Manure refinement affects apple rhizosphere bacterial community structure: a study in sandy soil

Abstract

We used DNA-based pyrosequencing to characterize the bacterial community structure of the sandy soil of an apple orchard with different manure ratios. Five manure percentages (5%, 10%, 15%, 20% and 25%) were examined. More than 10,000 valid reads were obtained for each replicate. The communities were composed of five dominant groups (Proteobacteria, Actinobacteria, Chloroflexi, Acidobacteria and Bacteroidetes), of which Proteobacteria content gradually decreased from 41.38% to 37.29% as manure ratio increased from 0% to 25%, respectively. Redundancy analysis showed that 37 classes were highly correlated with manure ratio, 18 of which were positively correlated. Clustering revealed that the rhizosphere samples were grouped into three components: low manure (control, 5%) treatment, medium manure (10%, 15%) treatment and high manure (20%, 25%) treatment. Venn analysis of species types of these three groups revealed that the bacteria community difference was primarily reflected by quantity ratio rather than species variety. Although greater manure content led to higher soil organic matter content, the medium manure improved soil showed the highest urease activity and saccharase activity, while 5% to 20% manure ratio improvement also resulted in higher bacteria diversity than control and 25% manure ratio treatment. Our experimental results suggest that the use of a proper manure ratio results in significantly higher soil enzyme activity and different bacteria community patterns, whereas the use of excessive manure amounts has negative effect on soil quality.

Figure 1. Urease and saccharase activities of soil with different manure ratios.

Each enzyme's highest activity levels was defined as 100%. Different letters means significantly different at P<0.05, urease is represented as capital letters.

Figure 2. Comparison of the bacterial communities at the phylum level.

Relative read abundance of different bacterial phyla within the different communities. Sequences that could not be classified into any known group were labeled “No_Rank”.

Figure 3. Redundancy analysis on the relative abundance of bacterial class using Canoco 4.5.

The canonical axis (horizontal axis) shows that with 21.5% total variability, the correlation with manure ratio is 0.909. The 37 classes that passed the criterion of having >21.5% variability are shown on the horizontal axis.

Figure 4. Genus level bacterial distribution among the six samples.

The phylogenetic tree was calculated using the neighbor-joining method, while the relationship among samples was determined using Bray distance and the complete clustering method. The heatmap plot was generated by Gene cluster and Tree View (written by Michael Eisen, Stanford University). Genuses with >0.1% relative abundances are quoted and the species were mean centered before clustering.

Figure 5. Shared species analysis of the different treatment groups.

The Venn diagram shows the unique and shared species among the different treatments.