FEMS Microbiology Reviews, fuab010, 45, 2021, 1–25 https://doi.org/10.1093/femsre/fuab010 Advance Access Publication Date: 4 February 2021 Review Article REVIEW ARTICLE The microbial dimension of submarine groundwater discharge: current challenges and future directions Clara Ruiz-González1, *,† , Valentı́ Rodellas2,‡ and Jordi Garcia-Orellana2,3,# 1 Institut de Ciències del Mar (ICM-CSIC), Passeig Marı́tim de la Barceloneta, 37-49, E08003 Barcelona, Spain, 2 Institut de Ciència i Tecnologia Ambientals (ICTA-UAB), Universitat Autònoma de Barcelona, E08193 Bellaterra, Spain and 3 Departament de Fı́sica, Universitat Autònoma de Barcelona, E08193 Bellaterra, Spain ∗ Corresponding author: Institut de Ciències del Mar (ICM-CSIC), Paseo Marı́timo de la Barceloneta, 37-49, E08003 Barcelona, Spain. Tel: 34-93-2309500, Ext. 6047; Fax: 34-93-2309555; E-mail: clararg@icm.csic.es One sentence summary: The authors review the available literature on the microbial aspects of submarine groundwater discharge, from the freshwater aquifers to the coastal ocean, and identify current challenges and future directions to foster knowledge on microbial ecology at the land–ocean interface. Editor: Prof. Corina Brussaard † Clara Ruiz-González, http://orcid.org/0000-0003-3568-4943 ‡ Valentı́ Rodellas, http://orcid.org/0000-0002-5896-9987 # Jordi Garcia-Orellana, http://orcid.org/0000-0002-0543-2641 ABSTRACT Despite the relevance of submarine groundwater discharge (SGD) for ocean biogeochemistry, the microbial dimension of SGD remains poorly understood. SGD can influence marine microbial communities through supplying chemical compounds and microorganisms, and in turn, microbes at the land–ocean transition zone determine the chemistry of the groundwater reaching the ocean. However, compared with inland groundwater, little is known about microbial communities in coastal aquifers. Here, we review the state of the art of the microbial dimension of SGD, with emphasis on prokaryotes, and identify current challenges and future directions. Main challenges include improving the diversity description of groundwater microbiota, characterized by ultrasmall, inactive and novel taxa, and by high ratios of sediment-attached versus free-living cells. Studies should explore microbial dynamics and their role in chemical cycles in coastal aquifers, the bidirectional dispersal of groundwater and seawater microorganisms, and marine bacterioplankton responses to SGD. This will require not only combining sequencing methods, visualization and linking taxonomy to activity but also considering the entire groundwater–marine continuum. Interactions between traditionally independent disciplines (e.g. hydrogeology, microbial ecology) are needed to frame the study of terrestrial and aquatic microorganisms beyond the limits of their presumed habitats, and to foster our understanding of SGD processes and their influence in coastal biogeochemical cycles. Keywords: aquatic prokaryotic communities; submarine groundwater discharge; coastal aquifers; subterranean estuaries; microbial diversity and ecology; ultrasmall prokaryotes INTRODUCTION yet they have been much less studied than surface aquatic ecosystems. Like in deep soils, deep marine sediments or the Aquifers contain a significant portion of the Earth’s freshwa- dark bathypelagic ocean (Acinas et al. 2019; Brewer et al. 2019; ter (∼23 × 106 km3 ; Gleeson et al. 2016) and a large frac- Wörmer et al. 2019), the absence of photosynthesis-derived tion of the global microbial diversity (Magnabosco et al. 2018), Received: 14 July 2020; Accepted: 28 January 2021  C The Author(s) 2021. Published by Oxford University Press on behalf of FEMS. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted reuse, distribution, and reproduction in any medium, provided the original work is properly cited. 1 2 FEMS Microbiology Reviews, 2021, Vol. 45, No. 5 labile organic carbon has forced groundwater organisms (mostly prokaryotic microorganisms—Bacteria and Archaea) to develop BOX 1. diverse strategies to live or persist (Griebler and Lueders 2009). Terms and definitions For example, heterotrophic groundwater microorganisms can Coastal aquifers: Coastal aquifers are frequently defined use ancient organic carbon from rocks (Nowak et al. 2017), as permeable geological units in the subsurface that can allochthonous organic carbon such as plant-derived material transmit significant flows of water and that are connected (Taubert et al. 2019), or can degrade organic contaminants with the sea (Duque et al. 2020). Coastal aquifers consist of (Griebler and Lueders 2009; Smith et al. 2015). Other groundwater unconsolidated surficial deposits (created by alluvial, lacus- microbes are capable of diverse chemoautotrophic metabolisms trine, deltaic, glacial or eolian processes; e.g. Fig. 1A), per- that allow them to fix inorganic carbon using energy from meable rocks (e.g. basalts, sandstones or limestones) or the oxidation of substrates such as ammonium, nitrite and heavily fractured systems (e.g. karstified carbonate rocks or reduced iron and sulfur compounds (Herrmann et al. 2015; volcanic systems; Fig. 1B), where cavities and pores between Anantharaman et al. 2016; Kumar et al. 2017, 2018; Probst et al. mineral grains provide open spaces for water storage and 2017, 2018). Groundwater microorganisms, however, are largely transmission (Freeze and Cherry 1979). under-represented in global diversity catalogs, as for exam- Subterranean estuary: The term subterranean estuary ple less than 2% of the public 16S ribosomal RNA (rRNA) gene was initially introduced to emphasize the importance of sequences (i.e. the most common taxonomic marker for iden- biogeochemical reactions occurring within coastal aquifers, tifying prokaryotic taxa, Appendix 1) were calculated to derive where meteoric groundwater mixes with the seawater that from groundwater organisms (Schloss et al. 2016). The mag- has entered into the coastal aquifer (Moore, 1999). The nitude of this knowledge gap has become evident in recent subterranean estuary is now considered as the part of years: 47 previously unknown phyla were found in a single the coastal aquifer that actively interacts with the ocean, aquifer (Anantharaman et al. 2016), highlighting a tremendous including thus the area of fresh groundwater that is con- potential for taxonomic and metabolic discovery in subsurface nected to the sea (i.e. groundwater that could discharge to ecosystems. the ocean), the mixing zone between meteoric and marine Among aquifers, those located in coastal areas are essen- groundwater, and the permeable sediments filled with com- tial freshwater sources for humanity given that half of the pletely saline groundwater that may exchange with overly- global population is concentrated along the oceans’ shoreline ing marine waters (Duque et al. 2020). (Small and Nichols 2003; Barragán and de Andrés 2015; Michael Submarine groundwater discharge (SGD): Groundwater et al. 2017). Coastal aquifers are also vulnerable ecosystems is often considered as ’any water in the ground’, regardless because the intense anthropogenic pressure is causing pollu- of its composition or origin, and thus it is synonymous with tion and salinization of coastal groundwater resources world- porewater (Burnett et al. 2003). Groundwater can thus con- wide (Ferguson and Gleeson 2012; Werner et al. 2013). Their sist of water of any salinity, ranging from freshwater from direct connection to the ocean through permeable sediments recharged aquifers on land to seawater that has invaded the or rocks (Fig. 1) allows the discharge of groundwater, a pro- subterranean estuary. Based on this concept, SGD is com- cess known as submarine groundwater discharge (SGD; Box monly defined as ’the flow of water through continental 1), but also the entrance of seawater. This often results in an and insular margins from the seabed to the coastal ocean, area of active mixing within the aquifer, referred to as ’subter- regardless of fluid composition or driving force’ (Burnett et ranean estuary’ (Moore 1999; Duque, Michael and Wilson 2020; al. 2003; Moore 2010; Taniguchi et al. 2019). SGD incorporates Box 1). Globally, SGD fluxes (∼1014 m3 yr−1 ) have been esti- thus water flows with different characteristics and is driven mated to be 3–4 times greater than the riverine water fluxes by different forces, including discharge of fresh groundwa- into the oceans (Kwon et al. 2014), representing a very impor- ter driven by the terrestrial hydraulic gradient, exchange tant source of dissolved nutrients and other chemical com- of seawater driven by the movement of the freshwater– pounds (e.g. metals, carbon, pollutants, greenhouse gases) to saltwater interface, or circulation of seawater driven by the coastal ocean (Moore 2010; Rodellas et al. 2015; Cho et al. tidal inundation or wave run-up, among others (Santos et 2018; Mayfield et al. 2021). SGD can comprise meteoric ground- al. 2012; Robinson et al. 2018). water originating through aquifer recharge (fresh groundwa- ter), seawater infiltrated into coastal aquifers and sediments (saline groundwater), or a mixture of both (brackish groundwa- ter) (Moore 2010; Fig. 1). At the global scale, SGD is mainly com- posed of saline groundwater, with fresh groundwater represent- Coastal aquifers and SGD are of special interest from ing only <1% of total SGD but sometimes being the dominant a microbiological perspective. First, coastal aquifers are component at a more local scale (Luijendijk, Gleeson and Moos- located at the land–ocean interface, a transition zone with dorf 2020). In general, most SGD to the coastal ocean occurs steep physicochemical gradients (e.g. freshwater–seawater, through slow and diffusive seeps through permeable sediments oxic–anoxic) that promote the establishment of spatially (Fig. 1A), but in systems with fractured rocks or preferential flow heterogeneous microbial communities (Héry et al. 2014; Beck paths (e.g. karstic carbonate or volcanic systems) groundwa- et al. 2017; Seibert et al. 2020). Second, given that groundwater ter can also be transferred to the coastal ocean through point- microbes control the cycling of nutrients, carbon and trace sourced seeps or submarine springs (Tovar-Sánchez et al. 2014; elements in aquifers (Hunter, Wang and Van Cappellen 1998; Fig. 1B). Griebler and Lueders 2009; Flynn et al. 2013; Griebler and Avramov 2015; Meckenstock et al. 2015; Hoffman et al. 2020), microbial activity in subterranean estuaries will determine the chemical composition of the groundwater reaching the ocean, but little is known regarding the metabolic functions of microorganisms in coastal aquifers (Santoro 2010; Seibert Ruiz-González et al. 3 Figure 1. Simplified scheme of SGD, including diffusive seepage through unconsolidated sediments (A) and point-sourced discharge, as in karstic or volcanic aquifers (B). Most of the groundwater flow in the latter systems is likely to occur through conduits, but it can also occur through the porous and thinly fissured rocks, developing mixing zones similar to systems with unconsolidated sediments (dotted lines) (B). (All illustrations by C. Ruiz-González). et al. 2020). Finally, despite the relevance of bacterioplank- nonetheless, a solid basis for anticipating the main challenges ton communities for marine food webs and biogeochemical that the study of the microbiology in subterranean estuaries processes (e.g. Whitman, Coleman and Wiebe 1998; Gasol et al. might face. Based on this knowledge, in this section we identify 2008), only a few studies have addressed marine bacterial taxo- four main challenges related to groundwater microbial idiosyn- nomic and functional responses to SGD (see references in Lecher crasies that future investigations should take into account for an and Mackey 2018). Improving our understanding on the micro- accurate understanding of microbial ecology at the land–ocean bial players along the entire aquatic continuum from freshwater interface. aquifers to the coastal ocean is essential to gain insight into Additional challenges related to sampling the subsurface the complex processes underlying coastal groundwater biogeo- or SGD sources are not discussed in this review. For example, chemistry, the potential variations in these processes due to aquifers need to be accessed through wells, piezometers, bore- global change, and their influence on marine ecosystems. holes, cores or excavations. This not only makes the collection The aim of this review is to provide an overview of the cur- of groundwater and aquifer solids technically and economically rent knowledge on microbial ecology along the land–ocean tran- challenging (Lehman 2007), but also precludes an accurate char- sition zone connected by SGD. We first identify current chal- acterization of the geologic, physicochemical and biological 3D lenges for the study of coastal groundwater microbiota based heterogeneity of subsurface ecosystems (Robinson et al. 2018; on recent discoveries in inland aquifers, which have been much Smith et al. 2018, Folch et al. 2020). Similarly, SGD is a much more intensely studied than coastal groundwater systems (see more diffuse and heterogeneous source of water to the ocean the section ‘Challenges for assessing microbial diversity and than riverine discharge, which translates into many uncertain- their function in the coastal subsurface: lessons learned from ties associated with the magnitude and spatiotemporal vari- inland aquifer research’). Second, we review the available lit- ability of this process (Burnett et al. 2006). Current and future erature on microbial diversity and ecology along subterranean efforts to improve the characterization of aquifers and ground- estuaries and the adjacent coast influenced by SGD (see the sec- water seeps to the ocean will be instrumental for developing the tion ‘Microbial ecology at the terrestrial–marine interface con- research lines identified in this review. nected by groundwater’). Finally, we provide future research directions and strategies that may foster our understanding of the microbial dimension underlying SGD processes (see the sec- Novel groundwater microbial taxa can be missed by tion ‘Future directions and research avenues’). current sequencing technologies The study of groundwater microorganisms has evolved progres- CHALLENGES FOR ASSESSING MICROBIAL sively together with the development of taxa identification tech- DIVERSITY AND THEIR FUNCTION IN THE niques (see Griebler and Lueders 2009 for an excellent review COASTAL SUBSURFACE: LESSONS LEARNED on groundwater prokaryotic diversity in inland aquifers since the earliest investigations). A compilation of studies examin- FROM INLAND AQUIFER RESEARCH ing microbial communities in coastal aquifers (Table S1, Sup- Despite the growing interest in inland groundwater microbiol- porting Information) evidences a clear technical shift mov- ogy in the last decades, coastal aquifers have received com- ing from approaches such as clone libraries, quantitative poly- parably little attention. A simple search in the ISI Web of merase chain reaction (qPCR) or fingerprinting of taxonomic or Science database looking for articles published between 2000 functional marker genes, to high-throughput sequencing (HTS) and 2020 containing ‘groundwater AND microbial∗ ’ in the title technologies such as 454 or Illumina (Logares et al. 2012; see retrieves 369 studies that show a clear increasing trend over time Appendices 1 and 2 for definitions of microbiology concepts (Fig. 2). Conversely, only 16 articles are recovered if we include and methodologies). Currently, most studies on coastal aquifer terms such as ‘submarine’, ‘seawater’, ‘subterranean estuar∗ ’ microbiota have been based on the high-throughput amplicon or ‘coastal aquifer’, highlighting a wide gap of knowledge on sequencing of the 16S rRNA gene (Table S1, Supporting Infor- the marine–groundwater interface from a microbial perspec- mation), which allows a much more detailed characterization tive. This growing body of inland groundwater research provides, of microbial communities than traditional approaches such as 4 FEMS Microbiology Reviews, 2021, Vol. 45, No. 5 Figure 2. Rising interest in groundwater microbiology over the last two decades, and scarcity of studies targeting coastal aquifers. Number of articles published per year (2000–2020) using the search terms ‘groundwater AND microbial∗ ’ (gray, n = 369) or adding ‘submarine’, ‘seawater’, ‘coastal aquifer∗ ’ or ‘subterranean estuar∗ ’ to the search term set (black, n = 16) in the title, as indexed in the ISI Web of Science. Only articles in English were considered (search performed in August 2020). clone libraries or fingerprinting techniques. However, this tech- which primers to use, precludes generalizations and presum- nique may be limited for the description of groundwater micro- ably renders a biased view of groundwater microbial inhabitants biota because it relies on the PCR amplification of particular in coastal areas. genes (e.g. the 16S rRNA gene) through the use of primers, which are designed based on the sequences available in public Ultrasmall prokaryotic groups are abundant in databases. Therefore, diversity recovery through primer-based groundwater ecosystems approaches can be biased against taxa that are less represented in public sequence catalogs, such as those from aquifers (and Another peculiarity of groundwater microbial communities is coastal aquifers in particular). In addition, primer pairs target- that they can shelter minuscule prokaryotes (Brown et al. 2015; ing different regions of the 16S rRNA gene can be biased against Luef et al. 2015; Proctor et al. 2018; Herrmann et al. 2019) that different taxa (Klindworth et al. 2013), preventing the direct com- may not be captured by current sampling strategies. Prokary- parison between studies using different primer pairs. Table S1 ote biomass for DNA extraction or microscopic visualization (Supporting Information) actually highlights a lack of consen- is often collected onto 0.2-μm-pore-size filters (Table S1, Sup- sus on primer usage, since up to 13 different 16S rRNA gene porting Information) and these cells can be smaller than this primer pairs have been used among the 25 studies describing pore size (Nakai 2020). The few groundwater studies (all con- groundwater prokaryotic diversity in coastal aquifers (Tables S1 ducted in inland aquifers) that used sequential filtration of water and S2, Supporting Information). These primer pairs differ in through 0.2- and 0.1-μm-pore-size filters unveiled that the frac- their coverage of some of the main phyla detected in ground- tion of small cells passing through the 0.2-μm filters is often water (Fig. 3A). Moreover, most fail to capture novel ground- dominated by members of the superphylum Ca. Patescibacteria water bacterial lineages discovered by studies applying primer- (Miyoshi, Iwatsuki and Naganuma 2005; Herrmann et al. 2019; free approaches (Fig. 3A–C), such as the superphylum Candida- He et al. 2021) or by the small Archaea belonging to the DPANN tus (Ca.) Patescibacteria (also known as candidate phyla radi- radiation (Castelle et al. 2018; He et al. 2021). The reconstruc- ation, CPR), which includes a high diversity of small unculti- tion of individual genomes from metagenomes (MAGs) or from vated bacterial groups (Brown et al. 2015; Luef et al. 2015; Anan- individually sorted cells (SAGs; Appendix 2) has yielded insight tharaman et al. 2016; Castelle et al. 2018; He et al. 2021; Tian into the metabolic potential of these uncultivated groundwa- et al. 2020; see the section ’Ultrasmall prokaryotic groups are ter groups, unveiling unusually reduced genomes and gaps in abundant in groundwater ecosystems’), or archaeal members of core metabolic potential that are typical of symbionts (Brown the DPANN radiation (Diapherotrites, Parvarchaeota, Aenigmar- et al. 2015; Luef et al. 2015; Youssef et al. 2015; Anantharaman chaeota, Nanoarchaeota, Nanohaloarchaea), also characterized et al. 2016; Castelle et al. 2018; Probst et al. 2018; He et al. 2021). by small cell sizes and genomes (Ludington et al. 2017; Castelle Members of Ca. Patescibacteria have been shown to be active in et al. 2018; He et al. 2021). The discovery of these groups has groundwater (Taubert et al. 2019) and seem to prefer a plank- been possible through techniques such as metagenomics and tonic lifestyle (Lazar et al. 2019), although their detection in metatranscriptomics that are not biased by the performance of larger planktonic size fractions (e.g. >2 μm) has been inter- primers (Appendix 2), but so far none of the microbial studies preted as evidence of episymbiosis (i.e. surface attachment to in subterranean estuaries has applied these techniques (Table other microorganisms; He et al. 2021). A recent study, however, S1, Supporting Information). Hence, despite the global distribu- claims that Ca. Patescibacteria and DPANN genomic features tion of the investigations on coastal aquifers conducted so far may not be indicative of symbiotic lifestyles but rather of an (Fig. 4), the unknown accuracy of the existing primers to recover ancestral absence of electron transport chains, and thus that groundwater diversity, together with the lack of consensus on they rely on fermentative metabolisms (Beam et al. 2020). In Ruiz-González et al. 5 Figure 3. Differential diversity coverage among the 16S rRNA gene primer pairs used in studies on subterranean estuaries (STE). The primers shown include those used in the studies included in Tables S1 and S2 (Supporting Information) to characterize prokaryotic diversity, which represent primers targeting only Archaea, only Bacteria or both (indicated by an asterisk). The different panels represent the diversity coverage of some of (A) the main bacterial phyla found in groundwaters, (B) the different phyla within the superphylum Ca. Patescibacteria (underlined in panel A) and (C) the main archaeal phyla known to be present in groundwater. The diversity coverage was tested in silico using the online tool TestPrime (Klindworth et al. 2013), by comparing the sequences captured by each primer pair against the SILVA taxonomic database, allowing for zero mismatches (i.e. no differences between the compared sequences). Primer sequences are indicated in Table S2 (Supporting Information). 6 FEMS Microbiology Reviews, 2021, Vol. 45, No. 5 Figure 4. Location of the studies assessing natural prokaryotic communities in coastal aquifers connected to the sea (’Aquifer’ studies, orange symbols) and the effects of SGD on natural marine bacterioplankton communities (’SGD effects’ studies, turquoise symbols). Microbial investigations in coastal aquifers are distinguished according to the typology of the aquifer (unconsolidated sediments vs carbonate or volcanic rocks; see Table S1, Supporting Information). Experimental and field studies of SGD effects on marine microbial communities are also distinguished (see Table S3, Supporting Information). Numbers correspond to the studied sites reported in Table S1 (Supporting Information), and letters correspond to the studies on SGD effects reported in Table S3 (Supporting Information). any case, despite their apparent importance in the subsurface, Such miniaturized cells could comprise a microbial ’seed the relevance of these ultrasmall prokaryotes in coastal aquifers bank’ (i.e. a reservoir of dormant bacteria that can reactivate remains unknown due to the previously mentioned primer lim- upon changes in conditions; Lennon and Jones 2011). In gen- itations for capturing these taxonomic groups (Fig. 3), and the eral, though, little is known about the relevance and extent lack of studies exploring the diversity within the ultrasmall of dormancy in groundwater systems. A recent experimental planktonic fraction (<0.2 μm) in subterranean estuaries (Table study suggested that bacteria in oligotrophic aquifers can be S1, Supporting Information; see the section ‘Microbial abun- locally co-limited by organic carbon and phosphorous, but that dance, diversity and environmental drivers along subterranean they can reactivate and reach high bacterial abundances when estuaries’). amended with different organic carbon sources and nutrients (Hofmann and Griebler 2018), or with soil dissolved organic Miniaturized inactive or dormant taxa may be matter (DOM) richer in labile compounds (Hofmann et al. 2020). Studies in deep marine sediments, where microbes face abundant in groundwater increasing starvation with depth, have shown that dormant Prokaryotes can also miniaturize in response to stress or star- endospore-forming cells can outnumber actively growing cells vation conditions (Velimirov 2001) becoming inactive or dor- (Wörmer et al. 2019) and that most microorganisms appear to mant (i.e. in a reversible state of low metabolic activity; Lennon be surviving instead of growing (Bradley, Amend and LaRowe and Jones 2011). Due to the typical scarcity of freshly produced 2019). Thus, rather than growth, maintenance mechanisms carbon in subsurface ecosystems (Hofmann and Griebler 2018), might be key to thrive in such oligotrophic habitats. Finally, groundwater microbes have probably evolved diverse strategies environmental stresses such as salinity changes also appear to to persist in resting or dormant states until more favorable con- trigger dormancy in bacteria (Sachidanandham and Yew-Hoong ditions occur. Given that an important fraction of the cells in Gin 2009; Aanderud et al. 2016), so this mechanism might allow aquifers are believed to be inactive (Griebler and Lueders 2009), repeated activation and inactivation of prokaryotes exposed it is possible that a fraction of the ultrasmall cell pool rep- to changing conditions caused by the mixing of seawater and resents miniaturized dormant taxa rather than obligate ultra- groundwater in coastal aquifers. microbacteria (sensu Nakai 2020). In support of this, the expo- Groundwater microbial communities might also include sure of bacterial groundwater isolates to the oligotrophic con- ’allochthonous’ microorganisms transported from external ditions typical of pristine aquifers caused a significant cell size sources that are hydraulically connected to the aquifer (e.g. soils, reduction (Herrmann et al. 2019). Also, we have observed that rivers, lakes; Hubalek et al. 2016; Graham et al. 2017; Meier et al. the incubation of 0.2-μm prefiltered groundwater from a coastal 2017; Herrmann et al. 2019, Fillinger et al. 2021). Although some Mediterranean aquifer results in pronounced increases in the of these transported microorganisms may comprise a travel- abundance, cell size and heterotrophic production of the bac- ling groundwater ’seed bank’, being able to colonize different teria passing through the 0.2-μm filter (Fig. 5), indicating that habitats within groundwater ecosystems (such as sediments or some cells had been miniaturized in situ (and were presumably rock surfaces; Griebler et al. 2002; Griebler, Malard and Lefébure dormant) likely due to the prevailing carbon scarcity. 2014; Fillinger et al. 2019b; Herrmann et al. 2019; see the section Ruiz-González et al. 7 Figure 5. Coastal aquifers hide miniaturized inactive taxa that can reactivate quickly upon changes in conditions. Increase in prokaryote abundances, average cell size and prokaryotic heterotrophic activity (prokaryotic production, measured as the incorporation of radioactive leucine) after an ∼4-day incubation (in situ temperature, dark conditions, no nutrient or carbon addition) of groundwater collected from a coastal alluvial aquifer (Argentona, western Mediterranean; see Folch et al. 2020). Groundwater samples representing unfiltered (dark blue) and filtered through a 0.2-μ m-pore-size filter (light blue) are distinguished. Prokaryotic abundance and cell size were estimated by flow cytometry (Gasol and del Giorgio 2000), and prokaryotic production was estimated measuring the incorporation of radiolabeled 3 H-leucine (Kirchman et al. 1985; Smith and Azam 1992). Note that bacteria passing through the 0.2-μ m filter became larger in size and increased their abundances and activity, suggesting that they were not true ultra-microbacteria, but rather miniaturized cells due to the prevalent conditions of organic carbon limitation. ’Groundwater biomass and activity is concentrated on aquifer which allows for an intense interaction between groundwa- solids’), others might just be allochthonous maladapted taxa ter and solids that influences groundwater chemistry and sur- able to persist inactively (or dead) during transit in groundwa- face characteristics of the geological matrix (Moore 2010; Tovar- ter, with no influence in this subsurface ecosystem. Sánchez et al. 2014). This physical structure of aquifers chal- The presence of inactive or allochthonous taxa may be of lenges the study and the sampling of aquifer microbiota, as special relevance in coastal aquifers, where communities may prokaryotes may not occur only as free-living planktonic cells be continuously subjected to a bidirectional inoculation of taxa but also attached onto solid aquifer surfaces as individual cells from the marine and fresh groundwater endmembers (Adyasari or aggregates (Smith et al. 2018). Actually, most groundwater et al. 2019). However, whether these transported microorgan- microbial biomass and activity seems to be concentrated on isms tolerate or inactivate upon such wide ranges of condi- solid surfaces (Griebler et al. 2002; Smith et al. 2018), often show- tions remains an open question, mostly because all studies ing a highly patchy distribution at the microscale due to the assessing microbial diversity have been based on sequencing microniches than can be established within pores and cavities of the 16S rRNA gene (Table S1, Supporting Information), which (Goldscheider, Hunkeler and Rossi 2006; Schmidt, Cuthbert and does not allow distinguishing between the active, inactive and Schwientek 2017; Smith et al. 2018). even dead components of communities. The recent finding that Not many studies have compared the free-living and extracellular or ’relic’ DNA can persist for long periods in soils the attached microbial groundwater components. In inland and sediments (Carini et al. 2016) further implies that some of aquifers, these investigations have shown that microbial com- the detected taxa might simply be remnants of formerly active munities adhered to karst sediments display higher activities communities, and the very small proportion of the total micro- than the suspended ones, likely due to a better access to organic bial biomass (0.02–6.36%) found to be active in a coastal allu- carbon and nutrients (Wilhartitz et al. 2009). The ratio of attached vial aquifer along a two-year period (Velasco Ayuso et al. 2010) versus free-living bacteria was shown to increase pronouncedly supports that only a small fraction of the detected diversity as oligotrophic conditions increased, ranging from ∼50:1 in a may be locally functioning in these coastal systems. Distin- contaminated site to >1600:1 in pristine groundwater (Griebler guishing between the active and the inactive (dormant or dead) et al. 2002). Attached microbial communities differ notably from groundwater microbial community members is therefore essen- the planktonic ones in their taxonomic composition, as has been tial, because the inactive pools of taxa may mask spatial or tem- shown by the direct sampling of microorganisms from aquifer poral patterns in taxonomic composition, obscuring our com- rocks and groundwater (Herrmann et al. 2017; Meier et al. 2017; prehension of the observed linkages between the environment, Lazar et al. 2019), or by incubation of sterilized sediments within the taxonomic structure of communities, and their functions wells (Zhou, Kellermann and Griebler 2012; Flynn et al. 2013) and (Niño-Garcı́a, Ruiz-González and del Giorgio 2016; Wisnoski et al. in experimental microcosms or mesocosms (Longnecker and 2020). Kujawinski 2013; Fillinger et al. 2019b). Results from these stud- ies show that groundwater transports taxa able to colonize sedi- Groundwater biomass and activity is concentrated on ments (Fillinger et al. 2019b) and that attached communities are aquifer solids more stable over time than planktonic microorganisms (Zhou, Kellermann and Griebler 2012), but more responsive to the pres- Compared with surface aquatic ecosystems, aquifers tend to ence of bacterial grazers (Longnecker and Kujawinski 2013). have much higher solid–water ratios and slower flow veloci- No study on coastal aquifers has directly compared ties (usually on the order of millimeters to meters per day), groundwater and sediment microbial communities (Table S1, 8 FEMS Microbiology Reviews, 2021, Vol. 45, No. 5 Supporting Information). However, prokaryotes living on sand biomass than ground- or porewater (see the section ’Groundwa- grains and in saline porewater were found to differ at fine ter biomass and activity is concentrated on aquifer solids’). taxonomic levels (Gobet et al. 2012), and communities from In terms of microbial community composition, the recent two planktonic size fractions (0.2 and 0.45 μm, which may application of high-throughput sequencing technologies (Table include small suspended particles) differed taxonomically in a S1, Supporting Information) has shown that subterranean estu- subterranean estuary (Chen et al. 2019b). Hence, targeting only aries hide heterogeneous and diverse prokaryotic communities the planktonic or the sediment-attached compartment may varying pronouncedly at small vertical and horizontal scales lead to a rather incomplete view of microbial ecology in inland (e.g. Héry et al. 2014; Unno et al. 2015; Ye et al. 2016; Beck et al. or coastal groundwater. Given the relevance of hydrologic 2017; Adyasari et al. 2019; Chen et al. 2019a; Hong et al. 2019; boundaries and flow paths for the assembly and functioning Adyasari et al. 2020; Chen et al. 2020b; Sola, Vargas-Garcı́a and of the biofilm communities in inland shallow aquifers (Smith Vallejos 2020). Such studies have generally reported a domi- et al. 2018), it is likely that subterranean estuaries are also nance of the bacterial phyla Proteobacteria (mostly Gamma- and characterized by a highly complex spatiotemporal structuring Alphaproteobacteria), Bacteroidetes and/or Actinobacteria, with of the sediment-attached microbial component. other phyla like Planctomycetes, Chloroflexi or the endospore- forming Firmicutes being locally important depending on the environmental conditions (Héry et al. 2014; McAllister et al. 2015; MICROBIAL ECOLOGY AT THE Unno et al. 2015; Ye et al. 2016; Sang et al. 2018, 2019; Chen et al. TERRESTRIAL–MARINE INTERFACE 2019a; Hong et al. 2019; Zhang et al. 2021). Taxa belonging to CONNECTED BY GROUNDWATER the ultrasmall Ca. Patescibacteria have rarely been observed, often comprising <2% of the 16S rRNA sequences (Héry et al. Environmental transitions or ecotones are relevant landscape 2014; Unno et al. 2015; Ye et al. 2016; Chen et al. 2019a), but components. Due to the mixing of resources and communi- locally reaching up to 10% (Adyasari et al. 2020). These abun- ties, these sites are often hotspots of biogeochemical activity, dances might be underestimates due to the previously men- where abrupt physicochemical changes or gradients determine tioned primer biases for capturing these novel groups (see the large variations in microbial community composition and activ- section ’Novel groundwater microbial taxa can be missed by cur- ity (e.g. McClain et al. 2003; Stegen et al. 2016). Subterranean rent sequencing technologies’; Fig. 3) and/or to the loss of ultra- estuaries (Box 1), which are ubiquitous along coastlines, repre- small cells during 0.2 μm filtration (Table S1, Supporting Infor- sent ideal systems to study microbial adaptations to the abrupt mation; see the section ’Ultrasmall prokaryotic groups are abun- environmental transitions created by the mixing of groundwa- dant in groundwater ecosystems’). ter and seawater, but they have been largely understudied com- Archaea are generally less abundant than bacteria, but they pared with surface waters, estuaries or inland aquifers. In this harbor functional groups such as methanogens, ammonia oxi- section, we review the current knowledge on microbial commu- dizers or nitrifiers that may play key roles in aquifer biogeo- nities from coastal aquifers based on the available studies (Table chemistry (Santoro et al. 2008; Davis and Garey 2018; Adyasari S1, Supporting Information; Fig. 4), highlighting what we have et al. 2020; see the section ’Microbial processes within subter- learned about the microbial diversity, drivers and functioning of ranean estuaries’). Archaea often comprise <1% of total prokary- these unexplored ecosystems. We also provide an overview of otic sequences in the planktonic compartment (Unno et al. 2015; the few studies exploring the effects of SGD on coastal bacterio- Adyasari et al. 2019) but they can reach higher abundances (5% plankton communities and the potential ecological implications up to 50%) under specific conditions in both groundwater (Sang (Table S3, Supporting Information; Fig. 4). et al. 2018; Adyasari et al. 2020) or sediments (Missimer et al. 2014; Hong et al. 2019). Euryarchaeota, Chrenarchaeota, Thau- Microbial abundance, diversity and environmental marchaeota and Ca. Woesearchaeota are generally the dominant archaeal phyla (Rogers and Casciotti 2010; Missimer et al. 2014; drivers along subterranean estuaries Unno et al. 2015; Davis and Garey 2018; Sang et al. 2018; Hong et al. Despite the recent rise in the number of microbial studies focus- 2019; Adyasari et al. 2020) but phyla such as Ca. Bathyarchaeota ing on aquifers connected to the sea (Table S1, Supporting Infor- and Ca. Parvarchaeota (the latter belonging to the DPANN radia- mation; Fig. 4), little is known regarding the abundance, diver- tion) have also been observed (Chen et al. 2019b; Hong et al. 2019). sity and the roles of the microorganisms inhabiting these sys- Temperature, salinity, pH, dissolved oxygen and nutrient tems. Only a few studies have quantified prokaryotic cell abun- concentration, redox conditions and DOM quality have com- dances or bulk heterotrophic production in subterranean estuar- monly been identified as potential environmental drivers of the ies (Table S1, Supporting Information), even though this is some- observed compositional variations in prokaryotic communities thing that is routinely done in microbial investigations on sur- from coastal aquifers (McAllister et al. 2015; Beck et al. 2017; Davis face waters. Applying epifluorescence microscopy to porewater and Garey 2018; Jiao et al. 2018; Adyasari et al. 2019, 2020; Chen samples in a sandy beach aquifer, Santoro et al. (2008) reported et al. 2019a; Hong et al. 2019; Jiang et al. 2020; Sola, Vargas-Garcı́a up to 1.4 × 106 cells mL−1 . Higher maximum abundances were and Vallejos 2020). Hydrological conditions and connectivity observed in a coastal alluvial aquifer that is well connected to with surrounding ecosystems, although less commonly studied, the surface, where large seasonal variations in cell numbers also appear as relevant drivers of the distribution of communi- (range 2 × 105 to >6 × 107 cells mL−1 ) were related to changes ties in these transition zones. For example, changes in micro- in temperature and nutrient concentrations (Velasco Ayuso et al. bial abundances, activity or composition have been reported in 2009a,b). These values fall all in the upper range of those typi- coastal aquifers driven by dynamic changes in environmental cally found in pristine inland groundwater (102 –106 cells mL−1 ; gradients due to tides or waves (Santoro et al. 2008; McAllister Griebler and Lueders 2009), but the scarcity of data prevents et al. 2015), seasonal changes in groundwater level (Menning et any accurate comparison. In sediments, up to 2.6 × 108 cells al. 2018; Chen et al. 2020a) or varying degrees of connectivity cm−3 were measured (Beck et al. 2017), implying that likewise with the surface or coastal seawater (Velasco Ayuso et al. 2009a,b, in inland aquifers, coastal sediments harbor higher prokaryotic 2010, 2011; Adyasari et al. 2019). Ruiz-González et al. 9 The reported changes in microbial community structure for example, is known to impose a strong filter for microbial cannot always be predicted from the mixing of freshwater communities, as large changes in microbial community com- and seawater, since the specific conditions in the mixing zone position are always found along salinity gradients (Lozupone promote the establishment of unique and diverse populations and Knight 2005). In a comprehensive review on the microbial of taxa that are rare elsewhere (Héry et al. 2014; Hong et al. 2019). nitrogen cycling in subterranean estuaries, Santoro (2010) sug- This is in accordance with experimental studies showing that gested that increases in salinity may favor metabolisms such the mixing of freshwater and marine bacterial assemblages as DNRA, sulfate reduction and carbon remineralization while can promote the growth of initially rare taxa (Shen, Jürgens decreasing denitrification, nitrification and methanogenesis. and Beier 2018; Rocca et al. 2020). As the activation of rare Both anammox and denitrification potential rates were shown taxa can impact relevant ecosystem processes (Sjo¨stedt et al. to decrease with increasing salinity in low permeable inter- 2012; Aanderud et al. 2015; Stegen et al. 2016), such hotspots of tidal sediments (Jiao et al. 2018). Also, higher rates of methano- microbial specialists along subterranean estuaries could have genesis within subterranean estuaries have been observed in strong implications in aquifer biogeochemistry. However, most fresher or brackish portions than in sulfate-rich saltwater sites of the microbial investigations in subterranean estuaries have (Brankovits et al. 2018; Pain, Martin and Young 2019), given that been restricted to a single sampling time or have low spatial methanogens may be outcompeted for chemical substrates by resolution (Table S1, Supporting Information), so we lack a com- sulfate-reducers (Whiticar 1999). prehensive understanding of the hydrologic and environmental Besides salinity, many other factors such as redox condi- drivers of these communities over time and space. tions can differentially affect the distribution of microbial func- tional groups. Slomp and Van Cappellen (2004) classified sub- terranean estuaries based on the redox state of the fresh and Microbial processes within subterranean estuaries saline groundwater that mix in the coastal zone, which has been Microbial activity in subterranean estuaries is mainly fueled by proposed as a useful framework for predicting biogeochemical the degradation of organic matter or the oxidation of inorganic cycles in different subterranean estuaries (Santoro 2010). How- compounds supplied by fresh groundwater and/or infiltrated ever, reality is extremely much more complex because there can seawater (Seibert et al. 2020; Fig. 6). For heterotrophic microor- be overlap in redox zones (Canfield and Thamdrup 2009) and ganisms, the principal pathways of organic matter degradation a large spatiotemporal heterogeneity in this metabolic zona- vary depending on the availability of electron acceptors, and tion at the microscale; for instance, discrete microenvironments include aerobic respiration, denitrification, manganese reduc- and microbial communities can be established within intersti- tion, iron reduction, sulfate reduction and methanogenesis tial pores (Schmidt, Cuthbert and Schwientek 2017; Smith et al. (Canfield and Thamdrup 2009). These pathways usually occur in 2018), and respiration processes can be spatially restricted to a predictable sequence along a groundwater flow path, which the fringes of contaminant plumes due to depletion of electron results in freshwater aquifers usually having a characteristic acceptors in the plume core (Meckenstock et al. 2015; Pilloni et al. spatial zonation of redox conditions and active microbial pop- 2019). In any case, in order to be able to predict the distribu- ulations (Hunter, Wang and Van Cappellen 1998). In coastal tion of different microbial groups and their functions, we need aquifers, this metabolic zonation may be much more dynamic to gain insight into their environmental tolerances, preferences given that the availability of organic matter, oxygen and other and metabolic capacities. electron acceptors and donors can change over time due to the dynamic mixing of fresh and marine waters and the complex interaction of terrestrial and marine forces (Slomp and Van Cap- Dynamics of microbial functional groups in coastal pellen 2004; McAllister et al. 2015; Robinson et al. 2018; Fig. 6). aquifers Hence, once the available oxygen supplied by fresh groundwa- ter or seawater is consumed by aerobic microorganisms, other The recent application of approaches such as metagenomics electron acceptors that are differentially delivered by fresh- or and metatranscriptomics, or stable isotope probing (SIP), has marine water will be used for anaerobic respiration by deni- yielded much insight into the metabolic potential of groundwa- trifiers, manganese- and iron-reducers, and finally by sulfate ter microbial communities from inland aquifers (see the section reducers and methanogens (McAllister et al. 2015; Beck et al. ’Improving the description of microbial diversity and function in 2017; Hong et al. 2019; Montiel et al. 2019). In turn, the reduced coastal aquifers’). On the contrary, most recent studies on sub- products of these reactions can be oxidized by other microorgan- terranean estuaries have exclusively focused on the taxonomic isms through processes such as anammox (anaerobic ammo- composition of communities (Table S1, Supporting Information), nium oxidation), nitrification, and iron, sulfur or methane oxida- which does not provide information on their metabolisms or tion, which are often coupled to CO2 fixation by chemolithoau- activity rates. Some of these studies have attempted to pre- totrophs (Santoro, Boehm and Francis 2006; Santoro et al. 2008; dict microbial function based on taxonomic identity (Table S1, McAllister et al. 2015; Fig. 6). Although accumulating evidence Supporting Information), unveiling a high diversity and highly from inland aquifers suggests that chemolithoautotrophy can be heterogeneous distributions of potential microbial metabolisms more important than previously believed in subsurface ecosys- (Davis and Garey 2018; Hong et al. 2019; Adyasari et al. 2020) tems (Alfreider, Schimer and Vogt 2012; Kellermann et al. 2012; and highlighting the importance of freshwater–saltwater tran- Herrmann et al. 2015; Hutchins et al. 2016; Jewell et al. 2016; sition zones as potential hotspots of functional diversity (San- Brankovits et al. 2017), the relative contribution of autotrophic toro, Boehm and Francis 2006; Hong et al. 2019). Although metabolisms to total carbon cycling is currently unknown in these metabolic predictions may be inaccurate or biased subterranean estuaries. because they depend on the sequenced genomes available (Sun, The physicochemical conditions imposed by the mixing Jones and Fodor 2020), the identification of particular taxa between fresh groundwater and marine seawater in the sub- with known metabolisms might provide useful clues into the terranean estuary lead to a dynamic and heterogeneous dis- processes that could be operating in a given area (Montiel tribution of functional capacities and metabolic rates. Salinity, et al. 2019). 10 FEMS Microbiology Reviews, 2021, Vol. 45, No. 5 Figure 6. Simplified representation of main biogeochemical processes mediated by prokaryotes in subterranean estuaries of importance for SGD chemical fluxes, and key functional marker genes. The different color fonts indicate metabolic pathways involved in the cycling of carbon (black), nitrogen (blue), manganese (orange), iron (green) and sulfur (purple). The oxidation of the reduced products deriving from these respiratory processes (i.e. NH4 + , Mn2+ , Fe2+ , H2 S) is often coupled to CO2 fixation by chemolithoautotrophs. Gray arrows denote electron acceptors; black solid arrows represent the transformations of chemicals. Marker genes used in groundwater studies to identify microbial functional groups involved in each process are indicated in bubbles (and are defined in Table 1). The most common products of biotic transformations in the subterranean estuary that might be delivered to the ocean through SGD are indicated inside the light blue arrows. This is an abstraction of the system, assuming that the freshwater–saltwater mixing zone behaves as a redox boundary between oxic fresh groundwater and saline groundwater of lower oxygen content, and that the circulation of oxygenated seawater through the surface layer of marine sediments (lighter area) also leads to a comparable sequence of transformations established along a depth profile (indicated as ’. . . ’). DNRA, dissimilatory nitrate reduction to ammonium; anammox, anaerobic ammonium oxidation; OM, organic matter. Other studies in subterranean estuaries have directly tracked because microbial studies in coastal aquifers have rarely cou- functional microbial groups using primer-based approaches pled microbial information with actual metabolic rates (Table S1, such as clone libraries or quantitative PCR (qPCR). These have Supporting Information). For example, the different responses reported pronounced spatial and temporal variations in the of ammonia-oxidizing bacteria or ammonia-oxidizing archaea abundance or diversity of genes such as the ammonia monooxy- to salinity resulted in large shifts in the ratio between the genase gene (amoA) involved in the aerobic oxidation of ammo- two along a subterranean estuary (Santoro et al. 2008), but the nia (Santoro et al. 2008; Rogers and Casciotti 2010; Beck et al. implications of this taxonomic shift for the in situ nitrification 2017), nitrate reductase encoding genes (nirK and nirS) involved rates remain unknown. Using stable nitrogen tracer (15 NO3 − ) in denitrification (Santoro, Boehm and Francis 2006; Jiao et al. assays, decreases in anammox and denitrification rates were 2018) or genes involved in sulfate reduction (aprA) (Beck et al. observed with increasing salinity, but these were not accompa- 2017) (Fig. 6; Table 1). Metabolic biomarkers such as ladderane nied by decreases in the estimated abundances of the respon- phospholipids or specific 16S rRNA gene primers have also been sible microorganisms (Jiao et al. 2018). Once again, PCR primer used to quantify anammox bacteria in subterranean estuaries limitations, presence of inactive taxa or different activity rates (Sáenz et al. 2012; Jiao et al. 2018) as well as respiratory quinone between taxonomically different groups may be among the rea- composition for distinguishing certain metabolisms such as aer- sons explaining this lack of coupling between microbial iden- obic methanotrophy (Brankovits et al. 2017). tity and function. In addition, there may be functional redun- Despite this accumulating knowledge, our understanding dancy among groundwater taxa, which implies that a given of the links between the taxonomic composition of a com- process could be maintained even if the main microbial play- munity and its biochemical outcome is still limited, in part ers are replaced (Pilloni et al. 2019). Given that sequencing Ruiz-González et al. 11 Table 1. Major metabolic pathways and some of the relevant marker genes targeted in groundwater studies. Groundwater Metabolic pathway Gene Protein studies Aerobic respiration cyoA Cytochrome bo(3) ubiquinol oxidase subunit 2 a cydA Cytochrome bd-I ubiquinol oxidase subunit 1 a ccoN Cytochrome c oxidase a,b coxA Cytochrome c oxidase subunit I a Autotrophic carbon cbbL Ribulose bisphosphate carboxylase large chain b,c,f fixation cbbM Ribulose bisphosphate carboxylase c,f cdhA Acetyl-CoA decarbonylase/synthase complex subunit α b cdhB Acetyl-CoA decarbonylase/synthase complex subunit ε b Nitrification amoA Ammonium monooxygenase subunit A a,d,e,i,j,k nxrA Nitrite oxidoreductase α subunit a,i Anammox hzsA Hydrazine synthase subunit A b hzsB Hydrazine synthase subunit B b hzsC Hydrazine synthase subunit C b hzoA Hydrazine oxidoreductase A b hzoB Hydrazine oxidoreductase B b Denitrification nirK Copper-containing nitrite reductase a,b,h,i,k nirS Nitrite reductase a,b,h,i nosZ Nitrous-oxide reductase a,b,i,k napA Periplasmic nitrate reductase a,b,i napB Periplasmic nitrate reductase, electron transfer subunit a,b,i narG Respiratory nitrate reductase 1 α chain a,b,i narH Respiratory nitrate reductase 1 β chain a,b,i narI Respiratory nitrate reductase 1 γ chain b,i DNRA nrfA Nitrite reductase (formate dependent) i,k nirB Nitrite reductase (NADH-dependent) large subunit i nirD Nitrite reductase (NADH-dependent) small subunit i Nitrogen fixation nifD Nitrogenase molybdenum-iron protein α chain a,i nifH Nitrogenase iron protein a,i,k nifK Nitrogenase molybdenum-iron protein β chain a,i Iron oxidation mtoA Decaheme c-type cytochrome b mtoB Decaheme-associated outer membrane protein b Iron/manganese mtrB Decaheme-associated outer membrane protein a oxidation/reduction mtrC Decaheme c-type cytochrome a Sulfur aprA Dissimilatory adenosine-5 -phosphosulfate reductase a,b,e oxidation/reduction Sulfate reduction dsrA Sulfite reductase α subunit a,b,k,l dsrB Sulfite reductase β subunit a,b Methanotrophy (aerobic) pmoA Membrane-bound particulate methane monooxygenase g,m mmoX Soluble methane monooxygenase g,k Methanotrophy (anaero- mcrA Methyl-coenzyme M reductase α subunit k,l,m bic)/methanogenesis Letters indicate examples of groundwater studies where the functional marker genes shown in Fig. 6 have been identified (or predicted from taxonomy as in ’i’, ’k’ and ’l’), plus some genes indicative of aerobic respiration and autotrophic carbon fixation. Studies performed in subterranean aquifers are indicated in bold (see also Table S1, Supporting Information), but examples of inland aquifer investigations where other marker genes have been targeted are also included. Anammox (anaerobic ammonium oxidation), DNRA (dissimilatory nitrate reduction to ammonium), RuBisCo (ribulose-1,5-bisphosphate carboxylase/oxygenase). (a) Lavy et al. 2019; (b) Jewell et al. 2016; (c) Herrmann et al. 2015; (d) Santoro et al. 2008; (e) Beck et al.2017; (f) Kellerman et al. 2012; (g) Shao et al. 2019; (h) Santoro et al.2006; (i) Adyasari et al.2020; (j) Rogers and Casciotti 2010; (k) Hong et al.2019; (l) Sang et al.2018; (m) Vigneron et al. 2017. techniques have replaced, rather than complemented, the use Microbial control of SGD-derived chemical fluxes of traditional bulk activity assays such as estimates of active biomass, heterotrophic production or exoenzymatic activity Ultimately, microbial activity, together with other geochemical (Table S1, Supporting Information), the gain in taxonomic res- reactions occurring in the subterranean estuary (e.g. desorption, olution has been accompanied by a notable loss of informa- dissolution, precipitation), will determine the chemical compo- tion on the actual functioning of communities and its reg- sition of the groundwater flowing through coastal aquifers into ulation. As noted by Smith et al. (2018), merging traditional the ocean (Moore 1999; Slomp and Van Cappellen 2004; Seibert activity measurements with novel sequencing technologies may et al. 2020; Fig. 6). Microbial activity in coastal sediments was provide the needed multifaceted view of these underexplored shown to reduce significantly anthropogenic nitrate through communities. denitrification and DNRA, acting as a natural filter, whereas 12 FEMS Microbiology Reviews, 2021, Vol. 45, No. 5 organic matter mineralization resulted in large amounts of has explored how much of this biologically used carbon derives ammonia and dissolved organic nitrogen to the ocean (Montiel from SGD. et al. 2019). Other studies have reported that processes such as Bacteria are also relevant competitors for inorganic nutri- methanogenesis and methanotrophy are important sources and ents, accounting for ∼40% of total uptake of inorganic phos- sinks of methane in subterranean estuaries, largely determining phorus, nitrogen and iron (Kirchman 2000). Bacterial ability to SGD methane fluxes to the ocean (Schutte et al. 2016; Brankovits outcompete eukaryotic phytoplankton may vary depending on et al. 2018; Pain, Martin and Young 2019), and that the transport the type of inorganic nutrients and carbon added, and hence of particulate marine matter through the seepage zone can fuel the chemical composition of the SGD may determine the bio- microbial dissolved organic carbon production, resulting in a net logical fate of nutrients or carbon. For example, bacteria out- release of SGD-driven DOM (Jiang et al. 2020). The different sea- competed diatoms in mesocosms with additions of phosphate sonal patterns shown by denitrification, DNRA or anammox in a and nitrate (with or without glucose), but if silicate was added, sandy subterranean estuary further illustrate that the multiple diatoms competed successfully with bacteria for the uptake of microbial processes involved in the cycling of a given element mineral nutrients (Havskum et al. 2003). In the study of Garcés (e.g. nitrogen) may be differentially regulated (Wong et al. 2020). et al. (2011) only a moderate response in the total number of Consequently, only a deep understanding of these processes and bacteria upon groundwater addition was reported after 3 days their natural controls will enable an accurate understanding or of incubation, probably caused by a better performance of phy- prediction of the potential changes in SGD-derived fluxes upon toplankton upon the increased availability of silicate. However, variations in environmental conditions. marine bacteria are the fastest responders to nutrient inputs (e.g. Lekunberri et al. 2012), so SGD-driven effects on bacterial abundance or activity may be missed if not sampled at the ade- quate temporal resolution. Influence of SGD on marine microbial communities Bacterial consumption of DOM and inorganic nutrients depends on community composition because different bacterial Since the awareness of the global relevance of SGD in coastal bio- taxa exhibit distinct metabolic capabilities and substrate prefer- geochemical processes, multiple investigations have explored ences, and not all bacteria are equally active in carbon consump- the effects of SGD on marine biota, including macro- and tion, respiration or biomass production (Cottrell and Kirchman microorganisms (see Lecher and Mackey 2018 for a review). 2000; Alonso-Sáez and Gasol 2007; Ruiz-González et al. 2012; Sar- Microbial-related SGD studies have focused mostly on the mento and Gasol 2012). In turn, bacterial community composi- responses of phytoplankton communities because of their key tion and activity are strongly shaped by environmental gradients role as marine primary producers, and have shown that the and biological factors such as viral infection or bacterivory, with nutrient inputs or salinity changes associated with SGD often some bacterial groups being more susceptible to these mortal- result in variations in the abundances and composition of plank- ity factors than others (Boavida and Wetzel 1998; Bouvier and tonic communities, promoting the growth of certain groups del Giorgio 2007; Mojica and Brussaard 2014; Teira et al. 2019). and sometimes causing eutrophication or harmful algal blooms Responses of heterotrophic bacteria to SGD may hence be either (Lecher and Mackey 2018; Adolf et al. 2019; Taniguchi et al. 2019; direct (if responding to nutrient or carbon inputs or changes Chen et al. 2020c). Despite the relevance of marine bacteria for in conditions such as salinity) but also indirect if SGD has an the functioning of coastal biogeochemistry (Gasol et al. 2008), influence on other planktonic organisms such as phytoplankton, much less is known about marine bacterioplankton taxonomic viruses or bacterial predators. As an example, large but transient and functional responses to SGD, as the few available studies increases in the abundance and activity of opportunistic marine have focused mostly on exploring changes in bulk abundances bacterial groups such as Roseobacter were observed upon exper- or activity of bacteria (Table S3, Supporting Information; Fig. 4). imental groundwater addition, which peaked in the first 24 h Photosynthetic unicellular cyanobacteria such as Synechococcus and decreased quickly afterward likely due to intense predation and Prochlorococcus are responsible for ∼25% of the primary pro- (Maister 2018). duction globally (Kirchman 2012) and can be important primary Depending on the type of discharge, microbial responses to producers also in coastal waters (Agawin et al. 2003; Scanlan et al. SGD may differ. Continuous diffusive SGD (Fig. 1A) may cre- 2009). Experimental addition of groundwater to marine commu- ate stable gradients of groundwater influence, leading to the nities resulted in fast abundance increases of Synechococcus in establishment of a succession in microbial communities like the Mediterranean Sea (Garcés, Basterretxea and Tovar-Sánchez those observed along surface estuarine plumes (Bouvier and del 2011), but not in other coastal settings (California and Hawaii) Giorgio 2002; Troussellier et al. 2002). Point-sourced SGD inputs where other phytoplankton groups were favored (Chamberlain (Fig. 1B) or discharge peaks following precipitation events (Kim et al. 2014; Lecher et al. 2015). These findings agree with exper- et al. 2003; Mejı́as et al. 2012), conversely, might trigger localized iments reporting variable responses of coastal Synechococcus to or ephemeral pulses of intense growth of taxa that may dis- different nutrient additions (Lekunberri et al. 2012), and high- proportionately contribute to carbon cycling, e.g. by channeling light that microbial responses to SGD will depend largely on large amounts of carbon toward higher trophic levels. However, the chemical composition of groundwater and specific local and groundwater is hardly ever included as a factor shaping bacte- biotic conditions. rial communities in coastal waters, highlighting the lack of inter- On the other hand, heterotrophic bacterioplankton largely action between microbial oceanographers and terrestrial ecolo- determine the flow and fate of carbon in the ocean. Community- gists or hydrogeologists. level processes such as bacterial respiration or biomass produc- tion thus have implications at the ecosystem level (del Giorgio and Williams 2005; Gasol et al. 2008), and appear to be affected by SGD-driven transport of microbial diversity into the SGD (Carlson and Wiegner 2016). The recurrent observation that coastal ocean coastal bacterial respiration rates often exceeds primary pro- duction indicates that bacteria rely on allochthonous (i.e. land- Given the high microbial diversity hidden in coastal aquifers derived) dissolved organic carbon (Duarte, Agustı́ and Vaqué and the magnitude of groundwater discharge at local and global 2004; Duarte and Prairie 2005), but to our knowledge no study scales, SGD may also influence coastal microbial communities Ruiz-González et al. 13 through the transport of groundwater prokaryotes. In freshwa- including the adjacent seawater and disregarding the land– ter ecosystems, the terrestrial–aquatic connectivity of microbial ocean microbial connectivity (Table S1, Supporting Information), assemblages has emerged as an important driver of microbial has precluded establishing microbial singularities or common- community composition and biogeography at the landscape alities across coastal aquifers or determining the magnitude of scale (Langenheder and Lindström 2019). For example, lake microbial fluxes between the aquifers and the sea. and river bacterial assemblages are strongly impacted by the dispersal of microorganisms from the surrounding terrestrial landscape or upstream water bodies, some of which seem FUTURE DIRECTIONS AND RESEARCH able to grow and dominate the aquatic environment (Crump, AVENUES Amaral-Zettler and Kling 2012; Ruiz-González, Niño-Garcı́a and Deepening our knowledge of the microbial dimension at the del Giorgio 2015a; Niño-Garcı́a, Ruiz-González and del Giorgio land–ocean interface will be beneficial from multiple points 2016). This SGD-driven microbial connectivity has mostly of view, ranging from the discovery of novel species and been addressed with regard to the transport of terrestrial metabolisms, to achieving more accurate predictions of SGD pathogenic or fecal indicator bacteria (FIB, i.e. enterococci and fluxes and their consequences in marine biogeochemical cycles. Escherichia coli) through coastal aquifers or sediments, which Based on the needs of the field identified in the previous sec- can have health implications (Yau et al. 2014; Vollberg et al. tions, here we propose five future research avenues and strate- 2019). Although some studies concluded that SGD may not gies that may help researchers advance toward a better under- be a main source of FIB (Knee et al. 2008), others reported a standing of the microbial dimension of SGD (Fig. 7). clear link between FIB and the concentration of SGD tracers such as radium isotopes (Paytan, Boehm and Shellenbarger 2004), as well as the presence of FIB and human enteric viruses Improving the description of microbial diversity and sourced by groundwater even 10 km offshore (Futch, Griffin and function in coastal aquifers Lipp 2010). Interestingly, different FIB can be transported from different sources (e.g. E. coli from sand washing and enterococci Given the high degree of taxonomic novelty found within single from groundwater; Russell et al. 2013) and different types of aquifers (Anantharaman et al. 2016), the study of the gene pools sediments can differentially retain fecal indicator bacteria or within both the active and the dormant fractions of communi- viruses (de Sieyes et al. 2016). Hence, processes operating at the ties from the undersampled coastal aquifers ensures the dis- land–ocean transition zone may modify not only the chemical covery of unimaginable living or surviving strategies, enriching but also the microbial composition of SGD. current sequence catalogs (Fig. 7.1). PCR-free sequencing tech- SGD discharging to the ocean is presumably also loaded with niques like metagenomics, metatranscriptomics or single cell the diverse microbial communities naturally inhabiting the sub- genomics (Appendix 2) have uncovered an enormous variety of terranean estuary or coastal sediments (see the section ’Micro- novel taxa and metabolisms in inland aquifers that are not cap- bial abundance, diversity and environmental drivers along sub- tured by current primer-based approaches (Jewell et al. 2016; Lau terranean estuaries’). Several hydrogeological characteristics et al. 2016; Probst et al. 2017, 2018; Bell et al. 2018), so their appli- (e.g. groundwater velocity, type of aquifer and discharge, mech- cation to coastal aquifers will offer key insight into the func- anisms driving the flow, degree of interaction between ground- tional potential hidden in these systems. The ongoing accumu- water and seawater) may determine the identity of the aquifer lation of genomic information from inland aquifers will facili- microbes arriving to the ocean, as for example not all bacte- tate the application of these techniques to coastal aquifer micro- rial taxa are equally mobilized from intertidal sands by seawa- biota by providing reference genomes and sequences. Moreover, ter (Boehm, Yamahara and Sassoubre 2014). To our knowledge, metagenomic information may also help in the design of new only two studies have suggested a potential role of SGD as con- PCR taxonomic or functional primers (e.g. Cornejo-Castillo 2017), veyor of groundwater microorganisms from coastal aquifers into which could be optimized for groundwater prokaryotic diversity. the marine environment: Lee et al. (2017) showed that both SGD Applied to the ultrasmall fraction of groundwater communities flow and tidal fluctuation determined the periodic occurrence of (<0.2 μm), these approaches will shed light on the unknown role freshwater-related bacteria in coastal bacterioplankton assem- of the minute Ca. Patescibacteria or the DPANN Archaea in sub- blages, and Menning et al. (2018) reported that changes in aquifer terranean estuaries. discharge controlled the microbial abundance and taxonomic Techniques that allow linking metabolic processes and activ- richness in a spring-fed estuary. ity rates to specific microbial groups offer ways to improve our It is still uncertain whether any of the groundwater or sedi- understanding of the role of different prokaryotic groups in the ment prokaryotes transported via SGD can thrive in seawater, cycling of substrates in coastal aquifers. For example, SIP of or whether marine taxa can survive within sandy sediments. DNA can identify microorganisms involved in the use of dif- However, there is experimental evidence that it is possible to ferent compounds (labeled with stable isotopes) through the retrieve living marine bacteria from sources such as freshwater, selective recovery of heavy isotope-enriched DNA: The micro- air or lake sediments (Comte et al. 2014; Langenheder et al. 2016) bial taxa involved in autotrophic fixation of CO2 (Lazar et al. and that mixing of fresh- and saline water promotes the growth 2017) or in the degradation of plant-derived polysaccharides of initially rare taxa (Shen, Jürgens and Beier 2018; Rocca et al. (Taubert et al. 2019) were identified in inland aquifers applying 2020). Also, active freshwater bacterial taxa were detected at the SIP. Also, SIP coupled to Raman microspectroscopy (Raman-SIP) marine extreme of an estuarine plume (Troussellier et al. 2002). unveiled the groundwater bacterial populations activated by dif- These findings suggest that coastal terrestrial and freshwater ferent plant-derived compounds using heavy water (D2 O) as a landscape components could represent reservoirs of microbial tracer (Taubert et al. 2018) and demonstrated complete denitri- taxonomic and functional diversity for marine communities, fication by a groundwater microbial community adding 15 NO3 − with yet unknown implications for marine community struc- (Kumar et al. 2018). ture and functioning. The fact that most coastal groundwater The relative contribution of different microbial taxa to sub- microbial research has been focused on single aquifers, rarely strate uptake can also be elucidated at the single-cell level 14 FEMS Microbiology Reviews, 2021, Vol. 45, No. 5 Figure 7. Future research directions to gain insight into the microbial dimension of SGD. Future studies should (A) improve the characterization of the taxonomic and functional diversity (including the ultrasmall, sediment-attached and previously unknown taxa) within subterranean estuaries; (B) elucidate the main environmental and hydrologic drivers of prokaryotic communities and their functions, considering the high spatiotemporal heterogeneity of these systems and features such as dormancy or dispersal of allochthonous taxa that may obscure links between community taxonomic composition (indicated by the color of prokaryotes), community functioning and external factors; (C) explore the role of biotic interactions (e.g. viruses, predators or symbiotic associations) and their implications for microbial community functioning; (D) incorporate knowledge on microbial processes and their drivers to constrain chemical SGD fluxes and their potential variations upon changes in conditions; and (E) gain insight into the responses of microbial communities to SGD, particularly bacterioplankton, focusing on the responses of specific microbial taxa, cascading effects and the dispersal of microorganisms from groundwater to the ocean. DOM, dissolved organic matter; DIC, dissolved inorganic carbon. through techniques such as microautoradiography (Brock 1967) in freshwater and marine systems (Pérez and Sommaruga 2006; or BONCAT (bioorthogonal non-canonical amino acid tagging) Ruiz-González et al. 2012), but only rarely in groundwater (Wil- (Hatzenpichler et al. 2014), which combined fluorescence in situ hartitz et al. 2009; Kellermann et al. 2012). More recently devel- hybridization techniques (FISH or CARD-FISH; Appendix 2) allow oped, BONCAT is based on the fluorescent labeling of the the visual identification of different active prokaryotic taxa. newly synthesized proteins. Combined with CARD-FISH or flow Microautoradiography has widely been used to quantify the cytometry (Sebastián and Gasol 2019), it allows visually identi- uptake of different radiolabeled compounds by bacterial groups fying active cells within specific taxonomic groups, making it a Ruiz-González et al. 15 promising tool for understanding the metabolic status of micro- analyses require much more initial biomass than primer-based bial inhabitants in coastal aquifers (Maister, 2018). sequencing, which may be logistically challenging due to the low The visualization of microorganisms and their activity is microbial abundances typically observed in groundwater (e.g. actually becoming increasingly needed in an era dominated 400–800 L was filtered for reconstructing genomes of ultrasmall by sequencing technologies, as it provides essential informa- cells by He et al. 2021). tion on microbial processes or features that cannot be derived Being able to distinguish between the truly active bacteria from sequencing data (Sebastián and Gasol 2019). For example, from those dormant, dead or allochthonous will also improve size-fractionated filtration for DNA sequencing does not allow our understanding of the role of different drivers on coastal unequivocally determining whether a given bacterial taxon is groundwater microbial communities. Experimental manipula- actually small, as filters may clog and retain small cells. Indeed, tions to promote or detect increases in the abundances of rare or we have found cells identified as ultrasmall Ca. Parcubacteria dormant taxa (Rajala and Bomberg 2017; Hofmann and Griebler (see the section ’Ultrasmall prokaryotic groups are abundant in 2018; Fig. 5), sequencing the RNA instead of the DNA to target the groundwater ecosystems’) onto the 0.2-μm filter in a coastal potentially active community (Aanderud et al. 2016; Wisnoski alluvial aquifer (Fig. 8A and B), which, to our knowledge, rep- et al. 2020), quantification of relic DNA (Carini et al. 2016; Lennon resents the first observation of ultrasmall cells in coastal saline et al. 2018), or approaches that allow linking taxonomy with groundwater. The comparison with the large Gammaproteobac- actual microbial activity (see the section ’Improving the descrip- teria detected in another subterranean estuary (Fig. 8C and D) tion of microbial diversity and function in coastal aquifers’) offer highlights the relevance of visualization for understanding the ways to learn about the metabolic status and role of different contribution of the different groundwater microbial groups to microbial taxa at the land–ocean transition zone. total biomass, for learning about their lifestyles (free-living, Flow cytometry may also be used to readily characterize attached), or for deciphering their contribution to community the physiological status of cells along the land–ocean transition activity (Fig. 8E; Sebastián and Gasol 2019). zones, such as through the nucleic acid double-staining proto- Finally, the sampling of prokaryotes attached to surfaces col (Gre´gori et al. 2001) that allows distinguishing dead or dam- should also be implemented in parallel to water sampling to aged cells from healthy cells in planktonic samples, or the detec- achieve the full picture of the groundwater microbiota (Smith tion of actively respiring cells stained with 5-cyano-2,3-ditolyl et al. 2018). Whereas characterizing the taxonomic composition tetrazolium chloride (Sieracki, Cucci and Nicinski 1999). Differ- of microbial communities in unconsolidated sediments is rel- entiating high versus low nucleic acid content cells (HNA vs atively easy if the sediment can be accessed, estimating the LNA) may help estimate variations in ultrasmall planktonic cells microbial biomass and activity associated with these surfaces along environmental gradients (Proctor et al. 2018). remains challenging (Smith et al. 2018). Recent protocols for the Considering the bidirectional dispersal of groundwater and direct quantification of prokaryotic abundance in groundwater marine species is also critical, as it may impact aquifer func- sandy sediments (Bayer et al. 2016) and the combination of tra- tioning inoculating viable taxa in both directions, or mask rel- ditional activity assays with new sequencing techniques per- evant patterns by delivering significant amounts of inactive formed at scales relevant for microbial processes (Smith et al. allochthonous taxa. Microbial source tracking methods, which 2018) open the door for the simultaneous assessment of the two allow assigning the most plausible habitat to a given taxon in microbial compartments. order to infer potential dispersal sources (Comte et al. 2017; Ortiz-Álvarez et al. 2019) could allow defining connectivity path- Expanding knowledge of chemical and hydrologic ways within aquifers and tracking the abundances of poten- drivers of microbial communities in subterranean tially inactive allochthonous taxa (e.g. Fillinger et al. 2021). The quantification of endospores as done in deep marine sediments estuaries (Wörmer et al. 2019) or of genes involved in dormancy (Lennon Most current information on groundwater microbial diversity and Jones 2011) may further help assess the extent of dormancy (in particular that from coastal aquifers) comes from snapshot and the factors governing activity in these dynamic environ- observations in specific locations (Table S1, Supporting Informa- mental transitions. Importantly, for any of these microbiologi- tion), so we lack a comprehensive view of the large-scale spatial cal investigations to yield accurate insight, they must be accom- patterns and the temporal dynamics of the microbial commu- panied by detailed hydrogeological assessments of the relevant nities in subterranean estuaries. Comparative studies covering characteristics of the system, including the physical forces driv- large environmental and spatial gradients are needed to gain ing the flow, the dynamic movement of the fresh–saline inter- insight into the most important drivers of groundwater micro- face, or the groundwater residence time in the subterranean bial communities and their functions (Fillinger, Hug and Griebler estuary. 2019a) but also studies exploring the highly dynamic spatiotem- poral heterogeneity within single coastal aquifers (McAllister Elucidating the role of biotic interactions as drivers of et al. 2015; Menning et al. 2018; Jiang et al. 2020; Fig. 7.2). This microbial communities in coastal aquifers requires overcoming the technical limitations described previ- ously, such as optimizing primer diversity coverage and reach- Besides environmental or hydrologic factors, biotic interactions ing a consensus on which primers to use to be able to compare are emerging as key drivers of food webs and biogeochemical among groundwater studies (see the sections ’Novel groundwa- cycles in ecosystems such as the ocean (Lima-Méndez et al. ter microbial taxa can be missed by current sequencing tech- 2015; Wilkins et al. 2019), but little is known about the interac- nologies’ and ’Improving the description of microbial diversity tions between indigenous microorganisms in the groundwater and function in coastal aquifers’). The lower cost and higher environment (Fig. 7.3). Some studies in inland aquifers have detection of rare prokaryotic taxa by primer-based approaches suggested that planktonic prokaryotes are more strongly than by metagenomics/metatranscriptomics make them use- controlled by viral lysis than by bacterivores (Wilhartitz et al. ful tools to gain insight into the mechanisms of assembly and 2013), and that viruses can influence carbon biogeochem- drivers of groundwater communities across large numbers of istry and prokaryotic community structure (Pan et al. 2014). samples. In addition, metagenomic and metatranscriptomic Predation by flagellates was shown to be intense and size 16 FEMS Microbiology Reviews, 2021, Vol. 45, No. 5 Figure 8. Visualization is essential to gain insight into the ecology of different groundwater microbial groups. Examples of microscopy pictures showing the pres- ence of ultrasmall Ca. Parcubacteria (A, B) or large Gammaproteobacteria (C–E) in two Mediterranean coastal aquifers in Spain [a limestone-karstified coastal aquifer (Garcia-Solsona et al. 2010) and an alluvial aquifer (Folch et al. 2020), respectively]. Micrographic images show: (A, C) all microbial cells stained with 4 ,6-diamidino- 2-phenylindole (DAPI); specific bacterial groups hybridized with the CARD-FISH probes OD1-289 (Gong et al. 2014) for Parcubacteria clades (B), and Gamma42a (Manz et al. 1992) for Gammaproteobacteria (D). (E) Translationally active cells (detected with BONCAT, see the section ’Improving the description of microbial diversity and function in coastal aquifers’). Scale bars are 20 μ m. Note the large difference in size between both groups, with obvious implications for their contributions to total biomass and activity, information that cannot be derived from sequencing approaches. Gammaproteobacteria (D) seem to be mostly concentrated on particles, as well as the majority of active cells (E). Ruiz-González et al. 17 selective in a contaminated aquifer (Kinner et al. 1998) and including the chemical cycling controlled by microbial commu- predatory bacteria (i.e. bacteria that feed on other bacteria) nities. Knowledge of microbial ecology in the subterranean estu- were isolated from a coastal alluvial aquifer (Banning, Casciotti ary thus needs to be coupled with estimates of biogeochemi- and Kujawinski 2010), yet the relevance of this mechanism cal and hydrogeological processes to unravel the role of subter- in shaping prokaryotic communities is largely unknown. ranean estuaries in controlling solute fluxes to the ocean; this is Interestingly, experimental exposure to protozoan grazers actually one of the major knowledge gaps in the SGD field high- caused changes in sediment-associated bacterial communities lighted by recent reviews (Robinson et al. 2018; Taniguchi et al. and increased the retention of 13 C-labeled organic carbon in 2019). groundwater sediments (Longnecker and Kujawinski 2013). This Most knowledge about element cycling in subterranean estu- suggests that these biotic interactions may impact not only the aries has been achieved by piecing together physicochemical microbial structure of communities but also biogeochemical snapshots obtained from profiles along cross-shore transects, processes in both the planktonic and the sediment-attached and by comparing nutrient concentrations in the fresh and compartments. marine endmembers (e.g. Gonneea et al. 2014; Reckhardt et al. Recently, some multi-approach studies have started to shed 2017). While these evaluations provide key information on the light into the complexity of trophic interactions within sub- behavior of chemicals in the subterranean estuary, they do not surface ecosystems. Combining isotopic composition of food always allow identifying the controlling mechanisms or predict- resources and consumers with information on food web struc- ing the potential consequences due to changes in the environ- ture, predator species diversity and organic matter, Hutchins ment (because different processes may yield similar chemical et al. (2016) unveiled that chemolithoautotrophic organic mat- snapshots). For example, coastal aquifers are especially sen- ter production supported complex macroinvertebrate food webs sitive to salinization (i.e. seawater intrusion) due to both sea and their species diversity in a groundwater system. Simi- level rise and excessive withdrawal of water for human pur- larly, a combination of geochemical, genomic and biomarker poses (Ferguson and Gleeson 2012; Werner et al. 2013; Retter, approaches revealed that carbon from methanotrophic bacteria Karwautz and Griebler 2021) and hence the salinity tolerances supported a large fraction of the diet of cave-adapted shrimps of different microbial taxa will control how a given process in a tropical subterranean estuary (Brankovits et al. 2017), high- respond to the expected salinity changes. Actually, seawater lighting the relevance of chemolithoautotrophic metabolisms intrusion has been shown to impact the distribution and activ- for sustaining life in these coastal systems. ity of microbial communities in coastal peat deposits, result- Microscopy may also help elucidate biological interactions ing in increases in the release of greenhouse gases (CH4 , CO2 ) within coastal groundwater microbial assemblages. Cryotrans- driven by SGD (Kreuzburg et al. 2020). Salinity increases might mission electron microscopy (cryo-TEM) was essential to deter- also enhance the delivery of NH4 + from groundwater to the mine that ultrasmall cells in an inland aquifer were metaboli- coast due to increased N mineralization and decreased nitri- cally active rather than starved and dormant, as they appeared fication (Santoro 2010). Investigations in anchialine systems infected by viruses or physically interacting with other bacterial (i.e. coastal sinkholes with a subterranean connection with the cells, further indicating a potential symbiotic lifestyle (Luef et al. ocean; see Fig. 1B) might shed light on some of the processes 2015; He et al. 2021). Also, using genomic information to design and the drivers operating in coastal underground ecosystems; FISH probes, Schwank et al. (2019) uncovered a host–symbiont their higher accessibility may allow a coverage of the spatial het- association between two uncultivated archaeal phyla in a deep erogeneity in environmental gradients and/or microbial dynam- aquifer. Due to the small genome of many of the novel phyla ics that is not feasible in most subterranean estuaries (e.g. Sey- discovered in groundwater, it has been hypothesized that many mour et al. 2007; González et al. 2011) or the sampling and func- groundwater groups, including members within Ca. Patescibac- tional exploration of surface-attached microbial communities teria, may have a symbiotic lifestyle (Nelson and Stegen 2015; (Haas et al. 2018). Youssef et al. 2015, but see Beam et al. 2020). This, coupled to In turn, prospection of subsurface microbial genes and the genomic evidence that groundwater microbial community metabolisms might help counteract our impact on groundwa- members are intimately linked through the exchange of chem- ter by means of microbially mediated management practices, ical substrates (Anantharaman et al. 2016; Lau et al. 2016), and like for example the degradation of pollutants (Posman, DeR- that most groundwater biomass is in the form of biofilms where ito and Madsen 2017) or those enhancing nitrogen removal cells are in close contact to one another (Smith et al. 2018), from aquifers (Ludington et al. 2017). Denitrification has long suggests that symbiotic relationships (either through obligate been considered as the main process removing nitrogen from or facultative syntrophy, mutualism or commensalism) might coastal aquifers, but other less studied metabolisms performed be essential in sustaining the microbiome and biogeochemical by largely unknown microbial players, e.g. DNRA and anammox, cycling in groundwater ecosystems. The study of biotic inter- seem to be important pathways of nitrogen loss in aquifers glob- actions within subterranean estuaries is, to our knowledge, a ally (Wang et al. 2020). In any case, it will not be possible to accu- largely unexplored but promising field of research. rately interpret microbial data without a good knowledge of the system (geological setting, geochemistry, hydrology). Attempts to join different approaches such as nutrient mass-balances, Incorporating microbial processes to constrain and stable isotopic signatures of water and sediments, information on DOM composition, redox conditions, water residence time better estimate SGD fluxes and microbial communities have provided useful insight into Integrating the microbial perspective into SGD studies should the different sources of chemical elements, the complexity of improve our capacity to constrain the magnitude of SGD-driven subsurface biogeochemical transformations and resulting SGD- solute fluxes and their ecological implications (Fig. 7.4). Accu- driven solute fluxes (Beck et al. 2017; Montiel et al. 2019). There- rate estimates of solute fluxes supplied by SGD (at local, regional fore, only through multidisciplinary efforts we may achieve a and global scales) require understanding the chemical trans- holistic view of the biotic interactions and biogeochemical pro- formations occurring to groundwater before it reaches the sea, cesses within coastal aquifers. 18 FEMS Microbiology Reviews, 2021, Vol. 45, No. 5 Deepening our understanding of SGD effects on marine might represent a diversity pathway not previously considered, food webs and biogeochemical cycles with potential implications for marine community structure and functioning. Finally, whether viruses, bacteria or protists deliv- Studies exploring the effects of SGD on the microbial compo- ered with coastal groundwater can directly interact with marine nents at the base of food webs (including bacterioplankton) will bacterioplankton communities remains completely unknown. provide useful insight into the ecological and biogeochemical Groundwater taxa arriving to the sea may be used as micro- consequences of SGD in the marine environment, improving bial ’indicators’ or tracers of groundwater inputs to the coastal our understanding of the microbial regulation of nutrient, car- zone. Recent studies have used microbial information to iden- bon and metal cycles in the coastal ocean (Fig. 7.5). As the ele- tify tidal or temporal variations of SGD (Lee et al. 2017), to distin- ment fluxes differ largely depending on the characteristics of the guish groundwater recharge sources to shallow wells (Higgings aquifer (e.g. natural and anthropogenic inputs into groundwater, et al. 2020), to uncover sporadic episodes of intense ground- geologic characteristics and type of discharge, water flow veloc- water influence (Kim et al. 2020) or to detect contamination ity; Tovar-Sánchez et al. 2014), an accurate understanding of the episodes, given that specific taxa can be used as biosensors microbial responses to SGD necessarily requires detailed evalu- even after a given pollutant has been fully degraded (Smith et al. ations of the magnitude of SGD-driven chemical fluxes. 2015). Similarly, the presence of marine microbial taxa in coastal Coupling microbial dynamics with estimations of SGD- aquifers can indicate active land–ocean connectivity, and bacte- driven solute fluxes derived from chemical tracers (e.g. radium rial taxa within Oceanospirillales and Alteromonadaceae (Chen isotopes, radon, methane, silica; Dulaiova et al. 2010; Rodellas et al. 2019a) or within Rhodobacteraceae and Flavobacteriaceae et al. 2017; Oehler et al. 2019) or the application of mixing models (Unno et al. 2015) have been proposed as potential indicators along salinity gradients (as in Stegen et al. 2016) may be useful of seawater intrusions in coastal aquifers. The identification of to identify and quantify localized microbial responses to SGD. pathogens or fecal bacteria may further allow the detection of Experimentation could allow differentiating between direct SGD wastewater contamination in coastal aquifers or in the adjacent effects on microbial communities (through the acquisition of seawater (Boehm, Shellenbarger and Paytan 2004; Yau et al. 2014; SGD solutes or changes in the environment), or indirect cas- Ahmed et al. 2019). As many of these ’indicators’ may not sur- cade effects such as bacterial responses to SGD-driven changes vive in marine systems, their detection will refine our capacity in phytoplankton production of DOM or grazer or viral activ- to distinguish between allochthonous and truly marine taxa. ity. Isotopic fractionation may provide valuable insight on the transfer of nutrients or carbon from groundwater into differ- ent marine microbial compartments (e.g. primary producers; CONCLUDING REMARKS Andrisoa et al. 2019), allowing to estimate the fate of the deliv- ered elements in coastal ecosystems. The degree of allochthony The submarine discharge of groundwater to the coastal ocean in aquatic bacteria (i.e. the contribution of terrestrial dissolved constitutes a terrestrial and marine linkage of high relevance for organic carbon to aquatic bacterial biomass) has been exten- the functioning of marine ecosystems, which is modulated to a sively studied in freshwater systems (e.g. Guillemette, McCal- large extent by the microbial processes occurring in the coastal lister and del Giorgio 2016), yet to our knowledge this has never aquifer. Despite the growing interest on groundwater microbiol- been explored in coastal bacterial communities influenced by ogy in recent years, coastal aquifers have received little atten- SGD. Finally, microbial transformations controlling dissolved tion from microbial studies in comparison to surface waters. In organic matter (DOM) quality within coastal aquifers (Chaillou, turn, SGD research has been historically dominated by studies Lemay-Borduas and Couturier 2016; Couturier, Nozais and Chail- performed from either a ’chemical oceanography’ or a ’hydro- lou 2016; Jiang et al. 2020) may in turn determine the taxonomic logical’ point of view, lacking a biological understanding of the or functional responses of planktonic marine bacterioplankton underlying processes. influenced by SGD, given that DOM quality is a main driver of Here, we have identified some of the current challenges and aquatic bacterioplankton structure and functioning (Landa et al. future directions that should be tackled for a deeper under- 2014; Ruiz-González et al. 2015b). Applying the above-mentioned standing of the microbial dimension underlying SGD processes. techniques to link taxonomy with function (see the section These include improving the characterization of groundwater ’Improving the description of microbial diversity and function in microbiota by overcoming current technical limitations and coastal aquifers’) along natural groundwater gradients or upon considering the specific characteristics of coastal aquifers (i.e. experimental groundwater additions will help identify specific dynamic mixing of water, chemicals and communities). In turn, responses of different marine bacterial groups to SGD. the evaluation of solute fluxes (e.g. nutrients, metals, carbon, On the other hand, the influence of the dispersal of ground- contaminants) supplied by SGD needs to rely on an accurate water microbial communities on coastal ecosystems could be understanding of the microbially mediated chemical cycling in resolved by means of experimental assessments of the capac- the coastal aquifer, requiring an improved knowledge on their ity of groundwater microorganisms to thrive in marine condi- drivers and metabolic capacities and responses. Finally, coastal tions (e.g. mixing water types; Maister 2018) or through trans- marine microbial ecology needs to incorporate the connectivity plant experiments (Shen, Jürgens and Beier 2018), or by track- with the continent, and in particular this largely overlooked ing the presence of groundwater taxa in marine sites. Microbial groundwater linkage; good knowledge of the hydrogeology and source tracking methods (see the section ’Expanding knowledge the chemical and microbial fluxes from the coast to the ocean of chemical and hydrologic drivers of microbial communities in will expand our understanding of the structuring and function subterranean estuaries’) have been applied to identify the origin of coastal planktonic communities, as well as the ecological of fecal bacteria in the surf zone (Russell et al. 2013) and could be consequences of SGD. Only by moving beyond the presumed used to detect groundwater taxa in coastal areas (or vice versa) limits of ecosystems and research fields (Xenopoulos et al. 2017; if simultaneously characterizing coastal aquifers and marine Kayler et al. 2019), we may obtain a comprehensive understand- sites (Unno et al. 2015; Adyasari et al. 2019). Specifically, the ing of the interlinked physical and biogeochemical processes SGD-driven transport of taxa able to thrive in coastal seawater occurring throughout the terrestrial–marine continuum. Ruiz-González et al. 19 ACKNOWLEDGMENTS Alfreider A, Schimer M, Vogt C. Diversity and expression of different forms of RubisCO genes in polluted groundwa- Particular thanks to Marta Sebastián and Andrea G. Bravo for ter under different redox conditions. FEMS Microbiol Ecol their useful and constructive comments on the manuscript, 2012;79:649–60. and to Josep M. Gasol, Olena Maister, Marc Diego-Feliu and Alonso-Sáez L, Gasol JM. Seasonal variations in the contri- Aaron Alorda-Kleinglass for their help during some of the first butions of different bacterial groups to the uptake of exploratory field surveys. We acknowledge support of the publi- low-molecular-weight compounds in northwestern Mediter- cation fee by the CSIC Open Access Publication Support Initiative ranean coastal waters. Appl Environ Microbiol 2007;73: through its Unit of Information Resources for Research (URICI). 3528–35. Anantharaman K, Brown CT, Hug LA et al. Thousands of micro- bial genomes shed light on interconnected biogeochemi- SUPPLEMENTARY DATA cal processes in an aquifer system. Nat Commun 2016;7: Supplementary data are available at FEMSRE online. 13219. Andrisoa A, Stieglitz TC, Rodellas V et al. Primary production in coastal lagoons supported by groundwater discharge and FUNDING porewater fluxes inferred from nitrogen and carbon isotope signatures. Mar Chem 2019;210:48–60. This work was funded by the Spanish Ministry of Science, Banning EC, Casciotti KL, Kujawinski EB. Novel strains isolated Innovation and Universities (MICINN) through the GRAMMI from a coastal aquifer suggest a predatory role for flavobac- project (RTI2018-099740-J-I00) and partly by the projects MEDIS- teria. FEMS Microbiol Ecol 2010;73:254–70. TRAES II (CGL2016-77122-C2) and OPAL (PID2019-110311RB-C21). Barragán JM, de Andrés M. Analysis and trends of the Additional financial support was provided by grants SPIP2020- world’s coastal cities and agglomerations. Ocean Coast Manag 02595 (Ministry for the Ecological Transition and the Demo- 2015;114:11–20. graphic Challenge (MITECO)) and 2017SGR/156 (Generalitat de Bayer A, Drexel R, Weber N et al. Quantification of aquatic Catalunya), with funding from the Spanish government through sediment prokaryotes: a multiple-steps optimization testing the ‘Severo Ochoa Centre of Excellence’ accreditation (CEX2019- sands from pristine and contaminated aquifers. Limnologica 000928-S).VR acknowledges financial support from the Beatriu 2016;56:6–13. de Pinós postdoctoral programme of the Catalan Government Beam JP, Becraft ED, Brown JM et al. Ancestral absence of electron (2017-BP-00334). This study contributes to the work carried out transport chains in Patescibacteria and DPANN. Front Micro- by the MERS research group 2017 SGR 1588. biol 2020;11:1848. Beck M, Reckhardt A, Amelsberg J et al. The drivers of biogeo- Competing Interests. The authors declare no competing financial chemistry in beach ecosystems: a cross-shore transect from interests. the dunes to the low-water line. Mar Chem 2017;190:35–50. Bell E, Lamminmäki T, Alneberg J et al. Biogeochemical cycling by a low-diversity microbial community in deep groundwater. REFERENCES Front Microbiol 2018;9:2129. Aanderud ZT, Jones SE, Fierer N et al. Resuscitation of the rare Boavida G, Wetzel RG. Inhibition of phosphatase activity by dis- biosphere contributes to pulses of ecosystem activity. Front solved humic substances and hydrolytic reactivation by nat- Microbiol 2015;6:24. ural UV. Freshw Biol 1998;40:285–93. Aanderud ZT, Vert JC, Lennon JT et al. Bacterial dormancy is more Boehm AB, Shellenbarger GG, Paytan A. Groundwater discharge: prevalent in freshwater than hypersaline lakes. Front Micro- potential association with fecal indicator bacteria in the surf biol 2016;7:853. zone. Environ Sci Technol 2004;38:3558–66. Acinas SG, Sánchez P, Salazar G et al. Metabolic architecture of Boehm AB, Yamahara KM, Sassoubre LM. Diversity and transport the deep ocean microbiome. bioRxiv 2019;635680. of microorganisms in intertidal sands of the California Coast. Adolf JE, Burns J, Walker JK et al. Near shore distributions of phy- Appl Environ Microbiol 2014;80:3943–51. toplankton and bacteria in relation to submarine groundwa- Bouvier T, del Giorgio PA. Compositional changes in free-living ter discharge-fed fishponds, Kona coast, Hawai’i, USA. Estuar bacterial communities along a salinity gradient in two tem- Coast Shelf Sci 2019;219:341–53. perate estuaries. Limnol Oceanogr 2002;47:453–70. Adyasari D, Hassenrück C, Montiel D et al. Microbial com- Bouvier T, del Giorgio PA. Key role of selective viral-induced mor- munity composition across a coastal hydrological system tality in determining marine bacterial community composi- affected by submarine groundwater discharge (SGD). PLoS tion. Environ Microbiol 2007;9:287–97. One 2020;15:e0235235. Bradley JA, Amend JP, LaRowe DE. Survival of the fewest: Adyasari D, Hassenrück C, Oehler T et al. Microbial commu- microbial dormancy and maintenance in marine sediments nity structure associated with submarine groundwater dis- through deep time. Geobiology 2019;17:43–59. charge in northern Java (Indonesia). Sci Total Environ 2019;689: Brankovits D, Pohlman JW, Ganju NK et al. Hydrologic controls 590–601. of methane dynamics in karst subterranean estuaries. Glob Agawin NSR, Duarte CM, Agusti S et al. Abundance, biomass Biogeochem Cycles 2018;32:1759–75. and growth rates of Synechococcus sp. in a tropical coastal Brankovits D, Pohlman JW, Niemann H et al. Methane- and ecosystem (Philippines, South China Sea). Estuar Coast Shelf dissolved organic carbon-fueled microbial loop supports Sci 2003;56:493–502. a tropical subterranean estuary ecosystem. Nat Commun Ahmed W, Hamilton K, Toze S et al. A review on microbial 2017;8:1835. contaminants in stormwater runoff and outfalls: poten- Brewer TE, Aronson EL, Arogyaswamy K et al. Ecological and tial health risks and mitigation strategies. Sci Total Environ genomic attributes of novel bacterial taxa that thrive in sub- 2019;692:1304–21. surface soil horizons. mBio 2019;10:e01318–9. 20 FEMS Microbiology Reviews, 2021, Vol. 45, No. 5 Brock TD. Bacterial growth rate in the sea—analysis by thymi- cluster consuming low- and high-molecular-weight dis- dine autoradiography. Science 1967;155:81–3. solved organic matter. Appl Environ Microbiol 2000;66: Brown CT, Hug LA, Sharon TI et al. Unusual biology across a 1692–7. group comprising more than 15% of domain Bacteria. Nature Couturier M, Nozais C, Chaillou G. Microtidal subterranean estu- 2015;523:208–11. aries as a source of fresh terrestrial dissolved organic matter Burnett WC, Aggarwal PK, Aureli A et al. Quantifying subma- to the coastal ocean. Mar Chem 2016;186:46–57. rine groundwater discharge in the coastal zone via multiple Crump BC, Amaral-Zettler LA, Kling GW. Microbial diversity in methods. Sci Total Environ 2006;367:498–543. arctic freshwaters is structured by inoculation of microbes Burnett WC, Bokuniewicz M, Huettel WS et al. Groundwater from soils. ISME J 2012;6:1629–39. and pore water inputs to the coastal zone. Biogeochemistry Davis M, Garey J. Microbial function and hydrochemistry within 2003;66:3–33. a stratified anchialine sinkhole: a window into coastal Canfield DE, Thamdrup B. Towards a consistent classification aquifer interactions. Water 2018;10:972. scheme for geochemical environments, or, why we wish the del Giorgio PA, Williams PJ. The global significance of respi- term ‘suboxic’ would go away. Geobiology 2009;7:385–92. ration in aquatic ecosystems: from single cells to the bio- Carini P, Marsden PJ, Leff JW et al. Relic DNA is abundant in soil sphere. In: del Giorgio PA, Williams PJ (eds). Respiration in and obscures estimates of soil microbial diversity. Nat Micro- Aquatic Ecosystems. New York: Oxford University Press, 2005, biol 2016;2:16242. 267–303. Carlson K, Wiegner T. Effects of submarine groundwater dis- de Sieyes NR, Russell TL, Brown KI et al. Transport of enterococci charge on bacterial growth efficiency in coastal Hawaiian and F+ coliphage through the saturated zone of the beach waters. Aquat Microb Ecol 2016;77:167–81. aquifer. J Water Health 2016;14:26–38. Castelle CJ, Brown CT, Anantharaman K et al. Biosynthetic Duarte CM, Agustı́ S, Vaqué D. Controls on planktonic capacity, metabolic variety and unusual biology in the CPR metabolism in the Bay of Blanes, northwestern Mediter- and DPANN radiations. Nat Rev Microbiol 2018;16:629–45. ranean littoral. Limnol Oceanogr 2004;49:2162–70. Chaillou G, Lemay-Borduas F, Couturier M. Transport and Duarte CM, Prairie YT. Prevalence of heterotrophy and atmo- transformations of groundwater-borne carbon discharging spheric CO2 emissions from aquatic ecosystems. Ecosystems through a sandy beach to a coastal ocean. Can Water Resour J 2005;8:862–70. 2016;41:455–68. Dulaiova H, Camilli R, Henderson PB et al. Coupled radon, Chamberlain SD, Kaplan KA, Modanu M et al. Biogeography of methane and nitrate sensors for large-scale assessment of planktonic and benthic cyanobacteria in coastal waters of groundwater discharge and non-point source pollution to the Big Island, Hawai’i. FEMS Microbiol Ecol 2014;89:80–8. coastal waters. J Environ Radioact 2010;101:553–63. Chen L, Hu BX, Dai H et al. Characterizing microbial diversity and Duque C, Michael HA, Wilson AM. The subterranean estuary: community composition of groundwater in a salt–freshwater technical term, simple analogy, or source of confusion? Water transition zone. Sci Total Environ 2019a;678:574–84. Resour Res 2020;56:e2019WR026554. Chen L, Zhang J, Dai H et al. Comparison of the ground- Ferguson G, Gleeson T. Vulnerability of coastal aquifers to water microbial community in a salt-freshwater mixing groundwater use and climate change. Nat Clim Chang zone during the dry and wet seasons. J Environ Manage 2012;2:342–5. 2020a;271:110969. Fillinger L, Hug K, Griebler C. Aquifer recharge viewed through Chen X, Cukrov N, Santos IR et al. Karstic submarine ground- the lens of microbial community ecology: initial disturbance water discharge into the Mediterranean: radon-based nutri- response, and impacts of species sorting versus mass effects ent fluxes in an anchialine cave and a basin-wide upscaling. on microbial community assembly in groundwater during Geochim Cosmochim Acta 2020c;268:467–84. riverbank filtration. Water Res 2021;189:116631. Chen X, Ye Q, Du J et al. Bacterial and archaeal assemblages from Fillinger L, Hug K, Griebler C. Selection imposed by local two size fractions in submarine groundwater near an indus- environmental conditions drives differences in micro- trial zone. Water 2019b;11:1261. bial community composition across geographically dis- Chen X, Ye Q, Sanders CJ et al. Bacterial-derived nutrient and tinct groundwater aquifers. FEMS Microbiol Ecol 2019a;95: carbon source-sink behaviors in a sandy beach subterranean fiz160. estuary. Mar Pollut Bull 2020b;160:111570. Fillinger L, Zhou Y, Kellerman C et al. Non-random processes Cho H-M, Kwon KG, Moosdorf EY et al. Radium tracing nutri- determine the colonization of groundwater sediments by ent inputs through submarine groundwater discharge in the microbial communities in a pristine porous aquifer. Environ global ocean. Sci Rep 2018;8:2439. Microbiol 2019b;21:327–42. Comte J, Berga M, Severin I et al. Contribution of different bac- Flynn TM, Sanford RA, Ryu H et al. Functional microbial diversity terial dispersal sources to lakes: population and commu- explains groundwater chemistry in a pristine aquifer. BMC nity effects in different seasons. Environ Microbiol 2017;19: Microbiol 2013;13:146. 2391–404. Folch A, del Val L, Luquot L et al. Combining Fiber Optic (FO-DTS), Comte J, Lindström ES, Eiler A et al. Can marine bacteria cross-hole ERT and time-lapse formation electrical conduc- be recruited from freshwater sources and the air? ISME J tivity to characterize and monitor a coastal aquifer. J Hydrol 2014;8:2423–30. 2020;588:125050. Cornejo-Castillo FM (2017). Diversity, Ecology and Evolution of Freeze RA, Cherry JA. Groundwater. Englewood Cliffs, N.J.: Marine Diazotrophic Microorganisms. Universidad Politécnica Prentice-Hall, 1979. de Cataluña. Ph.D. Thesis, Dec 2017 ( https://digital.csic.es/ Futch JC, Griffin DW, Lipp EK. Human enteric viruses in ground- handle/10261/164042). water indicate offshore transport of human sewage to coral Cottrell MT, Kirchman DL. Natural assemblages of marine reefs of the Upper Florida Keys. Environ Microbiol 2010;12: proteobacteria and members of the Cytophaga–Flavobacter 964–74. Ruiz-González et al. 21 Garcia-Solsona E, Garcia-Orellana J, Masqué P et al. An Havskum H, Thingstad TF, Scharek R et al. Silicate and labile assessment of karstic submarine groundwater and asso- DOC interfere in structuring the microbial food web via ciated nutrient discharge to a Mediterranean coastal area algal–bacterial competition for mineral nutrients: results of (Balearic Islands, Spain) using radium isotopes. Biogeochem- a mesocosm experiment. Limnol Oceanogr 2003;48:129–140. istry 2010;97:211–29. He C, Keren R, Whittaker ML et al. Genome-resolved metage- Garcés E, Basterretxea G, Tovar-Sánchez A. Changes in microbial nomics reveals site-specific diversity of episymbiotic communities in response to submarine groundwater input. CPR bacteria and DPANN archaea in groundwater ecosys- Mar Ecol Ser 2011;438:47–58. tems. Nat Microbiol 2021. https://doi.org/10.1038/s41564-020 Gasol JM, del Giorgio PA. Using flow cytometry for counting nat- -00840-5. ural planktonic bacteria and understanding the structure of Herrmann M, Opitz S, Harzer R et al. Attached and suspended planktonic bacterial communities. Sci Mar 2000;64:197–224. denitrifier communities in pristine limestone aquifers har- Gasol JM, Pinhassi J, Alonso-Sáez L et al. Towards a better under- bor high fractions of potential autotrophs oxidizing reduced standing of microbial carbon flux in the sea. Aquat Microb Ecol iron and sulfur compounds. Microb Ecol 2017;74:264–77. 2008;53:21–38. Herrmann M, Rusznya´k A, Akob DM et al. Large fractions of CO2 - Gleeson T, Befus KM, Jasechko S et al. The global volume fixing microorganisms in pristine limestone aquifers appear and distribution of modern groundwater. Nat Geosci 2016;9: to be involved in the oxidation of reduced sulfur and nitrogen 161–7. compounds. Appl Environ Microbiol 2015;81:2384–94. Gobet A, Boer SI, Huse SM et al. Diversity and dynamics of rare Herrmann M, Wegner C-E, Taubert M et al. Predominance of and of resident bacterial populations in coastal sands. ISME Cand. Patescibacteria in groundwater is caused by their pref- J 2012;6:542–53. erential mobilization from soils and flourishing under olig- Goldscheider N, Hunkeler D, Rossi P. Review: microbial bio- otrophic conditions. Front Microbiol 2019;10:51. cenoses in pristine aquifers and an assessment of investiga- Higgins MA, Robbins GA, Maas KR et al. Use of bacteria commu- tive methods. Hydrogeol J 2006;14:926–41. nity analysis to distinguish groundwater recharge sources to Gong J, Qing Y, Guo X et al. “Candidatus Sonnebornia yantaiensis”, shallow wells. J Environ Qual 2020;49:1530–40. a member of candidate division OD1, as intracellular bacteria Hofmann R, Griebler C. DOM and bacterial growth efficiency of the ciliated protist Paramecium bursaria (Ciliophora, Oligo- in oligotrophic groundwater: absence of priming and co- hymenophorea). Syst Appl Microbiol 2014;37:35–41. limitation by organic carbon and phosphorus. Aquat Microb Gonneea ME, Charette MA, Liu Q et al. Trace element geo- Ecol 2018;81:55–71. chemistry of groundwater in a karst subterranean estu- Hofmann R, Uhl J, Hertkorn N et al. Linkage between dissolved ary (Yucatan Peninsula, Mexico). Geochim Cosmochim Acta organic matter transformation, bacterial carbon production, 2014;132:31–49. and diversity in a shallow oligotrophic aquifer: results from González BC, Iliffe TM, Macalady JL et al. Microbial hotspots in flow-through sediment microcosm experiments. Front Micro- anchialine blue holes: initial discoveries from the Bahamas. biol 2020;11:543567. Hydrobiologia 2011;677:149–56. Hong Y, Wu J, Wilson S et al. Vertical stratification of sedi- Graham EB, Crump AR, Resch CT et al. Deterministic influences ment microbial communities along geochemical gradients exceed dispersal effects on hydrologically-connected micro- of a subterranean estuary located at the Gloucester Beach of biomes. Environ Microbiol 2017;19:1552–67. Virginia, United States. Front Microbiol 2019;9:1–11. Gre´gori G, Citterio S, Ghiani A et al. Resolution of viable and Hubalek V, Wu X, Eiler A et al. Connectivity to the surface membrane-compromised bacteria in freshwater and marine determines diversity patterns in subsurface aquifers of the waters based on analytical flow cytometry and nucleic acid Fennoscandian shield. ISME J 2016;10:2447–58. double staining. Appl Environ Microbiol 2001;67:4662–70. Hunter KS, Wang Y, Van Cappellen P. Kinetic modeling of Griebler C, Avramov M. Groundwater ecosystem services: a microbially-driven redox chemistry of subsurface environ- review. Freshw Sci 2015;34:355–67. ments: coupling transport, microbial metabolism and geo- Griebler C, Lueders T. Microbial biodiversity in groundwater chemistry. J Hydrol 1998;209:53–80. ecosystems. Freshw Biol 2009;54:649–77. Hutchins BT, Engel AS, Nowlin WH et al. Chemolithoautotro- Griebler C, Malard F, Lefébure T. Current developments in phy supports macroinvertebrate food webs and affects groundwater ecology—from biodiversity to ecosystem func- diversity and stability in groundwater communities. Ecology tion and services. Curr Opin Biotechnol 2014;27:159–67. 2016;97:1530–42. Griebler C, Mindl B, Slezak D et al. Distribution patterns of Héry M, Volant A, Garing C et al. Diversity and geochemical struc- attached and suspended bacteria in pristine and contam- turing of bacterial communities along a salinity gradient in a inated shallow aquifers studied with an in situ sediment carbonate aquifer subject to seawater intrusion. FEMS Micro- exposure microcosm. Aquat Microb Ecol 2002;28:117–29. biol Ecol 2014;90:922–34. Guillemette F, McCallister SL, del Giorgio PA. Selective consump- Jewell TNM, Karaoz U, Brodie EL et al. Metatranscriptomic evi- tion and metabolic allocation of terrestrial and algal car- dence of pervasive and diverse chemolithoautotrophy rele- bon determine allochthony in lake bacteria. ISME J 2016;10: vant to C, S, N, and Fe cycling in a shallow alluvial aquifer. 1373–82. ISME J 2016;10:2106–17. Haas S, de Beer D, Klatt JM et al. Low-light anoxygenic photo- Jiang S, Zhang Y, Jin J et al. Organic carbon in a seepage face of a synthesis and Fe-S-biogeochemistry in a microbial mat. Front subterranean estuary: turnover and microbial interrelations. Microbiol 2018;9:858. Sci Total Environ 2020;725:138220. Hatzenpichler R, Scheller S, Tavormina PL et al. In situ visu- Jiao L, Wu J, He X et al. Significant microbial nitrogen loss alization of newly synthesized proteins in environmental from denitrification and anammox in the land–sea inter- microbes using amino acid tagging and click chemistry. Env- face of low permeable sediments. Int Biodeterior Biodegradation iron Microbiol 2014;16:2568–90. 2018;135:80–9. 22 FEMS Microbiology Reviews, 2021, Vol. 45, No. 5 Kayler ZE, Premke K, Gessler A et al. Integrating aquatic and ter- Lau MCY, Kieft TL, Kuloyo O et al. An oligotrophic deep- restrial perspectives to improve insights into organic mat- subsurface community dependent on syntrophy is domi- ter cycling at the landscape scale. Front Earth Sci 2019;7, doi: nated by sulfur-driven autotrophic denitrifiers. Proc Natl Acad 10.3389/feart.2019.00127. Sci USA 2016;113:E7927–36. Kellermann C, Selesi D, Lee N et al. Microbial CO2 fixation poten- Lavy A, McGrath DG, Matheus Carnevali PB et al. Microbial tial in a tar-oil-contaminated porous aquifer. FEMS Microbiol communities across a hillslope–riparian transect shaped by Ecol 2012;81:172–87. proximity to the stream, groundwater table, and weathered Kim G, Lee KK, Park KS et al. Large submarine groundwa- bedrock. Ecol Evol 2019;9:6869–900. ter discharge (SGD) from volcanic island. Geophys Res Lett Lazar CS, Lehmann R, Stoll W et al. The endolithic bacterial 2003;30:2908. diversity of shallow bedrock ecosystems. Sci Total Environ Kim J, Choi H, Kim H et al. Using isotopes (strontium and radon) 2019;679:35–44. and microbial communities to quantify groundwater mix- Lazar CS, Stoll W, Lehmann R et al. Archaeal diversity and CO2 ing influenced by anthropogenic factors at riverside area. J. fixers in carbonate-/siliciclastic-rock groundwater ecosys- Hydrol 2020;581:124441. tems. Archaea 2017;2017:2136287. Kinner NE, Harvey RW, Blakeslee K et al. Size-selective pre- Lecher AL, Mackey K, Kudela R et al. Nutrient loading dation on groundwater bacteria by nanoflagellates in an through submarine groundwater discharge and phyto- organic-contaminated aquifer. Appl Environ Microbiol 1998;64: plankton growth in Monterey bay, CA. Environ Sci Technol 618–25. 2015;49:6665–73. Kirchman DL, Knees E, Hodson R. Leucine incorporation and Lecher AL, Mackey KRM. Synthesizing the effects of submarine its potential as a measure of protein synthesis by bacteria groundwater discharge on marine biota. Hydrology 2018;5:60. in natural aquatic systems. Appl Environ Microbiol 1985;49: Lee E, Shin D, Hyun SP et al. Periodic change in coastal 599–607. microbial community structure associated with submarine Kirchman DL. Processes in Microbial Ecology. New York, NY: Oxford groundwater discharge and tidal fluctuation. Limnol Oceanogr University Press, 2012. 2017;62:437–51. Kirchman DL. Uptake and regeneration of inorganic nutrients by Lehman RM. Understanding of aquifer microbiology is tightly marine heterotrophic bacteria. In: Kirchman DL (ed.). Micro- linked to sampling approaches. Geomicrobiol J 2007;24:331–41. bial Ecology of the Oceans, Hoboken, New Jersey, USA: Wiley & Lekunberri I, Lefort T, Romera-Castillo C et al. Relationship Sons, 2000. between induced phytoplankton blooms and the structure Klindworth A, Pruesse E, Schweer T et al. Evaluation of general and dynamics of the free-living heterotrophic bacterial com- 16S ribosomal RNA gene PCR primers for classical and next- munity. Mar Ecol Ser 2012;448:23–U58. generation sequencing-based diversity studies. Nucleic Acids Lennon JT, Jones SE. Microbial seed banks: the ecological and Res 2013;41:1. evolutionary implications of dormancy. Nat Rev Microbiol Knee KL, Layton BA, Street JH et al. Sources of nutrients and 2011;119:119–30. fecal indicator bacteria to nearshore waters on the North Lennon JT, Muscarella ME, Placella SA et al. How, when Shore of Kaua’i (Hawai’i, USA). Estuaries Coasts 2008;31: and where relic DNA affects microbial diversity. mBio 607–22. 2018;9:e00637–18. Kreuzburg M, Rezanezhad F, Milojevic T et al. Carbon release and Lima-Méndez G, Faust K, Henry N et al. Determinants of com- transformation from coastal peat deposits controlled by sub- munity structure in the global plankton interactome. Science marine groundwater discharge: a column experiment study. 2015;348:1262073. Limnol Oceanogr 2020;65:1116–35. Logares R, Haverkamp THA, Kumar S et al. Environmental Kumar S, Herrmann M, Blohm A et al. Thiosulfate- and microbiology through the lens of high-throughput DNA hydrogen-driven autotrophic denitrification by a microbial sequencing: synopsis of current platforms and bioinformat- consortium enriched from groundwater of an oligotrophic ics approaches. J Microbiol Methods 2012;91:106–13. limestone aquifer. FEMS Microbiol Ecol 2018;94:y141. Longnecker K, Kujawinski EB. Using stable isotope probing to Kumar S, Herrmann M, Thamdrup B et al. Nitrogen loss from characterize differences between free-living and sediment- pristine carbonate-rock aquifers of the Hainich Critical Zone associated microorganisms in the subsurface. Geomicrobiol J Exploratory (Germany) is primarily driven by chemolithoau- 2013;30:362–70. totrophic anammox processes. Front Microbiol 2017; Lozupone CA, Knight R. UniFrac: a new phylogenetic method 8:1951. for comparing microbial communities. Appl Environ Microbiol Kwon EY, Kim G, Primeau F et al. Global estimate of sub- 2005;71:8228–35. marine groundwater discharge based on an observation- Ludington WB, Seher TD, Applegate O et al. Assessing biosyn- ally constrained radium isotope model. Geophys Res Lett thetic potential of agricultural groundwater through metage- 2014;41:8438–44. nomic sequencing: a diverse anammox community domi- Landa M, Cottrell MT, Kirchman DL et al. Phylogenetic and struc- nates nitrate-rich groundwater. PLoS One 2017;12:e0174930. tural response of heterotrophic bacteria to dissolved organic Luef B, Frischkorn KR, Wrighton KC et al. Diverse unculti- matter of different chemical composition in a continuous vated ultra-small bacterial cells in groundwater. Nat Commun culture study. Environ Microbiol 2014;16:1668–81. 2015;6:6372. Langenheder S, Comte J, Langenheder S et al. Remnants of Luijendijk E, Gleeson T, Moosdorf N. Fresh groundwater dis- marine bacterial communities can be retrieved from deep charge insignificant for the world’s oceans but important for sediments in lakes of marine origin. Environ Microbiol Rep coastal ecosystems. Nat Commun 2020;11:1260. 2016;8:479–85. Magnabosco C, Lin L-H, Dong H et al. The biomass and biodiver- Langenheder S, Lindström ES. Factors influencing aquatic and sity of the continental subsurface. Nat Geosci 2018;11:707–17. terrestrial bacterial community assembly. Environ Microbiol Maister O. Effects of Submarine Groundwater Discharge on Coastal Rep 2019;11:306–15. Bacterioplankton Communities from the NW Mediterranean. M.Sc. Ruiz-González et al. 23 Thesis. Instituto de Cièncias del Mar(ICM-CSIC), Barcelona. Oehler T, Tamborski J, Rahman S et al. DSi as a tracer for subma- 2018. ( https://digital.csic.es/handle/10261/185172) rine groundwater discharge. Front Mar Sci 2019;6:563. Manz W, Amann R, Ludwig W et al. Phylogenetic oligodeoxynu- Ortiz-Álvarez R, Cáliz J, Camarero L et al. Regional commu- cleotide probes for the major subclasses of Proteobacteria: nity assembly drivers and microbial environmental sources problems and solutions. Syst Appl Microbiol 1992;15:593–600. shaping bacterioplankton in an alpine lacustrine district Mayfield KK, Eisenhauer A, Santiago Ramos DP et al. Groundwa- (Pyrenees, Spain). Environ Microbiol 2019;22:297–309. ter discharge impacts marine isotope budgets of Li, Mg, Ca, Pain AJ, Martin JB, Young CR. Sources and sinks of CO2 and Sr, and Ba. Nat Commun 2021;12:148. CH4 in siliciclastic subterranean estuaries. Limnol Oceanogr McAllister SM, Barnett JM, Heiss JW et al. Dynamic hydrologic 2019;64:1500–14. and biogeochemical processes drive microbially enhanced Pan D, Watson R, Wang D et al. Correlation between viral produc- iron and sulfur cycling within the intertidal mixing zone of a tion and carbon mineralization under nitrate-reducing con- beach aquifer. Limnol Oceanogr 2015;60:329–45. ditions in aquifer sediment. ISME J 2014;8:1691–703. McClain ME, Boyer EW, Dent CL et al. Biogeochemical hot spots Paytan A, Boehm AB, Shellenbarger GG. Bacterial contamination and hot moments at the interface of terrestrial and aquatic and submarine groundwater discharge: a possible link. Envi- ecosystems. Ecosystems 2003;6:301–12. ron Chem 2004;1:1–2. Meckenstock RU, Elsner M, Griebler C et al. Biodegradation: Pilloni G, Bayer A, Ruth-Anneser B et al. Dynamics of hydrology updating the concepts of control for microbial cleanup in and anaerobic hydrocarbon degrader communities in a tar- contaminated aquifers. Environ Sci Technol 2015;49:7073–81. oil contaminated aquifer. Microorganisms 2019;7:46. Meier A, Singh MK, Kastner A et al. Microbial communities in Posman KM, DeRito CM, Madsen EL. Benzene degradation carbonate rocks-from soil via groundwater to rocks. J Basic by a Variovorax species within a coal tar-contaminated Microbiol 2017;57:752–61. groundwater microbial community. Appl Environ Microbiol Mejı́as M, Ballesteros BJ, Antón-Pacheco C et al. Methodologi- 2017;83:e02658–16. cal study of submarine groundwater discharge from a karstic Probst AJ, Castelle CJ, Singh A et al. Genomic resolution of a cold aquifer in the Western Mediterranean Sea. J Hydrol 2012;464– subsurface aquifer community provides metabolic insights 5:27–40. for novel microbes adapted to high CO2 concentrations. Env- Menning DM, Carraher-Stross WA, Graham ED et al. Aquifer dis- iron Microbiol 2017;19:459–74. charge drives microbial community change in karst estuar- Probst AJ, Ladd B, Jarett JK et al. Differential depth distribu- ies. Estuaries Coast 2018;41:430–43. tion of microbial function and putative symbionts through Michael HA, Post VEA, Wilson AM et al. Science, society, and the sediment-hosted aquifers in the deep terrestrial subsurface. coastal groundwater squeeze. Sci Soc Coast Groundw Squeeze Nat Microbiol 2018;3:328–36. 2017;53:2610–7. Proctor CR, Besmer MD, Langenegger T et al. Phylogenetic clus- Missimer TM, Hoppe-Jones C, Jadoon KZ et al. Hydrogeology, tering of small low nucleic acid-content bacteria across water quality, and microbial assessment of a coastal allu- diverse freshwater ecosystems. ISME J 2018;12:1344–59. vial aquifer in western Saudi Arabia: potential use of coastal Pérez MT, Sommaruga R. Differential effect of algal- and soil- wadi aquifers for desalination water supplies. Hydrogeol J derived dissolved organic matter on alpine lake bacte- 2014;22:1921–34. rial community composition and activity. Limnol Oceanogr Miyoshi T, Iwatsuki T, Naganuma T. Phylogenetic character- 2006;51:2527–37. ization of 16S rRNA gene clones from deep-groundwater Rajala P, Bomberg M. Reactivation of deep subsurface microbial microorganisms that pass through 0.2-micrometer-pore-size community in response to methane or methanol amend- filters. Appl Environ Microbiol 2005;71:1084–8. ment. Front Microbiol 2017;8:431. Mojica KDA, Brussaard CPD. Factors affecting virus dynamics Reckhardt A, Beck M, Greskowiak J et al. Cycling of redox- and microbial host–virus interactions in marine environ- sensitive elements in a sandy subterranean estuary of the ments. FEMS Microbiol Ecol 2014;89:495–515. southern North Sea. Mar Chem 2017;188:6–17. Montiel D, Lamore AF, Stewart C et al. Natural groundwater Retter A, Karwautz C, Griebler C. Groundwater microbial com- nutrient fluxes exceed anthropogenic inputs in an ecolog- munities in times of climate change. Curr Issues Mol Biol ically impacted estuary: lessons learned from Mobile Bay, 2021;41:509–38. Alabama 2019;145:1–33. Robinson CE, Xin P, Santos IR et al. Groundwater dynamics in Moore WS. The effect of submarine groundwater discharge on subterranean estuaries of coastal unconfined aquifers: Con- the ocean. Ann Rev Mar Sci 2010;2:59–88. trols on submarine groundwater discharge and chemical Moore WS. The subterranean estuary: a reaction zone of ground inputs to the ocean. Adv Water Resour 2018;115:315–31. water and sea water. Mar Chem 1999;65:111–25. Rocca JD, Simonin M, Bernhardt ES et al. Rare microbial Nakai R. Size matters: ultra-small and filterable microorganisms taxa emerge when communities collide: freshwater and in the environment. Microbes Environ 2020;35:n/a. marine microbiome responses to experimental mixing. Ecol- Nelson WC, Stegen JC. The reduced genomes of Parcubacteria ogy 2020;101:e02956. (OD1) contain signatures of a symbiotic lifestyle. Front Micro- Rodellas V, Garcia-Orellana J, Masqué P et al. Submarine ground- biol 2015;6:713. water discharge as a major source of nutrients to the Niño-Garcı́a JP, Ruiz-González C, del Giorgio PA. Landscape-scale Mediterranean Sea. Proc Natl Acad Sci USA 2015;112:3926–30. spatial abundance distributions discriminate core from ran- Rodellas V, Garcia-Orellana J, Trezzi G et al. Using the radium dom components of boreal lake bacterioplankton. Ecol Lett quartet to quantify submarine groundwater discharge and 2016;19:1506–15. porewater exchange. Geochim Cosmochim Acta 2017;196: Nowak ME, Schwab VF, Lazar CS et al. Carbon isotopes of dis- 58–73. solved inorganic carbon reflect utilization of different carbon Rogers DR, Casciotti KL. Abundance and diversity of archaeal sources by microbial communities in two limestone aquifer ammonia oxidizers in a coastal groundwater system. Appl assemblages. Hydrol Earth Syst Sc 2017;21:4283–300. Environ Microbiol 2010;76:7938–48. 24 FEMS Microbiology Reviews, 2021, Vol. 45, No. 5 Ruiz-González C, Lefort T, Massana R et al. Diel changes in bulk Seymour J, Humphreys W, Mitchell J. Stratification of the micro- and single-cell bacterial heterotrophic activity in winter sur- bial community inhabiting an anchialine sinkhole. Aquat face waters of the northwestern Mediterranean Sea. Limnol Microb Ecol 2007;50:11–24 Oceanogr 2012;57:29–42. Shao Y, Hatzinger PB, Streger SH et al. Evaluation of methan- Ruiz-González C, Niño-Garcı́a JP, del Giorgio PA. Terrestrial ori- otrophic bacterial communities capable of biodegrading gin of bacterial communities in complex boreal freshwater trichloroethene (TCE) in acidic aquifers. Biodegradation networks. Ecol Lett 2015a;18:1198–206. 2019;30:173–90. Ruiz-González C, Niño-Garcı́a JP, Lapierre JF et al. The quality of Shen D, Jürgens K, Beier S. Experimental insights into the organic matter shapes the functional biogeography of bac- importance of ecologically dissimilar bacteria to commu- terioplankton across boreal freshwater ecosystems. Glob Ecol nity assembly along a salinity gradient. Environ Microbiol Biogeogr 2015b;11:11–2. 2018;20:1170–84. Russell TL, Sassoubre LM, Wang D et al. A coupled modeling and Sieracki ME, Cucci TL, Nicinski J. Flow cytometric analysis of molecular biology approach to microbial source tracking at 5-cyano-2,3-ditolyl tetrazolium chloride activity of marine Cowell Beach, Santa Cruz, CA, United States. Environ Sci Tech- bacterioplankton in dilution cultures. Appl Environ Microbiol nol 2013;47:10231–9. 1999;65:2409–17. Sachidanandham R, Yew-Hoong Gin K. A dormancy state Sjo¨stedt J, Koch-Schmidt P, Pontarp M et al. Recruitment of in nonspore-forming bacteria. Appl Microbiol Biotechnol members from the rare biosphere of marine bacterioplank- 2009;81:927–41. ton communities after an environmental disturbance. Appl Sang S, Dai H, Hu BX et al. The study of hydrogeochemical envi- Environ Microbiol 2012;78:1361–9. ronments and microbial communities along a groundwa- Slomp CP, Van Cappellen P. Nutrient inputs to the coastal ocean ter salinity gradient in the Pearl River Delta, China. Water through submarine groundwater discharge: controls and 2019;11:804. potential impact. J Hydrol 2004;295:64–86. Sang S, Zhang X, Dai H et al. Diversity and predictive metabolic Small C, Nichols RJ. A global analysis of human settlement in pathways of the prokaryotic microbial community along a coastal zones. J Coast Res 2003;19:584–99. groundwater salinity gradient of the Pearl River Delta, China. Smith D, Azam F. A simple, economical method for measur- Sci Rep 2018;8:17317. ing bacteria protein synthesis rates in seawater using 3H- Santoro AE, Boehm AB, Francis CA. Denitrifier commu- leucine. Mar Microb Food Webs 1992;6:107–14. nity composition along a nitrate and salinity gradi- Smith HJ, Zelaya AJ, De León KB et al. Impact of hydrologic ent in a coastal aquifer. Appl Environ Microbiol 2006;72: boundaries on microbial planktonic and biofilm communi- 2102–9. ties in shallow terrestrial subsurface environments. FEMS Santoro AE, Francis CA, de Sieyes NR et al. Shifts in the rela- Microbiol Ecol 2018;94:1–16. tive abundance of ammonia-oxidizing bacteria and archaea Smith MB, Rocha AM, Smillie CS et al. Natural bacterial com- across physicochemical gradients in a subterranean estuary. munities serve as quantitative geochemical biosensors. mBio Environ Microbiol 2008;10:1068–79. 2015;6:e00326–15. Santoro AE. Microbial nitrogen cycling at the saltwater– Sola F, Vargas-Garcı́a MC, Vallejos A. Interrelation prokary- freshwater interface. Hydrogeol J 2010;18:187–202. otic community-aquifer in a carbonate coastal environment. Santos IR, Eyre BD, Huettel M. The driving forces of porewater Aquat Sci 2020;82:13. and ground-water flow in permeable coastal sediments: a Stegen JC, Fredrickson JK, Wilkins MJ et al. Groundwater-surface review. Estuaries Coast Shelf Sci 2012;98:1–15. water mixing shifts ecological assembly processes and stim- Sarmento H, Gasol JM. Use of phytoplankton-derived dissolved ulates organic carbon turnover. Nat Commun 2016;7:11237. organic carbon by different types of bacterioplankton. Envi- Sun S, Jones RB, Fodor AA. Inference-based accuracy of ron Microbiol 2012;14:2348–60. metagenome prediction tools varies across sample types and Scanlan DJ, Ostrowski M, Mazard S et al. Ecological genomics functional categories. Microbiome 2020;8:46. of marine picocyanobacteria. Microbiol Mol Biol Rev 2009;73: Sáenz JP, Hopmans EC, Rogers D et al. Distribution of anaerobic 249–99. ammonia-oxidizing bacteria in a subterranean estuary. Mar Schloss PD, Girard RA, Martin T et al. Status of the Archaeal and Chem 2012;136–7:7–13. Bacterial census: an update. mBio 2016;7:e00201–16. Taniguchi M, Dulai H, Burnett KM et al. Submarine groundwa- Schmidt SI, Cuthbert MO, Schwientek M. Towards an inte- ter discharge: updates on its measurement techniques, geo- grated understanding of how micro scale processes shape physical drivers, magnitudes, and effects. Front Environ Sci groundwater ecosystem functions. Sci Total Environ 2017;592: 2019;7:141. 215–27. Taubert M, Stockel S, Geesink P et al. Tracking active groundwater Schutte CA, Wilson AM, Evans T et al. Methanotrophy controls microbes with D2 O labelling to understand their ecosystem groundwater methane export from a barrier island. Geochim function. Environ Microbiol 2018;20:369–84. Cosmochim Acta 2016;179:242–56. Taubert M, Stähly J, Kolb S et al. Divergent microbial commu- Schwank K, Bornemann TL V, Dombrowski N et al. An archaeal nities in groundwater and overlying soils exhibit functional symbiont–host association from the deep terrestrial subsur- redundancy for plant-polysaccharide degradation. PLoS One face. ISME J 2019;13:2135–9. 2019;14:e0212937. Sebastián M, Gasol JM. Visualization is crucial for understand- Teira E, Logares R, Gutierrez-Barral A et al. Impact of grazing, ing microbial processes in the ocean. Philos Trans R Soc B resource availability and light on prokaryotic growth and 2019;374:20190083. diversity in the oligotrophic surface global ocean. Environ Seibert SL, Degenhardt J, Ahrens J et al. Investigating the land– Microbiol 2019;21:1482–96. sea transition zone. In: Jungblut S, Liebich V, Bode-Dalby M Tian R, Ning D, He Z et al. Small and mighty: adaptation of super- (eds). YOUMARES 9 - The Oceans: Our Research, Our Future. phylum Patescibacteria to groundwater environment drives Switzerland: Springer Nature, 2020, 225–42. their genome simplicity. Microbiome 2020;8:51. Ruiz-González et al. 25 Tovar-Sánchez A, Basterretxea G, Rodellas V et al. Contribution Whitman WB, Coleman DC, Wiebe WJ. Prokaryotes: the unseen of groundwater discharge to the coastal dissolved nutrients majority. Proc Natl Acad Sci USA 1998;95:6578–83. and trace metal concentrations in Majorca Island: karstic vs Wilhartitz IC, Kirschner AKT, Brussaard CPD et al. Dynam- detrital systems. Environ Sci Technol 2014;48:11819–27. ics of natural prokaryotes, viruses, and heterotrophic Troussellier M, Schafer H, Batailler N et al. Bacterial activity and nanoflagellates in alpine karstic groundwater. Microbiology- genetic richness along an estuarine gradient (Rhone River open 2013;2:633–43. plume, France). Aquat Microb Ecol 2002;28:13–24. Wilhartitz IC, Kirschner AKT, Stadler H et al. Heterotrophic Unno T, Kim J, Kim Y et al. Influence of seawater intrusion prokaryotic production in ultraoligotrophic alpine karst on microbial communities in groundwater. Sci Total Environ aquifers and ecological implications. FEMS Microbiol Ecol 2015;532:337–43. 2009;68:287–99. Velasco Ayuso S, Acebes P, López-Archilla AI et al. Environ- Wilkins LGE, Leray M, O’Dea A et al. Host-associated micro- mental factors controlling the spatiotemporal distribution of biomes drive structure and function of marine ecosystems. microbial communities in a coastal, sandy aquifer system PLoS Biol 2019;17:e3000533. (Doñana, southwest Spain). Hydrogeol J 2009a;17:767–80. Wisnoski NI, Muscarella ME, Larsen ML et al. Metabolic Velasco Ayuso S, del Carmen Guerrero M, Montes C et al. Regu- insight into bacterial community assembly across ecosystem lation and spatiotemporal patterns of extracellular enzyme boundaries. Ecology 2020;101:e02968. activities in a coastal, sandy aquifer system (Doñana, SW Wong WW, Applegate A, Poh SC et al. Biogeochemical attenu- Spain). Microb Ecol 2011;62:162–76. ation of nitrate in a sandy subterranean estuary: insights Velasco Ayuso S, Guerrero MC, Montes C et al. Spatiotemporal from two stable isotope approaches. Limnol Oceanogr 2020;65: distribution of microbial communities in a coastal, sandy 3098–113. aquifer system (Doñana, SW Spain). Geobiology 2009b;7: Wörmer L, Hoshino T, Bowles MW et al. Microbial dormancy 66–81. in the marine subsurface: global endospore abundance and Velasco Ayuso S, López-Archilla AI, Montes C et al. Microbial response to burial. Sci Adv 2019;5:eaav1024. activities in a coastal, sandy aquifer system (Doñana Natural Xenopoulos MA, Downing JA, Kumar MD et al. Ecological and bio- Protected Area, SW Spain). Geomicrobiol J 2010;27:409–23. geochemical contrasts across the aquatic continuum. Limnol Velimirov B. Nanobacteria, ultramicrobacteria and starvation Oceanogr 2017;62:S3–14. forms: a search for the smallest metabolizing bacterium. Yau VM, Schiff KC, Arnold BF et al. Effect of submarine Microbes Environ 2001;16:67–77. groundwater discharge on bacterial indicators and swim- Vigneron A, Bishop A, Alsop EB et al. Microbial and isotopic mer health at Avalon Beach, CA, USA. Water Res 2014;59: evidence for methane cycling in hydrocarbon-containing 23–36. groundwater from the Pennsylvania Region. Front Microbiol Ye Q, Liu J, Du J et al. Bacterial diversity in submarine ground- 2017;8:1–12. water along the coasts of the Yellow Sea. Front Microbiol Vollberg F, Walther M, Gärdes A et al. Modeling the potential of 2016;6:1519. submarine groundwater discharge to facilitate growth of Vib- Youssef NH, Rinke C, Stepanauskas R et al. Insights into rio cholerae bacteria. Hydrology 2019;6:39. the metabolism, lifestyle and putative evolutionary his- Wang S, Zhu G, Zhuang L et al. Anaerobic ammonium oxidation tory of the novel archaeal phylum ‘Diapherotrites’. ISME J is a major N-sink in aquifer systems around the world. ISME 2015;9:447–60. J 2020;14:151–63. Zhang X, Qi L, Li W et al. Bacterial community variations with Werner AD, Bakker M, Post VEA et al. Seawater intrusion pro- salinity in the saltwater-intruded estuarine aquifer. Sci Total cesses, investigation and management: recent advances and Environ 2021;755:142423. future challenges. Adv Water Resour 2013;51:3–26. Zhou Y, Kellermann C, Griebler C. Spatio-temporal pat- Whiticar MJ. Carbon and hydrogen isotope systematics of terns of microbial communities in a hydrologically bacterial formation and oxidation of methane. Chem Geol dynamic pristine aquifer. FEMS Microbiol Ecol 2012;81: 1999;161:291–314. 230–42.