Taxonomic and Morphological Descriptors Reveal High Benthic Temporal Variability in a Mediterranean Marine Submerged Cave over a Decade

Depósito de investigación de la Universidad de Sevilla https://idus.us.es/ “This version of the article has been accepted for publication, after peer review (when applicable) and is subject to Springer Nature’s AM terms of use, but is not the Version of Record and does not reflect post-acceptance improvements, or any corrections. The Version of Record is available online at: http://dx.doi.org/10.1007/s10750-019-04005-2” Taxonomic and morphological descriptors reveal high benthic temporal variability in a Mediterranean marine submerged cave over a decade Juan Sempere-Valverde 1, Álvaro Sabino Lorenzo 1, Free Espinosa 1, Vasilis Gerovasileiou 2, Luis Sánchez-Tocino 3, Carlos Navarro-Barranco 1, 4 1 - Laboratorio de Biología Marina (LBM), Facultad de Biología, Universidad de Sevilla, Sevilla Spain. 2 - Institute of Marine Biology, Biotechnology and Aquaculture, Hellenic Centre for Marine Research, Heraklion, Crete, Greece. 3 - Departamento de Biología Animal, Facultad de Ciencias, Universidad de Granada, Granada, Spain. 4 - Departamento de Biología (Zoología), Universidad Autónoma de Madrid, Madrid, Spain. Keywords: Submerged Caves; Benthic Community; Morphological Descriptor; 1 Taxonomic Descriptor; Growth Forms 2 3 “This paper has not been submitted elsewhere in identical or similar form, nor will it 4 be during the first three months after its submission to Hydrobiologia.” 5 6 ABSTRACT 7 The effectiveness of monitoring programs in the marine environment is often limited by 8 the scarcity of information concerning medium/long-term time series observations 9 (years-decades) in complex and variable systems. Particularly, marine submerged caves 10 have received relatively low research effort, as they are restricted ecosystems difficult 11 to access. In this study, we describe the spatio-temporal variations of sessile benthic 12 assemblages along the exterior-interior axis of Cerro-Gordo submarine cave (Granada, 13 Spain), from year 2007 to 2016 using taxonomic (e.g. phylum) and morphological (i.e. 14 mounds and sheets) descriptors. In contrast to our initial expectations (higher stability 15 in the inner part of the cave), significant temporal variations of sessile community 16 structure and morphology occurred in both external and internal cave sectors. In 17 addition, higher spatial differences were encountered among opposite cave walls in the 18 inner dark sector, which is probably explained by small-scale variability in topography 19 and siltation rate. In the outer semi-dark sector, community was more heterogeneous, 20 although generally similar between opposite walls. Among the used descriptors, the 21 morphological one was able to detect greater spatio-temporal variation in the outer 22 semi-dark sector, while the taxonomic descriptor was more suitable in the inner dark 23 sector. These results could provide a basis for future monitoring programs. 24 25 INTRODUCTION 26 Spatio-temporal quantitative studies on marine benthic communities have been 27 useful to understand ecological resilience, ecosystem functioning and to provide 28 essential information for habitat conservation (Schneider, 1994). Nevertheless, some 29 areas of the marine environment, such as submerged caves, are often difficult to access 30 and have received limited research effort when compared to other habitat types 31 (Bussotti et al., 2006; Gerovasileiou et al., 2017). Marine submerged caves host unique 32 and vulnerable ecosystems and may function as ecological islands, hosting communities 33 sensitive to fragmentation, relict species, as well as communities similar to those found 34 in abyssal systems (Sarà, 1976; Harmelin et al., 1985; Boury-Esnault et al., 1993; Bianchi 35 et al., 1996; Rastorgueff et al., 2015; Gerovasileiou et al., 2016). In addition, submarine 36 caves are considered biodiversity hotspots (e.g. Mediterranean caves host 45.7% of all 37 poriferans and near 39.4% of all bryozoans species inhabiting this sea) that may provide 38 important refuges for the conservation of high biodiversity and sensitive key species 39 (Okamura et al., 2001; Gerovasileiou & Voultsiadou, 2012, 2014; Rosso & Di Martino, 40 2016; Garrabou et al., 2017; Rosso et al., 2018). In the Mediterranean Sea, these 41 fragmented habitats are widely distributed along rocky coastal areas and are present in 42 about 66% of the existing Marine Protected Areas (Abdulla et al., 2008; Giakoumi et al., 43 2013). Therefore, marine caves are listed in the European Community Habitats Directive 44 (Directive, 1992) and have been included (together with seamounts, underwater 45 canyons, aphotic hard beds and chemo-synthetic phenomena) in the “Dark Habitats 46 Action Plan” of the Barcelona Convention (UNEP-MAP-RAC/SPA, 2015) for the 47 conservation of sensitive species and habitats. 48 Decreasing illumination, water flow and nutrient supply towards the inner 49 confined cave sectors are the main factors influencing the structure of benthic 50 communities inhabiting submerged caves (Zabala et al., 1989; Bianchi & Morri, 1994; 51 Gerovasileiou & Voultsiadou, 2016). Therefore, blind-ending marine caves usually 52 exhibit a longitudinal gradient in the distribution of species, morphologies and feeding 53 strategies (Cinelli et al., 1977; Harmelin et al., 1985, 1986; Okamura et al., 2001; 54 Guarnieri et al., 2012; Gerovasileiou et al., 2017). However, each marine cave supports 55 a unique environment strongly depending of its topography (Harmelin, 1969; Bussotti 56 et al., 2006; Gerovasileiou et al., 2013; Gerovasileiou & Voultsiadou, 2016; Radolović et 57 al., 2015; Rosso et al., 2018). 58 Marine cave communities can be easily disturbed by natural and anthropogenic 59 alterations at global (e.g. climate change and water acidification) and regional/local 60 scales (e.g. storms, coastal pollution, coastal and marine infrastructure construction and 61 recreational diving – Sarà, 1976; Chevaldonné & Lejeusne, 2003; Parravicini et al., 2009, 62 2010; Di Franco et al., 2010; Giakoumi et al., 2013). These may affect benthic community 63 at both taxonomic (i.e. diversity, species composition and distribution) and functional 64 levels (i.e. species morphology and feeding strategies). Recently, the use of taxonomic 65 surrogates as well as morphology (i.e. body shape) have been proposed as descriptors 66 for monitoring potential impacts on cave benthic communities, requiring less 67 information, resources and taxonomic expertise (Parravicini et al., 2010; Guarnieri et al., 68 2012; Gerovasileiou et al., 2017; Nepote et al., 2017; UNEP-MAP-RAC/SPA, 2017; 69 Montefalcone et al., 2018). Such indicators would be useful to detect impacts such as 70 those caused by sediment re-suspension and mechanical impacts by divers, to which 71 erect morphologies are more fragile (Sala et al., 1996; Di Franco et al., 2010; Guarnieri 72 et al., 2012). Recently, the comparison of morphological data for sessile benthos, 73 collected within a 30-years trajectory from a semi-submerged cave subjected to several 74 pressures (e.g. heat waves and coastal interventions), revealed a decline in the cover of 75 3-dimensional morphologies with a parallel increase of encrusting poriferans and turf 76 (see Montefalcone et al., 2018). In this matter, the increase of water temperature is 77 expected to impact massive sponges, whose communities may be considered as poorly 78 resilient to water warming (Costa et al., 2018). 79 The effectiveness of marine monitoring programs is often limited by the scarcity 80 of information concerning the natural temporal changes present in complex and 81 variable systems (Magurran et al., 2010). In those cases, medium/long-term time series 82 observations (years-decades) are essential tools for addressing these issues. However, 83 temporal studies in marine caves (even at a short scale) are scarce (e.g. Harmelin, 1980; 84 Martí et al., 2004; Bussotti et al., 2006; Montefalcone et al., 2018; Costa et al., 2018). 85 Recent long-term studies in this habitat have been conducted in heavily disturbed areas, 86 so that temporal dynamics of sessile benthos are examined in response to close human 87 activities (Nepote et al., 2017; Montefalcone et al., 2018). The lack of temporal data 88 coupled with the high degree of natural variability among caves require intensive studies 89 in order to highlight the main natural patterns and to identify baselines for ecological 90 management (Bussotti et al., 2006; Navarro-Barranco et al., 2013). Therefore, attempts 91 to develop monitoring schemes and management programs in these habitats must cope 92 with a high degree of uncertainty (UNEP-MAP-RAC/SPA, 2017). 93 The general aim of the present study was to identify for the first time spatio- 94 temporal variations of sessile benthos over a decade in a submerged marine cave of the 95 Alboran Sea (Western Mediterranean), a marine area that has received little research 96 effort when compared to other Mediterranean areas (Gerovasileiou & Voultsiadou, 97 2014). Two main hypotheses were tested in this study: 98 1) Due to the higher environmental stability (i.e. temperature and water motion) 99 in the inner sectors of marine caves (Harmelin et al., 1985; Bianchi & Morri, 1994), which 100 has also been reported for the studied cave (Navarro-Barranco et al., 2012, 2015), 101 benthic communities in the inner dark sector will show less temporal variability than 102 those in the outer semi-dark sector. 103 2) Previous studies have highlighted the high degree of small-scale spatial 104 variability of sessile benthos in marine caves (e.g. Benedetti-Cecchi et al., 1998; Bussotti 105 et al., 2006; Dimarchopoulou et al., 2018). Therefore, small-scale differences among the 106 opposite vertical walls within each sector are expected to generate different temporal 107 patterns as well. 108 Additionally, the effective use of both taxonomic and morphological descriptors on 109 monitoring cave communities is also addressed in the present article. Patterns obtained 110 by each approach, as well as the amount of variability picked up, is compared in order 111 to evaluate the suitability of such descriptor to identify spatio-temporal responses of 112 sessile benthos in the studied cave. This is important for the development of easy and 113 efficient monitoring programs for vulnerable and understudied marine habitats such as 114 marine caves across different Mediterranean ecoregions (Gerovasileiou et al., 2017; 115 UNEP-MAP-RAC/SPA, 2017). 116 METHODS 117 Study site 118 The Alboran Sea (Western Mediterranean) is considered as a marine biodiversity 119 hotspot (Coll et al., 2010) with peculiar hydrodynamic conditions (such as events of local 120 upwelling) and higher primary productivity than other areas of the Mediterranean Sea 121 (see Delgado, 1990). The cave of Cerro-Gordo (also known as Cueva del Sifón or Cueva 122 de La Anémona) is located in the North Coast of the Alboran Sea, within Maro-Cerro 123 Gordo Natural Park, an Specially Protected Area of Mediterranean Importance (SPAMI - 124 UNEP-MAP-RAC/SPA, 2014) in Granada, Spain (Fig. 1a). 125 Cerro-Gordo Cave is a vertical straight fissure, perpendicular to the shoreline and 126 more than 100 m long. It is separated in two main chambers/sectors with different 127 environmental conditions (see Navarro-Barranco et al., 2012) by a marked narrowing at 128 a distance of 25 m from the entrance (see Fig. 1b). Sediment at the entrance of the cave 129 is mainly composed by medium-sized sands (Navarro-Barranco et al., 2012). The 130 entrance is totally submerged and is approximately 15 m wide and 15 m high (from the 131 cave ceiling, at 2 m depth, to the sandy bottom of the coastal cliffs at 15 m depth). Semi- 132 dark sector chamber gradually narrows down from the entrance to the aforementioned 133 strapping. It has a domed shape and an air chamber is found on the greatest part of the 134 ceiling. Up to 50% of the sediment at the bottom of this sector is composed by very fine 135 and fine sands (Navarro-Barranco et al., 2012). Light, although strictly reduced, is still 136 present in this sector (authors' personal observations). The inner dark sector is a 137 narrower, about 75 m long, blind-ending fissure, whose walls are nearly vertical in the 138 beginning and gradually rotate counter-clockwise. Dark sector walls narrow down when 139 reaching the ceiling, making access to divers impossible, and ends in a small internal air 140 chamber. Sediment at the bottom of this sector is dominated by silt and clay (Navarro- 141 Barranco et al., 2012). The cave receives freshwater infiltration through a complex 142 conduit network, and thus, the inner sector hosts a permanent halocline (Navarro- 143 Barranco et al., 2012). A dampening in temperature oscillations is also observed in the 144 inner parts of the cave (see Navarro-Barranco et al., 2015). Videos and more graphic 145 information about this marine cave are available at Sánchez-Tocino (2018). 146 Samples collection and treatment 147 Non-destructive photographic sampling was used for the study of sessile cave 148 benthos (Parravicini et al., 2009; Gerovasileiou & Voultsiadou, 2016; Dimarchopoulou et 149 al., 2018). Four randomly located replicate quadrats (20 x 20 cm) were photographed at 150 5 m intervals along the opposite walls (right and left) of the two cave sectors (5 sites in 151 the semi-dark and 7 in the dark sector, see Fig. 1b) at a depth range of 6-8 m. Samplings 152 were carried out with SCUBA diving in August of the years 2007, 2010 and 2016 in the 153 dark sector and 2008, 2011 and 2016 in the semi-dark sector. 154 Coverage data was obtained by overlaying and counting 50 random points over 155 each photoquadrat with photoQuad software (Trygonis & Sini, 2012). Benthic 156 community analysis was based on two descriptors: 1) morphological, representing 5 157 different growth forms of sessile biota, and 2) taxonomic, including 10 major taxonomic 158 groups visually identified in the photoquadrats (see Tab. 1). These descriptors have been 159 found to be suitable for the identification of spatio-temporal changes of sessile benthic 160 communities in other Mediterranean marine caves (Parravicini et al., 2010; Guarnieri et 161 al., 2012; Gerovasileiou et al., 2017; Nepote et al., 2017; Montefalcone et al., 2018). 162 Among the selected morphologies, encrusting bryozoans from the Flustrina Suborder 163 were further separated in prostrate sheets and spots (see definition in Okamura et al., 164 (Quinn, 1982; Sousa, 1984). These kinds of biotic interactions can be responsible, 301 without any environmental influence, for the unexpected high rate of temporal changes 302 in the inner part of the cave, although there are several sources of perturbation that 303 could also contribute to the temporal variability. For example, marine caves with 304 freshwater infiltrations receive organic matter from percolation through rock cracks and 305 fissures that may be of great importance in a highly oligotrophic ecosystem (Pohlman et 306 al., 1997; Rastorgueff et al., 2015). An increase in the presence of large-sized serpulids 307 in particular cave zones, as that observed in the dark sector in this study, has been 308 previously supposed to be correlated with freshwater inputs and related increase in 309 dissolved organic matter (Guido et al., 2013; Sanfilippo et al., 2015, 2017). Similarly, 310 some poriferans may directly benefit from dissolved organic matter; this has been 311 suggested to help their persistence in oligotrophic systems (de Goeij et al., 2013). 312 Nevertheless, the high coverage of poriferans observed in the dark sector during the 313 first year of study could also been favoured by the absence of competition with passive 314 filter feeders. Although nutrient loading associated with submarine groundwater 315 discharges could benefit some species, these benefits may be outweighed by negative 316 impacts associated to lowered pH (affecting mainly to calcifying organisms), 317 anthropogenic pollution (aquifer with bad status could act as a conduit of pollutants into 318 cave systems) or severe salinity changes that most fauna could not tolerate (Lecher & 319 Mackey 2018). Negative effects of this source of stress on benthic communities have 320 been already reported in marine caves (Iliffe et al., 1984; Guido et al., 2013; Radolović 321 et al., 2015). Specifically, in the studied area, Alberquillas aquifer (ES060MSBT060.063 322 water mass in MAPAMA 2018) has been assessed as “does not reach good condition” in 323 terms of water chemical condition or concentration of pollutants (Directive, 2000) but 324 the biological consequences of the quality and quantity of such groundwater discharges 325 have not yet been measured. In this matter, a great among-years rain variability 326 characterize Mediterranean climate patterns, which influences the size of groundwater 327 discharges. Therefore, the natural temporal variability in dark sector could be higher 328 than primarily expected, which reveals the importance of increase future research effort 329 on this topic. 330 Unstable weather conditions and recreational diving activities during the 331 examined timeframe could possibly induce recurrent turbidity events (sediment 332 resuspension), which may be responsible for sponge necrosis in the dark sector, where 333 sediment is thinner and more prone to resuspension (Di Franco et al., 2010; Navarro- 334 Barranco et al., 2012; Nepote et al., 2017). Storm events became more frequent from 335 2012 onwards (Puertos del Estado, 2018) and, although recreational diving is prohibited 336 in Cerro-Gordo Cave, it received a high number of unregulated visits for recreational and 337 research purposes during the timeframe of this study (Terrón-Sigler et al., 2008, 2016; 338 authors’ personal observations). Therefore, the observed increase of polychaetes and 339 brachiopods in the inner sector may have been favoured by the presence of large-sized 340 organic particles in the water column, which they can filter more efficiently (Riisgård & 341 Larsen, 2000; Mancenido & Gourvennec, 2007). This increase could also have been 342 favoured by the reduction of inter-specific competition with poriferans, as high 343 sedimentation levels can lead to clogging of their aquiferous system (Di Franco et al., 344 2010; Nepote et al., 2017). 345 The most studied disturbance caused by divers in marine caves is related to 346 physical contact and associated damage to sessile organisms with erect morphologies 347 (Sala et al., 1996; Di Franco et al., 2010; Guarnieri et al., 2012). Erect morphologies were 348 absent from the inner dark sector but, in the semi-dark one, domed morphologies 349 decreased through time. In any case, the large size of the outer sector chamber may 350 minimize divers’ physical contact and damage when compared to narrower caves, as is 351 the case of dark sector. Furthermore, sponges could have been affected by other 352 variables such as temperature increase (Parravicini et al., 2010; Rastorgueff et al., 2015; 353 Montefalcone et al., 2018; Costa et al., 2018). Recent studies showed that events of 354 seawater thermal increase affected predominantly massive morphologies in marine 355 caves (Parravicini et al., 2010; Gerovasileiou et al., 2018; Montefalcone et al., 2018). This 356 event could be detected by the morphological descriptor, since large mortalities of 357 massive poriferans, such as Ircinia variabilis (Schmidt, 1862) and Sarcotragus 358 fasciculatus (Pallas, 1766), were documented in 2010 and 2015 in our study area due to 359 the increase of water temperature (Maldonado et al., 2010; Sánchez-Tocino & Tierno de 360 Figueroa, 2016; Puertos del Estado, 2018). In Cerro-Gordo Cave, the major decrease of 361 poriferans was recorded in 2010-2011 and, later in 2016, sponge cover did not reach the 362 percentages found at the beginning of the study. 363 Spatial patterns of the morphological and taxonomic descriptors along the 364 longitudinal axis of Cerro-Gordo Cave were similar to those previously described for 365 blind-ending caves in the Mediterranean (see Rastorgueff et al., 2015 and references 366 therein). The predominance of prostrate sheet morphologies in the inner parts of the 367 cave has been suggested as an adaptation to a resource-restricted environment (Bibiloni 368 et al., 1989; Okamura et al., 2001). In addition, the presence of thin sediment (silt and 369 clay) on the bottom of the dark sector (see Navarro-Barranco et al., 2012) suggests that 370 water motion is low, although water circulation could still be occurring. Spatial 371 differences among the wall communities (alternative hypothesis 2) occurred within the 372 dark but not in the semi-dark sector, and are probably related to their inclination. This 373 would induce higher sediment deposition over left wall surfaces, causing a reduction in 374 the abundance of the brachiopod Novocrania anomala (O. F. Müller, 1776), which 375 inhabits mud-free surfaces (Logan et al., 2008; Rosso et al., 2015). Seemingly, small 376 changes in the slope of vertical walls that influence silt smothering and food availability 377 have been linked with decrease of bryozoan species richness and coverage in marine 378 caves (Harmelin, 1986). 379 Ecological indicators are simplifications of natural systems in many aspects such 380 as spatial heterogeneity (Turnhout et al., 2007). In this matter, the used morphological 381 descriptor (i.e. growth forms) proved to be a good indicator in the semi-dark cave sector 382 while the used taxonomic descriptor (i.e. major taxonomic groups) was more adequate 383 for describing variations in the dark sector. The efficiency of morphological descriptors 384 in characterizing spatial patterns and temporal dynamics in marine cave communities 385 has already been highlighted in previous studies (Guarnieri et al., 2012; Nepote et al., 386 2017; Gerovasileiou et al., 2017; Montefalcone et al., 2018). The higher accuracy of the 387 taxonomic descriptor in the inner dark cave sector is related to the locally lower 388 morphological diversity (Gerovasileiou & Voultsiadou, 2016). On the other hand, the 389 identification of taxa to lower taxonomic levels would probably increase the suitability 390 of the taxonomic descriptor in the semi-dark sector. In this regard, the identification at 391 the level of genus or species would require greater effort and taxonomic expertise, as 392 well as destructive sampling procedures. Nevertheless, it could provide useful 393 taxonomic information (Rosso et al., 2013a). 394 In conclusion, the spatio-temporal variation in Cerro-Gordo Cave reflects the 395 effect of complex context- and scale-dependent processes in submarine caves 396 (Harmelin, 1969; Bianchi et al., 1996; Bussotti et al., 2006; Radolović et al., 2015; 397 Gerovasileiou & Voultsiadou, 2016). Our study revealed variations in the taxonomic 398 composition and morphology of sessile benthic communities in both external and 399 internal cave sectors of Cerro-Gordo Cave within the last decade. The obtained results 400 provide valuable information for future monitoring programs of these vulnerable and 401 protected marine habitats. In agreement with previous studies, this study provides 402 evidence on the suitability of morphological descriptors to detect changes in a short 403 time-scale, at least in the outer cave sector (Parravicini et al., 2010; Rastorgueff et al. 404 2015; Nepote et al., 2017; Montefalcone et al., 2018). On the other hand, the use of the 405 taxonomic descriptors (even at high taxonomic levels) could be also recommended, 406 especially in the inner confined zones dominated by few morphological categories. 407 Nevertheless, results cannot be extrapolated to all marine caves, given the variable 408 environmental conditions and heterogeneity of this habitat type (Harmelin, 1969; 409 Bussotti et al., 2006; Gerovasileiou & Voultsiadou, 2016; Radolović et al., 2015; Rosso et 410 al., 2018), and further testing is required in order to evaluate the efficiency of the used 411 descriptors as surrogates of the sessile community structure in other caves of the 412 Mediterranean Sea. 413 ACKNOWLEDGEMENTS 414 We would like to thank Enrique Quintero-Torres, from the Venezuelan Institute 415 for Scientific Research, and César Megina, from the University of Seville, for their help 416 in data analysis. We are also grateful to Carlos González, Raúl Rodríguez Roig and “Open 417 Water La Herradura” diving center for their collaboration during field samplings. The 418 authors would also like to thank two anonymous reviewers for their constructive 419 comments. 420 REFERENCES 421 Abdulla, A., Gomei, M., Maison, É. & C. Piante, 2008. Status of Marine Protected Areas 422 in the Mediterranean Sea. IUCN, Málaga, and WWF, Paris. 423 Anderson, M. J., 2001. A new method for non‐parametric multivariate analysis of 424 variance. Austral ecology 26: 32–46. 425 Anderson, M. J., Diebel, C. E., Blom, W. M. & T. J. Landers, 2005. Consistency and 426 variation in kelp holdfast assemblages: spatial patterns of biodiversity for the 427 major phyla at different taxonomic resolutions. Journal of Experimental Marine 428 Biology and Ecology 320: 35–56. 429 Anderson, M. J., Gorley, R. N. & K. R. Clarke, 2008. PERMANOVA + for PRIMER. Guide to 430 software and statistical methods. Plymouth, UK. PRIMER-E Ltd. 431 Baldacconi, R. & E. Trainito, 2013. Spugne del Mediterraneo. Edizioni il Castello. 432 Benedetti‐Cecchi, L., Airoldi, L., Abbiati, M., & F. Cinelli, 1998. Spatial variability in the 433 distribution of sponges and cnidarians in a sublittoral marine cave with sulphur‐ 434 water springs. Journal of the Marine Biological Association of the United Kingdom 435 78: 43–58. 436 Bianchi C. N., Cattaneo‐Vietti, R., Cinelli, F., Morri, C. & M. Pansini, 1996. Lo studio 437 biologico delle grotte sottomarine del Mediterraneo: conoscenze attuali e 438 prospettive. Bollettino dei Musei e degli Istituti Biologici dell’Università di Genova 439 60/61: 41–69. 440 Bianchi, C. N. & C. Morri, 1994. Studio bionomico comparativo di alcune grotte marine 441 sommerse: Definizione di una scala di confinamento. Istituto Italiano di 442 Speleologia Memoria 6: 107–123. 443 Bibiloni, M. A., Uriz, M. J. & J. M. Gili, 1989. Sponge communities in three submarine 444 caves of the Balearic Islands (Western Mediterranean): adaptations and faunistic 445 composition. Marine Ecology 10: 317–334. 446 Boury‐Esnault, N., Harmelin, J. G. & J. Vacelet, 1993. Les abysses méditerranéennes à 447 vingt mètres de profondeur. La recherche 24: 848–851. 448 Bussotti, S., Terlizzi, A., Fraschetti, S., Belmonte, G. & F. Boero, 2006. Spatial and temporal 449 variability of sessile benthos in shallow Mediterranean marine caves. Marine 450 Ecology Progress Series 325: 109–119. 451 Carballo, J. L. & J. C. Garcia‐Gómez, 1994. Esponjas del Estrecho de Gibraltar y áreas 452 próximas, con nuevas aportaciones para la fauna Iberica [Porifera from the Straits 453 of Gibraltar and nearby areas, new species of the Iberian fauna]. Cahiers de 454 Biologie Marine 35: 193–211. 455 Carballo, J. L., Naranjo, S. & J. C. García‐Gómez, 1997. Where does the Mediterranean 456 Sea begin? Zoogeographical affinities of the littoral sponges of the Straits of 457 Gibraltar. Journal of Biogeography 24: 223–232. 458 Cebrián, P. E. & S. E. Ballesteros, 2003. Zonation patterns of benthic communities in an 459 upwelling area from de western Mediterranean (La Herradura, Alboran Sea). 460 Scientia Marina 68: 69–84. 461 Chevaldonné, P. & C. Lejeusne, 2003. Regional warming‐induced species shift in north‐ 462 west Mediterranean marine caves. Ecology Letters 6: 371–379. 463 Cinelli, F., Fresi, E., Mazzella, L., Pansini, M., Pronzato, R. & A. Svoboda, 1977. Distribution 464 of benthic phyto‐and zoocoenoses along a light gradient in a superficial marine 465 cave. Biology of benthic organisms. Pergamon Press, Oxford: 173–183. 466 Clarke, K. R. & R. N. Gorley, 2006. PRIMER v6: User manual/tutorial. PRIMER-E Ltd, 467 Plymouth, UK. 468 Coll, M., Piroddi, C., Steenbeek, J., Kaschner, K., Ben Rais Lasram, F., Aguzzi, J., 469 Ballesteros, E., Bianchi, C. N., Corbera, J., Dailianis, T., Danovaro, R., Estrada, M., 470 Froglia, C., Galil, B. S., Gasol, J. M., Gertwagen, R., Gil, J., Guilhaumon, F., Kesner‐ 471 Reyes, K., Kitsos, M. S., Koukouras, A., Lampadariou, N., Laxamana, E., López‐Fé 472 de la Cuadra, C. M., Lotze, H. K., Martin, D., Mouillot, D., Oro, D., Raicevich, S., 473 Rius‐Barile, J., Saiz‐Salinas, J. I., San Vicente, C., Somot, S., Templado, J., Turon, X., 474 Vafidis, D., Villanueva, R. & E. Voultsiadou, 2010. The biodiversity of the 475 Mediterranean Sea: estimates, patterns, and threats. PLoS ONE 5: e11842. 476 Costa, G., Betti, F., Nepote, E., Cattaneo‐Vietti, R., Pansini, M., Bavestrello, G. & M. 477 Bertolino, 2018. Sponge community variations within two semi‐submerged caves 478 of the Ligurian Sea (Mediterranean Sea) over a half‐century time span. The 479 European Zoological Journal 85: 382–392. 480 De Goeij, J. M., Van Oevelen, D., Vermeij, M. J., Osinga, R., Middelburg, J. J., de Goeij, A. 481 F. & W. Admiraal, 2013. Surviving in a marine desert: the sponge loop retains 482 resources within coral reefs. Science 342: 108–110. 483 Delgado, M., 1990. Phytoplankton distribution along the Spanish coast of the Alboran 484 Sea. Scientia Marina 54: 169–178. 485 Di Franco, A., Ferruzza, G., Baiata, P., Chemello, R. & M. Milazzo, 2010. Can recreational 486 scuba divers alter natural gross sedimentation rate? A case study from a 487 Mediterranean deep cave. ICES Journal of Marine Science 67: 871–874. 488 Dimarchopoulou, D., Gerovasileiou, V. & E. Voultsiadou, 2018. Spatial variability of 489 sessile benthos in a semi‐submerged marine cave of a remote Aegean Island 490 (eastern Mediterranean Sea). Regional Studies in Marine Science 17: 102–111. 491 García‐Gómez, J.C., 2015. A guide on environmental monitoring of rocky seabeds in 492 Mediterranean Marine Protected Areas and surrounding zones. RAC/SPA‐ 493 MedMPAnet Project, Tunis. 494 Directive, 1992. Council Directive 92/43/EEC of 21 May 1992 on the conservation of 495 natural habitats and of wild fauna and flora. Official Journal of the European 496 Communities. 497 Directive, 2000. Directive 2000/60/EC of the European Parliament and of the Council of 498 23 October 2000 establishing a framework for Community action in the field of 499 water policy. Official Journal of the European Communities. 500 Garrabou, J., Sala, E., Linares, C., Ledoux, J. B., Montero‐Serra, I., Dominici, J. M., Kipson, 501 S., Teixidó, N., Cebrian, E., Kersting, D. K. & J. G. Harmelin, 2017. Re‐shifting the 502 ecological baseline for the overexploited Mediterranean red coral. Scientific 503 Reports 7: 42404. 504 Gerovasileiou, V. & E. Voultsiadou, 2012. Marine Caves of the Mediterranean Sea: A 505 Sponge Biodiversity Reservoir within a Biodiversity Hotspot. PloS ONE 7: e39873. 506 Gerovasileiou, V. & E. Voultsiadou, 2014. Mediterranean marine caves as biodiversity 507 reservoirs: a preliminary overview. In Bouafif, C., Langar H. & A. Ouerghi (eds), 508 Proceedings of the 1st Mediterranean Symposium on the conservation of Dark 509 Habitats. RAC/SPA, Tunis: 45–50. 510 Gerovasileiou, V. & E. Voultsiadou, 2016. Sponge diversity gradients in marine caves of 511 the Eastern Mediterranean. Journal of the Marine Biological Association of the 512 United Kingdom 96: 407–416. 513 Gerovasileiou, V., Dailianis, T., Sini, M., Otero, M., Numa, C., Katsanevakis, S. & E. 514 Voultsiadou, 2018. Assessing the regional conservation status of sponges 515 (Porifera): the case of the Aegean ecoregion. Mediterranean Marine Science, 19: 516 526–537. 517 Gerovasileiou, V., Dimitriadis, C., Arvanitidis, C. & E. Voultsiadou, 2017. Taxonomic and 518 functional surrogates of sessile benthic diversity in Mediterranean marine caves. 519 PloS ONE 12: e0183707. 520 Gerovasileiou, V., Martínez, A., Álvarez, F., Boxshall, G., Humphreys, W., Jaume, D., 521 Becking, L., Muricy, G., van Hengstum, P., Dekeyzer, S., Decock, W., Vanhoorne, 522 B., Vandepitte, L., Bailly, N. & T. Iliffe, 2016. World Register of marine Cave Species 523 (WoRCS): a new Thematic Species Database for marine and anchialine cave 524 biodiversity. Research Ideas and Outcomes 2: e10451. 525 Gerovasileiou, V., Martínez, A., Álvarez, F., Boxshall, G., Humphreys, W.F., Jaume, D., 526 Becking, L.E., Muricy, G., van Hengstum, P.J., Yamasaki, H., Bailly, N. & T.M. Iliffe, 527 2019. World Register of Marine Cave Species (WoRCS). 528 http://www.marinespecies.org/worcs [last accessed on February 2019]. 529 Gerovasileiou, V., Trygonis, V., Sini, M., Koutsoubas, D. & E. Voultsiadou, 2013. Three‐ 530 dimensional mapping of marine caves using a handheld echosounder. Marine 531 Ecology Progress Series 486: 13–22. 532 Giakoumi, S., Sini, M., Gerovasileiou, V., Mazor, T., Beher, J., Possingham, H. P., Abdulla, 533 A., Çinar, M. E., Dendrinos, P., Gucu, A. C., Karamanlidis, A. A., Rodic, P., 534 Panayotidis, P., Taskin, E., Jaklin, A., Voultsiadou, E., Webster, C., Zenetos, A. & S. 535 Katsanevakis, 2013. Ecoregion‐based conservation planning in the 536 Mediterranean: dealing with large‐scale heterogeneity. PLoS ONE 8: e76449. 537 Gofas, S., Salas, C. & D. Moreno, 2011. Moluscos marinos de Andalucía. Málaga: 538 Universidad de Málaga, Servicio de Publicaciones e Intercambio Científico. 539 Grenier, M., Ruiz, C., Fourt, M., Santonja, M., Dubois, M., Klautau, M., Vacelet, J., Boury‐ 540 Esnault, N. & T. Pérez, 2018. Sponge inventory of the French Mediterranean 541 waters, with an emphasis on cave‐dwelling species. Zootaxa 4466: 205–228. 542 Guarnieri, G., Terlizzi, A., Bevilacqua, S. & S. Fraschetti, 2012. Increasing heterogeneity 543 of sensitive assemblages as a consequence of human impact in submarine caves. 544 Marine Biology 159: 1155–1164. 545 Guido, A., Heindel, D., Birgel, A., Rosso, A., Mastandrea, R., Sanfilippo, F., Russo & J. 546 Peckmann, 2013. Pendant bioconstructions cemented by microbial carbonate in 547 submerged marine caves (Holocene, SE Sicily). Palaeogeography, 548 Palaeoclimatology, Palaeoecology 388: 166–180. 549 Harmelin, J. G., 1969. Bryozoaires des grottes sous‐marines obscures de la région 550 marseillaise, faunistique et écologie. Téthys 1: 793–806. 551 Harmelin J. G., 1980. Etablissement des communautes de substrats durs en milieu 552 obscur. Resultats preliminaires d'une experience à long terme en Mediterranée. 553 Memorie di Biologia Marina e di Oceanografia 10: 29–52. 554 Harmelin, J.G., 1986. Patterns in the distribution of bryozoans in the Mediterranean 555 marine caves. Stygologia 2:10–25 556 Harmelin, J. G., Vacelet, J. & P. Vasseur, 1985. Les grottes sous‐marines obscures: un 557 milieu extreme et un remàrquable biotope refuge. Téthys 11: 214–229. 558 Hofrichter, R. (coord.), 2005. El mar Mediterráneo. Fauna, flora, ecología, Vol II/1: Guía 559 sistemática y de identificación. Ediciones Omega, Sabadell. 560 IBERFAUNA, 2018. El Banco de Datos de la Fauna Ibérica. Museo Nacional de Ciencias 561 Naturales (CSIC). http://iberfauna.mncn.csic.es/ [last accessed on November 562 2018] . 563 Iliffe, T. M., Timothy, D. J. & S. B. Martin, 1984. Organic pollution of an inland marine cave 564 from Bermuda. Marine environmental research. London 12: 173–189. 565 Lafargue, F., Ramos, A. A., Turon, X., Banaigs, B. & M. Wahl, 1986. The littoral ascidians 566 of the Spanish mediterranean. I. From Port Bou to the Islas Medas. Vie et Milieu 567 36: 133–139. 568 Lecher A. L. & K. R. M. Mackey, 2018. Synthesizing the effects of submarine groundwater 569 discharge on Marine Biota. Hydrology: 5‐60. 570 Logan, A., Tomašových, A., Zuschin, M. & B. Grill, 2008. Recent brachiopods from the Red 571 Sea and Gulf of Aden. Fossils and Strata 54: 299–309. 572 Magurran, A. E., Baillie, S. R., Buckland, S. T., Dick, J. M., Elston, D. A., Scott, E. M., Smith, 573 R. I., Somerfield, P. J. & A. D. Watt, 2010. Long‐term datasets in biodiversity 574 research and monitoring: assessing change in ecological communities through 575 time. Trends in Ecology and Evolution 25: 574–582. 576 Maldonado, M., Sánchez‐Tocino, L. & C. Navarro, 2010. Recurrent disease outbreaks in 577 corneous demosponges of the genus Ircinia: epidemic incidence and defense 578 mechanisms. Marine Biology 157: 1577–1590. 579 Mancenido, M. O. & R. Gourvennec, 2007. A reappraisal of feeding current systems 580 inferred for spire‐bearing brachiopods. Earth and Environmental Science 581 Transactions of the Royal Society of Edinburgh 98: 345–356. 582 MAPAMA, 2018. Estado de las masas de agua subterránea PHC 2009‐2015. [online] 583 Available at: http://sig.mapama.es/93/ClienteWS/redes‐seguimiento/ Dirección 584 General del Agua (DGA), Ministerio de Agricultura y Pesca, Alimentación y Medio 585 Ambiente (MAPAMA). Gobierno de España [last accessed on November 2018]. 586 Martí, R., Uriz, M. J., Ballesteros, E. & X. Turon, 2004. Benthic assemblages in two 587 Mediterranean caves: species diversity and coverage as a function of abiotic 588 parameters and geographic distance. Journal of the Marine Biological Association 589 of the United Kingdom 84: 557–572. 590 Montefalcone, M., De Falco, G., Nepote, E., Canessa, M., Bertolino, M., Bavestrello, G., 591 Morri, C. & C. N. Bianchi, 2018. Thirty year ecosystem trajectories in a submerged 592 marine cave under changing pressure regime. Marine Environmental Research 593 137: 98–110. 594 723 724 Fig. 3 Sizes of the square root of the estimates of components of variation for the terms 725 of a factorial PERMANOVA design including the fixed factors Wall and Year. Within each 726 sector, analyses for the morphological descriptor (Md), the taxonomic descriptor (Td), 727 as well as the combination of both descriptors (Md*Td), were carried out separately for 728 semi-dark and dark sectors (results of the PERMANOVAs for taxonomic and 729 morphological descriptors are given in Tab. 2 and 4). Note that the scale of the square 730 root of estimates of components of variation is equivalent to Bray-Curtis dissimilarity 731 percentages among replicates (see Anderson et al., 2008). 732 733 734 735 736 Fig. 4 Temporal variation of mean coverage and 95% confidence interval error bars for 737 prostrate bryozoans (spots and sheets morphologies) in the semi-dark and dark cave 738 sectors. 739 740 741 Fig. 5 Simple linear regressions for total biotic, brachiopods and flattened mounds 742 percent coverages with sediment percent coverage in the dark sector of Cerro-Gordo 743 Cave. 744 745 746 Tab. 1 Classification of Cerro-Gordo Cave taxa to the morphological and taxonomic 747 descriptors considered in this study. h = height; r = radius. 748 749 Morphological descriptor→ Prostrate sheets Flattened Mounds (h < r) Hemi-spherical mounds (h ≈ r) Domed mounds (h > r) Foliaceous/ branched Taxonomic descriptor↓ Macroalgae - Lithophyllum sp. - Peyssonnelia sp. - Mesophyllum spp. - Mesophyllum expansum - Bryopsis plumosa - Dictyota sp. - Gelidium sp. Foraminifera - Miniacina miniacea Porifera - Axinella sp. - Cliona sp. - Dendroceratida - Haplosclerida - Merliida - Plakinidae - Poecilosclerida - Spirastrellidae - Terpios sp. - Clathrina spp. - Dysidea sp. - Oscarella lobularis - Phorbas sp. - Acanthella acuta - Ircinia sp. - Chondrosia reniformis - Haliclona sp. - Petrosia ficiformis Anthozoa - Clavuriidae - Corynactis viridis - Paracyathus pulchellus - Parazoanthus axinellae - Aiptasia mutabilis - Astroides calycularis - Balanophyllia sp. - Caryophyllia sp. - Leptosammia pruvoti - Phyllangia americana mouchezii - Telmatactis forskalii Hydrozoa - Aglaophenia sp. - Eudendrium sp. - Sertularella sp. Polychaeta - Serpulidae - Sabella spallanzanii - Filograna implexa Mollusca - Lithophaga lithophaga - Vermetidae - Limidae - Thylacodes arenarius Brachiopoda - Novocrania anomala Bryozoa - Flustrina - Adeonella calveti - Amathia sp. - Chartella tenella - Myriapora truncata Ascidiacea - Diplosoma spongiforme - Ascidiiae - Phallusia sp. - Polycitor sp. - Aplidium sp. - Clavellina sp. - Stolonica sp. Tab. 2 PERMANOVA results and pair-wise comparisons for the morphological and 750 taxonomic descriptors percent coverage in the semi-dark cave sector. Y = Year; W = Wall; 751 L = left; R = right; df = degrees of freedom; SS = square sums; MS = mean square sums; 752 Res = residuals. 753 Semi-dark sector Source df MS Pseudo-F P(perm) Pair-Wise comparisons Morphological descriptor Year Wall P(perm) Year 2 12761 10.477 <0.01 2008 L = R 0.256 Wall 1 1698 1.394 0.225 2011 L ≠ R <0.05 Y*W 2 4729.7 3.883 <0.01 2016 L ≠ R <0.05 Res 114 1218 Total SS: 175,540 Year: 2011 ≠ (2008 = 2016) Taxonomic descriptor Year 2 3520.6 3.273 <0.01 Year: 2011 ≠ (2008 = 2016) Wall 1 1191.9 1.108 0.361 Y*W 2 1595.3 1.483 0.160 Res 114 1075.8 Total SS: 134,060 754 755 Tab. 3 Mean coverage of taxonomic and morphological descriptors in the outer semi- 756 dark cave sector. Coverage is given in integer numbers for Left and Right walls and the 757 years 2008, 2011 and 2016. * = coverage <1%; A = 2008; B = 2011; C = 2016; blank cell = 758 absent. 759 Semi-dark sector Morphological Prostrate sheets Flattened mounds Hemispherical mounds Domed mounds Foliaceous/ branched TOTAL (Taxonomic) Taxonomic Macroalgae Left 2 2 * * 2 2 * Right 2 1 5 2 1 5 Foraminifera Left * * * * * * Right * * * * Porifera Left 10 6 7 11 7 5 5 5 5 * 4 26 22 17 Right 8 5 11 10 5 6 5 5 5 2 2 25 17 22 Anthozoa Left 5 5 3 17 * 27 * 22 * 2 23 29 30 Right 13 4 15 6 11 4 13 4 1 20 32 19 Hydrozoa Left * * 2 * * 2 Right * 2 1 * 2 1 Polychaeta Left * 4 1 2 * 6 1 Right 1 3 4 * 2 * 2 5 4 Mollusca Left * 1 1 * * * * 1 1 Right * 1 1 * 1 * * 2 1 Brachiopoda Left * * Right * * Bryozoa Left * 1 4 5 5 5 6 5 Right 3 * * 5 10 6 8 10 6 Ascidiacea Left 2 1 2 * 2 1 2 Right 2 1 2 * * 2 1 2 TOTAL (Morphological) Left 20 19 13 28 8 32 6 27 5 * 5 5 8 7 60 69 58 Right 16 23 25 26 11 18 9 18 9 3 2 6 14 8 60 71 60 A B C A B C A B C A B C A B C A B C Cell shading represents SIMPER within-group similarity percentages: - 0 – 0.99% - 1 – 4.99% - 5 – 9.99% - 10 – 19.99% - 20 – 29.99% - 30 – 39.99% - 40 – 49.99% - 50 – 74.99% - 75 – 100% 760 761 762 Tab. 4 PERMANOVA results and pair-wise comparisons for morphological and taxonomic 763 descriptors percent coverage in the dark cave sector. Y = Year; W = Wall; df = degrees of 764 freedom; SS = square sums; MS = mean square sums; Res = residuals. 765 766 Dark sector Source df MS Pseudo-F P(perm) Pair-Wise comparisons Morphological descriptor Year 2 1146.4 2.51 <0.05 Year: 2007 ≠ 2010 ≠ 2016 Wall 1 9624.4 21.068 <0.01 Y*W 2 447.96 0.981 0.42 Res 186 449.3 Total SS: 97,781 Taxonomic descriptor Year 2 13265 18.75 <0.01 Year: 2007 ≠ 2010 ≠ 2016 Wall 1 14591 20.63 <0.01 Y*W 2 1420.1 2.008 0.06 Res 186 707.6 Total SS: 175,880 767 Tab. 5 Mean coverage of taxonomic and morphological descriptors in the inner dark 768 cave sector. Coverage is given in integer numbers for Left and Right walls and the years 769 2007, 2010 and 2016. * = coverage <1%; A = 2007; B = 2010; C = 2016; N/A = not 770 applicable; blank cell = absent. 771 Dark sector Morphological Prostrate sheets Flattened mounds Hemispherical mounds Domed mounds Foliaceous /branched TOTAL (Taxonomic) Taxonomic Macroalgae N/A Foraminifera N/A Porifera Left 22 13 12 3 3 4 * 1 25 16 17 Right 21 11 11 5 * 2 1 * 1 * 28 12 14 Anthozoa Left 3 * 3 Right 2 1 3 * 2 1 3 Hydrozoa N/A Polychaeta Left 24 31 43 24 31 43 Right 23 31 32 23 31 32 Mollusca Left 1 1 1 1 * 1 Right * 2 3 * * * 2 3 Brachiopoda Left 2 5 4 2 5 4 Right 11 19 21 11 19 21 Bryozoa Left * * * * * * Right 1 * 1 * Ascidiacea N/A TOTAL (Morphological) Left 50 44 56 5 8 8 * 1 * 55 52 65 Right 47 45 49 16 20 23 1 * 1 * * 65 66 73 A B C A B C A B C A B C A B C 772 Cell shading represents SIMPER within-group similarity percentages: - 0 – 0.99% - 1 – 4.99% - 5 – 9.99% - 10 – 19.99% - 20 – 29.99% - 30 – 39.99% - 40 – 49.99% - 50 – 74.99% - 75 – 100% - - - - - - 773 774