The unprecedented rate of CO2 increase in our atmosphere and subsequent ocean acidification (OA) threatens coastal ecosystems. To forecast the functioning of coastal seagrass ecosystems in acidified oceans, more knowledge on the long-term adaptive capacities of seagrass species and their epibionts is needed. Therefore we studied morphological characteristics of Posidonia oceanica and the structure of its epibiont communities at a Mediterranean volcanic CO2 vent off Panarea Island (Italy) and performed a laboratory experiment to test the effect of OA on P. oceanica photosynthesis and its potential buffering capacity. At the study site east of Basiluzzo Islet, venting of CO2 gas was controlled by tides, resulting in an average pH difference of 0.1 between the vent and reference site. P. oceanica shoot and leaf density was unaffected by these levels of OA, although shorter leaves at the vent site suggest increased susceptibility to erosion, potentially by herbivores. The community of sessile epibionts differed in composition and was characterized by a higher species richness at the vent site, though net epiphytic calcium carbonate concentration was similar. These findings suggest a higher ecosystem complexity at the vent site, which may have facilitated the higher diversity of copepods in the otherwise unaffected motile epibiont community. In the laboratory experiment, P. oceanica photosynthesis increased with decreasing pHT (7.6, 6.6, 5.5), which induced an elevated pH at the leaf surfaces of up to 0.5 units compared to the ambient seawater pHT of 6.6. This suggests a temporary pH buffering in the diffusive boundary layer of leaves, which could be favorable for epibiont organisms. The results of this multispecies study contribute to understanding community-level responses and underlying processes in long-term acidified conditions. Increased replication and monitoring of physico-chemical parameters on an annual scale are, however, recommended to assure that the biological responses observed during a short period reflect long-term dynamics of these parameters.
Citation: Guilini K, Weber M, de Beer D, Schneider M, Molari M, Lott C, et al. (2017) Response of Posidonia oceanica seagrass and its epibiont communities to ocean acidification. PLoS ONE 12(8): e0181531. https://doi.org/10.1371/journal.pone.0181531
Editor: Just Cebrian, University of Alabama, UNITED STATES
Received: January 31, 2017; Accepted: July 3, 2017; Published: August 9, 2017
Copyright: © 2017 Guilini et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All data can be accessed through the PANGAEA database: doi: 10.1594/PANGAEA.857983.
Funding: Financial support for this study was provided under the European Commission’s Framework Seven Programme Topic OCEAN.2010.3, project number 265847.T, “ECO2 - Sub-seabed CO2 Storage: Impact on Marine Ecosystems”. Katja Guilini additionally acknowledges the Flemish fund for Scientific Research (grant number: 1242114N) for funding. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Seagrasses, which are considered to be amongst the most productive ecosystems worldwide also support high biodiversity . Hence, their current worldwide decline at 7% loss per year is worrying [2,3]. Major threats to seagrass ecosystems and the services they provide include the immediate impacts of coastal development by growing human populations and the indirect impacts of the unprecedented increase in atmospheric CO2 concentrations through global warming  and ocean acidification . Evaluating the response of seagrass ecosystems to pressing environmental stressors is crucial for effective management of coastal regions in the future.
Human induced increases in atmospheric CO2 concentrations result in changes in ocean carbonate chemistry and pH, a process termed ocean acidification (OA). OA raises concerns about the health of ocean ecosystems, particularly the viability of calcifying organisms to cope with predicted seawater pH decreases of 0.06–0.32 units by the year 2100, depending on the considered CO2 emission scenario . The sole increase in partial pressure of CO2 (pCO2), however, is not expected to have a direct negative impact on seagrasses. On the contrary, it has been observed that when exposed to increased pCO2, seagrasses increase their photosynthetic rates [6–13], shoot productivity , density , leaf growth rates [8,16], flowering frequency , and biomass [6,11,15,18]. Environmental stressors such as nitrogen limitation  or an increase in turbidity [19,20], temperature [21–23] and grazing  that occur simultaneously, have been shown to counteract the direct positive effects on seagrasses. Moreover, non-calcifying epiphytes may respond positively to increased pCO2 and cause shading [25,26], while calcifying epiphyte coverage might decline under similar conditions and induce high light exposure . The reaction of epibiota generally depends on its ability to cope with a change in the inherent natural diurnal variability in seawater pH created by seagrass metabolism (ΔpH range: 0.06 to >1; [27–31]). Overall, it seems that the global effect of increases in pCO2 on seagrasses may be spatially heterogeneous, species-specific and related to the plant physiology structure . Particularly due to effects being dependent on geochemical site characteristics and resulting from the interplay of several physico-chemical stressors and interactions with other biota [9,12].
To forecast the functioning of marine seagrass ecosystems in acidified oceans, we need to increase our knowledge on the long-term adaptive capacities of different seagrass species to high pCO2 conditions, while taking into account the physico-chemical settings and their associated biota. Volcanic submarine CO2 vents may act as natural analogues to study the effects of long-term exposure to acidified seawater in a holistic approach. However, the high spatial and temporal variability in CO2 and the interference of other variables that may differ across the same gradient could hamper the detection of a reliable dose-response relationship [14,18,31,32]. More observations at different CO2 vent systems may, however, facilitate our understanding, particularly when these studies also consider laboratory and in situ pH manipulations that examine the same ecosystem-level measures.
Only a limited number of seagrass species have been examined at natural CO2 vents (i.e. Cymodocea nodosa, C. serrulata, Halophila ovalis, Posidionia oceanica) and these studies primarily investigated the effects on seagrass productivity and growth response [12,19,24]. Some studies have taken into account the effects on the associated sessile epiphytes, bryozoans and hydrozoans, the macro-invertebrates and the motile megafaunal grazers [4,14,18,33–36]. The response of motile meiofauna living in high abundances on the leaves or the vertical rhizomes  is still unexplored. Despite the valuable information previous CO2 vent studies have provided, sampling different biological elements at sites characterized by variable biogeochemistry (e.g. pH ranges) and sampling at different times, hampers the integration of data. This in turn hinders the interpretation of potential interacting processes and cascading responses over different levels of organization. It is therefore advisable to perform multispecies studies which aid in unraveling community-level responses and their underlying processes, permitting our comprehension of the populations’ ability to adapt and evolve .
In this study, we investigated the Mediterranean seagrass Posidonia oceanica (L.) Delile and its associated epibionts at a CO2 vent area in the vicinity of Bassiluzo Islet, off Panarea Island (Italy, Fig 1). Studies at complex natural CO2 vent environments require a description of the abiotic settings and dynamics. We therefore analyzed seawater and pore water from different heights above and below the surface of the seagrass beds and logged the diurnal variation in seawater pH, O2, water depth and oxygen reduction potential. P. oceanica and its associated sessile and motile epibionts were investigated by testing the following hypotheses: (H1) Long-term acidified seawater conditions result in an increase in P. oceanica leaf and shoot densities and leaf length. On the contrary, (H2) epibiont communities on P. oceanica are vulnerable to long-term acidified seawater conditions, and exhibit altered community structures and a reduction in abundances and diversities. Additionally, we performed a laboratory experiment to measure the potential buffering capacity of seagrass leaves and test if an increase in pCO2 enhances the photosynthetic rate of P. oceanica (H3).
(A) Map indicating the location of the two study sites east of Basiluzzo Islet (Mediterranean Sea): the reference site ‘REF’ (38°39.827’ N, 15°07.118’ E) and the CO2 vent site ‘CO2-R’ (38°39.749’ N, 15°07.132’ E). Software: ESRI—ArcGIS Desktop 10.1, map layers originating from ESRI—Europe—world countries 2008 and ESRI—ArcGIS online (layer package by fraciccio67)–comuni, (http://www.esri.com). (B) Overview of a P. oceanica seagrass meadow at CO2-R with visible gas bubbling. (C) P. oceanica leaf tip with indication of the examined part, i.e. half of the upper 10 cm of the leaf.
Materials and methods
The study area was located east of Basiluzzo Islet close to Panarea Island in the Mediterranean Sea (Fig 1). The area consists of patchily distributed seagrass meadows, typical sandy sediments and rocky outcrops. It covers locations where fluids and gases rich in CO2 (ca. 97%)  escape the seabed as a result of volcanic activity , as well as reference sites without any measurable evidence of seepage. To determine the biogeochemical characteristics of the seagrass meadow environment, sensors were used in situ and samples were obtained by SCUBA diving during the sampling campaign in 2012 (02.06–12.06). An exception is the pore water samples which were obtained in 2013 (01.06–14.06). Two sites (200 m apart) were the focus of this study: one with red sediment where evenly distributed gas seepage occurred (“CO2-R”, 38°39.749’ N, 15°07.132’ E, 13–14 m water depth) and a second site where sediments were colored grey and no gas seepage occurred (“REF”, 38°39.827’ N, 15°07.118’ E, 14–15 m water depth). Both study sites covered approximately 50 x 50 m. Water temperature at 1 m above the seafloor (asf) was 18.7 to 19.3°C, and P. oceanica meadows occurred in numerous patches at both sites. Sampling was performed after receiving permits of the Soprintendenza del Mare, Assessorato dei Beni culturali e del’Identità siciliana, Regione Sicilia, Palermo, Italy.
Seawater and pore water.
RBR-Dataloggers (XR-420 D; Ottawa, Canada) measured and logged pHNBS, O2, ORP (oxygen reduction potential) and pressure (tides) every three seconds, continuously for 15 days, in the midst of a seagrass meadow (at 2 cm asf) at both sites. These data series were averaged per 5 minutes. Additionally, seawater was sampled at 80 cm and 10 cm asf, i.e. above and within each seagrass patch, with a 5 L-Niskin bottle and glass syringes, respectively. Pore water was sampled with a stainless steel lancet and attached syringe with filter (22 μm) at 0 cm, -10 cm and -20 cm asf in the meadows. Triplicate seawater and pore water samples were collected on one occasion only and might therefore mask diurnal variation. Upon return to the laboratory, sub-samples were taken from the collected seawater samples using 10 mL syringes, attached to Rhizons MOM (19.21.21F, pore size 0.15 μm; Rhizosphere Research Products, Wageningen, Netherlands) to determine dissolved inorganic carbon (DIC), pHNBS, total alkalinity (TA), and nutrient concentrations (NH4+, PO43-, NO2-, NO3- + NO2-, Si). Dissolved silicate was measured photometrically according to Grasshoff et al. . NH4+, PO43-, NO2-, NO3- were measured spectrophotometrically with a continuous-flow analyzer (Bran & Lübbe GmbH, Norderstedt, Germany) using a variant of the method of Grasshoff et al. . Measurements of pHNBS were directly done in the field with a pH 96 by WTW (Wissenschaftlich-Technische Werkstätten GmbH, Weilheim, Germany) and an InLab Semi-Micro electrode by Mettler Toledo (Gießen, Germany). The probes were calibrated at ambient temperature with conventional buffer solutions by Mettler Toledo (pH 4.00 and 7.00). All pHNBS values were converted to pH values on the total scale (pHT), via a conversion from pHNBS to pH on seawater scale (pHSWS) according to Lewis & Wallace : aH = 10(-pHNBS) = fH * HSWS, where aH is the activity and fH is the activity coefficient of the H+ ion. Therefore, pHNBS = pHSWS—log10 fH. Takahashi et al.  provides a formula for the activity coefficient fH as a function of salinity (S) and temperature (T): fH = 1.2948–0.002036 * T + (0.0004607–0.000001475 * T) * S2. Finally, the conversion from pHSWS to pHT is described in Zeebe & Wolf-Gladrow . For DIC and TA, 2 mL seawater was stored into headspace-free glass vials at 4°C. DIC and TA were assessed via flow injection analysis  and two-point titration , respectively. Degree of saturation for calcite and aragonite (Ωcal and Ωar) were calculated based on TA and DIC data and in situ temperature and salinity conditions using the R package seacarb v 3.0.11 . The carbonate dissociation constants (K1 and K2) used were those of Roy et al. , Kf from Dickson and Riley  recommended by Dickson and Goyet  and Ks from Dickson . Average pH values were calculated based on averaged H+ concentrations and accompanied by their coefficient of variation (c.v.).
P. oceanica growth characteristics
SCUBA divers determined shoot density at replicated patches (n = 3) at both sites, using a circular metal frame with a diameter of 56.4 cm (i.e. 0.25 m2). After each count, the leaves of 10 shoots were randomly sampled (i.e. along a line each 5 cm, insuring low probability to match any natural pattern) by cutting them off at the leaf bases. All leaves of one shoot were carefully transferred into a plastic bag, sealed and transported. In the laboratory, the number of leaves per shoot were counted, both sides of all leaves were scanned wet in high resolution using a flatbed scanner to facilitate length and area measurements of each leaf using the image editing software ImageJ  and to also document epibiotic species. Subsequently, the leaves were stored in a 4% formaldehyde-seawater solution. The leaf area indices are found by multiplying the average leaf area per shoot by the estimated shoot density of each individual sample .
Epiphytic calcium carbonate (CaCO3) weight
Seagrass leaves (n = 5) from both sites were sampled and scanned as described before. The upper 5 cm were cut off, air dried and weighed before being washed in 0.1 M HCl. After all CaCO3 was dissolved, the leaves were air dried again and weighed to determine the CaCO3 weight per leaf area (mg cm-2).
Sessile epibiont community
From the 10 shoots that were collected per quadrat (n = 3), 5 were randomly chosen and used to examine the epibiotic community (i.e. 15 shoots site-1). Epibiota on the upper 10 cm of each leaf tip and on half of the area of the concave leaf side were analyzed (Fig 1). Only intact leaf tips were considered for epibiont abundance. All sessile organisms were identified to the most practical taxonomic level based on the high resolution scans of the fresh material and the preserved sample, using a microscope (5x10x). Biota or colonies with more than 1 mm in diameter or length were considered. Additionally, bite marks of macro-herbivores were recorded and quantified on all leaves from the 5 shoots, to estimate seagrass consumption. Three main consumers can be distinguished by their bite marks: the isopod Idotea sp. (Crustacea, Isopoda), the sea urchin Paracentrotus lividus (Echinodermata, Echinoidea), and the fish Sarpa salpa (Actinopterygii, Sparidae). The total number of leaves examined per site was influenced by the number of leaves per shoots (CO2-R: 195, REF: 165). Bite mark abundance was considered as number of bite marks per leaf.
Motile epibiont (meiofauna) community
Three replicate seagrass patches per site were sampled to study the meiofauna associated with seagrass leaves and vertical rhizomes, i.e. two subunits of the plants distinguished based on microhabitat structure . Approximately 12 to 18 leaves per sample were collected by placing a plastic bag over the leaves and gently cutting the leaves at the base before closing the bags with elastic bands. The remaining vertical rhizomes, which included the degrading leaf sheaths, were cut off and gently transferred into separate plastic bags, to minimize transfer through the water column. Upon return to the laboratory, both leaf and vertical rhizomes (further referred to as rhizomes) samples were sieved (32 μm) to eliminate the water. The retained material was stored in a 4% formaldehyde-seawater solution. Samples were then emptied over stacked 1 mm and 32 μm sieves and rinsed thoroughly. Meiofauna (size class 0.032–1 mm) were extracted from the samples through triple density centrifugation with the colloidal silica polymer LUDOX TM 40 (ρ: 1.18). All metazoan meiofauna were classified at higher taxon levels and counted under a Leica stereoscopic microscope (16x5x). ImageJ software was used to estimate the surface of scanned leaves to standardize meiofauna densities on the leaves (individuals 100 cm-2). To standardize the meiofauna densities on the rhizomes (individuals 10 cm-3), the volume of rhizomes was determined by the amount of water displaced when the rhizomes were submerged. Whenever possible, 120 copepods and 100 nematodes were randomly handpicked with a needle before identification to species (numeric) level. Diversity of the meiofauna at a higher taxon level and at nematode and copepod species level were assessed in terms of the Hill’s indices (H0, H1, H2, Hinf; ). The indices are sample-size sensitive and therefore analysed for leaves and rhizomes separately. The rarefaction index (ET(51) or ES(51); ), which is independent of sample size, was also calculated and analysed together for both rhizomes and leaves. Based on nematode mouth morphology, all identified individuals were classified in functional feeding groups according to Wieser . Nematode functional diversity was assessed by calculating the inverse trophic index (1/TI or 1/Σθ2, where θ is the percentage of each feeding group; ).
Seagrass shoots were collected by SCUBA divers at both sites and stored on shore in aerated tanks in seawater at in-situ conditions. Measurements were performed within 2 days after collection. For laboratory experiments, one shoot, that consisted of intact leaves and vertical rhizomes, was placed in a flow cell (volume: 1 L). Seawater, from the site of the plant’s origin, was pumped from an aerated recirculation tank (volume: 20 L) through the flow cell. One leaf was placed on two stoppers that were 5 to 8 cm apart and held in place with two small stoppers. Thus both sides of the leaf were freely exposed to the flowing water. Seawater temperature was held constant at 22°C and flow velocity was maintained at ca. 5 cm s-1. The light source was a fiber optic halogen lamp (Schott KL1500, Germany). Incident light intensity was 200 μmol photons m-2 s-1, quantified as scalar irradiance with a Biospherical Instruments meter (QSL-100, USA). Liquid membrane type pH microsensors and potentiometric O2 microsensors were prepared, calibrated and applied as described by Revsbech  and de Beer . The microsensors had a tip size of 10 μm and response times (t90) of less than 1 s. Profiles were measured by positioning the microsensors using a motorized micromanipulator. Net photosynthesis was determined as the oxygen flux (J) from the leaf surface, calculated from the interfacial oxygen gradient (dC/dX) using Fick's 1. law: J = D*dC/dX, where D is the diffusion coefficient for oxygen (i.e. 2.16 x 10−9 m2 s-1 at 22°C, salinity of 40). Net photosynthesis rates were determined on three leaves that were each exposed to pHT 7.6, 6.6 and 5.5, for a few hours per pH regime. The pH was reduced and stabilized by flushing the recirculation tank with CO2 gas and at each pH, 5 to 10 profiles were measured on the convex side of a leaf and averaged, after the shoots acclimatized to the manipulated conditions for 1 to 2 hours. Likewise, pH profiles were measured over a distance of 0 to 1000 μm off the leaves’ surface in both light (n = 2–3 for REF, n = 3 for CO2-R) and dark (n = 1 for REF) conditions at each pH. The pH in the flow cell was constantly monitored with a pH sensor and kept stable within 0.1 unit. All pHNBS values were converted to pHT as mentioned earlier and average pHT values were calculated based on averaged H+ concentrations and accompanied by their variation (c.v.).
To characterize the geochemical settings of the study sites, pH, O2, ORP and pressure in the water column collected by the RBR sensor were analyzed with non-parametric Wilcoxon tests, using R software . Additionally, univariate non-parametric permutational ANOVA (PERMANOVA) analyses (PRIMER v6 and PERMANOVA+ add-on software; ) were performed to test for the differences in environmental variables between study sites and positions in the water column or depth in the sediment. Therefore a two-factor design was used with the fixed categorical factors ‘site’ (CO2-R, REF) and ‘position’ (80 cm, 10 cm, 0 cm, -10 cm, -20 cm asf). The effects of long-term acidified seawater conditions on P. oceanica growth characteristics (H1) and sessile and motile epibiont community characteristics (H2) were tested with PERMANOVA analyses, considering the fixed categorical factor ‘site’ (CO2-R, REF) and, in case of the motile epibionts (meiofauna), also ‘habitat’ (leaves, rhizomes). PERMANOVA analyses were based on Euclidean distance similarity measures on untransformed, and occasionally log(x+1) transformed (in case of pore water, Ωca and Ωar) univariate data, on Bray-Curtis similarity measures on standardized, square root transformed higher meiofauna taxa, nematode and copepod species, and bite mark abundances and on Jaccard similarity measures on the presence-absence data of sessile epibiont taxa. Calculation of the Pseudo-F ratio and p value (null hypothesis are rejected based on a level of significance α = 0.05) required unrestricted permutation of raw data (univariate and one-factor multivariate) or 9999 permutations of the residuals under a reduced model (multiple-factors multivariate). Posteriori pair-wise tests were conducted where significant effects were found and only when a PERMDISP test confirmed the homogeneity of multivariate dispersions. Where only a restricted number of unique permutations was possible, p-values were obtained from Monte Carlo simulations. (Dis-)similarity between and among the sessile and motile epibiont communities were visualized by non-metric multidimensional scaling (NMDS) plots (in Supporting information). The variability among and between the sessile and motile epibiont communities was quantified with SIMPER analyses on presence absence data and standardized, square root transformed abundance data, respectively (in Supporting Information). The effect of elevated pCO2 on the photosynthetic rate of P. oceanica (H3) was tested with a linear regression, using R software . The data used for the linear regression contained no outliers and the residuals were normally distributed.
Seawater and pore water geochemistry
The RBR logger data showed depth variation over time related to the tidal cycle (Fig 2). Photosynthesis can explain the observed O2 fluctuations at both sites. A drop in seawater pH at low tide could be observed at CO2-R, while at REF the pH co-varied with photosynthesis. Daily pH variation (ΔpH) and minima (pHT-min) differed between REF and CO2-R (ΔpH: 0.12 ± 0.05 and 0.49 ± 0.33, respectively; pHT-min: 7.75 ± 0.12 and 7.23 ± 0.97, respectively, P(perm) = 0.0001; Table 1), while daily pH maxima (pHT-max) did not differ (7.87 ± 0.07 and 7.86 ± 0.06, respectively). The mean seawater pHT was also lower at CO2-R (7.72 ± 0.67) compared to REF (7.82 ± 0.11; Wilcoxon test, p < 2.2e-16). At CO2-R, O2 concentrations in the seagrass meadow reached higher values than at REF (0.268 ± 0.010 mM and 0.259 ± 0.017 mM, respectively; Wilcoxon test, p < 2.2e-16). ORP was also more variable at CO2-R (0.135 ± 0.029 mV) compared to REF (0.155 ± 0.022 mV) as a result of enhanced gas release and fluid seepage (including dissolved reductants that lowers the ORP) during low tide at CO2-R.
RBR sensor data were gathered over 14 to 15 days at the REF site (A) and the CO2-R site (B). The color legend is provided in the graph.
Seawater characteristics based on water sampling did not differ according to the sampling position, i.e. within or above the seagrass meadows. The values were therefore averaged per site and reported in Table 2. It is important to note that seawater and pore water samples were sampled only occasionally and that therefore these snapshot results could possibly mask diurnal fluctuations. Seawater salinity was 38 at both sites and nitrite+nitrate concentrations were always below detection limit. DIC and silicate concentrations in seawater were highest at CO2-R and differed from the values measured at REF (P(perm) ≤ 0.041; Table 1). TA, pH, phosphate and ammonium, and Ωcal and Ωar in the water column were similar at both sites. Ωcal and Ωar were on average 5.5 and 3.6, respectively. Pore-water characteristics (Table 2) were generally very variable and differed with depth in the sediment between the 0 cm layer and the deepest layer (-0.2 m bsf) for pH (P(perm) ≤ 0.05), Ωcal (P(perm) = 0.0283) and for Ωar (P(perm) = 0.0281). Pore water pH, Ωar and Ωcal were lower at CO2-R compared to REF (Table 2; P(MC) ≤ 0.03; Table 1).
Posidonia oceanica growth characteristics
The average shoot density (N m-2) and leaf density (N m-2 and N shoot-1) were highest at CO2-R (Table 2). The difference with REF was, however, not significant. Single sided surface area of leaves m-2 (Leaf Area Index, LAI) was higher at REF compared to CO2-R (P(MC) = 0.0161; Table 1).
Sessile epibiont community and bite marks
The amount of CaCO3, a proxy for the abundance of calcifying epibionts, on the upper 5 cm of the leaves did not significantly differ between the vent site (2.54 ± 1.75 mg cm-2) and REF (0.74 ± 0.52 mg cm-2). The ratio of the total number of calcifying versus non-calcifying sessile epibiont individuals (75: 25) at REF was higher compared to the ratio at CO2-R (84: 68). This indicates an increase in non-calcifying epibiont abundance at the vent site.
The epibiont community composition was found to differ between REF and CO2-R (S1 Fig, S1 Table; P(perm) = 0.0001; Table 1). Additionally, differences occurred in both the community of calcifying and non-calcifying epibionts (P(perm) = 0.0001; Table 1). Calcifying red algae, identified to the family Corallinaceae, and Miniacina miniacea (a colonial sessile foraminifera) were present in a few occasions at REF and were absent from CO2-R. Most calcifying species were, however, found at both sites, and only three calcifying bryozoan species exclusively occurred at CO2-R (Tubulipora sp., Cellepora sp., Collarina sp.). Amongst the non-calcifiers, no species exclusively occurred at REF, while 7 species were only found at CO2-R. These shifts in community composition resulted in a higher epibiont species richness at CO2-R compared to REF (20.3 ± 7.4 and 13.0 ± 2.7, respectively, P(MC) = 0.0218; Table 1). In particular, there was a higher species richness of non-calcifying epibionts at CO2-R versus REF (CO2-R: 4.53 ± 2.83, REF: 1.67 ± 1.11, P(perm) = 0.006; Table 1), while the number of calcifying epibiont species was similar (CO2-R: 5.60 ± 2.64, REF: 5.00 ± 2.04).
Bite marks were observed on the leaves at both sites. No significant differences were found between the sites in terms of total number of bite marks (CO2-R: 0.17 ± 0.07 marks leaf-1, REF: 0.27 ± 0.09 marks leaf-1) or absolute or relative abundance of species-specific bite marks. The majority of the bite marks at the vent site were attributed to the sea urchin Paracentrotus lividus (80%) and the fish Sarpa salpa (14%), while the isopod Idotea sp. bite marks were only found on two occasions (6%). 96% of the bite marks at REF originated from the P. lividus and 4% from S. salpa.
Motile epibiont (meiofauna) community
Total meiofauna density on the leaves or the rhizomes or the densities of separate meiofauna taxa did not differ between the sites (Table 3). An exception were the tanaidaceans who exclusively, though with low abundances, occurred on the rhizomes at CO2-R and not at REF (1.16 ± 0.68 ind. 10 cm-³, P(MC) = 0.047; Table 1). Distinct meiofaunal communities were found for the seagrass leaves and rhizomes at the higher taxon level and on nematodes at the species level (S1 Fig, S1 Table; P(perm) ≤ 0.0054; Table 1). Harpacticoid copepods were dominant in the leaf canopy, followed by nematodes, while the opposite occurred for the rhizomes (Table 3). Differences in nematode and copepod community composition were mainly attributed to shifts in the occurrences of less abundant species (i.e. < 5% relative abundance). A site effect on the community composition was only detected for nematode species (P(perm) = 0.022; Table 1). Chromadora sp.1 dominated at both REF and CO2-R, while a shift in relative abundances that occurred among the less dominant species changed the overall community composition. The difference found for copepod communities was attributed to non-homogeneous multivariate dispersion (Permdisp < 0.05). Meiofauna higher taxon and nematode and copepod species diversity (Hill’s indices; Table 3) on the leaves did not differ between sites. On the rhizomes, however, nematode and copepod species richness (H0; Table 3) was higher at CO2-R compared to REF (P(MC) ≤ 0.041; Table 1). The rarefaction index (ET(51) or ES(51)) was higher on the rhizomes compared to the leaves for nematode species, and higher at CO2-R compared to REF for the copepod species (P(perm) ≤ 0.027; Table 1). In terms of functioning, no differences in nematode trophic composition and trophic diversity were found (Table 3).
Posidonia oceanica net photosynthesis and pH profiles
In the laboratory experiment, we observed a trend in increasing net photosynthesis with decreasing seawater pH (adjusted R2 = 0.50, p = 9.93e-06; Fig 3). Moreover, under illuminated conditions, seawater pH was elevated with up to 0.4 and 0.5 units above ambient seawater values over a distance of 200μm from the leaf surface, when the seawater pH was decreased to 7.6 and 6.6, respectively (Fig 4, S2 Fig). At a seawater pH of 5.5 the pH close to the leaves was not elevated, due to strong buffering of the carbonate system when pH of the seawater approaches a pK value. Very little to no respiration was observed when leaves from REF were incubated under dark conditions as the surface pH was equal to seawater pH at all pH levels (Fig 4).
Net photosynthesis (J, mol m-2 s-1) of P. oceanica was measured on leaves from REF (dots, n = 3) and CO2-R (triangles, n = 3) in relation to manipulated seawater pHT (regression analysis, adjusted R2 = 0.50, p = 9.93e-06, no outliers, residuals normally distributed).
pHT profiles (averages ± s.d.) were measured on the convex side of P. oceanica leaves at seawater pHT 7.6 (triangle), 6.6 (square) and 5.5 (circle). Measurements started from the leaf surface (distance = 0 μm), and were performed in both light (black, n = 2–3) and dark (white, n = 1) conditions.
Typical of all highly productive ecosystems, variability in O2 concentrations and pH in the water centered on the P. oceanica meadows at our study site is a result of seagrass metabolic activity that follows a diurnal cycle . The pH pattern at the CO2 vents, however, is clearly modulated by the pressured release of CO2 gas during low tide episodes. A similar diurnal range in pH was measured in P. oceanica meadows at CO2 vents off Ischia, Italy . The daily pH variation at most vent sites near Ischia, however, was found to be two to three times higher when compared to pH measurements at Basiluzzo Islet vents (ΔpH: 1.0–1.6 versus 0.1–1.2, respectively). When considering the average pH values, the level of acidification in the seagrass meadows at the Basiluzzo Islet CO2 vents compared to the reference sites (ΔpH: 0.1) is relatively mild in comparison to the conditions found at other vent areas e.g. off Ischia (ΔpH: 0.2–0.8; ) and in Milne Bay, Papua New Guinea (ΔpH: 0.1–0.4  or 0.4–1.5 ). This characteristic, together with the absence of reactive gases (H2S, H2 and CO) and organic gas compounds (CH4, C2H6, C3H8), the lack of temperature anomalies of vented fluids and the vicinity of a reference area with similar environmental settings (supporting data based on field studies in 2012 and 2013 available via doi:10.1594/PANGAEA.871453), makes the study area close to Basiluzzo Islet a relatively good site to study pure CO2 effects. The site represents a realistic, near future scenario, that conforms to ranges of predicted open ocean surface pH reductions of 0.06 to 0.32 units by 2100 .
Several laboratory and mesocosm experiments and in situ studies at natural CO2 vents have demonstrated that excessive CO2 increases seagrass net photosynthesis [e.g. 6–11,13]. However, particularly for P. oceanica, findings are limited and not uniform. Increasing photosynthetic rates have been measured in laboratory experiments over pH ranges from 8.2 to 6.0 (unknown pH scale; ) and from 9.2 to 7.9 (unknown pH scale; ) or at a pHT of 7.3 . Cox et al. , however, found no difference in photosynthesis during the 4 months in situ experimental eFOCE (European Free Ocean Carbon Dioxide Enrichment) incubation, where pH was lowered by 0.26 units to pHT of 7.7. Also when considering photosynthetic efficiency and electron transport rates at the CO2 vents off Ischia (mean pHT: 8.2 versus 7.6; ) or in a laboratory experiment that simulated different levels of OA (380, 750 and 1000 ppm pCO2) over 1 to 3 weeks , no effects were found. Our laboratory experiment revealed an increasing trend in P. oceanica net photosynthesis with pCO2 increasing and pH decreasing over a pHT range of 7.6 to 5.5. Taking into account the reduced LAI at the vent site, it is unlikely that the higher O2 concentration measured in the meadows at CO2-R compared to REF are solely attributed to seagrass photosynthesis. Johnson et al.  observed an increase in epiphytic diatom abundance on artificial substrata placed along a CO2 gradient at the vent area off Vulcano Island (median pHNBS: 8.2–7.7). Diatom abundance at the water-sediment interface in a shallow subtidal area in the Adriatic Sea was also found to be linked to the availability of silicate . Likewise, diatom abundance on the seagrass leaves at the vents off Basiluzzo may have been stimulated by increased CO2 and silicate concentrations, the latter being released by the weathering of sediments. Their potential contribution to higher O2 concentrations at the vent site is, however, purely speculative as we didn’t quantify the epiphytic diatom abundance.
The overgrowth of seagrass leaves with algal epiphytes can initiate a negative growth response of seagrasses [25,64] by reducing the quantity and quality of light reaching the leaves. The overall plant performance during day-time is thus impeded, while an enhanced diffusive boundary layer reduces the exchange of gasses and nutrients with the ambient water column . This combined with the stimulation of seagrass growth, may explain the elevated P. oceanica shoot density at the CO2 vent site off Ischia when epiphytic coverage was severely reduced (pHT-mean: 7.60, pHT-min: 6.98; ). Under less severely acidified conditions at Ischia (pHT-min: 8.15–7.67), seagrass shoot density did not increase, though epiphyte coverage gradually decreased . The latter is partly in agreement with the findings near Basiluzzo Islet, where at average pHT-min of 7.23, shoot and leaf density also remained unaffected, despite an increase in non-calcareous epibiont abundance. It must be said that shoot density at REF had a relatively high variability compared to shoot densities of P. oceanica reported by Hall-Spencer et al.  and Balestri et al.  for the NW Mediterranean Sea. Therefore, we cannot exclude an effect of low replication in our study.
The significantly higher LAI, which indicates longer leaves, at REF compared to CO2-R corresponds to previous findings in the P. oceanica meadows at the CO2 vents off Ischia . The increased leaf erosion at Ischia was attributed to various grazers. Likewise, Cymodocea nodosa, also known as little Neptune seagrass, exposed to higher pCO2 at the Vulcano Island vent area were found to be more prone to grazing due to a reduction of phenolic substances in the seagrass leaves . We found no evidence of increased grazing based on the number of bite marks at Basiluzzo Islet. The dominant grazer, the sea urchin Paracentrotus lividus, prefers leaves covered with epibiota  and adult, thicker leaves , and Cox et al.  did measure an increase in P. oceanica leaf thickness after 4 months in-situ exposure to acidified conditions (pH lowered with 0.26 units to 7.7). Therefore, it could be that grazing at CO2-R was stimulated by the higher abundance of sessile non-calcifying epibionts we observed and perhaps an increased thickness of the leaves at the vent site, and that leaf tips eroded after fragmentation was triggered by consumption. Specific statements about the susceptibility of seagrasses to grazing are clearly not possible based on the number of bite marks only and require controlled cage experiments.
The CO2 vent conditions at Basilluzo Islet induced a shift in the sessile epibiont community composition, for both the calcifying and the non-calcifying group of organisms. Among the calcifiers, only the coralline red algae were obviously absent at the vent site. Alternatively, some calcifying bryozoan species (Tubulipora sp., Cellepora sp., Collarina sp.) appeared to benefit from the more acidic conditions and consequently the release of competitive pressure from formerly occurring or dominant species, as they exclusively occurred at the vent site. A similar community shift was also observed on P. oceanica leaves at the CO2 vents off Ischia (pHT-mean: 7.7 ) and following a 2- and 4-week incubation of P. oceanica leaves in the laboratory at pHT 7.0  and pHT 7.7 and 7.3 , respectively. Likewise, the diversity of crustose coralline algae that live on Enhalus acoroides, aslo known as tape seagrass, was reduced under the influence of the acidified conditions at vent sites in Papua New Guinea (pHT-median: 7.8 ). The high vulnerability of coralline algae versus the relatively high tolerance of at least some bryozoan species is related to the difference in their skeleton mineralogy. The relatively lower Mg content in the calcium carbonate skeleton of bryozoans makes them more resistant to chemical dissolution at low pH . At Basiluzzo Islet, the shift in the sessile epibiont community composition did, however, not affect the total abundance of calcifying epibionts; evidenced by the lack of variation in CaCO3 weight. Likewise, no change in CaCO3 mass and in percent coverage of invertebrate calcifiers and crustose coralline algae on P. oceanica occurred after 4 months of exposure to reduced pH conditions (pHT-mean: 7.75) in the FOCE system of Cox et al.  (NW Mediterranean Sea). Moreover, the exclusive occurrence of a number of sessile epibiont taxa at the CO2 vents near Basiluzzo Islet even resulted in an overall higher species richness under acidified conditions. Both these results contrast with findings from earlier studies on epibionts associated with natural and mimic P. oceanica seagrasses and artificial collector units, and studies on other benthic plant components and animals on hard substrata at the CO2 vent area off Ischia [18,34,69–71]. At Ischia, a general reduction in the abundance or a total disappearance of calcifying forms resulted in a drop in species number at the low pH stations (mean pHT varied from 6.6 to 7.4). Hall-Spencer et al.  indicated that the organisms with aragonite skeletons were absent at mean Ωar ≤ 2.5, which is a value lower than measured at our study sites (mean Ωar = 3.9 ± 0.7). Similarly, the percent cover of epiphytic calcareous algae decreased while fleshy macroalgal cover increased, though less substantial, from control towards vent sites in Milne Bay, Papua New Guinea (median Ωar = 3.5 versus 2.9 ). Similarly, in a laboratory experiment Cox et al.  observed a decrease of the carbonate content in epiphytes on P. oceanica under corrosive conditions (Ωar = 0.5–0.8) at pHT 7.32. Donnarumma et al.  reported that the implanted P. oceanica mimics at the CO2 vent sites off Ischia were dominated by filamentous algae and non-calcareous taxa such as hydroids and tunicates, which resulted in a poor and simplified assemblage. The relatively mild acidification effects observed at our study site, where the average pHT at the CO2 vent site is 7.72, which is 0.1 units lower compared to the reference site, suggests that the sessile epibiont communities can cope with a certain degree of OA. However, these results should be interpreted with caution because of the inherent natural variability in pCO2, pH and Ωar at shallow natural vents, the influences from non-acidified surrounding areas (e.g. recruitment) and incomplete understanding of historical biochemistry . The study of Cox et al. , however, where a portion of P. oceanica meadows was enclosed and exposed to control and acidified conditions (pHT-mean: 8.01 vs. 7.75) for 4 months, also concludes that negative impacts from OA on epiphytic communities were smaller than expected.
Under natural conditions, Mediterranean P. oceanica meadows are able to modify pH in the water column up to 0.5 pH units diurnally through photosynthetic activity and community metabolism [28,30]. Calcifying organisms may benefit from the modification of the carbonate system by these meadows as they provide a daily window of maximum Ωar where calcification is more cost efficient . This protective or buffering capacity might, however, be at risk when the health of seagrass meadows is altered by pressing environmental stressors . Apart from the unaltered leaf and shoot density and the increase in net photosynthesis at the CO2 vents off Basiluzzo Islet, the results from this study temper this concern, at least for the effect related to ocean acidification. The pH profiles measured on P. oceanica leaves in our laboratory experiment revealed that the exposure of epibionts to ambient acidified seawater conditions is restricted to the hours when no photosynthesis occurs. Under dark conditions the pH at the leaf surface equaled ambient seawater, as seagrass leaf respiration was very low. Seagrass photosynthesis buffered the leaf surface under illuminated conditions in the laboratory experiment. This in turn created a microenvironment where pH is elevated 0.2–0.4 to 0.3–0.5 units where ambient seawater pHT was lowered to 7.6 and 6.6. This period may be long enough and occurring often enough so that certain sessile or motile species can maintain calcification and/or normal metabolism. The diffusive boundary layer (DBL) in our experiments was 200 μm thick. The DBL may vary in nature depending on the leaf epiphyte cover or other factors that affect hydrodynamics. Based on our findings, we hypothesize that the thickness of the DBL and the periodical exposure to pH minima and pCO2 maxima, primarily determine the observed biotic responses at CO2 vents.
Several studies have demonstrated that the abundance and diversity of dominant motile invertebrates or meiofauna on marine vegetation are positively correlated with habitat complexity. This habitat complexity is measured by e.g. the biomass of epiphytic algae, percentage of leaf coverage and epiphyte structure, or seagrass density [37,73]. At sites where moderate CO2 seepage occurs at Basiluzzo Islet, more non-calcifying sessile epibionts and a higher overall sessile epibiotic species richness were detected compared to the reference site. This suggests a potentially more complex seagrass ecosystem when the surrounding seawater pH is decreased. This could explain the higher copepod diversity at CO2-R compared to REF. Conversely, no difference in the density and diversity of total meiofauna, nematodes in particular, was found between the sites. Also in terms of community composition, no major changes were detected on higher taxon level of meiofauna or on nematodes and copepods at the species level. These results suggest that meiofauna have little physiological intolerance to the low pH they are exposed to for at least part of the day. Garrard et al.  suggests the same for macro-invertebrates living in P. oceanica seagrass meadows at a CO2 vent off Ischia (pHT range: 8.2–7.1) after detecting increased abundances and unchanged diversity at the vent site. At Basiluzzo Islet, distinct differences in meiofauna community composition occurred between the leaves and rhizomes, which agrees with earlier findings on P. oceanica in non-acidified conditions near Ischia  and Corsica . Similarly, the highest total meiofauna density was found on the rhizomes and harpacticoid copepods dominated the meiofauna community in the leaf region [37,74]. Additionally, on the functional level of nematode communities, no site effect was detected at Basiluzzo Islet. Epistratum feeders, i.e. the group of nematodes that scrape the biofilm from surfaces or puncture and empty diatoms or microalgae , dominated at both sites, and on both seagrass parts with ≥ 65% relative abundance. This finding supports earlier suggestions of a close link with microbe and diatom based food chains which were based on the relationship between the abundance or biomass of nematodes and bacteria or diatoms on P. oceancia [37,75]. Similarly, based on stable isotope signatures, epiphytic biofilms have also been shown to contribute 70% of the diet of copepods living in close association with P. oceanica leaves . The success of invertebrates in tolerating a decrease in pH in seagrass meadows is suggested to be partly related to the high abundance of food. This may compensate the metabolic cost of OA tolerance .
The CO2 vent area at Basiluzzo Islet has allowed us to study the response of P. oceanica seagrass ecosystems to a level of OA that falls in the range of seawater pH reductions predicted to generally prevail by the year 2100. Despite the inherent short-comings of natural CO2 vents as analogues for OA and the limited replication in our study we hypothesize that the primary determinants of the observed biotic responses at CO2 vents may be the buffering capacity by the photosynthesis of P. oceanica, the thickness of the DBL, and the periodical exposure to pH minima and pCO2 maxima. Our findings suggest that with an average pH decrease of 0.1 units in seawater, changes in P. oceanica meadows will mainly occur on the level of the seagrass sessile epibiont community composition. However, P. oceanica seagrass productivity above the seafloor surface, net epibiont calcification, meiofauna community composition, abundance and function, and overall biodiversity will remain stable. The shorter leaves found at the vent sites, nevertheless, do seem to indicate an increased vulnerability to erosion, potentially by herbivores.
Although our study illustrated the relevance of an integrative ecosystem-based approach in studying long-term effect of OA at a natural analogue, it is likely that OA is not occurring in isolation and that biological responses may differ with the occurrence of additional stressors. Studying the effects of OA and potential interacting global stressors, such as temperature and oxygen, are restricted to short to medium-term laboratory experiments, but should ideally be supported by mesocosm (e.g. FOCE) and field studies at appropriate natural laboratories. Moreover, studies at natural laboratories should incorporate sufficient replication and monitoring of physico-chemical parameters on an annual scale to ensure that the biological responses observed during a short period reflect long-term dynamics of these parameters.
S1 Fig. Community structures of P. oceanica epibionts.
Non-metric multidimensional scaling (NMDS) ordination plots representing the community compositions of sessile epibiont taxa (A, 2D stress: 0.21), meiofauna at higer taxon level (B, 2D stress: 0.06), copepod species (C, 2D stress: 0.09) and nematode species (D, 2D stress: 0.12). The numbers provided to epibiont subsamples refer to the respective replicates (1 to 3). The percentage of similarity (SIMPER) is provided for each subset of samples.
S2 Fig. pH profiles in close vicinity of the Posidonia oceanica leaves (originating from CO2-R) when subjected to seawater acidification.
pHT profiles (averages ± s.d.) were measured on the convex side of P. oceanica leaves at seawater pH 7.6 (triangle), 6.6 (square) and 5.5 (circle). Measurements started from the leaf surface (distance = 0 μm), and were performed in light (black, n = 3) conditions.
S1 Table. Relative abundance of meiofauna, copepod, nematode and sessile epibiont taxa on the Posidonia oceanica rhizomes and leaves and in the vent site and reference site.
Meiofauna, copepod and nematode data are based on abundances, sessile epibiont data are based on presence/absence data.
The authors thank Stefanie Meyer for her assistance with processing the water samples during the sampling campaign at Panarea, and Boris Unger, Hanna Kuhfuss, Andreas Eich and Kristina Stemmer for support during counting, scanning and calculating the seagrass leafs. We also thank the Flanders Marine Institute (VLIZ), in particular Nathalie De Hauwere, for generating the bathymetric map, Dieter Wolf-Gladrow for his valuable feedback on the manuscript, and the anonymous reviewers for their time and valuable, constructive feedback.
- 1. Duarte CM. Seagrass Ecosystems. In: Levin SL, editor. Encyclopedia of Biodiversity. Academic Press, San Diego; 2001. pp. 254–268.
- 2. Waycott M, Duarte CM, Carruthers TJB, Orth RJ, Dennison WC, Olyarnik S, et al. Accelerating loss of seagrasses across the globe threatens coastal ecosystems. Proc Natl Acad Sci USA. 2009; 106: 12377–12381. pmid:19587236
- 3. Jordà G, Marbà N, Duarte CM. Mediterranean seagrass vulnerable to regional climate warming. Nat. Clim Chang. 2012; 2: 821–824.
- 4. Martin S, Rodolfo-Metalpa R, Ransome E, Rowley S, Buia MC, Gattuso JP, et al. Effects of naturally acidified seawater on seagrass calcareous epibionts. Biol Lett. 2008; 4: 689–692. pmid:18782731
- 5. IPCC. Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change. Stocker TF, Qin D, Plattner G-K, Tignor M, Allen SK, Boschung J, Nauels A, Xia Y, Bex V, Midgley PM, editors. Cambridge University Press; 2013.
- 6. Zimmerman RCA, Kohrs DGA, Steller DLB, Alberte RSA. Impacts of CO2 enrichment on productivity and light requirements of eelgrass. Plant Physiol. 1997; 115: 599–607. pmid:12223828
- 7. Invers O, Zimmerman RC, Alberte RS, Perez M, Romero J. Inorganic carbon sources for seagrass photosynthesis: an experimental evaluation of bicarbonate use in species inhabiting temperate waters. J Exp Mar Biol Ecol. 2001; 265: 203–217.
- 8. Jiang ZJ, Huang XP, Zhang JP. Effects of CO2 enrichment on photosynthesis, growth, and biochemical composition of seagrass Thalassia hemprichii (Ehrenb.) Aschers. J Integr Plant Biol. 2010; 52: 904–913. pmid:20883442
- 9. Alexandre A, Silva J, Buapet P, Björk M, Santos R. Effects of CO2 enrichment on photosynthesis, growth, and nitrogen metabolism of the seagrass Zostera noltii. Ecol Evol. 2012; 2: 2620–2630.
- 10. Campbell JE, Fourqourean JW. Mechanisms of bicarbonate use influence the photosynthetic carbon dioxide sensitivity of tropical seagrasses. Limnol Oceanogr. 2013; 58: 839–848.
- 11. Russell BD, Connell SD, Uthicke S, Muehllehner N, Fabricius KE, Hall-Spencer J. Future seagrass beds: Can increased productivity lead to increased carbon storage? Mar Poll Bull. 2013; 73: 463–469.
- 12. Apostolaki ET, Vizzini S, Hendriks IE, Olsen YS. Seagrass ecosystem response to long-term CO2 in a Mediterranean volcanic vent. Mar Environ Res. 2014; 99: 9–15. pmid:25081848
- 13. Ow YX, Collier CJ, Uthicke S. Response of three tropical seagrass species to CO2 enrichment, Mar Biol. 2015; 162: 1005–1017.
- 14. Cox TE, Schenone S, Delille J, Díaz-Castañeda V, Alliouane S, Gattuso J-P, Gazeau F. Effects of ocean acidification on Posidonia oceanica epiphytic community and shoot productivity. J Ecol. 2015; 103: 1594–1609.
- 15. Fabricius KE, Langdon C, Uthicke S, Humphrey C, Noonan S, De’ath G, et al. Losers and winners in coral reefs acclimatized to elevated carbon dioxide concentrations. Nat Clim Chang. 2011; 1: 165–169.
- 16. Thom RM. CO2-enrichment effects on eelgrass (Zostera marina L.) and bull kelp (Nereocystis luetkeana (Mert.) P. & R.). Water Air Soil Poll. 1996; 88: 383–391.
- 17. Palacios SL, Zimmermann RC. Response of eelgrass Zostera marina to CO2 enrichment: possible impacts of climate change and potential remediation of coastal habitats. Mar Ecol Prog Ser. 2007; 344: 1–13.
- 18. Hall-Spencer JM, Rodolfo-Metalpa R, Martin S, Ransome E, Fine M, Turner SM, et al. Volcanic carbon dioxide vents show ecosystem effects of ocean acidification. Nature. 2008; 454: 96–99. pmid:18536730
- 19. Marbà N, Duarte CM. Interannual changes in seagrass (Posidonia oceanica) growth and environmental change in the Spanish Mediterranean littoral zone. Limnol Oceanogr. 1997; 42: 800–810.
- 20. Guidetti P, Fabiano M. The use of lepidochronology to assess the impact of terrigenous discharges on the primary leaf production of the Mediterranean seagrass Posidonia oceanica. Mar Pollut Bull. 2000; 40: 449–453.
- 21. Mayot N, Boudouresque CF, Leriche A. Unexpected response of the seagrass Posidonia oceanica to a warm-water episode in the North Western Mediterranean Sea. C R Biol. 2005; 328: 291–296. pmid:15810553
- 22. Marbà N, Duarte CM. Mediterranean warming triggers seagrass (Posidonia oceanica) shoot mortality. Glob Chang Biol. 2010; 16: 2366–2375.
- 23. Hendriks IE, Olsen YS, Duarte CM. Light availability and temperature, not increased CO2, will structure future meadows of Posidonia oceanica. Aquat Bot. in press.
- 24. Arnold T, Mealey C, Leahey H, Miller AW, Hall-Spencer JM, Milazzo M, et al. Ocean acidification and the loss of phenolic substances in marine plants. PLoS ONE 7. 2012; e35107. pmid:22558120
- 25. Burnell OW, Russell BD, Irving AD, Connell SD. Seagrass response to CO2 contingent on epiphytic algae: indirect effects can overwhelm direct effects. Oecologia. 2014; 176: 871–882. pmid:25193313
- 26. Martínez-Crego B, Olivé I, Santos R. CO2 and nutrient-driven changes across multiple levels of organization in Zostera noltii ecosystems. Biogeosciences. 2014; 11: 7237–7249.
- 27. Schmalz RF, Swanson FJ. Diurnal variations in the carbonate saturation of seawater. J Sediment Petrol. 1969; 39: 255–267.
- 28. Invers O, Romero J, Pérez M. Effects of pH on seagrass photosynthesis: a laboratory and field assessment. Aquat Bot. 1997; 59: 185–194.
- 29. Semesi IS, Beer S, Björk M. Seagrass photosynthesis controls rates of calcification and photosynthesis of calcareous macroalgae in a tropical seagrass meadow. Mar Ecol Prog Ser. 2009; 382: 41–47.
- 30. Hendriks IE, Olsen YS, Ramajo L, Basso L, Steckbauer A, Moore TS, et al. Photosynthetic activity buffers ocean acidification in seagrass meadows. Biogeosciences. 2014; 11: 333–346.
- 31. Cox TE, Gazeau F, Alliouane S, Hendriks IE, Mahacek P, Le Fur A, Gattuso J-P. Effects of in situ CO2 enrichment on structural characteristics, photosynthesis, and growth of the Mediterranean seagrass Posidonia oceanica. Biogeosciences. 2016; 13: 2179–2194.
- 32. Kerrison P, Hall-Spencer JM, Suggett DJ, Hepburn LJ, Steinke M. Assessment of pH variability at a coastal CO2 vent for ocean acidification studies. Estuar Coast Shelf Sc. 2011; 94: 129–137.
- 33. Cox T, Nash M, Gazeau F, Déniel M, Legrand E, Alliouane S, Mahacek P, Le Fur A, Gattuso J-P, Martin S. Effects of in situ CO2 enrichment on Posidonia oceanica epiphytic community composition and mineralogy. Mar Biol. 2017; 164: 103.
- 34. Donnarumma L, Lombardi C, Cocito S, Gambi MC. Settlement pattern of Posidonia oceanica epibionts along a gradient of ocean acidification: an approach with mimics. Med Mar Sci. 2014; 15: 498–509.
- 35. Garrard SL, Gambi MC, Scipione MB, Patti FP, Lorenti M, Zupo V, et al. Indirect effects may buffer negative responses of seagrass invertebrate communities to ocean acidification. J Exp Mar Biol Ecol. 2014; 461: 31–38.
- 36. Hassenrück C, Hofmann LC, Bischof K, Ramette A. Seagrass biofilm communities at a naturally CO2-rich vent. Environ Microbiol Rep. 2015; 7: 516–525. pmid:25727314
- 37. Novak R. Spatial and seasonal distribution of the meiofauna in the seagrass Posidonia oceanica. Neth J Sea Res. 1982; 16: 380–388.
- 38. Nagelkerken I, Connell SD. Global alteration of ocean ecosystem functioning due to increasing human CO2 emissions. Proc Natl Acad Sci USA. 2015; 112: 13272–13277. pmid:26460052
- 39. Meyer S, Weber M, Guilini K, Bigalke N, Vielstädte L, Wenzhöfer F, et al. Investigation of benthic communities at submarine CO2 vents off Panarea Island (Italy): cruise report ECO2-3. Max Planck Institute for Marine Microbiology, Bremen, Germany / HYDRA Institute for Marine Sciences, Elba, Italy / UGent, Gent, Belgium / Geomar, Kiel, Germany; 2012.
- 40. Caramanna G, Fietzek P, Maroto-Valera M. Monitoring techniques of a natural analogue for sub-seabed CO2 leakages. Energy Procedia. 2011; 4: 3262–3268.
- 41. Grasshoff K, Ehrhardt M, Kremling K. Methods of seawater analysis. 2nd ed. Verlag Chemie, Weinheim, Weinheim/Deerfield Beach, Florida; 1983.
- 42. Lewis E, Wallace D. Program developed for CO2 system calculations. Carbon Dioxide Information Analysis Center, managed by Lockheed Martin Energy Research Corporation for the US Department of Energy Tennessee; 1998.
- 43. Takahashi T, Williams RT, Bos DL. Carbonate chemistry. In: Broecker WS, Spencer DW, Craig H, editors. GEOSECS Pacific Expedition, Volume 3, Hydrographic Data 1973–1974. National Science Foundation, Washington DC; 1982. pp. 77–83.
- 44. Zeebe RE, Wolf-Gladrow DA. CO2 in seawater: equilibrium, kinetics, isotopes. Elsevier Oceanography Series, Vol. 65, Elsevier, Amsterdam; 2001.
- 45. Hall PO, Aller RC. Rapid, small-volume, flow-injection analysis for ΣCO2 and NH4+ in marine and fresh waters. Limnol Oceanogr. 1992; 37: 1113–1119.
- 46. Edmond JM. High precision determination of titration alkalinity and total carbon dioxide content of sea water by potentiometric titration. Deep-Sea Res. 1970; 17: 737–750.
- 47. Lavigne H, Epitalon J-M, Gattuso J-P. seacarb: seawater carbonate chemistry with R. R package version 3.0. 2011. http://CRAN.R-project.org/package=seacarb.
- 48. Roy RN, Roy LN, Vogel KM, Porter-Moore C, Pearson T, Good CE, et al. The dissociation constants of carbonic acid in seawater at salinities 5 to 45 and temperatures 0 to 46°C. Mar Chem. 1993; 44: 249–267.
- 49. Dickson AG, Riley JP. The estimation of acid dissociation constants in seawater media from potentiometric titrations with strong base. I. The ionic product of water—KW. Mar Chem. 1979; 7: 89–99.
- 50. Dickson AG, Goyet C. Handbook of methods for the analysis of the various parameters of the carbon dioxide system in seawater, version 2. ORNL/CDIAC-74; 1994.
- 51. Dickson AG. Standard potential of the reaction: AgCl (s) + ½ H2 (g) = Ag (s) + HCl (aq), and the standard acidity constant of the ion HSO4- in synthetic seawater from 273.15 to 318.15 K. J. Chem Thermodyn. 1990; 22: 113–127.
- 52. Abramoff MD, Magalhaes PJ, Ram SJ. Image Processing with ImageJ. Biophotonics Int. 2004; 11: 36–42.
- 53. Watson DJ. Comparative physiological studies on the growth of field crops: I. Variation in net assimilation rate and leaf area between species and varieties and within and between years. Ann Bot. 1947; 11: 41–76.
- 54. Heip C, Herman P, Soetaert K. Indices of diversity and evenness. Océanis. 1998; 24: 61–87.
- 55. Hurlbert SH. The nonconcept of species diversity: a critique and alternative parameters. Ecology. 1971; 52: 577–586.
- 56. Wieser W. Die Beziehung zwischen Mundhöhlengestalt, Ernährungsweise und Vorkommen bei freilebenden marinen Nematoden. Eine ökologisch-morphologische Studie. Ark Zool. 1953; 4: 439–483.
- 57. Revsbech NP. An oxygen microelectrode with a guard cathode. Limnol Oceanogr. 1989; 34: 474–478.
- 58. de Beer D. Potentiometric microsensors for in situ measurements in aquatic environments. In: Buffle J, Horvai G, editors. In situ monitoring of aquatic systems: chemical analysis and speciation. Wiley & Sons: Chichester, UK; 2000. pp 161–194.
- 59. R Development Core Team. R: A Language and Environment for Statistical Computing. Vienna, Austria: the R Foundation for Statistical Computing. 2011. http://www.R-project.org/.
- 60. Anderson MJ, Gorley RN, Clarke KR. PERMANOVA+ for PRIMER: Guide to Software and Statistical Methods. PRIMER-E, Plymouth; 2008.
- 61. Duarte CM, Hendriks IE, Moore TS, Olsen YS, Steckbauer A, Ramajo L, et al. Is Ocean Acidification an Open-Ocean Syndrome? Understanding Anthropogenic Impacts on Seawater pH. Estuar Coast. 2013; 36: 221–236.
- 62. Johnson VR, Brownlee C, Rickaby REM, Graziano M, Milazzo M, Hall-Spencer JM. Responses of marine benthic microalgae to elevated CO2. Mar Biol. 2013; 160: 1813–1824.
- 63. Welker C, Sdrigotti E, Covellic S, Faganeli J. Microphytobenthos in the Gulf of Trieste (Northern Adriatic Sea): relationship with labile sedimentary organic matter and nutrients. Estuar Coast Shelf Sci. 2002; 55: 259–273.
- 64. Brodersen KE, Lichtenberg M, Paz L-C, Kühl M. Epiphyte-cover on seagrass (Zostera marina L.) leaves impedes plant performance and radial O2 loss from the below-ground tissue. Front Mar Sci. 2015; 2: 58.
- 65. Balestri E, Cinelli F, Lardicci C. Spatial variation in Posidonia oceanica structural, morphological and dynamic features in a northwestern Mediterranean coastal area: A multi-scale analysis. Mar Ecol Prog Ser. 2003; 250: 51–60.
- 66. Vergés A, Alcoverro T, Romero J. Plant defenses and the role of epibiosis in mediating within-plant feeding choices of seagrass consumers. Oecologia. 2011; 166: 381–390. pmid:21053016
- 67. Peirano A, Niccolai I, Mauro R, Bianchi CN. Seasonal grazing and food preference of herbivores in a Posidonia oceanica meadow. Sci Mar. 2001; 65: 367–374.
- 68. Milliman JD. Marine Carbonates. Berlin-Heidelberg-New York: Springer; 1974.
- 69. Cigliano M, Gambi MC, Rodolfo-Metalpa R, Patti FP, Hall-Spencer JM. Effects of ocean acidification on invertebrate settlement at volcanic CO2 vents. Mar Biol. 2010; 157: 2489–2502.
- 70. Kroeker KJ, Micheli F, Gambi MC, Martz TR. Divergent ecosystem responses within a benthic marine community to ocean acidification. Proc Natl Acad Sci USA. 2011; 108: 14515–14520. pmid:21844331
- 71. Porzio L, Hall-Spencer JM, Buia MC. Effects of ocean acidification on macroalgal communities. J Exp Mar Biol Ecol. 2011; 400: 278–287.
- 72. Koch M, Bowes G, Ross C, Zhang X-H. Climate change and ocean acidification effects on seagrasses and marine macroalgae. Glob Change Biol. 2013. 19: 103–132.
- 73. Mascart T, Lepoint G, Deschoemaeker S, Binard M, Remy F, De Troch M. Seasonal variability of meiofauna, especially harpacticoid copepods, in Posidonia oceanica macrophytodetritus accumulations. J Sea Res. 2015; 95: 149–160.
- 74. Mascart T, Lepoint G, De Troch M. Meiofauna and harpacticoid copepods in different habitats of a Mediterranean seagrass meadow. J Mar Biol Assoc UK. 2013; 93: 1557–1566.
- 75. Danovaro R. Detritus-bacteria-meiofauna interactions in a seagrass bed (Posidonia oceanica) of the NW Mediterranean. Mar Biol. 1996; 127: 1–13.