The first objective was to determine whether plant-soil feedbacks become more negative with increasing soil fertility, reflecting the increasing dominance of pathogenic over mutualistic plant-microbial interactions. This objective was achieved by carrying out field observations and collecting seeds and soils from 16 grasslands characterised by a range of plant species richness, primary productivity and soil properties. The seeds and soil material were used in a two-phase plant-soil feedback experiment. In the conditioning phase of the experiment, all species were grown as monocultures in soil collected from their home site. In the feedback phase of the experiment, each species was planted in soil previously occupied by the same species and soils of other species from the same site. An additional experiment ran parallel using the same soils from 16 grasslands but the conditioning and feedback phases were carried out using two common grassland species with commercially supplied seed. The aim of the additional experiment was to establish how plant-soil feedbacks varied along soil gradients in the absence of species turnover and local adaptation processes between plants and soil microbiota. Initial data analysis shows that plants from more productive sites, characterised by soil with low sand content and low plant species richness, experience negative growth response to the presence of their soil microbiomes (negative plant-soil feedback), while plants occupying less productive sites exhibit more positive feedbacks. We also found that such relationships are more clear when using homogeneous, commercially propagated seed material, while patterns become less clear when using locally abundant species and populations that have a history of co-existence in the soil tested. These findings indicate that although negative plant-soil feedbacks enhance local species co-existence, the most negative feedbacks are observed in species-poor ecosystems characterised by high soil fertility. This indicates that local species richness is primarily dependent on species pool size, which is determined by evolutionary and biogeographical history, rather than local plant-soil feedbacks. It is therefore necessary to consider plant-soil feedbacks within the wider context of abiotic and biotic filters that shape species pools across different temporal and spatial scales.
The second objective was to test if variation in soil properties leads to local adaptation within species and differential plant-soil feedbacks. We used populations of Festuca rubra, a common grass species, from 11 grasslands with contrasting soil properties to establish replicated contrasts between soils of different fertility and physical properties. Plants were grown in home live and sterilised soils and two “away” soils with contrasting soil properties that were previously occupied by either the same species (conspecific soil) or four other species (heterospecific soils). The initial data analysis shows that plant-soil feedbacks vary widely between populations and depending on soil type. Out of 11 populations, four showed significantly different feedbacks with their soil microbiomes depending on the history of co-existence. Two populations experienced more positive feedbacks in home than away soils, while the other two populations experienced more negative feedbacks from species-specific microbes in home than away soils. These findings indicate that plant-microbial interactions are subject to microevolutionary changes, and that local adaptation may lead to plant and microbial populations evolving towards either more mutualistic and antagonistic interactions. The direction of local adaptation towards enhanced mutualist or pathogenicity may depend on soil properties but may also be subject to co-evolutionary arms race and cyclical dynamics. We have obtained soil fungal sequencing data, and will link variation in the strength and direction of local adaptation to variation in the composition of fungal communities.
The third objective was to synthesise existing data and macroecological theory to build a predictive framework for predicting plant-soil feedbacks and their role in vegetation dynamics as a function of species traits, environmental context and biogeographical factors. This objective was achieved through collaborative work on review and conceptual papers. We used literature reviews and available data on plant-microbial associations across major fungal functional groups to advance conceptual understanding of how plant-soil feedbacks vary across species, habitats and ecosystems and which aspects of plant life history and plant-microbial interactions could help us explain and predict these feedbacks under current and future climatic conditions.
