An investigation of the ecology of rhizobia that nodulate white and subterranean clovers in response to soil pH : A thesis submitted in partial fulfilment of the requirements for the Degree of Doctor of Philosophy at Lincoln University

Abstract

White (WC) and subterranean (SC) clovers are widely used and economically important in New Zealand pasture systems. They are used as the base legume in grass-based pastures where they form an effective symbiosis with nitrogen-fixing rhizobia. Biological nitrogen fixation (BNF) by rhizobia is an important process for New Zealand agriculture, particularly in hill and high country areas where the use of nitrogen fertiliser is limited. In these environments, pH is often not optimal for plant and microbial growth. In particular, low (acidic) pH soils have a negative effect on the plant-rhizobia symbiosis and thereby reduce clover growth. Prior research has shown that pH affects the community structure, diversity and function of bacteria in the soil. However, there is limited research on the specific relationship between soil pH and the ecology of rhizobia, either when they are free-living or inhabiting the legume nodule. The overall goal of this research was to increase fundamental knowledge on the relationship between soil pH and the rhizobia nodulating subterranean and white clovers. This will improve our understanding of rhizobial ecology within the complex soil environment and contribute towards identifying strains tolerant of soil pH. In Chapter 2 the relationship between soil pH and the diversity of bacteria that inhabit the nodules of SC and WC across soils that encompassed a broad range of pH was investigated by amplicon sequencing (16S rRNA and nodC genes). The bacterial community was assessed in 5,299 nodules recovered from SC and WC planted in 44 soils that varied in their edaphic properties, including pH. As expected, fewer nodules were formed on both clovers at low soil pH. Rhizobia comprised ~92% of the total reads in both clovers, however several non-rhizobia genera were also present. Soil pH influenced the community structure of bacteria within the nodule. The alpha diversity of nodule microbiome in SC nodules was higher than in WC nodules and SC nodules also harboured a higher relative abundance of non-Rhizobium bacteria than WC. Beta diversity of Rhizobium and non-Rhizobium bacteria was influenced by clover species rather than edaphic factors. Sequence analysis of the nodC gene was used to investigate subspecies diversity in the rhizobia community within the nodules. This work revealed 353 unique nodC protein sequences within the 5,299 nodules, and these varied in their distribution in hosts and soil pH. There was no relationship between soil pH or other soil properties on the diversity and abundance of nodC protein in the nodules of SC or WC. In Chapter 3, pH adaptation by rhizobia strains in relation to the pH of their soil of origin was investigated. A high throughput pH bioassay was developed and optimised to test pH adaptation in a large number of strains (n = 299) in a time- and cost-efficient manner. Adaptation/tolerance by rhizobia strains to extremes of soil pH was a rare trait (≤ 6%). The optimal growth of strains at a particular media pH correlated (r = 0.6) with the pH of the soil the strains were isolated from. However, this association was strongly influenced by the growth of strains from soils of pH >7.0, especially in strains isolated from SC nodules. Naturalised strains that are pH adapted may compete well and/or persist in soils with pH <5.0 or >7.0 compared with introduced, commercial strains. Hence, this result identified the opportunity to improve commercial inoculants through the identification of strains of rhizobia that nodulate SC and WC, fix N and also have tolerance to soil pH extremes. Rhizobia occupy different habitats (bulk soil, rhizosphere soil and root nodules), which differ in their physicochemical conditions, and in the type of carbon sources present. Soil pH is also known to influence soil C cycles through an effect on microbial communities. For this reason, the divergent strains of WC identified from the pH assay were screened on 190 carbon sources (Biolog™) to determine their carbon utilisation profiles. This was done to determine whether pH-adaptation in some strains was linked to a competitive advantage in soils, based on metabolism of a diverse range of carbon sources. A very high degree of phenotypic variation was observed, with the 19 strains placed into 10 groups. Four of the strains used the most carbon sources and exhibited the highest growth (larger AUCs). Some carbon sources (e.g. D-glucosamine, laminarin, L-pyroglutamic acid) stimulated strain growth more than others (e.g. citric acid, tyramine, oxalic acid). A strong association was found between the phenotypic variation and genetic distance of the strains. However, there was no evidence of association of phenotypes with soil pH, soil order, sampling location, other soil physicochemical or climatic properties. Knowledge of C preference by pH adapted strains could support the production of strain-specific formulations that can stimulate maximum growth and assist the effective establishment of strains in the field. The legume nodule is a plant structure that is critically important in the global N cycle. This research investigated the community ecology of rhizobia and “other” bacteria recovered from legume nodules in a background of variable soil pH. The results produced new information on the relationship between soil pH, the nodule biome and pH adaptation by rhizobia. The nodule microbiome was influenced by external soil pH, especially in SC. Although pH adaptation was rare, soil pH did correlate with pH adaptation by strains of rhizobia, raising the potential for selection of strains better suited to sites with sub-optimal soil pH, such as the New Zealand high country. Adaptation to pH was not explained by the ability of strains to utilise specific C-sources, however, this work did highlight the potential for improving formulations by addition of specific C compounds. The strain collection produced by this work may also contain candidate strains with potential as commercial inoculants for soil sites with pH outside the optimum.