Use of intertidal areas by shrimps ( Decapoda ) in a brazilian Amazon estuary

The present work investigated the occupation and the correlation of the shrimp abundance in relation to environmental variables in different habitats (mangroves, salt marshes and rocky outcrops) in an Amazon estuary. The collections were made in August and November 2009, at low syzygy tide on Areuá Beach, situated in the Extractive Reserve of Mãe Grande de Curuçá, Pará, Brazil totaling 20 pools. In each environment, we recorded the physical-chemical factors (pH, salinity, and temperature) and measured the area (m2) and volume (m3) of every pool through bathymetry. The average pH, salinity, temperature, area and volume of tide pools were 8.75 (± 0.8 standard deviation), 35.45 (± 3), 29.49 °C (± 2.32), 27.41 m2 (± 41.18), and 5.19 m3 (± 8.01), respectively. We caught a total of 4,871 shrimps, distributed in three families and four species: Farfantepenaeus subtilis (98.36%) (marine) followed by Alpheus pontederiae (0.76%) (estuarine), Macrobrachium surinamicum (0.45%) and Macrobrachium amazonicum (0.43%) predominantly freshwater. The species F. subtilis and A. pontederiae occurred in the three habitats, whereas M. surinamicum occurred in salt marsh and rocky outcrop and M. amazonicum only in marisma. Temperature and pH were the most important environmental descriptors that significantly affected the density and biomass of shrimps.


INTRODUCTION
Estuaries are places to which many species migrate for reproduction or for developing stages of their life cycle and where many others species are residents.Some marine shrimps, specially Penaeidae family, are examples of temporary estuarine species.During the juvenile phase (post-larvae), they make use of marine currents and occupy these food-rich areas to ensure growth.Marine shrimps Farfantepenaeus subtilis, Litopenaeus schmitti and X. kroyeri spend about four months in estuarine environments (Martinelli 2005) and then return to the sea.
Sexual maturity is usually complete in the sixth month, when the individuals are ready for mating and complete their life cycle (Vinatea 2004).Some species of freshwater shrimps (Macrobrachium) also need to spend part of their life cycle in estuaries, since the larvae develop in brackish water (salinity from 12 to 16) until they reach the juvenile stage, when they initiate the migration to fresh water, where they become adults, sexually mature and ready for mating (SEBRAE/ES 2005).An example of this is Macrobrachium amazonicum (Heller 1862), a species with great plasticity whose populations widely inhabit Amazon estuaries and can also be exclusively a freshwater species.http://dx.doi.org/10.1590/0001-3765201420120039333-345 HEBERT A. SAMPAIO and JUSSARA M. MARTINELLI-LEMOS Intertidal zones are highly variable environments due to the action of tides and nycthemeral and seasonal variations, where inhabitants are subject to temperature and salinity variations, desiccation and hypoxia conditions (Horn et al. 1999).In this environment, tide pools can be formed during low tides, which according to Zander et al. (1999) are still waters dammed in depressions and cavities without direct communication with the sea.Survival conditions become critical in this environment due to increased temperature and salinity (Nybbaken 1997).Tide pools can determine local differences about the diversity of species (Araújo and Feitosa 2003), because they are environments with a complex biodiversity, which are internally subject to intense variations of physicalchemical conditions and biological interactions (Metaxas and Scheibling 1993).
The coastline of the State of Pará is surrounded by islands, bays and estuaries with a wide range of mangroves that correspond to 20% of Brazilian mangroves (Herz 1991).Most studies on decapod crustaceans from the coast of Pará refer to subtidal areas (e.g., Corrêa and Martinelli 2009, Nevis et al. 2009, Bentes et al. 2011, Oliveira et al. 2012, Silva and Martinelli-Lemos 2012, Cavalcante et al. 2012).Most published works related to intertidal areas of the coast of Pará were focused on ichthyofauna (e.g., Giarrizzo andKrumme 2007, 2009) and benthos (e.g., Rosa Filho et al. 2006).
Taking into consideration that tide pools are environments that can be formed in different locations with large variations of physicalchemical conditions, the present work aims to verify the occupation and the correlation of shrimp abundance in relation to environmental variables in different habitats (mangroves, salt marshes and rocky outcrops) found on Areuá Beach, Guarás Island, situated in the Extractive Reserve of Mãe Grande de Curuçá, State of Pará (coast of Brazilian Amazon) in order to identify the importance of the area for the species.

AREA OF STUDY
The study was carried out on the coast of Curuçá situated in the Extractive Reserve of Mãe Grande de Curuçá, Areuá Beach (Geographic coordinates: 35°00' 30.43'' S, 47°51'07.97''W), Guarás Island, northeast of the State of Pará, Brazilian equatorial region (Figure 1).
The temperature in this region is relatively constant, averaging 27 °C with a range of 6 °C.It has abundant rainfall exceeding 2,000 mm per year, with the rainiest season from December to June and the less rainy season from July to November (SEPOF-PA 2011).The Curuçá estuary has a perimeter, length and area of 133 km, 21 km and 200 km 2 , respectively (Giarrizzo and Krumme 2007).The estuary suffers a strong marine influence and is dominated by semi-diurnal macrotides, ranging from 2 to 4 m tall.This tidal influence reaches 12 to 16 km in land (Mácola and El-Robrini 2004).
The vegetation of the area is predominantly composed of mangrove forest and the prevailing species are: Rhizophora mangle followed by Avicennia germinans, in addition to some spots of salt marshes (Spartina spp.).Geologically, the island has Barreiras Group (Tertiary), Post-Barreiras and recent (Quaternary) sediments (Companhia Docas do Pará 2004).SAMPLING Samples were collected in August and November 2009, at low syzygy tide during the day.The collections were made in tide pools from three distinct habitats: rocky outcrops, salt marshes and mangroves (Figure 2), totaling 20 pools: eight in rocky outcrops, six in salt marshes and six in mangroves.Firstly, we recorded physical-chemical factors (pH, salinity and water temperature) in each habitat.The pH was determined using tapes (ColorpHast) with 0 to 14 marks; salinity was obtained in the laboratory with an optical SHRIMPS ON TIDE POOLS IN AN AMAZON ESTUARY refractometer (Atago); and the temperature was measured with a mercury thermometer (max 50 °C) at intervals of 30 minutes.
The area and volume of tide pools were determined using two three-meter rulers marked every 20 cm.The rulers were placed on the edges  of the tide pools for designing the "Cartesian plane" (X and Y), where in the Y axis we measured the depth each 20 cm.Data were entered in the Surfer @ 8.0 program (Golden Software Inc 2002), obtaining the maximum and minimum values of the area, volume and a 3D image of the pool.The shrimps were caught manually with a hand net (150 mm long; 101.6 mm opening; and 3 mm mesh) and packed in 70% alcohol.
We counted and identified the shrimps to the lowest taxonomic level according to Holthuis (1952), Pérez-Farfante (1978), Christoffersen (1984) and Melo (2003) at the laboratory.The weight of each individual was assessed with a precision scale accurate to 0.01 g (Marte).Subsequently, we measured the total length (TL) and the length of the cephalothorax (CL) with a precision digital caliper accurate to 1 mm (Vonder).Ecological descriptors of shrimps registered were: number of individuals, species and families; density (ind/m 3 ); and biomass (g/m 3 ).
The species accumulation curve, also known as collector's curve, was generated through the program PRIMER ® version 5.0 (Clarke and Warwick 1994) in order to verify the sufficiency of sampling, relating the number of species obtained to the increase in the collection effort.The nonparametric estimators of richness used were: Sobs (species observed); Chao1 (number of rare species); Chao2 (data of presence/absence taking into account the distribution of species among samples); Jacknife1 (species that only occurred in one sample); Jackknife2 (number of species occurring in only one sample and the number of species that occurred in exactly two samples); and Bootstrap (data based on the ratio of stands containing each of the species).
To test whether there was significant diffe rence among the medians of the abiotic factors of different habitats (rocky outcrops, salt marshes and mangroves) we performed a non-parametric KruskalWallis test with a significance level of 95% since the assumptions of normality and homoscedasticity were not met, even underwent transformations.An analysis of Spearman linear correlation was performed to verify the correlation of abiotic factors (pH, salinity, water temperature, area, and volume of tide pools) as independent variables, with the density and biomass of shrimps caught, regarded as dependent variables.Both analyses were carried out using the program Statistica ® version 7.0 (Statsoft 2004).

ABIOTIC FACTORS
The median of salinity was 35, with a maximum of 40 in the three habitats in November and a minimum of 31 in the salt marsh in August (Table I).The habitat of salt marsh had the lowest salinity, possibly due to the physical and morphometric characteristics of pools in this environment, which are more extensive and wooded, which reduces the incidence of sunlight, diminishing water evaporation and consequently increasing salinity.There were highly significant differences in the median of this factor among months (H = 30.32;p<0.01) (Figure 3a).Regarding the pH, the median was 9, with a maximum of 10 in August and November in the three habitats (mangrove, salt marsh and rocky outcrop) and a minimum of 7 in the rocky outcrop in August (Table I).There was significant difference of pH in August (H = 7.466; p<0.05) and November (H = 10.99;p<0.01) (Figure 3b).
The median of temperature was 29.23 °C, the maximum value was registered in August in the rocky outcrop (36.6 °C) and the minimum was 25.75 °C in the salt marsh in November (Table I).The values show significant difference of this factor in August (H = 9.95; p<0.01) and among habitats (H = 4.92; p<0.05) (Figure 3c).
The median of the areas measured of tide pools was 12.82 m 2 , with a maximum of 197.2 m 2 in the pool of salt marsh in November and a minimum of 1.91 m 2 in the rocky outcrop in August (Table I).There was no significant difference in the median of this factor (Figure 3d).The median of the volume of tide pools was 1.08 m 3 , of which the largest volume was found in the pool of the rocky outcrop (32.5 m 3 ) and the lowest in the pool of the mangrove (0.02 m 3 ); both values were recorded in November.There was significant difference between the median of the volumes of pools among habitats (H = 6.22; p<0.05) and measured in November (H = 13.51;p<0.01) (Figure 3e).

CARCINO-GROUP
We collected a total of 4,871 shrimps, distributed in three families and four species (Figure 4 and Table II), with Farfantepenaeus subtilis (Pérez-Farfante 1967) representing 98.36% of the total collection, followed by Alpheus pontederiae (Rochebrune 1883), 0.76%; Macrobrachium surinamicum (Holthuis 1948), 0.45%; and Macrobrachium amazonicum (Heller 1862), 0.43%.Palaemonidae was the most specious family and Penaeidae and Alpheidae occurred in all habitats.The species F. subtilis and A. pontederiae occurred in the three habitats, while M. surinamicum occurred in the rocky outcrop and the salt marsh and M. amazonicum only in the salt marsh.All species occurred in the two months with exception of M. amazonicum, which only occurred in August.
Regarding the species collected, two were predominantly freshwater species (M.amazonicum and M. surinamicum), one estuarine (A.pontederiae) and the other marine (F.subtilis).The lowest frequency was obtained in the rocky outcrop (0.78%) and the largest in the salt marsh (96.16%).
The median biomass of shrimps was 0.52 g/m 3 .The habitat with the largest and smallest biomass was the salt marsh, with 2,752.86 g/m 3 and 0.14 g/m 3 , respectively.There was significant difference in the median of the biomass of shrimps in August (H = 11.91;p<0.01) and November (H = 8.70; p<0.05) and among habitats (H = 15.34;p<0.01) (Figure 4 a).With regards to density (ind/m 3 ), the habitats had a median of 3 ind/m 3 , where the largest density (4,533.33ind/m 3 ) and the lowest density (1.44 ind/ m 3 ) both occurred in the salt marsh.Density differed among months (H = 11.04;p<0.01) and among habitats (H = 15.11;p<0.01) (Figure 4b), being higher in the salt marsh in August.
Maximum density (4,443.37ind/m 3 ) and biomass (2,734.75g/m 3 ) of F. subtilis occurred in the salt marsh (Figures 5h and g, respectively).M. amazonicum had the lowest density (17.24 ind/m 3 ) in the salt marsh (Figure 5d) and F. subtilis had the lowest biomass (0.14 g/m 3 ) in the salt marsh (Figure 5g).
The average CL, TL and total weight (TW) was 7.84 ± 2.57 mm, 33.54 ± 10.84 mm, and 0.34 ± 0.36 g, respectively (Tables III and IV).The species with the largest and smallest CL and TW was F. subtilis (Tables III and IV).The largest and smallest TL corresponded to M. amazonicum and A. pontederiae, respectively (Table IV).Shrimps with the smallest CL (2.66 mm) and TL (10 mm) were collected in the mangroves.Bigger and heavier shrimps (CL: 18.03 mm; TL: 77.1 mm; TW: 3.08 g) occurred in the pools of the salt marshes.The estimates of richness showed an interval of 4 for Jacknife1 and 2 and Chao1 and 2 to 4.04 for Bootstrap; indicating that we sampled between 99 to 100% of the species available to the collection method.Most estimators reached asymptote (Figure 6).

RELATIONSHIP BETWEEN ENVIRONMENTAL VARIABLES AND THE CARCINO-GROUP
The environmental variables did not show signifi cant correlation with density, nor with the biomass of shrimp species of mangrove pools (Table V).Among the abiotic factors studied (temperature, salinity, pH, area and volume of tide pools), the pH showed signifi cant negative correlation with the density and biomass of F. subtilis and positive correlation with the density of A. pontederiae in the rocky outcrop (Table VI) and the density of M. amazonicum in the salt marsh (Table VII).
The temperature correlated signifi cantly and negatively with the density and biomass of M. surinamicum and positively with the biomass of F. subtilis in the pools of the rocky outcrop (Table VI).Salinity was the only physical-chemical factor that did not correlate signifi cantly with the abundance of shrimp species (Tables VI and VII).

DISCUSSION
High temperatures in tide pools can be explained by high exposure to the sun during the period of low tides (Macieira 2008, Machado 2013)      can be explained by the lesser extent of these pools and less vegetation cover over them, since there are no shrubs protecting them from solar rays.However, the pools of the salt marshes with similar characteristics had the lowest temperature values.
The occurrence of predominantly freshwater species (M.amazonicum and M. surinamicum), estuarine (A.pontederiae) and marine (F.subtilis) in the salt marsh (96.16%) reflects the extreme ecological importance of this habitat, raising  questions for future management plans.This result is consistent with other studies that found high numbers of crustaceans in salt marshes, such as works by Kneib and Wagner (1994), which studied nektons in the Duplin River on the west side of Sapelo Island, Georgia, United States; Rozas and Minello (1998) who through a comparative study found that in an estuary in the south of Texas, United States, the total density of crustaceans was greater in vegetated habitats (salt marshes and marine angiosperms) than in non-vegetated habitats (shallow waters, <1 m depth); and Minello and Zimmerman (1992) who investigated the use of transplanted and natural salt marshes habitats by fish and crustaceans on the coast of Texas, United States, suggesting that density is correlated with the density of prey.
Farfantepenaeus subtilis represents a high ecological contribution to the aquatic food web, in addition to being one of the major resources    Guiana (Isaac et al. 1992).The northern limit of distribution of this species is represented by Antilles (Central America), with continuous occurrances in Cabo Frio, Rio de Janeiro (South America) as well.
The predominance and the high density of F. subtilis in salt marshes prove that they are important areas for recruitment of juveniles, corroborated by the presence of small sized shrimps.The predominantly freshwater species M. amazonicum is the main commercially exploited shrimp in estuaries and continental waters of the states of Pará and Amapá, by artisanal fisheries, where there is a significant market (OdinetzCollart 1987, Bentes et al. 2011).According to Brito and Furtado Júnior (2002), the Municipality of Curuçá in northeastern Pará, represents one of the most important marine and estuarine shrimp landing sites, where the average annual production was of approximately 30 tons between the years of 1997 and 2000.
Temperature and pH were the factors that most influenced the variation in the density and biomass of shrimps in tide pools of Areuá Beach, northeastern Pará, in the Brazilian equatorial region.The large temperature variation can lead these organisms to osmotic stress conditions being decisive in their occupation, in spite of the fact that crustaceans and estuarine organisms are eurythermal (Garcia and Le Reste 1987).However, this result differs from that obtained by Masunari et al. (1998) who described that the temporal density variation of most Decapoda species of an intertidal area on Farol Island, coast of Paraná (Brazilian subtropical region) did not show any correlation with temperature variation.However, this comparison should be interpreted with caution, since the environment of tide pools suffers greater amplitude of physical-chemical factors variation than intertidal zones.
Salinity was the only physical-chemical factor which was not significantly correlated with the abundance of shrimp species; however, according to Teixeira and Sá (1998), salinity appears to be the most important factor that acts as a regulator of the distribution and abundance of macrocrustaceans in estuarine complexes, even with highly significant difference in the median of this factor among months.
Most shrimps were captured in the juvenile phase, suggesting that tide pools are used as nurseries, refuge areas and shelters for shrimp species inhabiting the estuarine region, a fact seldom in scientific literature.Tide pools are environments that offer favorable resources for shrimps' development, being crucial for the maintenance and survival of this group.

Figure 1 -
Figure 1 -Location of the study area on Areuá Island, Curuçá, State of Pará, Brazilian Amazon.

Figure 2 -
Figure 2 -Location of the three sampled habitats: rocky outcrop, salt marsh and mangrove on Guarás Island, Areuá Beach, Municipality of Curuçá, Brazilian Amazon.

TABLE I Descriptive statistic of abiotic factors of tide pools on Guarás Island, Curuçá, State of Pará (Me = Mean; Md = Median; Min = Minimum; Max = Maximum; ST = Standard Deviation; CI = Confidence Interval; and SE = Standard Error).
SHRIMPS ON TIDE POOLS IN AN AMAZON ESTUARY . In the rocky outcrops, the highest temperatures recorded

TABLE II Composition of shrimp species on Areuá Beach, Guarás Island, Curuçá, State of Pará, Brazilian Amazon.
SHRIMPS ON TIDE POOLS IN AN AMAZON ESTUARY

TABLE V Spearman's correlation coefficient among abiotic factors (independent variables), density and biomass (dependent variables) of shrimp species from mangrove pools on Areuá Beach, Guarás Island, Curuçá, State of Pará, Brazilian Amazon
. SHRIMPS ON TIDE POOLS IN AN AMAZON ESTUARY for industrial fishing in the northern Brazilian coast, where its capture is performed in one of the most important shrimp fishing grounds of the world, which extends from Tutóia, in the State of Maranhão, to the border of Brazil and French