Department of Civil, Construction and Environmental Engineering, Marquette University, P.O. Box 1881, Milwaukee, WI, 53233, USA.
LBE, INRA, 102 Avenue des Etangs, Narbonne, F-11100, France.
Water Res. 2017 Mar 1;110:161-169. doi: 10.1016/j.watres.2016.12.010. Epub 2016 Dec 11.
A quantitative structure activity relationship (QSAR) between relative abundance values and digester methane production rate was developed. For this, 50 triplicate anaerobic digester sets (150 total digesters) were each seeded with different methanogenic biomass samples obtained from full-scale, engineered methanogenic systems. Although all digesters were operated identically for at least 5 solids retention times (SRTs), their quasi steady-state function varied significantly, with average daily methane production rates ranging from 0.09 ± 0.004 to 1 ± 0.05 L-CH/L-day (L = Liter of reactor volume) (average ± standard deviation). Digester microbial community structure was analyzed using more than 4.1 million partial 16S rRNA gene sequences of Archaea and Bacteria. At the genus level, 1300 operational taxonomic units (OTUs) were observed across all digesters, whereas each digester contained 158 ± 27 OTUs. Digester function did not correlate with typical biomass descriptors such as volatile suspended solids (VSS) concentration, microbial richness, diversity or evenness indices. However, methane production rate did correlate notably with relative abundances of one Archaeal and nine Bacterial OTUs. These relative abundances were used as descriptors to develop a multiple linear regression (MLR) QSAR equation to predict methane production rates solely based on microbial community data. The model explained over 66% of the variance in the experimental data set based on 149 anaerobic digesters with a standard error of 0.12 L-CH/L-day. This study provides a framework to relate engineered process function and microbial community composition which can be further expanded to include different feed stocks and digester operating conditions in order to develop a more robust QSAR model.
建立了相对丰度值与消化器甲烷产生速率之间的定量结构-活性关系（QSAR）。为此，将 50 个重复的厌氧消化器组（总共 150 个消化器）中的每个消化器都接种了来自规模化工程甲烷生成系统的不同产甲烷生物量样本。尽管所有消化器都以相同的方式至少操作了 5 个固体停留时间（SRT），但其准稳态功能变化很大，平均每日甲烷产生率从 0.09±0.004 到 1±0.05 L-CH/L-day（L = 反应器体积的升）（平均值±标准偏差）。使用超过 410 万个古菌和细菌的 16S rRNA 基因序列的部分序列分析了消化器微生物群落结构。在属水平上，观察到所有消化器中共有 1300 个操作分类单元（OTU），而每个消化器包含 158±27 OTU。消化器功能与典型的生物量描述符（如挥发性悬浮固体（VSS）浓度、微生物丰富度、多样性或均匀度指数）没有相关性。然而，甲烷产生率与一个古菌和九个细菌 OTU 的相对丰度显著相关。这些相对丰度被用作描述符，以开发一个多元线性回归（MLR）QSAR 方程，仅根据微生物群落数据预测甲烷产生速率。该模型基于 149 个厌氧消化器解释了实验数据集 66%以上的方差，标准误差为 0.12 L-CH/L-day。该研究提供了一个将工程化过程功能与微生物群落组成联系起来的框架，可以进一步扩展到包括不同的饲料和消化器操作条件，以开发更稳健的 QSAR 模型。
