Fish assemblages along a morphodynamic continuum on three tropical beaches

A comparative study involving three mesotidal beaches from reflective to estuarine states was conducted between October 2010 and September 2011 in the estuary of Mamanguape River, northeastern Brazil. The aim was to test for differences in fish assemblages along a morphodynamic continuum, particularly the effects of wave action. In total, 3,732 fish from 71 species were examined; Rhinosardinia bahiensis and Atherinella brasiliensis dominated the estuarine beach, whereas Trachinotus carolinus dominated the reflective beach. Fish assemblages were structured by spatial differences among the sites, and overall abundance, richness and biomass demonstrated a clear increasing trend from the reflective to the estuarine beaches, indicating a primary influence of wave exposure in the measured parameters. The Mamanguape River estuary plays a crucial ecological role as a nursery for several species that depend on its resources and have adapted to the high variability of its physical and chemical factors and the diversity of its habitats.


Introduction
The importance of shallow coastal biotopes in the maintenance of fish assemblages has been clearly demonstrated by studies conducted worldwide (Layman, 2000;Ramos et al., 2006;Kruger & Strydom, 2010;Woodland et al., 2012).These biotopes are often described as nursery habitats or recruitment areas for fish because they provide a great abundance of food resources and protection from predation.They may also act as a temporary habitat during a particular life cycle phase and/or as a permanent habitat for young-of-the-year, thereby ensuring the renewal of fish stock (McLachlan, 1983;Pessanha et al., 2000;Beck et al., 2001;Able, 2005).
Beaches are defined by McLachlan (1983) as a range of environments on a gradient from open sea to protected sands or mudflats in estuaries.The classification of beaches is performed using several models, including those that consider the beach morphodynamics (Wright & Short, 1984) and the influence of tides (Short, 1982;Masselink, 1993).In the case of reflective beaches, wave action generates physical stress at these sites that affects sedimentary structure and makes food resources available for fish that can adapt to these conditions (Calliari et al., 2003;Vasconcellos et al., 2007;Niang et al., 2010).Moreover, fish in such environments, with physical conditions created by tides and waves, possess adaptations that may be morphological (Palmeira & Monteiro-Neto, 2010), or even behavioural adaptations, such as migration to the intertidal zones or other habitats (Harvey, 1998).
In contrast, estuarine beaches are characterized as calmer environments that result from a reduced influence of wave energy (Bernabeu et al., 2012).Several fish species utilize these environments at some stage of their life cycle because of factors such as availability of food from high primary production; availability of refuge provided by structural complexity, especially for juveniles; high water turbidity; and a lower number of large carnivorous fish (Potter et al., 1986;Whitfield, 1999;Beck et al., 2001;Spach et al., 2003;Hajisamae et al., 2006;Paiva et al., 2008).
One of the primary bases for understanding fish assemblages is the knowledge of their distribution and abundance and their relationship with abiotic factors (Inoue et al., 2008), as well as factors that will limit or affect a change in population (Levin et al., 1997).Several studies from different regions have emphasized abiotic factors, such as temperature and salinity (Harrison & Whitfield, 2006), salinity and turbidity (Cyrus & Blaber, 1992;Barletta et al., 2005), turbidity (Cyrus & Blaber, 1987;Abou-Seedo et al., 1990), morphodynamic models (Inui et al., 2010), and input of freshwater and river flow (Barletta et al., 2005;Vinagre et al., 2009;Baptista et al., 2010;Castillo-Rivera, 2013) as factors that influence the abundance of young fish in these shallow water habitats.Additionally, factors such as recruitment, competition, predation, and structural complexity are listed as ecological processes that determine the structure of these communities, so that the fish assemblages are the result of the complex interaction of abiotic and biotic factors (Mariani, 2001;Laegdsgaard & Johnson, 2001).
In tropical northeastern Brazilian estuaries, information regarding juvenile fish composition, seasonal variations in abundance, and biomass in shallow water is still largely unavailable.This study aimed to describe the structure and dynamics of fish assemblages on three beaches in the Mamanguape River estuary, and to test the hypothesis that distinct fish assemblages occur along a morphodynamic continuum as the result of changes in environmental variables, particularly the effects of wave action.

Study site
The study was conducted in the Mamanguape River estuary (6º43'02"S 35º67'46"W), which is the second largest estuary in the state of Paraíba (northeastern Brazil).The estuary is located in an environmental protection area ("EPA") (CERHPB, 2004), whose goal is to protect coastal ecosystems as well as the marine manatee Trichechus manatus Linnaeus, 1758 (Mourão & Nordi, 2003).The climate is tropical and humid, as its annual total rainfall varies from 2000 mm to less than 30 mm.Temperatures are high, as the annual average temperature is between 24ºC and 26ºC (Macedo et al., 2010).The beaches display a mesotidal, semi-diurnal tidal regime.The mouth of the river forms a 6-km-wide bay that is nearly closed off by a coastal reef line, which results in calm and quiet waters (Paludo & Klonowski, 1999).In this area, well-preserved mangroves are present that are principally represented by Rhizophora mangle, Avicennia schaueriana, Avicennia germinans, Laguncularia racemosa, and Conocarpus erectus, which grow around the main channel and its intertidal creeks to encompass approximately 6,000 hectares, in addition to the remnants of the Atlantic rainforest (Rocha et al., 2008).

Sampling
Samples were collected between October 2010 and September 2011, on monthly daytime excursions during spring tides.Sampling was performed at three sites defined according to their locations relative to the mouth of the river (Fig. 1): Curva do Pontal Beach (CPO) (6º46'27"S 34º55'20"W ), which is situated in a protected estuarine region, and presents predominantly muddy sediment, calm waters and seagrass meadows; Pontal Beach (PON) (6º46'22"S 34º55'07"W), situated beyond the mouth of Mamanguape River, with sandy sediment and a barrier reef parallel to the shoreline which absorbs most of the wave energy; and Campina Beach (CAM) (6º48'43"S 34º54'49"W), which presents sandy sediment, gravels and high wave energy.
The fish were sampled using a beach seine net (10 m x 1.5 m, with an 8-mm mesh).The seine hauls were 30 m long, parallel to and close to the shore, and were taken out to a depth of approximately 1.5 m.This procedure was replicated five times at each site, separated 5 m apart to minimize the influence on the following haul.The fish collected were fixed in 10% formalin and later identified (Figueiredo & Menezes, 1978, 1980, 2000;Menezes & Figueiredo, 1980, 1985;Araujo et al., 2004), counted, measured (total length in mm), and weighed (g).
For each sample, environmental factors such as water temperature, salinity and transparency were also measured five times at each site, using a mercury thermometer, an optical refractometer and a Secchi disc, respectively.Samples from the sediment were collected in at each beach in five replicates.Granulometric analysis was performed according to Embrapa (1997), with the method using a combination of sieving and sedimentation steps and is designed to be used in conjunction with analysis of particulate organic matter (POM).Wave height and wave period were measured by visually estimating the height of the breaker waves during one minute for three times and the time interval between five successive breaker waves was recorded.

Data analysis
The morphological classification of beaches was performed according to the RTR (Relative Tide Range), which takes into account the average variation of tides and the wave height, whose reference values are: 3<RTR<7: group III (reflective state); 7<RTR<15: group II (intermediate state); RTR>15 (dissipative state) (Masselink 1993;Calliari et al., 2003).
The environmental variables and the fish abundance (weight and number) were transformed (log 10 x+1) to meet the assumptions of normality and homoscedasticity prior to comparison using one-way analysis of variance (ANOVA).Analyses were used for spatial comparison between abundance and environmental variables.A post hoc comparison of means was performed using the Tukey HSD test (Zar, 1982).
The data matrix to be analysed via multivariate techniques was created from the numerical abundance data for the species, using the Bray-Curtis similarity coefficient as a measure.An ANOSIM was used to test for differences in fish assemblage structure between beaches.When the ANOSIM test comparisons detected a significant difference, SIMPER was used to identify the species that were related to each beach type.These analyses were undertaken using PRIMER 6.0® statistical software.
Influences of environmental variables on fish assemblages were analysed using Canonical Correspondence Analysis (CCA), and the results were tested with the Monte Carlo test with 499 permutations.Direct analysis of the gradients through CCA clarified the role that environmental factors play in the structuring of fish assemblage by comparing species composition between sampling sites.A forward selection procedure was performed to identify variables that accounted for the greatest variation.The significance of these variables, defined as p < 0.05, was tested in Monte Carlo simulations.The Monte Carlo permutation test, at α = 0.05, was also used to test the statistical significance of the relationship between species and environmental variables.The software used to perform the analyses was CANOCO ® for Windows.
The analysis of the fish assemblages at the three beaches showed that there was a higher abundance, biomass and richness in the dissipative beach than at reflective and intermediary beaches, which demonstrates a significant increase at the site with the lowest exposure to wave activity (Fig. 3).This spatial difference was significant as determined by ANOSIM (Global R = 0.431; P> 0.1), with the dissipative beach being different from intermediary (Global R = 0.495; Spatially, no significant differences were found for temperature among the beaches.Highest salinity levels were recorded at Reflective Beach (41.3±0.3) and lowest at Dissipative Beach (11.3±0.9), the highest transparency values were found at Dissipative Beach (99.0±10.6)and lowest at Reflective Beach (10.0±0.6).The results were significantly different (Table 1).

Structure and composition
In total, 3,732 individuals belonging to 71 species (27 families) were collected during 174 beach seine hauls at beaches in the Mamanguape River estuary during the study period (Table 2).The family Carangidae had the largest number of species (7), followed by Sciaenidae (6) and Engraulidae, Gerreidae, Gobiidae, and Tetraodontidae (each with 5).The families with the highest number of individuals were Clupeidae, Atherinopsidae, Engraulidae, and Mugilidae, which collectively represented 79% of the total number of fish.The families Tetraodontidae, Clupeidae, Atherinopsidae, and Engraulidae, in decreasing order, contributed the most to the weight frequency and accounted for 73% of the total weight of the fish.Bahia sprat herring Rhinosardinia bahiensis (Steindachner, 1879) contributed the most to numerical abundance, as it accounted for 32% of the total, followed by the Brazilian silverside, Atherinella brasiliensis (Quoy & Gaimard, 1825), which represented 16% of the total abundance.The highest biomass contribution was from the puffer fish Sphoeroides testudineus (Linnaeus, 1758) which contributed 29% of the total biomass.Among the species, 18 occurred only once in the samples and 7 accounted for less than 0.01% of the biomass.
The families with the highest number of individuals in the dissipative beach were Clupeidae, Atherinopsidae, Engraulidae, and Mugilidae (81% of total number of individuals).In the intermediary beach, Engraulidae, Clupeidae, Haemulidae, Carangidae, and Sciaenidae were the abundant families (75%), whereas in the reflective beach,  Monte Carlo analysis revealing that RTR and characteristics of sediment (percentage of sand, silt, and clay) contributed most to species distribution.Samples representing the beaches were separated along the first axis.Axis 1 was positively correlated with RTR and silt clay, and negatively correlated with sand.

Discussion
Spatial differences in the structure for fish assemblages was demonstrated among the three tropical beaches that are strongly influenced by wave action, which is thought by many authors to be synonymous with exposure.This heterogeneous spatial distribution pattern of fish species showed a higher abundance, biomass and richness for the estuarine beach compared with the reflective beach with different species being characteristics of each beaches.These observations corroborate the hypothesis that the degree of exposure is a primary factor shaping local assemblages of fishes.This pattern has been demonstrated in other studies in tropical and subtropical regions (Romer, 1990;Clark, 1997;Jovanovic et al., 2007, Vasconcellos et al., 2007).
Estuarine beaches, which are also called "low energy" beaches, differ from the most exposed beaches in having additional features such as a narrow and abrupt sandy upper beach portion, a marked break in profile and a broad coastal plain with a smooth gradient (Rosa & Borzone, 2008).In estuarine beaches, the high fish abundance observed results from the aggregation of Clupeiforms species (such as R. bahiensis, A. clupeioides (Swainson, 1839), and L. grossidens in this study).Therefore, these beaches have been considered to be recruitment areas during the life cycles of these species (Oliveira-Silva et al., 2008;Hackradt et al., 2009).Moreover, the aggregation serves as a temporary refuge in calmer habitats close to the coast because high-energy sites do not permit such behaviour, mainly because of the high dynamic of the water column, which potentially results in high energy consumption (Felix et al., 2007).The use of calmer sites, such as those inside of bays and estuaries (Blaber & Blaber, 1980;Pessanha & Araújo, 2003) and even sheltered sandy beaches (Barreiros et al., 2004), plays an important recruitment role for pelagic species.P > 0.1) and reflective (Global R = 0.747; P > 0.1) beaches.SIMPER showed higher contributions from L. grossidens, R. bahiensis, A. brasiliensis, S. testudineus, and Hyporhamphus unifasciatus (Ranzani, 1841) for dissipative beach, whereas T. carolinus, Trachinotus goodei Jordan and Evermann, 1896, M. liza and P. virginicus showed higher contributions for intermediary and reflective beaches.

Influence of environmental factors
The first two axes from canonical correspondence analysis accounted for 63.1% of the cumulative percentage of variance for the environmental-species relationship (Fig. 4), with the     Nelson (2006).
and thus increases their success rate when catching suspended prey, as most of the juveniles are zooplanktivorous (Lasiak, 1986;Inoue et al., 2008).These characteristics indicate that estuarine beaches function more effectively than reflective beaches as nursery areas for a larger number of species.
Reflective beach has the lowest abundance and richness among the three studied beaches because of limiting factors such as increased wave energy and the high variability of abiotic factors.The degree of wave exposure is one of the characteristics that contributes most to determine fish For juvenile and young-of-the-year fishes, beaches located within estuaries provide an ideal site for growth and protection.At these beaches, a constant supply of food has been observed because of the primary production and the absence of large predators (Inui et al., 2010).As species indicative of sheltered environments, S. testudineus (Felix et al., 2007), M. liza (Pessanha & Araújo, 2003;Vasconcellos et al., 2007) and Bathygobius soporator (Valenciennes, 1837) (Oliveira-Silva et al., 2008) are prominent.Furthermore, the small size of the individuals increases their agility in shallow coastal water connectivity between the fish assemblage.This connectivity between beaches and estuarine environments has been described for subtropical environments (Monteiro-Neto et al., 2008) for species of Clupeiformes, and also in a study by Xavier et al. (2012) that observed connectivity between reef systems and mangroves within the same study area.
Results of this study indicate that two factors, namely the RTR and characteristics of sediment, play a major role determining spatial variations in fish abundance and community structure.Some other variables, e.g., salinity and transparency, also play an important role, but are probably of secondary importance.Species-specific preferences for varying combinations of these variables result in variations in the structure and overall abundance of fish assemblages in different areas.Overall, the following trends were observed: the dissipative beaches were found to harbor the most diverse assemblage of fishes, while the reflective beach harbored the lowest densities of fishes.assemblage structure, with typical of surf zones supporting high numbers of fish species, such as Trachinotus (T.carolinus and T. goodei) (Gomes et al., 2003;McLachlan & Brown, 2006), Menticirrhus [M.americanus (Linnaeus, 1758) and M. littoralis) (Layman, 2000;McLachlan & Brown, 2006) and Polydactylus (P.virginicus) (Gianinni & Paiva-Filho, 1995;McLachlan & Brown, 2006;Oliveira-Silva et al., 2008)].These species can be considered to be residents of the surf zones of reflective beaches, as indicated by McLachlan & Brown (2006).Few species can adapt and colonize highly dynamic environments such as the surf zones of reflective beaches (Vasconcellos et al., 2007), although these beaches provide protection from predation and a wide availability of zooplankton for feeding juvenile and larval fish (Lasiak, 1986).Wootton (1992) explains that fish from high-energy zones developed during their evolutionary history appropriate body shapes for minimizing energy costs by effectively counteracting the effects of waves, i.e., a more hydrodynamic form that minimizes the effects of wave energy.
The low number of individuals captured at the intermediate and reflective beaches indicates that these beaches act as nursery areas for a limited number of species.According to Beck et al. (2001), a characteristic that is considered primary for defining nursery areas is their great contribution to the production of individuals that recruit for adult populations.Nonetheless, some studies suggest that selective capture by the net, as well as the difficulty of seine hauling on high-exposure beaches, could influence the low abundance at reflective and intermediate beaches (Romer, 1990;Vasconcelos et al., 2007).
The beaches have functional ecological connections with adjacent ecosystems, such as estuaries and coastal lagoons, and may serve as passage routes and/or habitats, especially for juvenile individuals (Inui et al., 2010).On beaches adjacent to estuaries, the surf zone could provide water masses with low salinity that have been shown to be important in attracting juvenile fish, which are usually recruited within protected environments (Fisher & Pearcy, 1995).Therefore, the distribution of such species is determined according to a continuum or environmental gradient (herein represented by the gradual change in the degree of wave exposure along the estuary) where species turnover occurs along gradients of physical conditions, and where there are no well-defined barriers that isolate different communities (Ricklefs, 2003).The occurrence of certain species, such as M. littoralis, P. virginicus, and T. goodei in reflective and intermediate beaches, and A. brasiliensis and L. grossidens through all three beaches, demonstrates that there is connectivity between the beaches and the estuary, which may represent an open community pattern.Among other species that had a significant representation at the reflective beach in this study, M. liza and M. hospes are noteworthy because they had a high abundance of juveniles recruited at the estuarine beach, which suggests

Fig. 2 .
Fig. 2. Mean monthly variations of the environmental variables (±SE) for the study period.

Fig. 3 .
Fig. 3. Variation in the mean number of individuals (CPUE average ±SE), biomass (±SE) and number of species (±SE) at Mamanguape River estuary in Brazil.

Table 1 .
Results from ANOVA for comparisons of abiotic (temperature, salinity and transparency) and biotic (CPUE, Species number and Biomass) factors between beaches and hydrological regime in the Mamanguape River estuary, Brazil.**, p < 0.01; ns, not significant.

Table 2 .
Nelson (2006)ndance, biomass, frequency of occurrence, spatial distribution and maximum and minimum length of fish species captured on three beaches in the Mamanguape River estuary of Brazil.The species are ordered in accordance withNelson (2006).

Table 2 .
Cont.Numerical abundance, biomass, frequency of occurrence, spatial distribution and maximum and minimum length of fish species captured on three beaches in the Mamanguape River estuary of Brazil.The species are ordered in accordance with