Layout 1 Introduction While in marked decline worldwide, seagrasses are important structural species considered to be “ecosystem engineers”, hosting rich and abundant benthic assemblages and often representing a source of biodiversity for adjacent habitats (e.g. Surugiu et al., 2021; Barnes et al., 2022; García-Gómez et ARTICLE Macrobenthos of lagoon ecosystems: a comparison in vegetated and bare sediments Paolo Magni1,*, Maria Flavia Gravina1,2,3 1Consiglio Nazionale delle Ricerche, Istituto per lo Studio degli Impatti Antropici e Sostenibilità in Ambiente Marino (CNR-IAS), Località Sa Mardini, Torregrande, Oristano, Italy; 2Dipartimento di Biologia, Università di Roma “Tor Vergata”, Rome, Italy; 3Consorzio Nazionale Interuniversitario per le Scienze del Mare (CoNISMa), Rome, Italy ABSTRACT The classic paradigm, not always unequivocal though, that seagrass macrobenthic abundance and diversity are greater than those in adjacent unvegetated areas, was tested in a Mediterranean lagoon for which evidence is lacking. We compared the community structure and species composition of macrobenthic assemblages in three distinct areas of the Mistras Lagoon (Sardinia, Italy) dominated by i) the seagrass Cymodocea nodosa (Cym), ii) mixed-macrophyte/detritus (Mix), and iii) unvegetated sediments (Unv). Samplings were conducted in each area twice in spring (April and May) and twice in autumn (October and November) 2010. Multivariate analyses showed significant differences among the three areas and a marked separation between the two vegetated (Cym and Mix) and the unvegetated (Unv) areas. The top discriminating species, indicated by SIMPER analysis, were characterized by direct development lacking free living larvae. They included marine species, e.g. Cerithium lividulum, Microdeutopus gryllotalpa, Loripes orbiculatus and Gammarus aequicauda, at Cym vs Mix, whereas the pair discrimination of the areas Cym vs Unv and Mix vs Unv mostly depended on the marine Abra tenuis, the opportunistic Chironomidae and Capitella capitata, and the brackish Hydrobia acuta. Both vegetated areas showed a higher species number than the unvegetated area. Differently, abundance was higher in Unv than in Cym and Mix in all sampling dates, most notably in spring likely because of reproductive events, due to both opportunistic (C. capitata, chironomids) and halolimnobic (H. acuta, Cerastoderma glaucum) species. Overall, the present study demonstrated the importance of biological factors in structuring the macrobenthic assemblages of the Mistras Lagoon. Habitat- forming phanerogams hosted species-rich assemblages, species-specific reproductive cycles and adaptive strategies contributed to drive species colonization and abundance distribution. Both vegetated and unvegetated patches were found to contribute greatly to the local-scale heterogeneity of the habitat, highlighting the importance of coastal lagoons as hotspots for benthic biodiversity. For these reasons, lagoons must be considered major targets for conservation measures. Corresponding author: Paolo Magni, Consiglio Nazionale delle Ricerche, Istituto per lo Studio degli Impatti Antropici e Sosteni- bilità in Ambiente Marino (CNR-IAS), Località Sa Mardini, Tor- regrande, 09170, Oristano, Italy. E-mail: paolo.magni@cnr.it Key words: seagrass; benthic macroinvertebrates; biodiversity, coastal lagoons. Authors’ contributions: all the authors have made a substantive in- tellectual contribution. All the authors have read and approved the final version of the manuscript and agreed to be held accountable for all aspects of the work. Conflict of interest: the authors declare no potential conflict of in- terest. Funding: this publication is a result of MARBEFES (MARine Bio- diversity and Ecosystem Functioning leading to Ecosystem Serv- ices) project, funded by the European Union under the Horizon Europe Programme, “HORIZON-CL6-2021-BIODIV-01” Theme, Grant Agreement no. 101060937 [marbefes.eu]. Availability of data and materials: all data generated or analyzed during this study are included in this published article. Acknowledgments: we thank Giovanni Fenzi and Andrea Satta for their work in the field and in the laboratory and Messrs Manca of the Cooperativa Pescatori e Molluschicoltori of Mistras Lagoon for permission to conduct research on the Mistras Lagoon system. We gratefully acknowledge two anonymous Reviewers for provid- ing valuable comments to the original manuscript. Received: 29 December 2022. Accepted: 8 March 2023. Publisher’s note: all claims expressed in this article are solely those of the authors and do not necessarily represent those of their affil- iated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher. ©Copyright: the Author(s), 2023 Licensee PAGEPress, Italy Advances in Oceanography and Limnology, 2023; 14:11124 DOI: 10.4081/aiol.2023.11124 This work is licensed under a Creative Commons Attribution- NonCommercial 4.0 International License (CC BY-NC 4.0). No n- co mm er cia l u se on ly Macrobenthos of lagoon ecosystems: a comparison in vegetated and bare sediments 11 al., 2022). In intertidal and subtidal estuarine systems, several studies have been conducted on biodiversity differentials between seagrass and bare sediments, highlighting the prominent role of seagrass in enhancing the benthic biodiversity (Barnes and Barnes, 2012; Barnes, 2013a; Surugiu et al., 2021). Other studies have compared macrobenthic assemblages in different seagrass species (Leopardas et al., 2014; Barnes, 2020a, 2020b;), confirming the importance of the architectural complexity of seagrass meadows in supporting higher abundance, species richness and diversity. This said, it should also be acknowledged that the classic paradigm that seagrass supports macrobenthic assemblages with different composition and with greater abundance and diversity with respect to adjacent areas without this cover is not always unequivocal (see Nakaoka, 2005; Barnes, 2022 and references therein). For instance, departure from this paradigm is found in some intertidal and estuarine systems in relation to local environmental conditions and gradients, assemblage composition and biomass, predatory behaviour and ontogenetic diet shift, latitudinal gradients, rhizome biomass and seasonal cycle of macrophyte vegetation, etc. (see Barnes and Barnes, 2014; Włodarska-Kowalczuk et al., 2014; Magni et al., 2017 and references therein). These studies highlight the complexity of animal-plant interaction in seagrass ecology, which requires a detailed analysis of the local environmental peculiarities. Most importantly, this suggests that both vegetated and unvegetated areas need to be considered mutually for conservation purposes to prevent the potential loss of local biodiversity and to understand the ecological processes structuring the benthic assemblages in these complex habitats (Barnes, 2013b; Barnes and Hamylton, 2016; Barnes 2021). It has been shown that in coastal lagoons continuous stands of seagrass offer stability to well structured and taxa-rich macrozoobenthic assemblages (Sfriso et al., 2001; Tagliapietra et al., 2016). This can also be due to a series of additional factors, such as the shading effect and protection of seagrass meadows from high temperature in summer, especially in shallow waters, and a better oxygenation of the water and sediments in the presence of the seagrass (Tagliapietra et al., 2016). Moving away from the meadow to the bare sediment, the temporal fluctuations of the community structure become wider and the assemblages reduce their resistance and resilience, due to worsening or changing environmental conditions, such as increased sediment organic enrichment (Magni et al., 2015; Leopardas et al., 2016) or reduced structural complexity of the habitat (Bachelet et al., 2000; Leopardas et al., 2014), respectively. However, very few studies looking at seagrass vs bare sediment benthic biodiversity are available in coastal lagoons (Magni et al., 2017; Barnes, 2022) and, to our knowledge, no studies have compared the community structure of macrobenthic assemblages in multiple habitats, including vegetated and non-vegetated areas, within a Mediterranean lagoon system. The present study was carried out in one of these Mediterranean lagoons, the Mistras Lagoon, located along the western coast of Sardinia. We aimed to evaluate the differences in macrobenthic assemblages between seagrass-dominated and unvegetated seabeds by analyzing and comparing the community structure and species composition of macrobenthic assemblages in three distinct areas, within the same lagoon, characterized by the dominance of i) the seagrass Cymodocea nodosa, ii) mixed-macrophyte/detritus, and iii) unvegetated sediments. Materials and Methods The Mistras Lagoon The Mistras Lagoon is a semi-enclosed coastal barrier system (~4.5 km2, mean depth <1 m) located on the western coast of Sardinia Island (Italy, Western Mediterranean Sea) (Figure 1). Connected to the Gulf of Oristano through its mouth, it is part of the so-called Oristano Lagoon-Gulf (OLG) system which comprises the Gulf of Oristano (150 km2; maximum depth 24 m), several salt marshes and the lagoons of Cabras, Santa Giusta, S’Ena Arrubia, Corru S’Ittiri and San Giovanni- Marceddì (Magni et al., 2008a). These lagoons are characterized by eutrophic conditions, with the dominance of small sized phytoplankton species and the occurrence of harmful algal species (e.g. Cabras Lagoon) or the major presence of macrophyte (e.g. S’Ena Arrubia and Santa Giusta lagoons) (Padedda et al., 2012; Satta et al., 2014; Pulina et al., 2018). Sediments are generally organically enriched, leading to impoverished benthic communities (Magni et al., 2005; Brundu and Magni, 2021) and occasional dystrophic events (Murenu et al. 2004; Magni et al., 2008b). On the contrary, the Mistras Lagoon is characterized by oligotrophic waters (Sechi, 1982; Specchiulli et al., 2018) and sandy sediments with a low content of Organic Matter (OM), especially in the unvegetated areas, owing to negligible riverine discharge and low sediment supply (Pascucci et al., 2018). Major pollution sources are absent, leading a low-pressure index compared to other Sardinian lagoons (Gravina et al., 2020). During dry periods with high evaporation rates in summer, satellite ponds and natural salt flats are formed in shallower and confined areas where salinity can exceed 50 psu. Fishing activity in the Mistras Lagoon is carried extensively using a capture system made by workers and fish traps named “bertovelli”. The Mistras Lagoon is listed in the Ramsar Convention on Wetlands (D.M. 04/03/82) and is a Special Protection Area under the Nature 2000 Network (ZPS ITB034006, SIC ITB030034), thus subjected to local protection measures. In the lagoon there are meadows of various species of aquatic phanerogams, including the dominant C. nodosa followed by Zostera noltii and Ruppia maritima, and several macroalgal species, such as Gracilaria sp., Ulva sp. and Enteromorpha sp., (Camarda, 1995). The lagoon is surrounded by luxuriant glasswort dominated by Salicornia fruticosa, with the presence of Obione portulacoides, Salsola soda, Limonium vulgare, Aeluropus litoralis, Arthrocnemum glaucum, Atriplex sp. and Scirpus sp. (Fenzi, 2013). Study areas, sampling strategy and sample treatment Three different sampling areas were selected based on the morphology of the lagoon and the presence and distribution of vegetation. In particular, the seagrass area (hereafter, Cym) was located in the southwest sector of the lagoon and was dominated by C. nodosa meadows, with scattered patches of Z. noltii. The No n- co mm er cia l u se on ly P. Magni and M.F. Gravina12 second vegetated area, named Mixed (hereafter, Mix), was located closer to the lagoon’s mouth and was characterized by sparse patches of both C. nodosa and Z. noltii, the red alga Gracilaria sp. developing from autumn to spring, and vegetation-derived detritus. The third area was unvegetated (hereafter, Unv) and was located in the north-eastern and most confined sector of the lagoon. This area was characterized by sandy sediments and low OM content, without vegetation except sporadic patches of R. maritima and Enteromorpha sp. in spring (Figure 1). Sampling was conducted in each of the three selected areas, i.e. Cym, Mix and Unv, twice in spring, i.e. April and May, and twice in autumn, i.e. October and November, 2010. In each area and each date, 3 stations were randomly selected 10’s meters apart and at each station four replicate samples, meters apart, were randomly collected using a box-corer (170 cm2, penetration depth 15 cm), for a total of 144 samples. Each sample was sieved on a 0.5 mm mesh-size and the residue was preserved in a 4% buffered formaldehyde solution for the determination of macrozoobenthos. In the laboratory, macrobenthic organisms from each sample were sorted, counted and identified to the species level whenever possible using a stereomicroscope. Data analysis Macrozoobenthic data were analyzed by means of ordination technique Non-Metric Multidimensional Scaling (nMDS) using the mean abundance of four replicates for each of the three stations in each area and each sampling date based on (dis)similarity Bray-Curtis measure. Significance tests for differences among spatial and temporal replicates within and between areas were performed by means of the ANOSIM test. The SIMPER test was carried out in order to evaluate the taxa contributing most to (dis)similarity between and within temporal and spatial groups of replicates. In order to analyze spatiotemporal changes in the macrobenthic assemblages, the following biotic variables were computed: species richness (number of species), total abundance (N, number of individuals m-2), diversity (H’), calculated by the Shannon-Wiener index, and evenness by the Pielou index. Significance of the biotic variables was tested by means of a two-way PERMANOVA. Results Macrobenthic assemblages In the present study, 6524 individuals belonging to 33 taxa were classified, of which 39.9 % were molluscs, followed by gastropods (16.1%), crustaceans (15,5%) and polychaetes (14.9%). The ten most abundant taxa, accounting for 88.8% of the total abundance, included the bivalves Abra tenuis (26.3%), Cerastoderma glaucum (6.6%) and Loripes orbiculatus (4.6%), the gastropods Cerithium lividulum (8.6%) and Hydrobia acuta (7.6%), the polychaetes Capitella capitata (7.7%) and Perinereis rullieri (3.9%), the amphipods Microdeutopus gryllotalpa (6.5%) and Gammarus aequicauda (4.4%), and Chironomidae (12.7%). Figure 1. Location of the Mistras Lagoon and the three study areas: Cymodocea nodosa-dominated (Cym) area, mixed-macrophyte/de- tritus (Mix) area, and unvegetated sediment (Unv) area. No n- co mm er cia l u se on ly Macrobenthos of lagoon ecosystems: a comparison in vegetated and bare sediments 13 Multivariate analyses The one-way ANOSIM test revealed significant differences among the three areas and a marked separation between the two vegetated (Cym and Mix) and the unvegetated (Unv) areas, with a global R=0.74, and a within and between mean ranks of 156.3 and 388.5 (p=0.0001), respectively. The nMDS ordination model confirmed this separation, with Unv being most isolated irrespective of seasonal changes. There was instead some overlap between Cym and Mix as a result of a high similarity between the two biocoenoses (Figure 2). Furthermore, on a seasonal basis, the two autumn samplings at Unv were more dispersed than those at Cym and Mix, due to a major reduction of both abundance and species richness and diversity. The percentage contribution to the average dissimilarity between the sampling areas, measured by the SIMPER analysis, was reported for the top discriminating species accounting for up to 95% of the cumulative percentage (Table 1). These latter included marine species, e.g. C. lividulum, M. gryllotalpa, L. orbiculatus and G. aequicauda, at Cym vs Mix, whereas the pair discrimination of the areas Cym vs Unv, and Mix vs Unv mostly depended on the marine species Abra tenuis, the opportunistic taxa, e.g. Chironomidae and C. capitata, and the brackish species H. acuta. Two-way PERMANOVA conducted on the biotic variables showed significant differences in the species richness, abundance, diversity and evenness between areas and months (Table 2). Figure 2. Non-Metric Multidimensional Scaling (nMDS) ordi- nation plot of macrozoobenthic abundance data for the three in- vestigated areas of the Mistras Lagoon: white symbols refer to Cym, gray to Mix and black to Unv. Circles indicate April, in- verted triangles May, squares October and triangles November (each symbol represents the mean of 4 replicates, see M&M) (Stress=0.1). Figure 3. Mean values (n=12 ± standard error, SE) of the total abundance of individuals (N. ind. 170 cm–2), species richness (S), Shan- non-Wiener diversity index and Evenness Pielou index in the three study areas in the Mistras Lagoon. A, April; M, May; O, October; N, November. No n- co mm er cia l u se on ly P. Magni and M.F. Gravina14 Table 1. List of the top discriminating taxa with contribution more than 5% by SIMPER analysis at the three investigated areas, C. no- dosa-dominated (Cym), mixed-macrophyte/detritus (Mix) and Unvegetated (Unv). Species category: O, opportunistic; C, common in coastal sheltered waters; B, brackish; M, properly marine. Cym vs Mix Average Contribution Cumulative Average Average dissimilarity % % abundance abundance Cym Mix Cerithium lividulum M 8.39 13.71 13.71 7.77 3.88 Microdeutopus gryllotalpa C 8.36 13.67 27.37 6.46 2.23 Loripes orbiculatus C 6.98 11.41 38.78 3.83 2.40 Gammarus aequicauda C 6.50 10.62 49.41 3.50 2.33 Abra tenuis M 4.74 7.74 57.15 1.63 1.67 Sphaeroma serratum M 4.60 7.52 64.67 4.00 0.25 Nainereis laevigata C 3.93 6.42 71.08 2.19 0.56 Perinereis rullieri M 3.66 5.98 77.06 2.06 3.29 Abra segmentum C 3.59 5.86 82.92 1.08 2.15 Chironomidae O 2.11 3.45 86.37 0.19 1.04 Tanais dulongii O 1.33 2.17 88.54 1.15 0.06 Cyathura carinata B 0.94 1.53 90.07 0.73 0.00 Gibbula adansoni M 0.85 1.39 91.46 0.08 0.42 Cerastoderma glaucum B 0.79 1.30 92.76 0.23 0.35 Phylo norvegicus C 0.61 1.00 93.76 0.65 0.02 Dispio uncinata M 0.56 0.91 94.67 0.00 0.25 Pirenella conica B 0.51 0.84 95.51 0.00 0.27 Cym vs Unv Average Contribution Cumulative Average Average dissimilarity % % abundance abundance Cym Mix Abra tenuis M 26.91 27.93 27.93 1.63 32.40 Chironomidae O 16.39 17.01 44.94 0.19 16.00 Capitella capitata O 8.23 8.54 53.48 0.02 10.30 Hydrobia acuta B 8.15 8.46 61.94 0.00 10.30 Cerithium lividulum M 8.11 8.42 70.36 7.77 0.00 Cerastoderma glaucum B 6.08 6.31 76.67 0.23 8.33 Microdeutopus gryllotalpa C 5.50 5.71 82.38 6.46 0.13 Loripes orbiculatus C 3.17 3.29 85.68 3.83 0.00 Sphaeroma serratum M 2.95 3.06 88.74 4.00 0.00 Gammarus aequicauda C 2.64 2.74 91.47 3.50 0.17 Perinereis rullieri M 2.34 2.43 93.90 2.06 0.00 Nainereis laevigata C 2.05 2.13 96.03 2.19 0.00 Mix vs Unv Average Contribution Cumulative Average Average dissimilarity % % abundance abundance Cym Mix Abra tenuis M 28.62 30.90 30.90 1.67 32.40 Chironomidae O 17.80 19.22 50.13 1.04 16.00 Hydrobia acuta B 8.86 9.57 59.70 0.00 10.30 Capitella capitata O 8.72 9.41 69.11 0.19 10.30 Cerastoderma glaucum B 6.51 7.03 76.14 0.35 8.33 Cerithium lividulum M 4.75 5.13 81.27 3.88 0.00 Perinereis rullieri M 4.00 4.32 85.59 3.29 0.00 Gammarus aequicauda C 2.81 3.03 88.62 2.33 0.17 Abra segmentum C 2.72 2.94 91.56 2.15 0.00 Microdeutopus gryllotalpa C 2.57 2.78 94.34 2.23 0.13 Loripes orbiculatus C 2.32 2.51 96.85 2.40 0.00 No n- co mm er cia l u se on ly Macrobenthos of lagoon ecosystems: a comparison in vegetated and bare sediments 15 Univariate analyses The mean number of species was highest in Cym in April (17.3±0.6) and lowest in Unv in November (4.0±0.0) (Figure 3). Overall, both vegetated areas (Cym and Mix) showed a higher species richness than Unv, yet with a progressive reduction from April to November. In particular, species number was significantly higher in spring and autumn in Cym, and significantly higher in April than in the other three dates in Mix. Instead, species number was constantly lowest in Unv where it did not show significant differences between dates. Similarly, the diversity index H’ was generally higher in Cym and Mix than in Unv which showed no significant differences between dates. Differently, the mean abundance was highest in Unv in May (144.7±32.5) and lowest in Mix in October (11.7±2.5), with a marked reduction in autumn especially in Unv. Overall, abundance was higher in Unv than in Cym and Mix in all sampling dates. Accordingly, evenness was lowest in Cym in April and May, and highest in Unv in April (0.87±0.08), while Mix showed the highest eveness values in May, October and November (Figure 3). Discussion The present study demonstrated that both seagrass-dominated and mixed/detritus beds hosted a higher macrobenthic species richness and diversity than unvegetated sediments, the latter being characterized by very different assemblages in terms of community structure and species composition. On the contrary, the abundance was considerably higher in the bare sediments than in the vegetated areas. Peaks in the number of individuals in Unv were due to both opportunistic (Capitella capitata, chironomids) and halolimnobic (Hydrobia acuta, Cerastoderma glaucum) species, which most notably occurred in spring, likely as a result of reproductive events. Notwithstanding an overall decline from spring to autumn, the abundance remained higher in Unv than in Cym and Mix, indicating that the unvegetated areas sustain large populations. The lowest evenness values recorded in Unv in autumn indicated that variations in the community structure are mainly due to changes in the abundance rather than in the number of species which in turn remained similar over time. Similarly, Barnes (2021) found that in the Knysna estuarine bay (South Africa) the endangered dwarf-eelgrass, Zostera capensis, supports fewer animals than adjacent unvegetated areas. This reportedly unusual situation was explained by a probable of lack of bioturbators in bare sediments, whose activity may cause a significant reduction of surface and near- surface animals. In the Mistras Lagoon, marked differences between areas were also indicated by the seasonal changes of the macrobenthic assemblages. As highlighted by the nMDS plot, both spring and autumn replicates in the two vegetated areas were spatially very close one another, with a marked distinction between the two seasons. The observed decline in the number of species and individuals at Cym and Mix over the months can be explained by the expected seasonal cycles of lagoonal environments (Magni et al., 2005; 2015; Tagliapietra et al., 2016), including the extent of algae and plant cover providing important food sources, especially in spring, for many benthic species (Magni et al., 2008b; Como et al., 2012; Zheng et al., 2020). Differently, the autumn replicates of the unvegetated area appeared to be much more separated one another than the spring ones. The larger spread of the autumn replicates in Unv than in Cym and Mix can be related to a greater suffering and a slower recovery of the benthic assemblage in Unv from environmental stress occurring in the Mistras Lagoon in summer, when a significant increase in temperature and salinity is recorded (unpublished data). On the contrary, the function of seagrass meadows in offering protection from “natural stress” was clearly expressed by the benthic assemblage at Cym that displayed greater spatial and temporal stability of abundance and composition (Tagliapietra et al., 2016; see also Barnes, 2023), and high levels of resilience to the harsh environmental conditions found in summer. In the Mistras Lagoon, faunal differences in terms of species composition were especially due to several marine species and species typical of coastal sheltered waters (Gravina et al., 2020), which were found much more numerous in the two vegetated areas. As indicated by the SIMPER analysis, most of the species that contributed up to the cumulative 95% were amphipods (M. gryllotalpa, G. aequicauda), isopods (S. serratum, C. carinata), tanaids (T. dulongii) and two bivalves, L. lacteus and A. tenuis (see Currás and Mora, 1996; Holmes et al., 2004), all being characterized by direct development lacking free living Ta bl e 2. R es ul ts o f t w o- w ay P E R M A N O V A a m on g ar ea s an d da te s (p er m ut at io n N : 9 99 9, d f = D eg re es o f F re ed om ). S pe ci es R ic hn es s A bu nd an ce D iv er si ty (H ’) E ve nn es s (P ie lo u) So ur ce S um o f d f M ea n P se ud o p S um o f df M ea nP se ud o p S um o f df M ea n P se ud o p S um o f df M ea n P se ud o p s qu ar es sq ua re -F s qu ar es s qu ar e - F sq ua re s s qu ar e - F sq ua re s s qu ar e -F A re as 1. 00 2 0 .5 0 4 2. 97 0 .0 00 1 1 .5 0 2 0 .7 5 2 2. 23 0 .0 00 1 0 .4 3 2 0. 21 3 2. 87 0. 00 01 0 .0 4 2 0 .0 2 5. 21 0 .0 09 4 D at es 0. 43 3 0 .1 4 1 2. 40 0 .0 00 3 1 .3 8 3 0 .4 6 1 3. 63 0 .0 00 1 0 .2 6 3 0. 09 1 3. 45 0. 00 01 0 .0 4 3 0 .0 1 3. 58 0 .0 23 7 In te ra ct io n 0 .1 8 6 0 .0 3 2. 56 0 .0 38 6 1 .0 8 6 0 .1 8 5 .3 7 0. 00 01 0 .0 9 6 0. 01 2 .2 6 0 .0 58 1 0 .0 5 6 0 .0 1 2. 27 0 .0 61 9 R es id ua l 0. 28 2 4 0 .0 1 0 .8 1 2 4 0. 03 0 .1 6 2 4 0 .0 1 0 .0 9 2 4 0 .0 0 To ta l 1. 89 3 5 4. 76 3 5 0. 93 3 5 0 .2 2 3 5 No n- co mm er cia l u se on ly P. Magni and M.F. Gravina16 larvae. They also included the nonplanktotrophic gastropod C. lividulum which is able of crawling directly on the bottom after hatching (Garilli and Galletti, 2006; Evangelisti et al., 2016). Only a few of the main discriminating species revealed by the SIMPER analysis were opportunistic and brackish/halolimnobic species which contributed to characterize the unvegetated area. They comprised chironomids, C. capitata, H. acuta and C. glaucum, the latter two species found >10 times more abundant in the unvegetated area. Marked differences between areas were also highlighted by the distribution of two congeneric species, Abra segmentum and Abra tenuis, which relates to their different biological cycles and larval phases. In particular, A. segmentum, characterized by a long-life pelagic larval phase typical of bivalves, was dominant and exclusive to the two vegetated areas. This suggests that the nearby marine environment can be a source of new recruits, while too confined unvegetated areas may be hardly colonisable by this bivalve. Differently, A. tenuis was dominant in very high numbers in the unvegetated area indicating its ability to maintain a considerable population in confined areas due to its direct development which does not require the marine environment as an external donor (Holmes et al., 2004). A similar adaptive strategy was shown by opportunistic (C. capitata, chironomids) and halolimnobic (H. acuta, C. glaucum) species which were able to occupy very confined habitats due to their reproduction and adaptation modes. These results show that only a portion of marine species occurring in the Mistras Lagoon can be sustained by larval supply from the marine environment, while another pool of species lacking direct connectivity with the sea, but characterized by direct development, is able to self-sustain their populations. The results of the present study add new knowledge to the spatial and temporal distribution of benthic assemblages in vegetated coastal environments, contributing to the understanding of differences (or similarities) between seagrass and bare sediments. It follows conspicuous research on macrobenthic seagrass ecology conducted in the intertidal zone (e.g., Barnes and Barnes, 2012, 2014; Barnes, 2020a, 2020b), with fewer comparative studies carried out in coastal marine habitats (Włodarska-Kowalczuk et al., 2014; Surugiu et al., 2021) and coastal lagoons (Magni et al., 2017; Barnes, 2022; Hu et al., 2022). Furthermore, although several ecological studies have been carried out on macrophyte-derived detritus (e.g. see review by Lepoint and Hyndes, 2022), few works have examined the benthic assemblages in leaf litter and respective mother plants in the same environmental settings (Ólafsson et al., 2013). One of these studies was conducted outside the Mistras Lagoon in the Gulf of Oristano (Como et al., 2008). The authors found that two seagrasses, Posidonia oceanica and Cymodocea nodosa, had similar species composition and differed from a third habitat, i.e. leaf litter beds, for the exclusive presence of hard-bottom species, such as the tunicate Phallusia fumigata, and seagrass-associated species, such as the polychaete Syllis garciai and the decapod Paguristes syrtensis. In contrast, the leaf litter sediments were characterized by the bivalves A. alba and C. glaucum, not found in the seagrasses, and by the bivalves L. lacteus and Ruditapes decussatus, and the polychaete Prionospio multibranchiata, exclusive to leaf litter beds. Conclusions The present study demonstrated the importance of biological factors in structuring the macrobenthos of the Mistras Lagoon in both vegetated and non-vegetated areas. The presence of habitat-forming phanerogams, such as C. nodosa, Z. noltii and Ruppia maritima, sustained well-structured and diverse benthic assemblages that were lacking in bare sediments. Our results agree with those reported from other lagoons and sheltered coastal areas worldwide (Bachelet et al., 2000; Sfriso et al. 2001; Blanchet et al., 2005; Brito et al., 2005; Yamada et al., 2007; Barnes and Barnes, 2012; Magni et al., 2015, Tagliapietra et al., 2016; Magni et al., 2017; Surugiu et al., 2021). Furthermore, the life-history traits of the species, such as species-specific reproductive cycles and adaptive strategies (Nonnis Marzano et al., 2010; Cardone et al., 2014; Gravina et al., 2020), were important drivers for the colonization of benthic species in the confined areas. We anticipate that, in addition to biological factors, environmental constraints, such as confinement, are also to be considered to fully explore the reasons for differences in the benthic community of the Mistras Lagoon. In particular, the marked isolation of the unvegetated area and the limited water exchange with the sea and other areas of the lagoon, may enhance the “natural stress” due to larger variations in abiotic variables, e.g. salinity. This could affect both the flora and the fauna of the lagoon and will be the focus of a subsequent study. Finally, the present study demonstrates that both vegetated and unvegetated patches contribute greatly to the local-scale heterogeneity of the lagoon ecosystems, highlighting the peculiarity and importance of these habitats as hotspots of benthic biodiversity. For these reasons, coastal lagoons must be considered major targets for conservation measures. References Bachelet G, de Montaudouin X, Auby I, Labourg PJ, 2000. Seasonal changes in macrophyte and macrozoobenthos assemblages in three coastal lagoons under varying degrees of eutrophication. ICES J. Mar. Sci. 57:1495-506. Barnes RSK, 2013a. Distribution patterns of macrobenthic biodiversity in the intertidal seagrass beds of an estuarine system, and their conservation significance. Biodivers. Conserv. 22:357–72. Barnes RSK, 2013b. Spatial stability of macrobenthic seagrass biodiversity. Mar. Ecol. Progr. Ser. 493:127–39. Barnes RSK, 2020a. Do different sympatric seagrasses support macrobenthic faunas of differing composition, abundance, biodiversity or patchiness? Mar. Environ. Res. 160:104983. Barnes RSK, 2020b. Do species display characteristic intraspecific levels of patchiness in a given habitat type? The case of intertidal seagrass macrobenthos. Mar. Biol. 167:1-12. Barnes RSK, 2021. Patterns of seagrass macrobenthic biodiversity in the warm-temperate Knysna estuarine bay, Western Cape: a review. Aquat. Ecol. 55:327–45. Barnes RSK, 2022. Biodiversity differentials between seagrass and adjacent bare sediment change along an estuarine gradient. Estuar. Coast. Shelf Sci. 274:107951. Barnes RSK, 2023. Seagrass macrobenthic biodiversity does not No n- co mm er cia l u se on ly Macrobenthos of lagoon ecosystems: a comparison in vegetated and bare sediments 17 vary in conformity with a leaky-lagoonal confinement gradient. Mar. Environ. Res. 185:105897. Barnes RSK, Barnes MKS, 2012. Shore height and differentials between macrobenthic assemblages in vegetated and unvegetated areas of an intertidal sandflat. Estuar. Coast. Shelf Sci. 106:112–20. Barnes RSK, Barnes MKS, 2014. Biodiversity differentials between the numerically dominant macrobenthos of seagrass and adjacent unvegetated sand in the absence of sandflat bioturbation. Mar. Environ. Res. 99:34–43. Barnes RSK, Claassens L, Seath J, 2022. Where ecologically ‘tis better to go brown than green: enhanced seagrass macrobenthic biodiversity within the canals of a brownfield coastal marina. Biodivers. Conserv. 31:2981–97. Barnes RSK, Hamylton S, 2016. On the very edge: faunal and functional responses to the interface between benthic seagrass and unvegetated-sand assemblages. Mar. Ecol. Prog. Ser. 553:33–48. Blanchet H, de Montaudouin X, Chardy P, Bachelet G, 2005. Structuring factors and recent changes in subtidal macrozoobenthic communities of a coastal lagoon, Arcachon Bay (France). Estuar. Coast. Shelf Sci. 64:561-76. Brito MC, Martin D, Núñez J, 2005. Polychaetes associated to a Cymodocea nodosa meadow in the Canary Islands: assemblage structure, temporal variability and vertical distribution compared to other Mediterranean seagrass meadows. Mar. Biol. 146:467–81. Brundu G, Magni P, 2021. Context-dependent effect of serpulid reefs on the variability of soft-bottom macrobenthic assemblages in three Mediterranean lagoons (Sardinia, Italy). Estuar. Coast. Shelf Sci. 262:107589. Camarda I, 1995. Un sistema di aree di interesse botanico per la salvaguardia della biodiversità floristica della Sardegna. Bollettino della Società Sarda di Scienze Naturali, 30:245-95. Cardone F, Corriero G, Fianchini A, et al., 2014. Biodiversity of transitional waters: species composition and comparative analysis of hard bottom communities from south-eastern Italian coast. J. Mar. Biol. Ass. UK 94:25–34. Como S, Magni P, Baroli M, et al., 2008. Comparative analysis of macrofaunal species richness and composition in Posidonia oceanica, Cymodocea nodosa and leaf litter beds. Mar. Biol. 153:1087-101. Como S, Magni P, Van Der Velde G, et al., 2012. Spatial variations in δ13C and δ15N values of primary consumers in a coastal lagoon. Estuar. Coast. Shelf Sci. 115:300-8. Currás, A, Mora J, 1996. Aspectos de la dinámica de poblaciones de Loripes lacteus (Linnaeus, 1758) (Mollusca, Bivalvia) en fondos Zostera spp. de la ría del Eo (NO de España). Nova Acta Científica Compostelana (Bioloxía) 6:167-78. Evangelisti F, Bonfitto A, Morassi M, Sabelli B, 2016. How many native Cerithium species in the Mediterranean Sea. An integrative taxonomic approach. J. Mollusc. Stud. 82:292–304. Fenzi GA, 2013. Variabilità spazio-temporale dei popolamenti macrozoobentonici in relazione ai cambiamenti stagionali in un ambiente lagunare (Mistras, Sardegna Occidentale) con basso impatto antropico ed elevata stabilità ecologica. Tesi dottorale, Università degli Studi della Tuscia, Viterbo, Italy. García-Gómez G, García-Herrero Á, Sánchez N, et al., 2022. Meiofauna is an important, yet often overlooked, component of biodiversity in the ecosystem formed by Posidonia oceanica. Invert. Biol. 141:e12377. Garilli V, Galletti L, 2006. Taxonomical characters for distinguishing Cerithium lividulum Risso, 1826, and C. renovatum Monterosato, 1884 (Gastropoda, Caenogastropoda, Cerithiidae). Basteria, 70:109-22. Gravina MF, Cabiddu S, Como S, et al, 2020b. Disentangling heterogeneity and commonalities in nanotidal Mediterranean lagoons through environmental features and macrozoobenthic assemblages. Estuar. Coast. Shelf Sci. 237:106688. Holmes SP, Dekker R, Williams DI, 2004. Population dynamics and genetic differentiation in the bivalve mollusc Abra tenuis: aplanic dispersal. Mar. Ecol. Prog. Ser. 268:131–40. Hu C, Liu Y, Yang X, et al., 2022. Functional trait responses of macrobenthic communities in seagrass microhabitats of a temperate lagoon. Mar. Pollut. Bull. 177:113491. Leopardas V, Honda K, Go GA, et al., 2016. Variation in macrofaunal communities of sea grass beds along a pollution gradient in Bolinao, northwestern Philippines. Mar. Poll. Bull. 105:310-8. Leopardas V, Uy W, Nakaoka M, 2014. Benthic macrofaunal assemblages in multispecific seagrass meadows of the southern Philippines: variation among vegetation dominated by different seagrass species. J. Exp. Mar. Biol. Ecol. 457:71-80. Lepoint G, Hyndes GA, 2022. Tropicalization of seagrass macrophytodetritus accumulations and associated food webs. Front. Mar. Sci. 9:943841. Magni P, Como S, Cucco A, et al., 2008a. A multidisciplinary and ecosystemic approach in the Oristano Lagoon-Gulf System (Sardinia, Italy) as a tool in management plans. Transit. Water Bull. 2:41-62. Magni P, Como S, Kamijo A, Montani S, 2017. Effects of Zostera marina on the patterns of spatial distribution of sediments and macrozoobenthos in the boreal lagoon of Furen (Hokkaido, Japan). Mar. Environ. Res. 131C:90-102. Magni P, Micheletti, S, Casu D, et al., 2005. Relationships between chemical characteristics of sediments and macrofaunal communities in the Cabras lagoon (western Mediterranean, Italy). Hydrobiologia 550:109-15. Magni P, Rajagopal S, van der Velde G, et al., 2008b. Sediment features, macrozoobenthic assemblages and trophic relationships (δ13C and δ15N analysis) following a dystrophic event with anoxia and sulphide development in the Santa Giusta lagoon (western Sardinia, Italy). Mar. Pollut. Bull. 57:125-36. Murenu M, Olita A, Sabatini A, Follesa MC, Cau A, 2004. Dystrophy effects on the Liza ramada (Risso, 1826) (Pisces, Mugilidae) population in the Cabras lagoon (central-western Sardinia). Chem. Ecol. 20(S1):S425–S433. Nakaoka M, 2005. Plant-animal interactions in seagrass beds: ongoing and future challenges for understanding population and community dynamics. Popul. Ecol. 47:167–77. Nonnis Marzano C, Gravina MF, Fianchini A, et al., 2010. The macrozoobenthos of the Karavasta lagoon system (Albania): local assemblages and geographical comparisons. Mar. Ecol. 31:622–32. Ólafsson E, Aarnio K, Bonsdorff E, Arroyo NL, 2013. Fauna of the green alga Cladophora glomerata in the Baltic Sea: density, diversity, and algal decomposition stage. Marine Biology 160: 2353–62. No n- co mm er cia l u se on ly P. Magni and M.F. Gravina18 Padedda BM, Pulina S, Magni P, et al., 2012. Phytoplankton dynamics in relation to environmental changes in a phytoplankton-dominated Mediterranean lagoon (Cabras Lagoon, Italy). Adv. Limnol Oceanogr. 3:147-169. Pascucci V, De Falco G, Del Vais C, et al., 2018. Climate changes and human impact on the Mistras coastal barrier system (W Sardinia, Italy). Mar. Geol. 395:271-84. Perez M, Duarte CM, Romero J, et al., 1994. Growth plasticity in Cymodocea nodosa stands: the importance of nutrient supply. Aquat Bot 47:249–64. Pulina S, Satta CT, Padedda BM, et al., 2018. Seasonal variations of phytoplankton size structure in relation to environmental variables in three Mediterranean shallow coastal lagoons. Estuar. Coast. Shelf Sci. 212:95-104. Ramsar Convention, 1971. Convention on wetlands of international importance especially as waterfowl habitat. Ramsar (Iran), 2 February 1971. UN Treaty Series No. 14583. As amended by the Paris Protocol, 3 December 1982, and Regina Amendments, 28 May 1987. Available from: http://www.ramsar.org/key_conv_e.html. Satta CT, Anglès S, Garcés E, et al, 2014. Dinoflagellate cyst assemblages in surface sediments from three shallow Mediterranean lagoons (Sardinia, North Western Mediterranean Sea). Estuar. Coast. 37:646-63. Sechi N, 1982. Lo stato trofico di alcuni stagni salmastri costieri della Sardegna. Boll. Soc. Sarda Sci. Nat. 21:285-95. Sfriso A, Birkemeyer T, Ghetti PF, 2001. Benthic macrofauna changes in areas of Venice lagoon populated by seagrasses or seaweeds. Mar. Environ. Res. 52:323-49. Specchiulli A, Cilenti L, D’Adamo R, et al, 2018. Dissolved organic matter dynamics in Mediterranean lagoons: The relationship between DOC and CDOM. Mar. Chem. 202:37-48. Surugiu V, Teacâ A, Svedu I, Quijón PA, 2021. A Hotspot in the Romanian Black Sea: Eelgrass Beds Drive Local Biodiversity in Surrounding Bare Sediments. Front. Mar. Sci. 8:745137. Tagliapietra D, Pessa G, Cornello M, et al., 2016. Temporal distribution of intertidal macrozoobenthic assemblages in a Nanozostera noltii-dominated area (Lagoon of Venice). Mar. Environ. Res. 114:31-9. Włodarska-Kowalczuk M, Jankowska E, Kotwicki L, Balazy P, 2014. Evidence of season-dependency in vegetation effects on macrofauna in temperate seagrass meadows (Baltic Sea). PLoS One 9:e100788. Yamada K, Hori M, Tanaka Y, et al., 2007. Temporal and spatial macrofaunal community changes along a salinity gradient in seagrass meadows of Akkeshi-ko estuary and Akkeshi Bay, northern Japan. Hydrobiologia 592:345-58. Zheng X, Como S, Huang L, Magni P, 2020. Temporal changes of a food web structure driven by different primary producers in a subtropical eutrophic lagoon. Mar. Environ. Res. 161:105-28. No n- co mm er cia l u se on ly