Introduction

Organic carbon (OC), ubiquitous in wetland soils, is important for food web dynamics (rapid carbon dynamics) and carbon sequestration (long-term carbon dynamics). Labile OC serves as the base of the food web, providing nutrients and energy to higher tropic levels (Kwak and Zedler 1997). Additionally, the waterlogged conditions and rapid accumulation of sediments can allow OC, especially recalcitrant carbon, to be buried and stored for significant time periods (Chmura et al. 2003, Dodla et al. 2012, Hopkinson et al. 2012). As a result, coastal salt marshes store up to 1700 g/m2/yr of OC, making them one of the most carbon-rich environments on Earth (Mcleod et al. 2011). Half of all marine carbon burial occurs in wetlands, even though wetlands occupy only 0.2% of the area available for marine carbon burial (Duarte et al. 2013). Due to the high amount of stored carbon, coastal marshes are considered blue carbon ecosystems leading to intense study of marsh carbon burial rates over the past several decades (Chmura et al. 2003, Duarte et al. 2005, McLeod et al. 2011, Ouyang and Lee 2014).

Most of the carbon found in salt marsh soils has been attributed to plants (macrophytes) (Chmura et al. 2003, Ouyang and Lee 2014). Belowground biomass, in the form of roots and rhizomes, contributes OC directly to sediments, while above-ground biomass can decay on the surface, be exported by tides, or be buried. Although salt marsh plants are probably the main contributor to this carbon pool, algae may be a significant source of OC in salt marsh sediments. Indeed, stable carbon isotopes values of marsh sediments have indicated that a major source of carbon in the sediment may be from planktonic or benthic photosynthetic microorganisms (Middelburg et al. 1997). Furthermore, stable isotopes have identified biofilms as an important source of carbon in salt marsh food webs (Galvan et al. 2008; Nelson et al. 2019; Johnson et al. 2019; Lee et al. 2021). Microphytobenthos or biofilms, have been suggested to be a major contributor to the carbon storage in marsh systems (Connor et al. 2001). Additionally, while marsh productivity is often driven by plants, gross primary production by biofilms can be similar to that of plants. For example, Zedler (1980) found that biofilm net primary production was 0.8 to 1.4 times the aboveground production, while Gallagher and Daiber (1974) found that algal production beneath salt marsh vegetation was ~1/3 of the net production by the plants.

Benthic photosynthetic biofilms, composed of algae, bacteria and fungi, are typically found as patchy mats on marsh surfaces and intertidal zones worldwide (Decho 2000). In this study, we focus on biofilms primarily composed of diatoms and their extracellular polymeric substances (EPS). Living diatom-based biofilms, because of their light requirements, are limited to the top several millimeters of the sediment surface, but have been shown to have some vertical motility (MacIntyre et al. 1996, Kingston 1999). The movement of diatoms in the sediment in response to tidal fluctuations (Paterson 1989 and light availability (Perkins et al. 2001) is facilitated by the secretion of EPS; in order to move vertically to optimize abiotic conditions, the diatoms are propelled by secretions of EPS (Edgar and Pickett-Heaps 1983), creating a network of EPS in the sediment matrix (Smith and Underwood 1998).

The net primary production of biofilms may be greater than 90% of their gross primary production (Pomeroy 1959), suggesting that most of the carbon biofilms created is not respired, but instead is available for decomposition, transfer to other trophic levels, or burial. Although the organic material produced by biofilms, particularly the EPS, is relatively labile compared to marsh plants (McKew et al. 2011), the sheer volume of carbon produced by the rapid turnover rate of these microorganisms may contribute significantly to the marsh sediment carbon pool. In marshes, biofilms are either decomposed by heterotrophic bacteria, buried, resuspended, or consumed by other organisms (Middelburg et al. 2000). Furthermore, biofilms can be a CO2 sink on the sediment surface, suggesting that they can accumulate carbon (Chen S. et al. 2019). Biofilms that are rapidly buried may decompose slower in an anaerobic environment than at the surface, allowing greater carbon preservation.

Biofilms exist in a delicate balance with sediment deposition. If sedimentation rates are too low, biofilms will be exposed to oxic conditions, resulting in more rapid decomposition and less burial of carbon. On the other hand, if sediment deposition rates are too high, biofilms may be buried, unable to reach the surface and photosynthesize, fix carbon, and reproduce (Miller et al. 1996, Jesus et al. 2009, Pivato et al. 2019). The existence of a maximum sedimentation threshold for biofilm survival has been postulated even within the context of stromatolite growth (Grotzinger and Knoll 1999), but it has never been tested experimentally.

Here we hypothesize that at some intermediate sediment deposition rate, the burial of biofilm OC is maximized. The purpose of this study is twofold. First, we test whether benthic biofilms can accumulate carbon in muddy sediments under favorable conditions (light exposure, nutrient supply, and in the absence of grazing). Second, we test how sedimentation rate affects the rate of biofilm carbon accumulation.

Methods

Laboratory Set Up

A homogenized bentonite-mud slurry (125 g/L bentonite, bentonite powder (Natures Oil, Wyoming USA) 35 psu Instant Ocean seawater (Spectrum Brands, Inc., Blacksburg, VA, USA)) was poured into plastic cylinders (height = 20 cm, diameter = 9.5 cm; Fig. 1). The cylinders were placed on orbital shakers (orbital diameter = 0.5 cm, 100 RPM) and allowed to settle to create a sediment bed ~10 cm thick with an overlying water column of ~10 cm. The water column was then changed weekly using a peristaltic pump to avoid disturbing the bed surface. The replacement medium was a solution of deionized water, Instant Ocean salts (to achieve a salinity of 35 psu), and a diluted f/2 medium (Bigelow Laboratory), which provided the necessary nitrogen (10 μm, same order of magnitude as world rivers (Sprague et al. 2011)), phosphorus, silica, vitamins and trace metals for growth (N:P:Si = 24.4:1:2.9). Each cylinder was inoculated with a sample of biofilm scraped from the surface of a salt marsh in Cocodrie, Louisiana (USA). Prior to inoculation, the biofilm was examined under a microscope and qualitatively determined to be diatom-dominated. Once inoculated, the cylinders were exposed to a 12-hour light/dark cycle using grow lights (Agrobrite, 120 V, 60 Hz high output fluorescent lighting system, 6400 °K, providing ~2000 lx at the bed surface or ~46 µmol/s/m2). The sides of the containers were covered with dark paper to ensure light came only from the provided source. Control containers did not receive the inoculum, were treated with 150 uL of bleach, and kept in the dark to prevent biofilm growth. The cylinders were kept on the orbital shaker, which provided a gentle agitation and promoted vertical mixing of the water column. The biofilms were allowed to establish for two weeks without any disturbance.

Fig. 1
figure 1

Plastic cylinders used for the experiment after 14 weeks of the experiment. (A) Side view showing the vertical accumulation of sediments. The parallel layers in the sediment, starting at about half of the sediment column, are from individual sedimentation events and subsequent growth of biofilm. (B) Plane-view of growth experiment. Light brown color is indicative of diatom-based biofilm

Full size image

The sedimentation experiment began after the observed colonization of the sediment surface by biofilms (two weeks of growth). A slurry of bentonite clay mixed with the medium was added according to five sedimentation rates (Table 1), ranging from 12 to 189 mm/yr. These rates represent very high mineral deposition rates compared to field measurements and represent areas such as newly-forming deltas (Shields et al. 2017). Following two weeks of growth, there were 11 weeks of sedimentation events, and sampling occurred one week after the final sedimentation event.

Table 1 List of treatments, or sedimentation rates used in this experiment
Full size table

The benefit of using the bentonite powder is that it is entirely abiotic and does not contain carbon; any carbon that we measured was produced in situ. There were two replicates of each treatment except the low sedimentation rate, in which there was only one container. Additionally, there was one control (no biofilm) for each of the sedimentation rates.

Sampling and Analyses

Biofilm growth was monitored using a pulse-amplitude modulation (PAM) fluorometer throughout the duration of the experiment every other day. PAM fluorescence values have been used as a proxy for chl a and biomass of biofilms in previous studies (Honeywill et al. 2002, Jesus et al. 2005, Murphy et al. 2009, Orvain et al. 2014), and has the advantage of being not destructive. Every other day, a grid containing 13 equi-spaced points was placed over each container and PAM fluorescence was measured at each point. The fluorescence were proxies for relative growth within the experiment, but not for actual biomass. Bed heights (difference in bed height from the beginning of the experiment i.e., accumulated sediment and biofilm height) were also measured and recorded throughout the experiment.

One week following the last sedimentation event, the sediment in each cylinder was analyzed to calculate the total amount of organic matter (OM), OC, and chl a accumulated throughout the sediment column. Operationally, these measurements were performed by separating the top six centimeters of the sediment column – which encompassed the whole layer in which biofilm grew – into two layers (0-3 and 3-6 cm).

Each layer was then homogenized and subsampled for bulk density and water content, chl a analysis (EPA Method 445.0), loss on ignition (LOI), and total organic carbon (TOC) (Ramnarine et al. 2011). Bulk density, the mass of sediment divided by the volume (g/m3), is a measure of soil density and was used for calculating accretion rates and translating between a carbon content and a carbon density. Three separate methods were used to identify the amount of organic material from the biofilms: chl a, LOI and TOC. Each method measures a different parameter related to biofilm growth (chlorophyll content, organic content, and organic carbon content, respectively). For LOI analysis, which determines the amount of OM via combustion of the sample, the samples were burned at 400 °C for 8 h (Dean 1974; Hoogsteen et al. 2015). As bentonite clay has high structural water content (Hoogsteen et al. 2015) and our samples had relatively low amounts of OM, the mass lost in the control samples (no biofilm) was subtracted from all samples to account for the loss of this structural water during the LOI procedure. For chl a analysis, the samples were extracted using acetone and the fluorescence was measured at 485 nm wavelength (Turner Designs TD-700 Fluorometer). To determine TOC, samples were fumigated to remove carbonates, dried, and analyzed using a Costech 1040 CHNOS Elemental Combustion system (Harris et al. 2001).

The total amount of chl a, OM, and TOC in each layer was then summed together and divided by the duration of the experiment and the surface area, thus obtaining accumulation rates per unit of area. For all treatments except the highest accretion rate, all accumulated sediment was contained in the top layer (0-3 cm). For the treatment with the highest accretion rate, which accreted a total of 4.3 cm during the experiment, the accumulated material was in both the top (0-3 cm) and bottom (3-6 cm) layers. Therefore, for this treatment it was possible to compare the carbon density per unit volume for the two layers. For this comparison we corrected the measurements in the bottom layer for the dilution effect (i.e., 1.3 cm of the bottom layer were from the accumulated sediment, while 1.7 cm of the bottom layer were from the underlying bentonite bed and thus diluted the carbon density).

Control containers were monitored to ensure no growth, but PAM measurements were limited to prevent cross-contamination. No growth or color changes were observed in the control containers throughout the experiment. TOC and chl a were not measured in the control containers.

The OM and OC data were fit according to the form:

$${C}_{acc}={C}_{max}\left(1-\text{exp}\left(-\frac{D}{a}\right)\right)$$
(1)

where \({C}_{acc}\) is the accumulation rate of OM or OC, \({C}_{max}\) is the maximum rate of accumulation of OM or carbon mediated by sediment deposition, \(D\) is the deposition rate, and \(a\) is a fitting parameter.

Results

PAM Fluorescence and Vertical Accretion

Fluorescence values increased approximately two weeks following inoculation in all treatments (Fig. 2). While there was substantial variation in fluorescence over time, the general trends across all treatments (increase in fluorescence two weeks after inoculation, with average values of ~150 for all treatments) indicate that biofilm growth in each treatment was similar. The change in bed height compared to initial bed height in each container demonstrated that the addition of bentonite increased the height of the sediment column and the rate of height increase depended on the amount of sediment added (S1). The vertical accretion rates ranged from 4 mm to 45 mm, for the lowest and highest mineral sedimentation rate respectively over the duration of the experiment. Following each sediment addition, there was an initial increase in bed height and then a slight decrease due to the deposition and compaction of the added sediment.

Fig. 2
figure 2

Monitoring of fluorescence (unitless) over the 14-week experiment. Colors represent the five different sedimentation rates (see Table 1). Fluorescence measurements are consistent across treatments. Vertical lines indicate when sediment was added to the experiment

Full size image

Chl a

Sediment chl a accumulation rate increased with increasing vertical accretion (Fig. 3A). The containers with the lowest vertical accretion had an average chl a accretion rate of 123 mg/cm2/yr and the containers with the highest vertical accretion rate had an average chl a accretion rate of 534 mg/cm2/yr. For the highest sedimentation rate, chl a values were similar between the top (0-3 cm) and bottom (3-6 cm) sediment layers (Table 2).

Fig. 3
figure 3

Chl a (A), OM (LOI) (B), and OC (TOC) (C) values for the content at the end of the experiment for each of the five sedimentation rates. All three metrics show an increase with equivalent vertical accretion rate. Lines in (B) and (C) show best fit to Eq. 1. (B) R2 = 0.85 and (C) R2 = 0.76. Measurement error is in the Supplemental Information (S2)

Full size image
Table 2 Chl a, OM (LOI), and OC (TOC) values for the top (0-3 cm) and bottom (3-6 cm) layers of the samples with the highest accretion rate
Full size table

OM and OC

As sedimentation rate increased, more OM was stored in the sediment (Fig. 3B). The average rate of OM accretion for the highest vertical accretion rate was 456 g/m2/yr, which is over five times the average rate of OM accretion measured in the containers with the lowest vertical accretion rates (86 g/m2/yr). The sedimentation rate was 16 times higher in the treatment with the highest vertical accretion rate compared to the lowest. Similarly, the rate of carbon accretion increased with increasing rates of vertical accretion (Fig. 3C). The containers with the lowest accretion rates accumulated 31 g/m2/yr C, while those with the highest accretion rate accumulated 211 g/m2/yr C. OC was similar between the top (0-3 cm) and bottom (3-6 cm) layers, while OM was higher in the top layer compared to the bottom layer (Table 2).

We fit the exponential model to the OM (LOI) and OC (TOC) datasets (Eq. 1, Fig. 3) with the assumption that there is little to no accumulation of OM or C from biofilms without sediment deposition, as without burial the labile OM from biofilms will decompose and will not to contribute to OM/OC accumulation. As accumulation rates increase, the rates of C production increase decreases (Fig. 3B, C). For OM, we found that the maximum rate of OM accumulation, \({C}_{max}\), was 534 g/m2/yr. In terms of OC, \({C}_{max}\) was determined to be 201 g/m2/yr C.

Discussion

The Potential for Biofilm Carbon Accumulation

The carbon accretion rates (CAR) from this study are comparable with those observed in marshes worldwide. We found rates of 100-200 g/m2/yr C with moderate to high accretion rates, while worldwide rates for marshes range from 100 to 300 g/m2/yr C, depending on the latitude and vegetation type, amongst other variables (Ouyang and Lee 2014). Our results demonstrate that under favorable conditions (light, nutrients, no grazing or competition), biofilms have the potential to produce soil carbon at the same order of magnitude of what is observed in marshes worldwide.

Previous experiments have shown that much of the carbon from biofilms is in the form of extra-polymeric substances (EPS), and that this material is rapidly degraded (Guarini et al. 2000, de Brouwer and Stal 2001). These experiments looked at the surface biofilm and the associated carbon, and not at the biofilm carbon with time or depth. However, the similar values of chl a and TOC with depth (Table 2) indicate that the biofilm material is present not only on the sediment surface but also throughout the sediment, giving rise to the potential for carbon storage. Our experiment did not show the ability to store carbon over decadal to centennial time scales due to logistical restraints. Yet, recent studies (Unger et al. 2016) showed that even labile carbon can be stored at depth and for greater than 50 years in marsh sediment, enhanced by high sedimentation rates. The combination of decreased decomposition of labile carbon and the increased OC production by biofilms with high sedimentation rates indicate that biofilms may substantially contribute to marsh carbon storage.

Sedimentation Rate Increases Carbon Accumulation

Our experiment clearly shows that the rate of chl a and carbon accumulation increases with the rate of sedimentation. A possible explanation for this trend is that sedimentation stimulates biofilm production by providing additional nutrients. However, this hypothesis is not likely given the abundance of nutrients in the water column; none of these experiments were nutrient limited and therefore a small increase in nutrients from the addition of bentonite should not have increased carbon production significantly.

Another explanation for the increase in OC accumulation with sedimentation rate is that sedimentation could provide additional space (volume) that the biofilms are able to fill as they grow upward towards the light source. Sedimentation necessitates vertical movement by the photosynthetic organisms, and thus causes an increase in OM production (Pinckney and Zingmark 1993). Diatoms have been shown to migrate in sediments in short time frames, largely as a response to light (Paterson 1989, Underwood and Kromkamp 1999). As a mechanism of migration, diatoms use their organic secretions (EPS) to aid in their vertical movement (Underwood et al. 1995, Smith and Underwood 1998). With higher sedimentation rates, the diatoms need to migrate further and therefore secrete more organic material. Furthermore, as diatoms migrate, dead cells remain scattered through the sediment (Debenay et al. 2007); with increased sedimentation and increased migration, the amount of carbon from dead cells would also increase. Ultimately, the more volume of sediment present for biofilms to grow upon leads to higher amounts of OM production by the biofilms.

Furthermore, sedimentation may affect the “age” of the biofilm, and therefore change its metabolism. The physiological state of biofilm changes over time (Sutherland et al. 1998), with lower rates of photosynthesis (Serodio et al. 2005) and higher EPS production for more mature biofilms (Orvain et al. 2003). We find that early in the experiment (weeks 2.5-7), fluorescence measurements (i.e. rates of photosynthesis) has no statistical relationship with sedimentation rates, but late in the experiment (weeks 7-14), fluorescence values are linearly related to sedimentation rate (Fig. 4). In fact, at low sedimentation rates, fluorescence values are lower during the later stage of the experiment, supporting the hypothesis of decreased rates of photosynthesis with time (Serodio et al. 2005). Conversely, with high sedimentation rates, fluorescence rates remain high. Our results suggest that sedimentation may constantly “reset” the biofilm age and allow it to grow as in the early stage of development, allowing for the production of more carbon and increased carbon in the sediments.

Fig. 4
figure 4

Average PAM fluorescence value by sedimentation rate in each container (replicates for all except the lowest accretion rate) for weeks 2.5-7 (A) and 7-14 (B). There was no statistically significant relationship between fluorescence and sedimentation rate in the beginning of the experiments, but in weeks 7-14, there was a significant linear relationship (y=0.41x+98, R2 = 0.58, p=0.018)

Full size image

High rates of carbon accumulation have been previously related to high mineral suspended sediment supply (Connor et al. 2001), and therefore increased marsh accretion rates (Kirwan and Megonigal 2013). While Connor and others (Connor et al. 2001) are reporting CARs from all carbon sources, they suggest that at low elevations, where sediment accretion rates are higher, biofilms may be a factor influencing carbon accumulation. We find in our experiments that OM from biofilms agree with the relationship between high suspended sediment, high sedimentation rates, and high rates of OC burial.

The limited number of replicates (two replicates per treatment, except one replica for the lowest accretion rate) make it difficult to perform statistical analyses, for example comparing the results of two different accretion rates. However, we obtained a robust result by considering the results as a continuous function of the accretion rate (thus considering nine datapoints, Fig. 3). All regression analyses suggest that higher sedimentation rates do allow for more biofilm growth (chl a), and more OC, and OM accumulation.

Limits on C Accumulation by Biofilms

The consistent trend in all metrics of biofilm growth (chl a, LOI and TOC) confirm that biofilm grown under favorable conditions can maintain itself and even thrive under sedimentation rates nearly 16 times the natural rate along the Gulf Coast (Cahoon 2010). Although our results suggest that a constant level of OC accumulation can be reached for arbitrary high sedimentation rates, this is likely not the case. We expect that there is a sedimentation maximum which the biofilms would not be able to recover from (Grotzinger and Knoll 1999), thus limiting its ability to accumulate carbon. Ultimately, at some deposition rate, the biofilms would not be able to reach the sediment surface, or not be able to colonize, grow and reproduce quickly enough on the surface to contribute to carbon accumulation. At very high sedimentation rates, OM and carbon accretion rates would likely decline quickly as less of the biofilm is able to reestablish on the sediment surface. This could also be the case depending on how sediment was delivered to the marsh; for the same annual deposition rate the sediment deposition could be gradual (i.e. press disturbance) or delivered from large storm events (i.e. pulse disturbance) (Turner et al. 2006). It is possible that biofilms may not be able to recover from pulse deposition events (e.g., several centimeters of deposition within hours).

An unexpected result of this experiment was that the biofilms were incredibly resilient and able to grow despite large sedimentation rates (applied as a press disturbance, i.e., at weekly intervals). Following each sedimentation event, the biofilms colonized the new sediment-water interface very quickly, within 24-48 h. Indeed, PAM fluorescence (Fig. 2) did not decrease following the sedimentation events, even though these measurements were taken 24-48 h following such an event. The mineral sedimentation rates tested in this experiment exceed most sedimentation rates for coastlines worldwide and were done episodically. As the biofilms were able to grow in these extreme conditions, biofilms in nature would likely be able to withstand normal sedimentation.

Consequences for Natural Systems

The importance of increased sedimentation rates on the productivity of salt marsh biofilm is particularly relevant for coastal restoration projects. Some methods of marsh restoration projects, including sediment diversions (e.g.: Elsey-Quirk et al. 2019) and thin-layer sediment deposition (e.g.: Ford et al. 1999), involve the introduction of high rates of sedimentation to marshes. For example, in a restored marsh in the Bay of Fundy, high sedimentation rates and high carbon accumulation rates were measured prior to the establishment of marsh vegetation (Wollenberg et al. 2018). Wollenberg and others (Wollenberg et al. 2018) suggested that the high C accumulation prior to vegetation is allochthonous. However, given the results of our experiment, biofilm productivity could explain high rates of carbon accumulation prior to the establishment of marsh vegetation.

While in this study, we focused on the role of biofilm OM in salt marsh sediments, biofilms can also be an important source of carbon in tidal flats. There are substantial data gaps in our understanding of how much carbon is stored in tidal flats (Lovelock and Reef 2020), and it is possible that these systems may play a large role in coastal carbon storage (Lovelock and Duarte 2019). As there is no vascular vegetation, the primary autochthonous carbon in tidal flats is biofilms. Thus, quantifying the amount of carbon in tidal flats from biofilms will improve our understanding of this potential carbon sink.

Future Directions

Future studies should improve the ability to individuate the source of the carbon in marsh sediments (Macreadie et al. 2019). This could help to quantify the impact of biofilms in terms of OC in nature and reconcile our laboratory results with field results. A combination of approaches, including stable isotopes (Choi et al. 2001, Gebrehiwet et al. 2008, Galvan et al. 2008, Tanner et al. 2010), organic biomarkers (Spohn and Giani 2012, Johnson et al. 2019), and environmental DNA (Reef et al. 2017) will yield a better understanding of the source of carbon in marsh sediments (Geraldi et al. 2019). For example, studies that have used an increased suite of isotopic signatures were more successful in identifying biofilms (Moncreiff and Sullivan 2001, Hondula and Pace 2014, Duarte et al. 2018). In the field of ecology, stable isotopes have been a useful tool in identifying the relative contributions of carbon sources in food webs (Nelson et al. 2019); a similar approach could be used to differentiate carbon sources in marsh sediment. These tools have been primarily used to map out food webs, but expanding their use to identify carbon sources can help quantify the contribution of biofilms to salt marsh carbon in the field.

Furthermore, there is a need to conduct more laboratory experiments including additional factors, such as grazing. Biofilms are an important component of the diet of grazing macrofauna in coastal ecosystems (Daggers et al. 2020). However, while we demonstrate that high sedimentation promotes biofilm carbon accumulation, little work has been done on how sedimentation rate affects grazers. In sediment-addition restoration projects, snail growth rates were highest with intermediate sediment addition (Stagg and Mendelssohn 2012). It is unclear whether the higher sedimentation rates will allow more of the biofilms to be buried and protected from grazing, or if bioturbation could increase and overall grazing may increase. The strength and direction of this feedback will impact how much biofilm carbon is able to be stored in salt marsh sediments in real settings.

Another important aspect to investigate is the fate of resuspended biofilms. Previous studies have focused on the transfer of biofilm OM to consumers in the water column and adjacent habitats from consumers (Carlton and Hodder 2003) or resuspension events (Ubertini et al. 2012; Savelli et al. 2019). While it is clear that biofilm resuspension dynamics are important, the ultimate fate of the resuspended biofilm carbon is not well understood. Much of the resuspended biofilm OM is likely consumed or decomposed, but some of the biofilm may be redeposited and subsequently buried and stored in the sediments. For example, recent flume experiments (Chen X. et al. 2019) found that resuspended biofilms allowed for faster biofilms recovery and suggested that repeated erosion redistributed surface biofilms deeper in the bed. They argued that this is important for sediment stabilization, but we posit that it would also be important for carbon storage.

Conclusions

Benthic biofilms in coastal environments are resilient and able to flourish under high sedimentation rates, given ample nutrients and light. These experiments clearly demonstrate that biofilms have the potential to contribute to carbon accumulation in salt marsh sediments. Based on the results presented here, biofilms have the potential to accumulate as much carbon in soils as what is typically measured in salt marshes. While this carbon is labile and may not be stored on a centennial to millennial timescale, it likely plays an important role in the carbon cycle in the marsh.

All analyses validate our hypothesis that higher sedimentation rates increase biofilm carbon accumulation. A sedimentation threshold above which biofilms cease to grow and to accumulate carbon might still exist, but it would be relatively high (i.e., >200 mm/yr). The results of this experiment represent the upper bounds of OC accumulation by biofilms, as they were grown under favorable conditions over a short timescale. Further experiments should quantify the role of grazing in limiting biofilm carbon accumulation, and how this effect changes as a function of the sedimentation rate.