Abstract:[Objective] To assess the effect of different dilution gradients of dilution-plate method on the number and composition of soil bacteria, and to study the differences in soil bacterial communities under distinct land use scenarios by comparing plate-dilution method and high-throughput sequencing. [Methods] With four types of soil (secondary forest soil, healthy banana soil, diseased banana soil and paddy soil) collected, soil suspensions of five dilution gradients (10-1–10-5) were prepared. Classic plate-dilution method was employed to obtain culturable bacteria, followed by colony-counting and DNA extraction. In addition, the total bacterial DNA in background soils was extracted. Then high-throughput sequencing of 16S rRNA gene was performed to study the diversity of bacterial communities at different dilution gradients and in background soils, proportion of culturable bacteria in total soil bacteria, and species difference. [Results] The highest increase in soil respiration was found after the conversion of secondary forest to banana plantations, and both dilution-plate method and real-time fluorescence quantitative PCR found the highest decrease in bacterial counts. However, the results of these two methods were not entirely consistent for other land use changes affecting soil bacterial populations. The Chao 1 index of the soil culturable bacterial community was reduced by 86%–98% compared with that of the total background soil bacteria. Dilutions of 100–100 000 times (10-2–10-5) lowered the Chao 1 index of the culturable bacterial community by 35%–60% compared with dilutions of 10 times of soil suspensions. Gradient dilution also reduced the species diversity. Dilution-plate method detected a total of 315–401 microbial genera, with only 21–38 genera common to all dilution gradients, compared with 92–210 genera unique to the 10-1 gradient and 2–59 genera unique to the 10-2–10-5gradients. The proportions of culturable bacteria in the four types of background soils were in the range of 16.1%–47.7% (phylum level) and 7.4%–30.9% (genus level). The number of significantly increased and decreased species detected by plate-dilution method was only 9.7% and 22.9%, respectively of the results by high-throughput sequencing. The two methods identified a number of common divergent species, including Bacillaceae, Micrococcaceae, Microbacterium, Comamonadaceae and Burkholderiales. [Conclusion] The highest percentage of culturable species in total soil bacteria was 47.74% and 30.90% at phylum and genus levels, respectively. The community diversity of culturable species was much lower than that of the background soil bacteria. The proportion and diversity of culturable bacteria was significantly higher in the 10-1 dilution gradient than in the 10-2–10-5 gradients, while there was no remarkable difference between the 10-2–10-5 gradients. Although the dilution-plate method greatly underestimated the differences in bacterial communities between land use practices, common differential bacterial taxa were found by the culture-free and culturable methods. Compared to the background abundance of bacteria in the four soils, the solid plates were significantly enriched in Pseudomonas and Flavobacterium, with fold increases of 2 233–5 805 and 43–4 506. Culture-independent molecular techniques and classical culturable methods should be coupled in the future research to deepen the exploration of microbial resources in complex environments.
