Metagenomics to determine land use effects on soil ecosystem services

Electronic versions

Documents

  • Briony Jones

    Research areas

  • Microbial ecology, Soil, Land use, Traits, Metagenomics, Assembly, 16S rRNA, Grasslands, Arable, pH, UGRASS, Countryside survey, Doctor of Philosophy (PhD)

Abstract

Agricultural intensification is exerting increasing pressure on our soils and there is growing recognition of the trade-offs in balancing human nutritional needs with longer term degradation of soil ecosystem services. It is well established that soil microorganisms play a significant role in driving numerous soil processes, including biogeochemical cycling, however we lack a detailed understanding of the biodiversity and functioning of these organisms in response to land use change. The development of molecular methodologies has provided new insights into microbial community change across different soils and land use systems, but advanced approaches are now needed to synthesise findings across studies and build a more predictive framework to specifically link land use change to change in microbial taxonomy and function. Initial work examined bacterial taxonomic responses to soil pH, as pH is both highly influential over soil microbial communities and heavily associated with land use practices such as fertilization and liming. I modelled pH responses of several thousand bacterial taxa from a large amplicon dataset consisting of > 1000 soils across Britain and developed a novel database/web application enabling querying of 16S sequences to obtain associated ecological information for novel soil taxa, in addition to taxonomy. Importantly this work also demonstrated that taxonomic shifts in soil bacterial communities can be predicted based solely on soil pH information. In subsequent chapters I explored how such predictable changes in taxonomy link to functional change, through exploiting a large metagenomic dataset covering various land use intensification contrasts (grassland and arable) at a range of locations in the UK. Here I found that whilst both taxonomic and functional composition of microbial communities were largely driven by soil pH they also varied with land use intensity. The relationship between soil pH and specific microbial functions was explored through examining relative abundances of a key organic matter decomposition gene (β-glucosidase) within long term pH manipulated grassland plots. Here I found there were increased relative abundance of Acidobacteria β-glucosidase genes in more acidic soils alongside shifts in related glycoside hydrolase families in response to soil pH, demonstrating pH has not only an important influence on bacterial taxonomy, but also important soil functions relating to carbon cycling. I then used novel genomic assembly and binning methods to demonstrate that numerous Thaumarchaeota (also known as Thermoproteota) bins were important discriminators of intensive arable soils based on random forest analyses. Within short read analyses I found that nitrate reductase subunits were important in distinguishing land use and were consistently in higher abundance within high intensity soils. Other denitrification genes were statistically significant indicators of arable soils specifically, including nitrite reductase and nitric oxide reductase genes. Significant grassland indicators included numerous nitrogen fixation genes. Both sulfur and phosphorus metabolism also demonstrated shifts in genes in response to land use with findings collectively indicating differential nutrient acquisition strategies in grassland and arable soils, potentially due to a reliance on nutrients derived from fertilisers within arable soils. Coupling of short read functional indicators with indicators of taxon based on metagenomic bins enabled insights into how land use driven changes in microbial taxa relate to functional change at the community level; and further emphasised the power of a binning approach to link taxa to environmental responses and functions within a land use change context. Through using these novel bioinformatics approaches there is now the opportunity to enhance the database system I developed early in my research, to capture novel ecological information on whole bacterial genomes. This will further enable an improved predictive understanding of the resilience of soil microbial functioning both to land use and wider environmental change.

Details

Original languageEnglish
Awarding Institution
Supervisors/Advisors
  • Robert Griffiths (External person) (Supervisor)
  • David Jones (Supervisor)
Thesis sponsors
  • Bangor University
  • Envision DTP
Award date15 Jun 2021