Meet Me in the Middle: Median Temperatures Impact Cyanobacteria and Photoautotrophy in Eruptive Yellowstone Hot Springs

ABSTRACT Geographic isolation can be a main driver of microbial evolution in hot springs while temperature plays a role on local scales. For example, cyanobacteria, particularly high-temperature Synechococcus spp., have undergone ecological diversification along temperature gradients in hot spring outflow channels. While water flow, and thus temperature, is largely stable in many hot springs, flow can vary in geysing/eruptive hot springs, resulting in large temperature fluctuations (sometimes more than 40°C). However, the role of large temperature fluctuations in driving diversification of cyanobacteria in eruptive hot springs has not been explored. Here, we examined phototroph community composition and potential photoautotrophic activity in two alkaline eruptive hot springs with similar geochemistry in the Lower Geyser Basin in Yellowstone National Park, WY. We observed distinct cyanobacterial amplicon sequencing variants (ASVs) consistent with allopatry and levels of light-dependent inorganic carbon uptake rates similar to other hot springs, despite large temperature fluctuations. Our data suggest median temperatures may drive phototroph fitness in eruptive hot springs while future studies are necessary to determine the evolutionary consequences of thriving under continuously fluctuating temperatures. We propose that large temperature swings in eruptive hot springs offer unique environments to examine the role of allopatry versus physical and chemical characteristics of ecosystems in driving cyanobacterium evolution and add to the debate regarding the ecology of thermal adaptation and the potential for narrowing niche breadth with increasing temperature. IMPORTANCE Hot spring cyanobacteria have long been model systems for examining ecological diversification as well as characterizing microbial adaptation and evolution to extreme environments. These studies have reported cyanobacterial diversification in hot spring outflow channels that can be defined by distinct temperature ranges. Our study builds on these previous studies by examining cyanobacteria in geysing hot springs. Geysing hot springs result in outflow channels that experience regular and large temperature fluctuations. While community compositions are similar between geysing and nongeysing hot spring outflow channels, our data suggest median, rather than high, temperature drives the fitness of cyanobacteria in geysing hot springs. We propose that large temperature swings may result in patterns of ecological diversification that are distinct from more stable outflows.

IMPORTANCE Hot spring cyanobacteria have long been model systems for examining ecological diversification as well as characterizing microbial adaptation and evolution to extreme environments. These studies have reported cyanobacterial diversification in hot spring outflow channels that can be defined by distinct temperature ranges. Our study builds on these previous studies by examining cyanobacteria in geysing hot springs. Geysing hot springs result in outflow channels that experience regular and large temperature fluctuations. While community compositions are similar between geysing and nongeysing hot spring outflow channels, our data suggest median, rather than high, temperature drives the fitness of cyanobacteria in geysing hot springs. We propose that large temperature swings may result in patterns of ecological diversification that are distinct from more stable outflows.
FC exhibits a more chaotic eruption periodicity-106 min on average, ranging from 25 min to .12 h-but maintains a steady temperature/outflow rate ;68% of the time ( Fig. 1; Fig. S2). JJ exhibits a more regular eruption periodicity-88 min on average, ranging from 76 to 103 min (13), with a continuous but fluctuating discharge ;54% of the time ( Fig. 1; Fig. S3). At FC, phototrophs were first visible in the center of the south outflow channel ;8 m from the source (here designated "FC hot"). Temperatures at FC hot varied by 40.5°C during a 4-h observation period: median of 56.0°C, with maximum of 70.0°C and minimum of 29.5°C (Fig. 1). Downstream from the photosynthetic fringe (;14 m from the source, here designated "FC cool"), water reached a median of 40.0°C over a 4-h period (maximum = 60.0°C, minimum = 29.0°C). At JJ, phototrophs were first visible in the center of the north outflow channel ;24 m from the source (here designated "JJ hot"). At JJ hot, temperatures varied by 38.0°C during a 4-h observation period: median of 61.5°C, with maximum of 75.0°C and minimum of 37.0°C ( Fig. 1). Further downstream (;60 m from the source, here designated "JJ cool"), the median was 42.5°C (maximum of 52.0°C, minimum of 33.0°C). Temperatures deeper in the phototrophic mats were muted compared to that of the water at the mat-water interface ( Fig. 1): at a depth of ;1 cm in the JJ hot mats, the median was 58.5°C, with maximum of 67.5°C and minimum of 40.5°C.
Despite temperature fluctuations of up to 40°C, diversity and the composition of putative phototrophs in the geysing sites were similar to those in nongeysing sites (e.g., references 14 to 16): richness and diversity were lower in phototrophic mats near the upper temperature limit of photosynthesis ( Fig. 2A), and at 97% sequence identity (defined as operational taxonomic units [OTUs]), sequences assigned to Chloroflexi (Roseiflexus and Chloroflexus), Cyanobacteria (Synechococcus and "Candidatus Gloeomargarita"), and Chlorobi ("Candidatus Thermochlorobacteriaceae bacterium GBChlB") were abundant. Notably, sequences affiliated with other cyanobacteria, including "Candidatus Gloeomargarita," Geitlerinema PCC-8501, Leptolyngbya FYG, and Pseudanabaenaceae, were recovered only from the "cool" sites, consistent with increasing diversity with decreasing temperature.
Temperature selects for distinct cyanobacterial ecotypes in nongeysing outflows (e.g., A9 and A ecotypes occur at higher temperatures while B9 and B are observed at lower temperatures [17]). However, in our geysing outflows, all but one of the most abundant Synechococcus cyanobacterial ecotypes (identified as amplicon sequence variants [ASVs]) shared the highest sequence identity with the B9 ecotype. This indicates that median temperature (e.g., 56.0°C at FC hot and 61.5°C at JJ hot) drives ecotype differentiation in fluctuating systems despite regular exposure to higher temperatures that select for distinct ecotypes in nongeysing systems (e.g., A9 and A ecotypes [17]). With a few exceptions (e.g., ASV00002 and ASV00003), the ASVs from JJ and FC were distinct from each other while ASVs from "hot" and "cool" sites within the same hot spring outflow were also distinct (Fig. 2C). These data are consistent with a role for both geographic isolation and temperature in driving diversification and provide a framework to further examine allopatry versus physical and chemical characteristics in driving cyanobacterial evolution and diversification under continuously fluctuating temperatures.
We hypothesized that relatively stable temperatures at FC would result in higher rates of photoautotrophy (based on light-dependent C assimilation rates) compared to JJ and that the large fluctuation in temperatures at both would result in lower photoautotrophy rates compared to steady-temperature sites. We performed microcosm assays by placing mats and water from hot and cool sites at FC or JJ in sealed serum vials that were amended with NaH 13 CO 3 following the methods in reference 14. To test our hypotheses, vials were incubated under the following conditions: (i) "in situ"-vials placed at the sample location, experiencing fluctuating temperatures (Fig. 1); (ii) "steady"-vials placed in nearby noneruptive hot springs meant to mimic lower temperatures observed at each site (FC cool and JJ hot). As expected, in situ rates were higher at FC than at JJ (Fig. 2D). For in situ versus steady, the C assimilation rate for the JJ hot mat held at a steady low temperature (steady in Fig. 2D; 28.1 mg C uptake/g C  Table S1. biomass/h) was lower than that for in situ microcosms while the C assimilation rates between in situ and steady treatments at FC cool were indistinguishable. Overall, lightdependent C assimilation rates at both eruptive sites were lower than rates observed for alkaline phototrophic communities collected from springs with similar temperature and pH in YNP (14,15). For example, in previous studies of phototrophic mats, filaments, and biofilms from nongeysing alkaline hot springs with similar pHs in YNP (e.g., pH 7 to 9), observed light-dependent C assimilation rates ranged from 658.3 to 3813.8 mg C uptake/g C biomass/h (14,15).
We propose that eruptive hot springs are an overlooked but key ecosystem for examining outstanding questions regarding the ecophysiology of hot spring cyanobacteria including whether adaptation to increasingly higher temperatures results in narrowing niche breadth (3,18), the roles of temperature and allopatry in driving diversification, and how Cyanobacteria adapt to high, fluctuating temperatures. Our data indicate stable temperatures might drive higher fitness: light-dependent C assimilation rates were higher at FC which, while more chaotic in eruption periodicity, supported outflows with stable temperatures 68% of the time compared to more regular    Table S2.) Details of the methods are provided in Text S1. eruptivity but continuous temperature variation observed at JJ (changing discharge ;54% of the time). In addition, we recovered sequences most closely related to B9, a lower-temperature cyanobacterial ecotype, across a broad niche breadth (at least in terms of temperature). Thus, while median rather than maximum temperature appears to drive cyanobacterial diversification in geysing outflows, the full range of adaptation to high temperature in hot spring Synechococcus, particularly in ecotypes from geysing systems, warrants further investigation. Indeed, there is rich history of previous studies on cyanobacterial ecotypes and thus an established comparative framework for examining the evolutionary history and ecophysiology of ecotypes in geysing systems through characterization of new isolates and genomic and metagenomics approaches. Data availability. All sequence data including raw reads with quality scores for this study have been deposited in the NCBI Sequence Read Archive (SRA) database under the BioProject number PRJNA756970. Library designations are provided in Table S3.

SUPPLEMENTAL MATERIAL
Supplemental material is available online only. TEXT S1, PDF file, 0.1 MB.