From January to December 2012, 50 specimens of Nacella magellanica were collected each month on the rocky intertidal shore of Punta Ninfas, Chubut (42°58’42”S, 64°18’33”W). The analysis were made only with limpets above the maturity size [1.73 cm high in females and 1.5 cm in males, sensu Nieto Vilela (2014)]. Unfortunately, no individuals could be collected during the month of May because the extreme weather conditions made it impossible to access to the sampling site. Punta Ninfas is located in the open sea, about 100 km from the city of Puerto Madryn, an area lacking fishing activity and inhabited by a stable community of marine elephants (Mirounga leonina) and particularly frequented by outdoor tourists.

All the specimens collected were taken directly to the laboratory and 30 were fixed in Bouin’s fluid for 48 h and stored in 70% (v/v) aqueous ethanol for histology. Because physical procedures enhance spawning in other molluscs (Velasco et al. 2007, Aji 2011), no sand depuration was performed and the biofilm over the shell was not removed. For each sampling, morphometric measurements were recorded on 30 individuals with callipers (precision 0.1 mm) to establish the height, length and width. To determine the total weight, each shell was removed and the shell and body weights were recorded on a digital balance (precision ±0.001 g). Sex was determined on the basis of gonadal frotis under an optical microscope and related to gonadal colour: females were green and males yellow (see Results). From the total of sampled individuals, 5 specimens were used for nutritional values and 5 were used for ash determination. The 10 remaining specimens were used as a back-up, because in the process of shell extraction the tissue was often broken, making the specimens useless for nutritional composition determination or histological gonadal analysis.

Each month, only around 10 individuals were suitable for gonadal determination (Table 1). Gonads were processed following standard histological procedures to study the annual cycle, as described by several authors (Ramón and Amor 2002, Horn et al. 2005, Mazurkiewicz and Pokryszko 2005). The gonad was sectioned in anteroposterior slices (toward both front and back), embedded in paraffin and sectioned at 5 to 7 µm with a digital microtome (Leica), then heated for 24 h at 60°C, and finally stained with haematoxylin and eosin (Gabe 1968).

Each sample was observed by light microscopy (Zeiss Axiostar, Germany) and photographed with a digital camera (Sound Vision 2.0). For each female, the gonadal cycle was estimated by oocyte diameter measurements (length and width mean) in all cells with a visible nucleolus using the ZEN 2.3.0.13 program. Between 10 and 100 oocytes with a visible nucleolus were measured in every female. The gametogenic cycle of both sexes was divided into four stages (Table 2) according to the stage of the reproductive cells; the classification was based on previous studies on Patagonian gastropods (Morriconi 1999, Bigatti et al. 2008, Averbuj et al. 2010) (Fig. 1).

For nutritional value determination, five randomly selected individuals with a size range of 0.9 to 2.7 cm height were analysed monthly. The determination was performed in males and females together, because both sexes are collected for consumption at the same time due to the absence of external sexual dimorphism. Total body masses were placed on aluminium film and heated at 60°C to constant weight. Dried tissues were triturated before sample processing to obtain a composite sample and an average monthly value for protein, lipid and carbohydrate content.

These determinations were performed by colorimetry according to the respective protocols of Lowry et al. (1951) using albumin as standard, Zöllner and Kirsch (1962) using cholesterol as standard and Fraga (1956) using glucose as standard. The measurements were made in a Hewlett Packard Model 8452A spectrophotometer, and the results were expressed as percent dry weight of tissues.

For ash determination, a single aliquot of wet body mass of five individuals was first heated at 60°C until the dried weight (DW) became constant. Thereafter, the sample was calcinated in a muffle oven at 550°C for 12 h and the weight of the resulting ash (CW) was recorded. The percent ash content was calculated as (CW/DW)×100. The estimation of moisture content was performed with the same aliquot of the monthly wet sample by recording the initial weight (IW) and the value obtained after heating at 80°C to constant weight (CW). The percent moisture content was then expressed as ((IW-CW)/IW)×100.

Because the parametric assumption was not met, differences between weight and length in males and females were evaluated with Wilcoxon and Mann-Whitney tests. Differences between average oocyte size and seasonal biochemical comparisons were assessed with a Kruskal-Wallis test. For significant results (P<0.05) a pairwise comparison test of subgroups was applied. The sex ratio was determined and compared with a 1:1 proportion by a chi-squared test. All the testes and modal determination were performed with the Statistica 7.0 software package (StatSoft, Inc., Tulsa).

To infer whether the sea surface temperature (SST) in °C and the gametogenic cycle were associated, we made in situ measurements in each sampling event and estimated the data from standard mapped satellite images (SMIs, NOAA, http://oceancolor.gsfc.nasa.gov/) using those of seasonal satellite-derived SSTs from MODIS Aqua for the period January to December 2012. The images have a spatial resolution of approximately 9.2 km. The global SMI data were subsampled from the region bounded by latitude from 33° to 43°S and longitude from 49° to 65°W. The satellite data were extracted from the sea 2 km off the coast. To evaluate seasonal variation in oocyte diameter, we performed a Kruskal-Wallis test and a pairwise comparison test of ranks.

We found a gonad coloration pattern across the reproductive cycle; in the case of females, green gonads recorded during maturation become darker in totally spawned specimens. In males, the gonad was yellow upon maturing and in ripe (fully mature) specimens, but brown in individuals after gamete release. Of 30 fixed specimens, only a mean of 10 were suitable for gonadal determination, because of rupture of several slides caused by sand content in the gut or because they were full of parasites Gymnophalloides (Table 1). Castration due to digenean parasites (Gymnophalloides), a phenomenon well described previously for this species (Cremonte et al. 2013, Bagnato et al. 2015), was detected in 16% of the analysed individuals (Table 3).

The monthly value of the oocyte modal diameter, from January to December 2012, ranged from 12 µm in January and February to 145 µm in October. We observed two modal diameters in several months due to the presence of growing and ready-to-spawn oocytes (Table 3). The low oocyte modal diameter observed in January, February and March (12-16 µm) may correspond to a phase of recovering after spawning, with resorption and gamete developing, and the gonad shows a developing stage in March, with oocytes with a modal diameter of 129 µm. This stage continued until September, with an increase in the oocyte size until spawning; in October a second developing period started and continued with the increase and spawning (Table 3, Fig. 1). Oocytes with mode diameters bigger than 140 µm seemed to be mature and ready to spawn when we recorded them free in the gonadal lumen.

We observed that female gamete liberation is low in January, March and April and increases in August, continuing until the end of the year (Fig. 2). The male gametogenic cycle was different to the female one (Fig. 3). The first individuals in the ripe stage appeared from April to December, with a maximum in November. Gamete release in males occurred throughout most of the year, except in June. Thereafter, resorption was found throughout the year except in June and October.

Although histological sections of N. magellanica showed that individuals were in more than one gonadal stage (Figs 2 and 3) at the same time, we were able to determine a relationship between the oocyte sizes and the seasons. With a rank comparison we found three groups (selected by oocyte size) that were significantly different (P<0.05): A, summer oocytes with a low size; B, autumn oocytes with a medium size; and C, spring and winter oocytes grouped together.

Sea temperature values recorded in situ were similar to those recorded by satellite (Fig. 4). Throughout the sampling period the measurements made by the MODIS Aqua satellite recorded an average water surface temperature of 12.8°C (max. 17.13°C, min. 8.72°C) and a mean photoperiod of 13.0 h (max. 16.3 h, min. 10.06 h). The increase in photoperiod started after 21 June and reached a maximum of 16.3 h of light in December, while the SST began to increase in August and reached its maximum value of 17°C in January (satellite information, Fig. 4). The spawning started in August and continued until December, with the concomitant increase in the water temperature (Fig. 4). During the autumn and winter (March to August), the oocytes became mature, with diameters greater than 140 µm before gamete release in spring (September, October and November). Between the seasons, the average size of the oocytes was significantly higher in winter (120.72 µm) and spring (122.54 µm) (P<0.05). The smallest mean value was recorded during the summer (32.4 µm).

Finally, the female spawning period for N. magellanica was associated with a decrease in body weight (Fig. 5). During the month of July, when the gonads were mature and ready to spawn, the body weight underwent a slight increase, and decreased later during the spring spawning peak.

Results of biochemical analysis are shown in Table 3. The average nutritional values (dry weight) were 29.8±3.25% proteins, 2.71±0.12% lipids, and 1.8±0.26% carbohydrates. The average ash content was 25.0±2.2% and the average moisture 81.8±0.6%. The abnormally high percentages of ashes are related to the abundant content of inorganic sediment particles within the organism´s digestive system. Average seasonal concentrations showed no significant variations for proteins, but for lipids the winter average was significantly lower than that of the other seasons (P<0.05), while for carbohydrates the winter and spring averages were significantly lower and higher, respectively, than those of the other seasons (P<0.05).

We found a body weight variation in males and females across the sample period, though female body weight remained higher than male body weight during most of the year. Proteins were high in April and June, lipids remained low throughout the year with a slight decrease during the spawning period (Table 3), and carbohydrates increased during the spawning period with a maximum in October (Fig. 5).