Soil microbial legacy of plant diversity and drought from Phase
I
Amplification of 16S rRNA yielded 18,990 bacterial OTUs. The db-RDA
showed that soil bacterial community structure was significantly
influenced by soil moisture content (F1 = 23.91,P < 0.001, Fig. 2a) and plant species richness
(F1 = 4.06, P < 0.001, Fig. 2a),
indicating that drought and plant diversity caused significant
alterations to the soil bacterial communities and potentially creating
legacies for future plant-soil feedbacks. These factors explained
21.86% of variation in bacterial community composition. In general,
drought treatments significantly decreased bacterial diversity (Table
S3). There was a marginally significant relationship between soil
bacterial diversity and plant species richness in ambient treatments
(P =0.081) and there was no evidence of a relationship observed
in drought treatment (P = 0.108, Fig. 2b). However, bacterial
diversity was greater in ambient than drought treatments at monoculture
and two-species mixtures (Fig. 2b).
Amplification of ITS2 yielded 3444 fungal OTUs. Soil moisture treatment
(F 1 =12.94, P < 0.001, Fig. 2a)
and plant species richness (F 1 = 3.78, P< 0.001, Fig. 2a) had significant influences on soil fungal
community composition, explaining 14.32% of variation in fungal
community composition, and again producing soil fungal legacies from
plant diversity and drought. Drought treatments significantly decreased
soil fungal diversity (Table S3). However, there was a positive
relationship between soil fungal diversity and plant species richness
under ambient conditions (r = 0.45, P < 0.001),
while no significant relationship was observed under drought (P =
0.257, Fig. 2c), resulting in a significant plant species richness ×
soil moisture treatment interaction in Phase I (Table S3).
Drought significantly reduced the richness and relative abundance of AMF
(Table S3). Drought further affected the relationship between AMF and
plant species richness, resulting in a significant plant species
richness × soil moisture interaction in Phase I for both AMF OTU
richness and relative abundance (Table S3). AMF richness increased with
plant species richness in both moisture treatments while the plant
richness effects were more pronounced in the ambient treatments
(r= 0.85, P < 0.001 for ambient and r = 0.19,P =0.002 for drought treatments, Fig. 2d), and the relative
abundance of AMF was positively related to plant species richness only
in the ambient treatments (r = 0.45, P < 0.001,
Fig. 2e). There was a negative relationship between OTU richness of
fungal pathogens and plant species richness in the ambient treatment
(r = 0.13, P = 0.010, Fig. 2f) but there was no
significant relationship observed in the drought treatment (P =
0.802, Fig. 2f), resulting in a significant plant species richness ×
soil moisture interaction (Table S3). The relative abundance of fungal
pathogens increased with increasing plant species richness in ambient
conditions (r = 0.92, P = 0.008, Fig. 2g) while there was
no significant relationship in drought conditions (P = 0.111,
Fig. 2g). Drought treatments significantly increased the relative
abundance of fungal pathogens in all plant species richness levels
(Table S3).