Rhizosphere microbes influence one another, forming extremely complex webs of interactions that may determine plant success. Identifying the key factors that structure the fungal microbiome of the plant rhizosphere is a necessary step in optimizing plant production. In a long-term field experiment conducted at three locations in the Canadian prairies, we tested the following hypotheses: (1) diversification of cropping systems influences the fungal microbiome of the canola (Brassica napus) rhizosphere; (2) the canola rhizosphere has a core fungal microbiome, i.e., a set of fungi always associated with canola; and (3) some taxa within the rhizosphere microbiome of canola are highly interrelated and fit the description of hub taxa. Our results show that crop diversification has a significant effect on the structure of the rhizosphere fungal community but not on fungal diversity. We also discovered and described a canola core microbiome made up of one zero-radius operational taxonomic unit (ZOTU), cf. Olpidium brassicae, and an eco-microbiome found only in 2013 consisting of 47 ZOTUs. Using network analysis, we identified four hub taxa in 2013: ZOTU14 (Acremonium sp.), ZOTU28 (Sordariomycetes sp.), ZOTU45 (Mortierella sp.) and ZOTU179 (cf. Ganoderma applanatum), and one hub taxon, ZOTU17 (cf. Mortierella gamsii) in 2016. None of these most interacting taxa belonged to the core microbiome or eco-microbiome for each year of sampling. This temporal variability puts into question the idea of a plant core fungal microbiome and its stability. Our results provide a basis for the development of ecological engineering strategies for the improvement of canola production systems in Canada.