Population structure and reproductive aspects of puffer fish Sphoeroides annulatus ( Jenyns , 1842 ) ( Osteichthyes : Tetraodontidae ) , landed in Teacapán , Sinaloa , Mexico

The puffer fish Sphoeroides annulatus is an important target species for the artisanal fishing fleets of NW Mexico. To obtain information on population structure of the local stock, we determined the total length and total weight (TL and TW) ranges, sex ratio and reproductive stages of 745 specimens of this species, landed from May 2010 to April 2011 in Teacapán, Sinaloa, NW Mexico. TL ranged from 15 to 40 cm and TW from 100 to 1600 g. There were no differences between mean TL (27.41 ± 4.14 cm) and TW (534.5 ± 226.0 g) of males and females respectively. Sex ratio was not significantly different (χ2 = 0.03, P > 0.05) from 1F:1M. The length-weight relationship for both sexes was TW = 0.044TL, R = 0.895. The value of the slope b was significantly lower than 3 (P < 0.05), and indicated negative allometric growth. The distribution of maturity stages indicated one main reproductive period from June to September and one less intense, from November to December for females, and in December for males. Size at first maturity (L50%) of females was 26.52 cm and that of males was 27.41 cm.


INTRODUCTION
The puffer fish Sphoeroides annulatus (Jenyns, 1842) is a common species in marine and coastal waters of the eastern Pacific from southern California to Peru, including the Gulf of California (Bussing, 1995;Amezcua-Linares, 1996), which has become a target species for the artisanal fishing fleets of the Mexican NW, for its high market price due to the quality of its meat (Chávez-Sánchez et al., 2008;Aguilera & Duncan, 2009).
According to Arreguín-Sánchez & Arcos-Huitrón (2011), this resource has been fully exploited during four of the five years for which data on landing volumes are available for the Gulf of California.Since this information concerns the 1979-1988 period, and because of the increased fishing effort devoted to this species, it is conceivable that captures have reached or possibly exceeded sustainable levels.
However, in spite of its increasing commercial interest, information on this species is scant.Particularly, studies on its reproductive biology in coastal Sinaloa waters are scarce and partially contradictory, possibly because of geographic or interannual differences of climatic and oceanographic conditions.According to Castellanos-Rodriguez et al. (1982) and Komar (2001), S. annulatus has a short reproductive season lasting two and three months, from April-May to June, whereas Sánchez-Cárdenas et al. (2007, 2011) described two annual reproductive periods.The first was longer and more important; it lasted from April to August with intense spawning from June to August, while the second lasted between October and November.
Most of these results were obtained using the landings of the Mazatlán artisanal fishing fleet.In this work we analyze the information on size distribution, biometric relationships and sexual maturity obtained from landings of the Teacapán fishing fleet between 2010 and 2011.

MATERIALS AND METHODS
Information was obtained approximately every second week, between May 2010 and April 2011, at Playa Sur (Southern Beach), which is located immediately to the south of the town of Teacapán and it is the main landing site of the local artisanal fishing fleet (Fig. 1).Sample size depended on the size of the catches, and was at least 20% of all puffer fish landed.Total length and weight (TL and TW) were determined to the nearest 1 mm and 1 g with a measuring board and a digital scale, respectively.Sex and maturity stages were determined using the morphochromatic maturity scale by Sánchez-Cárdenas et al. (2007), which recognizes five stages (resting, early developing, late developing, ripe and spawned) for females, and four (resting, developing, ripe and spent) for males.
The normality of size and weight distribution was verified with Kolmogorov-Smirnov tests, and the data were used to calculate length-weight relationships separately for each sex and for the total sample, adjusting the data to the model: TW = a TL b (Pauly, 1983) The frequencies of the maturity stages of each sex were determined monthly (Salgado-Ugarte et al., 2005), and the size at first maturity was calculated as in Gunderson et al. (1980), using the logistic model: P x = 1/1 + e a + bTL e a+bx In this model, P x is the proportion of organisms at size x with clear signs of gonadic activity (stages III, IV and V: advanced development, mature and spawned for females, respectively) and stages II, III and IV (in development, mature and ejaculate), in the case of males.
The mean TL and TW values of each sex were compared using Mann-Whitney tests, because the data were not normal (Kolmogorov-Smirnov test).The allometry of TW-TL relationships were determined by Student's t tests, and χ² tests served to determine whether monthly and yearly sex ratios were different from the theoretical 1:1 value.All tests were performed with α = 0.05 (Zar, 1996).

RESULTS
Data were obtained for a total of 745 organisms (375 males and 370 females, 50.3 and 49.5%, respectively).TL ranged from 15 to 40 cm; 56.2% of the females and 63% of the males were within the range 26 to 32 cm.The respective mean values were 27.5 ± 3.8 cm for males and 27.3 ± 4.5 cm for females, with no significant difference between sexes (Mann-Whitney, P = 0.966).The global mean value for both sexes was 27.41 ± 4.14.The size distribution was bimodal, and the modal values were 20 and 28 cm (Fig. 2).
TW ranged from 100 to 1600 g, with 61.69% of the catches between 300 and 700 g.The mean values were 528.5 ± 209.3 g for males and 531.3 ± 249.7 g  for females and the means were not significantly different (Mann-Whitney, P = 0.896).Considering both sexes, mean TW was 534.5 ± 226.0 g and the only modal value was 600 g (Fig. 3).
The relationships between TW and TL for males and females were described by the equations: TW=0.036TL 2.87 , R 2 = 0.921 (females) TW=0.058TL 2.73 , R 2 = 0.860 (males) The values of both slopes were significantly lower than 3 (P < 0.05), indicating negative allometric growth for both sexes.The equation calculated using all data was: TW = 0.044TL 2.815 , R 2 = 0.895 and the slope was again significantly lower than 3 (P < 0.05) (Fig. 4).
The F: M ratios calculated for sampling date were not significantly different from the expected 1:1 value.The annual F: M ratio was 0.987 F: 1 M.
Between >75 and >90% of the females were in maturity stages III to V from June through August.Stage V was dominant (87.5%) in August and it was present until December, but with lower percentages  There are also discrepancies concerning the type of growth: while we found negative allometric growth for both sexes, Sánchez-Cárdenas (2005) determined negative allometric growth for females but not for males, and Morán-Angulo (2009) found isometric growth for both sexes.As for size ranges, these differences could be explained by the different origin of the samples, or by differences in the sexual maturity cycle, since these could modify the value of the indicator of isometric-allometric growth.The overall 0.99H: 1M proportion found in this study is in agreement with Sánchez-Cárdenas (2005) for the whole annual cycle, but differs from the information obtained at the peak of the spawning season (June) in Mazatlán, where males were dominant over females (1M: 0.48H), which could be due to an haremic behavior, not observed in this work, but similar to that reported for other Tetraodontid species (Kobayashi, 1986;Gladstone & Westoby, 1988).
The main spawning period found in this work was from June to August, followed by a minor peak in winter.Although the separation between first and second peak (December instead of October-November) was between one and two months longer than that observed by Sánchez-Cárdenas (2005) and Sánchez-Cárdenas et al. (2007), this partial agreement indicates that there are no differences in the reproductive behavior of this species in the two sampling areas.In both, the most intense reproductive activity coincides with the whole summer season, whereas the indications of Castellanos-Rodriguez et al. (1982) and Komar (2001) suggest a far shorter reproductive season, with peak activity limited to May and June.

Figure 3 .
Figure 3. Frequency distribution of total weight by sex of Sphoeroides annulatus, in Teacapán, Sinaloa, Mexico.