Monitoring soil microorganisms with community-level physiological profiles using Biolog EcoPlates™ in Chaohu lakeside wetland, east China

Under the circumstance of wetland degradation, we used Biolog EcoPlates™ method to investigate the impact of ecological restoration on the function of topsoil microbial communities by monitoring their metabolic diversity around Chaohu lakeside wetland. Four restoration patterns including reed shoaly land (RL), poplar plantation land (PL), abandoned shoaly grassland (GL) and cultivated flower land (FL) were selected. The result showed a rapid growth trend at the initial stage of incubation, following the fastest change rate at 72 h in both dormant and growing seasons, and the AWCD values of RL pattern was the highest at the detection points of each culture time, while the GL were the lowest. The calculation of diversity indicators also displayed significant lower McIntosh index in dormant season and Shannon-Wiener index in growing season in GL than in the others (P < 0.05). Carbohydrates and carboxylic acids were found to be the dominant substrates used in dormant season, whereas amino acids, polymers and phenolic acids were increasingly utilized by the microbial communities in growing season. We observed soil total potassium as the key factor that significantly affected the utilization efficiency of different carbon sources in both seasons (P < 0.05).


36
As one part of the terrestrial carbon pool, wetlands play a crucial role in global 37 carbon cycling process [1]. Soil is the main component of the wetland ecosystem, and After removing the litter layer, we collected several soil subsamples using a soil 106 auger (6 cm in diameter), mixed together as one sample in each plot, and three plots 107 were selected for biological replicates of each pattern. The samples were stored into 108 sealed polyethylene bags and transported to the laboratory in a cooler box with ice bags.
Analyzer (EA 3000, Vector, Italy). DOC was determined using a TOC auto-analyzer 136 The EcoPlates to be tested were prepared in the following way: 10 g of fresh soil 137 was weighed and put into a 250 mL triangle bottle. Then we added 100 mL sterilized 138 0.85% NaCl solution, shaking for 30 min (speed at 170 r·min -1  In the process of Biolog data analysis, the AWCD value was calculated to reflect

Soil characteristics
There was no significant difference in the SWC of the surface soil in dormant 167 season between GL, FL, RL and PL patterns, while in growth season, the SWC of RL 168 and PL were significantly higher than that of GL (P < 0.05) (Table 3). Overall, the SWC 169 of GL was the lowest in the growing season, and FL obtained the highest value in the 170 dormant season, which was about 4.6 times of the lowest. Soil pHs in GL and PL in 171 dormant season were significantly lower than those in FL and RL, while in growing 172 season, the pH was obviously higher in RL than in the other three patterns (P < 0.05).

173
The contents of NH 4 + -N, NO 3 --N and DOC were the highest in PL in the dormant season, 174 whereas in the growth season, the highest values of NH 4 + -N and NO 3 --N were in FL, 175 and of DOC was in PL (P < 0.05).

176
The contents of SOC and TN were significantly higher in FL than in the other three     carbon substrates (Fig 2A). During the growing season, the relative utilization rate of 225 phenolic acids in GL was observed to be far more less than that in other three patterns, 226 while the rate on carboxylic acids was significantly higher compared with RL, FL and 227 PL ( Fig 2B). The dotted circles represent the 95% confidence interval.

244
The load values of 31 carbon sources on the two principal components were shown 245 in Table 5. It can be seen that there were 4 types of carbon sources significantly 246 correlated with PC1, and 10 types with PC2 (P < 0.05). Consequently, carbohydrates 247 were the main carbon sources that distinguished the soil microbial metabolic 248 characteristics from differential restoration patterns in dormant season.  sources highly correlating with PC1, and 9 types with PC2 (P < 0.05, Table 5).   Table 6).

278
Furthermore, the plot also showed that various types of carbon-source utilized by 279 microorganisms in different patterns were quite similar, and concentrated in the center 280 of the axis (Fig 5). Based on the distribution distance between sites and carbon sources, Carbon-source code is the same as in Table 5.   in RDA ordination plot displayed that the substrates utilized by microorganisms in 309 different patterns were relatively discrete, and the sources that may be used by certain 310 communities were largely scattered around the positive half of the RDA1 axis (Fig 6).

311
Overall, amino acids, polymers and phenolic acids were increasingly utilized in RL, PL 312 and FL patterns, which were different from the metabolic characteristics in GL.  (Table in S1 Table). Carbon-source code is the same as in Table 5.

327
In the present study, the Biolog EcoPlates TM method was used to monitor the soil Consequently, not only the easily degraded compounds but also the complex ones were 368 consumed to satisfy the metabolic requirement of soil microorganisms. Adam et al. [12] 369 also reported that in spring and summer, the frequency of nitrogen-rich carbon sources All relevant data are within the paper and its Supporting Information files.

447
The authors have declared that no competing interests exist.  Table 3.