To determine the size of the first juvenile, plankton samples were collected from the coastal area near Comodoro Rivadavia (Chubut, Argentina) in October-November 2008. Decapodites were randomly selected and transferred into individual bowls with 150 ml of seawater at 8°C. Three times a week, seawater was changed, and recently hatched Artemia nauplii were supplied. When the decapodites moulted, the CL of juveniles was measured from the posterior edge of the orbit to the posterior mid-dorsal edge of the carapace, using a stereoscopic microscope with a graduated ocular (±0.01 mm). Juveniles were kept in the laboratory until the next moult to estimate intermoult duration and size of the second juvenile instar. Seawater was changed once a week and juveniles were fed three times a week with the isopods Exosphaeroma sp. and commercial fish food.

Squat lobsters were sampled monthly from boats using a bottom net with a 2-m net mouth and a 10-mm mesh size. Samplings were carried out between June 2008 and December 2009 at a depth ranging between 30 and 97 m in areas adjacent to the city of Comodoro Rivadavia. In the laboratory, the CL was measured using a digital caliper (±0.01 mm) or using a stereoscopic microscope with a graduated ocular (±0.01 mm) for smaller juveniles. Moult activity was analysed monthly in randomly selected field subsamples of ten males and ten females. Moult stages were identified, observing the changes in the distal edge of the exopodite of the second maxilliped according with the criteria used by Tapella (2002). The distal edge of the exopodite was removed and observed under a microscope (100×).

Specimens that were presumed to be in premoult according to carapace colour and hardness were placed in individual plastic mesh cages in aquaria with variable conditions fixed to environmental conditions of temperature, salinity and light:dark duration. Food (isopods Exosphaeroma sp. and commercial fish food) was supplied three times a week, and seawater was changed once a week. Premoult and postmoult CL were recorded to calculate moult increments

MI=CL_pos–CL_pre.

The relative increment was also calculated as

RI=(CL_pos–CL_pre)/CL_pre.

To minimize a possible effect of culture, only squat lobsters that moulted in the first 15 days after the sampling were considered. However, after the moult, specimens were maintained in the laboratory to estimate intermoult duration.

A regression was made between CL_pre and growth indicators (CL_pos, MI and RI). Analysis of covariance (ANCOVA) was used to determine whether regression equations differed between sexes. Homogeneity of slopes was previously tested. Comparison was considered significant when P<0.05.

Length-frequency histograms were built using CL data. Bhattacharya’s method was applied to length-frequency analysis, using FiSAT II (Gayanilo et al. 2005).

The gonadal development of the young of the year was analysed in field samples from June to October and laboratory assays. Based on size-frequency distribution and available data about the reproduction of M. gregaria (Vinuesa 2007), we considered that ovigerous females smaller than 13 mm CL were mainly specimens 0+. Additional samplings in the reproductive period were carried out in 2010 and used to analyse gonadal development and fecundity. After morphometric and sex data were recorded, subsample specimens (n=30) were dissected and the gonads were removed. The presence of spermatophores in the deferent duct and of vitellogenic oocytes was considered as an indicator of gonadal maturity in males and females, respectively. A smear of the deferent ducts was observed under a microscope at 400× magnification. In females, the diameter of 40 oocytes was measured using a stereoscopic microscope with an ocular micrometer to the nearest 0.01 mm. Oocytes were classified as previtellogenic and vitellogenic according to earlier works (Tapella et al. 2002, Vinuesa 2007). The presence of eggs in females was also recorded.

The analysis of field samples did not preclude the possibility that the smallest specimens of the previous cohort could be involved in the analysis, particularly at the end of the first year, so laboratory experiments were performed. In 2009, young of the year (n=100) of both sexes with CL ranging between 7.91 and 9.42 mm were placed in individual plastic mesh cages in aquaria. The experiments started in May and continued until October according to the breeding period reported in an earlier work (Vinuesa 2007). Ten males and ten females were sampled monthly and then measured and dissected. Gonadal development and moult stage were evaluated according to the methods and criteria mentioned above. Moult increments were recorded during the experiments to evaluate differences in growth of mature and immature individuals. Analysis of variance (ANOVA) was performed to compare moult increments in the two groups. Homogeneity of variance and normality assumptions of ANOVA were previously tested. Comparison was considered significant when P<0.05.

Realized fecundity (Ramírez Llodra 2002) and reproductive output (RO) (Hines 1982) were analysed in females 0+. Squat lobsters were measured as mentioned above and the egg masses were detached from pleopods. Fecundity was estimated using gravimetric procedures described by Vinuesa (2007). Egg masses were oven-dried at 60°C for 48 h and then weighed with an analytical balance (0.0001 g). A fraction of eggs (no less than 300) was used.

The weights of egg dry masses and squat lobster dry mass were used to estimate RO. Thirty squat lobsters of each year were dried and weighed according to the procedures mentioned above. These procedures were also used in older females to compare fecundity and RO in both groups.

Piecewise regression variance was used to evaluate differences in fecundity between females 0+ and older females. ANOVA was used to evaluate differences in RO between two groups of females. Homogeneity of variance and normality were previously tested. Comparison was considered significant when P<0.05.

The size of postlarvae ranged between 1.98 and 2.69 mm (2.38±0.15 mm CL). First juvenile CL varied between 2.07 and 3.26 mm (2.45±0.17 mm CL). Most postlarvae moult to juveniles in mid-November. Juveniles moulted again 15-20 days after the settlement. The size of the second juvenile instar ranged between 2.56 and 4.05 (3.45±0.48 mm CL).

Moult activity was higher during summer and early autumn, while at the end of winter (August-September) most squat lobsters were in intermoult. Moulting activity started again in spring. The moult cycle was similar between females and males (Fig. 2).

Growth indicators (CL_pos, MI and RI) were related to CL_pre and were similar in males and females (Table 1). In agreement with Hiatt’s criteria (Kurata 1962), the young of the year of both sexes showed positive allometric growth (b>1). Moult increment was higher in larger-size individuals (0+). However, the relative increment was higher in small juveniles and negatively related to CL_pre (Fig. 3). Intermoult duration was also negatively related to CL (R²=0.76; F_{(1, 49)}=158.6; p<0.01). In smaller-size squat lobsters (≈4 mm CL) intermoult lasted approximately 15 days, whereas in larger-size individuals (≈11 mm CL) intermoult lasted 45-57 days (Fig. 4).

Decapodite moults suggest that the first juveniles settle in November. In field samples, early juveniles were observed in January. In agreement with the size of first juvenile and moult increment observed, the smallest juveniles found in field samplings would correspond to third crab instar. After settlement, juveniles showed fast growth until winter. Growth restarted in September-October (Fig. 5). Seven size classes were identified from January to October. The size frequency distribution was similar in males and females. Differences in size between males (12.56±0.44 mm CL) and females (11.90±0.76 mm CL) appeared at the end of the first year after the settlement (Table 2).

Males 0+ showed spermatophores in the deferent duct during all the reproductive periods from June to October. At the beginning of the breeding period, nearly 70% of males had mature gonads. In August 2008, all males 0+ analysed had spermatophores in the deferent duct, while the maximum proportion of mature males was 90% in 2009 and 86.6% in 2010 (Fig. 6). In the laboratory experiments, only 40% of males had mature gonads in June. As was observed in the 2008 field samples, in the laboratory experiments, all males were mature in August (Fig. 7a). During the experiments, males moulted at the end of June and again at the end of September.

At the beginning of the reproductive period (early June), females 0+ showed no mature gonads. In 2010, a low fraction of females 0+ showed mature gonads. However, samples were carried out at the end of June. The percentage of females 0+ with mature gonads increased throughout the reproductive season until August, and then the fraction of females with vitellogenic oocytes decreased. Ovigerous females 0+ appeared in August in 2008 and 2009 (no data of July), while in 2010 ovigerous females appeared in July. However, only ≈50 % of females 0+ carried eggs (Fig. 8). In early June, females 0+ showed oocytes in previtellogenesis. These oocytes were always present from June to October. Females with vitellogenic oocytes were observed in August-September in all years. In October, when the mating period finished, only females with previtellogenic oocytes were observed. Oocyte size and development were similar in the field and laboratory cultures (Fig. 9). Females in the laboratory experiments moulted in the same periods as males. In August-September, when vitellogenic oocytes were found, a high proportion of intermoult females were recorded (Fig. 7b). In the moult of June, immature females had moult increments 29% higher than females that later showed mature gonads (Table 3). The moult increments of this experiment were not used in growth analysis due to the long culture time. However, the moult increment was used to evaluate the effects of growth increment in future gonad development. Intermoult duration in this experiment was longer than the moult increment experiment, which was performed in a non-reproductive period.

The size of ovigerous females 0+ ranged between 8.96 and 12.87 mm CL, while fecundity varied between 30 (CL=9.83 mm) and 339 eggs (CL=11.18 mm). The fecundity-size relationship of females 0+ was significantly different than the fecundity of larger-size cohort (Fig. 10). Significant differences were found in RO between females 0+ (6.03±2.21) and older females (11.26±4.38) (Table 4).