Direct contribution of invertebrate holobionts to methane release from coastal sediments

Sediment macrofauna play a vital role in sustaining aquatic food webs and biogeochemical cycles. Previous research demonstrated that bioturbation indirectly affects methane (CH4) dynamics through mobilization of porewater and alteration of microbial processes in the surrounding sediment. However, little is known on the direct contribution of macrofauna holobionts (the assemblage of invertebrate host and associated microbiome) to biogeochemical fluxes. Here, we investigated how 19 taxa of macrofauna holobionts, from different estuarine habitats spanning 40° to 63° latitude, directly contribute to CH4 fluxes. Deep burrowing infauna and deposit feeders were responsible for the highest CH4 production, whereas epifauna and filter feeders promoted oxidative CH4 consumption. Among the different environmental parameters, salinity was inversely correlated with CH4 production by macrofauna holobionts, with the process suppressed at high salinity (≥ 33). This study provides empirical evidence on how functional traits and environmental factors influence sediment invertebrates' contribution to CH4 fluxes.

oxidative CH 4 consumption.Among the different environmental parameters, salinity was inversely correlated with CH 4 production by macrofauna holobionts, with the process suppressed at high salinity (≥ 33).This study provides empirical evidence on how functional traits and environmental factors influence sediment invertebrates' contribution to CH 4 fluxes.
Macrofaunal activity profoundly impacts biogeochemical processes and microbial diversity through sediment reworking during the process of feeding or locomotion, that is, bioturbation (Kristensen and Kostka 2005;Laverock et al. 2010), or via biodeposition and nutrient excretion (Vanni and McIntyre 2016).In addition, benthic invertebrates represent distinct geochemical niches inhabited by specific microbial communities with variable degrees of fidelity to their animal host.The assemblage of the invertebrate host and its associated microbial communities is referred to as the holobiont (Bordenstein and Theis 2015;Dittami et al. 2021).Recent studies show that holobionts promote biogeochemical processes such as methane (CH 4 ) production, dinitrogen fixation, and nitrate reduction (Bonaglia et al. 2017;Zilius et al. 2020;Marzocchi et al. 2021).
In shallow and productive lagoons and estuaries, the release of reduced compounds as CH 4 is mostly driven by the benthic compartment (Canfield et al. 2005;Reeburgh 2007;Roth et al. 2022).CH 4 emissions from coastal aquatic ecosystems to the atmosphere are critical in global CH 4 budgeting, especially from anthropogenically impacted systems (Zheng et al. 2022).Estuaries and lagoons emit 0.29-0.63Tg CH 4 yr À1 (Borges and Abril 2011;Rosentreter et al. 2021).Recent estimates suggest that global estuaries emit 0.25 Tg CH 4 yr À1 as median (Rosentreter et al. 2023).
Until recently, biogenic CH 4 production in marine systems was exclusively attributed to methanogenic Archaea under anoxic conditions in sediments and bottom waters (Reeburgh 2007).CH 4 production in oxic marine waters has generally been credited to methanogenesis occurring in the digestive tracts of organisms such as macrozooplankton (Reeburgh 2007) or, recently, to methylphosphonate decomposition in phosphate-stressed oceanic waters (Karl et al. 2008;Repeta et al. 2016).Benthic invertebrates and their microbiome can also mediate CH 4 production, as they ingest large quantities of particulate organic carbon and have anoxic digestive tracts (Stief and Eller 2006;Stief et al. 2009;Bonaglia et al. 2017).However, the actual contribution that macrofauna holobionts from different functional groups have on CH 4 fluxes is currently unknown.The environmental conditions that control CH 4 production in these holobionts are also unresolved.
Macrofauna are often grouped based on feeding guilds (e.g., filter-feeders, grazers, and predators; Welsh 2003) or/and their bioturbation mode in sediments (i.e., surface-and deep-burrowers, sediment mixers; Mermillod-Blondin and Rosenberg 2006).Among different invertebrate functional groups, surface and deep burrowers generally inhibit the production and release of reduced substances in the upper sediment layer by stimulating oxidative processes (Kristensen and Kostka 2005).However, it has been recently shown that opportunistic deep burrowers can strongly stimulate anaerobic processes such as nitrate ammonification (Bonaglia et al. 2013;Benelli et al. 2019) and methanogenesis (Bonaglia et al. 2017), and promote reduced substance release (Kauppi et al. 2018).Epifauna, such as small sediment remixers crustaceans, generally favor oxidation processes by enhancing oxygen (O 2 ) penetration depth (Welsh 2003;Bonaglia et al. 2019).
A few studies have addressed the impact of worm and bivalve bioturbation on CH 4 release.Nogaro and Burgin (2014) showed that CH 4 release does not significantly increase along with tubificid-oligochaete abundance in organic-rich estuarine sediments.Opposite is the situation of urban wetlands, where CH 4 fluxes are strongly correlated with tubificid abundance (Mehring et al. 2017).Experimental work suggested that Baltic Sea bivalves may induce a seven-to ten-fold increase of CH 4 efflux compared to sediment without macrofauna (Bonaglia et al. 2017).To our knowledge, no systematic studies have been conducted to investigate the direct CH 4 production by benthic invertebrates and to quantify their impact on benthic CH 4 fluxes.
In this study, disparate benthic invertebrate species belonging to different functional groups (e.g., deep and surface burrowers, surface sediment remixers, and epifauna) were incubated to quantify their holobiont-associated CH 4 production.Our goals were to: 1. Quantify CH 4 production in macrofauna holobionts.2. Explore if patterns in CH 4 production are explained by macrofauna functional groups.3. Establish relationships between CH 4 production and environmental factors (e.g., salinity).4. Assess the relative contribution of macrofauna holobionts to CH 4 fluxes in estuarine systems.

Sampling sites and experimental approach
The study was conducted in four coastal shallow systems (Curonian Lagoon, Sacca di Goro Lagoon, Oristano lagoons, and Öre estuary), characterized by contrasting salinities and situated along a latitudinal gradient ranging from 40 to 63 N (Fig. 1; Table S1).In each system, the dominant sediment invertebrate species were selected and incubated to determine CH 4 production rates.O 2 respiration and NH 4 + excretion were also measured.We report these rates as both individual rates and mass-standardized rates sensu Maciute et al. (2021).Overall, individual and mass-standardized CH 4 production rates (IPR and MPR), O 2 respiration rates (IRR and MRR, respectively), and NH 4 + excretion rates (IER and MER), were quantified in 103 measurements, for a total of 262 incubated individuals belonging to 19 macrofaunal taxa (Table S2).

Individual animal incubation setup
Invertebrate of similar weight classes were collected by hand or box-corer and immediately after collection, organisms were placed in a climate-controlled room at in situ temperature and identified to the lowest possible taxonomic level.Individuals were incubated in 22 mL glass microcosms filled with 0.22 μm twice-filtered in situ water without headspace.The present experimental approach uses filtered water to target microbial activity in the host digestive system while eliminating the effect of ingestion process.Glass microcosms were equipped with optode sensor spots for continuous monitoring of O 2 (FireStingO2, PyroScience GmbH).The incubations were carried out in up to six replicate microcosms with animals and in two controls only with filtered water.For small animals where low respiratory rates were expected, microcosms received more than one individual (up to 7 ind.per microcosm).Assuming linearity in animals CH 4 production (Bonaglia et al. 2017), we used a start/end incubation approach.Incubations were carried out in the dark, at constant temperature, with a preincubation period of 1 h, and lasted from 0.8 h (large animals) to 8.5 h (small animals) depending on the O 2 consumption rates.Incubated animals never consumed more than 40% of the initial O 2 availability.Two water aliquots were collected for CH 4 and NH 4 + analysis, one at the beginning (from the batch of filtered water) and one at end of the incubation (from each microcosms).The sample for CH 4 was transferred to 12 mL Exetainers (Labco, UK), allowing overflow and fixed with 100 μL of 7 M ZnCl 2 to stop microbial activity.The sample for NH 4 + was filtered (Frisenette GF/F filters) into a 6-mL PE test tube.Biomass was determined as dry weight (DW) or as shell-free dry weight (DWSF, for bivalves) after desiccation at 70 C until constant mass.All rates were then massnormalized (g DW or g DWSF ).

Chemical analyses
Optode sensors were calibrated at oxygen-free conditions and at 100% air saturation.Dissolved NH 4 + concentrations were determined by a continuous flow analyzer (San++, Skalar) using standard colorimetric methods (Grasshoff et al. 1983).Dissolved CH 4 concentrations were analyzed with gas chromatography (GC).All vials were headspaced with 2 mL pure N 2 , shaken and left to equilibrate overnight.1 mL headspace aliquots were manually injected into a GC 8A (Shimadzu Corp.) equipped with a Porapak N column (80-100 μm mesh) and a flame ionization detector.For calibration, certified standards at atmospheric concentration of 1.9 ppm and at 4.99 ppm CH 4 (AirLiquide Gas AB) were injected for every 10 samples.We used the ideal gas law (PV = nRT) to convert ppm concentrations into molar concentrations (μmol CH 4 L À1 ) and corrected concentrations in the seawater samples by considering the Bunsen solubility coefficient of CH 4 at the specific analytical conditions of temperature and salinity (Wiesenburg and Guinasso 1979).

Data analysis and statistics
Individual respiration rates were calculated from linear regression analysis of the solute (O 2 ) vs. time Eq.1: where IRR (μmol O 2 ind.À1 d À1 ) is the respiration of O 2 ; slope (μmol O 2 L À1 d À1 ) is the slope of the linear regression between O 2 concentration and time; V (L) is the water volume in the microcosm; N is the number of incubated animals per microcosm.S1.
Individual NH 4 + excretion and CH 4 production were calculated from the changes in concentrations (NH 4 + and CH 4 ) over time (2): Equations 1 and 2 were also used to calculate massstandardized rates by replacing N with total animal biomass.Rates measured in control vials were used to correct rates in vials with animals.
Regression models established equations to predict the relationship between individual/mass-standardized rates and predictors (e.g., salinity), with student's t-test quantifying relationship significance.Nonparametric Kruskal-Wallis test was performed to compare between functional and taxonomic groups.The significance level was set at p = 0.05.Statistical analyses were performed in SigmaPlot 14.0.The individual and interactive effects of the factors Feeding guild and Living habit were tested by PERMANOVA (PRIMER 6 + permanova addon) on the resemblance matrix (Euclidean distance) of the MPR rates after Monte Carlo permutation (n = 9999; Table S3).All data and metadata are available in Politi et al. (2023).Rates are reported as average AE standard error.

Macrofaunal metabolism and CH 4 production rates
Individual respiration rates (IRR) ranged between 0.7 and 105.6 μmol O 2 ind.À1 d À1 , whereas individual excretions rates (IER) ranged between 0.2 and 12.8 μmol NH 4 + ind.À1 d À1 (Politi et al. 2023;Fig. 2a,b).Oxygen respiration to NH 4 + excretion ratios varied between 3 and 18 (Table S2).Individual CH 4 production rates (IPR) included also negative data suggesting uptake or oxidation and ranged between À0.15 and 0.59 nmol CH 4 ind.À1 d À1 (Politi et al. 2023).Results from this study align with the widely accepted statement that body mass is one of the main factors controlling animal metabolism (Gerlach et al. 1985;Gillooly et al. 2001).Indeed, we found significant positive correlations between macrofauna biomass and both O 2 respiration and NH 4 + excretion (Fig. 2a,b).In contrast to animal metabolism, both CH 4 IPR and MPR did not correlate with animal biomass (Fig. 2c; Table S4).This is not surprising as CH 4 is produced by microbial metabolism associated with macrofauna hosts (Bonaglia et al., 2017) and such production ).Different colors represent macrofauna with different living habits (= living at different depth in sediment).The equations of the regression lines are reported with the correlation coefficient (R 2 ) and p value.The sample size is lower (64 out of 103 observations) for panel c as negative rates (CH 4 uptake) were excluded.For visualizing all 103 observations refer to the non-logged dataset presented in Fig. S1.
is regulated by different mechanisms (Schmitt-Wagner et al. 2003;Cambon-Bonavita et al. 2021).Also in terrestrial invertebrates, the host biomass is a poor predictor of CH 4 production rates (Crutzen et al. 1986).
Lower CH 4 production by epifauna can be expected as this group generally consists of large and/or mobile invertebrates with multiple oxic surfaces on their bodies, which include shell, carapace, and gills.These oxic surfaces are often colonized by biofilm-forming bacteria sustaining CH 4 oxidation and other oxidative processes (Watsuji et al. 2014;Zilius et al. 2020;Cambon-Bonavita et al. 2021).CH 4 production by macrofauna is related to the methanogenic activity of Archaea associated with macrofauna host (Bonaglia et al. 2017).Differences in CH 4 production may depend upon the abundance and activity of these microorganisms in the host digestive system.Previous studies have demonstrated that the invertebrate gut is anoxic or partially anoxic (i.e., only intestine tracts as midgut and hindgut) (Stief and Eller 2006;Poulsen et al. 2014).Such tracts represent favorable microniches for methanogenic microorganisms and other chemoautotrophs (Aubé et al. 2022).Bonaglia et al. (2017) showed that, in aquatic invertebrates, the copies of marker mcrA gene, encoding CH 4 production in Archaea, were not associated with ingested food, but rather with symbionts colonizing the inner, anoxic tracts of the digestive system.
With respect to putative relationships between invertebrates diet and CH 4 production, Crutzen et al. (1986) reported that many soil insects and other terrestrial invertebrate primary consumers have a great potential role in CH 4 production.Brune (2014) suggested that in terrestrial invertebrates, such as termites, the increased methanogens diversity was likely related to the higher availability of substrate in gut.
Here, the anaerobic food chain involves the degradation of organic compounds, where the end products of one group act as substrates for the subsequent group in the chain.Archaea produce CH 4 utilizing the primary by products of anaerobic organic matter breakdown, which include hydrogen, carbon dioxide, and acetate (Hoppert et al. 2013).Termites, which are consumers of plant detritus as other primary saprotrophs, have truly anoxic digestive tracts (Brune 2014).As such, they harbor methanogenic symbiotic microbes that use the byproducts of anaerobic chemical alkaline hydrolysis of cellulose (Lemke et al. 2003;Schmitt-Wagner et al. 2003)  Average and horizontal standard error bars (n = 4-6 replicas per taxa) are presented as dots for each taxon and as squares for the species grouped by their (a) taxonomic group (Mollusca, Gastropoda, Crustacea, and Anellida/Artropoda); (b) living habits (epifauna, surface remixer/burrower, and deep burrower); (c) feeding guilds (filter feeders, predators, grazers/scrapers, and deposit feeders).
chironomid larvae) feed on refractory organic carbon and on microorganisms often buried in strictly anoxic and chemically reduced sediment horizons, hosting an active community of methanogenic Archaea (Bonaglia et al. 2017).
To our knowledge, only three studies have quantified CH 4 production rates directly in benthic invertebrates from marine environment (Bonaglia et al. 2017;McCarthy et al. 2019;Ray et al. 2019).The Baltic Proper invertebrates Limecola balthica and Marenzelleria arctia were shown to release 2.15 nmol and 0.08 nmol CH 4 ind.À1 d À1 , respectively.Our results do not significantly deviate for M. arctia (0.07 AE 0.07 nmol CH 4 ind.À1 d À1 ), whereas they were substantially lower for L. balthica (0.04 AE 0.02 nmol CH 4 ind.À1 d À1 ), probably because our bivalves from the Gulf of Bothnia were found in surface sediments while those from the Baltic Proper lived buried in deeper sediment layers (down to 5 cm depth) in sulfidic sediments (Bonaglia et al. 2017).Ray et al. (2019) and McCarthy et al. (2019) found that oysters do not significantly release CH 4 , probably due to specific environmental conditions (see next section).

Environmental factors driving CH 4 release by macrofauna
Methanogenesis is the biological production of CH 4 that typically occurs only when more favorable electron acceptors, including sulfate, are depleted.Sulfate is one of the most abundant ions in the marine environment whereas its concentrations are generally low in inland waters; strong concentration gradients are generally found along the freshwater-seawater continuum.Salinity can therefore be used as a proxy for sulfate availability.Under high salinity conditions, sulfate availability outcompetes methanogens for substrate (Reeburgh 2007).We therefore hypothesized that high salinity would limit or even prevent CH 4 production in invertebrate holobionts.We found that increasing salinity significantly reduced (p < 0.01) massstandardized CH 4 production (MPR) (Fig. 4 and Table S4).
Thus, our data suggest that sulfate concentrations contribute to the regulation of the methanogenic activity within invertebrate holobionts.This may explain why Ray et al. (2019) did not report any CH 4 production in aquaculture oysters incubated at high salinities (29-31 range).It is striking that all seven macrofaunal taxa living at high salinities (33-40 range) in Oristano lagoons had no CH 4 production (À0.14 AE 0.30 nmol CH 4 g DW À1 d À1 ).Since these invertebrates also have anoxic niches inside their bodies, facilitating alternative anaerobic processes (Stief et al. 2009), it is likely that sulfate reduction may have outcompeted methanogenesis.We speculate that mass-standardized CH 4 production rates in macrofauna holobionts may undergo pronounced seasonal variations.For example, average rates of 151 AE 37 nmol CH 4 g DW À1 d À1 in the Sacca di Goro may be temporally enhanced in spring, when this estuarine system is strongly flushed by riverine inputs and the whole system becomes freshwaterdominated (Magri et al. 2020).Interestingly, other parameters such as the respiration to excretion ratio (O 2 : NH 4 + ) and temperature were not significantly correlated with massnormalized CH 4 production (Table S4).Respiration to excretion ratio was tested as a proxy for macrofauna food quality (i.e., lower ratio characterizes more labile food).We expected longer residence time in the digestive tract and higher CH 4 production associated with more recalcitrant food source (i.e., resulting in high respiration to excretion ratios).The absence of such correlation may be due to methodological constraints: we measured metabolic rates after macrofauna preincubation in filtered water and we did not characterize the feces and pseudofeces N content.
In the aquatic environment, methanogenesis and methanotrophy are tightly tied to temperature (Schulz et al. 1997;Li et al. 2021).Considering that our results provide a snapshot of the holobionts activity during the warm season (May-August), we conclude that salinity may have an important regulatory effect on CH 4 production.Further experiments are needed to explore temperature regulation mechanisms (Reeburgh 2007).We also suggest that future studies should better characterize food quality and investigate the effects of extreme events (e.g., freshwater floods and extreme droughts) on CH 4 production in macrofauna holobionts, including organisms reared in aquaculture systems.

Rate upscaling and environmental implications
Our results indicate that most macrofauna taxa (15 out of 19) release CH 4 in coastal ecosystems (Fig. 5).We took a closer look at whether animal associated CH 4 production can play a substantial role in budgets by upscaling the animal contribution to the actual ecosystem flux for two cases, where actual ecosystem CH 4 emission rates were reported in literature.First, we analyzed the case of T. philippinarum in the Sacca di Goro Lagoon.(Leip 2000), after accounting for oxidation processes taking place in the sediment.
In the central part of the brackish Baltic Sea, sedimentary CH 4 emissions are estimated to be 12 μmol CH 4 m À2 d À1 (Bonaglia et al. 2017).The most abundant macrofauna taxa in this environment are M. arctia and L. balthica, which can reach densities up to 4000 ind.m À2 (Gogina et al. 2016).Bonaglia et al. (2017) performed an upscaling exercise comparable to that we carried out for T. philippinarum and found that polychaetes and bivalves may contribute up to 9.5% of the total CH 4 flux in the Baltic Sea, which is in the same order of magnitude of holobionts' contribution in the Sacca di Goro (Adriatic Sea).We conclude that in brackish to almost freshwater environments (salinity between 0.2 and 16.9) animal CH 4 release plays a substantial role in budgets, especially when a few taxa are present at very high densities and biomasses, such as monoculture farming sites characterized by low salinity.S1.Invertebrate illustrations were created by IAN/UMCES (ian.umces.edu/media-library)

Fig. 1 .
Fig. 1.Map showing the geographical locations of the four investigated coastal shallow systems: Öre Estuary (Sweden), Curonian Lagoon (Lithuania), Sacca di Goro Lagoon, and Oristano lagoons (Italy).Note that Oristano lagoons include Cabras and Mistras lagoons as they are hydrologically connected.For additional information on the study sites refer to Fig. 5 and TableS1.

Fig. 2 .
Fig. 2. Regression between log-transformed rates of individual O 2 respiration (IRR, a), NH 4 + excretion (IER, b) and CH 4 production (IPR, c; μmol or nmol ind.À1 d À1 ), and log-transformed individual biomass (g DW ind.À1 ).Different colors represent macrofauna with different living habits (= living at different depth in sediment).The equations of the regression lines are reported with the correlation coefficient (R 2 ) and p value.The sample size is lower (64 out of 103 observations) for panel c as negative rates (CH 4 uptake) were excluded.For visualizing all 103 observations refer to the non-logged dataset presented in Fig. S1.
Clam cultivation sites cover more than one third of the lagoon surface at biomasses equivalent to 528 g DW m À2 (Welsh et al. 2015).At our incubation conditions T. philippinarum produced 214 AE 15 nmol CH 4 g DW À1 d À1 .We used this production rate to compare the invertebrates' flux with the total CH 4 flux.The upscaled flux (biomass Â MPR) resulted in 113 μmol CH 4 m À2 d À1 , which is a maximum estimation of the direct contribution of the farmed T. philippinarum in the Sacca di Goro.In this context, clams potentially contribute $ 6% to the overall benthic flux of 1968 μmol CH 4 m À2 d À1

Fig. 5 .
Fig. 5. Geographical position of sampling sites with indicated the dominant macrofaunal taxa.Note that Oristano lagoons include Cabras and Mistras lagoons as they are hydrologically connected.Colored arrows indicate the average production of CH 4 as follows.Green arrow: Mass-standardized production rates (MPR) < 0.34 nmol CH 4 g DW À1 d À1 or CH 4 uptake.Yellow arrow: 6.0 < MPR < 70.7 nmol CH 4 g DW À1 d À1 , Orange arrow: 113.9 < MPR < 188.4 nmol CH 4 g DW À1 d À1 and Red arrow: 185.7 < MPR < 340.1 nmol CH 4 g DW À1 d À1 .Additional details on the study sites are reported in TableS1.Invertebrate illustrations were created by IAN/UMCES (ian.umces.edu/media-library) C f and C i (μmol or nmol L À1 ) are the final and initial concentrations of the chemical species; V (L) is the water volume in the microcosm; N is the number of incubated animals per microcosm; and t (day) is the incubation time.Positive values represent production while negative values represent uptake.