An approach to evaluate potentiality for biological nitrogen fixation in soil: metagenomic analysis of nifh sequences in dna and rna extracted from argentinean soils under different agricultural management

PRISCILA A CALDEROLI; FILIPE BEHRENDS KRAEMER; O. MARIO AGUILAR; COLLAVINO M.M

Simposio; The 12th Dahlia Greidinger Memorial Symposium 2016; 2016

Biologically available N is often a limiting nutrient in agriculture. Biological reduction of molecular N2 gas to assimilable ammonium (BNF) is considered the main route by which fixed nitrogen enters the biosphere by natural processes (Galloway et al., 2004). BNF, catalyzed only by prokaryotic cells, requires nitrogenase, a two component complex enzyme. The nifH gene, which encodes Component II, is highly conserved across the bacterial and archaeal domains.Since 1980, agriculture has rapidly expanded in Argentina, replacing grasslands, with the widespread adoption of limited tillage systems, particularly no-till with crop rotation (Kovalevski & García, 2007). Within the context of a global project on soil biology, a large effort was made to provide microbial indicators of the sustainability of agricultural no-till systems in the Central Pampas Argentina. This study is a specific survey of only the N2-fixing microbial communities in soil, using deep pyrosequencing of the nifH gene. Firstly we aimed to examine the abundance, diversity and structure of diazotrophic communities in a gradient of Argentinean agricultural soils subjected to the following no-till treatments: intensive crop rotation, nutrient replacement and minimal use of agrochemicals (good agricultural practices, GAP); soybean monoculture, low nutrient replacement and high agrochemical use (poor agricultural practices, PAP), and a grassland soil used as a reference natural environment (NE).A dataset of 87020 nifH reads and 16782 unique nifH protein sequences obtained over 2 years from four locations across a gradient of agricultural soil types in Argentina were analysed. At 98% similarity, the 1558 operational taxonomic units (OTUs) for the entire study were assigned to 17 subclusters of the four major nifH clusters previously defined by Zehr et al. (2003). Phylogenetic analysis revealed an expected high proportion of Alpha-, Beta- and Deltaproteobacteria, mainly relatives to Bradyrhizobium, Methylosinus/Methylocystis and Pelobacter/Geobacter, but a surprising paucity of Gammaproteobacteria. A significant difference in diazotrophic diversity was observed across the four locations but not among management treatments. Analysis of variance and stepwise regression modelling suggested location and treatment-specific influences of soil type on diazotrophic community composition and organic carbon concentrations on nifH diversity. nifH gene abundance, determined by quantitative real-time polymerase chain reaction, was higher in agricultural soils than in non-agricultural soils, and was influenced by soil chemistry under intensive crop rotation but not under monoculture. Active diazotrophic populations were also examined by deep sequencing of cDNA-nifH amplicons from RNA extracted from the Argentinean soils subjected to the two previously described no-till practices and sampled each at 10 and 20 cm depth. Most abundant OTUs where related to Geobacter, Bradyrhizobium, Nostoc and Anabaena. The composition is similar to that observed in the potential community, except cyanobacteria which represented less than 1% of total ADN-nifH sequences. In cultivated soils, a significant change is observed in the composition of the community in the depths analyzed. For example, Geobacter OTUs increased their relative proportion in the 20 cm whereas Bradyrhizobium displays the opposite trend. On the other hand, relative abundance of Cyanobacteria and Desulfovibrionales were significant (>1%) only in the first 10 cm (1B 33%) and 20 cm (3B 15%). Contrarily, communities at both depths in grassland soil (NE) were similar, with high proportion of Geobacter and Rhizobiales. Using the OTU richness and diversity estimators, NE was the more diverse and both estimators were lower at deeper sampling in pristine and GAP, whereas increased in the 20 cm depth of PAP.Differences in diversity across sites appear to be mainly associated with the statistical distribution of the pores, hydrophobicity, structural stability and levels of soil organic carbon. Correlations suggest that diversity increased along the occurrence of different niches e.g. pore diameter, and their structural stability.In contrast to the diazothropic community detected in DNA from soil, we conclude that diversity and richness of the active populations seem to be affected by agricultural practices, and also by soil depth and soil parameters.