Search

Genome duplication is associated with multiple changes at different levels, including interactions with pollinators and herbivores. Yet little is known whether polyploidy may also shape belowground interactions.Methods: To elucidate potential ploidy‐specific interactions with arbuscular mycorrhizal fungi (AMF), we compared mycorrhizal colonization and assembly of AMF communities in roots of diploid and tetraploid Centaurea stoebe s.l. (Asteraceae) co‐occurring in a Central European population. In a follow‐up greenhouse experiment, we tested inter‐cytotype differences in mycorrhizal growth response by combining ploidy, substrate, and inoculation with native AMF in a full‐factorial design.Key Results: All sampled plants were highly colonized by AMF, with the Glomeraceae predominating. AMF‐community composition revealed by 454‐pyrosequencing reflected the spatial distribution of the hosts, but not their ploidy level or soil characteristics. In the greenhouse experiment, the tetraploids produced more shoot biomass than the diploids did when grown in a more fertile substrate, while no inter‐cytotype differences were found in a less fertile substrate. AMF inoculation significantly reduced plant growth and improved P uptake, but its effects did not differ between the cytotypes.Conclusions: The results do not support our hypotheses that the cytotype structure in a mixed‐ploidy population of C. stoebe is mirrored in AMF‐community composition and that ploidy‐specific fungal communities contribute to cytotype co‐existence. Causes and implications of the observed negative growth response to AMF are discussed.

Soil chemical properties. Data represent means and standard errors from four time points (n = 24). (PDF 11 kb)

A) Daily mean temperatures in 5 cm soil depth in course of the year. The arrows indicate approximate soil sampling date. B) Monthly and yearly mean temperatures in degrees of Celsius in 5 cm soil depth in control and snow-manipulated dry and wet tundra sites. (PDF 315 kb)

Identification of the 30 most abundant bacterial OTUs from dry and wet tundra soil according to SILVA database, their relative abundance in control (C) and snow manipulated plots (S) across the plant growing seasons (June, July, September and October) and seasonal average (SA). Data of relative abundance are expressed as means from 6 (24 for SA) replicates, standard errors are shown in italic. Statistically significant differences between control and snow manipulated plots in the individual seasons are highlighted (DESeq2, Benjamini-Hochberg correction, p

Venn diagrams showing shifts in (a) fungal and (b) bacterial community composition. Proportion of shared and unique OTUs across four time points is displayed (n = 24). OTUs with > 0.01% abundance were considered as present. D-dry tundra, W-wet tundra, C-control, S-snow manipulation. (PDF 53 kb)

OTU accumulation curves expressed as the number of OTUs by number of reads from sequencing. (PDF 307 kb)

Relative abundance of archaea in prokaryotic sequence pool in dry (A) and wet (B) tundra soil in control (green) and snow-manipulated (red) plots by season and as a seasonal average (SA). The data represent the means with standard errors (n = 6), for seasonal average (n = 24). (PDF 57 kb)

Identification of the 30 most abundant fungal OTUs from dry and wet tundra soil according to UNITE database, their relative abundance in control (C) and snow manipulated plots (S) across the plant growing seasons (June, July, September and October) and seasonal average (SA). Data of relative abundance are expressed as means from 6 (24 for SA) replicates, standard errors are shown in italic. Statistically significant differences between control and snow manipulated plots in the individual seasons are highlighted (DESeq2, Benjamini-Hochberg correction, p

Heatmap depicting the relative abundances of the 20 most dominant fungal genera in dry (a) and wet (b) tundra soil in control (C) and snow-manipulated (S) plots across the seasons. The data represent mean abundances (n = 6). Statistically significant effect (DESeq2, Benjamini-Hochberg correction, P

Proportion of sequences belonging to different bacterial functional groups in dry (D) and wet (W) tundra soil in control (C) and snow-manipulated (S) plots. The data represent the means with standard errors (n = 6). (PDF 51 kb)