Fig. S1. Locations of the 11 CTD casts collected for this study in 2009 and 2010. Mound Culebra (Station 5; see Table S1): AT15-44 CTD10, 1515 m seafloor depth. Jaco Scarp (Station 6): AT15-44 CTD12, 2233 m; AT15-59 CTD11, 1783 m; and AT15-59 CTD12, 753 m. Mound Quepos (Station 4): AT15-44 CTD7, 1184 m. Mound 11 (Station 7): AT15-59 CTD5, 1004 m. Mound 12 (Station 1): AT15-44 CTD1, 1019 m; AT15-59 CTD1, 995 m; and AT15-59 CTD7, 993 m. Off-mound NE of Mound 12 (Station 2): AT15-44 CTD4, 813 m and off-mound SW of Mound 12 (Station 3): AT15-44 CTD5, 1727 m.


Fig. S2. Cu-MMO-encoding community profile analysis for Mound 12 (2010). Two casts during cruise AT15-59 were obtained 3 days apart at similar coordinates above Mound 12. The similarity between these independent depth profiles demonstrates that MISA analysis reproducibly assesses Cu-MMO-encoding gene distributions. Asterisks denote samples with supplemental clone library information.

A. AT15-59 CTD1, collected 6 January 2010. MISA profile in the methane-enriched bottom waters indicates a more diverse cluster of Cu-MMO-encoding genotypes than higher in the water column.

B. AT15-59 CTD7, collected 9 January 2010. Cu-MMO-encoding genotype population shifts were reproducible through water depth in these independent casts.


Fig. S3. Cu-MMO-encoding community profile analysis for Mound 12 and Mound Quepos (2009).

A. Mound 12, AT15-44 CTD1, collected 21 February 2009. MISA profiles through water depth indicated similar Cu-MMO-encoding genotype population shifts as measured throughout the CR-OMZ. A shift in OPU1 genotypes was detected at 200 and 350 m water depth and verified with clone library analysis (asterisk denotes sample with clone library information).

B. AT15-44 Alvin dive AD4503, collected 24 February 2009. AD4503 water sample indicated that MISA peaks characteristic of Groups X and W were present near a tube worm community associated with CTD1; Groups W and X were not detected in significant relative abundance elsewhere in the CR-OMZ.

C. AT15-44 CTD7, collected above Mound Quepos on 27 February 2009. The population shift in OPU1 genotypes measured above Mound 12 was also observed above Mound Quepos at ∼300 m water depth, as indicated by MISA analysis.

D. AT15-44 Alvin dive AD4512, representative of the seafloor lying within the oxygen minimum at the Quepos landslide. Bottom water within the oxygen minimum indicated that low oxygen concentrations may be a driving factor for a Cu-MMO-encoding genotype community composition dominated by OPU3, despite proximity to the seafloor.


Fig. S4. Distance tree of Methylococcales-related 16S rRNA genes recovered in this study, inferred by Neighbour-joining with the Jukes and Cantor model used to estimate distances using the ARB database SSURef-108-SILVA-NR ( and the provided bacterial filter. Clones from this study are shown in boldface type; 99% identical 16S rRNA gene sequences have been recovered from environments similar to the Costa Rica OMZ, including within low-oxygen waters in the northern Pacific and within the Gulf of Mexico following the 2010 Deepwater Horizon oil spill. Bootstrap values were obtained in PAUP* 4.0b10 by transferring aligned and masked sequences from ARB and then running three algorithms (i) Neighbour-joining, Jukes-Cantor correction, 1000 replicates, (ii) Parsimony, 100 replicates and (iii) Maximum Likelihood, 10 replicates. Nodes with bootstrap support of 60 or greater are show as closed circles when supported by all three methods and open circles when supported by two methods. Sequences from Aquifex pyrophilus (M83548), Hydrogenobacter thermophilus (AP011112) and Hydrogenobaculum acidophilum (D16296) served as an out-group to root the tree. Scale bar represents 0.10 substitutions per site. Sequences for 16S rRNA genes recovered in this study were deposited to GenBank (accession numbers JX569113–JX569145).


Fig. S5. Abundances of 16S rRNA gene sequences relevant to groups OPU1 and OPU3 obtained using qPCR on water samples from AT15-59 CTD11. Oxygen and methane concentrations are indicated; methane concentrations stated as a range of values indicate multiple sequential water samples in which methane concentrations were measured in the indicated range. Taqman assays were performed in triplicate; standard deviation was typically 10–30% of averaged abundance estimate. 16S rRNA gene candidates were present through depth with trends similar to those measured for the pmoA gene targets for groups OPU1 and OPU3 (Fig. 2), typically at higher absolute abundances.


Fig. S6. Reduction in OPU abundance in two casts in the microoxic zone. In casts with high-frequency sampling within the microoxic core of the OMZ (three of 11 casts in this study: AT15-59 CTD7, 11 and 12), groups OPU1 and OPU3 show a reduction in abundance within the microoxic zone (∼400 mbsl in each cast).


Fig. S7. Temporal stability of OPU1, OPU3 and a previously undescribed Methylococcaceae-related Cu-MMO-encoding sequence near Jaco Scarp. AT15-44, CTD12 collected 4 March 2009 and AT15-59, CTD11 collected 11 January 2010 are both located along Jaco Scarp, approximately 3.5 km apart. The distinct MISA pattern for both samples indicates a persistence of the three genotypes at Jaco Scarp year-over-year.


Fig. S8. Absolute abundances of Cu-MMO-encoding gene targets determined by qPCR for OPU1, OPU3 and a water-column pxmA sequence with increasing depth at Jaco Scarp (CTD11). Primers to recover pxmABC from the water column targeted conserved sites in pxmA and pxmC and are listed in Table S1. A Taqman assay to enumerate the recovered pxmA sequence was developed using Primer Express v2.0. The pxm sequence detected in this study was present in comparable abundances as group OPU1 in CTD11 and at all stations in the study site (not shown).


Fig. S9. Relationship between Groups OPU1 and OPU3 abundances and methane concentration. Methane measurements are described in Table S1; methane was detectable in 24 samples on cruise AT15-59. Groups OPU1 and OPU3 were most abundant in samples with methane values below ∼50 nM. The box and whisker plots provide statistics for OPU1 and OPU3 abundance in samples (n = 33) with methane concentrations below the detection limit (0.5 nM).


Fig. S10. Cu-MMO-A phylogeny. Sequences analysed in this study from the marine water column are indicated by red text. pmoA gene sequences from Groups OPU1 and OPU3, and another previously undescribed pmo sequence recovered near Jaco Scarp, affiliated with pmoA sequences from Methylococcaceae, which is consistent with a role in methane oxidation. Cu-MMO sequences from the marine water column affiliated with a variety of proteobacterial Cu-MMO-encoding lineages implicated in the oxidation of ammonia, methane or higher-order hydrocarbons.


Table S1. Summary of CTD casts in this study.


Table S2. Primers and PCR conditions used in this study.


Table S3. Pearson product moment correlation between Cu-MMO abundances and key environmental variables.


Table S4. 16S rRNA genes recovered in this study.

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.