Legume symbiotic rhizobia and total bacteria number
The rhizotron experiment lasted for 15 days. As this was not long enough for chickpea plants to form visible nodules, an additional pot experiment was established using the same chickpea genotypes and soils. Nodules were collected and cleaned after 5 weeks growth. We performed 16S rRNA gene amplicon shotgun sequencing of chickpea clean nodules and found ASV29417 consistently accounted for 90%-95% relative abundance in all the samples and considered it as the chickpea symbiotic rhizobia ASV (ASVcsr).
For the abundance of ASVcsr relative to the bulk soil, interactions between soil and other factors were not detected (Table 1). The enrichment of ASVcsr in the endosphere was greater than in the rhizosphere (P < 0.01, 4.02 vs 0.67 log2 fold changes to bulk soil, Fig. 3c). In the rhizosphere of the apical root, no variation of ASVcsr abundance among host genotypes was observed, while the EF genotype had more ASVcsr at the rhizosphere of basal zone than LF2 did. In the apical root endosphere, the genotype LF2 was colonised by the highest abundance of ASVcsr (82 vs 3.2-5.5 based on log2 FC to bulk soil), while this large genotypic difference was minimal in the basal zone. It should be noted that the ASVcsr was one of the core ASVs at the endosphere of LF2 apical roots, but not the core ASV in other niches or genotypes (Fig. 4b).
Bacterial number was estimated by the copy number of bacterial 16S rRNA genes using qPCR. In the endosphere, there was a significant interaction between genotype and longitudinal root segment (P < 0.01, Table S2); the root tip of EF genotype was colonised by more bacteria than the LF genotypes (Fig. 3d), while genotypic variation in the basal zone was not observed. In contrast, bacterial number in the rhizosphere was only influenced by soil type (P < 0.01, Table S2 and Fig. S5).