Ecological processes governing the assembly of bacterial root nodule communities in rooibos (Aspalathus linearis)

Abstract

Rhizobia are bacteria specialized in making atmospheric nitrogen available to legumes, playing a key role in the support of legume growth and establishment in managed and natural ecosystems. The relationship between rhizobial diversity and plant performance is not yet clear. Many studies show lower rhizobial diversity is more beneficial to plant growth, but at the same time soils and legume roots contain large amounts of rhizobial and other root-associated bacterial diversity. This suggests we need to better understand the ecological factors that drive the diversity and composition of these microbes in space and time. Understanding the ecological processes driving the diversity of rhizobia and other root-associated bacteria is a prerequisite to potentially manage their beneficial functions to plants. This thesis explores the biogeography and ecological drivers of root nodule bacterial diversity across temporal and spatial scales in rooibos (Aspalathus linearis), a South African-endemic legume crop. The main goal is to understand which factors determine rhizobial diversity and composition at the plant, field, and regional scales, and in both cultivated and wild rooibos populations. Rooibos is adapted to acidic and sandy soils very poor in mineral nutrients and organic matter, and grows under low rainfall conditions (150-450 mm y-1. Rooibos has multiple adaptations to withstand mineral nutrient limitation, among which rhizobial symbioses play a central role. Rooibos associates to a diverse range of rhizobial genera, Mesorhizobium being the dominant group, but frequently interacting with Bradyrhizobium, Rhizobium, and Burkholderia. The plant is particularly interesting to study ecological drivers of rhizobial diversity because 1) plantations lie adjacent to wild populations across the landscape; 2) the plant grows in a climatically and edaphically heterogeneous region with high speciation rates; and 3) the crop has been recently domesticated (~70 years), providing a good analog to study rhizobial diversity before and after crop domestication. This thesis is based on a dual marker approach, in which root nodule bacterial communities of rooibos were characterized using a functional and a taxonomic gene markers. We sequenced a 455 bp-long fragment of the rhizobial gene nodA (encoding an N-acyltransferase important to induce root nodulation), and a 817 bp-long fragment of the taxonomic marker gene gyrB (encoding a topoisomerase common to all bacterial taxa). This combination ensured a majority of the bacterial community of rooibos root nodules could be described. The thesis starts providing a framework to assess how root microbial symbioses differ between cultivated and the ancestral counterparts of legumes. Given the coexistence of cultivated and wild populations, rooibos appears as a good study case upon which the framework could be developed. This framework provided theoretical and experimental basis to assess the ability of different cultivars to associate to, and benefit from, rhizobia from their wild ancestors. This provided a baseline from which a pot experiment growing rooibos on soils from cultivated and uncultivated soils was conceived. The aim of this experiment was to address how geographical and soil abiotic factors drive the diversity of root nodule bacteria in rooibos seedlings. The addition of sheep dung in the experiment revealed the diversity of rhizobia colonizing root nodules was weakly dependant on the soil environment, and strongly related to stochastic assembly processes driven by root biomass and geographic isolation. Rhizobial community composition displayed a strong geographical pattern, whereby different farms contained distinct rhizobial communities, and these became more different with geographic distance. Instead, the communities from plants grown on cultivated and uncultivated soils had low levels of dissimilarity. In a field survey, root nodules from cultivated and wild rooibos plants were collected, and their diversity was related to rooibos foliar nutrient concentrations, plant genotype, geographical location, and soil nutrient concentrations. This revealed that rooibos is nodulated by a core of dominant and widespread Mesorhizobium strains, but whose dominance depends on the geographical location more than any other factor. Organic rooibos cultivation maintained high levels of root nodule diversity at the field level, but had a homogenizing effect at the regional scale. There were established weak but significant links between rooibos foliar nutrient concentrations and rhizobial community structure, suggesting different rooibos populations may harbour distinct root nodule communities. In the pot experiment soils from cultivated and wild rooibos populations were mixed, as a way to increase the available rhizobial pools in the soil. This so-called community coalescence events could introduce rhizobial diversity from wild rooibos populations to arable land after seedling transplantation. This revealed that fertilization, despite not changing the composition of rhizobial communities, promoted the dominance of a larger number of taxa. It had a synergistic effect with soil mixing, whereby a Mesorhizobium OTU that was rare in plants grown on soils from cultivated and wild rooibos populations became dominant only after combining soil mixing and fertilizer addition. Higher soil rhizobial diversity, however, had no influence on any plant response. In summary, through the description of the predominant root nodule bacterial groups, this thesis shows that rooibos root nodule diversity is geographically structured, and weakly affected by the measured environmental factors. Rooibos cultivation appears as a homogenizing force of rhizobial diversity at regional scales, while soil mixing appears as an alternative to associate rooibos seedlings to rhizobia from unmanaged populations locally. Overall, this thesis shows that the ecological drivers of root nodule diversity differ across scales: at the root level, diversity increases probabilistically with root biomass; at the population level, local soil abiotic factors and potentially plant ecotypes define the composition of root nodule communities; and regionally, barriers to dispersal seem to drive divergence between rooibos root nodule communities. Over time, some evidence suggests rhizobial diversity is filtered out as the post-fire succession proceeds in the progress of rooibos from seedling to maturity. Future research should address how agronomic practices in rooibos farming can promote potential benefits of rhizobial symbioses under changing climatic conditions.