A) Map of the coring locations within the Plum Island Estuary Long Term Ecological Research (PIE LTER) site and B) Details of core sampling and the bioinformatic approaches used in this study. The arrows flow from the sampling approach through genome reconstruction, network analysis, identification of a cooccurring subnetwork, and the metabolic complementarity and variant detection of that subnetwork. The primary bioinformatic tool is shown for each step.

The tree at the center of the display shows the phylogenetic relationship among all MAGs.

The innermost ring is a histogram of the single copy gene completion estimates (range = 0-100%), followed by redundancy (range = 0-10%), and total length of the genome (range = 0-7.6Mbp). The “Depth proportion” ring shows the proportion of the total reads mapped to a MAG from each depth. The outermost ring shows the phylum level taxonomy as a color bar with matching text. Arrows are positioned adjacent to phyla names with split positions in the tree.

Summary of MAG relative abundance and ordination.

In the ordination panel A) each point represents a sample colored according to depth and the shape indicates creek location. B) The results of dispersion tests and Andonis2 PERMANOVA. C) Phylum level composition for each core, with cores collected from Sweeney Creek on the left and West Creek on the right.

A.) Random matrix theory derived cooccurrence network, based on MAG relative abundance. The point size scales with genome size, the color indicates the phylum classification, and the shape identifies the module. All connections are significant and positive. Ellipses are used to clarify the modules. B.) Cumulative relative abundance of MAGs in network modules with eight or more members. The color of the bars are used to draw attention to the patterns and align with the module ellipses C.) The number of complete pathways plotted for each MAG and box-whisker plots for the median, 25th and 75th percentile lines for each of the modules of interest.

Functions detected within each of the network modules, with individual functions on the left and functional category on the right.

The bars are colored according to the functional category. Each of the separate bar plots represents the proportion of MAGs that contain the function within a network module. The modules are listed in the same order as Fig. 4, where the collective relative abundance distribution of MAGs goes from shallower to deeper (left to right).

The percent relative abundance of MAGs in the Bathyarchaeia BA1 subnetwork community relative to all mapped reads in the metagenome and metatranscriptome samples in each of the two deepest cores.

Only MAGs identified as abundant (mean %relative abundance > 0.1) were shown for the metagenome data and only those MAGs were displayed for the metatranscriptome data. This represents 25 out of the 33 MAGs in the subnetwork.

Bathyarchaeia BA1 subnetwork of metagenomic (MG) and metatranscriptomic (MT) functions.

A.) The phylogenetic relationship among MAGs according to single copy core genes with family and phylum level classification. Stars are used to highlight several MAGs that have high metabolic weight in the MG and MT data. B.) Heat map for the metabolic weight contribution in the MG and MT to each of the functional categories displayed on the right-hand side of the figure. The blue bars represent the cumulative metabolic weight of the functional category within either the MG or MT data. C.) The relative abundance of the starred MAGs from panel A. Corresponding cores are shown for each dataset and labeled within each individual panel. Relative abundance in this case is the percentage of reads mapped to that MAG relative to all mapped reads.

Variability ((number of variants/genome length)*1000) and sample specific variant clustering of six MAGs from the subsurface core community (A-F).

The analysis was limited to samples where the mean coverage was > 15X. Each panel contains a plot of the variability with depth. Separate lines are drawn for each of the two cores (blue = Sweeney Creek, green = West Creek). Ordination based on SNV variant frequency at 5000 randomly chosen sites in the genome. Geometric shapes and lines to the centroid are shown for the two cores and color coded according to creek. Statistics for the ordination are located in Table S8.