Impoverished Mobile Epifaunal Assemblages Associated with the Invasive Macroalga Asparagopsis Taxiformis in the Mediterranean Sea

Depósito de investigación de la Universidad de Sevilla https://idus.us.es/ “This is an Accepted Manuscript of an article published by Elsevier in: MARINE ENVIRONMENTAL RESEARCH on 2018, available at: https://doi.org/10.1016/j.marenvres.2018.07.016” Impoverished mobile epifaunal assemblages associated with the invasive 1 macroalga Asparagopsis taxiformis in the Mediterranean Sea 2 3 Carlos Navarro-Barranco1,2*, Marta Florido2, Macarena Ros3, Pablo González-Romero2, José 4 Manuel Guerra-García2. 5 1Departamento de Biología (Zoología), Campus de Cantoblanco, Universidad Autónoma de Madrid, 6 28049, Madrid, Spain. 7 2Laboratorio de Biología Marina, Dpto. Zoología, Facultad de Biología, Universidad de Sevilla. Avda Reina 8 Mercedes 6, 41012 Sevilla, Spain. 9 3Departamento de Biología, CASEM, Facultad de Ciencias del Mar y Ambientales, Campus Universitario de 10 Puerto Real, 11510, Puerto Real, Cádiz, Spain. 11 12 * Corresponding author. E-mail: carlosna[email protected]s. Tel: +34 954556229. 13 14 Abstract 15 There is an increasing concern about the ecosystem consequences of altering macroalgal 16 assemblages. Many macrophytes are foundation species in coastal habitats, supporting much of 17 the biodiversity of these ecosystems by providing essential resources such as food and habitat. 18 The addition of invasive species strongly contributes to habitat modification, but the bottom-up 19 impacts of non-native macroalgae on higher trophic levels remains difficult to predict. The main 20 aim of this study was to evaluate the effects of the invasive macroalga Asparagopsis taxiformis 21 on biodiversity by comparing the mobile epifaunal assemblages inhabiting this species to the 22 dominant native species Halopteris scoparia. This is the first comprehensive study of the effects 23 of this widespread invasive species in higher trophic levels. A hierarchical sampling design with 24 two different spatial scales was conducted to explore the constancy of the patterns observed. 25 Fifty-nine peracarid species were found, accounting for more than 90% of all organisms. A. 26 taxiformis hosted an impoverished epifaunal assemblage in comparison to that associated with 27 the native seaweed, showing significantly lower values of diversity, abundance and number of 28 epifaunal species in all study locations. The epifaunal community structure (both in terms of 29 species composition, variability among samples and relative abundance of the species) was also 30 different. Our results highlighted the strong influence of A. taxiformis in the resident community, 31 with differences among the two macroalgae in all the parameters considered. Finally, our results 32 also reflect a biotic homogenization of the epifaunal assemblages associated to A. taxiformis, a 33 scarcely explored consequence of invasive processes in marine environments. Future studies 34 exploring the cascading effects of the observed changes in the epifaunal assemblages would be 35 necessary in order to estimate system responses to macroalgal invasions. 36 Keywords: Biological invasions, Non-native macroalgae, Epifauna, Asparagopsis taxiformis, 37 Peracarida, Biotic homogenization 38 Introduction 39 Marine macroalgae are widely recognized as ecosystem engineers and/or foundation 40 organisms in many habitats since they transform simple surfaces into structured environments 41 that support hundreds of species (both sessile and mobile) (Bruno et al. 2003; Dijkstra et al. 42 2012; Thomsen et al. 2017). The associated communities, mainly composed by small 43 invertebrates, in turn, support diverse and productive food webs (Taylor & Cole 1994; Taylor 44 1997; Thornber et al. 2017 and references therein). Thus, it is expected that changes to the 45 composition of foundation species (either the addition or removing of seaweed species) will lead 46 to major consequences on ecosystem integrity and functioning (Crooks et al. 2002; Gribben et 47 al. 2009; Ramus et al. 2017). 48 Together with climate change, overfishing and habitat destruction, the introduction of 49 invasive species has been identified as one of the most important threats to marine biodiversity 50 in coastal habitats (Vitousek et al. 1997; Halpern et al. 2008). Current global estimates of 51 introduced macroalgae range from 163 to more than 300 species (Stachowicz et al. 2002; 52 Williams & Smith 2007). All the major groups of macroalgae are represented and their invasion 53 rate has increased in most ecosystems, becoming a very complex and global problem 54 (Stachowicz et al. 2002; Williams & Smith 2007). 55 The ecological impacts of currently known introduced macroalgal species have been 56 quantified from just a small fraction of species (<7%) (Thomsen et al. 2017). Most of these 57 studies have been focused on how non-indigenous species (NIS hereafter) may displace resident 58 primary producers (e.g. space monopolization, change in assemblage composition). Less effort 59 has been devoted to explore the influence of algal invasions on higher trophic levels (Davidson 60 et al. 2015). Recent meta-analyses have revealed the generally negative impact of NIS 61 macroalgae on biodiversity at the same trophic level (via competitive effects), as well as the 62 inability to identify general patterns and mechanism of impact at other trophic levels (Thomsen 63 et al. 2014; Maggi et al. 2015). Bottom up effects can result from diverse and complex 64 mechanisms, so that the influence of NIS macroalgae is often species-specific, depending on the 65 identity of both the invading and native species) (Maggi et al. 2015). 66 67 Almost half of the known algal introductions are in the Mediterranean Sea and NE 68 Atlantic, with most of the NIS being rhodophytes, transported by hull fouling and aquaculture 69 activities (Williams & Smith 2007). Among them, the red alga Asparagopsis taxiformis (Delile) 70 Trevisan de Saint-Léon, 1845 is considered one of the 100 “Worst invasives in the Mediterranean 71 Sea” (Streftaris & Zenetos 2006). This macroalga embodies six cryptic mitochondrial lineages 72 whose geographical distribution, taxonomic and native/introduced status requires revision 73 (Zanolla et al. 2014, Dijoux et al. 2014; Andreakis et al. 2016). Although several of those lineages 74 coexist in the Mediterranean Sea, the dominance and invasive status of the so-called “lineage 75 2” in western Mediterranean is clear (Andreakis et al. 2004, 2007; Altamirano et al. 2008; Zanolla 76 et al. 2017a). Gametophytes of A. taxiformis in most of the Mediterranean countries were first 77 recorded during the last few decades, namely in Spain (Ballesteros & Rodriguez-Prieto 1996), 78 Italy (Barone et al. 2003) or Greece (Tsiamis & Panayotidis 2007). It was first reported in South 79 Iberian Peninsula in 2001 (Báez et al. 2001) and, since then, it has quickly and widely expanded 80 its distribution range, forming conspicuously established stands all over the coast of the Alboran 81 Sea (Altamirano et al. 2008). Previous studies suggest A. taxiformis lineage 2 as the most invasive 82 cryptic taxonomical unit within the genus, with an expected risk of range expansion in tropical 83 areas (Zanolla et al. 2017a). Although its presence has been related with negative effects in 84 native community, most of its ecological impacts remain unknown (Zanolla et al., In press). There 85 have been no studies to explore the ability of free-living epifauna to colonize A. taxiformis, 86 required to assess the consequences of this non-indigenous species on native invertebrate 87 biodiversity. 88 89 Macrofaunal communities associated with marine macroalgae are usually dominated by 90 crustaceans, particularly amphipods and isopods from the superorder Peracarida (Taylor 1997; 91 Frediksen et al. 2005; Guerra-García et al. 2011). They are widely distributed in a range of marine 92 habitats, often playing a key role in trophic interactions, as grazers of macroalgae or their 93 epiphytes and as food for predatory fish (Duffy & Hay, 2000; Poore et al. 2014, Jiménez-Prada 94 et al. 2015). Because of their dominant role within epifaunal assemblages, diversity of life 95 strategies and sensitivity to environmental changes, peracarids are commonly used to monitor 96 the environmental quality of marine habitats (Sánchez-Moyano & García-Gómez 1998; Guerra- 97 García et al. 2009; Thiel 2011). Many studies assessing the impacts of non-indigenous species on 98 the biodiversity of resident communities have used high taxonomic levels of classification (e.g. 99 family, order). For example, all amphipods species (as well as isopods or tanaids) are often 100 grouped together as a simple taxonomic unit (e.g. Raffo et al. 2009; Thomsen et al. 2013; Dijkstra 101 et al. 2017), even though these higher taxa do comprise species with different life cycles and 102 trophic guilds (e.g., predators, herbivores, filter feeders and deposit feeders). Indeed, studies 103 including species-level resolution often reveal more complex impacts due to the specificity of 104 responses (Maggi et al. 2015). Mobile invertebrates inhabiting seaweeds usually are not highly 105 host plant specialist (very few species are found on just one species of macroalgae) and their 106 response (negative, neutral or positive) to seaweed introduction usually are hardly predictable. 107 Generally, non-indigenous species that increase habitat complexity acts as better ecosystem 108 engineers and tend to cause abundances and/or species richness to rise. On the contrary, other 109 related traits such as low epiphytic load, presence of chemical defenses or non-perennial life 110 cycle, tend to have the opposite effect (Viejo et al. 1999; Crooks et al. 2002) 111 Thus, the main aim of this study is to evaluate the effect on epifaunal assemblages of 112 the red alga A. taxiformis. To achieve that, we characterize the peracarid assemblage inhabiting 113 this invasive species and compare with that inhabiting the dominant, sympatric native 114 macroalga Halopteris scoparia (Linnaeus) Sauvageau, 1904. Both species are erect, flexible and 115 among the most common macroalgae on shallow rocky bottoms in the study area. Macroalgae 116 from the genus Halopteris support diverse epifaunal assemblages due to their dense 117 ramification, perennial behavior and accumulation of large amount of detritus (Sánchez- 118 Moyano et al. 2000; Vázquez-Luis et al. 2012). On the contrary, species belonging to genus 119 Asparagopsis have a seasonal growth form and produce halogenated secondary metabolites, 120 which can act as herbivore deterrents and defenses again sessile epiphytes (Altamirano et al. 121 2008; Manilal et al. 2010). Thus, our initial hypothesis is that A. taxiformis would host an 122 impoverished mobile epifaunal assemblage with a lower abundance of organisms and fewer 123 species in comparison to the native Halopteris. Samples were collected at different spatial scales 124 to explore the constancy of the patterns obtained and test whether the establishment of the 125 invasive macroalga results in a biotic homogenization of the associated assemblages. 126 127 Material and Methods 128 129 Study area and macroalgae characterization 130 131 Sampling was conducted in the Mediterranean coast of South Spain (Alboran Sea) during 132 July 2014. The increasing distributional range of Asparagopsis taxiformis and their invasive 133 status in the Alboran Sea has been previously established by Altamirano et al. (2008). The high 134 biodiversity of the benthic communities in this region is mainly determined by the heterogeneity 135 of habitats, the overlapping of Mediterranean and Atlantic species and the upwelling of deep 136 water in the coastal area (Cebrián & Ballesteros, 2004). Six different rocky shore stations were 137 considered along 25 km of coast: “Marina del Este” (ME; 36°43’18’’N, 3°43’40’’W), “La 138 Herradura” (LH; 36°44’16’’N, 3°45’16’’W), “Cantarriján” (CN; 36°44’15’’N, 3°46’39’’W), “El 139 Cañuelo” (CA; 36°44’’32’’N, 3°47’17’’W), “Molino de Papel” (MO; 36°45’3’’N, 3°48’59’’W) and 140 “El Morche” (EM; 36°44’22’’N, 4°0’6’’W). A. taxiformis is characterized by a triphasic and 141 heteromorphic life cycle. In the present study, we focused on the gametophyte stage of the 142 macroalgae, which have been proved to be the dominant stage in the study area and thus likely 143 to represent the changed habitat due to invasion (Zanolla et al. 2017b). 144 To quantify the abundance of each algal species, fifteen random photographic samples 145 of 0.25m2 were taken at each station in order to estimate the coverage of A. taxiformis and H. 146 scoparia. Pictures were later analyzed by dividing each 50x50cm square into 25 subquadrats and 147 counting the number of quadrats where the species were present. Mean coverage values in 148 percentage (± SE) of each macroalgae per station were obtained. 149 150 Epifaunal sampling design 151 152 To test the hypothesis that the epifaunal assemblages inhabiting A. taxiformis would 153 differ from H. scoparia, we sampled both species from three different sites (separated by tens 154 of meters) at each station. Four replicates (separate by meters) of each macroalga per site were 155 collected. Seaweed individuals at depth of 1.5-3 m were carefully detached from the 156 substratum, enclosed in a plastic container in order to minimize loss of organisms and preserved 157 in 70% ethanol. At the laboratory, each sample was washed through a 0.5 mm mesh sieve with 158 fresh water to collect all mobile macrofauna. Organisms were sorted into higher taxa and 159 counted. Additionally, all peracarid individuals were further identified to species level using a 160 binocular microscope. 161 The seaweed volume of each sample was measured by the displacement of a known 162 volume of water when the macroalga was introduced. Both wet and dry weight (after placing 163 the samples in a stove to 70°C for 24h) of each replicate were also measured. Volume was used 164 as a proxy of habitat quantity (all densities values were as nº of individuals/1000 ml) since both 165 weight measures were highly correlated with volume values. 166 Macroalgal complexity is known to correlate with abundance and richness of epifaunal 167 assemblages (i.e. Warfe et al. 2008; Torres et al. 2015). To test whether habitat complexity could 168 explain differences in epifaunal assemblages of A. taxiformis and H. scoparia, we calculated the 169 fractal measures of both area (Da) and perimeter (Dp) in five samples of each species, following 170 the methodology described by McAbendoth et al. (2005). Although quite related, the 171 information provided by these two parameters is slightly different; higher Dp values indicate a 172 higher degree of convolution of macrophyte edge and Da is an estimate of the way the 173 macroalga stands are divided up (higher values are associated with more and smaller divisions) 174 (McAbendroth et al. 2005). These fractal indices have been proved to be a useful parameter to 175 understand the response of epifaunal assemblages associated with native and invasive 176 macroalgae (Veiga et al. 2014; Suárez-Jiménez et al. 2017). Given the small body size of most of 177 epifaunal organisms, it might be expected that the relevant scale of macroalga complexity 178 should be also small. Thus, two different scales were considered: the morphological complexity 179 of the whole branches (low magnification) and detailed structures at the tip of the branch 180 (higher magnification) 181 182 Data analyses 183 184 Peracarid species richness, density and Shannon-Wiener diversity values (Shannon- 185 Weaver 1963) were calculated for each sample. The influence of seaweed identity on these 186 parameters across all sampling stations was tested using a three way analysis of variance 187 (ANOVA) with the following factors: Alga (Al) was a fixed factor with two levels (A. taxiformis 188 and H. scoparia); Station (St) was a random factor, orthogonal with Algae and with six levels 189 (each rocky shore location considered in the study); Site (Si) was a random factor, nested within 190 Station and with three levels (n=4). Heterogeneity of variance was checked using Cochran’s test, 191 and transformations were applied when necessary (Underwood 1997). The same analysis was 192 also conducted individually for the standardized abundance (individuals/1000 ml) of each 193 peracarid species. When ANOVA indicated a significant difference for a given factor, the source 194 of difference was identified using the Student-Newman-Keuls (SNK) tests. 195 The structure of the peracarid assemblage was contrasted considering the species 196 abundance matrix for all sampling stations using a permutational multivariate analysis of 197 variance (PERMANOVA), following the same design used for the univariate analyses. Two 198 different analyses were conducted, based on Bray-Curtis similarity matrix generated using both 199 untransformed and presence/absence data. Terms found to be significant in the full model were 200 examined individually using appropriate pair-wise comparisons. Monte Carlo P-values were 201 used in pair-wise test due to low number of unique permutations (Anderson 2005). Finally, the 202 analysis CLUSTER was also conducted using the mean values per site to determine the spatial 203 patterns of the peracarid assemblage inhabiting each macroalga. The permutation test SIMPROF 204 was used to assess clusters with a significant internal structure (at 5% level). A permutation 205 analysis of multivariate dispersions (PERMDISP) was also carried out to test differences in the 206 variation of the peracarid assemblage among both macroalgae and a non-parametric 207 multidimensional scaling (nMDS) to visualize those differences in structure and variability 208 between the two seaweed host species. 209 Univariate analyses were implemented using GMAV-5 (Underwood et al. 2002), while 210 multivariate analyses were done using the PRIMER v. 6 + PERMANOVA package (Anderson et al. 211 2008). 212 213 Results 214 215 Habitat structure 216 217 Both species were the dominant macroalgae in the study area. Coverage of Halopteris 218 scoparia ranged from 4.4% (in Cantarriján) and 62.5% (in El Morche), while coverage of 219 Asparagopsis taxiformis ranged from 3.2% in El Morche and 19% in la Herradura (Figure 1). 220 Regarding complexity measures, H. scoparia showed significantly higher values for fractal 221 dimension based on perimeter (Dp) at low magnification (F = 17.39; P = 0.003) and based on 222 area (Da) at both low (F = 14.01; P = 0.006) and high magnification (F = 8.71; P = 0.018) (Figure 223 2). 224 225 Epifaunal assemblages 226 227 A total of 30,363 specimens were collected in the study, belonging to the phyla 228 Arthropoda, Mollusca, Annelida and Echinodermata. In both algae, the vast majority of 229 individuals were from the superorder Peracarida, accounting for 93.7% of all organisms in H. 230 scoparia and 92% in A. taxiformis. Within the peracarids, Amphipoda was the dominant order 231 both in number of species (44) and number of individuals (97%), followed by Isopoda (10 species; 232 1,3% total abundance), Tanaidacea (4; 1,2%) and Cumacea (1; 0.5%). Table 1 presents the mean 233 abundance (± SE), frequency of occurrence and full name of all peracarid species found during 234 the whole study in each macroalga. Dominant species were the amphipods Stenothoe 235 monoculoides, Hyale camptonyx, Aora spinicornis, Gammaropsis palmata, Gammaropsis 236 maculata and Caprella acanthifera. 237 Asparagopsis taxiformis supported significantly lower density, species richness and 238 diversity of peracarids than H. scoparia. The absence of any interactions between algal species 239 and other factors indicates the spatial constancy of this pattern (Figure 3, Table 2). Density and 240 species richness varied among stations, and there were no significant differences in the 241 epifaunal assemblage at lower spatial scales (i.e. among sites) (Table 2). 242 The species composition of the peracarid assemblage varied significantly among 243 macroalga species, stations and sites, with interactions between all factors significant (using 244 both untransformed data and presence/absence matrix) (Table 3). These results reveal a high 245 heterogeneity of responses in the structure of the assemblage. Pair-wise analyses indicate that 246 the structure significantly differed between the native and the invasive species in 17 out of 18 247 study sites for untransformed data, and 14 out of 18 sites for presence/absence. 248 The nMDS ordination showed the separation between both macroalgae, as well as the 249 greater variability in the epifaunal assemblage hosted by A. taxiformis (Figure 4). PERMDISP 250 analyses indicated that differences in variability among substrates are significant (P < 0.001). 251 Despite the higher degree of homogeneity showed by H. scoparia, CLUSTER analysis revealed 252 the existence of a defined spatial structure, highlighting significantly different groups which 253 often match with different sampling stations. In the opposite way, this spatial differentiation is 254 not evident for A. taxiformis, where all samples constitute a single group (Figure 5). 255 256 Discussion 257 258 Mobile epifauna impoverishment 259 260 The mobile epifauna inhabiting the invasive alga, Asparagopsis taxiformis, is significantly 261 impoverished in comparison to the co-occuring native macroalga Halopteris scoparia. The 262 smaller species richness, abundance of organisms and diversity associated with A. taxiformis 263 was a constant pattern across small (sites) and medium (stations) spatial scales. 264 Previous comparative studies dealing with mobile macrofauna associated to invasive 265 and native macrophytes highlighted that the consequences and magnitude of seaweed 266 introduction are heterogeneous. Thus, those responses range from negative (e.g. Raffo et al. 267 2009; Janiak & Whitclatch 2012) to neutral (e.g. Wikstrom & Kautsky 2004; Buschbaum et al. 268 2006) and positive (e.g. Schmidt & Scheibling 2006; Guerra-García et al. 2012). Although many 269 different factors have been proposed to explain the magnitude and direction of these effects (or 270 the lack of those), algal morphological complexity has been often highlighted as highly relevant 271 (Gestoso et al. 2012; Veiga et al. 2016). In recent work comparing epifaunal assemblages on the 272 widely invasive kelp Undaria pinnatifida and other native seaweed species, it was concluded 273 that the morphology complexity of the alga (rather than its native or invasive status) was the 274 key factor to predict the impacts of invasive macroalgae (Suárez-Jiménez et al. 2017). Our 275 complexity measurements support these conclusions. Among all the parameters considered, 276 only the fractal dimension based on perimeter at high magnification showed similar values in 277 both algae, while the native algae exhibit significant higher values in the other three 278 measurements. This results indicate a higher abundance of smaller gaps in -Halopteris branches 279 when observed at both low and high scale, which can be related with a higher capacity of the 280 species to provide refuge to epifauna. However, both macroalgae show a complex three- 281 dimensional structure, which could not be properly taken into account with any of the current 282 estimate of fractal dimension based on image analyses, since they force plants onto a two- 283 dimensional plane. 284 Other macroalgae traits commonly related with negative effects over the epifaunal 285 assemblage are low epiphytic load, presence of chemical defenses and seasonal life cycle. All 286 these (often linked) characteristics are present in A. taxiformis and might be responsible of their 287 impoverished epifauna. The patterns in the epifaunal assemblage could be also mediated by 288 top-down processes; for instance, decreased values of abundance, richness, etc. could be 289 associated to increased predation pressure on Asparagopsis stands. However, the proved 290 deterrent activity of some bioactive metabolites synthesized by A. taxiformis (already proposed 291 for the control of herbivorous and predaceous fishes) suggests that this is not the most likely 292 scenario (Manilal et al. 2010). 293 294 Species composition 295 296 The epifaunal community structure is also significantly different in both macroalgae. 297 There are many species whose abundance is significantly higher in Halopteris (Table 1). On the 298 contrary, only two species (Elasmopus vachoni and Quadrimaera inaequipes) showed 299 significantly higher densities in A. taxiformis. These species have been already reported in other 300 invasive macrophytes such as Asparagopsis armata Harvey, 1855 and Caulerpa cylindracea 301 Sonder, 1845 (Vázquez-Luis et al. 2009; Soler-Hurtado & Guerra-García 2010) and its higher 302 abundance on A. taxiformis could reflect the availability of some epifaunal taxa (generally 303 opportunistic and highly tolerant species) to take advantage of the new habitat provided by non- 304 native species and being benefited by his presence (Thomsen et al. 2017). 305 Most of the exclusive species found during the study (either in Halopteris or 306 Asparagopsis) were present in very low abundances, so that there is no room for speculation 307 about the ecological processes involved. In fact, it is considered that hard substrate epifaunal 308 taxa generally do not show strong habitat specialization; in addition, previous studies have also 309 highlighted that differences in epifaunal assemblages between native and invasive macroalgae 310 occurs mainly in terms of abundance rather than of species composition (Taylor 1997; Gestoso 311 et al. 2010). However, the analysis conducted here with presence/absence data showed 312 differences in species composition between native and invasive macroalgae in most locations, 313 revealing that the impact of A. taxiformis in the assemblage is stronger than those produced by 314 other introduced seaweed species. 315 316 Biotic homogenization 317 318 As well as for other parameters, differences in the multivariate dispersion of fauna 319 associated with invasive and native stands will depend on the macroalgae species compared 320 (Engelen et al. 2013). In our case, the higher similarity between samples of Halopteris is also 321 reflected in the higher frequency of occurrence of most of the species. Considering as “common 322 species” those present in at least 75% of the macroalga samples, 11 peracarid species were 323 commonly found in H. scoparia while only two were observed in the invasive. Despite being 324 more homogeneous, there were differences in the peracarid fauna associated with the native 325 species among several locations (i.e. El Morche, Cañuelo, Marina del Este); on the contrary, 326 these spatial differences were not evident for Asparagopsis samples. Thus, the spreading of A. 327 taxiformis not only translate in an impoverished epifaunal assemblage but also in their 328 homogenization. The loss of taxonomical or functional singularity is a common pattern on 329 invaded systems and it is a recent emergent area in conservation biology (McKinney & Lockwood 330 1999, Olden 2006). In the marine realm, biotic homogenization as a result of NIS has been 331 already suggested but its effect in key groups of small and highly diverse groups as peracarids 332 remains unexplored (Galil 2007; de Amicis & Foggo 2015). Both the high heterogeneity among 333 samples and the biotic homogenization associated to A. taxiformis could be explained by 334 impediment factors for the development of a mature and well structured community (e.g. the 335 lack of a perennial life cycle or the absence of shared evolutionary relationship between the 336 macroalga and native epifaunal species). 337 338 Implications for conservation and conclusions 339 340 Monitoring the epifaunal temporal dynamics would be advisable in order to obtain an 341 improved understanding of the consequences of the invasion. Episodes of explosive growing and 342 dominance of A. taxiformis (accounting more than 90% of the algal biomass) have been reported 343 in our study area and, due to its seasonal life cycle, these growth episodes can be followed by 344 sudden drops of abundance (Zanolla et al. 2018). Therefore, epifaunal resilience would be 345 conditioned by their response to high variability in the coverage of A. taxiformis. Previous studies 346 pointed out that recovery of the faunal diversity associated to seaweed after disturbance events 347 can be a complex and long-term process (Lambert et al 2014, Marzinelli et al. 2015). However, 348 the high colonization capability of peracarid populations (Poore 2005) and the absence of a 349 complete competitive displacement of the native macroalgae by A. taxiformis could reduce the 350 diversity loss. 351 Overall, our results pointed out the strong influence of A. taxiformis on the mobile 352 epifaunal assemblage associated, with significant differences regarding all parameters 353 considered: species richness, abundance, diversity, community structure (both in terms of 354 species composition and relative abundance of the species), variability and spatial structure. The 355 observed patterns, together with the high dominance of the two algae here considered in 356 606 607 0 10 20 30 40 50 60 70 ME LH CN CA MO EM Coverage (%) Stations Halopteris Asparagopsis 608 R² = 0,985 R² = 0,739 -1 0 1 2 3 4 5 6 7 8 0 5 10 15 20 25 0 5 10 15 20 25 Dry weight Wet weight (g) Volume (ml) Wet weight Dry weight Lineal (Wet weight) Lineal (Dry weight) R² = 0,983 R² = 0,840 0 2 4 6 8 10 12 0 5 10 15 20 25 30 35 40 0 5 10 15 20 25 30 35 Dry weight (g) Wet weight (g) Volume (ml) A) B) 609 610 0 0,5 1 1,5 2 Da Dp Da Dp Asparagopsis Halopteris HIGH MAGNIFICATION LOW MAGNIFICATION n.s. n.s. n.s. * Fractal dimension 0 5 10 15 20 25 ME LH CN CA MO EM Number of species / replicate 0 20000 40000 60000 80000 100000 ME LH CN CA MO EM Number of individuals / 1000ml 0 0,5 1 1,5 2 2,5 ME LH CN CA MO EM Shannon-Wiener values / replicate Stations Asparagopsis taxiformis Halopteris scoparia A) B) C) 611 612 Macroalgae Halopteis scoparia Asparagopsis taxiformis 2D Stress: 0,21 Samples 100 80 60 40 20 Similarity H1 H2 H3 H4 H5 A1 Macroalgae Halopteis scoparia Asparagopsis taxiformis Station Marina del Este Cantarriján Cañuelo Molino de Papel El Morche La Herradura 613 Asparagopsis taxiformis Halopteris scoparia Species Abundance (individuals/1000ml) Frequency of ocurrence (%) Abundance (individuals/1000ml) Frequency of ocurrence (%) AMPHIPODA Ampelisca serraticaudata 48 1 1 Amphilochus neapolitanus 13275 44 35 112377 210** 88 Amphilochus piccadurus 857 9 3 473 4 4 Ampithoe ramondi 11890 100 17 11411 46 32 Aora spinicornis 78698 145 88 241055 503 96 Apherusa mediterranea 182 3 1 4276 23 15 Apherusa spp. 43774 236 49 48497 128 57 Apocorophium acutum 3757 17 17 45906 158* 58 Atylus swammerdami 278 3 3 Caprella acanthifera 190974 552 86 80873 208 64 Caprella hirsuta 31544 266 17 594 4 6 Caprella liparotensis 11645 42 26 1333 15 3 Caprella mitis 8453 40 15 222248 1206 17 Dexamine spiniventris 9640 28 33 39985 67*** 89 Dexamine spinosa 7617 30 24 28677 50** 81 Elasmopus vachoni 37047 116* 50 13028 46 33 Ericthonius cf. argenteus 5619 27 17 10144 56 22 Eusiroides dellavallei 0 3073 19 13 Gammarella fucicola 1210 8 7 670 6 4 Gammaropsis maculata 117789 503 65 199133 358* 90 Gammaropsis palmata 32455 114 38 302102 624*** 89 Guernea coalita 274 3 3 2537 12 18 Hyale camptonyx 25662 172 44 332095 579** 97 Hyale sp. 1679 15 4 19159 116 10 Iphimedia minuta 0 167 2 1 Ischyrocerus inexpectatus 11322 45 28 15182 58 35 Jassa slateryi 7649 45 21 7346 30 22 Lembos websteri 9888 103 13 5976 32 19 Leucothoe spinicarpa 275 3 3 355 4 3 Liljeborgia sp. 53 1 1 Lyssianassidae 74 1 1 729 4 8 Maera sp. 358 3 4 Metaphoxus sp. 48 1 1 Microdeutopus cf. chelifer 36834 75 65 400883 1083 96 Microprotopus longimanus 3810 16 21 116512 245 78 Peltocoxa marioni 250 3 1 809 7 6 Perionotus testudo 2256 19 4 Phtisica marina 18976 60 46 42861 111 65 Podocerus variegatus 27423 101 44 40059 76* 79 Pseudoprotella phasma 1583 9 10 617 6 4 Quadrimaera inaequipes 42313 176** 46 2581 23 8 Stenothoe catai 8792 38 21 7328 59 11 Stenothoe monoculoides 88193 189 60 485226 863** 100 Stenothoe tergestina 20007 78 26 25811 188 24 ISOPODA Astacilla cf. Axelli 167 2 1 Cymodoce cf. truncata 125 2 1 2970 17 11 Dynamene magnitorata 3753 16 21 7869 35 28 Janira maculosa 2669 13 13 729 5 6 Janiropsis breviremis 417 4 3 Joeropsis brevicornis 154 2 1 Paranthura costana 608 4 6 2236 9 17 Paranthura nigropunctata 1987 9 13 10689 21* 56 Stenosoma nadedja 11567 60 26 Uromuna cf. petiti 9031 47 21 TANAIDACEA Tanais dulongii 5637 24 19 38532 117 54 Zeuxo cf. normani 100 1 1 Zeuxo sp. 154 2 1 Chondrochelia cf. savignyi 286 4 1 167 2 1 CUMACEA Cumella cf. limicola 297 3 4 Table 1. Abundance (Mean SE) and Frequency of ocurrence of the peracarid species associated with the invasive and native macroalgae. The abundance of those species with significant higher values in one seaweed are higlighted in grey, as well as those species with frequency of ocurrence of at least 75%. * = P< 0.05; ** = P< 0.01; *** = P< 0.001. 614 615 616 Species richness Abundance Diversity Source of variation df MS F P MS F P MS F P Algae = Al 1 1534.028 73.13 <0.001*** 263504.543 32.98 0.002** 3.953 17.35 0.009** Station = St 5 51.550 5.84 0.006** 34781.920 20.36 <0.001*** 0.343 2.34 0.105 Site = Si 12 8.833 0.79 0.658 1108.368 1.18 0.304 0.147 1.58 0.109 Al x St 5 20.978 1.15 0.388 7989.835 3.06 0.052 0.228 1.82 0.184 Al x Si (St) 12 18.278 1.64 0.092 2608.388 1.81 0.056 0.125 1.35 0.202 Residual 108 11.157 1443.088 0.093 Results of three way ANOVA for species richness (n of species/replicate), abundance (n of individuals/1000ml) and diversity values (Shannon- Wiener values/replicate) . No transformation were necessary for species richness and diversity values. Abundance data were Sqrt (x+1) transformed. ** = P< 0.01; *** = P< 0.001. Untransformed data Presence/abscence Source of variation df MS Pseudo-F P(perm) MS Pseudo-F P (perm) Algae = Al 1 58648.0 6.831 <0.001*** 30848.0 11.414 <0.001*** Station = St 5 15675.0 4.672 <0.001*** 8120.8 4.837 <0.001*** Site (Station) = Si(St) 12 3355.5 2.140 <0.001*** 1678.6 2.256 <0.001*** Al x St 5 8586.0 2.999 <0.001*** 2702.7 1.859 0.024 * Al x Si(St) 12 2862.9 1.826 <0.001*** 1453.8 1.954 <0.001*** Residual 108 1567.9 744.0 Results of the PERMANOVA analysis conducted for the peracarid community, based on Bray-Curtis similarities of untransformed and presence/abscencetransformeddata. Significance P values were obtained by computin 9999 permutations. * = P < 0.05; *** = P< 0.001.