Tidal saltmarshes provide a number of ecosystem services which are valuable to humans, including coastal protection from storms, carbon sequestration, and denitrification (Barbier et al., 2011). These habitats also act as nurseries for commercially and recreationally important species (Boesch and Turner, 1984; Zimmerman et al., 2002; Mackenzie and Dionne, 2008). However, salt marshes across the globe are in decline due to a variety of factors (MEA, 2005; UNEP, 2006).

The marshes of Louisiana were initially built by yearly flooding of the Mississippi River over thousands of years (Fisk and McFarlan, 1955) but have lost an area the size of the state of Delaware since 1932 (Couvillion et al., 2017) due to flood control structures preventing these floods and sediment accretion. River sediment compacts readily causing subsidence, which together with sea-level rise leads to land loss and salt water intrusion (Delaune and Pezeshki, 1994; Yuill et al., 2009). In some areas of Louisiana marsh, subsidence may exceed 1 cm per year, which is among the highest rates of subsidence along the Gulf Coast of the United States (Penland and Ramsey, 1990). In addition, marshland is lost due to erosion exacerbated by hurricanes and oil spills (Palaseanu-Lovejoy et al., 2013; Rabalais and Turner, 2016). For example, oiling after the 2010 Deepwater Horizon oil spill of Louisiana’s salt marshes, which are dominated by Spartina alterniflora (Loisel.) and Juncus roemerianus (Scheele), led in some areas to complete loss of vegetation, which resulted in shearing of the marsh edge and accelerated land loss after hurricanes (Lin and Mendelssohn, 2012; Mendelssohn et al., 2012; Silliman et al., 2012; Zengel et al., 2015; Khanna et al., 2017).

A comprehensive Louisiana Coastal Master Plan pledging $50 billion to 79 coastal restoration and protection projects through the next half century has been compiled by the Coastal Resource Management Agency to counteract the alarming rate of loss of Louisiana marshes (Coastal Protection and Restoration Authority of Louisiana, 2007; Coastal Protection and Restoration Authority of Louisiana, 2012; Peyronnin et al., 2013; Coastal Protection and Restoration Authority of Louisiana, 2017). Included in these projects are several sediment and freshwater diversions planned to divert flow from the Mississippi River into the marshes to promote land accretion and reduce saltwater intrusion into historically freshwater marshes (Elsey-Quirk et al., 2019). The first of these major sediment diversions is the Mid-Barataria Sediment Diversion, which was slated to begin construction in 2022 but is now delayed. Environmental impact studies of these diversions will require the establishment of baselines of the composition and distribution of organisms in the marsh. Because salinity is one of the dominant environmental variables in salt marshes (Chapman, 1977; Holland et al., 1990), detailed inventories of different salinity zones are vital to measuring effects of freshwater diversions, saltwater intrusion and other environmental disturbances on the ecology of the marsh community.

Invertebrate organisms are often used for ecological monitoring because they have a high diversity of species, life histories and niches and are responsive to environmental change such as salinity shifts. Therefore, Aker (Aker, 2020) established detailed year-long insect inventories across different salinity zones in Louisiana marshes and identified bioindicator taxa for salinity zones. In addition, sediment-dwelling infauna acts as a linkage between microbe and macrofauna food webs and between benthic and pelagic food webs (Coull, 1999; Posey et al., 2005; Rozas and Minello, 2011; Schratzberger and Ingels, 2018). Meiofauna are infauna typically ranging from 45 to 500 μm (Giere, 2009). The intergenerational time of many groups of meiofauna is quite short, with multiple generations per year, allowing rapid response in both diversity and abundance of major taxa to environmental changes (Warwick, 1981; Kennedy and Jacoby, 1999; Montagna et al., 2002; Cai et al., 2012; Ngo et al., 2013; Zeppilli et al., 2015). Therefore, meiofauna are often used as bioindicators of marsh health, especially with regards to pollution responses (Fleeger et al., 1982; Fleeger and Chandler, 1983; Fleeger, 1985; Fleeger et al., 2007; Fleeger et al., 2015; Semprucci et al., 2015; Bam et al., 2018). However, soil infauna is also likely to be affected by salinity changes. Meiofauna abundance has been shown to increase in salt marshes as salinity increases due to the influx of marine species (Montagna et al., 2002). Salinity was also a major factor impacting benthic infauna across marsh zones from tidal flats to high marsh of the Atlantic coast (MacKenzie et al., 2015; van Regteren et al., 2020).

While meiofauna have been studied in both Gulf and Atlantic coastal marshes, these studies tend to identify meiofauna to major taxonomic groups such as nematodes, harpacticoid copepods, and oligochaetes (Bell, 1979; Coull and Bell, 1979; Fleeger, 1985; Kennedy and Jacoby, 1999; MacKenzie et al., 2015; van Regteren et al., 2020) or a limited number of well described species (Fleeger et al., 2015; Fleeger et al., 2018; Fleeger et al., 2019). Meiofauna diversity in general remains poorly known at the species level (Snelgrove et al., 1997; Gielings et al., 2021). This is largely due to the effort and technical experience required to identify meiofauna to the species level with classical morphological methods. Traditional studies of meiofauna also are limited by capture of certain groups, especially the smaller, soft bodied organisms such as the Platyhelminthes, Rotifera, and Gastrotricha, due to sampling bias and preservation issues (Kennedy and Jacoby, 1999; Carugati et al., 2015; Leasi et al., 2018). To overcome these limitations, studies on meiofauna are increasingly using DNA metabarcoding methods for rapid, cost-effective identification of meiofauna over traditional methods (Bik et al., 2012; Brannock et al., 2014; Brannock and Halanych, 2015; Creer et al., 2016; Husseneder et al., 2022, reviewed in Gielings et al., 2021). Metabarcoding is used to assess diversity via high-throughput sequencing of amplified DNA markers and assign taxonomic information by comparison to sequences of identified organisms in reference databases (Taberlet et al., 2012). Metabarcoding captures and analyzes DNA from all living organisms in a sample without the need of isolating them, but also from genetic material shed into the environment (environmental DNA, Zaiko et al., 2018). The use of DNA sequencing methods detects even rare taxa and cryptic life stages and allows diversity comparisons even when only a fraction of the species is described and, thus, often exceeds conventional diversity assessments in terms of taxonomic resolution, precision and sensitivity (Zaiko et al., 2018). Thus, metabarcoding is an excellent tool in ecological monitoring (Chariton et al., 2015; DiBattista et al., 2020).

The purpose of this study was to describe metazoan community diversity consisting of meiofauna and environmental DNA at different salinity zones in Louisiana estuaries to detect changes in these communities caused by the planned freshwater diversion projects or other unforeseen environmental disturbances including saltwater intrusion and pollution. Communities were described from three salinity zones in two Louisiana bays, Barataria Bay and Caillou Bay, which previously were selected for the collection of baseline insect community data across the salinity zones (Aker, 2020). Samples were collected from four distances from the marsh edge to capture diversity and test whether there would be distance-dependent differences in metazoan composition. The timing of this study was critical for the collection of a baseline taxa and diversity inventory since Barataria Bay likely will be immediately affected by the Mid-Barataria Sediment Diversion project when it is initiated (Coastal Protection and Restoration Authority of Louisiana, 2007; Mccall and Pennings, 2012; Minowa and Garaffoni, 2017).

Coastal marsh ecosystems, especially within Louisiana, are at risk due to numerous natural and anthropogenic environmental stressors (Yuill et al., 2009; Wong et al., 2010; Lin and Mendelssohn 2012; Kirwan et al., 2016). Assessment of the impact of these stressors on marsh communities and habitat remediation via restoration projects must take changes in salinity levels and resulting shifts in marsh communities into account. Therefore, detailed inventories of organisms relative to salinity zones are needed as baselines to measure impacts of environmental changes on marsh community health.

Soil meiofauna and macrofauna (collectively, infauna) have been studied for decades in Louisiana’s coastal marshes and inventories based on morphological classification are relatively complete for major taxa (Fleeger, 1985; Fleeger et al., 2015; Fleeger et al., 2018; Fleeger et al., 2019). However, morphological identification is likely to underestimate meiofaunal diversity (Gielings et al., 2021) and only few studies included eukaryote metabarcoding to describe infauna diversity in Louisiana’s tidal marshes (Husseneder et al., 2022). Moreover, species inventories of Louisiana’s tidal marshes lack detailed metadata concerning salinity, which is expected to be one of the major factors changing due to sea level rise and freshwater diversions (Das et al., 2012). This project was designed to provide baselines of metazoan biodiversity based on meiofauna and environmental DNA across salinity zones to monitor changes in marsh habitat primarily due to salinity changes with saltwater intrusion or freshwater influx when the planned Mid-Barataria Sediment Diversion commences (Coastal Protection and Restoration Authority of Louisiana, 2007; Coastal Protection and Restoration Authority of Louisiana, 2012; Coastal Protection and Restoration Authority of Louisiana, 2017; Peyronnin et al., 2013). Freshwater flow has been shown to drastically change meiofauna populations (Montagna and Kalke 1992; Montagna et al., 2002).

In summary, our study showed that salinity, but not bay or distance from the marsh edge significantly impacted metazoan Alpha-diversity within samples. Salinity was also a major factor accounting for 20% of the variation in community composition among samples, with many frequent OTUs being uniquely present or absent in salinity zones. Bay and distance from the marsh edge as well as interactions among the factors only had minor impacts on community differentiation (all <8%).

Differences in metazoan communities among bays might be driven by differences in the plant communities between the bays. For example, Marionina spp. (Annelida), which are likely associated with Spartina stems (Healy and Walters, 1994), were more common in the Barataria Bay, which had more Spartina ground cover than Caillou Bay (Aker, 2020). Long-term impacts of the Deepwater Horizon oil spill might also explain some differences between bays. Taxa with slow recovery due to poor dispersal activity including the kinorynchs (Fleeger et al., 2018; Fleeger et al., 2019), might still endure reduced abundance in Barataria Bay, which had been more heavily impacted by the Deepwater Horizon oil spill than Caillou Bay (Michel et al., 2013). Kinorhyncha detected by metabarcoding made up less than 1% of the total OTUs and were assigned as Echinoderes sp. Echinoderes were an indicator for oil pollution, since they were absent from the heavily oiled marsh sites following the Deepwater Horizon oil spill (Fleeger et al., 2015; Fleeger et al., 2018). Even though our sites were not oiled, area wide population reduction for some species has been observed even at sites not directly oiled (Husseneder et al., 2022).

Previous studies reported distance effects with higher invertebrate densities near the marsh-water interface due to larval recruitment and food deposited by inundation (Kneib, 1984; MacKenzie et al., 2015) and clear differentiation of benthos communities across marsh zones from the tidal flats landward to the pioneer zone, low marsh and high marsh (MacKenzie et al., 2015; van Regteren et al., 2020). Dominance of annelids and mollusks in the pioneer zone and tidal flats reported by van Regentren et al. (2020) supports the higher frequency of detection of these phyla at our sample sites close to the marsh edge. In contrast, MacKenzie et al. (2015) found higher species richness in the high marsh compared to low marsh but this was due to sampling across larger distances and elevations compared to our study. Overall, distance effects were much less pronounced in our metabarcoding study and none of the frequently detected OTUs were uniquely present or absent at a certain distance. Our study used much shorter transects, which explains why less community differentiation was detected compared to other studies across marsh zones. Moreover, with the low elevation gain of Louisiana marshes even our sites furthest from the marsh edge (10 m) received frequent inundation.

Similar to our study, salinity was a major explanatory variable of benthic macro- and meiofauna community variation in previous studies. Generally, macro- and meiofauna diversity increases with higher salinity along marine salinity gradients (Broman et al., 2019). Studies in saltmarshes of the East coast and estuaries of the Texas Gulf coast found weak correlations of salinity to observed soil invertebrate communities (MacKenzie et al., 2015; Van Diggelen and Montagna, 2016). However, in our metabarcoding study, the relationship of overall metazoan diversity and salinity was U-shaped with the lowest values in mid salinity zones. In particular, insects and annelids had the lowest detection frequency in mid salinity. This pattern has been observed frequently in estuarine systems (Deaton and Greenberg, 1986).The mid salinity zone in our study averaged approximately 8.5 ppt, which is close to the “Artenminimum” zone between 5 and 8 ppt where overall species numbers are lowest in estuaries and bodies of brackish water but numbers of brackish water species are highest (Remane, 1931; Deaton and Greenberg, 1986). The low species diversity in the mid salinity zone has been hypothesized to occur because speciation rates in dilute brackish environments are lower than for freshwater and marine species and few animals acquire successful physiological adaptations to tolerate the range of conditions in this habitat (Deaton and Greenberg, 1986).

In traditional studies of tidal marsh fauna, nematodes are typically among the most abundant non-insect invertebrate group (Bell, 1979; Fleeger, 1985; Alves et al., 2013; MacKenzie et al., 2015), which was consistent with nematodes being the most frequently detected group in our study. Nematodes are known for their ability to tolerate extreme environmental conditions (Zeppilli et al., 2018). Previous studies showed a positive correlation of nematode diversity or abundance with salinity gradients of intertidal rivers and the Baltic Sea (Tran et al., 2018; Broman et al., 2019). However, no such relationship was detected with our incidence data across similar salinity ranges. The most frequently detected nematode, S. parasitifera a marine microbivore in the family Desmodoridae also commonly found in European estuaries (Soetaert et al., 1995; Hodda, 2011), was detected across all three salinity zones. Five nematode OTUs of the frequent group were never detected in samples from the high salinity zone: E. filiformis, two Prismatolaimus sp. OTUs, Anoplostoma sp., and Prodorylaimus sp. We predict that salinity reduction following the opening of the Mid-Barataria freshwater diversion will lead to range expansion of these OTUs into formerly high salinity areas. The opposite is expected for taxa predominantly found in high salinity, such as the plant-parasitic root-knot nematode M. spartinae (Plantard et al., 2007) and an OTU of the genus Prochaetosoma, which contains nematode species known from sub-tidal zones (Rho et al., 2010). Notably, some nematodes can survive in hypersaline zones with salinities higher than seawater (Zeppilli et al., 2018).

Arthropoda were the second most abundant group in the dataset after the nematodes. Similar to the nematode distribution, arthropod incidences were approximately evenly distributed across the different bays and distances from marsh edges, but Arthropods were more commonly detected in the low salinity zone than in the mid and high salinity zone. This was mainly caused by non-insect arthropods. Frequently detected non-insect arthropods that were possible bioindicators for low salinity included Collembola assigned as ambiguous taxa and C. sverdrupi as well as ostracods of the Cyprididae and Cypria sp.

Insects are typically among the most common organisms in the low and high marsh (MacKenzie et al., 2015; van Regteren et al., 2020). To confirm taxonomic assignments of the insect OTUs and compare incidences to organismal abundance, we compared our data to those of Aker (2020), who conducted a year-long sampling of insects (2018-2019) using sweep net techniques at the same sites and salinity zones as sampled in this study. While insect abundance cannot be directly compared between metabarcoding and sweep net collection studies due to differences in sampling strategy and data representation, broader insect community comparisons can be made. Aker’s studies found that insect abundance increased dramatically with salinity while insect family-level biodiversity decreased as salinity increased, which mirrored the decrease in plant species. In contrast, the DNA-based incidences and number of insect OTUs and families was U-shaped across salinity zones with slightly lower numbers detected at the mid salinity zone and high salinity zones.

Over half of the insect OTUs detected were unique to a single salinity zone. Being unique for a salinity zone alone could not unequivocally identify these insect OTUs as salinity bioindicators due to their low detection rate; however, closely related species have been confirmed independently as salinity bioindicators (Aker, 2020). Ischnura ramburii (Coenagrionidae, Odonata) and C. obscura (Derbidae, Hemiptera) were identified as bioindicators of low salinity by Aker (2020). Therefore, the metabarcoded OTUs from the families Derbidae (C. obscura) and Coenagrionidae (ambiguous taxa) likely represent bioindicators of low-salinity insect communities because they were detected at the low salinity sites within a month of their peak abundances described by Aker (2020). Biting midges of the genus Culicoides (Ceratopogonidae) are abundant in marshes along the Atlantic (Kneib, 1984; MacKenzie et al., 2015) and Gulf coast (Aker, 2020). The Culicoides OTU detected predominantly in mid and high salinity zones is likely the highly abundant high salinity bioindicator identified in Aker’s studies, Culicoides mississippiensis.

Fly larvae are common in saltmarsh macrofaunal samples but are typically rare in meiofauna assemblages (Kneib, 1984). In our study, five Diptera OTUs belonging to gall midges (Cecidomyiidae) and tabanids (Chrysops and Tabanus spp.) were common among all salinity zones. Gall midges of salt marsh succulents have been described from Australia (Veenstra-Quah et al., 2007), but were not found in Aker’s sweep net surveys because larvae are found in plant tissue and dispersing adults might only be present for a limited time. The detection rate of tabanids across samples and their wide distribution across salinity zones is remarkable and supports the prevalence of tabanid larvae in salt marsh sediment where they are top invertebrate predators (Kneib, 1984; Husseneder et al., 2022). All tabanid OTUs belonged to the known estuarine species of Tabanidae (Tabanus nigrovittatus, T. acutus, T. hinellus, and C. flavidus) collected from the same coastal sites by Aker (2020) with sweep nets and Davis (2022) using tabanid targeted trap techniques (Hribar et al., 1991). Tabanus nigrovittatus is the saltmarsh greenhead horse fly that showed severe population crashes after the Deepwater Horizon oil spill and is thus regarded as a bioindicator of oil pollution (Husseneder et al., 2016; Husseneder et al., 2018; Husseneder et al., 2022). Although direct comparison of the different detection methods is difficult, the sweep net and metabarcoding inventories represent complementary aspects of the marsh community. The advantage of the use of metabarcoding over traditional insect surveys by collection is that insect taxa can be monitored without the need for catching and identifying intact organisms. This makes metabarcoding especially useful for high throughput detection of insects including those living in inaccessible niches or outside of the daily or annual activity period.

Similar to the insects, the frequency of detection of the third most numerous group, i.e., the Annelida, across salinity zones was also U-shaped with lower numbers of annelid incidence in the mid salinity zone compared to low and high salinity zones. The annelid species M. southerni and D. aestuarina were never detected in the high salinity zone samples and are likely low salinity bioindicators. The organisms that these OTUs represent may experience range expansions associated with decreased salinity resulting from the future Mid Barataria freshwater diversion. The polychaete M. aestuarina, which is commonly represented among both macrofauna and meiofauna in marshes of the Southeastern U.S. (Bell, 1979; Kneib, 1984) and is a well-known bioindicator species for the impact of oil spills (Fleeger et al., 2015; Fleeger et al., 2018) was detected in the high salinity zone but in rare incidences. Polychaetes typically show the highest abundance in the tidal flats with higher soil salinity and moisture than the low marsh where oligochaetes have their peak abundance (MacKenzie et al., 2015; van Regteren et al., 2020). The limited detection of Manayunkia is likely due to the processing of the samples, since this species was among the most frequent OTUs in unprocessed soil at three marsh islands in Barataria Bay surrounding Bay Jimmy while in samples sieved for meiofauna it belonged to the infrequent group (Rayle, 2021).

The Platyhelminthes were the fourth most commonly detected and numerous group of OTUs. The Platyhelminthes are not commonly detected or reported in traditional studies of meiofauna (Fleeger et al., 2018), presumably since traditional sampling and extraction methodology destroys the soft bodied members of the meiofauna such as the Platyhelminthes (Carugati et al., 2015). However, they are abundantly represented in environmental DNA studies (Fonseca et al., 2010). Detection frequency of platyhelminths decreased with increasing salinity. Two Platyhelminthes (O. truncula and R. simplex) were never detected in the high salinity zone and, thus, are likely indicators of reduced salinity, which might be able to expand into the areas currently designated as high salinity when the Mid-Barataria freshwater diversion opens. In contrast, two members of the genus Macrostomum were never detected in the low salinity zone and are, thus, potential candidates for indicators of saltwater intrusion. The remarkable ability of saline acclimation of intertidal infauna has been studied in the model M. lignano, which was the most abundant flatworm OTU in our study. Macrostomum was able to survive in high salinity through upregulation of antioxidant enzymes to combat oxidative stress from increased respiration rates to meet the energy demand for osmoregulation (Rivera-Ingraham et al., 2016). In low salinity M. lignano decreased respiration and activity in a hyposmotically induced metabolic shutdown similar to that in other intertidal invertebrates like the marsh periwinkle snail.

The Gastrotricha were the fifth most numerous but the seventh most commonly detected group and were more prevalent in the low salinity zone than in the other two salinity zones. Specifically, the OTU assigned as C. novenarius can be considered as a low salinity bioindicator which might experience a range expansion with freshwater diversions. The Mollusca were the sixth most numerous group of OTUs and the fifth most commonly detected. Most of the conspicuous mollusk species in southeastern US saltmarshes are large and are only represented by traces of their environmental DNA in our data (Kneib, 1984). Most mollusk OTUs were roughly evenly distributed across salinity zones. Kinorhyncha (Echinoderes sp.) were never detected in the low salinity zone and were most prevalent in the high salinity zone. Kinorhynchs of the genus Echinoderes are believed to have special osmoregulatory adaptations provided by an enlarged nephridial sieve plate that enables them to tolerate salinity fluctuations and hypersaline conditions (Zeppilli et al., 2018). Therefore, increasing kinorhynch populations may be an early indicator of high salinity through saltwater intrusion.

This metabarcoding study provided a comparison of compositional inventories and diversity of soil organisms among salinity zones in Louisiana tidal marshes based on sequence data from metazoan meiofauna and environmental DNA. The advantage of the use of metabarcoding over traditional surveys by collection is that taxa can be monitored based on their DNA without the need for having intact organisms, keys and expertise for time-consuming morphologically identification. This makes metabarcoding especially useful for high throughput detection of diversity including taxa that are not well described or difficult to collect because they are small, fragile, cryptic or inactive during the collection period (Gielings et al., 2021). However, the disadvantage is that it is unknown how long environmental DNA can be detected in the absence of living organisms (Harrison et al., 2019; Sakata et al., 2020), which might create background noise for monitoring current state and functional changes in ecosystems. In addition, number of sequence reads correlates only weakly with organism numbers or biomass (Egge et al., 2013) in multicellular organisms, so population abundance is more accurately determined by collections of organisms via traditional methods. Ideally, metabarcoding would not replace traditional sampling methods but rather add a complementary view of the marsh soil community with increased sampling depth and diversity detection. In summary, using environmental DNA we found numerous meiofaunal taxa in Louisiana salt marshes that are potentially limited in distribution by specific salinity conditions. These data are not only useful by providing potential indicator taxa for salinity changes from future events such as freshwater releases or sea-level rise, but also for predicting what meiofaunal communities may look like before these changes occur.