Glyphosate rhizoremediation strategies for soils under intensive agricultural use

Abstract

The Argentinian Humid Pampa is the main agricultural region of the country and one of the most important land fields in South America. The agricultural production model that prevails entails, among other problems, an intensive use of inputs, mainly agrochemicals, meaning putting at risk both human and ecosystems health. In the past years, the heavy dependence on glyphosate by the agricultural sector faced the growing concern in the environment, social and human health setting a growing dispute in the country. Considering the wide extent of the territory, the need to preserve the soil texture and quality, while not interfering with the main economic activity, lowcost and on-site remediation techniques arises as the best option to remove glyphosate in soils and thus, avoid the contamination of water bodies. That is why rhizoremediation, which is the use of plants and root associated microorganisms to remove pollutants from soil, emerges as the most promising phytoremediation technology. In this context, the main objective of this thesis work is to develop the knowledge to design a glyphosate rhizoremediation system to reduce its associated impact in the Argentinean Humid Pampa soils under intensive agricultural exploitation. To achieve this goal, glyphosate tolerance testing methods were developed and applied on different plant species with commercial value. Then, phenotypic and genotypic characterization were made on isolated glyphosate degrading and tolerant bacterial strains. Two of the most promising strains were selected for whole genome sequencing. Additionally, soil and rhizosphere samples from which bacteria have been obtained were also characterized based on its physicochemical and microbial community properties. Finally, different plantmicroorganism interaction experiments were conducted, together with a microcosms assay, where different plant-bacteria combinations were exposed to an agronomic dose of glyphosate in soil. Among all tested plant species, L. corniculatus showed the most suitable tolerance to glyphosate, since it was able to grow in a maximum bioavailability medium with 5.0 mg Kg-1 of a commercial glyphosate product, and to endure a typical product application of 700 g ha -1 . Besides, methods such as measurement of shikimate and the impact of the analysis using a visual scoring should also be considered for future applications in field trials for their simplicity, costs and the provided information. A culture collection was obtained from different sample types, and physicochemical and community structure characterization was also performed. When explaining the observed differences in the microbial communities, it was found that the rhizosphere effect is more important than the soil characteristics or the periodically glyphosate use in the plot. Almost all the isolated microorganisms corresponded to the class Alphaproteobacteria. These microorganisms fulfilled the requirements of having a greater glyphosate tolerance in several orders of magnitude, to the maximum reported glyphosate concentration in agronomic soils, the capability of using glyphosate as the sole source of phosphorus, which is crucial in soils with very low contents in bioavailable phosphorus, and different in vitro plant growth promotion abilities, showing all the strains no less than three positive results on each test. The whole genome sequencing of two of the most promising strains, Ochrobactrum sp. P6BS-III and Rhizobium sp. P44RR-XXI was performed. The use of different methodologies of in silico genome-to-genome comparison was used to obtain complementary information regarding the genome structure of Rhizobium sp. P44RR-XXIV. At the same time, Ochrobactrum sp. P6BS-III genome contributed with a large number of unique genes to the genus, many of them having no assigned function. Finally, plant-microorganism interaction studies were carried out using glyphosate tolerant and degrading bacteria strains together with L. corniculatus, which consisted of inoculation tests on vertical agar plates, imaging studies using different microscopy methods and a microcosm assay. In vertical agar plates assays, strains showed a high growth when inoculated with the plant, without causing any sign of damage on them. The microscopy images obtained showed the different colonization patterns of Ochrobactrum sp. P6BS-III and Rhizobium sp. P44RR-XXIV in roots. The microcosm assay revealed that L. corniculatus inoculated with Ochrobactrum sp. P6BS-III caused the degradation of almost all the amount of 5 mg Kg-1 of glyphosate in soil after 20 days post application, considering this time adequate for the proposed remedial system.