Final Report Summary - MICROBS (Assessing the effects of climate change on soil ecosystem functioning using an established alpine Microbial Observatory)
Contemporary climate change is taking place at an alarming pace. One effect of this is changes in species assemblages, often associated with a loss of biodiversity. In spite of their critical ecological importance, relatively little is known about such changes in soil microbial communities. Microorganisms are known to adapt faster to local conditions than plants and other macro-organisms, making them ideal as indicators for soil and ecosystem health. Mountain areas are particularly well suited for changes in microbial communities related to climate, thanks to their steep altitudinal gradients. This also makes them more sensitive, increasing the risk for loss of biodiversity and perturbations of ecosystem function, with possible global implications. Advancements in DNA sequencing technology now allow in-depth studies of microbial community composition and function. Taking advantage of this, MicrObs was designed to investigate the potential for utilising microbial communities as a bioindicator of soil health along an altitude gradient, established as part of the Microbial Observatory in “Ordesa and Monte Perdido” National Park in the Pyrenees. The study also included a transplantation experiment, designed to assess how rapidly compositional or functional changes may occur.
The altitudinal gradient utilised in MicrObs ranges from 1,500 to 2,600 meters above sea level. Samples were taken annually from 2012 to 2014 in late August or early September, at 100 meter altitude intervals. Other samples were taken in order to assess spatial heterogeneity, at two altitudes, and to compare the rhizosphere and bulk soil horizons. Soil samples were also taken to assess temporal variability, including from snow-covered sites. Microbial community structure was determined using amplicon sequencing targeting prokaryotes, eukaryotes and fungi (targeting 16S and 18S ribosomal RNA, and ITS, respectively). Laboratory protocols for amplicon sequencing using these taxonomic markers with the MiSeq platform of Illumina were developed, along with protocols for downstream bioinformatic analyses. Other biological soil parameters such as ATP content, bacterial vs. fungal abundance (qPCR) and functional physiological profiles (Biolog EcoPlatesTM) were also measured, in addition to soil physicochemical properties, including nutrient concentrations. During the third year, temperature was monitored hourly at all sampling sites using on-site sensors, and at several sites during the second year.
At order rank, prokaryotic communities were dominated by Rhizobiales, Acidobacteriales and Chthoniobacterales, whereas Mortierellales, Archaeorhizomycetales and Agaricales were the most common fungal taxa. Fungal community structure assessed by ITS and 18S agreed well with each other, sharing the same three dominating taxa, a mutual mantel R statistic of 0.81 (p<0.001) and strongly correlated diversity estimates (rarefied OTU richness, sigma=0.91 p=2x10-13). These are important methodological results, since no study has applied and compared this combination of primers to the same samples previously. Further, prokaryotic, fungal and total eukaryotic community structure appeared significantly correlated as well as fungal, protist and metazoan communities assessed by 18S (p<0.001). Diversity estimates also correlated strongly between prokaryotes and fungi, as well as all prokaryotes and total eukaryotes (p<0.0001). This illustrates the importance of interactions between these life-forms, including trophic links, mutualistic and competitive interactions, including parasitism; the later suggested by presence of taxa known to be predominantly parasitic.
The majority of taxa were shared between altitudes, although relative abundances varied considerably. The results show that the structures of both prokaryotic and fungal communities were significantly correlated to altitude (and thus average yearly temperature). However, other parameters appeared to have a stronger effect on community structure, namely sampling year, C/N (carbon-to-nitrogen) ratio, and soil composition (in terms of sand, silt and clay). Other parameters correlated significantly with community structure, although their relative importance compared to altitude remain to be determined, due to incomplete sampling. A remaining challenge is also to study the direct and indirect (ecologically mediated) effects of temperature and meteorological variations on nutrient composition. Nutrient levels varied strongly between years and sampling occasions. In addition to meteorological factors, temporal differences between years may be due to random or periodic fluctuations. As for spatial variability, a grid experiment indicated that physicochemical parameters varied with increasing strength over relatively short distances (4 – 17 m), on a level comparable to that between altitudes.
Because the effect of the level of heterogeneity (temporal and spatial) and the relatively stronger effect of other variables, only one individual taxon abundance was found to be directly correlated to altitude. However, several taxa were linked to C/N-ratio and potassium concentration. In conclusion, these results provide insights into the challenges associated with heterogeneity in natural soil habitats at different spatial and temporal scales. A practical consequence of this is the need for a larger number of spatial replicates, in order to further study the probable effects of climate change on the soil ecosystem. Although disappointing for the utility of microbial communities as biological indicators for climate change, this is an important methodological insight. More importantly, MicrObs demonstrates that it is possible to study the probable effects of climate change on microbial community structure, using carefully chosen altitude gradient as proxy for temperature. The strong interactions implicated between prokaryotes, fungi and protists also show great promise for further studies. The relationships between several taxa with the C/N ratio and potassium concentration are other ecological insights warranting further study. Thus, we believe the results generated will be very valuable for the design and interpretation of future work in the interface of climate change and soil microbial ecology.
The results of the transplantation experiment were not conclusive, probably due to the influence of the artificial barriers constituted by the cylinders used for this purpose. Further, the planned RNA-based work targeting effects on protein-coding transcripts was not successful, due to technical problems, as well as the high levels of heterogeneity identified. Other experiments were instead added, including identification of differences between rhizosphere and bulk soil, the effect of snow cover, and differences in gene abundance related to carbon cycling using a novel targeted metagenomics approach. We plan to continue these analyses, which are not yet completed, as well as dissemination of results generated so far. The fellow (Anders Lanzén) has secured additional research grants from the Spanish Government and will thus be able to stay in the host group for, at least, two years from the autumn of 2015. Apart from dissemination at conferences and a master thesis directed, a project website and blog has been made available at http://microbsproject.blogspot.com. A ten minute video presentation was also produced. English and Spanish versions are available on the project website.
Apart from MicrObs, Anders Lanzén also collaborated in two related projects. This resulted in a publication and a manuscript in review. In the later, soil microbial communities of mountain pastures were investigated using a similar amplicon sequencing approach. He has also continued his collaboration with scientists at the University of Bergen regarding methodological aspects of environmental genomics. Finally, MicrObs has successfully provided Anders Lanzén with an important opportunity to develop and strengthen his conceptual and practical skills in soil microbial ecology, including sequence data analysis, study design, field work and laboratory techniques. It has also strengthened his skills in supervision (thanks to the integration of a MSc student), project management, grant preparations, dissemination, and last but not least cultural skills including Basque and Spanish language proficiency.
The altitudinal gradient utilised in MicrObs ranges from 1,500 to 2,600 meters above sea level. Samples were taken annually from 2012 to 2014 in late August or early September, at 100 meter altitude intervals. Other samples were taken in order to assess spatial heterogeneity, at two altitudes, and to compare the rhizosphere and bulk soil horizons. Soil samples were also taken to assess temporal variability, including from snow-covered sites. Microbial community structure was determined using amplicon sequencing targeting prokaryotes, eukaryotes and fungi (targeting 16S and 18S ribosomal RNA, and ITS, respectively). Laboratory protocols for amplicon sequencing using these taxonomic markers with the MiSeq platform of Illumina were developed, along with protocols for downstream bioinformatic analyses. Other biological soil parameters such as ATP content, bacterial vs. fungal abundance (qPCR) and functional physiological profiles (Biolog EcoPlatesTM) were also measured, in addition to soil physicochemical properties, including nutrient concentrations. During the third year, temperature was monitored hourly at all sampling sites using on-site sensors, and at several sites during the second year.
At order rank, prokaryotic communities were dominated by Rhizobiales, Acidobacteriales and Chthoniobacterales, whereas Mortierellales, Archaeorhizomycetales and Agaricales were the most common fungal taxa. Fungal community structure assessed by ITS and 18S agreed well with each other, sharing the same three dominating taxa, a mutual mantel R statistic of 0.81 (p<0.001) and strongly correlated diversity estimates (rarefied OTU richness, sigma=0.91 p=2x10-13). These are important methodological results, since no study has applied and compared this combination of primers to the same samples previously. Further, prokaryotic, fungal and total eukaryotic community structure appeared significantly correlated as well as fungal, protist and metazoan communities assessed by 18S (p<0.001). Diversity estimates also correlated strongly between prokaryotes and fungi, as well as all prokaryotes and total eukaryotes (p<0.0001). This illustrates the importance of interactions between these life-forms, including trophic links, mutualistic and competitive interactions, including parasitism; the later suggested by presence of taxa known to be predominantly parasitic.
The majority of taxa were shared between altitudes, although relative abundances varied considerably. The results show that the structures of both prokaryotic and fungal communities were significantly correlated to altitude (and thus average yearly temperature). However, other parameters appeared to have a stronger effect on community structure, namely sampling year, C/N (carbon-to-nitrogen) ratio, and soil composition (in terms of sand, silt and clay). Other parameters correlated significantly with community structure, although their relative importance compared to altitude remain to be determined, due to incomplete sampling. A remaining challenge is also to study the direct and indirect (ecologically mediated) effects of temperature and meteorological variations on nutrient composition. Nutrient levels varied strongly between years and sampling occasions. In addition to meteorological factors, temporal differences between years may be due to random or periodic fluctuations. As for spatial variability, a grid experiment indicated that physicochemical parameters varied with increasing strength over relatively short distances (4 – 17 m), on a level comparable to that between altitudes.
Because the effect of the level of heterogeneity (temporal and spatial) and the relatively stronger effect of other variables, only one individual taxon abundance was found to be directly correlated to altitude. However, several taxa were linked to C/N-ratio and potassium concentration. In conclusion, these results provide insights into the challenges associated with heterogeneity in natural soil habitats at different spatial and temporal scales. A practical consequence of this is the need for a larger number of spatial replicates, in order to further study the probable effects of climate change on the soil ecosystem. Although disappointing for the utility of microbial communities as biological indicators for climate change, this is an important methodological insight. More importantly, MicrObs demonstrates that it is possible to study the probable effects of climate change on microbial community structure, using carefully chosen altitude gradient as proxy for temperature. The strong interactions implicated between prokaryotes, fungi and protists also show great promise for further studies. The relationships between several taxa with the C/N ratio and potassium concentration are other ecological insights warranting further study. Thus, we believe the results generated will be very valuable for the design and interpretation of future work in the interface of climate change and soil microbial ecology.
The results of the transplantation experiment were not conclusive, probably due to the influence of the artificial barriers constituted by the cylinders used for this purpose. Further, the planned RNA-based work targeting effects on protein-coding transcripts was not successful, due to technical problems, as well as the high levels of heterogeneity identified. Other experiments were instead added, including identification of differences between rhizosphere and bulk soil, the effect of snow cover, and differences in gene abundance related to carbon cycling using a novel targeted metagenomics approach. We plan to continue these analyses, which are not yet completed, as well as dissemination of results generated so far. The fellow (Anders Lanzén) has secured additional research grants from the Spanish Government and will thus be able to stay in the host group for, at least, two years from the autumn of 2015. Apart from dissemination at conferences and a master thesis directed, a project website and blog has been made available at http://microbsproject.blogspot.com. A ten minute video presentation was also produced. English and Spanish versions are available on the project website.
Apart from MicrObs, Anders Lanzén also collaborated in two related projects. This resulted in a publication and a manuscript in review. In the later, soil microbial communities of mountain pastures were investigated using a similar amplicon sequencing approach. He has also continued his collaboration with scientists at the University of Bergen regarding methodological aspects of environmental genomics. Finally, MicrObs has successfully provided Anders Lanzén with an important opportunity to develop and strengthen his conceptual and practical skills in soil microbial ecology, including sequence data analysis, study design, field work and laboratory techniques. It has also strengthened his skills in supervision (thanks to the integration of a MSc student), project management, grant preparations, dissemination, and last but not least cultural skills including Basque and Spanish language proficiency.