Arbuscular Mycorrhizal Fungi (AMF, Glomeromycota) and class 4 Fungal Endophytes (FEs, Dikarya) are plant root mutualists having a wide host spectrum, but with very different behaviours and evolutionary histories. The numerous services they provide to plants explain the great scientific interest they have aroused. The efficiency of these services appear to be extremely dependent on the environment. These environmental conditions can define a mutualistic fungus ecological niche. My thesis project focuses on the study of the dispersal, infectivity and competitiveness of mutualistic fungi. I measured the diversity and abundance of mutualistic fungi according to several environmental factors, notably the identity of hosts, with Marker Gene Metabarcoding approaches using High Throughput Sequencing technologies. These approaches suffer from many technical and fundamental biases studied in specialized literature, but rarely addressed in papers performing diversity analyses. I designed a Python library named EGRETTA to handle these biases. EGRETTA can be implemented in Python and R programmers’ scripts. It includes classes performing database annotation, parsing common formats and Newick trees, curing sequencing data, computing phylogenies and phylogenetic diversity studies.
Corn has a prime significance in French agriculture but was domesticated in southwest Mexico. The main project of my thesis was to compare the populations of mutualistic fungi living with corn in the two continents separated 150 my ago, and to examine the Zea mays (Zm) importance to their evolutionary niche. Are Mexican populations of corn mutualistic fungi better fitted to colonize these plant roots and to provide their services? The low biological diversity of crop plots, the varying degree of maize domestication and the numerous cultural practices highly complicate the question. In the first place, I used another context to train and develop my EGRETTA library. Bromeliaceae display a considerable diversity of life forms, and closely related species can be strictly epiphytic or terrestrial, creating a marked ecological gradient. I optimized EGRETTA’s curation and annotation processes with reads from 17 bromeliad species roots, and built a phylogenetic tree of their symbionts. EGRETTA’s classes developed for the database annotation were also able to automatically retrieve 8 genes from public repositories covering the Bromeliaceae family and to construct the family phylogenetic reconciled tree. We found a dual association of mycorrhizal and non-mycorrhizal fungi in both terrestrial and epiphytic plants.
With this newly gained bioanalytical experience, I analysed fungal sequences sampled from 35 maize plots around Mexico Ciudad and Toulouse and established that the composition of AMF and FEs in soils correlated more with physico-chemical parameters and cultivation practices than with the country of origin. The AMF species diversity in roots of Zm, weeds and cover crops was strongly different, although it was very similar in Zm roots from Mexico and France, with Rhizophagus irregularis (Ri) as the most abundant species. At the intraspecific level, we found a distinct population in each country for several fungal species. We isolated several Ri individuals from our plots and multiplied them in vitro along with 4 Ri strains from collections. I developed a tool for EGRETTA for genome comparison, which identified 146 polymorphic loci Rhizophagus irregularis specific. We used the 8 best candidates, along with the ITS2 haplotype, to characterize our strains. In order to test the competitiveness of several strains of Rhizophagus irregularis in interaction with maize, we identified 5 strain-specific primer pairs for qPCR quantification. |