Arbuscular mycorrhizal fungi in oat-pea intercropping

Biomass yield and WUE

Biomass increased with precipitation, as expected and reported in Lee³. There was less water uptake in intercropping compared to sole oat and pea. In general, intercropping represented the median of the two sole cropping treatments, where oat had the highest biomass and WUE while pea had the lowest, and where pea had the highest mineral content and oat had the lowest³. Intercropping resulted in advantages in forage yield stability and was not associated with changes to the AMF community.

Alpha diversity

We found differences in AMF species richness estimates in the roots across treatment combinations (i.e., intercropping systems × N fertilizer rate) in 2019 (Chao1, p < 0.05, Supplementary Table 1). In oat, a decrease in AMF richness occurred in soils receiving N fertilizer, while an increase in AMF richness occurred in pea with the N fertilizer application (Fig. 1B, Chao1). The richness of AMF species remained relatively similar in oat-pea irrespective of N fertilizer addition (Fig. 1B, Chao1), and there were no differences in alpha diversity in bulk soils (Fig. 1A, Supplementary Table 1). No significant differences in species richness were observed in 2020 (Fig. 2A, Supplementary Table 1). Instead, we observed a difference in AMF species diversity in the roots across the cropping systems, a pattern not evident in 2019 (Shannon, Supplementary Table 1). Pea and oat-pea had the highest AMF diversity, while oat had the lowest (Fig. 2B, Shannon).

Community composition

Differences in AMF community composition between the two study years in the bulk soil were significantly associated with forage nutritive indicators across all experimental units (Fig. 3A), specifically K, Na, and P concentrations in harvested forages (Ps < 0.05, model P < 0.001). In the root compartment, the AMF community composition between the two study years were also significantly associated with forage nutritive indicators (Fig. 3B), specifically neutral detergent fibre (NDF), calcium (Ca), potassium (K), magnesium (Mg), and sodium (Na) concentrations, biomass productivity as well as WUE (Ps < 0.05, model P < 0.001). Differences in AMF community composition across cropping systems within 2019 were significantly associated with forage indicators (Fig. 3C), specifically biomass, crude protein, NDF, WUE, and most mineral contents (Ca, K, Mg, Na) in the harvested forage (Ps < 0.05, model P < 0.001). In 2020, while the differences in AMF community composition between cropping systems were significantly associated with forage nutritive indicators (model P < 0.01), the fitted vectors did not significantly correlate with any of the AMF community composition of each cropping system (Fig. 3D).

NMDS plots showed spread in root AMF community composition between cropping options (Fig. 4A,B), which were found to be significant using PERMANOVA. Furthermore, the pairwise community composition comparisons indicated that the AMF communities were significantly different across each of the three cropping systems (Ps < 0.05; Table 2, Supplementary Table 2).

Root and soil AMF communities were also different (Supplementary Table 2). AMF alpha diversity between the two compartments was significantly different in both years (Supplementary Table 3). Not only was there a difference in richness of AMF species (Chao1, P < 0.05, Supplementary Table 3), but there was also a difference in species diversity (Evenness, Shannon, InvSimpson, Ps < 0.05, Supplementary Table 3). NMDS plots indicated no overlap in AMF community between the roots and soil compartment (Fig. 4C,D) with PERMANOVA confirming a significant difference between two compartments (Supplementary Table 2). Noticeably, this effect was observed when we pooled the AMF communities across all experimental samples; the similarities were greater between compartments (roots vs. bulk soil) rather than between years (2019 vs. 2020) (Supplementary Fig. 1).

While the soil community alpha diversity did not differ significantly (Supplementary Table 1), we did observe overlapped community composition when comparing between intercropping systems (Supplementary Fig. 2A,B), as well as in N fertilizer rates (Supplementary Fig. 2C–F).

Diversispora abundance in 2019 was negatively correlated with both biomass production and WUE, while positively correlated to crude protein content in the forage (Ps < 0.05, Fig. 5A). Conversely, crude protein was negatively correlated with Paraglomus in 2019, compared to Diversispora (P < 0.05). In 2020, Diversispora abundance was positively correlated with forage Mg concentration (P < 0.05, Fig. 5A). There were significant correlations with specific mycorrhizal genera in the roots in 2019 (Ps < 0.05, Fig. 5B); Diversispora abundance was negatively correlated with forage biomass production while positively correlated with forage Ca and crude protein contents (Ps < 0.01), and Paraglomus abundance was negatively correlated with Ca while Claroideoglomus abundance was positively correlated with Ca (Ps < 0.05). In the same year, NDF and WUE were positively correlated with Paraglomus abundance, but negatively correlated with Diversispora and Claroideoglomus (Ps < 0.05, Fig. 5B). No significant correlations in both soil and root compartments could be found in 2020 (Fig. 5C).

In terms of relative abundance, 80.14% of the AMF community was assigned to phylum Glomeromycota in 2019, and 67.74% in 2020 (Fig. 6; Supplementary Fig. 3). Differential abundance analysis was implemented to see how cropping systems could influence specific genera. There were significant treatment impacts observed within the soil and root compartments (Ps < 0.05; Supplementary Tables 4, 5, 6, 7), as well as significant differences when comparing the bulk soil vs. root compartments (Ps < 0.05; Supplementary Table 8; Fig. 6). However, there was no significant effects of N fertilizer addition on AMF community within the soil or root compartments (Supplementary Tables 4, 5, 6, 7).

In the roots compartment in 2019, a proportional difference was observed in genera Claroideoglomus and Diversispora between oat and pea (p < 0.01, Supplementary Table 7). There was significantly higher content of Claroideoglomus and Diversispora in pea compared to oat (Fig. 6B). In the same year, between pea and oat-pea, a proportional difference was observed in Paraglomus (Supplementary Table 5). There was significantly higher content of Paraglomus in oat-pea compared to pea (Fig. 6B). In the soil compartment in 2020, a proportional difference was observed in Archaeospora between pea and oat-pea (P < 0.01, Supplementary Table 5). There was a significant higher presence of Archaeospora in oat-pea compared to pea alone (Fig. 6C). In the roots compartment in 2020, Diversispora abundance was higher in pea compared to oat, while Paraglomus abundance was highest in oat (Ps < 0.05; Supplementary Table 7, Fig. 6D). In terms of compartmental differences (roots vs. bulk soil), Ambispora and Archaeospora were found to be significantly different in both years, Claroideoglomus and Glomus in 2019, and Paraglomus in 2020 (Supplementary Table 8). In both years, there was a higher abundance of Acaulospora and Archaeospora in the bulk soil compartment compared to the roots compartment (Ps < 0.05; Supplementary Table 8, Supplementary Fig. 4). There was a higher abundance of Claroideoglomus and Glomus in the root compartment in 2019 (Ps < 0.05; Supplementary Table 8, Supplementary Fig. 4A), and there was a higher abundance of Paraglomus in the root compartment in 2020 (P < 0.05; Supplementary Table 8, Supplementary Fig. 4B).

Root colonization

In the counting of hyphae, arbuscule, and vesicle infection in roots, we did not observe any interactions between cropping systems and N fertilizer addition (data not shown). Likewise, there were no significant effects of cropping systems or N fertilizer addition.

Source: Ecology - nature.com