We had a very low sequencing success rate and this appeared to be due to contamination of many samples by saprotrophic other fungal infection. Our sampling was done in mid summer, and this may be a time of high root turnover. The levels of nitrogen fertilization applied in this study termed “high” and “low” would be considered “extremely high” and “very high” in natural settings. The highest rates of nitrogen deposition in the eastern US are generally below 20 kg N/ha/yr , and many forests considered to be exhibiting nitrogen saturation associated decline are below 15 kg N/ha/yr . The high and low levels of fertilization employed here are 20 and 6 fold higher than the atmospheric deposition levels in this part of the northeast. Because there was no evidence of an interactive effect of nitrogen treatment and horizon with regards to community composition, we will first discuss our findings in the context of nitrogen addition experiments and then discuss the implications of our findings on horizon preference. There is some evidence that deciduous forests may respond differently to nitrogen deposition than coniferous forests , so, when possible, we will focus on studies done in deciduous forests. We will also not discuss studies that focus exclusively on ectomycorrhizal sporocarp inventories. Sporocarp inventories formed the majority of early studies on the effects of acid deposition on ectomycorrhizal communities, and could be considered the canary in the coal mine that has spurred 2 decades of research following the seminal work by Arnolds . However, there is ample evidence that sporocarp abundance may not be reflective of mycorrhizal abundance .
We will address the results of our colonization intensity measurements in the context of rooting distribution,vertical grow system ecosystem biogeochemistry and ecosystem responses to nitrogen deposition.Our findings of a significant shift in ectomycorrhizal species composition are generally in agreement with other investigations on the effects of N deposition on ectomycorrhizal commmmunities. Avis et al. and Lucas and Casper found that nitrogen fertilization significantly impacted ECM community composition in oak forests of the eastern US, as have numerous studies in coniferous forests . Our results stand apart from those of Avis et al. and Lucas and Casper in that they observed significant effects on N fertilization on ECM community composition at N fertilization levels between 20 and 35 kg/ha/yr and in as little as two years, while we found no significant effects of 18 years of 50 kg/ha/yr of N fertilization on ectomycorrhizal community composition. Avis et al. also failed to find a significant community change after 18 years of fertilization at 54 kg/ha/yr. The majority of N addition studies on coniferous plots have found a significant impact of 30S50 kg N/ha/yr on ECM community composition , though Ishida and Nordin failed to find an effect of 12 years of 50 kg N/ha/yr on ECM community composition in a spruce forest. Wallenda and Kottke identified a general threshold for 20S30 kg N/ha/yr before marked changes in the ECM community composition are likely to be observed, although they caution that forest specific factors may change this threshold a great deal in either direction, and at the time of their review, there were no studies on the effects of N addition on ECM communities in deciduous forests.
Our findings of a marked reduction in ECM diversity with high N fertilization are also in line with the findings of Avis et al , and Lucas and Casper , which found significantly reduced ECM diversity in their high N treatments, though again, they noted significant decreases in ECM diversity at levels comparable to our “low” N treatment, while we did not. Avis et al. failed to find a significant effect of N addition on ECM diversity. Looking at studies in coniferous forests, a reduction in ECM diversity with N addition seems to be the general finding , but studies have also found no reduction in ECM diversity in forests fertilized with moderate or very high levels of N addition . We could clearly identify certain species to be nitrophilic or nitrophobic but we could not make such characterizations for any higher level phylogenetic groups other than the Clavulinaceae. We found that the Clavulinaceae were generally nitrophobic and this is in agreement with the findings of Avis et al. . Among our 5 most abundant Lactarius species, one was nitrophilic, two were more common in the low nitrogen treatment, and nitrogen had no effect on the abundance of the other two. Lactarius&quietus was also quite abundant in Avis et al.’s study on oak, but they found no consistent reaction to nitrogen fertilization. In their study across a depositional gradient in an Alaskan spruce forest Lilleskov et al. found Lactarius&theiogalus to be dramatically nitrophilic, shifting from 7% to 69% of all root tips from the low to high end of their nitrogen gradient; we found L.&theiogalus to be mildy nitrophilic, occurring in greatest abundance in the low N treatment. Cox et al.’s study across european Picea&abies forests identified the genera Lactarius and Thelepohra/Tomentella to be nitrophilic.Our most abundant species was Cenococcum&geophilum, and it was termed nitrophobic due to its sharply reduced abundance in the high N treatment. Avis et al. found Cenococcum abundance across N treatments variable, with one TRFLP type being nitrophilic and another nitrophobic.
Lucas and Casper found Cenoccocum&geophilum abundance increased on oak roots in response to nitrogen fertilization . Avolio et al. looked at ECM community composition on pine and oak seedlings in response to nitrate fertilization and found that Cenococcum abundance increased markedly in response to N fertilization on oaks, and decreased markedly in response to N fertilization on pines. Cenococcum geophilum is widely thought to be acryptic species complex , and we examined the possibility that we had sub-taxa within C.&geophilum that might have a distinct affinity for nitrogen or horizon. Clustering our C.&geophilum sequences at different % similarities did yield different clusters,macetas cuadradas but did not divide this large group according to any horizon or nitrogen affinity. This inability to define Cenococcum’s niche is not uncommon; in a 2007 commentary, Dickie identified Cenococcum as “one notable exception to the rule of niche differentiation”.The ECM communities in the organic and mineral soil were quite distinct, both across all treatments and within a given nitrogen treatment. Our findings of different ECM communities in the mineral and organic horizon is common to other studies which have examined the ECM communities in organic and mineral soil separately , though to date few studies have done so. We found evidence of increased diversity in the mineral soil. Rosling et al. found higher diversity in the mineral soil, while Dickie et al. found lower diversity in the mineral soil, and Scattolin et al. found moderately increased diversity of ECM species in the mineral soil. All three of these studies were done in coniferous forests. Certain species in our study exhibited clear preferences for one horizon or another, and these trends were also applicable to higher phylogenetic classifications. In particular the Russulaceae exhibited a preference for the organic horizon while members of the genus Inocybe, the order Agaricales, and the families Tricholomataceae and Clavulinaceae were significantly more abundant in the mineral soil. In contrast to our findings, Baier et al. and Scattolin et al. found that Lactarius and Russula were generally more abundant in the mineral soil, though they found individual species within both genera that were more abundant in the organic soil; both studies were done in high elevation coniferous forests. Tedersoo et al. found, as we did, that the Agaricales exhibited strong preference for mineral soil, though they also found the Clavulinaceae to be significantly more abundant in the organic horizon, in contrast to our findings.
In general, our data suggests greater specialization in the mineral horizon than in the organic horizon. Of the 65 species we found, 31 were found only in the mineral soil, while only eight were found exclusively in the organic horizon. Rosling et al. also found a higher proportion of species occurring exclusively in mineral soil.Across all treatments, the ectomycorrhizal colonization intensity was very similar between the mineral and organic horizon, and this stands in some contrast to the published literature. Very few studies have exhaustively sampled the ectomycorrhizal community in the mineral soil , but those that have, have found as many, if not more, ECM in the mineral soil as in the organic soil . More studies have sampled the upper layers of the mineral soil, and ECM colonization intensity, measured as either percent root length colonized or percentage of fine root tips that are mycorrhizal, is generally lower in the mineral soil .However, Scattolin et al. found that the A and B horizons in a montane spruce forest in Italy had significantly higher mycorrhizal colonization intensity than the organic soil, so our finding of equal or slightly elevated colonization intensity in the mineral soil is not unprecedented. Magill et al. assessed the fine root biomass in the same plots we did two years before we sampled these stands. They found that between 55% and 62% of all fine roots were found in the mineral soil, though they only sampled the top 20cm of mineral soil. Thus, our findings of equivalent or higher colonization intensity in the mineral soil can be interpreted as equivalent or higher mycorrhizal biomass in the mineral soil. Nitrogen addition had no effect on the colonization intensity in the low N treatment or in the organic horizon of the high N treatment, and increased the percentage root length colonized in the mineral soil in the high N treatment. According to Magill et al. , the fine root biomass in the organic horizon in these same stands is 25% and 27% lower in the low and high N treatments, respectively. Fine root biomass in the top 20 cm of the mineral horizon has not been significantly affected by nitrogen treatment, thus the percentage of fine roots found in the mineral horizon increased marginally from 55% to 62% with nitrogen addition . We can thus infer that there is a significantly higher fraction of total ectomycorrhizae in the mineral soil in the high N treatment. There is considerable variation in the literature in reported responses of ECM colonization and fine root biomass to nitrogen addition. The consensus seems to be that nitrogen addition decreases both, but the large number of reports finding the opposite suggests that characteristics of the individual forest being examined may be important to consider. In a meta analysis of 6 studies conducted in 14 forest stands, Treseder et al. found a moderate decrease in %RLC of 5.8% with nitrogen fertilization, but also noted that the responses to N were very heterogeneous. In a meta analysis of boreal forests’ responses to nitrogen addition, Cudlin et al. found a not significant decrease of 10% in percent mycorrhizal colonization. Carfrae at al. and Koren and Nylund found significantly increases in %RLC with nitrogen fertilization. Wöllecke et al. noted a sharp decrease in % RLC with nitrogen fertilization, but this was much more pronounced for the organic horizon than the mineral horizon, also indicating increased fraction of total mycorrhizal activity in the mineral soil under nitrogen addition. To the authors’ knowledge no published studies have examined the effects of nitrogen on ECM colonization intensity in deciduous forests. The literature on the effects of nitrogen on rooting activity is no more consistent. While, many micro and mesocosm studies have demonstrated a decreased ratio of root biomass to shoot biomass with increased N availability , and the mechanisms for this are well understood , the effect is generally driven by increases in above ground biomass and the effects of nitrogen fertilization on below ground biomass in forests vary considerably. Cudlin et al. found an increase in root biomass of 10% in their meta analysis of 22 forest studies, while Ostonen et al. found a 20% decrease in specific root length in a meta analysis of 54 studies, both of these studies reported such high variation in responses to N that neither trend was significant.The altered mycorrhizal colonization we observed in the high N plot may be an adaptation to limitation or colimitation by phosphorous or base cations.