Abundance and spatio-temporal distribution of the amphidromous shrimp Macrobrachium olfersii (Caridea: Palaemonidae) along the Ribeira de Iguape River (São Paulo, Brazil)

In this study we investigated the distribution of Macrobrachium olfersii (Wiegmann, 1836) along ~150 km of the Ribeira de Iguape river, São Paulo, Brazil. We compared the abundance and spatio-temporal distribution, and checked for differences in size and proportion of each sex in the collections, using two sampling methods. Shrimps were collected monthly at four sites (Eldorado, Sete Barros, Registro, and Iguape), from January to December 2007, using traps and sieves. We obtained a total of 23,818 individuals. The abundance was significantly higher at the Iguape-site, which was the closest to the estuary. There was a positive cross-correlation between abundance and rainfall, indicating an increase in abundance with a decrease in rainfall. The body size increased significantly upstream, suggesting a juvenile upstream migration, controlled by the rainfall regime and the amphidromous behavior of M. olfersii . More than 95% of the individuals were captured by sieving through the marginal vegetation of the river. The average size and sex ratio of each sample varied depending on the capture method: traps captured more and larger males than the sieve. Therefore, we recommend the combined use of these methods to obtain a better coverage of the population biology of freshwater shrimps.

Shrimps of the genus Macrobrachium Spence Bate, 1868 can be separated into two groups, depending on the environment where they spend most of their life cycle. The first group is composed of freshwater species that have an abbreviated or direct larval development (Magalhães and Walker, 1988;Jalihal et al., 1993;Hancock, 1998). The second group has amphidromous species that migrate twice during their life cycle: first as planktonic larvae that float passively towards an estuary, and later as juveniles that migrate upstream in rivers and creeks to live there as adults (Bauer and Delahoussaye, 2008;Bauer, 2011).
The Ribeira de Iguape river catchment area has a high species richness of decapod crustaceans, including five shrimp species of the genus Macrobrachium (see Rocha and Bueno, 2004). Four of them are amphidromous (Macrobrachium acanthurus (Wiegmann, 1836), Macrobrachium carcinus (Linnaeus, 1758), Macrobrachium heterochirus (Wiegmann, 1836), and Macrobrachium olfersii (Wiegmann, 1836), and the other one lives exclusively in freshwater (Macrobrachium potiuna (Müller, 1880)). Macrobrachium olfersii, popularly known as "Pitu", "Bristled River Shrimp" or "Buchura River Prawn" (Rossi et al., 2016), has a wide geographic distribution, occurring from the southeast of the United States to the south of Brazil (Holthuis and Provenzano, 1970;Melo, 2003;Rossi and Mantelatto, 2013). Due to the physiological need for brackish water, it is restricted to coastal basins and to the lower courses of rivers and streams that flow directly into the sea (Holthuis and Provenzano, 1970). In these environments, M. olfersii is found in the marginal vegetation and in crevices between trunks and rocks (Barros, 1995;Melo, 2003).
Amphidromous shrimps can undertake upstream migrations of hundreds of kilometers (Rome et al., 2009). In many cases, the shrimp's distribution is influenced by the degree of conservation of the environments found along the migration course (Snyder et al., 2013). During the migration, they can be exposed to stressful conditions due to: 1) the suppression of riparian forest and siltation of rivers and streams, both which lead to loss of microhabitats (Iwata et al., 2003;Teixeira and Couto, 2012); 2) fisheries, that can lead to declining populations (Hein et al., 2011;Snyder et al., 2013); and 3) dams, that can change and/or disrupt species migration routes (March et al., 2003;Greathouse et al., 2006). Some of these anthropogenic impacts have been observed in the catchment area of Ribeira de Iguape river, namely fisheries (Bertini and Valenti, 2010), suppression of riparian vegetation (Leonardo et al., 2008;Iori et al., 2019) and contamination by pesticides, especially carbofuran, which is one of the most commonly used agrochemicals employed in the banana plantations in the region (Marques et al., 2007a;2007b).
The influence of environmental degradation in the abundance and distribution of M. olfersii has been recorded by several authors. In the Mangues river, south of Bahia (Brazil), Teixeira and Couto (2012) noticed that the shrimps were absent in places lacking the riparian vegetation and where the riverbed was altered. Pescinelli et al. (2016) also attributed the disappearance of M. olfersii from the Taquaral river (Ubatuba, Brazil), over a five-year interval, to anthropogenic impacts. In Sarapiquí (Costa Rica), Snyder et al. (2013) compared the relative abundance of M. olfersii over a 20-year gap between the predisturbance and post-disturbance and observed strong changes in the abundance in the places affected by the anthropogenic impacts.
The distribution of shrimps in freshwater environments is also driven by environmental factors. In the case of M. olfersii the main known factors driving the populations are temperature (Müller and Prazeres, 1992), conductivity, river flow, and substrate type (Snyder et al., 2016). In addition, the distribution of shrimps in a water body is also related to resources such as organic detritus, phytoplankton and zooplankton, which are used by different species during their life cycles (Hobbs and Hart, 1982;Teixeira and Sá, 1998;Müller et al., 1999;Almeida et al., 2008).
Previous studies about the distribution patterns and ecology of M. olfersii addressed only populations from relatively small rivers (Müller and Prazeres, 1992;Ammar et al., 2001;Mossolin and Bueno, 2002;Teixeira and Couto, 2012;Pescinelli et al., 2016). Thus, in this study we investigated the distribution of M. olfersii in the Ribeira de Iguape river, covering four sites along an ~150 km transect from the estuary. We analyzed the abundance and spatio-temporal distribution using two capture methods. Due to Macrobrachium olfersii in Ribeira de Iguape River Nauplius, 28: e2020017 the high habitat heterogeneity found in freshwater environments, there are many different methods that can be used to capture shrimps, such as traps, sieves, nets, electrofishing, and manual capture (Abele and Blum, 1977;Fièvet et al., 1996;Garcia et al., 2003;Rocha and Bueno, 2004;Rocha, 2010;Copatti et al., 2016). We also analyzed differences in the size and proportion of each sex in the samples, depending on the capture method. We expected to find a similar abundance across all sampling sites, since there are no physical barriers that could hamper the migration, and the degree of anthropogenic impact is similar along the whole river extension.

Study Area
The region of Vale do Ribeira is located south of the state of São Paulo and east of the state of Paraná. This region harbors an important remnant of the Atlantic Forest. About 40% of the vegetation cover of this remnant belongs to São Paulo State and was appointed as the largest continuous reserve inside the national territory (Silva Matos and Bovi, 2002). The Ribeira de Iguape river catchment area (23°50'S 46°50'W; 25°30'S 50°00'W) is approximately 25,000 km 2 and is one of the largest on the Brazilian coast (Mahiques et al., 2014). According to Köppen's climate classification, the climate is humid subtropical, and the average annual temperature is 21°C (Rolim et al., 2007). The average rainfall is 2200 mm yr -1 and there is a marked rainy season between November and February (Mahiques et al., 2014). The river outflow, in its lower course, varies from ~300 to more than 1200 m 3 s -1 , and this fluctuation is strongly influenced by the humid subtropical climate (Mahiques et al., 2014).
The riverbed of Ribeira de Iguape is rocky along the rapids, and sandy/loamy in the backwaters. Depth varies with rainfall in the range of 1−8 meters (DAEE: http://www.hidrologia.daee.sp.gov.br/). The most conspicuous problem is that the riparian vegetation has been replaced by banana plantations and pastures along almost all the river's extension (Leonardo et al., 2008). These land-use changes introduced agrochemicals into the water (Marques et al., 2007b), and the suppression of vegetation plus the extraction of sand has led to siltation (Lelles et al., 2005).

Data Collection
Samples of M. olfersii were taken monthly, from January to December 2007, at four sites distributed along the river. These sites were selected to characterize the migration and abundance of M. olfersii along the river's extension. The site Iguape was located 72 km from the estuary, and the remaining sites were located upstream: Registro (114 km), Sete Barras (138 km) and Eldorado (155 km) (Fig. 1). The river margins were similar in these four sites, with pastures and banana plantations. In the river margins the presence of grasses was common (Brachiaria subquadripara (Trin.) Hitchc, Brachiara platyphylla Sukachev, and Panicum dichotomiflorum Michx) and in Iguape there was a high amount of aquatic plants (floating water hyacinth, Eichhornia crassipes (Mart.) Solms). Between Eldorado and Iguape, the river receives two tributaries, the Etá river, near Sete Barras, and Juquiá river, the main tributary, near Registro ( Fig. 1).
In each site we took samples from two places, 2 km apart from each other, using a combination of minnow traps and a sieve (Fig. 2). Twelve minnow traps (1 m length, 30 cm diameter, and 8 mm mesh size) were baited with small banana pieces and/or cow bones and installed in each site. They were spread among the marginal vegetation (n = 6 traps) and in the river bottom (n = 6 traps) and retrieved after 24 h. In addition, two people sampled the river margin (against the current) for 20 min with a single sieve (0.5 m 2 , 5 mm mesh size), focusing on the semisubmerged vegetation where shrimps are known to shelter . The collected shrimps were placed in thermal boxes with crushed ice and transported to the laboratory, where they were kept frozen until the analysis. All specimens were identified according to Melo (2003).
In each site we measured the water temperature and salinity with a thermometer and a refractometer, respectively. Rainfall data for each region for 2007 were obtained from the Integrated Agrometeorological Information Center (CIIAGRO) of the Agronomic Institute (IAC).

Macrobrachium olfersii in Ribeira de Iguape River
Nauplius, 28: e2020017 To analyze the shrimp size, we did a subsampling of the total amount of M. olfersii collected per site and month. In samples containing up to 80 shrimps, all individuals were analyzed. If samples had more than 80 shrimps, a subsample was randomly taken, by adapting the procedures of Wenner et al. (1991) and , as follows: 80 individuals were selected from samples containing 80−160 shrimps; 50% of the individuals were taken from samples with 160−320; 25% were taken from samples with 320−500, and 10% were taken from samples with more than 500 shrimps.
All individuals of M. olfersii from each subsample were sexed based on the examination of secondary sexual characters. Males were recognized by the presence of the appendix masculina on the endopodite of the second pair of pleopods, and females by the absence of it. The carapace length (CL) was taken as the distance between the orbital angle and the posterior margin of the carapace. In individuals > 10.0 mm the CL was measured with a caliper (to the nearest 0.1 mm). Individuals < 10.0 mm CL were measured to the nearest 0.025 mm under the stereomicroscope using a graduated ocular micrometer (Olympus, SZX7®). After evaluating all individuals, we concluded that the size of the smallest male bearing an appendix masculina was 4 mm. Thus, all shrimps smaller than 4 mm were considered as juveniles, following Pescinelli et al. (2016).

Statistical analysis
To analyze the distribution and abundance of M. olfersii, we recorded the total number of shrimps per site, month, and capture method (trap or sieve). The abundance was compared between seasons, as follows: Summer (January to March), Autumn (April to June), Winter (July to September) and Spring (October to December).
To verify the possible relationship between the variation in total abundance per month and the monthly average water temperature and rainfall in the region, a time series cross-correlation analysis was used. Six time-intervals were considered (-6 to 6), and the significant values in the correlogram were used to verify the magnitude and lag in the covariance between the time series (Chatfield, 2004).
To compare the total abundance of shrimps between sites and seasons a generalized linear model (GLM) with gamma distribution and log-link function was fitted to the data (Venables and Dichmont, 2004). This model was chosen among others because it had the lowest Akaike Information Criterion (AIC) (Akaike, 1974). The same GLM was used to check for differences in size between the sexes and capture methods (trap or sieve).
The body size of M. olfersii was compared between sites and months using the non-parametric Kruskal-Wallis test, followed by Dunn's multiple comparison test. In addition, the body size was also compared between the two capture methods using the Mann-Whitney test (Sokal and Rohlf, 2001). Pearson's chisquared test with residue analysis was used to check for biases in the sex ratio, depending on the capture method (trap and sieve) (Sokal and Rohlf, 2001). All statistical analyses took into account a significance level of 95%.

Environmental factors
Salinity was zero (0‰) at all sites and there was no tidal influence, even at Iguape, which was close to the estuary. Mean water temperature in the four sites ranged from 17.9 ± 0.18°C (July; Iguape) to 30.6 ± 0.07°C (March; Sete Barras) (Fig. 3A). The lowest rainfall (10.6 ± 7.06 mm) was recorded in June, and the highest, in January (251.9 ± 52.59 mm) (Fig. 3B). In July, a high volume of rainfall was observed, with a mean of 151.2 ± 25.83 mm, which is considered atypical for the region (Fig. 3B).

Spatial and temporal distribution
The total number of M. olfersii individuals captured in the four sites along the Ribeira de Iguape river was 23,818, corresponding to 1,594 specimens from Eldorado, 2,333 from Sete Barras, 2,201 from Registro, and 17,690 from Iguape. The GLM indicated a significant difference in the total abundance between sites and seasons (p<0.05), but there was no interaction between these factors (Fig. 4A, B). The abundance at Iguape (closest to the estuary) was statistically higher than in the other regions, farther from the river mouth (Eldorado, Sete Barras and Registro) (Post hoc LSD test; p<0.05) (Fig. 4A). The highest abundance was recorded in autumn (Post hoc LSD test; p<0.05) (Fig. 4B).

Macrobrachium olfersii in Ribeira de Iguape River
Nauplius, 28: e2020017 Regarding the relationship between abundance, temperature and rainfall (Fig. 5A, B), the crosscorrelation indicated a relationship with rainfall, demonstrating that, when rainfall decreases the number of individuals increased (3 lag time; r = 0.659; p<0.05).

Individual size and capture methods
A high disparity was observed in relation to the two capture methods: 3.1% of the total number of individuals was captured in traps (total = 741 individuals; Eldorado = 205; Sete Barras = 172; Registro = 200; and Iguape = 164), whereas 96.9% (23,077 shrimps) was obtained by sieving (Eldorado = 1,389 ind.; Sete Barras = 2,161 ind.; Registro = 2,001 ind.; and Iguape = 17,526 ind.). Due to the subsampling, the size of 9,691 individuals from the four sites were measured and analyzed (8,962 captured by sieving and 729 captured by the traps) (Tab. 1). The overall size of M. olfersii differed statistically between sites and increased towards the farthest site (Kruskal-Wallis, H = 2,051; p<0.05) (Fig. 6). Between November and February, i.e., from the end of spring until the end of summer, individuals were statistically larger than in the other months (Kruskal-Wallis, H = 2,684; p<0.05) (Fig. 7).
The GLM revealed that the shrimp size differed between the capture methods. Also, there was a significant interaction between the factors "capture methods" and "sex" (p<0.05). There were no differences in size when shrimps were captured by sieving (Post hoc Sidak test; p>0.05); in traps, however,

Macrobrachium olfersii in Ribeira de Iguape River
Nauplius, 28: e2020017 the males were larger than the females (Post hoc Sidak test; p<0.05). The between-sex comparison indicated that traps selected for largest individuals, irrespective of sex (Post hoc Sidak test; p<0.05) (Fig. 8). The overall size of individuals captured by the traps (16.4 ± 2.69 mm CL) was statistically higher than those from the sieve (4.2 ± 2.36 mm CL) (Mann-Whitney U = 37,530; p<0.05).
The statistical analysis of the contingency table indicated that the sex ratio depended on the capture method (Pearson Chi-Square = 45,817; p<0.001).
In the traps, the number of males was higher than expected (observed: 498, expected: 419), and the opposite happened with the females (observed: 231; expected: 310). In the sieve, the number of males (1,248) was lower than expected (1,327), and the number of females (1,059) was higher than expected (980). Thus, the ratio of males:females was higher in the traps.   disCussion Macrobrachium olfersii was extremely abundant in the Ribeira de Iguape river. It was found throughout all the studied sites, including Eldorado, which is ~155 km upstream the river mouth. Within the Vale do Ribeira, M. olfersii was also found in small rivers of the Juréia-Itatins Ecological Station, Campina do Encantado State Park, Cardoso Island State Park, Cananéia, and in Comprida Island, all located within the state of São Paulo (Rocha and Bueno, 2004). Furthermore, Rocha and Bueno (2004) reported the existence of M. olfersii specimens from Eldorado (MZUSP 13568) and from Iporanga (MZUSP 13549) (which is a city located ~80 km upstream of Eldorado), and deposited in the "Museu de Zoologia da Universidade de São Paulo" (MZUSP). They concluded that these populations established there through the migration of adults, but that their larvae are unlikely to contribute to the recruitment process due to the long distance to the estuary, which is needed to complete the life cycle.
Our results on the spatial distribution of M. olfersii indicate that they are more abundant in Iguape, near the river mouth. In Eldorado we found juveniles and adults, suggesting that M. olfersii individuals are capable of migrating upstream and reaching this region as juveniles, after going through the long passive migration towards the estuary during the larval phase.  observed the same migration pattern in the sympatric species M. acanthurus (in the same study area), indicating that these species are capable to undertake a long migration and establish far away from the estuary. The larval migration back to the estuary is possible because the reproductive peak of M. acanthurus takes place in summer, when rainfall is abundant and increases the current's speed. Under these conditions, the larvae from Eldorado can reach the estuary in the region of Iguape within ~37 h of passive transport.
The temporal f luctuation observed in the abundance of M. olfersii can also be explained by its amphidromous behavior. In summer, when the precipitation was high, there was a decrease in abundance and the population was composed of large individuals. During the low precipitation recorded in autumn, there was a peak in abundance, and the population was composed of smaller individuals, indicating an upstream migration of juveniles. This relationship between reproduction and recruitment and the rainy period was also observed in the northern coast of São Paulo, in M. olfersii (cf. Mossolin and Bueno, 2002;Pescinelli et al., 2016) and in M. acanthurus (cf. . The migration of M. olfersii during the early juvenile stages was previously observed in Tacarigua, a coastal lagoon in eastern Venezuela, where it occurred year-round but was more intense in January−February (Gamba and Rodríguez, 1987). There, late post larvae or early juveniles of M. olfersii form visible parallel bands at the margins of the canal during migration (Gamba and Rodríguez, 1987). In Brazil, Pompeu et al. (2006) observed the mass migration of juvenile M. carcinus at dams of the Mucuri river (Minas Gerais state) using the mechanisms built for fish transposition. This mass migration is related to the reproductive strategy of this species. Since reproduction occurs in the upper river and adults do not need to reach the coast, downstream larval drift and upstream migration of juveniles are the only migratory movements necessary for the maintenance of this species in this basin (Pompeu et al., 2006). Upstream migrations of juvenile amphidromous shrimps were also observed in many other locations around the world, such as India (Ibrahim, 1962), Caribbean and other tropical islands (Holmquist The boxplots show the median: 75% and 25% quartiles; the whiskers indicate the non-outlier range; GLM with a posteriori Sidak test: capital letters compare the sexes by capture method, whereas small caps indicate the comparison of capture methods by sex at p<0.05. Nauplius, 28: e2020017 et al., 1998;March et al., 2003;Covich et al., 2006), Japan (Hamano and Hayashi, 1992), Australia (Lee and Fielder, 1979;Novak et al., 2017a;2017b), United States (Bauer and Delahoussaye, 2008), Mexican Pacific coast (Rodríguez-Uribe et al., 2014), and Costa Rica (Snyder et al., 2016).

Macrobrachium olfersii in Ribeira de Iguape River
Despite the intense anthropogenic impacts observed along the Ribeira de Iguape river (Marques et al., 2007a;2007b;Leonardo et al., 2008), which included water contamination and suppression of the riparian vegetation, the upstream migration of M. olfersii still occurred. This statement is based on observations made during other collection campaigns performed at the same study sites from 2017 to 2019. In other localities of Brazil and Costa Rica, however, the occurrence and abundance of M. olfersii have been negatively affected by the lack of riparian vegetation, morphological changes in the riverbed, contamination by pesticides, and shrimp fisheries (Teixeira and Couto, 2012;Snyder et al., 2013;Pescinelli et al., 2016). In these localities, the rivers studied were small. The Ribeira de Iguape river, on the other hand, is large and has a higher carrying capacity. Its large dimensions may have contributed to minimize the negative effects of the anthropogenic impacts on the shrimps, allowing the maintenance of the population along the river course.
Changes in the riparian vegetation can increase the erosion of stream banks, sedimentation, and water flow, which in turn decrease the habitat heterogeneity and affect the abundance and diversity of decapods (Pérez-Reyes et al., 2016). Nonetheless, we can infer that the studied population of M. olfersii is relatively resistant and/or resilient to the disturbances that affect the Ribeira de Iguape river, which are constant and have occurred for a long time in the region. The same can be said about M. acanthurus, which is also abundant in the studied region . When a species is resistant its abundance stays relatively stable upon disturbances, whereas in a resilient species the abundance decreases but can recover fast to the original, pre-disturbance levels (Holling, 1973). The resistance and/or resilience of the M. olfersii population and its upstream migration along the 155 km of the Ribeira de Iguape river may be explained by the presence of grasses and macrophytes along the river margins, which are used as shelters by the juveniles, during the upstream migration.
The marginal vegetation and macrophytes can help to minimize the anthropogenic impacts, since they are important resources to many vertebrate and invertebrate species, including freshwater shrimps. This vegetation constitutes refuge against predators (West and Williams, 1986;Rozas and Odum, 1988;Castellanos and Rozas, 2001;Almeida et al., 2008;Beatty et al., 2011;Soares et al., 2015) and places for reproduction and feeding (Montoya, 2003). Species of the genus Macrobrachium are omnivores (Abele and Blum, 1977;Odinetz-Collard, 1988) and can feed directly on the roots of water hyacinths (Green et al., 1976). The periphytic algae (mainly diatoms and Cyanophyceae) attached to the roots likely enhance the quality of this nutritional resource (Montoya, 2003). Thus, M. olfersii may benefit from the presence of this marginal vegetation and use it as shelter and food along its upstream migration.
According to Snyder et al. (2016), the long-distance (>100 km) upstream migratory behavior of freshwater shrimps like M. olfersii has a great influence on their distribution and abundance patterns. It is driven by factors that operate at local (substrate and discharge) and landscape (conductivity) scales. Thus, depending on the characteristics of the region and river from which the samples are taken, abiotic factors and/or anthropogenic impacts may have a great influence on the abundance and distribution. However, the reproductive strategy (amphidromous behavior), the rainfall regime, and the presence of marginal and submerged vegetation can also be considered as key factors. In this context, Cantrell (1985) and Montoya (2003) mentioned also the important role of drifting or windblown vegetation in spreading invertebrates and shrimps from one part of a water body to another.
Our comparison of sampling methods revealed that sieves were more effective and captured more than 95% of the total individuals. Moreover, the body size varied with the method as sieves captured smaller individuals compared to traps. Thus, the use of a single method would affect the estimation of mean and maximum size of the individuals. Although comparative studies like ours with freshwater shrimps are scarce, previous studies reported similar patterns. For example, Rocha Macrobrachium olfersii in Ribeira de Iguape River Nauplius, 28: e2020017 (2010) mentioned this issue when studying M. olfersii and M. acanthurus in a river within the Juréia-Itatins Ecological Station (in the same basin of Ribeira de Iguape). Differences in body size depending on the sampling methods have also been observed in other decapods, such as aeglids (Dalosto et al., 2014;Copatti et al., 2016).
Our results indicated that the capture method is also biased regarding the sex, since traps captured more and larger M. olfersii males compared to the sieve. Traps captured 315 males larger than 16 mm CL, whereas the sieve captured only five. Large M. olfersii males inhabit the bottom of the river, and are known to be territorialists and have well-documented aggressive behavior (see Karplus and Barki, 2019), and seem to be attracted by the baited traps more efficiently. They live in deeper places, among rocks and large tree trunks and come out at night to feed, and then are caught in the traps. On the other hand, females and juveniles inhabit the marginal vegetation where they find shelter and food (Pyron et al., 1999;Montoya, 2003;Almeida et al., 2008), and are more likely to be caught by sieving. Similar behaviors that can lead to a higher capture of males by traps, have also been demonstrated in the mud shrimp Neotrypaea uncinata (H. Milne Edwards, 1837) (Garcia et al., 2003), in the crayfish Cherax cainii (Beatty et al., 2011) and many aeglids (Bueno et al., 2007;Teodósio and Masunari, 2009;Grabowski et al., 2013;Dalosto et al., 2014;Copatti et al., 2016).
According to Rocha (2010), the sex ratio of M. acanthurus changed depending on the sampling method: sieves captured more females, while traps captured more males, as demonstrated with M. olfersii in the present study. Therefore, the sampling method influences the composition of the collected shrimps, because if the biased sex ratio was a true characteristic of the population, it would have been detected by both methods (Rocha, 2010). Thus, the combined use of both methods helps to cover all microhabitats occupied by the different demographic groups, as discussed by Copatti et al. (2016) for Aegla georginae Santos and Jara, 2013.
In conclusion, M. olfersii shrimps are very abundant in Ribeira de Iguape river, where they undertake a long-distance upstream migration. This species can be considered resistant and/or resilient to the anthropogenic impacts that have been affecting the region. The long-distance upstream migration can be accomplished due to the amphidromous behavior, rainfall regime of the region, and presence of grasses and macrophytes by the river margins. The two sampling methods are selective and lead to samples with different sex ratio and body size. Thus, the best way of studying the population biology of freshwater shrimps is to combine both methods (trap and sieve). The combination of methods can provide a more accurate overview of the population structure of these shrimps.

aCKnowledgeMenTs
We are grateful to the FAPESP (Fundação de Amparo à Pesquisa do Estado de São Paulo) for providing financial support (05/56131-0). We are also thankful to students and technicians for their help during fieldwork and Dr. Luis Carlos Ferreira de Almeida for his help with the statistical analysis. Sampling was conducted according to the São Paulo state and Brazilian federal laws.