Reproductive cycle of Mytella guyanensis ( Lamarck , 1819 ) in a Marine Reserve ( RESEX Bay of Iguape ) , Bahia , Brazil

Mytella guyanensis, consumed and commercialized in coastal regions of Brazil, is one of several bivalve species of socioeconomic interest for coastal communities. Besides serving as a source of income and subsistence for these communities, it also contributes to their food security as it is a source of proteins and micronutrients. Thus, the reproductive cycle of this species was studied aiming to contribute to food security and its preservation. Samples were collected monthly, between March 2014 and March 2015, in a natural stock (12°38’50”S; 38°51’43”W) in a Marine Reserve (RESEX Bay of Iguape) (community Engenho da Ponte), Bahia, Brazil. Mytella guyanensis is collected by women on site, where the artisanal fishing of this resource is performed without following any specific handling procedure. Also, empirical evidence indicates overexploitation. The specimens collected were measured along the anterior-posterior axis (length), and after macroscopic analysis they were fixed in Davidson solution, processed by routine histology techniques and stained with Harris haematoxylin and eosin (H&E). The macroscopic analysis showed sexual dimorphism, with the male and female gonads presenting a milky-white and orange colour, respectively. A 1:1 sex ratio (M: F) was observed and reproduction of the species was continuous all year round. March, April, July and August were the months with highest values of gamete elimination. We suggest that a M. guyanensis management plan should restrict capture during these months, in order to sustainably regulate exploitation of this food resource in this reserve.


Introduction
Studies on the reproductive biology of economically important aquatic species are essential to comprehend its life history (Garner et al., 1999) and are important in management actions.Studies supply information about the period of gamete release, sex ratio, and reproductive potential, subsidizing actions to maintain natural stocks and the preservation of the species (Araújo, 2001).According to Carranza et al. (2009) studies report declining stocks of bivalves across Brazilian's Northeast coast, mainly due to overexploitation and the loss of habitat of species.
The Mytilidae family (Mollusca: Bivalvia) includes various species with socioeconomic interest, being a source of income and livelihood for coastal communities (Nishida and Leonel, 1995;Carranza et al., 2009), which contributes to their food security, because they are sources of proteins as well as micronutrients, such as vitamin A, Iron, and Zinc (Santos et al., 2014).Furthermore, mussels may sometimes be the only affordable source of protein available.One of these species is Mytella guyanensis (Lamarck, 1819), popularly known as "bacucu", "bico de ouro", and "sururu" (Rios, 1994;Pinto and Boehs, 2008).This mytilid is distributed from Mexico to Peru, on the Pacific Ocean, and from Venezuela to Brazil, on the Atlantic (Rios, 2009), inhabiting intertidal regions of mangroves where it uses the byssus to attach to underground roots.It is euryhaline, i.e., it resists wide variation in salinity and is dioecious (Sibaja, 1986).
The Reserva Extrativista Marinha Baía de Iguape (RESEX), a marine reserve, and as such to which is prioritized the sustained exploitation of resources within the limits that do not compromise environmental sustainability (Brasil, 2000), was created in August 11 th , 2000, it has an area of 10.082.45 ha, and is located in the municipalities of Maragogipe and Cachoeira, in the region of Bahia's Recôncavo, located 111 km from the capital of the state of Bahia, Salvador.In the community Engenho da Ponte, north of the Reserve, women of the community collect M. guyanensis for sale and as food for their own families.However, in spite of being an area of environmental protection, the species is collected without any control by the local population, i.e., the artisanal fishing of this resource is done without following any specific handling procedure (Prost, 2010).Thus, empirical evidence demonstrates that the overexploitation combined with easily accessible natural stocks, are the main factors that affect this species in this Reserve.
Considering the economic importance represented by M. guyanensis for the communities from this RESEX, this study aimed to determine the reproductive cycle of the species in this marine reserve, with a view to contributing to food security of the communities living there, and to preserve it as a resource.

Material and Methods
Samples of M. guyanensis were collected from a natural stock from the RESEX Bay of Iguape at the coordinates 12°38'50"S and 38°51'43"W, in the community Engenho da Ponte, Cachoeira, Bahia, Brazil.Every month, between March 2014 and March 2015, 30 adult mussels, with length above 40 mm (Sibaja, 1986), were collected manually during periods of low spring tide.The samples were placed in buckets and transported to the Laboratory of Science and Technology in Health of the Federal University of Recôncavo da Bahia (Santo Antônio de Jesus, Bahia), for measurement, weighing, and histological processing.On each sampling occasion, water temperature and salinity were measured in situ, with the assistance of a multiparameter Water Quality Meter AK88 (Akso  ) (N = 12).Accumulated rainfall over the last 30 days, in mm, were obtained from the National Institute for Space Research (INPE, 2014) for each sampling occasion.Identification of the tidal stage was made consulting the Brazilian Navy Tide Table of the Directorate of Hydrography and Navigation (Brasil, 2015).
The specimens were measured for shell length (anterior-posterior axis), with the usage of a Vernier calliper with a precision of 0.01 mm and weighed with a precision of 0.001g, and afterwards were opened and examined macroscopically for colour and gonad texture, as well as for the identification of macroscopic parasites or signs of disease.Immediately afterwards, specimens were fixed in Davidson's solution (Shaw and Battle, 1957).After a 24-hour period, the specimens were transferred to ethanol 70%, processed according to classic histological technique, with dehydration in a series of increasing alcohol concentrations, diafanization in xylene and embedding in paraffin.Thin-sections of 5 µm were stained with Harris's haematoxylin and eosin (H&E), followed by the preparation of slides.The histological cuts were examined using an optical microscope (Leica DM 500  ), through which sex and the gonadal maturation stages were determined, based on previous studies (Araújo, 2001;Carpes-Paternoster, 2003;Gomes et al., 2009;Luz and Boehs, 2011).Representative phases of each stage of the reproductive cycle were photographed with a camera attached to the microscope.
Temporal variation in the sex ratio was estimated by the absolute frequency of male and female mussels, verifying the hypothesis of a difference in sex ratio from 1:1 across time through the application of the χ 2 test.The Kruskal-Wallis test was used to indicate possible differences in gamete release during the period of this study (significance limit: α = 0.05).The Kolmogorov-Sminornoff test was used to verify the normality of the environmental data distribution (water temperature and salinity, and rainfall), and the Spearman coefficient was used to establish possible correlations between gamete release and water temperature and salinity, and rainfall.All statistical analyses were carried out using the software SPSS v. 17.All the mussels sampled in this study were collected under licence according to the current legislation (ICMBio-42081-1).

Environmental data
The average water temperature was 29.7 °C (± 2.45 SD), with the highest temperature (33.6 °C) in February 2015 and the lowest (25.5 °C) in April 2014 (Figure 1).The average salinity was 27.2 PSU (Practical Salinity Unit) (± 3.55 SD).The lowest value (21.1 PSU) was registered in August 2014 and the highest (34.4 PSU) in November 2014 (Figure 2).The average rainfall was 101.17 mm, with the lowest values in November (37.1 mm) and December 2014 (48.3 mm) and the highest in June (159.26mm) and August 2014 (209.03mm) (Figure 3).The low salinity in August coincided with the rainiest month in this period, as the highest salinity (November 2014) coincided with the driest period.

Morphometry
The specimens measured between 41.1 and 68.6 mm (average = 50.64mm ± 4.98 SD, n = 363).The macroscopic analyses showed evidence of sexual dimorphism, with 1 2 the male gonad presenting a milky-white colour and the female gonad an orange colour.The percentage of males (M) was 47.93% and the females (F) was 49.31%, resulting in a sex ratio (M: F) of 1:1 (χ2 = 16,757; p>0.05).The month in which the greatest variation of the sex ratio occurred was November 2014 (Table 1).Six cases of hermaphroditism (1.1%) were observed in March, July, and October 2014 (Figure 4).Sex was not determined in six animals (1.65%) due to parasitic castration by a digenetic trematode (Table 1).
Histological analysis allowed concluding that the reproductive cycle of M. guyanensis in the Marine Reserve Bay of Iguape occurred continuously across the year, without a well-defined rest period.Peak numbers of females with full gametes occurred in March, October, and December 2014, and in January and March 2015.Partial release with  Reproductive cycle of Mytella guyanensis gamete proliferation occurred in almost every month for females especially between April and September 2014, and except for March 2015 (Figure 5), and for males the peaks occurred in March, April, July, August, October, and December 2014 and March 2015 (Figure 6).Follicles containing oocytes in lysis were visualized in females with gamete repletion.The period of total release for males and females did not coincide (Figures 5 and 6).
All hermaphrodites presented partial release of gametes.Total release was less prevalent in males than partial release with proliferation (Figure 6).Despite the presence of release peaks of gametes in different times of the year (Figures 5-6), the Kruskal-Wallis test showed that these periods did not differ significantly from rates of release that occurred in other months, both in males (p = 0.576; n=104,) and in females (p = 0.161; n=54).
At the full gamete stage, the follicles were expanded and juxtaposed for both sexes, presenting mature cells (Figures 7C and 7D; Figures 8C and 8D).On the partial release stage with proliferation, the follicles presented lumens with visible spaces formed by gamete release, both in males and females (Figures 7E and 7F; Figures 8E and 8F).
The Spearman test indicated no significant correlation between the stages of gamete release and environmental variables: temperature (r s =0.004; p= 0.958), salinity (r s =0.010; p= 0.904) and precipitation (r s =0.080; p= 0.310), considering both males and females together (n = 162).

Discussion
Mytella guyanensis presented sexual dimorphism, related to the colour of the gonads between the sexes in the Caribbean region, as reported by Cruz and Villalobos (1993).However, this fact was not observed in other locations (Carpes-Paternoster, 2003;Luz, 2009;Gomes et al., 2009).Though separate sexes (dioecism) is the dominant situation for mytilids (Coe, 1943), cases of hermaphroditism were reported in the literature on surveys conducted for this species in other Brazilian coastal regions (Carpes-Paternoster, 2003;Gomes et al., 2009), which agrees with the findings of this study, and which may be associated with environmental factors, such as sudden variation in salinity and pollution (Ceuta et al., 2010).
As for the sex ratio, the ratio of male to female of M. guyanensis did not differ from 1:1, corroborating with affirmations of Coe (1943) for marine bivalves, and was similar to that found in other locations, such as in the Ochoa estuary, Costa Rica (Cruz and Villalobos, 1993); in Tavares River, Santa Catarina Island (Carpes-Paternoster, 2003); in Cachoeira River, Ilhéus, Bahia (Luz, 2009) and in the Caeté estuary, Bragança, Pará (Gomes et al., 2009).Significant differences were found in the sex ratio for this species in a study made by Adorno (2003), in Oruabo Farm, Bahia's Recôncavo, with male predominance.However, the method used by this author (macroscopic observation followed by puncture and fresh analysis of gonadal material), in agreement with what was observed by Cruz and Villalobos (1993), is not very reliable, as sexual dimorphism is not always so evident in this species (Gomes et al., 2009).
Various studies (Loosanoff and Davis, 1952;Galtsoff, 1964;Grotta and Lunetta, 1980) showed that abiotic factors (mainly temperature and salinity) control gametogenesis, because they influence transfer mechanisms of stored reserves in the digestive gland to the gonads, in addition to determining the qualitative and quantitative composition of the plankton, which is the mussels' food, supply the nutritive demand for these animals.High temperatures stimulate gametogenesis, promoting maturation of the gonads and gametes proliferation and spawning, whereas low temperatures usually inhibit the initial process of maturation (Gil and Thomé, 2004).Based on this, the high and stable temperatures throughout the year in the region of study favoured the continuous reproductive cycle.According to Boehs (2000), bivalves of tropical regions, where the seasons are well defined, tend to present continuous reproduction, with peaks of gamete release.This was observed for M. guyanensis in Brazil's North (Gomes et al., 2009) and Northeast (Luz, 2009) regions, as well as for other bivalve species, such as Anomalocardia brasiliana (Lavander et al., 2011), Crassostrea rhizophorae (Lenz and Boehs, 2011), Tagelus plebeius (Ceuta andBoehs, 2012), Perna viridis (Acosta et al., 2012), and Iphigenia brasiliana (Silva et al., 2012).
Regarding salinity, Leonel et al. (1983), in a study of variation in salinity, showed experimentally that A. brasiliana resisted salinity variations ranging from 17 to 42, with optimal value of 22, values in the range found in this study.Near the limits of tolerance, bivalves close the valves, reducing the filtering and feeding rate, which affects gametogenesis.For M. guyanensis in the region of study, there was no significant correlation between the gamete release and rainfall, which preponderantly determines the salinity levels.
Most intense peaks of partial release with proliferation were less frequent in females than in males, and the same occurred regarding the duration of the period, a result that coincides with the ones found by Mzighani (2005) for Anadara antiquare, with females proliferating more slowly than males.Oogenesis is, in general, a slower process than spermatogenesis and requires higher energy expenditure, since the female gametes usually need to accumulate reserves for the first stages of embryo development, and that is why they tend to be substantially bigger compared to the male gametes (Gosling, 2003).
Continuous reproduction of M. guyanensis in the Marine Reserve Bay of Iguape promotes constant renovation of the population, allowing, at first, the exploration of this species throughout the year.Therefore, a management plan is necessary to assure the conservation of the mussel beds in the region and the sustainable exploitation of this food resource.This plan should include the creation of a closed season during the months with high release of gametes (March, April, July and August) and restrictions regarding the harvesting of small M. guyanensis throughout the year.
In agreement with Morgan (2008) on the importance of studies regarding reproduction for species conservation, the results of the present study may subsidize a regulation plan of the intensity of use of M. guyanensis in the RESEX Bay of Iguape in order to reduce the anthropogenic impacts on this food resource in this reserve.

Figures 1- 3 .
Figures 1-3.Values of temperature (°C), water salinity (PSU) and rainfall level from the collection area of the bivalve Mytella guyanensis, from March 2014 to March 2015.

Figure 4 .
Figure 4. Hermaphroditism in Mytella guyanensis from the Marine Reserve in the Bay of Iguape, Bahia, Brazil.In the central region, the presence of male follicles (MF) containing cells of male germ line and adjacent to these female follicles (FF) containing oogonia and oocytes.Bar: 100 μm.Staining: H.E.