Marine Environmental Research 119 (2016) 222e237

Contents lists available at ScienceDirect

Marine Environmental Research

journal homepage: www.elsevier.com/locate/marenvrev

Assessment the short-term effects of wrack removal on supralittoral using the M-BACI design on Atlantic sandy beaches of Brazil and Spain

* Jenyffer Vierheller Vieira a, ,Ma Carmen Ruiz-Delgado b,Ma Jose Reyes-Martínez b, Carlos Alberto Borzone a, Angelico Asenjo c, Juan Emilio Sanchez-Moyano d, Francisco Jose García-García b a Departamento de Ci^encias da Terra, Universidade Federal do Parana, Centro de Estudos do Mar, Av. Beira Mar, s/n CEP 83255-000, Pontal do Sul, Pontal do Parana, Parana, Brazil b Departamento de Sistemas Físicos, Químicos y Naturales, Universidad Pablo de Olavide, ES-41013 Sevilla, Spain c Departamento de Biologia e Zoologia, Instituto de Bioci^encias, Universidade Federal de Mato Grosso, Av. Fernando Correa da Costa, 2367, CEP 78060-900 Cuiaba, Mato Grosso, Brazil d Departamento de Zoología, Universidad de Sevilla, Av. Reina Mercedes 6, 41012 Sevilla, Spain article info abstract

Article history: Wrack removal has been adopted indiscriminately, with no previous assessment of the ecological im- Received 18 March 2016 plications for sandy beach ecosystem. This study evaluated, through an M-BACI design, the effect of Received in revised form wrack removal on supralittoral arthropods on Atlantic sandy beaches receiving different types of wrack: 4 June 2016 mangrove propagules (Brazil), seagrasses and macroalgae (Spain). Impacted plots were contrasted with Accepted 10 June 2016 controls in 8 successive periods before and after experimental wrack removal. After the disturbance, Available online 13 June 2016 drastic decreases in the densities of the amphipod Platorchestia monodi, coleopterans Cleridae, Nitidu- lidae and Phaleria testacea (Brazilian beaches) and amphipod Talitrus saltator (Spanish beaches) were Keywords: Macroinvertebrates detected in the impacted plots. The recovery patterns of arthropods might be related to wrack features fi Wrack subsidies (amount, composition, and degradation) combined with density and species-speci c strategies (e.g. Disturbance mobility, feeding preferences) in each Atlantic region. The temporary suppression of wrack and its Experimental design associated fauna can have potential effects on the wrack-derived process and food-web structure on Supratidal sandy beaches. Beaches © 2016 Elsevier Ltd. All rights reserved. Atlantic coast Ecosystem management

1. Introduction mechanical operations (Davenport and Davenport, 2006; Dugan and Hubbard, 2010), which remove all litter generated by human Beaches worldwide are important spaces for leisure of the local activity, as well as wrack debris (Colombini et al., 2011; Defeo et al., population, tourists and recreational users (Defeo et al., 2009). To 2009; Dugan et al., 2003; Llewellyn and Shackley, 1996). The satisfy this public demand, local authorities have promoted and complete removal of wrack has attracted the interest of scientists in supported actions that attract and ensure the welfare of all beach understanding the ecological implications inherent to this man- users (Davenport and Davenport, 2006). Cleaning or grooming the agement practice (Dugan et al., 2003; Fairweather and Henry, beach is conducted, to improve the aesthetics, amenity and utility 2003; Gilburn, 2012; Llewellyn and Shackley, 1996). of these systems (Fairweather and Henry, 2003; Noriega et al., From an ecosystem perspective, wrack debris is a key element 2012). This management strategy involves several approaches for the maintenance of biodiversity (Harris et al., 2014) and func- that range from simple manual collection (using rakes), to tioning of sandy beaches (Barreiro et al., 2011; Defeo et al., 2009). Wrack deposits may be composed of several types of organic ma- terials (i.e., marine macrophytes, macroalgae, or propagules from * Corresponding author. mangroves) (Barreiro et al., 2011; Colombini and Chelazzi, 2003; E-mail address: [email protected] (J.V. Vieira). http://dx.doi.org/10.1016/j.marenvres.2016.06.007 0141-1136/© 2016 Elsevier Ltd. All rights reserved. J.V. Vieira et al. / Marine Environmental Research 119 (2016) 222e237 223

Gonçalves and Marques, 2011; Ince et al., 2007; Ruiz-Delgado et al., the whole supralittoral assemblages in the impacted plots 2014). Since sandy beaches have low in situ primary productivity, compared to control plots; 2) wrack removal would lower density their food webs are supported by allochthonous organic debris of supralittoral populations, particularly those species that use imported from the sea and coastal areas (Colombini and Chelazzi, wrack as food and/or shelter; 3) different recovery patterns of 2003; Nel et al., 2014). Besides being a significant food source, supralittoral arthropods in response to wrack removal are expected wrack debris provides beach fauna with a hospitable microhabitat on sandy beaches located in both Atlantic regions (southern Brazil for refuge, reproduction and growth (Colombini and Chelazzi, and south-western Spain) related to the differences in community 2003; Ruiz-Delgado et al., 2014). However, most beach users composition of species and composition of wrack debris. This work consider wrack debris as useless debris, or as an unpleasant is meant to help elucidate the effect of wrack removal at commu- disturbance (Fairweather and Henry, 2003). Usually this perception nity and population scales. is strongly influenced by the disagreeable odor from its decompo- fl sition, which attracts swarms of beach ies and buzzards 2. Materials and methods (Davenport and Davenport, 2006; McLachlan and Brown, 2006). Wrack removal may cause ecological problems by disrupting 2.1. Study area pathways of decomposition and nutrient exchange between ma- rine and terrestrial ecosystems. This exchange forms the basis for This experimental approach was conducted on four sandy primary production and food chains of nearshore waters (Dugan beaches located in two geographical regions (Fig. 1): southern et al., 2011; Barreiro et al., 2013). Moreover, this activity can alter Brazil (Parana State) and south-western Spain (Atlantic coast of the composition of supralittoral invertebrates (such as crustaceans Cadiz), in order to investigate, in a general way, the effect of wrack and ), and therefore, affect beach trophic dynamics by removal (i.e. mangrove propagules, seagrasses and macroalgae) on reducing prey availability to higher trophic levels (bottom-up ef- local and global scales. fects), such as shorebirds, lizards, and rodents (Dugan et al., 2003; The coast of Parana has a humid subtropical climate and semi- Fairweather and Henry, 2003; Gilburn, 2012; Llewellyn and diurnal tides with spring-tide ranges up to 1.7 m (Lana et al., 2001). Shackley, 1996; Martin et al., 2006). Wrack removal also alters the Along this microtidal coast, two beaches were selected for this physical characteristics of the beach environment, mainly sediment study. Assenodi (253502400 S; 482200400 W) is an intermediate to properties, beach morphology, and morphodynamics and prevents dissipative, wave-dominated beach with fine sands and a gentle dune formation (Malm et al., 2004; Ochieng and Erftemeijer, 1999; slope Cem (253402400 S; 482001300 W) is a low-energy reflective fi Piriz et al., 2003). These physical modi cations can cause increasing beach, modified by tides with fine sands and a steep slope (Table 1). fi erosion of the beach pro le and loss of the frontal dune (Nordstrom Both beaches are bordered by typical coastal sand dune vegetation. et al., 2000). These beaches received wrack inputs composed by mangrove Most sandy beach studies related to human impacts have used propagules of Laguncularia racemosa, Avicennia shaueriana and ‘ ’ compare and contrast designs (e.g. Schlacher et al., 2008). In this Rizophora mangle from the estuarine system of Paranagua Bay type of design, the pre-disturbance situation is unknown, and in- (Borzone and Rosa, 2009; Rosa et al., 2007). ferences are made by simple spatial comparison between previ- The Atlantic coast of Cadiz has a dry-summer subtropical ously disturbed and undisturbed areas (Underwood, 2000). climate and semidiurnal tidal regime with a range up to 3.2 m Manipulative experiments are more suitable to determine cause- (Benavente et al., 2002). Levante (363303700 N; 61302700 W) located effect relationships between a disturbance and biological vari- in the outer zone of Cadiz Bay, is a dune-backed, dissipative beach. ables (Glasby and Underwood, 1996). The M-BACI design (multi- It is a wide beach, characterized by a gentle slope and fine-sized variate before and after/control and impact) is considered the most sand (Table 1). During the experiment, this beach received inputs appropriate sampling strategy for evaluating planned impacts of the seagrasses Cymodocea nodosa and Zostera noltii from seagrass (Downes et al., 2004; Underwood, 2000). This design includes beds located around Cadiz Bay. Cortadura (362805800 N; multiple control and impacted locations, which allow differenti- 61507700 W), situated in the southern part of Cadiz Bay, is an in- ating between the effects of impact and the background environ- termediate beach, backed by foredunes and low non-vegetated mental variation. Moreover, the treatments are compared in dune ridges. This beach is narrower than Levante beach and has a multiple sampling dates before (baseline samples) and after the beach profile with a gentle slope and fine sand (Table 1). Cortadura impacting activity. Therefore, this design ensures the correct beach receives inputs of brown macroalgae, such as Dictyopteris interpretation of the interactions between locations and sampling membranacea and Cladostephus spongiosus, several species of red times (Downes et al., 2004). algae, such as Halopithys incurva and Chondria dasyphylla, and fi This study provides the rst assessment of the short-term effect green algae, such as Codium decorticatum and Codium fragile from of wrack removal on the supralittoral arthropods assemblages us- nearby rocky shores and subtidal habitats. ing a field-based experiment, following the M-BACI design. This experimental approach was performed on four sandy beaches 2.2. Experimental design and field procedures located on both sides of the Atlantic Ocean (southern Brazil and south-western Spain), which differ in wrack composition and On two Brazilian beaches (Assenodi and Cem), a field experi- morphodynamic characteristics in order to understand the effects ment was conducted between 24 May and 4 July 2012, whereas on of wrack removal on local and global scales. Moreover, the M-BACI two Spanish beaches (Levante and Cortadura) an analogous design allowed us to evaluate the temporal patterns of arthropods experiment was performed between 2 October and 12 November (community and population level) in impacted plots after the 2012. During this period, climatic conditions were quite similar at removal of wrack debris and compare those with patterns of both regions (Table 1) and we expected a great number of species occurrence of fauna in adjacent control plots. associated with wrack debris (e.g. Gonçalves and Marques, 2011). In this context, three hypotheses were experimentally tested: 1) Moreover, in these months of the year, the four beaches do not reduction in the stranded wrack biomass by experimental removal receive tourists and visitors. Consequently, the beaches are not would lower density and diversity as well as change the structure of exposed to other disturbances such as human trampling, vehicle 224 J.V. Vieira et al. / Marine Environmental Research 119 (2016) 222e237

Fig. 1. Geographic location of the study site, highlighting the four sandy beaches sampled: (i) Assenodi and (ii) Cem located in South Brazil (Parana State). (iii) Levante and (iv) Cortadura situated on the south-western Spain (Atlantic coast of Cadiz). Satellite image from 2003 extracted from Google Earth. Photo credits (i-iv): Vieira, J.V. and Ruiz-Delgado, M.C.

Table 1 Summary of the physical characteristics of each surveyed beach together with i.e. four sampling days before ( 16, 6, 3, and 1) and four seasonal climate data (temperature and precipitation, mean for 1961e90) for sampling days after (þ1, þ3, þ6, and þ16) the wrack removal southern Brazil (May to July) and south-western Spain (October to November). (Fig. 2). At each impacted plot, all macroscopic wrack debris on the Beach Southern Brazil South-western Spain beach surface (supratidal and intertidal zone) was removed using garden rakes. Wrack debris present at the 2 m intervals bordering Assenodi Cem Levante Cortadura plots was also eliminated. Beach length (m) 2100 1000 1000 2480 Stranded wrack was removed daily for 10 consecutive days Intertidal width (m) 110 53 105 85 (between 1 and þ1 sampling times), to ensure a significant Slope ( ) 1.26 2.40 2.03 2.19 Mean sand grain size (mm) 0.27 0.23 0.18 0.23 reduction of wrack biomass available in impacted plots. By contrast, Wave period (s) 7.00 4.90 5.00 7.00 the control plots were not manipulated and represent the undis- Wave height (m) 1.00 0.25 0.75 0.65 turbed condition of wrack-associated fauna. In this way, we test ’ Dean s parameter 5.53 1.76 6.20 3.37 whether differences in supralittoral arthropods between plots Mean spring tide range (m) 1.70 1.70 2.00 2.00 Relative tide range (m) 1.70 6.80 5.71 3.08 (three impacted vs. three control) are detected only after the wrack Seasonal mean temperature (C) 18.00 16.83 removal. Total seasonal precipitation (mm) 271.90 276.40 On each sampling date, wrack coverage was measured from photographs taken within six 1 1 m quadrats placed randomly in each plot (three control and three impacted). On these photo- graphed quadrats, supralittoral arthropods associated with wrack traffic, and other recreational pursuits. At each beach the same and those that had burrowed underneath the wrack patches were sampling design (M-BACI) was adopted, i.e. multiple control and collected. Biological samples were randomly collected from wrack impacted plots were compared on multiple sampling dates before patches by pushing a core (15 cm in diameter and 20 cm in depth) and after wrack removal. vertically through the wrack mat into the sediment to collect wrack For this design, an experimental 100-m transect (alongshore) debris as well as wrack-associated macrofauna (0.02 m2 surface was designated and divided into six plots, three control (C), and area). At the free end of the corer, a plastic bag was used to prevent three impacted (I), each 15 m long (alongshore), extending from the mobile fauna (mainly insects and amphipods) from escaping. All base of the dune to the highest mark of the strandline biological samples (six samples per plot - three control and three (width ¼ 40 m). The plots were 2 m apart, using an interspersed impacted - making a total of 36 samples at each beach per sampling design (Hurlbert, 1984) in which the first plot along the shore on day) were placed in 70% ethanol. In the laboratory, samples were each beach was randomly assigned to a treatment (impacted or washed to separate the fauna and the wrack. All organisms retained control), and the remaining plots were alternately assigned one of in a sieve of 0.5 mm mesh were sorted, identified, and counted the two treatments. During the field experiment, all samplings under a stereomicroscope. Wrack debris from each sample was were conducted at low tide. dried (60 C) to a constant weight (g dw). Biological responses among the three control and three Wrack coverage was estimated as the surface area covered by impacted plots were monitored in pre-established temporal scales, wrack deposits on the photographed quadrats (1 1 m) using J.V. Vieira et al. / Marine Environmental Research 119 (2016) 222e237 225

Fig. 2. Multiple beforeeafter controleimpact (M-BACI) design used in field experiment and in data analysis. T ¼ Time (days); C (control) and I (impacted) plots; n ¼ numbers of samples in each plot and date of sampling.

ImageJ program (Abramoff et al., 2004). For an estimate of the disturbance (i.e. wrack removal). Taking into account differences in wrack biomass (g dw m 2) in each plot, the percentage of coverage wrack subsidies (quantity and composition) and morphodynamic was applied to values of wrack weight (g dw) per corer of each state among the evaluated beaches (detailed in the Study area), all sampling date. For each biological sample, the assemblage de- analyses were separately performed for each beach (Assenodi, Cem, scriptors were calculated: total density (expressed as total number Levante and Cortadura). of ind. m 2 of the surface covered by wrack deposits), species Permutational multivariate analysis of variance (Anderson, richness (number of taxa) and diversity (ShannoneWiener Index). 2001) was applied to test differences among supralittoral For the physical characterization, three equidistant and across- assemblages. PERMANOVA tests were performed fol- shore transects were placed in a 100 m long-shore area at each lowed by a posteriori pair-wise comparisons whenever significant beach. Each transect comprised 10 equidistant points, from the interactions (p < 0.05) were detected between the terms of interest. base of the dune to the swash zone. At each sampling point, sedi- All taxa that were found in less than 10% of the samples (out of a ment samples were collected with a plastic tube (3.5-cm diameter) total of 288 samples) were excluded from the analysis, to reduce buried at a depth of 20 cm. Moreover, the following environmental the number of zeros in the data matrix. Non-metric multidimen- variables were recorded: beach width, measured as the distance sional scaling (n-MDS), derived from similarity matrices using the between the base of the dune and the lower swash level; slope Bray-Curtis index, was used to visualize the variation trends of the estimated using Emery’s profiling technique (Emery, 1961); mean supralittoral arthropod assemblages (Clarke and Warwick, 1994). grain size was determined by sieving dry sediment through a Average values of six replicates of each plot (three controls and graded series of sieves (5, 2, 1, 0.5, 0.25, 0.125, and 0.063 mm (Folk three impacted) were considered for the n-MDS. Furthermore, the and Ward, 1957); wave height was visually recorded and wave similarity percentage routine (SIMPER) was used to identify the period was estimated with a stopwatch. These measurements were contribution of abundant taxa to the total dissimilarity in the used to calculate the following indices that describe beach mor- assemblage structure among the terms of interest (i.e. (Tr Pe) or phodynamic state: Dean’s parameter (U)(Short and Wright, 1983) Tr Ti(Pe)). and relative tide range (RTR) (Masselink and Short, 1993). Four-factor mixed ANOVAs (analyses of variance) were used to test the hypothesis related to differences in the wrack biomass and assemblage descriptors (total density, species richness, and di- 2.3. Data analysis versity). Moreover, the same model was applied to examine dif- ferences in the density of numerically dominant taxa (representing The model formulated for multivariate and univariate analyses over 70% of the assemblage ¼ frequency) between impacted and considered four factors: Treatment (Tr; two levels: control and control plots after wrack removal. Therefore, significant in- impacted, fixed and orthogonal), Period (Pe; two levels: before and teractions that included the mentioned above terms (i.e. (Tr Pe) after, fixed and orthogonal), Plot (Pl; three levels: random and or Tr Ti(Pe)) were analysed using Student-Newman-Keul (SNK) nested within Treatment), Time (Ti; four levels, fixed and nested tests (Underwood, 1997). The homogeneity of variances was tested within Period). According to the principles of M-BACI model with Cochran’s test, and the data were transformed when neces- (Downes et al., 2004), the effect that eliminating stranded wrack sary. Untransformed data were used when transformations failed to biomass (impact) exerts on supralittoral assemblages is identifiable stabilize heterogeneous variances, but in those cases, the chance of as the interaction between treatments and periods (Tr Pe) or as Type I error was reduced, by using the significance level (a)of differences among treatments at any particular time after the p ¼ 0.01 (Underwood, 1997). experimental wrack removal (Tr Ti(Pe)). Therefore, these in- Multivariate analyses were performed using PRIMER v.6.1 with teractions are the only terms of interest to quantify the effect of this 226 J.V. Vieira et al. / Marine Environmental Research 119 (2016) 222e237 the PERMANOVA þ add-on (PRIMER-E Ltd.) (Anderson et al., 2008). Brazilian beaches. From these, 10 taxa were found exclusively at Univariate analyses and graphs were carried out using R pro- Assenodi beach, 8 taxa were found only at Cem beach, and 19 taxa gramming language (R Core Team, 2012) combined with GAD were found at both beaches (Table A1). At Assenodi beach, the most (Sandrini-Neto and Camargo, 2012) and Sciplot (Morales, 2012) abundant species were the staphylinid Bledius bonariensis (48.50%), packages. the talitrid amphipod Platorchestia monodi (23.10%) and beetle families Carabidae (9.52%) and Nitidulidae (2.54%). On the other 3. Results hand, P. monodi (36.64%) followed by nitidulids (24.22%), clerids (8.60%) and the tenebrionid Phaleria testacea (3.79%) were the most 3.1. Spatio-temporal changes in wrack biomass representative taxa at Cem beach. On Spanish beaches, 27,817 (2,840 at Levante and 24,977 at In the pre-disturbance samplings, the wrack biomass in control Cortadura) individuals belonging to 35 taxa were sampled during plots was similar to those of the impacted plots (SNK test p > 0.05; this experiment. The two beaches shared 21 wrack-associated taxa, Fig. 3). This pattern was observed in all evaluated beaches. with 12 found exclusively at Levante beach and 2 found only at On Brazilian beaches, wrack biomass varied between control Cortadura beach (Table A2). Wrack deposits of Levante and Corta- and impacted plots only after experimental manipulation (signifi- dura beach were dominated by the talitrid amphipod Talitrus sal- cantly Tr Ti(Pe) interaction; Table 2). At Assenodi beach, the mean tator (63% and 95%, respectively). Other numerically dominant taxa of wrack biomass was significantly lower in the impacted than in were the dipteran Brachycera (10%), the tenebrionid Phaleria the control plots at all sampling times after wrack removal bimaculata (6%), and Brachycera larvae (6%) at Levante beach, and (þ1, þ3, þ6 and þ16 days, SNK test p < 0.001; Fig. 3). At Cem beach the staphylinid Aleocharinae sp. 3 (2%) at Cortadura beach. this difference was observed only on day þ1 (SNK test p < 0.001; Fig. 3). 3.3. Spatio-temporal changes in structure of the supralittoral Statistical analyses also detected differences in the wrack arthropod assemblages biomass (significant Tr Ti(Pe) interaction; Table 2) in the pos- disturbance samplings on both Spanish beaches. At Levante At Assenodi beach, an effect of wrack removal was detected on beach, wrack biomass proved significantly lower in the impacted the supralittoral arthropod assemblage structure (i.e. significantly than in the control plots on day þ1 and þ3 (SNK test p < 0.001; Tr Pe interaction PERMANOVA test; Table 3). Before wrack-debris Fig. 3). From day þ6 no significant difference was found in this removal, the assemblages in control plots were similar to those of variable between impacted and control plots (SNK test p > 0.05; the impacted plots (pairwise comparisons p > 0.05; Table 3 and Fig. 3). At Cortadura beach, the decrease in wrack biomass in the Fig. 4). In contrast, the assemblage structure differed between impacted plots was recorded only on day þ1 (SNK test p < 0.001; impacted and control plots after impact (pairwise comparisons Fig. 3). In the other sampling times, there was no difference in p < 0.001; Table 3 and Fig. 4). On this beach, SIMPER analysis wrack biomass between control and impacted plots (SNK test indicated that wrack removal caused 79.55% dissimilarity between p > 0.05; Fig. 3). control and impacted plots. The most important taxa contributing to this dissimilarity were B. bonariensis (34%), P. monodi (32%), and 3.2. Composition and structure of supralittoral arthropod carabids (14%) (Table A3). assemblages PERMANOVA test also detected an effect of wrack removal in assemblage structure at Cem beach, although this response was During the field experiment, 9,924 individuals (5,680 at Asse- observed only on day þ1 (i.e. Tr Ti(Pe) interaction; Table 3). The nodi and 4,244 at Cem) belonging to 37 taxa were sampled in dissimilarity between control and impacted plots on day þ1was

Fig. 3. Temporal changes in wrack biomass (g dw/m2) in the control (black column) and impacted (white column) plots, from before (16, 6, 3, 1 days) to after (þ1, þ3, þ6, þ16 days) the wrack removal for the two Brazilian beaches, Assenodi (i) and Cem (ii) and two Spanish beaches, Levante (iii) and Cortadura (iv). Data are mean of wrack biomass ± S.E. (n ¼ 18). Significant differences between plots (control and impacted) at any time was also represented (SNK tests, ***p < 0.001). J.V. Vieira et al. / Marine Environmental Research 119 (2016) 222e237 227

Table 2 Summary of the ANOVAs for the M-BACI model for wrack biomass from the two Brazilian beaches: Assenodi (i) and Cem (ii) and two Spanish beaches, Levante (iii) and Cortadura (iv). Significant terms of interest (a ¼ 0.05) are highlighted in bold.

(i) Assenodi (ii) Cem

df MS FP MS FP

Treatment ¼ Tr 1 106.90 11.78 0.026 13.60 43.29 0.003 Period ¼ Pe 1 361.50 77.52 <0.001 710.41 790.66 <0.001 Plot ¼ Pl(Tr) 4 9.60 23.13 <0.001 0.31 0.678 0.608 Time ¼ Ti(Pe) 6 2.40 6.68 <0.001 43.93 48.81 <0.001 Tr Pe 1 69.65 14.95 0.018 10.58 11.78 0.027 Tr Ti(Pe) 6 1.78 4.96 0.002 4.56 5.07 0.002 Pl(Tr) Pe 4 4.66 13.00 <0.001 0.90 1.00 0.428 Pl(Tr) Ti(Pe) 24 0.36 0.92 0.580 0.90 1.94 0.007 Residuals 240 0.39 0.46 Transformation (4th root) (4th root)

(iii) Levante (iv) Cortadura

df MS FP MS FP

Treatment ¼ Tr 1 16.59 26.66 0.007 11.98 22.51 0.010 Period ¼ Pe 1 0.09 0.47 0.530 122.23 259.80 <0.001 Plot ¼ Pl(Tr) 4 0.62 3.91 0.004 0.53 0.82 0.513 Time ¼ Ti(Pe) 6 6.43 19.38 <0.001 15.76 21.57 <0.001 Tr Pe 1 11.05 57.07 0.002 1.31 2.80 0.170 Tr Ti(Pe) 6 3.44 10.38 <0.001 2.95 4.04 0.006 Pl(Tr) Pe 4 0.19 0.58 0.677 0.47 0.64 0.636 Pl(Tr) Ti(Pe) 24 0.33 2.08 0.003 0.73 1.13 0.315 Residuals 240 0.16 0.65 Transformation (4th root) log(xþ1)

97.74% (Fig. 4). According to SIMPER analysis, P. monodi (47%), experiment, no significant (p > 0.05) differences were detected in nitidulids (16%), clerids (14%) and P. testacea (11%) were responsible impacted plots respect to control plots (Table 3; Fig. 4). by these differences (Table A3). From day þ3 until the end of the On the other hand, supralittoral arthropod assemblages were

Table 3 Summary of the PERMANOVAs for the M-BACI model for supralittoral artropods assemblages from the two Brazilian beaches, Assenodi (i) and Cem (ii) and two Spanish beaches, Levante (iii) and Cortadura (iv). For pairwise comparisons: “s” indicates p < 0.05 and “¼” indicates p > 0.05. Significant terms of interest (a ¼ 0.05) are highlighted in bold.

(i) Assenodi (ii) Cem

df MS Pseudo-F P(MC) MS Pseudo-F P(MC)

Treatment ¼ Tr 1 27,656 63.60 0.007 16,632 5.42 <0.001 Period ¼ Pe 1 52,578 12.86 <0.001 7603 26.00 <0.001 Plot ¼ Pl(Tr) 4 7682.2 4.21 <0.001 3067.8 1.32 0.114 Time ¼ Ti(Pe) 6 24,992 10.80 <0.001 14,827 4.84 <0.001 Tr Pe 1 20,172 4.94 0.002 10,613 3.63 0.006 Tr Ti(Pe) 6 2216.8 0.96 0.553 6543.1 2.14 <0.001 Pl(Tr) Pe 4 4087.3 2.24 <0.001 2927.1 1.26 0.153 Pl(Tr) Ti(Pe) 24 2313.6 1.27 0.012 3061.4 1.32 0.004 Residuals 240 1823.2 2317.8

Pairwise test Condition Condition

Tr Ti(Pe) Tr Pe Before: I ¼ C T 16: C ¼ I T ¡3: C ¼ I T 6: C ¼ I T ¡1: C ¼ I After: C s I T þ1: C s I T þ6: C ¼ I T þ3: C ¼ I T þ16: C ¼ I

(iii) Levante (iv) Cortadura

df MS Pseudo-F P(MC) MS Pseudo-F P(MC)

Treatment ¼ Tr 1 1763.2 1.06 0.417 2946.6 1.14 0.360 Period ¼ Pe 1 16,727 4.77 0.005 67,825 47.33 <0.001 Plot ¼ Pl(Tr) 4 1668.7 0.79 0.733 2579.1 1.52 0.079 Time ¼ Ti(Pe) 6 19,637 7.34 <0.001 22,679 7.28 <0.001 Tr Pe 1 3124.3 0.89 0.503 1441.3 1.01 0.429 Tr Ti(Pe) 6 3950.2 1.48 0.072 2601.1 0.83 0.698 Pl(Tr) Pe 4 3503.3 1.65 0.034 1433.1 0.85 0.646 Pl(Tr) Ti(Pe) 24 2673.6 1.26 0.030 3117.3 1.84 <0.001 Residuals 240 2117.7 1695.3 228 J.V. Vieira et al. / Marine Environmental Research 119 (2016) 222e237

Fig. 4. Non-metric multidimensional scaling (nMDS) of supralittoral arthropod assemblages comparing the control (circles) and impacted (triangles) plots from before (1, 3, 6, and 16 days, white symbols) to after (þ1, þ3, þ6, and þ16 days, black symbols) the wrack removal for the two Brazilian beaches, Assenodi (i) and Cem (ii) and two Spanish beaches, Levante (iii) and Cortadura (iv). Dotted circles represent the clusters detected in the PERMANOVAs tests (c.f. Table 3).

Table 4 Summary of the ANOVAs for the M-BACI model for assemblage descriptors (total density (a), species richness (b) and diversity index (c)) from the two Brazilian beaches: Assenodi (i) and Cem (ii) and two Spanish beaches, Levante (iii) and Cortadura (iv). Significant terms of interest (a ¼ 0.05) are highlighted in bold.

(i) Assenodi (ii) Cem (iii) Levante (iv) Cortadura

df MS FP MS FP MS FP MS FP

(a) Total density Treatment ¼ Tr 1 2.31 0.75 0.437 61.05 11.13 0.029 5.70 13.96 0.002 10.08 2.57 0.184 Period ¼ Pe 1 137.52 39.61 0.003 22.00 7.55 0.051 0.52 0.18 0.693 421.54 54.32 0.002 Plot ¼ Pl(Tr) 4 3.09 3.38 0.010 5.49 2.48 0.045 0.41 0.16 0.960 3.93 1.19 0.315 Time ¼ Ti(Pe) 6 18.24 10.59 <0.001 14.22 3.92 0.007 15.04 5.23 0.001 101.44 12.87 <0.001 Tr Pe 1 0.20 0.06 0.822 23.79 8.17 0.046 0.01 0.00 0.961 7.25 0.93 0.388 Tr Ti(Pe) 6 1.87 1.09 0.398 10.35 2.85 0.031 3.38 1.18 0.352 6.59 0.84 0.554 Pl(Tr) Pe 4 3.47 2.02 0.124 2.91 0.80 0.535 2.86 1.00 0.428 7.76 0.98 0.435 Pl(Tr) Ti(Pe) 24 1.72 1.88 0.009 3.63 1.64 0.034 2.87 1.10 0.343 7.88 2.39 <0.001 Residuals 240 0.92 2.22 2.61 3.30 Transformation (4th root) (4th root) log(xþ1) (4th root)

(b) Species richness Treatment ¼ Tr 1 45.92 26.14 0.007 55.13 5.85 0.073 0.06 2.32 0.202 0.46 4.80 0.094 Period ¼ Pe 1 306.28 70.68 0.001 72.00 22.91 0.009 0.67 14.83 0.018 0.48 1.36 0.308 Plot ¼ Pl(Tr) 4 1.76 0.84 0.500 9.43 3.21 0.014 0.03 0.22 0.924 0.10 0.52 0.723 Time ¼ Ti(Pe) 6 6.22 2.37 0.062 23.73 4.74 0.003 0.57 5.20 0.001 5.48 25.38 <0.001 Tr Pe 1 19.53 4.51 0.101 13.35 4.25 0.108 0.03 0.72 0.443 0.68 1.93 0.238 Tr Ti(Pe) 6 4.74 1.80 0.141 9.98 2.00 0.106 0.12 1.09 0.398 0.43 2.01 0.104 Pl(Tr) Pe 4 4.33 1.65 0.195 3.14 0.63 0.647 0.05 0.41 0.780 0.35 1.64 0.198 Pl(Tr) Ti(Pe) 24 2.63 1.26 0.193 5.01 1.71 0.024 0.11 0.93 0.564 0.22 1.18 0.265 Residuals 240 2.09 2.94 0.12 0.18 Transformation none none (2th root) (2th root)

(c) ShannoneWiener Index Treatment ¼ Tr 1 4.45 38.37 0.003 8.36 6.92 0.058 0.02 0.25 0.646 0.10 0.18 0.695 Period ¼ Pe 1 26.24 38.51 0.003 13.53 19.10 0.012 5.25 34.07 0.004 5.28 23.00 0.009 Plot ¼ Pl(Tr) 4 0.12 0.40 0.807 1.21 3.04 0.018 0.09 0.22 0.926 0.56 2.68 0.032 Time ¼ Ti(Pe) 6 1.07 2.56 0.046 3.28 5.69 <0.001 3.68 8.70 <0.001 3.04 9.67 <0.001 Tr Pe 1 3.66 5.37 0.081 0.75 1.06 0.361 0.01 0.03 0.868 0.79 3.44 0.137 Tr Ti(Pe) 6 0.78 1.85 0.132 1.66 2.88 0.030 1.05 2.50 0.052 0.21 0.67 0.673 Pl(Tr) Pe 4 0.68 1.63 0.200 0.71 1.23 0.326 0.15 0.36 0.832 0.23 0.73 0.581 Pl(Tr) Ti(Pe) 24 0.42 1.46 0.083 0.58 1.46 0.083 0.42 1.00 0.464 0.31 1.52 0.063 Residuals 240 0.29 0.40 0.42 0.21 Transformation none none none none J.V. Vieira et al. / Marine Environmental Research 119 (2016) 222e237 229 structurally similar between impacted and control plots on Spanish Regarding the individual average density of the dominant taxa beaches over the time after the experimental wrack removal (i.e. no on Brazilian beaches (Table A1), ANOVAs revealed that the density significant interaction Tr Pe or Tr Ti(Pe) at Levante and Cor- of amphipods (P. monodi) varied significantly between impacted tadura beach; Table 3; Fig. 4). and control plots after wrack removal (Tr Pe interaction; Table 5a) at Assenodi beach. Thus, the density of P. monodi was significantly lower in the impacted plots compared with control plots during all 3.4. Spatio-temporal changes in assemblage descriptors and in the pos-disturbance samplings (þ1, þ3, þ6, and þ16 days; SNK test density of the more abundant taxa p < 0.01; Fig. 5a). Moreover, after experimental manipulation, nitidulid individuals disappeared from the impacted plots (Fig. 5b). fi On Brazilian beaches, the impacted plots registered signi cantly On the other hand, the density of carabids and B. bonariensis lower density and diversity values respect to control plots one day showed higher values in impacted than in control plots (Fig. 5c, d). after the wrack removal at Cem beach (Tr Ti(Pe) interaction; However, significant interactions between experimental plots and < Table 4a and c; SNK test p 0.001, Fig. A1d, f). This pattern was also periods (Tr Pe) or any time sampling (Tr Ti(Pe)) were not observed at Assenodi beach, although this effect was not statisti- detected (Table 5c, d). fi cally signi cant (Table 4a, c; Fig. A1a, c). Overall, on Spanish bea- The wrack removal also had an effect on the population density fi > ches no signi cant differences (p 0.05) were found in none of the of three main taxa at Cem beach (P. monodi, nitidulids, and univariate descriptors between control and impacted plots P. testacea; Table 5e, f, h, respectively). Immediately after the impact e (Table 4a c; Fig A2a-f]. (day þ1), P. monodi and Nitidulidae disappeared from the impacted plots, whereas in the control plots maintained high density values (Fig. 5e, f, respectively). The mean density P. testacea density was Table 5 significantly lower in the impacted plots than in controls on Summary of the ANOVAs for the M-BACI model for density of numerically dominant days þ1orþ6 (SNK test p < 0.05; Fig. 5h). Moreover, a lower taxa from the two Brazilian beaches. a-d: taxa recorded at Assenodi Beach (i). e-h: þ taxa recorded at Cem Beach (ii). Significant terms of interest (a ¼ 0.05) are high- density of clerids was detected in the impacted plots on day 1 lighted in bold. (Fig. 5g), although this pattern was not statistically significant (i.e. no significant Tr Pe or Tr Ti(Pe) interaction; Table 5g). (i) Assenodi (ii) Cem The most abundant taxon at Levante (Spanish beach), T. saltator df MS FP MS FP (Table A2), showed a clear difference (impacted vs. control plots) (a) P. monodi (e) P. monodi over different times after wrack removal (i.e. Tr Ti(Pe) interac- Treatment ¼ Tr 1 164.28 11.39 0.028 1355.52 9.69 0.036 tion; Table 6a). The density of T. saltator was significantly lower in Period ¼ Pe 1 516.87 85.20 <0.001 2211.22 210.66 <0.001 the impacted than in control plots on day þ1 (SNK test p < 0.001; Plot ¼ Pl(Tr) 4 14.42 4.02 0.004 139.87 1.72 0.146 þ þ þ Time ¼ Ti(Pe) 6 1.60 0.42 0.862 974.25 6.24 <0.001 Fig. 6a), while on day 3, 6 and 16 no difference were detected Tr Pe 1 178.27 29.39 0.006 1636.65 155.92 <0.001 (SNK test p > 0.05; Fig. 6a). At Cortadura beach, T. saltator showed a Tr Ti(Pe) 6 7.39 1.92 0.120 268.35 1.72 0.160 different response from that noted at Levante beach. On this beach, Pl(Tr) Pe 4 6.07 1.67 0.213 10.50 0.07 0.991 no significant pattern of declining density was detected after wrack Pl(Tr) Ti(Pe) 24 3.86 1.07 0.375 156.25 1.92 0.007 Residuals 240 3.59 81.18 removal (Table 6d; Fig. 6d). Meanwhile, for other taxa such as Transformation log(xþ1) (2th root) Brachycera and P. bimaculata (Levante beach) and Aleocharinae sp. 3 (Cortadura beach), no significant differences were detected be- (b) Nitidulidae (f) Nitidulidae Treatment ¼ Tr 1 37.60 1.92 0.238 18.60 2.37 0.199 tween the impacted and control treatments after the wrack Period ¼ Pe 1 43.42 5.33 0.082 60.11 42.60 0.003 removal (i.e. no significant Tr Ti(Pe) or Tr Pe interaction; Plot ¼ Pl(Tr) 4 19.59 6.88 <0.001 7.84 2.80 0.028 Table 6b, c, e; Fig. 6b, c, e, respectively). Time ¼ Ti(Pe) 6 8.71 2.78 0.034 12.51 2.64 0.042 Tr Pe 1 34.07 4.48 0.110 12.48 8.84 0.041 Tr Ti(Pe) 6 6.68 2.13 0.086 7.34 1.60 0.206 4. Discussion Pl(Tr) Pe 4 8.15 2.60 0.061 1.41 0.30 0.877 Pl(Tr) Ti(Pe) 24 3.13 1.10 0.346 4.74 1.70 0.030 4.1. Effects of wrack removal on supralittoral arthropods Residuals 240 2.85 2.83 þ Transformation log(x 1) (4th root) Our experimental design (M-BACI), applied both multivariate (c) Carabidae (g) Cleridae and univariate analyses, allowed us to establish a relationship be- Treatment ¼ Tr 1 21.89 5.58 0.078 44.44 5.53 0.078 tween the disturbance (i.e., wrack removal) and supralittoral ar- Period ¼ Pe 1 23.20 4.07 0.114 92.13 28.89 0.006 thropods. Although the removal of wrack debris along the Plot ¼ Pl(Tr) 4 3.93 1.00 0.412 8.03 1.88 0.114 Time ¼ Ti(Pe) 6 89.75 14.10 <0.001 32.35 3.95 0.007 strandline is assumed to be a source of disturbance for wrack- Tr Pe 1 10.47 1.84 0.247 17.16 5.38 0.081 associated fauna, due to the reduction in food sources and micro- Tr Ti(Pe) 6 2.70 0.42 0.856 4.50 0.55 0.765 habitat refuges (e.g. Dugan et al., 2003; Gilburn, 2012), no experi- Pl(Tr) Pe 4 5.70 0.90 0.482 3.19 0.39 0.814 mental field studies have evaluated the direct effects of this Pl(Tr) Ti(Pe) 24 6.40 1.61 0.040 8.18 1.92 0.008 Residuals 240 3.95 4.26 disturbance on supralittoral arthropods. Transformation log(xþ1) log(xþ1) The results of the M-BACI experiment indicated that the removal of wrack change the structure of the supralittoral arthro- (d) B. bonariensis (h) P. testacea fi Treatment ¼ Tr 1 48.35 1.66 0.267 5.45 3.07 0.155 pods assemblages (according with our rst hypothesis), although Period ¼ Pe 1 108.08 20.07 0.011 0.58 0.11 0.754 this response pattern varies between beaches. At Assenodi (Bra- Plot ¼ Pl(Tr) 4 29.10 9.21 <0.001 1.78 0.47 0.761 zilian beach) the assemblage structure differed between impacted ¼ < Time Ti(Pe) 6 133.22 26.58 0.001 6.02 2.08 0.093 and control plots in all post-disturbance samplings, while at Cem Tr Pe 1 30.10 5.59 0.077 18.11 3.51 0.134 þ Tr Ti(Pe) 6 1.46 0.29 0.935 9.28 3.21 0.019 beach this effect was observed only on 1 day. Changes in the Pl(Tr) Pe 4 5.39 1.07 0.391 5.15 1.78 0.165 structure of assemblages were greatly influenced by the response of Pl(Tr) Ti(Pe) 24 5.01 1.58 0.045 2.89 0.76 0.788 some taxa. After the manipulation of wrack debris, drastic de- Residuals 240 3.16 3.81 creases in the densities of the amphipod Platorchestia monodi and Transformation log(xþ1) log(xþ1) sap beetle Nitidulidae were detected in the impacted plots on both 230 J.V. Vieira et al. / Marine Environmental Research 119 (2016) 222e237

Fig. 5. Temporal changes in density of numerically dominant taxa in the control (black column) and impacted (white column) plots, from before (16, 6, 3, 1 days) to after (þ1, þ3, þ6, þ16 days) the wrack removal for the two Brazilian beaches. aed: taxa recorded at Assenodi Beach (i). eeh: taxa recorded at Cem Beach (ii). Data are mean of taxa density ± S.E. (n ¼ 18). Significant differences between plots (control and impacted) at any time was also represented (SNK tests, *p < 0.05; ***p < 0.001).

Brazilian beaches, in agreement with our second hypothesis. impacted plots can be directly related to higher prey availability (in Similar responses were observed for other taxa (i.e. beetles Cleridae this case B. bonariensis). and the Tenebrionidae, Phaleria testacea), particularly at Cem beach. On Spanish beaches the disturbance (i.e. wrack removal) was Staphylinids (as Bledius bonariensis) and carabids were abundant not severe enough to change the overall structure of assemblages. taxa at Assenodi beach showing response patterns opposite to However, at the population level, wrack removal caused a reduction those described for other taxa. Density of these two coleopterans in the amphipod Talitrus saltator in impacted plots from Levante was higher in impacted than did control plots after the disturbance. beach (second hypothesis). The different response of this species at Larvae and adults of B. bonariensis build their tracks from the the two evaluated beaches might be related to the beach mor- supralittoral to the upper mesolittoral area (Gandara-Martins et al., phodynamic as well as composition of wrack debris as a recent 2010; Vianna and Borzone, 2015) and, thus are found in the wrack study has demonstrated (Ruiz-Delgado et al., 2016). This study has debris, as well as in bare-sand areas. These coleopterans feed on showed that the abundance of supralittoral air-breathing crusta- microalgae in the sediment (Herman, 1986) and are usually asso- ceans in the intermediate (Cortadura beach) was higher that ciated with the debris to only prevent dehydration (Ruiz-Delgado dissipative condition (Levante beach), as observed in our experi- et al., 2014). Such characteristics might explain a less dependence ment. Moreover, the high availability of stranded macroalgae of wrack debris than others species (such as P. monodi), and thus (mainly brown algae) at Cortadura beach could concentrate high density values in impacted plots although these plots regis- T. saltator individuals on wrack deposits, searching for food, since tered low values of wrack biomass after wrack removal. The brown algae are preferred by this species (Adin and Riera, 2003; abundant presence of these staphylinids might attract carnivorous Olabarria et al., 2009). In contrast, beaches subsidized by sea- beetles (McLachlan and Brown, 2006), such as carabids, which are grasses, such as Levante beach, could concentrate low density of present in the control plots and at other neighbouring sites. Carabid T. saltator, since this species does not feed directly on seagrass beetles are regarded as the main predators of Bledius (Colombini leaves (Adin and Riera, 2003; Colombini et al., 2009). These factors and Chelazzi, 2003; Herman, 1986), and their increase density in might explain the higher density of T. saltator associated with wrack J.V. Vieira et al. / Marine Environmental Research 119 (2016) 222e237 231

Table 6 Summary of the ANOVAs for the M-BACI model for density of numerically dominant taxa from the two Spanish beaches. aec: taxa recorded at Levante Beach (iii). e-f: taxa recorded at Cortadura Beach (iv). Significant terms of interest (a ¼ 0.05) are highlighted in bold.

(iii) Levante (iv) Cortadura

df MS FP MS FP

(a) T. saltator (d) T. saltator Treatment ¼ Tr 1 53.46 0.84 0.411 1125000 0.02 0.885 Period ¼ Pe 1 683.62 2.86 0.166 2287133889 67.20 0.001 Plot ¼ Pl(Tr) 4 63.58 0.90 0.464 4795434 2.39 0.052 Time ¼ Ti(Pe) 6 835.33 12.16 <0.001 511458843 9.19 <0.001 Tr Pe 1 65.58 0.27 0.628 23006806 0.68 0.457 Tr Ti(Pe) 6 219.39 3.19 0.019 56868889 1.02 0.435 Pl(Tr) Pe 4 238.94 3.48 0.022 34033056 0.61 0.658 Pl(Tr) Ti(Pe) 24 68.72 0.97 0.501 55625090 2.80 <0.001 Residuals 240 70.53 19896278 Transformation (2th root) none

(b) Brachycera (e) Aleocharinae sp. 3 Treatment ¼ Tr 1 2.83 17.13 0.014 0.70 0.16 0.711 Period ¼ Pe 1 3.44 2.70 0.176 4.52 3.69 0.127 Plot ¼ Pl(Tr) 4 0.17 0.09 0.985 4.35 2.23 0.066 Time ¼ Ti(Pe) 6 31.57 16.69 <0.001 53.00 19.25 <0.001 Tr Pe 1 3.16 2.48 0.190 5.79 4.72 0.095 Tr Ti(Pe) 6 4.16 2.20 0.078 2.58 0.94 0.488 Pl(Tr) Pe 4 1.27 0.67 0.617 1.22 0.45 0.775 Pl(Tr) Ti(Pe) 24 1.89 1.05 0.405 2.75 1.41 0.101 Residuals 240 1.80 1.95 Transformation (4th root) (4th root)

(c) P. bimaculata Treatment ¼ Tr 1 0.01 0.01 0.915 Period ¼ Pe 1 21.89 12.71 0.023 Plot ¼ Pl(Tr) 4 0.69 0.61 0.657 Time ¼ Ti(Pe) 6 19.74 8.24 <0.001 Tr Pe 1 0.51 0.30 0.615 Tr Ti(Pe) 6 1.85 0.77 0.600 Pl(Tr) Pe 4 1.72 0.72 0.600 Pl(Tr) Ti(Pe) 24 2.40 2.12 0.002 Residuals 240 1.13 Transformation (4th root)

patches at Cortadura than Levante beach and, consequently the wrack deposits, and therefore the age of the wrack is limited by the smaller response of this amphipod on the former than the latter intervals between beach-cleaning activities. In this regard, the beach. wrack deposits available on day þ16 (two weeks after the distur- Overall, the recovery of the fauna in the impacted plots was bance) presumably had different microclimatic conditions (espe- rapid, although the recovery time varied between beaches in each cially low moisture) compared to that deposited on days þ3 study region, according with our third hypothesis. Our results and þ6. Consequently, populations of Tenebrionidae species that suggest that recovery time can be closely related to the wrack- prefer aged and dry wrack deposits (Jaramillo et al., 2006; Ruiz- debris characteristics (e.g. the amount, composition, and degrada- Delgado et al., 2014, 2015) might show a slower recovery than tion), coupled with the behavioural and biological strategies the those of talitrids, which are typical species associated with new (mobility, burrowing, feeding preferences) of each species. For and fresh wrack debris (Colombini et al., 2000; Lastra et al., 2008; instance, the recovery of supralittoral arthropods in the impacted Ruiz-Delgado et al., 2015). Future studies on the behavioural stra- plots at Cem beach was faster than at the Assenodi beach. Differ- tegies of the beetles Nitidulidae and Cleridae, as well as their ences in the wrack biomass between the two beaches might explain relationship with habitat features (e.g. microclimatic conditions) these results, since the recovery of species coincides with the re- are needed to understand the response patterns following a accumulation of new debris on the beach. The availability of disturbance in the strandlines. wrack debris has been reported as the major factor structuring A notable result was the different recovery pattern of the two supralittoral arthropod assemblages and populations (Dugan et al., species of talitrid amphipods (i.e. T. saltator on Spanish beaches and 2003; Gonçalves and Marques, 2011; Ince et al., 2007). Thus, a P. monodi on Brazilian beaches) to the same disturbance (i.e. wrack significant and continuous input of new debris on beaches, as removal). T. saltator recovered more quickly (during the first 3 days) recorded at Cem beach, might promote a rapid recovery of the than P. monodi (within a 16-day period). Some studies on T. saltator populations after wrack removal. have reported its behavioural strategies (i.e. orientation, zonation, Besides the availability of wrack, the recovery of supralittoral burrowing, and mobility) to cope with beach disturbances (Bessa arthropods species should also be related to age of the debris. In et al., 2013; Fanini et al., 2005). In our case, the great ability of fact, the recovery of the tenebrionidae P. testacea population at Cem this species to move long distances, due to its effective jumping, beach was slower (i.e. day þ16) than that of other taxa (day þ3 for crawling and burrowing abilities (Colombini et al., 2013), might Nitidulidae and Cleridae, and day þ6 for P. monodi). As was re- enable the displacement of sandhoppers to nearby undisturbed ported by Gilburn (2012), wrack removal also affects the ageing of zones, and consequently its rapid recovery after the impact. By 232 J.V. Vieira et al. / Marine Environmental Research 119 (2016) 222e237

Fig. 6. Temporal changes in density of numerically dominant taxa in the control (black column) and impacted (white column) plots, from before (16, 6, 3, 1 days) to after (þ1, þ3, þ6, þ16 days) the wrack removal for the two Spanish beaches. aec: taxa recorded at Levante Beach (iii). dee: taxa recorded at Cortadura Beach (iv). Data are mean of taxa density ± S.E. (n ¼ 18). Significant differences between plots (control and impacted) at any time was also represented (SNK tests, ***p < 0.001).

contrast, the inability to burrow, previously reported for P. monodi reservoirs (Colombini et al., 2011). This might constitute a mitiga- (Serejo, 2004; Stock, 1996) and its reduced mobility (personal tion strategy for the negative effect of wrack removal on supra- observation), make this amphipod more susceptible to the elimi- littoral arthropod populations, as our results suggest. However, this nation of wrack debris. Moreover, the beachhopper, P. monodi,is strategy was tested at an intermediate spatial scale (i.e. 100 m restricted to wrack deposits, where it takes refuge to avoid desic- along-shore), and therefore field experiments at larger spatial cation (Borzone and Rosa, 2009; Rosa et al., 2007). Alternatively, scales are needed, to determine the length of the uncleaned zones T. saltator which is a burrowing species, does not require a specific that are required to effectively conserve biodiversity. habitat (e.g. wrack debris) to satisfy its physiological needs Another important mitigation strategy might be to educate (Colombini et al., 2013), and therefore, can be considered to be beach users. It is essential that people (residents, visitors, and de- partially tolerant to this disturbance. cision makers) understand the ecological value of wrack debris to coastal wildlife, and that this organic material be left on some 4.2. Beach management implications sections of beaches (Colombini et al., 2011). A combined approach that incorporates spatial zoning and environmental education of all To many beach ecologists, the ideal management scenario is to stakeholders could provide a viable alternative to current man- introduce alongshore zoning, which provides the appropriate agement. For this, beach ecologists and coastal recreational man- planning of the shoreline (Defeo et al., 2009; Dugan et al., 2003; agers should work together and plan sustainable management McLachlan and Defeo, 2013; McLachlan et al., 2013). Overall, the strategies that do not inflict generating environmental and eco- spatial pattern of the recreational use of beaches is determined by nomic losses (Colombini et al., 2011). development, transport facilities, and popularity (de Ruyck et al., In summary, our results demonstrated that some species appear 1997). This human behaviour promotes a concentration of users to be more sensitive to wrack removal than others and its recovery in relatively small areas (Schlacher and Thompson, 2012), and thus patterns varied between beaches at each region. In fact, no beach is we recommend the manual removal of debris only in these areas. equal, so the local dimensions should be considered in human The adoption of this management strategy (spatial zoning) re- disturbances assessments (i.e. wrack removal) such as: 1) beach sults in the maintenance of uncleaned zones, within which the type; 2) wrack debris features and 3) specific density and compo- debris can remain on some sections of the beach year-round. When sition of individuals associated with wrack debris. Moreover, the the entire beach is cleaned, which is common practice in many amount and dynamic of the stranding wrack and species-specific coastal regions of Brazil and Spain (and other countries), the effects behavioural and biological strategies might determine the of wrack removal are more severe, since the disappearance of some response of supralittoral arthropod assemblages to the same species (particularly amphipods) has been documented (Gilburn, disturbance (i.e. wrack removal) at each Atlantic region. Coastal- 2012). In this way, uncleaned sectors might act as species zone management based on robust scientific data is essential to J.V. Vieira et al. / Marine Environmental Research 119 (2016) 222e237 233 ensure sustainable use of goods and services, as well as to conserve valho for map elaboration and graphical edition, and David Nesbitt the unique biota on sandy beaches (McLachlan and Defeo, 2013; Nel for English revision. J.V. Vieira was supported by the “Coordenaçao~ et al., 2014). Therefore, studies that contribute to understanding the de Aperfeiçoamento de Pessoal de Ensino Superior”, CAPES, of bottom-up effects associated with wrack removal and the potential Brazil (CAPES/DGU n 206/09) and M.C. Ruiz-Delgado was sup- consequences for the functioning of sandy beaches are extremely ported by the Spanish Ministry of Education via a predoctoral grant valuable and can serve as useful information for decision-makers (FPU, AP-2009-3906). This work was supported by the incentive and coastal managers. program to Excellent Research Projects, financed by the Regional Government of Andalusia (P09-HUM-4717) and by the Hispanic- Brazilian Program of interuniversity cooperation of Ministry of Acknowledgements Education and Science of Spain (PHB-2008-0132).

Special thanks to authorities of the Natural Park “Los Torunos~ ” for permission and its staff for facilities and help during the field Appendix experiment. We also thank all participants in the fieldwork for their valuable efforts. We specially thank Leonardo Sandrini-Neto and Maurício Carmargo for statistical support, Fabiano Grecco de Car-

Fig. A1. Temporal changes in total density (a, d), species richness (b, e) and diversity index (c, f) in the control (black column) and impacted (white column) plots, from before (16, 6, 3, 1 days) to after (þ1, þ3, þ6, þ16 days) the wrack removal for the two Brazilian beaches, Assenodi (i) and Cem (ii). Data are mean of assemblage descriptors ± S.E. (n ¼ 18). Significant differences between plots (control and impacted) at any time was also represented (SNK tests, ***p < 0.001). 234 J.V. Vieira et al. / Marine Environmental Research 119 (2016) 222e237

Fig. A2. Temporal changes in total density (a, d), species richness (b, e) and diversity index (c, f) in the control (black column) and impacted (white column) plots, from before (16, 6, 3, 1 days) to after (þ1, þ3, þ6, þ16 days) the wrack removal for the two Spanish beaches, Levante (iii) and Cortadura (iv). Data are mean of assemblage descriptors ± S.E. (n ¼ 18).

Table A1 Total abundance (n indv.), frequency (% assemblage) and constancy (% samples out of a total of 288) of arthropods collected in wrack deposits from the two Brazilian beaches, Assenodi (i) and Cem (ii).

Taxa (i) Assenodi beach (ii) Cem beach

Abundance Frequency Constancy Abundance Frequency Constancy

SubP. Crustacea Ord. Amphipoda Talitridae Platorchestia monodi 1,312 23.10 200 1,555 36.64 197 Ord. Isopoda Tylidae Tylos niveus 1 0.02 1 1 0.02 1 SubP. Hexapoda SubClas. Collembola 200 3.52 82 135 3.18 55 Ord. Coleoptera Coleoptera larvae 411 7.24 137 555 13.08 112 Staphylinidae Bledius sp1 45 0.79 30 9 0.21 8 Bledius sp2 5 0.09 4 eee Bledius bonariensis 2,756 48.50 222 22 0.52 20 Bledius hermani eee 32 0.75 27 Bledius fernandezi 4 0.07 4 1 0.02 1 Gabronthus eee 3 0.07 3 Heteroceridae Heteroceridae larvae 68 1.20 32 eee Efflagitatus freudei 21 0.37 19 eee Tenebrionidae Tenebrionidae larvae 1 0.02 1 70 1.65 56 Phaleria testaceae 6 0.11 5 161 3.79 79 Carabidae 541 9.52 155 eee Cleridae 17 0.30 6 365 8.60 128 J.V. Vieira et al. / Marine Environmental Research 119 (2016) 222e237 235

Table A1 (continued )

Taxa (i) Assenodi beach (ii) Cem beach

Abundance Frequency Constancy Abundance Frequency Constancy

Curculionidae 9 0.16 9 8 0.19 8 Nitidulidae 144 2.54 71 1,028 24.22 192 Scarabaeidae, Aphodiinae 8 0.14 8 16 0.38 15 Ord. Dermaptera riparia 7 0.12 6 2 0.05 2 Ord. Diptera Subord. Brachycera 71 1.25 52 205 4.83 97 SubP. Chelicerata Ord. Acariformes 2 0.04 2 42 0.99 35 Ord. Araneae Aranae sp1 1 0.02 1 eee Aranae sp2 eee 2 0.05 2 Aranae sp3 eee 1 0.02 1 Aranae sp4 1 0.02 1 eee Lycosidae 43 0.76 41 22 0.52 20 Linyphiidae Linyphiidae sp1 eee 4 0.09 4 Linyphiidae sp2 1 0.02 1 eee Labicymbium 1 0.02 1 1 0.02 1 Salticidae Salticidae sp1 1 0.02 1 eee Salticidae sp2 1 0.02 1 eee Salticidae sp3 1 0.02 1 eee Salticidae sp4 1 0.02 1 1 0.02 1 Theridiidae Theridiidae sp1 eee 1 0.02 1 Theridiidae sp2 eee 1 0.02 1 Theridiidae sp3 eee 1 0.02 1 Total abundance 5,680 4,244 Total number of species 29 27

Table A2 Total abundance (n indv.), frequency (% assemblage) and constancy (% samples out of a total of 288) of arthropods collected in wrack deposits from the two Spanish beaches, Levante (iii) and Cortadura (iv).

Taxa (iii) Levante (iv) Cortadura

Abundance Frequency Constancy Abundance Frequency Constancy

SubP. Crustacea Ord. Amphipoda Talitridae Talitrus saltator 1,784 62.82 234 23,764 95.14 278 Ord. Isopoda Tylidae Tylos europaeus eee 108 0.43 56 Armadillidiidae Armadillidium 4 0.14 3 eee SubP. Hexapoda Ord. Coleoptera Coleoptera larvae 110 3.87 65 154 0.62 65 Staphylinidae Aleocharinae Aleocharinae sp1 30 1.06 21 104 0.42 34 Aleocharinae sp2 eee 1 0.00 1 Aleocharinae sp3 62 2.18 43 552 2.21 99 Oxytelinae Carpelimus 18 0.63 7 1 0.00 1 Staphylininae Cafius 13 0.46 9 17 0.07 13 Cafius xantholoma 7 0.25 7 13 0.05 13 Tenebrionidae Phaleria bimaculata 155 5.46 59 90 0.36 44 Carabidae 21 0.74 21 6 0.02 6 Curculionidae 7 0.25 7 1 0.00 1 Chrysomelidae 6 0.21 6 1 0.00 1 Elateroidea 4 0.14 4 2 0.01 2 (continued on next page) 236 J.V. Vieira et al. / Marine Environmental Research 119 (2016) 222e237

Table A2 (continued )

Taxa (iii) Levante (iv) Cortadura

Abundance Frequency Constancy Abundance Frequency Constancy

Hydrophilidae 2 0.07 2 eee Scarabaeidae 6 0.21 6 eee Histeridae 7 0.25 7 9 0.04 8 Ord. Diptera Diptera larvae 172 6.06 34 40 0.16 24 Diptera pupa 38 1.34 29 18 0.07 12 Subord. Brachycera 285 10.04 83 35 0.14 29 Subord. Nematocera 43 1.51 31 55 0.22 43 Ord. Hemiptera Hemiptera 8 0.28 6 eee Subord. Heteroptera Miridae 4 0.14 4 eee Tingidae 2 0.07 2 eee Pteromalidae 11 0.39 11 eee Mesoveliidae 4 0.14 4 1 0.00 1 Saldidae 12 0.42 10 eee Cymicidae 3 0.11 3 eee Ord. Thysanoptera 1 0.04 1 eee SubP. Chelicerata Ord. Araneae Agelinidae 1 0.04 1 eee Dictynidae 1 0.04 1 eee Gnaphosidae 12 0.42 10 1 0.00 1 Lyniphiidae 4 0.14 4 3 0.01 3 Salticidae 3 0.11 3 1 0.00 1 Total abundance 2,840 24,977 Total number of species 33 23

the northern coast of Brittany (France). Estuar. Coast. Shelf Sci. 56, 91e98. Anderson, M.J., 2001. A new method for non-parametric multivariate analysis of variance. Austral Ecol. 26, 32e46. Table A3 Anderson, M.J., Gorley, R.N., Clarke, K.R., 2008. PERMANOVA for PRIMER: Guide to SIMPER analysis to evaluate the contributions of taxa to dissimilarities between Software and Statistical Methods. PRIMEReE Ltd, Plymouth, United Kingdom. control and impacted plots after wrack removal for the two Brazilian beaches: Barreiro, F., Gomez, M., Lastra, M., Lopez, J., de la Huz, R., 2011. Annual cycle of wrack Assenodi (i) and Cem (ii). supply to sandy beaches: effect of the physical environment. Mar. Ecol. Prog. Ser. 433, 65e74. (i) Assenodi beach Barreiro, F., Gomez, M., Lopez, J., Lastra, M., de la Huz, R., 2013. Coupling between macroalgal inputs and nutrients outcrop in exposed sandy beaches. Hydro- Groups: Control after & impacted after; average dissimilary ¼ 79.55% biologia 700, 73e84. Benavente, J., Del Río, L., Anfuso, G., Gracia, F.J., Reyes, J.L., 2002. Utility of mor- Species after period Control plot Impacted plot Contrib % Cum. % phodynamic characterisation in the prediction of beach damage by storms. Av. Abund Av. Abund J. Coast. Res. 36, 56e64. Bessa, F., Rossano, C., Nourisson, D., Gambineri, S., Marques, J.C., Scapini, F., 2013. Bledius bonariensis 113.89 318.06 34.25 34.25 Behaviour of Talitrus saltator (Crustacea: Amphipoda) on a rehabilitated sandy Platorchestia monodi 269.44 54.17 31.84 66.09 beach on the European Atlantic coast (Portugal). Estuar. Coast. Shelf Sci. 117, Carabidae 45.14 116.67 13.52 79.61 168e177. Coleoptera larvae 21.53 19.44 5.59 85.21 Borzone, C.A., Rosa, L.C., 2009. Impact of oil spill and posterior clean-up activities on Nitidulidae 30.56 0.00 4.03 89.24 wrack-living talitrid amphipods on estuarine beaches. Braz. J. Oceanogr. 57 (4), Bledius sp1 13.19 9.03 2.97 92.21 315e323. Clarke, K.R., Warwick, R.M., 1994. Change in Marine Communities: an Approach to (ii) Cem beach Statistical Analysis and Interpretation. PRIMER-E, Plymouth, UK. Groups: Control after (þ1 day) & impacted after (þ1 day); average Colombini, I., Aloia, A., Fallaci, M., Pezzoli, G., Chelazzi, L., 2000. Temporal and spatial use of stranded wrack by the macrofauna of a tropical sandy beach. Mar. dissimilary ¼ 97.74% Biol. 136, 531e541. Species þ1 day (after period) Control plot Impacted plot Contrib % Cum. % Colombini, I., Chelazzi, L., 2003. Influence of marine allochthonous input on sandy beach communities. Oceanogr. Mar. Biol. An Annu. Rev. 41, 115e159. Av. Abund Av. Abund Colombini, I., Fallaci, M., Chelazzi, L., 2011. Terrestrial macroinvertebrates as key elements for sustainable beach management. J. Coast. Res. Spec. Iss. 61, 24e35. Platorchestia monodi 433.33 0.00 46.94 46.94 Colombini, I., Fallaci, M., Gagnarli, E., Rossano, C., Scapini, F., Chelazzi, L., 2013. The Nitidulidae 150.00 0.00 15.53 62.47 behavioural ecology of two sympatric talitrid species, Talitrus saltator (Mon- Cleridae 94.44 5.56 13.68 76.15 tagu) and Orchestia gammarellus (Pallas) on a Tyrrhenian sandy beach dune Phaleria testacea 133.33 0.00 11.19 87.34 system. Estuar. Coast. Shelf Sci. 117, 37e47. Tenebrionidae larvae 22.22 16.67 4.90 92.25 Colombini, I., Mateo, M.A., Serrano, O., Fallaci, M., Gagnarli, E., Serrano, L., Chelazzi, L., 2009. On the role of Posidonia oceanica beach wrack for macro- invertebrates of a Tyrrhenian sandy shore. Acta Oecol. 35, 32e44. Davenport, J., Davenport, J.L., 2006. The impact of tourism and personal leisure transport on coastal environments: a review. Estuar. Coast. Shelf Sci. 67, 280e292. Defeo, O., McLachlan, A., Schoeman, D.S., Schlacher, T., Dugan, J., Jones, A., Lastra, M., Scapini, F., 2009. Threats to sandy beach ecosystems: a review. Estuar. Coast. References Shelf Sci. 81, 1e12. De Ruyck, A.M.C., Soares, A.G., McLachlan, A., 1997. Human recreational patterns on Abramoff, M.D., Magalhaes, P.J., Sunanda, J.R., 2004. Image processing with ImageJ. beaches with different levels of development. Trans. R. Soc. S. Afr. 52, 257e276. Biophot. Int. 11 (7), 36e42. Downes, B.J., Barmuta, L.A., Fairweather, P.G., Faith, D.P., Keough, M.J., Lake, P.S., Adin, R., Riera, P., 2003. Preferential food source utilization among stranded mac- Mapstone, B.D., Quinn, G.P., 2004. Monitoring Ecological Impacts: Concepts and roalgae by Talitrus saltator (Amphipoda, Talitridae): a stable isotopes study in Practice in Flowing Waters. Cambridge University Press, UK. J.V. Vieira et al. / Marine Environmental Research 119 (2016) 222e237 237

Dugan, J.E., Hubbard, D.M., 2010. Loss of coastal strand habitat in Southern Cali- and recreation:guidelines for optimising management strategies for multi- fornia: the role of beach grooming. Estuar. Coast. 33, 67e77. purpose use. Ocean Coast. Manag. 71, 256e268. Dugan, J.E., Hubbard, D.M., McCrary, M.D., Pierson, M.O., 2003. The response of Morales, M., 2012. Sciplot: Scientific Graphing Functions for Factorial Designs macrofauna communities and shorebirds to macrophyte wrack subsidies on (Available on CRAN). exposed sandy beaches of southern California. Estuar. Coast. Shelf Sci. 58, Nel, R., Campbell, E.E., Harris, L., Hauser, L., Schoeman, D.S., McLachlan, A., du 25e40. Preez, D.R., Bezuidenhout, K., Schlacher, T.A., 2014. The status of sandy beach Dugan, J.E., Hubbard, D.M., Page, H.M., Schimel, J.P., 2011. Marine macrophyte wrack science: past trends, progress, and possible futures. Estuar. Coast. Shelf Sci. 150, inputs and dissolved nutrients in beach sands. Estuar. Coast. 34, 839e850. 1e10. Emery, K.O., 1961. A simple method of measuring beach profiles. Limnol. Oceanogr. Nordstrom, K.F., Lampe, R., Vandemark, L.M., 2000. Reestablishing naturally func- 6, 90e93. tioning dunes on developed coasts. Environ. Manag. 25, 37e51. Fairweather, P.G., Henry, R., 2003. To clean or not to clean? Ecologically sensitive Noriega, R., Schlacher, T.A., Smeuninx, B., 2012. Reductions in ghost crab pop- management of wrack deposits on sandy beaches. Ecol. Manag. Restor. 4, ulations reflect urbanization of beaches and dunes. J. Coast. Res. 28, 123e131. 227e228. Ochieng, C.A., Erftemeijer, P.L.A., 1999. Accumulation of seagrass beach cast along Fanini, L., Cantarino, C.M., Scapini, F., 2005. Relationships between the dynamics of the Kenyan coast: a quantitative assessment. Aquat. Bot. 65, 221e238. two Talitrus saltator populations and the impacts of activities linked to tourism. Olabarria, C., Incera, M., Garrido, J., Rodil, I.F., Francesca, R., 2009. Intraspecific diet Oceanologia 47, 93e112. shift in Talitrus saltator inhabiting exposed sandy beaches. Estuar. Coast. Shelf Folk, R.L., Ward, W.C., 1957. Brazos River Bar: a study in the significance of grain size Sci. 84, 282e288. parameters. J. Sediment. Petrol. 27, 3e26. Piriz, M.L., Eyras, M.C., Rostagno, C.M., 2003. Changes in biomass and botanical Gandara-Martins, A., Borzone, C.A., Rosa, L.C., Caron, E., 2010. Ocorrencia^ de tres^ composition of beach-cast seaweeds in a disturbed coastal area from Argentine especies do genero^ Bledius Leach, 1819 (Coleoptera, Staphylinidae, Oxytelinae) Patagonia. J. Appl. Phycol. 15 (1), 67e74. nas praias arenosas expostas do Parana, Brasil. Braz. J. Aquat. Sci. Technol. 14 (2), R Core Team, 2012. R: a Language and Environment for Statistical Computing. R 23e30. Foundation for Statistical Computing. Vienna, Austria. http://www.R-project. Gilburn, A.S., 2012. Mechanical grooming and beach award status are associated org/. with low strandline biodiversity in Scotland. Estuar. Coast. Shelf Sci. 107, 81e88. Rosa, L.C., Aluizio, R., Borzone, C.A., 2007. Talitridae (Amphipoda, Crustacea) nas Glasby, T.M., Underwood, A.J., 1996. Sampling to differentiate between pulse and praias estuarinas da baía de Paranagua, sul do Brasil: distribuiçao~ espacial e press perturbations. Environ. Monit. Assess. 42, 241e252. abundancia.^ Braz. J. Aquat. Sci. Technol. 11, 37e41. Gonçalves, S.C., Marques, J.C., 2011. The effects of season and wrack subsidy on the Ruiz-Delgado, M.C., Reyes-Martínez, M.J., Sanchez-Moyano, J.E., Lopez-P erez, J., community functioning of exposed sandy beaches. Estuar. Coast. Shelf Sci. 95, García-García, F.J., 2015. Distribution patterns of supralittoral arthropods: wrack 165e177. deposits as a source of food and refuge on exposed sandy beaches (SW Spain). Harris, L., Campbell, E.E., Nel, R., Schoeman, D., 2014. Rich diversity, strong ende- Hydrobiologia 742, 205e219. mism, but poor protection: addressing the neglect of sandy beach ecosystems Ruiz-Delgado, M.C., Vieira, J.V., Reyes-Martínez, M.J., Borzone, C.A., Sanchez- in coastal conservation planning. Divers. Distrib. 20 (10), 1120e1135. Moyano, J.E., García-García, F.J., 2016. Wrack removal as short-term disturbance Herman, L.H., 1986. Revision of Bledius. Part IV. Classification of species, groups, for Talitrus saltator density in the supratidal zone of sandy beaches: an exper- phylogeny, natural history, and catalogue (Coleoptera, Staphylinidae, Oxy- imental approach. Estuar. Coast. 39, 1113e1121. telinae). Bull. Am. Mus. Nat. Hist. 184, 1e368. Ruiz-Delgado, M.C., Vieira, J.V., Veloso, V.G., Reyes-Martínez, M.J., Sallorenzo, A.I., Hurlbert, S.H., 1984. Pseudoreplication and the design of ecological field experi- Borzone, C.A., Sanchez-Moyano, J.E., García-García, F.J., 2014. The role of wrack ments. Ecol. Monogr. 54, 187e211. deposits for supralittoral arthropods: an example using Atlantic sandy beaches Ince, R., Hyndes, G.A., Lavery, P.S., Vanderklift, M.A., 2007. Marine macrophytes of Brazil and Spain. Estuar. Coast. Shelf Sci. 136, 61e71. directly enhance abundances of sandy beach fauna through provision of food Sandrini-Neto, L., Camargo, M.G., 2012. GAD: an R Package for ANOVA Designs from and habitat. Estuar. Coast. Shelf Sci. 74, 77e86. General Principles (Available on CRAN). Jaramillo, E., de la Huz, R., Duarte, C., Contreras, H., 2006. Algal wrack deposits and Schlacher, T.A., Schoeman, D.S., Dugan, J., Lastra, M., Jones, A., Scapini, F., macroinfaunal arthropods on sandy beaches of the Chilean coast. Rev. Chil. Hist. McLachlan, A., 2008. Sandy beach ecosystems: key features sampling issues, Nat. 79 (3), 337e351. management challenges and climate change impacts. Mar. Ecol. 29, 70e90. Lana, P.C., Marone, E., Lopes, R.M., Machado, E.C., 2001. The subtropical estuarine Schlacher, T.A., Thompson, L., 2012. Beach recreation impacts benthic invertebrates complex of Paranagua Bay, Brazil. Ecol. Stud. 144, 131e145. on ocean exposed sandy shores. Biol. Conserv. 147, 123e132. Lastra, M., Page, H.M., Dugan, J.E., Hubbard, D.M., Rodil, I.F., 2008. Processing of Serejo, C., 2004. Talitridae (Amphipoda, Gammaridea) from the Brazilian coastline. allochthonous macrophyte subsidies by sandy beach consumers: estimates of Zootaxa 646, 1e29. feeding rates and impacts on food resources. Mar. Biol. 154, 163e174. Short, A.D., Wright, L.D., 1983. Physical variability of sandy beaches. In: Llewellyn, P.J., Shackley, S.E., 1996. The effects of mechanical beach-cleaning on McLachlan, A., Erasmus, T. (Eds.), Sandy Beaches as Ecosystems. Dr. W. Junk invertebrate populations. Br. Wildl. 7, 147e155. Publishers, The Hague, The Netherlands, pp. 133e144. Malm, T., Raberg, S., Sabine, F., Carlsson, P., 2004. Effects of beach cast cleaning on Stock, J.H., 1996. The genus Platorchestia (Crustacea, Amphipoda) on the Mid beach quality, microbial food web, and littoral macrofaunal biodiversity. Estuar. Atlantic islands, with description of a new species from Saint Helena. Misc. Zool. Coast. Shelf Sci. 60, 339e347. 19.1, 149e157. Martin, K., Speer-Blank, T., Pommerening, R., Flannery, J., Carpenter, K., 2006. Does Underwood, A.J., 1997. Experiments in Ecology: Their Logical Design and Interpre- beach grooming harm grunion eggs? Shore Beach 74, 17e22. tation Using Analysis of Variance. Cambridge University Press, Cambridge. Masselink, G., Short, A.D., 1993. The effect of tide range on beach morphodynamics Underwood, A.J., 2000. Importance of experimental design in detecting and and morphology: a conceptual beach. J. Coast. Res. 9, 785e800. measuring stresses in marine populations. J. Aquat. Ecosyst. Stress Recovery 7, McLachlan, A., Brown, A.C., 2006. The Ecology of Sandy Shores, second ed. Academic 3e24. Press, New York. Vianna, A.X.M.R., Borzone, C.A., 2015. Distribuiçao~ espacial e temporal de especies McLachlan, A., Defeo, O., 2013. Coastal beach ecosystems. Encycl. Biodivers. 2, de Bledius (Coleoptera, Staphylinidae) em praias arenosas no sul do Brasil. 128e136. Iheringia Serie Zool. 105 (1), 122e132. McLachlan, A., Defeo, O., Jaramillo, E., Short, A.D., 2013. Sandy beach conservation