Spatial variability and

Elsevier Editorial System(tm) for Marine Pollution Bulletin Manuscript Draft Manuscript Number: Title: Spatial variability and response to anthropogenic pressures of assemblages dominated by a habitat forming seaweed sensitive to pollution (northern coast of Alboran Sea). Article Type: Research Paper Keywords: Cystoseira, Spatial variability, Rocky shores, Habitat forming species, Anthropogenic pressures, Oceanography and community patterns, Seaweeds. Corresponding Author: Mr. Ricardo Bermejo, Corresponding Author's Institution: Universidad de Cádiz First Author: Ricardo Bermejo Order of Authors: Ricardo Bermejo; Gina de la Fuente; Eduardo RamírezRomero; Juan José Vergara; Ignacio Hernández Abstract: The Cystoseira ericaefolia group is conformed by three species: C. tamariscifolia, C. mediterranea and C. amentacea. These species are among the most important habitat forming species of the upper sublittoral rocky shores of the Mediterranean sea and adjacent Atlantic coast. This species group is sensitive to human pressures and therefore is currently suffering important losses. This study aimed to assess the influence of anthropogenic pressures, oceanographic conditions and local spatial variability in assemblages dominated by C. ericaefolia. The results showed the absence of significant effects of anthropogenic pressures or its interactions with environmental conditions in the Cystoseira assemblages. This fact was attributed to the high spatial variability, which is most probably masking the impact of anthropogenic pressures. The results also showed that most of the variability occurred on at local levels. A relevant spatial variability was observed at regional level, suggesting a key role of oceanographic features in these assemblages. Suggested Reviewers: Luisa Mangialajo ECOMERS, Nice Sophia Antipolis University [email protected] First author of the article: Mangialajo, L., Chiantore, M., CattaneoVietti, R., 2008. Loss of fucoid algae along a gradient of urbanisation, and structure of benthic assemblages. Mar. Ecol. Prog. Ser. 358, 63–74. This article is cited several times along the manuscript. Fernando Tuya Postdoctoral researcher, Universidad de Las Palmas de Gran Canaria [email protected] Author of the article: Tuya, F., Haroun, R., 2006. Spatial patterns and response to wave exposure of shallow water algal assemblages across the Canarian Archipelago: a multi-scaled approach. Mar. Ecol. Prog. Ser. 311, 15–28.

Manuscript (clean - no tracked changes) Click here to view linked References

Spatial variability and response to anthropogenic pressures of 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

assemblages dominated by a habitat forming seaweed sensitive to pollution (northern coast of Alboran Sea). 1

2

3

1

Ricardo Bermejo , Gina de la Fuente , Eduardo Ramírez-Romero , Juan J. Vergara , and Ignacio Hernández 1

1

Departamento de Biología. Área de Ecología. Facultad de Ciencias del Mar y Ambientales,

Universidad de Cádiz. 11510 Puerto Real. Cádiz. Spain. 2

Dipartimento di Scienze della terra, dell'ambiente e della vita (DISTAV). Università degli Studi

di Genova. Corso Europa, 26 16132 Genova (Italy). 3

GEOMAR Helmholtz Centre for Ocean Research Kiel, Experimental Ecology (Foodwebs)

Düsternbrooker Weg 20, D-24105 Kiel, Germany.

Corresponding author: Ricardo Bermejo Departamento de Biología (Área de Ecología). Facultad de Ciencias del Mar y Ambientales. Universidad de Cádiz. Campus de Río San Pedro, 11510 Puerto Real (Cádiz, Spain). Telephone: +34-956016029 / Fax: +34-956 016019 e-mail: [email protected]

ABSTRACT

The Cystoseira ericaefolia group is conformed by three species: C. tamariscifolia, C. mediterranea and C. amentacea. These species are among the most important habitat forming species of the upper sublittoral rocky shores of the Mediterranean Sea and adjacent Atlantic coast. This species group is sensitive to human pressures and therefore is currently suffering important losses. This study aimed to assess the influence of anthropogenic pressures, oceanographic conditions and local spatial variability in assemblages dominated by C. ericaefolia. The results showed the absence of significant effects of anthropogenic pressures or its interactions with environmental conditions in the Cystoseira assemblages. This fact was attributed to the high spatial variability, which is most probably masking the impact of anthropogenic pressures. The results also showed that most of the variability occurred on at local levels. A relevant spatial variability was observed at regional level, suggesting a key role of oceanographic features in these assemblages.

Keywords: Cystoseira, Spatial variability, Rocky shores, Habitat forming species, Anthropogenic pressures, Oceanography and community patterns, Seaweeds.

1.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

Introduction

Rocky shores around the world are increasingly being subjected to a variety of anthropogenic stresses, acting at different spatial and temporal scales (Airoldi and Beck, 2007; Coll et al., 2010), and producing noteworthy shifts between alternative stable states in marine ecosystems (Knowlton, 2004; Viaroli et al., 2008; Orfanidis et al., 2011). Recent research has been focused on identifying and assessing the effects of local anthropogenic pressures on the littoral community (e.g. Diez et al., 1999; Arévalo et al., 2007; Seridi et al., 2007) and its interaction with global stressors as ocean warming and acidification (e.g. Brown et al., 2013; Strain et al., 2015). This information is especially important for management, since local anthropogenic stressors are more easily amendable by management and conservation actions (Russell and Connell, 2012; Brown et al., 2013). However, these anthropogenic stressors are superimposed on the stress caused by natural environmental factors and it is difficult to distinguish their contribution (Crowe et al., 2000; Bermejo et al., 2013). Furthermore, these pressures act together, causing non additive (i.e. synergistic, antagonistic) or additive effects on littoral communities (Knowlton and Jackson, 2008; Brown et al., 2013; Strain et al., 2015). These facts make the discrimination between the two, and the achievement of effective management actions, more difficult. In this sense, proper experimental designs at different spatial scales are useful to estimate the effects produced by anthropogenic stressors on the community, and to identify putative non-additive interactions with environmental conditions.

Canopy forming brown seaweeds which belong to the orders Laminariales, Tilopteridales and Fucales, are among the main habitat forming species on most temperate rocky shores (Lüning, 1990). Currently, many of these species are suffering strong declines in their populations worldwide, which have been attributed to local and global stressors in the context of global change (Steneck et al., 2002; Fernández, 2011; Mineur et al., 2015). In this sense, coastal development has been pointed out as one of the most important factors explaining the loss of habitat forming macrophytes, mainly as a consequence of the increase in water turbidity and eutrophication as well as other habitat related changes (Airoldi and Beck, 2007; Mangialajo et al., 2008). These losses are causing relevant deleterious consequences for local economies and biodiversity (Serio et al., 2006; Voerman et al., 2013). In fact, habitat destruction or degradation is considered the most important threat to the diversity, structure, and functioning of marine coastal ecosystems and to the goods and services they provide (Lotze et al., 2006; Airoldi and Beck, 2007; Coll et al., 2010).

Significant losses of Cystoseira forests (Fucales) have been reported (Thibaut et al., 2005; Serio et al., 2006) and several species of this genus have been identified sensitive to human disturbances in the Mediterranean Sea (Arévalo et al., 2007; Seridi et al., 2007; Mangialajo et al., 2008) and proximate Atlantic coasts (Diez et al., 1999) along different pollution gradients. On these coasts, Cystoseira ericaefolia group is among the most important marine habitat

forming species on littoral and upper sublittoral rocky shores. This group is conformed by three

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

closely related species: C. amentacea, C. mediterranea and C. tamariscifolia (Giaccone and Bruni, 1971; Amico et al., 1985; Gómez-Garreta et al., 1994); which are not possible to be assigned unambiguously on the basis of their morphology (Ballesteros and Catalán, 1981; Gómez-Garreta et al., 1994). These species can constitute extensive and dense forests in wave exposed or moderately exposed places, in the littoral and upper sublittoral zone (Barceló-Martí et al., 2000; Rodríguez-Prieto et al., 2013), where they are a key element of the landscape (Ballesteros et al., 2007; Bermejo et al., 2013; Thibaut et al., 2014). The Cystoseira forests form a complex 3-dimensional physical structure, providing a complex habitat for other algae, invertebrates and fishes (Bellan and Bellan-Santini, 1972; Bulleri et al., 2002; Cheminée et al., 2013) and thus playing an essential role in the conservation of the biodiversity and ecosystem functioning (Ballesteros, 1989; Giaccone et al., 1994).

Due to the ecological importance of assemblages dominated by Cystoseira spp. and the decline of their populations within the past decades, these species have recently been protected in the Mediterranean Sea (Annex II of the Barcelona Convention, COM/2009/0585 FIN), their assemblages being considered as habitats of community interest by the UE (Directive 92/43/EEC; Annex I, "Rocky reefs"). Furthermore, as a consequence of the sensitivity of these species to a variety of anthropogenic pressures, they are considered as a relevant biological element by different Atlantic (Díez et al., 2012) and Mediterranean (Orfanidis et al., 2003; Ballesteros et al., 2007) indices to assess the ecological status in the context of the Water Framework Directive (WFD; Directive 2000/60/EC). Previous data of assemblages dominated by Cystoseira ericaefolia species are available in different localities along the eastern (e.g. Ballesteros et al., 1984; Ballesteros, 1988; Boisset and Gómez-Garreta, 1989) and western Mediterranean Sea (Soltan et al., 2001; Benedetti-Cecchi et al., 2001; Bulleri et al., 2002). However, in the Alboran Sea, the most western ecoregion of the Mediterranean Sea, these studies are scarce, being too local (Ballesteros and Catalán, 1981) or general (Ballesteros and Pinedo, 2004).

The position of the Alboran Sea, in the transition towards the Atlantic Ocean, and its special orographic features, produces complex oceanographic and meteorological conditions (GarcíaLafuente et al., 1998; García-Lafuente and Ruiz, 2007), which determine the distribution and the structure of the littoral and sublittoral communities (Conde, 1989; Bermejo et al., 2013). This area can be considered as a soft transition between the Atlantic Ocean and the Mediterranean Sea (Báez et al., 2004; Ballesteros and Pinedo, 2004), and three biological subregions matching with regional oceanographic patterns can be identified. This feature suggests a key role of regional oceanography in the structure and composition of littoral communities along the northern coast of this sea (Bermejo et al., 2015).

In this context, the main objectives of the study were: i) to assess the influence of anthropogenic

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

pressures, oceanographic conditions and local spatial variability in determining the taxonomic and functional composition of littoral assemblages dominated by C. ericaefolia species along the rocky shores of the northern coast of the Alboran Sea; and ii) to establish a suitable baseline describing the structure and composition of assemblages dominated by C. ericaefolia species along the studied coasts. To accomplish these goals an observational experiment was developed, taking advantage of the complex oceanographic conditions and the wide range of anthropogenic disturbances that can be found along this coast. Due to the lack of previous data for the studied localities, all comparisons were developed between localities under different levels of anthropogenic pressures in each of the biological subregions previously identified along these coasts.

2. Materials and methods

2.1 Study area

The northern coast of the Alboran Sea is mainly oriented in a longitudinal direction (from aprox. 5º 30' W to 0º 30' W, and from 36º N to 37º 30' N; Fig. 1), which reduces the influence of other confounding climatic factors that covariate with latitude (e.g. day length, solar irradiation, temperature), being environmental conditions mainly driven by local oceanographic conditions (Conde, 1989; Bermejo et al., 2015). Three coastal subregions with different oceanographic dynamics have been identified on its northern coast (Bermejo et al., 2015; Fig. 1): i) western Alboran (from site "Mi" to "TQ"), where littoral environmental conditions are determined by the existence of a semi-permanent upwelling, which determines a lower mean seawater temperature (c.a. 17°C) and richer nutrient waters; ii) eastern Alboran (from site "GV" to "ST"), where the prevailing coastal conditions are determined by the dynamics of Mediterranean Surface Waters (MSW), it shows the broadest thermal amplitude (from 25ºC in summer to 14ºC in winter) and the most oligotrophic character (Parada and Canton, 1998; Baldacci et al., 2001); and iii) central Alboran (from site "Ar" to "CR"), characterized by alternative episodes of upwelling and the presence of MSW, specially during summer. These alternative episodes are driven by winds, and produce short-time oscillations in temperature and nutrient availability, which should cause an acute stress for benthic organisms (Bermejo et al., 2015).

Diverse anthropogenic pressures related to high-density population (e.g. in sites "CH" or "TQ"), intensive agriculture (sites "CR", and "GV"), industrial activities or marine traffic impact these coasts, affecting the development of biological communities (Díaz-de Alba et al., 2011; Alonso Castillo et al., 2013; Bermejo et al., 2013, 2014). However, the studied area also comprises different protected areas under low anthropogenic pressures such as the Natural Parks of "El Estrecho" (Western Alboran; site "Mi") and "Cabo de Gata" (eastern Alboran, sites "Co" and "Ne"), or the Site of Community Importance "Calahonda-Castell de Ferro" (central Alboran, site

"Ri"), which show a high percentage of cover by natural lands and are far from important

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

pollution sources (Fig. 1).

2.2 Anthropogenic pressures assessment

Anthropogenic pressures were assessed based on the land use (e.g. Lopez-Royo et al., 2009; Nikolić et al., 2013) along a two kilometre long and half a kilometre wide stretch following the 2

surrounding coast to the sampling site (ca. 1 km ). Three types of land uses were considered: natural, agricultural and urban. For each type of land use, the percentage of cover was calculated from the total area of the stretch considered using the Quantum Geographical Information System (QGIS). The geographical data needed were obtained from the CORINE land cover maps of Spain (National Geographical Institute, http://www.ign.es/ign/layoutIn/corineLandCover.do). For statistical analysis, sampling sites were classified as follows: under "None/Low" anthropogenic pressures, when the percentage of natural land was higher than 60% and urbanized lands lower than 30%; or under "Moderate/High" anthropogenic pressures, when the percentages of natural land were lower than 60% or urbanized land higher than 30%.

2.3 Sampling of "Cystoseira ericaefolia" assemblages

The samples of assemblages dominated by Cystoseira ericaefolia were taken between 2011 and 2012, at 18 sites along the northern coast of the Alboran Sea (Fig. 1). To diminish the effects of seasonal variability, field samplings were carried out during late spring and mid summer (June-August). 2

Three or four replicates of quadrats of 17x17 cm separated between five and ten metres were taken in assemblages dominated by C. ericaefolia (i.e. stands of few squared metres or dense meadows; the cover of C. ericaefolia in collected quadrats was always >70%). In each quadrat all organisms were scraped off the surface using a hammer and chisel. To decrease community variability due to environmental factors, quadrats were taken in the horizontal intertidal (slope between 0 and 30º) on the upper level of the sub-littoral zone, avoiding very sheltered zones. The samples were kept in 5% formalin. Subsequently, samples were sorted out in the laboratory. Sessile organisms were classified as epiphytes and non-epiphytes, macrophytes were identified to the genus/species level, and invertebrates to order/family level. The taxonomical algal nomenclature followed AlgaeBase (Guiry and Guiry, 2014). The abundance of each taxa was calculated by transforming the fresh weight into volume assuming three different densities based on empirical own data: i) 1.05 mg/mL for flesh seaweeds (e.g. Ceramium spp., Ulva spp., Plocamium spp.); ii) 1.10 mg/mL for thick cartilaginous/leathery seaweeds (e.g. Cystoseira spp., Halophytis spp., Gymnogongrus spp.); and iii) 1.55 mg/mL for calcareous seaweeds (Corallina spp., Jania spp., Mesophyllum spp.). In the case of species with low

volume, algae were sorted out in the three density groups and were weighted together, being

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

the final volume estimated based on the percentage of surface occupied by each species on a petri dish, estimated under a dissecting microscope (Nikon SMZ645). Finally, all the species were classified in seven functional/morphological groups (Littler et al., 1983), which were: i) sheet-like algae; ii) filamentous algae; iii) coarsely branched algae; iv) leathery cartilaginous algae; v) articulate corallines algae; vi) crustose algae; and vii) filter-feeders (Table S1).

2.4 Data analysis

Statistical analyses were performed using the Vegan package for R and PERMANOVA + add on PRIMER 6 (Plymouth Routines in Multivariate Ecological Research) software. In all statistical analyses, significance was set at p-value < 0.05 probability, and when it was necessary they were based on 9999 permutations.

2.4.1 Assessing the effects of local and regional factors on the structure of "Cystoseira ericaefolia" assemblages.

To assess the effects of regional and local pressures on the structure and composition of littoral assemblages dominated by C. ericaefolia species, a permutational multivariate analysis of variance (PERMANOVA; Anderson et al., 2008) was performed. The analysis considered a three-way model, where "Oceanographic Region" (OR; three levels: western, central and eastern) and "Anthropogenic Pressures" (AP; two levels: None/Low, and Moderate/High) were treated as fixed factors, and "Site" was treated as a random factor nested within the interaction "OR x AP" as not all AP situations were present in all sites. The volume of Cystoseira ericaefolia was considered as a covariable, and a type I sum of squared was applied (Anderson et al., 2008). Additionally, a distance-based test for homogeneity of multivariate dispersion (PERMDISP; Anderson et al., 2008) and a non-metric multidimensional scaling (nMDS) analysis (Clarke and Warwick, 2001) were performed to interpret and visualize data patterns. These analyses were based on a Bray-Curtis similarity matrix (Bray and Curtis, 1957) between samples, being these data previously four-root-transformed to reduce the high contribution of the most abundant taxa (Clarke and Warwick 2001). Cystoseira ericaefolia was excluded from the species list because its abundance was considered as a covariable. When necessary, terms were pooled as suggested by Anderson et al. (2008).

2.4.2. Identifying emerging traits of "Cystoseira ericaefolia" assemblages.

To identify the taxa that most contributed to the dissimilarities between the different levels of significant fixed factors, an analysis of species contribution to similarity (SIMPER; Clarke and Gorley, 2006) was carried out.

On the other hand, a three-way permutational univariate analysis of the variance (PERANOVA)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

was applied to each of the different morphological/functional groups, the species number, the epiphytes abundance and the Simpson evenness (1-λ'), to test the same hypotheses described above for multivariate data. In this case, a PERANOVA was used instead of an ANOVA because even after log-transformation, some variables (e.g. filter feeder or crust seaweeds abundances) did not fit the normality and/or homoscedasticity assumptions. These PERANOVA analyses were based on Euclidean distances, being abundance data previously squared root transformed.

Finally, Spearman's correlations (rS) among the abundances of the different morphological/functional groups were measured in order to identify possible biological interactions between organisms.

3. Results

A total of 137 taxa were identified in the fifty-nine quadrats conducted at the eighteen studied sites (Table S2). Eastern Alboran showed the highest species richness (100), followed by western Alboran (84 identified taxa) and Central Alboran (79 identified taxa). However, there were marked differences in the sampling effort between oceanographic subregions (western Alboran -4 sites/13 quadrats-, central Alboran -5 sites/17 quadrats-, and eastern Alboran -9 sites/29 quadrats-) so, when data was standardised considering only four sites, Western Alboran showed the highest species richness (84 taxa), followed by Central (71±3) and Eastern Alboran (67±6 taxa), which yielded similar values. The most frequent taxa in C. ericaefolia assemblages from the northern coast of the Alboran Sea were: Lithophyllum incrustants (76.91% of the samples), Corallina spp. (76.01%), Jania rubens (73.26%), Halopteris scoparia (72.69%), Ceramium secundatum (69.59%), Mytilidae (63.87%), and Herposiphonia secunda (54.18%).

3.1 Assessing the effects of local and regional factors on the structure and composition of "Cystoseira ericaefolia" assemblages.

The PERMANOVA results indicated that assemblages dominated by C. ericaefolia differed significantly between the different sites and oceanographic regions (OR), being the differences significant for all pairwise comparisons between subregions (Table 1). In addition, a significant effect of the covariable Cystoseira canopy (Ca) was also identified. No significant effects were found for the factor anthropogenic pressures (AP) and the interaction between "OR" and "AP". The PERMDISP test for the oceanographic regions found significant differences in the dispersion (F2,56: 9.396; p(perm)