What is correlation and causation?
Correlations can be made in a few ways. One is by noticing that two environments are different due to one observed difference. An example of this could be two wastewater sequence batch reactors (SBR), one reactor has Enhanced Biological Phosphorus Removal (EBPR) occurring in it and the other does not have this process occurring. When examined, a difference in bacterial species would be found, thus a correlation would be made that the species of bacteria the EBPR reactor has and the other does not have, must be the bacteria that causes EBPR. Another type of correlation can be made through data analysis, for example an increase in putative Phosphorus Accumulating Organisms (PAOs) leading to an increase in phosphorus removal. The data presents a correlation between the two so it can be assumed that they have a relationship.
Causation involves experimentation and detailed analysis. For a correlative study to become a causative study it must – or attempt to – understand the method by which the correlation occurs. This could be achieved by gaining knowledge of a metabolic pathway by which a process occurs, or identifying the organism(s) responsible for the metabolic pathway by providing genomic or physical evidence that supports the correlation made.
An Example Study
A study titled ‘A Critical Assessment of the Microorganisms Proposed to be Important to Enhanced Biological Phosphorus Removal in Full-Scale Wastewater Treatment Systems’ by Stokholm-Bjerregaard et al. (2017), uses correlations to draw conclusions.
In this study, the abundance and diversity of a set of proposed glycogen accumulating organisms (GAOs) and PAOs was recorded by taking samples four times per year over a period of 9 years. The samples were put through DNA extraction and amplicon sequencing to organise into genus level. The study found that the microbial community composition was quite stable over time.
The classification of PAOs and GAOs in this study relies on previous studies, most of which were conducted by members of this study. There is limited evidence of the accuracy of such classification. Furthermore, the conclusions of the study are solely based on a genus level abundance statistic. Often it was assumed that the most numerically abundant genera were the most important bacteria in the EBPR process, however, no correlation between bacterial abundance and plant phosphorus removal performance was made or could be found in the data. Additionally, low abundance PAOs were assumed “numerically unimportant” to the EBPR process.
In order to attempt to become a causative study, it is suggested to further investigate the metabolic pathways of the putative PAOs and GAOs. Such pathways can be studied via means of transcriptomics or metabolomics, combined with database comparison in order to identify the key organisms of EBPR and the reaction pathways. Further experimentation should be conducted to quantify the importance of abundance of certain bacterial genera to EBPR. PAOs and GAOs could potentially be identified via comparison with polyphosphate and glycogen identifying stains and FISH probe images. Such a process is conducted in a study by Crocetti et al. (1999) in ‘Identification of Polyphosphate-Accumulating Organisms and Design of 16S rRNA-Directed Probes for Their Detection and Quantitation’.
References
Crocetti, G., Hugenholtz, P., Bond, P., Schuler, A., Keller, J., Jenkins, D., Blackall, L. (1999). Identification of Polyphosphate-Accumulating Organisms and Design of 16S rRNA-Directed Probes for Their Detection and Quantitation. American Society for Microbiology, 66 (3), 1175-1182. doi: 10.1128/AEM.66.3.1175-1182.2000
Stokholm-Bjerregaard, M., McIlroy, S. J., Mierychlo, M., Karst, S. M., Ablertsen, M., Nielsen, P. H. (2017). A Critical Assesment of the Microorganisms Proposed to be Important to Enhanced Biological Phosphorus Removal in Full-Scale Wastewater Treatment Systems. Frontiers in Microbiology, 8(718), 1-18. doi: 10.3389/fmicb.2017.00718
Microbiomes in Aerobic Watewater Treatment 