FIGURE 2.

Example of using iterative operations of the Framework for Integrated, Conceptual, and Systematic Microbial Ecology (FICSME) to study nitrous oxide off-gassing. (A) Example experimental cycle and (B) overall FICSME process diagram. (A) We exemplify the process of applying FICSME with an open question in microbial ecology about determining the biogeochemical controls on nitrous oxide off-gassing from nitrate-contaminated sediments. The figure depicts one experimental cycle consisting of a research question or problem, hypothesis, series of five experiments across three scales, integration of data into the model, evaluation of results, opportunity to iterate, or move to outcome. For further details, please see the Tutorial in the Supplementary Material, which includes Supplementary Figure 1 and Supplementary Tables 2, 3. (B) Applying the FICSME follows the same guiding principles as the scientific method but incorporates consideration of the FICSME terms at every step. First, the researcher determines what problem they want to study and poses a research question. Then, the researcher will state a specific testable hypothesis, as the FICSME can be used iteratively to address multiple hypotheses and processes that may constitute a larger overarching research question (Step 1). Second, the researcher selects the FICSME terms that are needed to test their hypothesis; this may include removing irrelevant terms from the FICSME or adding terms from other models as appropriate. Then the researcher performs a literature review and checks databases for existing results and data that may satisfy a selected term. The researcher should then populate the FICSME selected terms with these data and identify the knowledge gaps. Next, the researcher will design experiments to fill the identified knowledge gaps and populate the corresponding terms in the FICSME. Each experiment follows the general flow of stating an experimental hypothesis, testing, predicting, and evaluating the result (Step 2). The experiment can be conducted at field scale or in situ (Step 2a); at the mesocosm level, which can occur in the field or in the laboratory (Step 2b); or at the isolated molecules level in the laboratory (Step 2c). The FICSME workflow can start at any of these levels and can iterate from one level to any other level (horizontal double arrows). Within each level of experimentation, there are three categories of experiments that can be performed, again in any order, and all might not be required to obtain resolution sufficient for the research question. The three categories are (1) survey or identification and quantification (Steps 2a.1, 2b.1, and 2c.1), (2) dynamics and kinetics (Steps 2a.2, 2b.2, and 2c.2), and (3) interactions and connections (Steps 2a.3, 2b.3, and 2c.3) and are defined for each level of analysis in the figure. Third, the data are collected and the results of individual experiments are evaluated, the data are integrated across scales and techniques, and the total findings are populated into the FICSME (Step 3). Fourth, the collective understanding is used to pose a mechanism giving rise to the target phenotype (Step 4). The mechanism should be tested by performing an experiment from Step 2. This will likely require several iterative cycles to refine the model and prediction. Once the mechanism accurately predicts the system well enough, then the researcher can stop; or fifth, use the quantitative results from the FICSME workflow to intervene in the system to induce the outcome that solves the initial problem identified at the beginning (Step 5).