, combined with high diversity and spatial variability. Previously, most studies were limited to morphological typing. Unfortunately, morphotypes were not easily comparable among studies, and often lead to false lumping and splitting of taxa (Mehmann et al., 1995). These problems have recently been overcome by the development of PCR-based molecular identification methods. Several studies have applied these methods to the question of below-ground community composition in response to N fertilization or over N deposition gradients, but the study by Peter et al. (2001) covers new territory. The authors used a carefully designed sampling regime to enable them to determine relative frequencies of different ectomycorrhizal fungi with minimal disturbance of the site. They also utilized rapidly expanding sequence databases for fungal identification, and an innovative approach for precise quantification of DNA restriction fragment sizes. As a result, they identified a large proportion of fungi present on root tips to the family level or lower. By examining ectomycorrhizal communities before and for 2 yr after fertilization in N-treated and control plots, they have been able to track short-term community changes, and compare them with the trends in sporocarp production in the same plots. What they found is that there is a change in ectomycorrhizal fungal communities as seen both above- and below-ground, but that the below-ground response is less pronounced. Whereas in the sporocarps there was a significant decline in species richness and diversity, below-ground there was no loss of diversity in response to the treatment. However, certain taxa declined in frequency in response to N inputs (e.g. Russula spp.), whereas others did not (e.g. Tylospora asterophora). This study provides the best evidence of short-term below-ground community response to N inputs. Previous short-term studies have shown little or no change below-ground (Saunders et al., 1996; Kårén & Nylund, 1997; Brandrud & Timmermann, 1998; Jonsson et al., 2000). The ability of Peter et al. (2001) to detect a change may be due to two factors. First, their additions of N were larger than those in previous studies, perhaps leading to more rapid change. Second, the combination of molecular identification methods with the above-mentioned experimental design may have given the authors more sensitivity to detect small changes than in previous studies. The smaller change that Peter et al. (2001) report in the ectomycorrhizal fungal community below-ground relative to sporocarps can be interpreted in two ways: as a system operating with no lags, in which the community shows little response because N does not affect the below-ground species composition, but rather only affects allocation to sporocarp production; or as a system which exhibits lags, in which the below-ground community shows less short-term response than the sporocarp community, but significant long-term response. Two molecular studies of the effects of long-term inputs on below-ground communities suggest that the latter model is more likely. In both a fertilization study (Kårén, 1997), and an N deposition gradient study (Lilleskov et al., 2001a), high N inputs are associated with strikingly similar changes in ectomycorrhizal fungal communities, resulting in the loss of many ectomycorrhizal fungal taxa and a shift in the dominants. It is important to know how much we can generalize these results. Arnolds (1991) found that diversity of sporocarps had declined more for conifer-associated than for deciduous-associated ectomycorrhizal fungal taxa. Along a north–south transect in Europe Taylor et al. (2000) found a negative relationship between morphotype richness and soil inorganic N in spruce stands, and a weaker positive relationship between these variables in beech stands. Given that all other below-ground studies of N effects have been carried out in conifer forests, the work by Taylor et al. (2000) underscores the need for more conifer–deciduous tree comparisons. Although evidence is accumulating that N inputs can cause changes in ectomycorrhizal fungi communities, the proximal mechanisms for this change are unclear. Some possibilities include the large carbon cost of assimilation of inorganic N into amino acids under high N conditions (Wallander, 1995); changes in host-plant nutrition and subsequent shifts in host-plant carbon allocation and receptivity to ectomycorrhizal fungi; and N-mediated declines in soil pH, base cation availability and toxic metal availability. Given the difficulties involved in defining individuals of ectomycorrhizal fungi at the root level, we also cannot say whether observed changes in species abundance occur via clonal expansion or spore colonization. The N-induced change in sporocarp production could affect below-ground community composition, if certain species require spore inputs for local maintenance. For example, one might expect that species colonizing mature forest stands would be effective at vegetative spread, but it appears that some Russula species from mature forests actually have numerous small genets, suggesting either that genets expand very slowly, or that continued colonization by spores may be important (Redecker et al., 2001). Given that Russula sporocarp production declines with increasing N inputs (Peter et al., 2001; Wallenda & Kottke, 1998; Mehmann et al., 1995; Lilleskov et al., 2001b), this raises the possibility that reduced spore inoculum may be one mechanism leading to below-ground decline of Russula. Perhaps the biggest question is, what functional effect does ectomycorrhizal fungal community change have on forest ecosystems? N appears to limit plant growth in many temperate and boreal forests dominated by ectomycorrhizal trees, and its availability varies over a broad range of temporal and spatial scales. So it is not surprising that the organisms that supply the majority of N to these trees – the ectomycorrhizal fungi – possess a broad physiological potential for N uptake, utilizing a range of inorganic and organic N sources, and supplying this N to plant hosts (Smith & Read, 1997). Recent work suggests that there may be a decline in the importance of taxa utilizing organic N with increasing soil N availability (Lilleskov et al., 1998; Taylor et al., 2000). There are many other ways that ectomycorrhizal fungi community function might change as community structure changes. It seems likely that carbon supply and nutrient availability interact to select functionally distinct fungi. As nutrients other than N become limiting, do we see functional shifts towards species that specialize on the new limiting resources? As carbon supply to roots declines, do we see a shift toward taxa that are more carbon efficient, or toward taxa that are carbon parasites? Can we disentangle the effects of changing resources and changing ectomycorrhizal fungal communities on plant growth? How do shifts in community composition and changing carbon allocation below ground affect mycorrhizosphere organisms, and below-ground food webs? These questions remain largely unanswered. The study by Peter et al. (2001) reinforces our belief that they are worth asking. *Author Correspondence(tel +1510 643–5483; fax +1510 642–4995; emaillillesko@uclink4.berkeley.edu)