RESUMO
Much remains unknown about the relationships between microbial community structure and anaerobic digester function. However, knowledge of links between community structure and function, such as specific methanogenic activity (SMA) and COD removal rate, are valuable to improve anaerobic bioprocesses. In this work, quantitative structure-activity relationships (QSARs) were developed using multiple linear regression (MLR) to predict SMA using methanogen community structure descriptors for 49 cultures. Community descriptors were DGGE demeaned standardized band intensities for amplicons of a methanogen functional gene (mcrA). First, predictive accuracy of MLR QSARs was assessed using cross validation with training (n = 30) and test sets (n = 19) for glucose and propionate SMA data. MLR equations correlating band intensities and SMA demonstrated good predictability for glucose (q(2) = 0.54) and propionate (q(2) = 0.53). Subsequently, data from all 49 cultures were used to develop QSARs to predict SMA values. Higher intensities of two bands were correlated with higher SMA values; high abundance of methanogens associated with these two bands should be encouraged to attain high SMA values. QSARs are helpful tools to identify key microorganisms or to study and improve many bioprocesses. Development of new, more robust QSARs is encouraged for anaerobic digestion or other bioprocesses, including nitrification, nitritation, denitrification, anaerobic ammonium oxidation, and enhanced biological phosphorus removal.