Timescales of variation in diversity and production of bacterioplankton assemblages in the Lower Mississippi River.

Rivers are characterized by rapid and continuous one-way directional fluxes of flowing, aqueous habitat, chemicals, suspended particles, and resident plankton. Therefore, at any particular location in such systems there is the potential for continuous, and possibly abrupt, changes in diversity and metabolic activities of suspended biota. As microorganisms are the principal catalysts of organic matter degradation and nutrient cycling in rivers, examination of their assemblage dynamics is fundamental to understanding system-level biogeochemical patterns and processes. However, there is little known of the dynamics of microbial assemblage composition or production of large rivers along a time interval gradient. We quantified variation in alpha and beta diversity and production of particle-associated and free-living bacterioplankton assemblages collected at a single site on the Lower Mississippi River (LMR), the final segment of the largest river system in North America. Samples were collected at timescales ranging from days to weeks to months up to a year. For both alpha and beta diversity, there were similar patterns of temporal variation in particle-associated and free-living assemblages. Alpha diversity, while always higher on particles, varied as much at a daily as at a monthly timescale. Beta diversity, in contrast, gradually increased with time interval of sampling, peaking between samples collected 180 days apart, before gradually declining between samples collected up to one year apart. The primary environmental driver of the temporal pattern in beta diversity was temperature, followed by dissolved nitrogen and chlorophyll a concentrations. Particle-associated bacterial production corresponded strongly to temperature, while free-living production was much lower and constant over time. We conclude that particle-associated and free-living bacterioplankton assemblages of the LMR vary in richness, composition, and production at distinct timescales in response to differing sets of environmental factors. This is the first temporal longitudinal study of microbial assemblage structure and dynamics in the LMR.

Introduction

In small streams, because of frequent and pronounced environmental disturbances in phyclass="Chemical">sical aclass="Chemical">nd chemical coclass="Chemical">nditioclass="Chemical">ns, variatioclass="Chemical">n iclass="Chemical">n microbial assemblage structure may be uclass="Chemical">nrelated to timescale so that assemblages sampled closer iclass="Chemical">n time may be as disclass="Chemical">n class="Chemical">similar as those sampled months apart [1]. In less stochastically disturbed aquatic systems, however, microbial assemblages appear to vary more predictably, and over the same temporal scales in which there is variation in diversity and/or activity of annual plant and animal assemblages [2]. For example, seasonally recurrent bacterioplankton assemblages have been observed in temperate marine environments [3, 4], lakes [5, 6], and even large rivers [7-10] associated with variation in day length, water temperature, hydrology, and nutrient concentrations.

Large river ecosystems of temperate zones are characterized by substantial temporal variation in nutrient and suspended sediment loads that is governed by their individual hydrographical underpinnings [11, 12]. At any given class="Chemical">site withiclass="Chemical">n these systems, eclass="Chemical">nviroclass="Chemical">nmeclass="Chemical">ntal fluctuatioclass="Chemical">n may be class="Chemical">n class="Disease">abrupt and unpredictable over brief periods of time responding to local storm events, or relatively gradual and deterministic due to climatic changes in temperature and/or precipitation within and among regional watersheds. Temporal dynamics of bacterial communities have been well described for many aquatic ecosystems, yet temporal variability in bacterioplankton assemblages of large rivers remains understudied. This is a significant gap in our knowledge of large river ecology, because of the importance of large rivers as conduits of nutrients to the sea [13]; because, as in other environments, bacteria are the most versatile and presumably the most important catalysts of biogeochemical transformations [14]; and because bacteria can reproduce rapidly and their community composition respond to environmental changes on a short-term basis [15].

From previous studies of the Misclass="Chemical">sisclass="Chemical">n class="Chemical">sippi River network, a system of multiple linked large rivers, we observed consistent and pronounced spatial variation in bacterioplankton assemblages. At a microhabitat level, assemblages attached to suspended particles (i.e. particle-associated bacterioplankton) were richer in bacterial operational taxonomic units (OTUs), and distinct in composition compared to free-living bacterioplankton [16, 17]. At a regional level, assemblages in major tributaries of the Mississippi River—the Illinois, Missouri, and Ohio rivers—were distinct in composition, presumably due to selection by particular environmental conditions of each river [16, 17]. Within the Mississippi River itself, planktonic microbial assemblages flowing downstream exhibited relatively large shifts in diversity after mixing at major confluences, while varying more gradually with increasing distance from confluences [17]. Clearly, as for other aquatic ecosystems, environmental selection processes structure bacterioplankton assemblages of this river network. However, in what taxonomic groups, of what magnitude, over what temporal scales, and in response to exactly what factors do assemblage changes occur? For instance, if one were to sample continuously over time at a single location in a large river water-column, in what respects and in concert with what environmental conditions, would the microbial plankton community vary? These questions address the relative importance to microbial community diversity and activity of stochastic variation over short time periods compared to over longer timeframes, in the context of an ecosystem marked by continuous, directional fluxes of water, chemicals, suspended materials, and microorganisms.

To address these questions, we class="Chemical">documeclass="Chemical">nted variatioclass="Chemical">n iclass="Chemical">n alpha diverclass="Chemical">n class="Chemical">sity (within-sample richness of OTUs) and beta diversity (between-sample differences in composition) within and between particle-associated and free-living bacterioplankton assemblages over a range of temporal scales at a single site in the main channel of the Lower Mississippi River, the final segment of the largest river system in North America. Assemblages were collected on a daily and weekly basis in summer, and monthly over a year. Additionally, on each sampling date, we measured bacterial production and environmental variables. From these measurements, we determined the relationship of timescale to variation in assemblage diversity and production, and identified the strongest environmental correlates of variation. We hypothesized that bacterioplankton diversity and production of the LMR would vary less over shorter timescales and more over longer timescales, in relationship to gradual change in factors such as temperature, suspended sediments, algal biomass, and nutrient concentrations.

Materials and methods

Sample site and water collection

The Lower Misclass="Chemical">sisclass="Chemical">n class="Chemical">sippi River (LMR) was sampled on 23 dates between February 2013 and January 2014 (Fig 1A), near mid-channel directly off Mhoon Landing (34°44'35.59" N 90°26'58.03" W), near Tunica, Mississippi, USA (Fig 1B). Mhoon Landing is 76 river kilometers (rkm) below Memphis, Tennessee, and 426 rkm below Cairo, Illinois, where the Ohio River joins the Mississippi River, forming the LMR. At the Mhoon Landing sampling location the river is turbulent and deep (>7 m) with little evidence of vertical stratification in dissolved chemistry [18], and discharge generally ranges from roughly 7,000 to 27,000 m3 s-1 [19] depending on time of year (Fig 1A).

Fig 1

Hydrograph of discharge of the Lower Mississippi River at Mhoon Landing, Mississippi between February 2013 and January 2014 (A). Points on hydrograph represent sample dates. Monthly sample dates (n = 12) are labeled by date, while horizontal bars indicate weekly (3 June to 15 July 2013, n = 7) and daily sampling (24 June to 1 July 2013, n = 8) periods. Discharge measurements were calculated using gage height data collected daily by the U.S. Army Corps of Engineers at Helena, Arkansas located 40 rkm below Mhoon Landing. Map of a portion of the Mississippi River Basin indicating sample location (Mhoon Landing) relative to Memphis, Tennessee, and major river tributaries (B).

Hydrograph of discharge of the Lower Misclass="Chemical">sisclass="Chemical">n class="Chemical">sippi River at Mhoon Landing, Mississippi between February 2013 and January 2014 (A). Points on hydrograph represent sample dates. Monthly sample dates (n = 12) are labeled by date, while horizontal bars indicate weekly (3 June to 15 July 2013, n = 7) and daily sampling (24 June to 1 July 2013, n = 8) periods. Discharge measurements were calculated using gage height data collected daily by the U.S. Army Corps of Engineers at Helena, Arkansas located 40 rkm below Mhoon Landing. Map of a portion of the Mississippi River Basin indicating sample location (Mhoon Landing) relative to Memphis, Tennessee, and major river tributaries (B).

Sampling spanned three temporal scales (Fig 1A). Samples were collected once monthly, near the beginning of each calendar month, from 2 February 2013 to 11 January 2014, for a total of 12 monthly samples. At a finer scale, samples were collected weekly from 3 June to 15 July 2013, for a total of seven weekly samples. Finally, samples were collected daily from 24 June to 1 July 2013, for a total of eight daily samples. We chose to sample frequently during summer because this is a period of high bacterial production [18], and potentially a period in which a high degree of short-term temporal variation could be detected. On each date, sampling occurred between 10:00 and 13:00 h, and class="Chemical">water was collected from mid-river at a depth of 0.5 m. Sterilized 1-L class="Chemical">n class="Chemical">Nalgene sample bottles (n = 3) were used to collect water for chemical analyses and heterotrophic bacterial production, and sterilized 500-mL Nalgene sample bottles (n = 3) were used to collect water to analyze bacterioplankton assemblage structure. All bottles were stored in coolers containing river water to maintain ambient temperature during transportation to the laboratory (0.5–1.5 h) for additional measurements, sample fractionation, and preservation.

This field study did not involve endangered and protected species, and all samples used in this study were collected from a public river n class="Chemical">waterway for which permisclass="Chemical">n class="Chemical">sion to obtain samples was not required.

Environmental measurements

class="Chemical">Water temperature was measured iclass="Chemical">n the field uclass="Chemical">n class="Chemical">sing a Hawkeye Digital Sonar H22PX-B. In the laboratory, sub-samples (100–200 mL) were filtered through ashed 47-mm diameter, Whatman GF/F filters. For preservation, filters and filtrates were frozen at -60°C or -20°C, respectively. Samples remained frozen < 18 months prior to testing. Total suspended sediment (TSS) concentrations were measured gravimetrically on filters after drying at 60°C. Chlorophyll a (Chla) concentrations were assayed by spectrophotometry of pigments extracted in 90% NH4OH-buffered acetone for 24 h at 5°C [20]. Total dissolved organic C (DOC) and total dissolved N (TDN) were measured in filtrates using a Shimadzu Total Organic Carbon Autoanalyzer, while total dissolved P (TDP) concentrations were assessed using standard spectrophotometric methods [20]. Units for these environmental measures pertinent to all analyses are given in Fig 2.

Fig 2

Environmental variables measured in Lower Mississippi River water between February 2013 and January 2014.

Abbreviations: Temp, water temperature; TSS, total suspended solids; Chla, chlorophyll a; DOC, total dissolved organic carbon; TDN, total dissolved nitrogen; TDP, total dissolved phosphorus. Except for water temperature, parameter measurements are presented as means (± SE) for each date, n = 2–3. For clarity, sample dates are connected by lines. These lines are not intended to convey patterns of variation at shorter time intervals than what is shown.

Abbreviations: Temp, class="Chemical">water temperature; TSS, total suspeclass="Chemical">nded solids; class="Chemical">n class="Chemical">Chla, chlorophyll a; DOC, total dissolved organic carbon; TDN, total dissolved nitrogen; TDP, total dissolved phosphorus. Except for water temperature, parameter measurements are presented as means (± SE) for each date, n = 2–3. For clarity, sample dates are connected by lines. These lines are not intended to convey patterns of variation at shorter time intervals than what is shown.

DNA extraction and sequencing

From the 500-mL sample bottles, 100-mL subsamples were removed for serial filtration (<5 mm Hg vacuum). Subsamples were initially passed through sterile Millipore 3-μm pore-class="Chemical">size polyclass="Chemical">n class="Chemical">carbonate filters, and the filtrate immediately filtered through sterile Millipore 0.22-μm pore-size polyethersulfone filters. Particles collected in the first filtration include particle-associated cells, cells, or colonies >3 μm in size (hereafter referred to as particle-associated cells). Particles collected in the second filtration step include smaller (0.22–3 μm) bacteria, assumed to be mostly free-living [16, 17]. Filters were stored at -20°C before molecular processing. Samples remained frozen < 18 months prior to testing.

Dclass="Chemical">NA was extracted from filters uclass="Chemical">n class="Chemical">sing PowerWater DNA isolation kits (MoBio, Carlsbad, California). The bacterial 16S rRNA gene was amplified and sequenced using methods modified from Kozich et al. [21], and described previously [17, 22]. Briefly, DNA was amplified using standard forward (5’-GTGCCAGCMGCCGCGGTAA) and reverse (5’-GGACTACHVGGGTWTCTAAT) primers adapted with dual-index barcodes for Illumina MiSeq next generation sequencing [21], and run through 30 cycles of denaturation (95°C) for 20 s, annealing (55°C) for 15 s, and elongation (72°C) for 2 min, and a final elongation (72°C) for 10 min. Negative (no template) controls were used in all amplifications and consistently gave negative results. Such negative amplifications were also used as blanks in sequencing, yielding no sequence data. Positive controls were not needed as we have used these procedures successfully for a variety of sample types [17, 23, 24]. PCR products were normalized by sample using SequalPrep Normalization Plates (Life Technologies, Grand Island, New York), pooled, and sequenced using an Illumina MiSeq platform located at the Molecular and Genomics Core Facility at the University of Mississippi Medical Center. All sequences can be accessed in the NCBI SRA database under the BioProject ID PRJNA358603.

Sequence processing

Sequence data were processed uclass="Chemical">siclass="Chemical">ng the bioiclass="Chemical">nformatics software mothur [25] by a procedure modified from Payclass="Chemical">ne et al. [17]. Briefly, the class="Chemical">n class="Chemical">SILVA rRNA database (release 119) was used to align sequences with reference V4 sequences [26], and all unaligned sequences were discarded in addition to homopolymers >8 bp. Before classification, sequences differentiated by ≤2 bp were merged, and potential chimeras identified by UCHIME [27] removed. Sequences were classified using the RDP database (Release 11, September 2016) [28]. Non-bacterial lineages (e.g. Archaea, Eukarya, and mitochondria) were then removed. As RDP classification does not distinguish between cyanobacteria and chloroplast lineages at the phylum-level, chloroplast sequences were removed in a subsequent step (see below). Finally, all remaining sequences were clustered into OTUs based on ≥97% similarity.

Sequence data were processed further and analyzed in R verclass="Chemical">sioclass="Chemical">n 3.5.1 [29]. OTU aclass="Chemical">nd taxoclass="Chemical">nomy tables geclass="Chemical">nerated by mothur were imported iclass="Chemical">nto R aclass="Chemical">nd merged with eclass="Chemical">nviroclass="Chemical">nmeclass="Chemical">ntal metadata uclass="Chemical">n class="Chemical">sing the microbiome analysis software phyloseq version 1.14.0 [30]. OTUs identified as belonging to chloroplast lineages were removed from the dataset.

Alpha diverclass="Chemical">sity (i.e. richclass="Chemical">ness of bacterial OTUs withiclass="Chemical">n samples) was determiclass="Chemical">ned from aclass="Chemical">n uclass="Chemical">ntrimmed dataset (i.e. coclass="Chemical">ntaiclass="Chemical">niclass="Chemical">ng class="Chemical">n class="Chemical">singleton OTUs) using the phyloseq function “estimate_richness”. Beta diversity (i.e. differences in assemblage composition) was evaluated after removal of OTUs with fewer than one read in 10% of the samples (i.e. potentially erroneous and rare OTUs) were removed from the dataset. OTU counts were then normalized using edgeR [31].

Bacterial production measurements

Bacterial production was determined based on radiolabeled isotope incorporation. class="Chemical">Leucine (class="Chemical">n class="Chemical">3H-leucine) and thymidine (3H-thymidine) (Moravek Biochemicals) at specific activities of approximately 100 Ci mmol-1 were used to determine synthesis rates of proteins and DNA, respectively [20]. Production of the total assemblage was measured using whole-water samples, while production of free-living cells was measured in sample water filtered through sterile 47-mm diameter, 3-μm pore-size Millipore polycarbonate filters [18].

Production measurements were made uclass="Chemical">siclass="Chemical">ng a microceclass="Chemical">ntrifuge procedure modified from Kirchmaclass="Chemical">n [32]. Triplicate bulk aclass="Chemical">nd filtered class="Chemical">n class="Chemical">water samples (1.5 mL) were added to 2-mL microcentrifuge vials along with a saturating concentration of 60 nM 3H-leucine or 3H-thymidine [18]. A control tube for every treatment was prepared by adding trichoroacetic acid (TCA) immediately after isotope addition (see below). Thus, there were a total of 16 vials used per sample event. Incubations were initiated in the field beginning immediately after sample collection. Vials were incubated in river water at ambient temperature for 1 h, then placed on ice for 5 min, after which 94 μL of 80% TCA was added to halt isotope uptake. In the laboratory, vials were centrifuged at 18,000 rpm for 10 min, and the supernatant removed. Cold 5% TCA (1 mL) was then added to each vial followed by vortexing, centrifugation, and removal of supernatant. Finally, 1 mL of ice-cold 80% ethanol was added, followed by the washing steps above. Pellets were dried at room temperature overnight, and 1 mL of Fisher ScintiSafe Plus 50% scintillation fluid added to vials, followed by further vortexing. Radioassays were run on a Perkin-Elmer Tri-Carb 2810 TR liquid scintillation counter. Radioisotope-uptake calculations for 3H-leucine representing biomass production, and 3H-thymidine representing cell reproduction, were made as explained in Wetzel and Likens [20]. Production of all cells (whole-water) and free-living cells (<3-μm fraction) was determined directly, while production of particle-associated cells was determined by difference.

Statistical analysis

Univariate statistics were performed un class="Chemical">siclass="Chemical">ng the package car [33], while multivariate statistics were performed uclass="Chemical">n class="Chemical">sing either phyloseq or vegan version 2.5–3 [34]. Graphics were generated using ggplot2 version 2.1.0 [35].

Levene’s Test was used to detect homogeneity of variance in bacterial alpha diverclass="Chemical">sity betweeclass="Chemical">n particle-associated aclass="Chemical">nd free-liviclass="Chemical">ng samples. Variaclass="Chemical">nce iclass="Chemical">n assemblage alpha diverclass="Chemical">n class="Chemical">sity, beta diversity, and production between samples collected over daily (24-Jun– 1-Jul, n = 8), weekly (3-Jun– 15-Jul, n = 7), and monthly (2-Feb– 11-Jan, n = 12) sampling intervals were shown using boxplots. Mood’s median tests were used to compare the medians. Post-hoc tests were run using the function “pairwiseMedianTest” in the rcompanion package [36].

Beta diverclass="Chemical">sity was quaclass="Chemical">ntified uclass="Chemical">n class="Chemical">sing Bray-Curtis dissimilarity matrices. To visualize whether bacterial samples collected closer in time were more similar in composition, mean pairwise dissimilarities were plotted against Euclidian distances in sample date. Differences in composition between particle-associated and free-living samples were also visualized using non-metric multidimensional scaling (NMDS) ordinations. Envfit (package vegan) analysis was then used to determine abundant bacterial OTUs that correlated with separation of samples in NMDS space.

Permutational multivariate analyclass="Chemical">sis of variaclass="Chemical">nce (fuclass="Chemical">nctioclass="Chemical">n “adoclass="Chemical">nis” iclass="Chemical">n the package vegaclass="Chemical">n) was used to test for class="Chemical">n class="Chemical">significant differences in beta diversity between groups of samples (e.g. between particle-associated and free-living, or between samples collected at daily, weekly, and monthly timescales) [37]. Permutated distance-based test for homogeneity of multivariate dispersion (function “PERMDISP2” in the package vegan) was then used to test for significant differences in the variance in beta diversity between sample groupings [38].

Environmental drivers of particle-associated and free-living beta diverclass="Chemical">sity were determiclass="Chemical">ned by model selectioclass="Chemical">n uclass="Chemical">n class="Chemical">sing corrected Akaike information criterion (AICc) [39] in the software Plymouth Routines in Multivariate Ecological Research (PRIMER) 7.0 [40]. Environmental variables in models included: temperature, TSS, Chla, DOC, TDN, TDP, and discharge. Prior to AICc analysis, a cross-correlation matrix analysis of candidate predictors was performed. Predictors having a correlation coefficient ≥ 0.8 were not both included in the model for community composition. Relative variable importance (RVI) scores were calculated for each environmental variable based on appearance in the AICc-best models, and a pseudo-R2 was calculated for the best models to quantify their fit to the data. Variables that had RVI > 0.5 were considered most important.

Samples were also used to assess patterns of variation in relative abundances of bacterial OTUs. Plots were created in package ggplot2 uclass="Chemical">siclass="Chemical">ng the fuclass="Chemical">nctioclass="Chemical">n “stat_smooth”. Local polyclass="Chemical">nomial regresclass="Chemical">n class="Chemical">sion fitting (function “loess” in the package ggplot2) was used to display patterns of variation in relative abundances. 95% confidence intervals were plotted around regression lines.

Results

Patterns in the river environment

Over the course of the study, class="Chemical">water temperature raclass="Chemical">nged from 5°C oclass="Chemical">n 11-Jaclass="Chemical">nuary to 30°C oclass="Chemical">n 29-Juclass="Chemical">ne aclass="Chemical">nd 9-September (Fig 2). TSS coclass="Chemical">nceclass="Chemical">ntratioclass="Chemical">ns peaked duriclass="Chemical">ng high discharge oclass="Chemical">n 6-May aclass="Chemical">nd 8-July, while class="Chemical">n class="Chemical">Chla concentrations were at a maximum during low discharge on 5-November. TDN corresponded closely to the pattern in the river hydrograph (r = 0.70, p = 0.01). DOC (r = 0.42, p = 0.26) and TDP (r = 0.33, p = 0.28), in contrast, did not vary with discharge. Seasonal and annual variability of these variables in the LMR are tightly coupled with climatic and hydrologic conditions inherent to the river’s large watershed, as documented previously [18, 41, 42].

To compare patterns in the timescales of variation, for each environmental variable we calculated the coefficient of variation (CV) for measurements taken over daily (24-Jun– 1-Jul, n = 8), weekly (3-Jun– 15-Jul, n = 7), and monthly (2-Feb– 11-Jan, n = 12) sampling intervals. For all variables, relative variation increased with timescale of measurement (Table 1).

Table 1

Coefficient of variation (%) in environmental variables at daily, weekly, and monthly timescales.

Variable	Daily (n = 8)	Weekly (n = 7)	Monthly (n = 12)
Temp	4	9	55
TSS	14	41	64
Chla	27	33	44
DOC	6	7	13
TDN	2	11	51
TDP	8	23	23
Discharge	12	20	60

Abbreviations: Temp, water temperature; TSS, total suspended solids; Chla, chlorophyll a; DOC, total dissolved organic carbon; TDN, total dissolved nitrogen; TDP, total dissolved phosphorus.

n represents the number of dates per sampling interval.

Abbreviations: Temp, n class="Chemical">water temperature; TSS, total suspeclass="Chemical">nded solids; class="Chemical">n class="Chemical">Chla, chlorophyll a; DOC, total dissolved organic carbon; TDN, total dissolved nitrogen; TDP, total dissolved phosphorus.

Patterns in bacterial alpha diversity

A total of 4,774,499 bacterial sequences were recovered from particle-associated and free-living bacterial fractions, corresponding to 43,289 bacterial OTUs. High-quality sequence reads for individual sample sets ranged between 1,136 and 457,102 sequences. On all dates, bacterial alpha diverclass="Chemical">sity (i.e. richclass="Chemical">ness of OTUs) was greater withiclass="Chemical">n particle-associated compoclass="Chemical">neclass="Chemical">nts compared to the free-liviclass="Chemical">ng couclass="Chemical">nterpart (raclass="Chemical">nge = 1.1 class="Chemical">n class="Species">to 7.1 times), with peaks of richness for both fractions in mid-summer (Fig 3). However, the degree of variation in richness over the year was not significantly different between the different components of the microbial community (Levene’s Test, p = 0.264).

Fig 3

Temporal patterns in bacterioplankton alpha diversity measured using richness of OTUs.

Differences in richness of OTUs in (A) particle-associated and (B) free-living bacterioplankton assemblages collected on 23 dates from February 2013 to January 2014.

There was no class="Chemical">sigclass="Chemical">nificaclass="Chemical">nt differeclass="Chemical">nce iclass="Chemical">n mediaclass="Chemical">n particle-attached richclass="Chemical">ness (Mood’s mediaclass="Chemical">n tests: p < 0.001) amoclass="Chemical">ng daily, weekly, aclass="Chemical">nd moclass="Chemical">nthly timescales (Fig 4A). Furthermore, there was a class="Chemical">n class="Chemical">similar degree of variability for richness of this fraction among all timescales. There was more variability in free-living richness at short time intervals (i.e. daily and weekly timescales) (Fig 4A), but there was no significant difference among richness medians.

Fig 4

Boxplots showing variance in bacterial assemblage (A) OTU richness (B) Bray–Curtis dissimilarity, and (C) production (3H-leucine) among sampling timescales. Boxes show medians (dark lines), averages (diamonds), and inter-quartile ranges. Whiskers indicate data within 3X inter-quartile ranges, and points are outliers. Letter(s) above boxes indicate the groups of samples that are significantly different in their medians (Mood’s median tests: p < 0.001). Sample sizes are presented for each timescale.

Boxplots showing variance in bacterial assemblage (A) OTU richness (B) Bray–Curtis disclass="Chemical">similarity, aclass="Chemical">nd (C) productioclass="Chemical">n (class="Chemical">n class="Chemical">3H-leucine) among sampling timescales. Boxes show medians (dark lines), averages (diamonds), and inter-quartile ranges. Whiskers indicate data within 3X inter-quartile ranges, and points are outliers. Letter(s) above boxes indicate the groups of samples that are significantly different in their medians (Mood’s median tests: p < 0.001). Sample sizes are presented for each timescale.

Patterns in bacterial beta diversity

In general, both particle-associated and free-living components were more class="Chemical">similar iclass="Chemical">n compoclass="Chemical">n class="Chemical">sition on daily and weekly timeframes than on a monthly timeframe. However, the pattern was not linear for either group. Instead, dissimilarity exhibited a roughly parabolic pattern (Fig 5). Assemblages became increasingly dissimilar in composition with separation in time up to six months, after which the trend was for a gradual decrease in dissimilarity. If we disregard year, these trends indicate that assemblages occurring closer in time, whatever the time of year, are increasingly alike in composition. Furthermore, this pattern of nonlinearity shows that the LMR microbiome varies along seasonal gradients.

Fig 5

Temporal patterns in bacterioplankton beta diversity measured using Bray-Curtis dissimilarity.

Relationships between (A) particle-associated and (B) free-living dissimilarities and interval of time between sample dates. Points represent pairwise dissimilarities calculated from bacterioplankton assemblages collected between 1 and 343 days apart.

Relationships between (A) particle-associated and (B) free-living disclass="Chemical">similarities aclass="Chemical">nd iclass="Chemical">nterval of time betweeclass="Chemical">n sample dates. Poiclass="Chemical">nts represeclass="Chemical">nt pairwise disclass="Chemical">n class="Chemical">similarities calculated from bacterioplankton assemblages collected between 1 and 343 days apart.

While particle-associated and free-living assemblages were distinct in compoclass="Chemical">sitioclass="Chemical">n (adoclass="Chemical">nis: R2 = 0.08, p < 0.001), they were class="Chemical">n class="Chemical">similarly variable in composition (PERMDISP2, p = 0.172). For both particle-associated and free-living components, there was less variability in beta diversity at a daily timescale compared to a weekly timescale, with the greatest variability occurring at a monthly timescale (Fig 4B). Furthermore, median beta diversity values increased significantly (p < 0.001) with the increase in sampling interval for both components.

The best models selected by class="Disease">AICc explaiclass="Chemical">ned 49% aclass="Chemical">nd 38% of the variatioclass="Chemical">n iclass="Chemical">n particle-associated aclass="Chemical">nd free-liviclass="Chemical">ng assemblage compoclass="Chemical">n class="Chemical">sition, respectively (Table 2). The best model explaining variation in particle-associated beta diversity included water temperature as the primary factor (RVI = 0.93) and TDN was also important (RVI = 0.62). Water temperature was the main factor (RVI = 0.81) in the model explaining variation in free-living assemblages, followed by Chla (RVI = 0.53).

Table 2

Summary of results of relationships between variation in environmental variables and bacterial assemblage beta diversity including relative importance of variables based on model selection using AICc (Akaike’s Information Criterion corrected for small samples).

Analysis	Relative variable importance and sum of Akaike weights (sum wi) for each variable	Pseudo-R2 of AICc-best model
Particle-associated	Temperature (sum wi = 0.93) >	0.49
	TDN (sum wi = 0.62) >
	Discharge (sum wi = 0.43) >
	TDP (sum wi = 0.42) >
	Chla (sum wi = 0.38) >
	DOC (sum wi = 0.31) >
	TSS (sum wi = 0.28)
Free-living	Temperature (sum wi = 0.81) >	0.38
	Chla (sum wi = 0.53) >
	Discharge (sum wi = 0.45) >
	TDP (sum wi = 0.41) >
	TDN (sum wi = 0.40) >
	DOC (sum wi = 0.40) >
	TSS (sum wi = 0.33)

Variables in bold type had a sum of Akaike weight (sum wi) greater or equal to 0.5 and thus were considered relatively important.

Variables in bold type had a sum of Akaike weight (sum wi) greater or equal to 0.5 and thus were conn class="Chemical">sidered relatively importaclass="Chemical">nt.

Patterns in relative abundances of bacterial taxa

At a broad taxonomic level, particular bacterial phyla exhibited distinct patterns in their proportional abundance over the year (Fig 6). Proportions of Proteobacteria were fairly constant over much of the sampling period, but trended upward from class="Chemical">November to Jaclass="Chemical">nuary iclass="Chemical">n both particle-associated aclass="Chemical">nd free-liviclass="Chemical">ng compoclass="Chemical">neclass="Chemical">nts. Relative abuclass="Chemical">ndaclass="Chemical">nces of other phyla, iclass="Chemical">n coclass="Chemical">ntrast, were more closely related to seasoclass="Chemical">nal chaclass="Chemical">nges iclass="Chemical">n class="Chemical">n class="Chemical">water temperature and/or the river hydrograph. Sequences classified as Bacteroidetes and Verrucomicrobia were abundant in assemblages collected in cooler water in spring and winter. Decreased proportions of these taxa, in particular Bacteroidetes, in warm river conditions corresponded with increased proportions of Acidobacteria in summer, and Planctomycetes throughout summer and fall. Cyanobacteria increased in proportion in late-summer and into early fall when the river was at a minimum in discharge, TSS load, and turbidity. Proportions of Actinobacteria increased from late-summer to winter, after which they strongly dominated free-living assemblages during the period of least discharge from mid-July to December. However, members of this phylum were much less abundant in particle-associated assemblages sampled during this time.

Fig 6

Temporal patterns in relative abundances of bacterial phyla sequenced from particle-associated and free-living bacterioplankton assemblages.

Lines were made using local polynomial regression fitting (loess). Shading around lines indicate 95% confidence intervals.

Lines were made un class="Chemical">siclass="Chemical">ng local polyclass="Chemical">nomial regresclass="Chemical">n class="Chemical">sion fitting (loess). Shading around lines indicate 95% confidence intervals.

A class="Chemical">NMDS ordiclass="Chemical">natioclass="Chemical">n coclass="Chemical">nfirmed these seasoclass="Chemical">nal patterclass="Chemical">ns of chaclass="Chemical">nge iclass="Chemical">n compoclass="Chemical">n class="Chemical">sition of particle-associated and free-living bacterioplankton assemblages (Fig 7). Particle-associated assemblages separated in time in a roughly clockwise pattern in NMDS space, from winter to spring to summer to fall, revealing changes in composition over time in a gradual manner. While a cyclical pattern was not apparent for the free-living fraction, the ordination shows that both particle-associated and free-living assemblages collected nearly a year apart trended towards increased similarity in composition.

Fig 7

A NMDS ordination showing seasonal changes in composition of particle-associated and free-living bacterioplankton assemblages.

Stress for the ordination equaled 0.11. Arrows indicate bacterial OTUs correlated (Envfit analysis: R2 = 0.55–0.72, p = 0.001) with the ordination. Identifications of OTUs (RDP classification) are as follows: (OTU02 and OTU04) order Actinomycetales (Actinobacteria); (OTU08) family Comamonadaceae (Betaproteobacteria); (OTU13) class Betaproteobacteria (Proteobacteria); (OTU21) Prosthecobacter (Verrucomicrobia); (OTU32) Methylophilus (Proteobacteria); (OTU41 and OTU53) Flavobacterium (Bacteroidetes); (OTU060) phylum Bacteroidetes; and (OTU66) family Cytophagaceae (Bacteroidetes). Complete identifications of OTUs and specific R2 values of correlations are presented in Table 3.

class="Disease">Stress for the ordiclass="Chemical">natioclass="Chemical">n equaled 0.11. Arrows iclass="Chemical">ndicate bacterial OTUs correlated (Eclass="Chemical">nvfit aclass="Chemical">nalyclass="Chemical">n class="Chemical">sis: R2 = 0.55–0.72, p = 0.001) with the ordination. Identifications of OTUs (RDP classification) are as follows: (OTU02 and OTU04) order Actinomycetales (Actinobacteria); (OTU08) family Comamonadaceae (Betaproteobacteria); (OTU13) class Betaproteobacteria (Proteobacteria); (OTU21) Prosthecobacter (Verrucomicrobia); (OTU32) Methylophilus (Proteobacteria); (OTU41 and OTU53) Flavobacterium (Bacteroidetes); (OTU060) phylum Bacteroidetes; and (OTU66) family Cytophagaceae (Bacteroidetes). Complete identifications of OTUs and specific R2 values of correlations are presented in Table 3.

Table 3

OTUs that correlated (Envfit analysis, R2) with bacterioplankton assemblages plotted in NMDS space.

OTU	Phylum	Class	Order	Family	Genus	R2
OTU02	Actinobacteria	Actinobacteria	Actinomycetales			0.65
OTU04	Actinobacteria	Actinobacteria	Actinomycetales			0.70
OTU08	Proteobacteria	Betaproteobacteria	Burkholderiales	Comamonadaceae		0.72
OTU13	Proteobacteria	Betaproteobacteria				0.67
OTU21	Verrucomicrobia	Verrucomicrobiae	Verrucomicrobiales	Verrucomicrobiaceae	Prosthecobacter	0.68
OTU32	Proteobacteria	Betaproteobacteria	Methylophilales	Methylophilaceae	Methylophilus	0.64
OTU41	Bacteroidetes	Flavobacteriia	Flavobacteriales	Flavobacteriaceae	Flavobacterium	0.57
OTU53	Bacteroidetes	Flavobacteriia	Flavobacteriales	Flavobacteriaceae	Flavobacterium	0.58
OTU60	Bacteroidetes					0.57
OTU66	Bacteroidetes	Cytophagia	Cytophagales	Cytophagaceae		0.55

OTUs were classified using the RDP database (release 11, September 2016).

Envfit analyn class="Chemical">sis ideclass="Chemical">ntified several OTUs that were correlated (R2 ≥ 0.55) with bacterioplaclass="Chemical">nktoclass="Chemical">n assemblages collected iclass="Chemical">n spriclass="Chemical">ng aclass="Chemical">nd wiclass="Chemical">nter (Fig 7; Table 3). These OTUs were related to Bacteroidetes (OTU41, OTU53, OTU60, aclass="Chemical">nd OTU66), Betaproteobacteria (OTU08 aclass="Chemical">nd OTU32), aclass="Chemical">nd Verrucomicrobia (OTU21). The associatioclass="Chemical">ns betweeclass="Chemical">n bacterial OTUs aclass="Chemical">nd assemblages collected iclass="Chemical">n summer were weaker iclass="Chemical">n comparisoclass="Chemical">n, however, free-liviclass="Chemical">ng assemblages iclass="Chemical">n late-summer aclass="Chemical">nd fall correlated with OTUs ideclass="Chemical">ntified to the Acticlass="Chemical">nobacteria order Acticlass="Chemical">nomycetales (OTU02 aclass="Chemical">nd OTU04) aclass="Chemical">nd aclass="Chemical">n uclass="Chemical">nclasclass="Chemical">n class="Chemical">sified member of Betaproteobacteria (OTU13).

OTUs were clasn class="Chemical">sified uclass="Chemical">n class="Chemical">sing the RDP database (release 11, September 2016).

Patterns in bacterial production

Rates of whole-class="Chemical">water bacterial productioclass="Chemical">n measured by the two radioisotopes were very class="Chemical">n class="Chemical">similar, ranging over the year from about 30 to 300 nmol C L-1 h-1 (Fig 8). The temporal pattern correlated strongly with temperature, R2 = 0.68 and 0.78 for 3H-leucine and 3H-thymidine incorporation, respectively (p < 0.001 for each), increasing from spring through late summer, and declining to minimum values in winter. Particle-associated production was usually much greater than for free-living cells. On average, attached bacteria represented 87.9% (standard error = 2.3%) of new biomass measured by 3H-leucine uptake (protein synthesis), and 89.3% (standard error = 2.7%) measured by rates of 3H-thymidine uptake (cell division) in whole-water.

Fig 8

Rates of bacterial production measured from whole-water, and from particle-associated and free-living cells between February 2013 and January 2014 using a 3H-leucine and b 3H-thymidine.

Rates of production are presented as means (± SE) for each date, n = 2–3.

There was less variability in whole-n class="Chemical">water aclass="Chemical">nd particle-associated productioclass="Chemical">n at daily aclass="Chemical">nd weekly timescales compared to the moclass="Chemical">nthly timescale, while there was a class="Chemical">n class="Chemical">similar amount of variability in free-living production among all timescales (Fig 4C).

Discussion

The phyclass="Chemical">sical eclass="Chemical">nviroclass="Chemical">nmeclass="Chemical">nt aclass="Chemical">nd associated placlass="Chemical">nktoclass="Chemical">n commuclass="Chemical">nities of flowiclass="Chemical">ng waters are coclass="Chemical">nticlass="Chemical">nuously iclass="Chemical">n dowclass="Chemical">nstream flux. Heclass="Chemical">nce, at a particular rivericlass="Chemical">ne locatioclass="Chemical">n, placlass="Chemical">nktoclass="Chemical">n assemblages could diverge rapidly iclass="Chemical">n diverclass="Chemical">n class="Chemical">sity and metabolic activity in response to flow-mediated immigration and emigration. Adding to the potential for rapid change in community diversity with flow rate is the reproductive potential of resident biota. Having potentially high rates of turnover, while also subject to continuous downstream flux, the bacterioplankton microbiome of a particular river location potentially could vary as much on the order of days or weeks as among months or seasons. However, in contrast to low-order streams and rivers, the immense volume of large rivers may buffer these systems from rapid environmental or biological variation. In that case, we would expect microbiome assemblage structure and function to vary slowly, following seasonal or annual patterns in regional environmental drivers, rather than transiently-acting factors associated with random local disturbances. We documented temporal patterns of variability in bacterioplankton microbiome structure and production at a single location on the LMR over a range in timescales, from days up to a year. Our time-nested sampling design and results allow us to assess the extent to which constant habitat turnover and environmental variation drives community change.

Differences in particle-associated and free-living alpha diverclass="Chemical">sity betweeclass="Chemical">n aclass="Chemical">ny two days or weeks were ofteclass="Chemical">n as great or greater thaclass="Chemical">n betweeclass="Chemical">n aclass="Chemical">ny two moclass="Chemical">nths across the sampliclass="Chemical">ng period. A poteclass="Chemical">ntial explaclass="Chemical">natioclass="Chemical">n for this patterclass="Chemical">n may be that temporal variability betweeclass="Chemical">n days iclass="Chemical">n microbiome richclass="Chemical">ness was obscured by more ficlass="Chemical">ne-scale temporal aclass="Chemical">nd spatial heterogeclass="Chemical">neity. Although the LMR is turbuleclass="Chemical">nt aclass="Chemical">nd geclass="Chemical">nerally well mixed, because of its high eclass="Chemical">nergy aclass="Chemical">nd complex curreclass="Chemical">nts (that may iclass="Chemical">nclude gyres, eddies, aclass="Chemical">nd upwelliclass="Chemical">ng) patchiclass="Chemical">ness is posclass="Chemical">n class="Chemical">sible at local and sub-daily scales.

However, while differences in bacterial OTU richness were not predictable based on time interval of sampling for either component of the river microbiome, richness of OTUs was always greater in the particle-associated fraction compared to free-living assemblages. This observation is conclass="Chemical">sisteclass="Chemical">nt with those made previously aloclass="Chemical">ng the leclass="Chemical">ngth of the Misclass="Chemical">n class="Chemical">sissippi in mid-summer 2013 [17], highlighting that suspended particles are important microhabitat “hotspots” for bacterial production [18, 43], organic matter transformations [43, 44], and species richness in large river systems.

In contrast to temporal patterns in alpha diverclass="Chemical">sity, we fouclass="Chemical">nd that beta diverclass="Chemical">n class="Chemical">sity of both particle-associated and free-living assemblages varied least on a daily sampling basis, more on a weekly basis, and most between samples separated by monthly intervals. At longer timescales, bacterioplankton assemblages separated by roughly six months were the most distinct from each other in composition, while those separated by more than six months up to a year gradually converged towards similarity. This parabolic pattern of community assemblage differences aligns partly with temperature being an important driver of community assembly in the LMR and other temperate aquatic environments [3-7]. However, in addition to temperature, shifts in composition were related to variability in dissolved N (highest in spring) and chlorophyll a (highest in late summer), indicating that fluctuations in nutrients contribute to seasonality of the river microbiome, and suggesting that the composition of bacterioplankton assemblages of such large rivers [7-10] may be predictable depending on the interaction of the temperature and nutrient regimes.

Patterns in compoclass="Chemical">sitioclass="Chemical">n were associated with chaclass="Chemical">nges iclass="Chemical">n the relative abuclass="Chemical">ndaclass="Chemical">nces of bacterial taxa importaclass="Chemical">nt iclass="Chemical">n other large river systems [7–10, 16, 17, 45–50]. The priclass="Chemical">ncipal eclass="Chemical">nviroclass="Chemical">nmeclass="Chemical">ntal correlate of chaclass="Chemical">nge iclass="Chemical">n proportioclass="Chemical">n of most phyla iclass="Chemical">n particle-associated aclass="Chemical">nd free-liviclass="Chemical">ng assemblages was temperature, to which Acidobacteria aclass="Chemical">nd Placlass="Chemical">nctomycetes respoclass="Chemical">nded poclass="Chemical">n class="Chemical">sitively, and Bacteroidetes and Verrucomicrobia responded negatively.Taxa identified as Actinobacteria responded positively to low river flow, and contributed to substantial differences in the free-living microbiome between mid-summer and fall. Actinobacteria were observed previously during mid-July in 2012 in major tributaries of the Mississippi [16], and during mid-July in 2013 along a 1,300 stretch of the Mississippi itself [17], to be in much higher proportions in free-living assemblages than in the particle-associated microbiome. These studies indicate that during low flow conditions aquatic members of Actinobacteria (e.g. order Actinomycetales) are consistently prominent within free-living assemblages. These taxa may be more competitive when discharge is low due to a reduction in the immigration of allochthonous bacteria from terrestrial sources [51], and/or as a consequence of increased time in transit [9, 47–50].

Differences in beta diverclass="Chemical">sity of assemblages were maximized at arouclass="Chemical">nd 180 days apart iclass="Chemical">n sampliclass="Chemical">ng, regardless of the times of year beiclass="Chemical">ng compared, while differeclass="Chemical">nces iclass="Chemical">n bacterial alpha diverclass="Chemical">n class="Chemical">sity did not vary with time interval. This is likely because microbiome composition varied along seasonal transitions in temperature as well as dissolved N and chlorophyll a concentrations, while bacterial richness oscillated unpredictably at short timescales. Bacterial production, in contrast, while ranging the most between cold and warm months, was nearly identical in spring and fall, indicating the dominant influence of water temperature on microbial metabolic activity. However, this was the case only for particle-attached assemblages, as production of free-living cells did not vary with changes in the environment. These results suggest that bacterial diversity and production in the LMR respond to different sets of drivers, resulting in different patterns of variation both within the river microbiome and across time.

Conclusions

In this study, we found that variation in microbiome richness was unrelated to the timescale of change in the river environment, suggesting there is a high degree of local spatial variation in richness at any given moment in time. In contrast, variation in microbiome compoclass="Chemical">sitioclass="Chemical">n, as well as particle-associated productioclass="Chemical">n, was clearly related to temporal chaclass="Chemical">nges iclass="Chemical">n the river eclass="Chemical">nviroclass="Chemical">nmeclass="Chemical">nt. While productioclass="Chemical">n was driveclass="Chemical">n almost excluclass="Chemical">n class="Chemical">sively by water temperature, the parabolic pattern of variation in dissimilarity indicates that composition was driven by changes in temperature interacting with temporal variation in other environmental factors having a strong seasonal pattern such as dissolved N and chlorophyll a concentrations. Our results indicate that temporal variability in composition of the LMR microbiome is not random; rather, there is successional change over monthly to seasonal timescales, with gradual divergence up to 180 days, followed by gradual reassembly thereafter up to at least 360 days distance in time.

9 Jan 2020

n class="Chemical">PONE-D-19-32405

Timescales of variation in divern class="Chemical">sity aclass="Chemical">nd productioclass="Chemical">n of bacterioplaclass="Chemical">nktoclass="Chemical">n assemblages iclass="Chemical">n the Lower Misclass="Chemical">n class="Chemical">sissippi River

n class="Disease">PLOS Oclass="Chemical">n class="Chemical">NE

Dear Dr. Ochs,

Thank you for submitting your manuscript to class="Disease">PLOS Oclass="Chemical">n class="Chemical">NE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

Both reviewers found the study to be well described and have suggested only minor revin class="Chemical">sioclass="Chemical">ns to better coclass="Chemical">ntextualize the study iclass="Chemical">n the existiclass="Chemical">ng literature aclass="Chemical">nd provide clarificatioclass="Chemical">ns to the methods/aclass="Chemical">nalyclass="Chemical">n class="Chemical">sis.

We would appreciate receiving your revised manuscript by Feb 23 2020 11:59PM. When you are ready to submit your reviclass="Chemical">sioclass="Chemical">n, log oclass="Chemical">n to https://www.editorialmaclass="Chemical">nager.com/poclass="Chemical">ne/ aclass="Chemical">nd select the 'Submisclass="Chemical">n class="Chemical">sions Needing Revision' folder to locate your manuscript file.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter.

To enhance the reproducibility of your results, we recommend that if applicable you depoclass="Chemical">sit your laboratory protocols iclass="Chemical">n protocols.io, where a protocol caclass="Chemical">n be asclass="Chemical">n class="Chemical">signed its own identifier (DOI) such that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). This letter should be uploaded as separate file and labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original vern class="Chemical">sioclass="Chemical">n. This file should be uploaded as separate file aclass="Chemical">nd labeled 'Revised Maclass="Chemical">nuscript with Track Chaclass="Chemical">nges'.

An unmarked vern class="Chemical">sioclass="Chemical">n of your revised paper without tracked chaclass="Chemical">nges. This file should be uploaded as separate file aclass="Chemical">nd labeled 'Maclass="Chemical">nuscript'.

Please note while forming your response, if your article is accepted, you may have the opportunity to make the peer review history publicly available. The record will include editor decin class="Chemical">sioclass="Chemical">n letters (with reviews) aclass="Chemical">nd your respoclass="Chemical">nses to reviewer commeclass="Chemical">nts. If eligible, we will coclass="Chemical">ntact you to opt iclass="Chemical">n or out.

We look forward to receiving your revised manuscript.

Kind regards,

Christopher Staley, Ph.D.

Academic Editor

n class="Disease">PLOS Oclass="Chemical">n class="Chemical">NE

Journal Requirements:

When submitting your revin class="Chemical">sioclass="Chemical">n, we class="Chemical">need you to address these additioclass="Chemical">nal requiremeclass="Chemical">nts.

1. Please ensure that your manuscript meets class="Disease">PLOS Oclass="Chemical">n class="Chemical">NE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

http://www.journals.class="Disease">plos.org/class="Chemical">n class="Disease">plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and http://www.journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. Thank you for stating the following in the Acknowledgments Section of your manuscript:

"Field sampling asn class="Chemical">sistaclass="Chemical">nce was provided by Derrick Bussaclass="Chemical">n aclass="Chemical">nd Bram Stoclass="Chemical">ne. Laboratory

asn class="Chemical">sistaclass="Chemical">nce was provided by Tricia Lipsoclass="Chemical">n aclass="Chemical">nd Dr. Beth Baker. We thaclass="Chemical">nk Jasoclass="Chemical">n Hoeksema, Chaz

Hyseni, and Jarrod Sackreiter for consultation on statistics. The UMMC Molecular and

Genomics Facility is supported, in part, by funds from the n class="Chemical">NIGMS, iclass="Chemical">ncludiclass="Chemical">ng Misclass="Chemical">n class="Chemical">sissippi INBRE

(P20GM103476), Center for Psychiatric n class="Chemical">Neuroscieclass="Chemical">nce (CPclass="Chemical">n class="Chemical">N)-COBRE (P30GM103328),

Oben class="Chemical">sity, Cardioreclass="Chemical">nal aclass="Chemical">nd class="Chemical">n class="Disease">Metabolic Diseases- COBRE (P20GM104357) and Mississippi Center

of Excellence in Perinatal Research (MS 465 -CEPR)-COBRE (P20GM121334). Funding for the

study was provided by NSF DEB 1049911."

We note that you have provided funding information that is not currently declared in your Funding Statement. However, funding information should not appear in the Acknowledgments section or other areas of your manuscript. We will only publish funding information present in the Funding Statement section of the online submisn class="Chemical">sioclass="Chemical">n form.

Please remove any funding-related text from the manuscript and let us know how you would like to update your Funding Statement. Currently, your Funding Statement reads as follows:

'n class="Chemical">CAO

NSF DEB 1049911

n class="Chemical">Natioclass="Chemical">nal Scieclass="Chemical">nce Fouclass="Chemical">ndatioclass="Chemical">n

https://www.nsf.gov/

The funders had no role in study den class="Chemical">sigclass="Chemical">n, data collectioclass="Chemical">n aclass="Chemical">nd aclass="Chemical">nalyclass="Chemical">n class="Chemical">sis, decision to publish, or preparation of the manuscript.'

Please provide an amended Funding Statement that declares *all* the funding or sources of support received during this specific study (whether external or internal to your organization) as detailed online in our guide for authors at http://journals.n class="Disease">plos.org/class="Chemical">n class="Disease">plosone/s/submit-now

Please state what role the funders took in the study.  If any authors received a salary from any of your funders, please state which authors and which funder. If the funders had no role, please state: "The funders had no role in study declass="Chemical">sigclass="Chemical">n, data collectioclass="Chemical">n aclass="Chemical">nd aclass="Chemical">nalyclass="Chemical">n class="Chemical">sis, decision to publish, or preparation of the manuscript."

c. Please include your amended statements within your cover letter; we will change the online submisn class="Chemical">sioclass="Chemical">n form oclass="Chemical">n your behalf.

3. We note that you have indicated that data from this study are available upon request. class="Disease">PLOS oclass="Chemical">nly allows data to be available upoclass="Chemical">n request if there are legal or ethical restrictioclass="Chemical">ns oclass="Chemical">n shariclass="Chemical">ng data publicly. For more iclass="Chemical">nformatioclass="Chemical">n oclass="Chemical">n uclass="Chemical">nacceptable data access restrictioclass="Chemical">ns, please see http://jourclass="Chemical">nals.class="Chemical">n class="Disease">plos.org/plosone/s/data-availability#loc-unacceptable-data-access-restrictions.

In your revised cover letter, please address the following prompts:

a) If there are ethical or legal restrictions on sharing a de-identified data set, please explain them in detail (e.g., data contain potentially senn class="Chemical">sitive iclass="Chemical">nformatioclass="Chemical">n, data are owclass="Chemical">ned by a third-party orgaclass="Chemical">nizatioclass="Chemical">n, etc.) aclass="Chemical">nd who has imposed them (e.g., aclass="Chemical">n ethics committee). Please also provide coclass="Chemical">ntact iclass="Chemical">nformatioclass="Chemical">n for a data access committee, ethics committee, or other iclass="Chemical">nstitutioclass="Chemical">nal body to which data requests may be seclass="Chemical">nt.

b) If there are no restrictions, please upload the minimal anonymized data set necessary to replicate your study findings as either Supporting Information files or to a stable, public repoclass="Chemical">sitory aclass="Chemical">nd provide us with the relevaclass="Chemical">nt URLs, DOIs, or accesclass="Chemical">n class="Chemical">sion numbers. For a list of acceptable repositories, please see http://journals.plos.org/plosone/s/data-availability#loc-recommended-repositories.

We will update your Data Availability statement on your behalf to reflect the information you provide.

[n class="Chemical">Note: HTML markup is below. Please do class="Chemical">not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclun class="Chemical">sioclass="Chemical">ns?

The manuscript must describe a technically sound piece of scientific research with data that supports the concluclass="Chemical">sioclass="Chemical">ns. Experimeclass="Chemical">nts must have beeclass="Chemical">n coclass="Chemical">nducted rigorously, with appropriate coclass="Chemical">ntrols, replicatioclass="Chemical">n, aclass="Chemical">nd sample class="Chemical">n class="Chemical">sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: Yes

**********

2. Has the statistical analyn class="Chemical">sis beeclass="Chemical">n performed appropriately aclass="Chemical">nd rigorously?

Reviewer #1: Yes

Reviewer #2: Yes

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The class="Disease">PLOS Data policy requires authors to make all data uclass="Chemical">nderlyiclass="Chemical">ng the ficlass="Chemical">ndiclass="Chemical">ngs described iclass="Chemical">n their maclass="Chemical">nuscript fully available without restrictioclass="Chemical">n, with rare exceptioclass="Chemical">n (please refer to the Data Availability Statemeclass="Chemical">nt iclass="Chemical">n the maclass="Chemical">nuscript PDF file). The data should be provided as part of the maclass="Chemical">nuscript or its supporticlass="Chemical">ng iclass="Chemical">nformatioclass="Chemical">n, or depoclass="Chemical">n class="Chemical">sited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: No

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

class="Disease">PLOS Oclass="Chemical">n class="Chemical">NE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: “Timescales of variation in diverclass="Chemical">sity aclass="Chemical">nd productioclass="Chemical">n of bacterioplaclass="Chemical">nktoclass="Chemical">n assemblages iclass="Chemical">n the lower Misclass="Chemical">n class="Chemical">sissippi River” seeks to describe patterns in particle-associated and free-living microbial assemblages at three different timescales using next generation sequencing combined with water physical and chemical paired measurements. The study is well organized and clearly written. Findings should be of interest to PLOS readers.

Dclass="Chemical">NA extractioclass="Chemical">n aclass="Chemical">nd sequeclass="Chemical">nciclass="Chemical">ng quality coclass="Chemical">ntrols. Authors performed Dclass="Chemical">n class="Chemical">NA extractions followed by 16S rRNA gene amplification. Were any field blanks, method blanks, no template controls, or positive controls used during 16S rRNA gene amplification? Did authors include a method blank during sequencing? If so, please list controls and results in manuscript. If not, please mention that these controls were not included along with rationale.

Screen for covariate auto correlation. Authors do not report any correlation testing amongst covariates to identify potential confounding auto correlation. Please conduct correlation analyn class="Chemical">sis amoclass="Chemical">ng covariates used for statistical testiclass="Chemical">ng aclass="Chemical">nd report results. If some covariates eclass="Chemical">nd up beiclass="Chemical">ng auto correlated, theclass="Chemical">n repeat respective statistical tests with oclass="Chemical">nly covariates class="Chemical">not auto correlated aclass="Chemical">nd revise maclass="Chemical">nuscript as class="Chemical">needed.

Minor Comments:

Line 112: Please include the maximum length of time samples remained frozen prior to testing. For example, (< xx months).

Line 127: Please include the maximum length of time samples remained frozen prior to testing. For example, (< xx months).

Line 174: Do you mean sampling “period” or “event”? I think you mean “event”, please clarify.

Line 249: Please provide range of high-quality sequence reads for individual sample sets.

Line 437: Please revise statement to, “This is likely because microbiome…”

Figure 1 caption: Please provide the USGS gage number in the caption description.

Reviewer #2: The manuscript by Payne et al describes temporal variability in bacterioplankton community structure and function in a river ecosystem. This research covers several important concepts that add valuable information to class="Chemical">sigclass="Chemical">nificaclass="Chemical">nt kclass="Chemical">nowledge gaps iclass="Chemical">n the literature iclass="Chemical">ncludiclass="Chemical">ng. Firstly, it liclass="Chemical">nks measuremeclass="Chemical">nts of both structure aclass="Chemical">nd fuclass="Chemical">nctioclass="Chemical">n, which is class="Chemical">needed iclass="Chemical">n more studies examiclass="Chemical">niclass="Chemical">ng microbiome-ecosystem iclass="Chemical">nteractioclass="Chemical">ns. Secoclass="Chemical">ndly, it addresses temporal variability, which is class="Chemical">not well uclass="Chemical">nderstood iclass="Chemical">n microbial ecology, especially at the microbiome scale. Aclass="Chemical">nd thirdly, it addresses river ecosystems, which represeclass="Chemical">nt a dyclass="Chemical">namic aclass="Chemical">nd uclass="Chemical">nidirectioclass="Chemical">nal flowiclass="Chemical">ng system that are quite differeclass="Chemical">nt from more stable microbiome habitats that are commoclass="Chemical">nly studied such as hosts, soils, aclass="Chemical">nd blue class="Chemical">n class="Chemical">water marine systems. The manuscript is also well written and the results are presented in a manner that is broadly valuable beyond those interested in the Lower Mississippi River. Adding the distinction between particle-associate and free-living organisms is also an important contribution. I have the following suggestions to improve and clarify the manuscript prior to publication:

The introduction is well written and relevant. However the discusclass="Chemical">sioclass="Chemical">n of previous refereclass="Chemical">nces seems a little thiclass="Chemical">n. For example, liclass="Chemical">nes 57-66 represeclass="Chemical">nt just speculatioclass="Chemical">n aclass="Chemical">nd hypotheclass="Chemical">n class="Chemical">sis exploration on the part of the authors. I would prefer to see this space devoted to a little more detail about what was learned in previous studies related to these questions and what knowledge gaps still remain that are being addressed here.

There is also not really any discusn class="Chemical">sioclass="Chemical">n iclass="Chemical">n the iclass="Chemical">ntroductioclass="Chemical">n related to “eclass="Chemical">nviroclass="Chemical">nmeclass="Chemical">ntal chaclass="Chemical">nge” aclass="Chemical">nd the factors (temp, class="Chemical">n class="Chemical">chla, nutrients, etc.) that were measured in the study. What knowledge gap is being addressed here?

L 75-78: The hypotheses are also somewhat vague. What does “scale with time” mean? In relation to what type of environmental change that is being measured here? This is written as one hypothen class="Chemical">sis, but seems to actually be at least three.

L 100-102: I assume the more intenn class="Chemical">sive sampliclass="Chemical">ng was coclass="Chemical">nducted iclass="Chemical">n the summer due to higher biomass/productivity/etc? It would be good to provide a brief ratioclass="Chemical">nale.

Are the sequence data being made publicly available?

The data analyn class="Chemical">sis sectioclass="Chemical">n is very well explaiclass="Chemical">ned.

I suggest looking for opportunities to reduce wordiness in some of the results. E.g., L262: “more variability”; L273: “were more n class="Chemical">similar”; L288: “class="Chemical">n class="Chemical">similarly variable in composition”

L280: I might be misn class="Chemical">siclass="Chemical">ng somethiclass="Chemical">ng, but it doesclass="Chemical">n’t seem like you caclass="Chemical">n commeclass="Chemical">nt oclass="Chemical">n somethiclass="Chemical">ng aclass="Chemical">nythiclass="Chemical">ng happeclass="Chemical">niclass="Chemical">ng “oclass="Chemical">n aclass="Chemical">n aclass="Chemical">nclass="Chemical">nual baclass="Chemical">n class="Chemical">sis” based on the patterns in a single year.

I’m not sure I understand the rationale for how the discharge data are used. Unless I’m misn class="Chemical">siclass="Chemical">ng somethiclass="Chemical">ng, those data are class="Chemical">not used iclass="Chemical">n the model selectioclass="Chemical">n. Why class="Chemical">not? Aclass="Chemical">nd if they areclass="Chemical">n’t used there, why iclass="Chemical">nclude them?

The figures overall are very well done.

**********

6. n class="Disease">PLOS authors have the optioclass="Chemical">n to publish the peer review history of their article (what does this meaclass="Chemical">n?). If published, this will iclass="Chemical">nclude your full peer review aclass="Chemical">nd aclass="Chemical">ny attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

[class="Chemical">NOTE: If reviewer commeclass="Chemical">nts were submitted as aclass="Chemical">n attachmeclass="Chemical">nt file, they will be attached to this email aclass="Chemical">nd accesclass="Chemical">n class="Chemical">sible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files to be viewed.]

While reviclass="Chemical">siclass="Chemical">ng your submisclass="Chemical">n class="Chemical">sion, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email us at figures@plos.org. Please note that Supporting Information files do not need this step.

2 Mar 2020

n class="Chemical">PONE-D-19-32405

Timescales of variation in divern class="Chemical">sity aclass="Chemical">nd productioclass="Chemical">n of bacterioplaclass="Chemical">nktoclass="Chemical">n assemblages iclass="Chemical">n the Lower Misclass="Chemical">n class="Chemical">sissippi River

n class="Disease">PLOS Oclass="Chemical">n class="Chemical">NE

Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: “Timescales of variation in diverclass="Chemical">sity aclass="Chemical">nd productioclass="Chemical">n of bacterioplaclass="Chemical">nktoclass="Chemical">n assemblages iclass="Chemical">n the lower Misclass="Chemical">n class="Chemical">sissippi River” seeks to describe patterns in particle-associated and free-living microbial assemblages at three different timescales using next generation sequencing combined with water physical and chemical paired measurements. The study is well organized and clearly written. Findings should be of interest to PLOS readers.

Dclass="Chemical">NA extractioclass="Chemical">n aclass="Chemical">nd sequeclass="Chemical">nciclass="Chemical">ng quality coclass="Chemical">ntrols. Authors performed Dclass="Chemical">n class="Chemical">NA extractions followed by 16S rRNA gene amplification. Were any field blanks, method blanks, no template controls, or positive controls used during 16S rRNA gene amplification? Did authors include a method blank during sequencing? If so, please list controls and results in manuscript. If not, please mention that these controls were not included along with rationale.

Regarding the use of controls during 16S rRn class="Chemical">NA geclass="Chemical">ne amplificatioclass="Chemical">n, we have added the followiclass="Chemical">ng explaclass="Chemical">natioclass="Chemical">n withiclass="Chemical">n the methods sectioclass="Chemical">n:

class="Chemical">Negative (class="Chemical">no template) coclass="Chemical">ntrols were used iclass="Chemical">n all amplificatioclass="Chemical">ns aclass="Chemical">nd coclass="Chemical">nclass="Chemical">n class="Chemical">sistently gave negative results. Such negative amplifications were also used as blanks in sequencing, yielding no sequence data. Positive controls were not needed as we have used these procedures successfully for a variety of samples types (12, 18, 19).

12. Payne JT, Millar JJ, Jackson CR, Ochs CA (2017) Patterns of variation in diverclass="Chemical">sity of the Misclass="Chemical">n class="Chemical">sissippi river microbiome over 1,300 kilometers. PLoS One 12: e0174890. doi: doi.org/10.1371/journal.pone.0174890

18. Shirur KP, Jackson CR, Goulet TL (2016) Len class="Chemical">sioclass="Chemical">n recovery aclass="Chemical">nd the bacterial microbiome iclass="Chemical">n two Caribbeaclass="Chemical">n gorgoclass="Chemical">niaclass="Chemical">n corals. Mar Biol 163:238 doi:10.1007/s00227-016-3008-6

19. Weingarten EA, Atkinson CA, Jackson CR. (2019) The class="Species">gut microbiome of freshclass="Chemical">n class="Chemical">water Unionidae mussels is determined by host species and is selectively retained from filtered seston. PLoS One 14: e0224796. doi: doi.org/10.1371/journal.pone.0224796.

Screen for covariate auto correlation. Authors do not report any correlation testing amongst covariates to identify potential confounding auto correlation. Please conduct correlation analyn class="Chemical">sis amoclass="Chemical">ng covariates used for statistical testiclass="Chemical">ng aclass="Chemical">nd report results. If some covariates eclass="Chemical">nd up beiclass="Chemical">ng auto correlated, theclass="Chemical">n repeat respective statistical tests with oclass="Chemical">nly covariates class="Chemical">not auto correlated aclass="Chemical">nd revise maclass="Chemical">nuscript as class="Chemical">needed.

As suggested, we conducted a cross-correlation matrix analyn class="Chemical">sis of predictors. The table of correlatioclass="Chemical">n results is below. A priori, we assumed a correlatioclass="Chemical">n coefficieclass="Chemical">nt of 0.80 iclass="Chemical">ndicaticlass="Chemical">ng auto-correlatioclass="Chemical">n. Based oclass="Chemical">n this aclass="Chemical">nalyclass="Chemical">n class="Chemical">sis, we do not think that statistical tests need to be repeated.

We added the following statement to the Methods section: “A cross-correlation matrix analyn class="Chemical">sis of caclass="Chemical">ndidate predictors was performed. Predictors haviclass="Chemical">ng a correlatioclass="Chemical">n coefficieclass="Chemical">nt ≥ 0.8 were class="Chemical">not both iclass="Chemical">ncluded iclass="Chemical">n the model for commuclass="Chemical">nity compoclass="Chemical">n class="Chemical">sition.”

We think that this analyn class="Chemical">sis aclass="Chemical">nd statemeclass="Chemical">nt is sufficieclass="Chemical">nt to address poteclass="Chemical">ntial coclass="Chemical">nfouclass="Chemical">ndiclass="Chemical">ng auto correlatioclass="Chemical">n amoclass="Chemical">ng variables. We iclass="Chemical">nclude the matrix table below for the sake of review, but we do class="Chemical">not thiclass="Chemical">nk it is class="Chemical">necessary to add it to the maclass="Chemical">nuscript.

Temp TSS n class="Chemical">Chla C..mmol. class="Chemical">n class="Chemical">N..mmol. P..mmol. Discharge

Temp 1.00

TSS 0.06 1.00

n class="Chemical">Chla 0.10 -0.20 1.00

C..mmol. 0.60 0.23 -0.15 1.00

N..mmol. 0.57 0.48 -0.49 0.57 1.00

P..mmol. 0.63 0.17 -0.34 0.47 0.65 1.00

Discharge 0.21 0.62 -0.52 0.42 0.70 0.33 1

Please note that for the reviclass="Chemical">sioclass="Chemical">n we replaced class="Chemical">nutrieclass="Chemical">nt ratios (C/class="Chemical">n class="Chemical">N, N/P, C/P) as predictors with concentrations of individual nutrients (DOC, TDN, TDP). This change strengthens the predictive capability of the model, eliminates the redundancy of considering both C/N and N/P as predictors (because N varies much more than C or N), and has only a minor effect on multivariate model output. Temperature and N (replacing C/N) remain the most important predictors of particle-associated communities, and Temperature and Chla remain the most important predictors of free-living composition (Table 2).

Minor Comments:

Line 112: Please include the maximum length of time samples remained frozen prior to testing. For example, (< xx months).

We kept samples frozen < 18 months prior to testing. Changes were made within the text.

Line 127: Please include the maximum length of time samples remained frozen prior to testing. For example, (< xx months).

We kept samples frozen < 18 months prior to testing. Changes were made within the text.

Line 174: Do you mean sampling “period” or “event”? I think you mean “event”, please clarify.

We meant sampling event, not period. Changes were made within the text.

Line 249: Please provide range of high-quality sequence reads for individual sample sets.

The range of high-quality sequence reads for individual sample sets is now provided within the manuscript.

Line 437: Please revise statement to, “This is likely because microbiome…”

We made this revin class="Chemical">sioclass="Chemical">n withiclass="Chemical">n the text.

Figure 1 caption: Please provide the USGS gage number in the caption description.

Gage height data was collected at a station run by the U.S. Army Corps of Engineers, not the USGS. We provided coordinates for this gage station (34°44'26.79" n class="Chemical">N 90°26'42.52" W) iclass="Chemical">n the captioclass="Chemical">n descriptioclass="Chemical">n for clarificatioclass="Chemical">n.

Reviewer #2: The manuscript by Payne et al describes temporal variability in bacterioplankton community structure and function in a river ecosystem. This research covers several important concepts that add valuable information to class="Chemical">sigclass="Chemical">nificaclass="Chemical">nt kclass="Chemical">nowledge gaps iclass="Chemical">n the literature iclass="Chemical">ncludiclass="Chemical">ng. Firstly, it liclass="Chemical">nks measuremeclass="Chemical">nts of both structure aclass="Chemical">nd fuclass="Chemical">nctioclass="Chemical">n, which is class="Chemical">needed iclass="Chemical">n more studies examiclass="Chemical">niclass="Chemical">ng microbiome-ecosystem iclass="Chemical">nteractioclass="Chemical">ns. Secoclass="Chemical">ndly, it addresses temporal variability, which is class="Chemical">not well uclass="Chemical">nderstood iclass="Chemical">n microbial ecology, especially at the microbiome scale. Aclass="Chemical">nd thirdly, it addresses river ecosystems, which represeclass="Chemical">nt a dyclass="Chemical">namic aclass="Chemical">nd uclass="Chemical">nidirectioclass="Chemical">nal flowiclass="Chemical">ng system that are quite differeclass="Chemical">nt from more stable microbiome habitats that are commoclass="Chemical">nly studied such as hosts, soils, aclass="Chemical">nd blue class="Chemical">n class="Chemical">water marine systems. The manuscript is also well written and the results are presented in a manner that is broadly valuable beyond those interested in the Lower Mississippi River. Adding the distinction between particle-associate and free-living organisms is also an important contribution. I have the following suggestions to improve and clarify the manuscript prior to publication:

The introduction is well written and relevant. However the discusclass="Chemical">sioclass="Chemical">n of previous refereclass="Chemical">nces seems a little thiclass="Chemical">n. For example, liclass="Chemical">nes 57-66 represeclass="Chemical">nt just speculatioclass="Chemical">n aclass="Chemical">nd hypotheclass="Chemical">n class="Chemical">sis exploration on the part of the authors. I would prefer to see this space devoted to a little more detail about what was learned in previous studies related to these questions and what knowledge gaps still remain that are being addressed here.

There is also not really any discusn class="Chemical">sioclass="Chemical">n iclass="Chemical">n the iclass="Chemical">ntroductioclass="Chemical">n related to “eclass="Chemical">nviroclass="Chemical">nmeclass="Chemical">ntal chaclass="Chemical">nge” aclass="Chemical">nd the factors (temp, class="Chemical">n class="Chemical">chla, nutrients, etc.) that were measured in the study. What knowledge gap is being addressed here?

We added a paragraph within the Introduction (second paragraph of Introduction) that discusses environmental factors that are characteristic of large river systems, and could vary at different timescales, and that were examined in this study.

L 75-78: The hypotheses are also somewhat vague. What does “scale with time” mean? In relation to what type of environmental change that is being measured here? This is written as one hypothen class="Chemical">sis, but seems to actually be at least three.

We replaced “scale with time” with “change more over longer timescales” to clarify our hypothen class="Chemical">sis. We also explicitly stated eclass="Chemical">nviroclass="Chemical">nmeclass="Chemical">ntal factors that were measured duriclass="Chemical">ng this study.

L 100-102: I assume the more intenn class="Chemical">sive sampliclass="Chemical">ng was coclass="Chemical">nducted iclass="Chemical">n the summer due to higher biomass/productivity/etc? It would be good to provide a brief ratioclass="Chemical">nale.

We added the following as an explanation for intenn class="Chemical">sive sampliclass="Chemical">ng duriclass="Chemical">ng summer: “We chose to sample frequeclass="Chemical">ntly duriclass="Chemical">ng summer because this is a period of high bacterial productioclass="Chemical">n (Ochs et al. 2010), aclass="Chemical">nd poteclass="Chemical">ntially a period iclass="Chemical">n which a high degree of short-term temporal variatioclass="Chemical">n could be detected.”

Are the sequence data being made publicly available?

Yes, all sequences can be accessed in the n class="Chemical">NCBI SRA database uclass="Chemical">nder the BioProject ID PRJclass="Chemical">n class="Chemical">NA358603.

The data analyn class="Chemical">sis sectioclass="Chemical">n is very well explaiclass="Chemical">ned.

Thank you!

I suggest looking for opportunities to reduce wordiness in some of the results. E.g., L262: “more variability”; L273: “were more n class="Chemical">similar”; L288: “class="Chemical">n class="Chemical">similarly variable in composition”

Changes were made within the text to reduce wordiness.

L280: I might be misn class="Chemical">siclass="Chemical">ng somethiclass="Chemical">ng, but it doesclass="Chemical">n’t seem like you caclass="Chemical">n commeclass="Chemical">nt oclass="Chemical">n somethiclass="Chemical">ng aclass="Chemical">nythiclass="Chemical">ng happeclass="Chemical">niclass="Chemical">ng “oclass="Chemical">n aclass="Chemical">n aclass="Chemical">nclass="Chemical">nual baclass="Chemical">n class="Chemical">sis” based on the patterns in a single year.

“Reassembling on an annual ban class="Chemical">sis” was removed from the text to reflect that patterclass="Chemical">ns were observed withiclass="Chemical">n a class="Chemical">n class="Chemical">single year.

I’m not sure I understand the rationale for how the discharge data are used. Unless I’m misn class="Chemical">siclass="Chemical">ng somethiclass="Chemical">ng, those data are class="Chemical">not used iclass="Chemical">n the model selectioclass="Chemical">n. Why class="Chemical">not? Aclass="Chemical">nd if they areclass="Chemical">n’t used there, why iclass="Chemical">nclude them?

In the original manuscript, we included discharge in Figure 1 to display the dynamic hydrological nature of the river. Although we did not include these data in model selection, the reviewer is correct that discharge could be an important predictor. Therefore, we have revised our model selection procedure to include discharge.

The figures overall are very well done.

Thank you!

Submitted filename: Response to Reviewers submit20Feb20.n class="Chemical">docx

Click here for additional data file.

13 Mar 2020

Timescales of variation in divern class="Chemical">sity aclass="Chemical">nd productioclass="Chemical">n of bacterioplaclass="Chemical">nktoclass="Chemical">n assemblages iclass="Chemical">n the Lower Misclass="Chemical">n class="Chemical">sissippi River

n class="Chemical">PONE-D-19-32405R1

Dear Dr. Ochs,

We are pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it complies with all outstanding technical requirements.

Within one week, you will receive an e-mail containing information on the amendments required prior to publication. When all required modifications have been addressed, you will receive a formal acceptance letter and your manuscript will proceed to our production department and be scheduled for publication.

Shortly after the formal acceptance letter is sent, an invoice for payment will follow. To ensure an efficient production and billing process, please log into Editorial Manager at https://www.editorialmanager.com/pone/, click the "Update My Information" link at the top of the page, and update your user information. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@n class="Disease">plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to enable them to help maximize its impact. If they will be preparing press materials for this manuscript, you must inform our press team as soon as posn class="Chemical">sible aclass="Chemical">nd class="Chemical">no later thaclass="Chemical">n 48 hours after receiviclass="Chemical">ng the formal acceptaclass="Chemical">nce. Your maclass="Chemical">nuscript will remaiclass="Chemical">n uclass="Chemical">nder strict press embargo uclass="Chemical">ntil 2 pm Easterclass="Chemical">n Time oclass="Chemical">n the date of publicatioclass="Chemical">n. For more iclass="Chemical">nformatioclass="Chemical">n, please coclass="Chemical">ntact oclass="Chemical">nepress@class="Chemical">n class="Disease">plos.org.

With kind regards,

Christopher Staley, Ph.D.

Academic Editor

n class="Disease">PLOS Oclass="Chemical">n class="Chemical">NE

Additional Editor Comments (optional):

Reviewers' comments:

18 Mar 2020

n class="Chemical">PONE-D-19-32405R1

Timescales of variation in divern class="Chemical">sity aclass="Chemical">nd productioclass="Chemical">n of bacterioplaclass="Chemical">nktoclass="Chemical">n assemblages iclass="Chemical">n the Lower Misclass="Chemical">n class="Chemical">sissippi River

Dear Dr. Ochs:

I am pleased to inform you that your manuscript has been deemed suitable for publication in class="Disease">PLOS Oclass="Chemical">n class="Chemical">NE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please notify them about your upcoming paper at this point, to enable them to help maximize its impact. If they will be preparing press materials for this manuscript, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@n class="Disease">plos.org.

For any other questions or concerns, please email n class="Disease">plosoclass="Chemical">ne@class="Chemical">n class="Disease">plos.org.

Thank you for submitting your work to n class="Disease">PLOS Oclass="Chemical">n class="Chemical">NE.

With kind regards,

n class="Disease">PLOS Oclass="Chemical">n class="Chemical">NE Editorial Office Staff

on behalf of

Dr. Christopher Staley

Academic Editor

n class="Disease">PLOS Oclass="Chemical">n class="Chemical">NE