­Figure 3 – Top-down nutrient treatments inhibit and enrich specific microbes and their metabolism. A) Microbiome composition of microcosms with and without treatment across days 3 – 14. B) H₂S headspace measurements of microcosms after 14 days of treatment, T-tests performed with respect to None except where indicated. Error bars represent standard deviation, n=4 (except for “None”, n=3); * = p < 0.05, ** = p < 0.005, and *** = p < 0.001. Enterobacteriaceae * andLachnospiraceae * indicate unspecified genera within these two Families.

To evaluate how persistent the impacts of these treatments are, we passaged the communities into media that contained the same, different, or no supplements. After three days of growth, microcosms were passaged three separate times into fresh media with and without supplements. Subcultures of untreated communities with either no supplement or only nitrate produced moderate amounts of H₂S although lower than the untreated parent culture, while molybdate treatment of untreated subcultures matched that of the initial molybdate-treated cultures with little to no H₂S formation after the passages (≤7 ppm H₂S) (Figure 4, Supplemental Figure 5A ). When cultures were passaged from nitrate treatments and selection pressure was removed or kept the same (Figure 4B ; no supplement or NO₃^-1 only, respectively), the H₂S production resumed and remained a little lower than the level of the parent culture (~100 ppm). This suggests that the nitrate treatment may have enriched for H₂S-reducing bacteria or at best suppressed their metabolism initially. However, if nitrate selection is applied to molybdate cultures, H₂S levels remained below that of nitrate treatment of the native microbiome. Therefore, we suspect that the reduced alpha diversity of molybdate treatment (Figure 2 ) reflects a direct inhibition or loss of many of the sulfur-reducing species, which are either temporarily limited or possibly enriched by the nitrate treatment. As before, molybdate was able to limit the H₂S production of the sulfate-reducing communities to very low levels (Figure 4B ), further demonstrating that it was effective even against actively H₂S-generating communities like those from the nitrate-treated group that otherwise produce H₂S (Figure 4C ). Molybdate-treated subcultures, regardless of parent culture, generated less H₂S (Supplemental Figures 5A and D ). Interestingly, when molybdate microcosms were subcultured and had their selection pressure removed (Figure 4C , Supplemental Figure 5B and C ), H₂S was produced although at low levels (<25 ppm) with each sequential passage producing more H₂S demonstrating that continuous application of this selection pressure is needed to permanently suppress H₂S. This suggest that H₂S-generation is inhibited by molybdate, but the associated microbes may still be present at low levels. Lastly, subcultures of microcosms generated from combination treatment (Figure 4D ) behaved like molybdate subcultures for combined supplement and molybdate passages. However, after three passages, H₂S generation of these communities treated with nitrate began to return to higher levels. This suggests that the supplemented nitrate substrate in the combination treatment may have continued enriching for nitrate consumers, specifically some that harbor both nitrate and sulfate reducing pathways, so that sulfate reduction resumed once the nitrate was consumed. Throughout the subculture cultivations, the microbial community compositions were similar to their respective nutrient screening microcosms which were dominated by Enterobacteriaceae, Bacteroides, and Lachnoclostridium (Figure 3, Supplemental Figure 6 ). Ultimately, the combination of molybdate and nitrate best limits H₂S (Supplemental Figure 5E ) and provides some redundancy of inhibitors in the case one of the selection pressures is resisted or lost; however, molybdate is the most important for effectively limiting H₂S production from these oil well communities (Supplemental Figure 5D ). From this in vitro screening system, we showed that oil well microbial communities can be engineered and controlled from the top down using nitrate and molybdate, structural analogs of sulfate, to inhibit sulfate-reducing bacteria and H₂S generation.