Reproductive patterns of Aratus pisonii (Decapoda: Grapsidae) from an estuarine area of São Paulo Northern Coast, Brazil

The tree crab Aratus pisanii (H. Milne Edwards 1837) is a tropical sesarminae species widely distributed in the western Atlantic from central Florida to Brazil, arid in the eastem Pacific from Nioaragua to Peru (Melo 1996), being commonly found 00 the upper HUoral zone in IUangroves.

The tree crab Aratus pisanii (H. Milne Edwards 1837) is a tropical sesarminae species widely distributed in the western Atlantic from central Florida to Brazil, arid in the eastem Pacific from Nioaragua to Peru (Melo 1996), being commonly found 00 the upper HUoral zone in IUangroves.
Many aspects concerning to the Ji fe cycle and population biology. of A. pisonii have been studíed in severalenvironments.(Hartnoll1965, Wamer 1967, Díaz andConde and Díaz 1989a b). Tbose studies baverevealed the large plasticity ofthe life historyJeatures shown by t.his species' populations inhabiting different environments.
It ís well known that reproductive periodici ty may vary along a latítudinalgradient asa function of envíronmental changes. The relative importance of a given factor as cueing the initi atiQnof reproductive activities varíes among species ,and' habitats. The identification of this factor can clarify the cau §es of annuaJ ábim dance fluctuations in natural populations.
The frequency of . o, ccurrence a nd the repro ductive period oi the tree mangrove crab A. pisonii in subtropical estuanne mangrove sys tem (230 29' S, 45Ciío' W) near the australlimit of its Western AUantic distribution are ana}yzed in .this study to provide information for f�her comparative work.

MATERlALS AND METHODS
Ubatuba is located in the northern coast of Sao Paulo State, Brazil. The climate in tbis region is tropical humid tending to subtropícal. Rainfall is more intense from September lO March (spring to summer). The "Comprido" and "Escuro" streams are the freshwater resource of the area and they drain into the Fortaleza Bay. The salinity varies widely due the mixed tide regime in that region.
Monthly samples were carried out from January 1993 to June 1994 in the fringe area of rnangrove near to the water, where the abundance of A. pisonii had been previously observed to be bigher than in ¡nlet areas. The animal s were cap tured by hand from the roots and branches of the mangrove trees. Samplings consisted of 1 hour catch-effort sessions conducted by two people during low lide. AH crabs were sexed and pres ence or absence of eggs in the females was recorded. Their carapace width (CW ) was mea sured to the nearest 0.1 mm using a Vernier caliper. The crabs were grouped into eleven size classes from 4.1 lO 26.0 mm Cw.
Recruitment pulses were verified in frequency of individuals measuring from 4.1 to 8.0 mm Cw. Indíviduals smaller than 4.1 mm CW were no! considered due to the difficulty to determine accurately their sexo The reproductive period was determined based on ovigerous female frequency in relation to mature female along the year. Since the small est ovigerous females obtained measured 15.0 mm CW, only crabs grouped in the 14.1-16.0 mm size class or larger were regarded mature. Reproductive periodicity was statistically exam jned by means of binomial proportion analyses (Goodman 1965). The sex ratio in each month and size class were examined by this analysis too. The significance level adopted for all analyses were 5%.
Due to the terrestrial habits of tbis species the association between proportíon of ovigerous females and two factors (monthly average air temperature and rainfall) was analyzed. Temperature and rainfall data were obtained in a nearby (4 Km) meteorological station of the University of Sao Paulo.

RESULTS
During the study period, a total of 1 078 crabs were obtained; 489 males, 589 females, of which 131 were ovigerous.
The size of the individuals ranged from 4.2 to 25.9 mm CW. The overall size frequency dis tribution presented an unímodal and slightly asymmetric shape skewed to the left (Fig. 1). Average size of both males and females was 16.8 mm Cw.  Overall sex-ratio was 1: 1.2, not differing statistically from the 1 : 1 ratio. Remarkable deviations from 1 : 1 ratio were registered in January, February and May 1993 and in January 1994, favouring females. Only in May and June 1994 males outnumbered females (Table 1). Deviations were observed in sorne síze class es, these differences being statístically signifi cant in sorne cases (Table 2). Females were more numerous in the intermediate size classes, from 14.1 to 18.0 mm CW. Immature individu als from 4.1 to 14.0 mm CW did not present dif ferences in sexual proportions. In large crabs from 22 to 24 mm CW, males were significant ly more numerous than females (Goodman's test, p < 0.05). Only a few specimens larger than 24.0 mm were obtaíned.

DISCUSSION
The Oyeran size frequency distribution of A. pisonii is unimodal and asymmetric which usu ally reflects a population in equilibrium (Hartnoll and Bryant 1990). The modal varia tions in the size frequency distribution along the year were mainly caused by recruitment pulses, which were more evident in June 1993, May and June 1994 (late autumn and early winter). However, juveniles could be also found in other months, but in smaller frequencies. This con spicuous recruitment was probably a result of intense reproductive activity occured in March 1993 andFebruary andMarch 1994 (summer). It can be supposed that the main period of megalopa settlement occurred from late sum mer to earIy autumn, since the larval develop ment of this species includes four zoeal stages and a megalopa instar in approximately one month (Warner 1968).
This reproduction and recruitment pattem could be a result of the geographical location of the studied area, a transitional faunistic region with tropical humid climate tending to subtropi cal. In this latitude, intensive temperature changes are likely to occur more often than in the tropics. In this study the temperature was very correlated to the ovigerous rate as had been observed in other studies as the main factor influ encing the reproduction (Fusaro 1980, Jones and Simons 1983, Siddiqui andAhmed 1992. An extended breeding season with a peak of higher activity during the rainy season was observed in the present study, as those obtained by Díaz and Conde (1989) and Conde and Díaz (1989a). Spawning in the rainy season may pro vide a selective advantage to intertidal decapods populations sínce that periods of higher rainfan rate can cause changes in the salinity of water (Pillay andNair 1971, DeVries et al. 1983) and also promote an increase of nutrients concentra tion, favouring the development of plank totrophic larvae (Siddiqui and Ahmed 1992), and an ¡ncrease of primary productivity and ses ton availability in the estuary (Rodriguez and Conde 1989). In the populatíon studied herein, higher flow rate and seawards water velocity during the rainy season, are conditions that probably prevent zoeae from stranding and osmoregulatory stress.
In this study the sex-ratio does not differ from the 1 : 1 in the smalIest classes but it does differ in the intermediate size classes favouring females and also in larger crabs favouring males, according to the "anomalous" pattem describedby Wenner (1972). The highest female proportion in intermediate size classes (from 14 mm CW), match the size in which an important part of the population is sexually active. For A. pisonii, Warner (1967) suggests differential growth rates in which males reach large sizes fas ter due to the higher reproductive effort of females which do not molt when they are incubating eggs.
According to Giesel (1972) variations in the sex-ratio occur primarily during the main sea sonal changes of environmental conditíons. Warner (1967) reported that during the repro ductive period of A. pisonii, females migrate from the interior of the mangrove to water edge increasing the relative female frequency. Peculiar characteristics of mangrove fringes, such as humidity conditions provide more suit able grounds to egg development and larval release (Emmerson 1994). Furthermore, female migration towards the fringes, possibly enhances the chances of mate encounters. Christy and Salmon (1984) assumed that sex ratio is associated to the mating system of each species, but Willson and Pianka (1963) pointed out that this is not a cause-effect relationship. It could be suggested that female-biased sex-ratio observed herein may be related to a poligynous mating system in which males could mate with different females increasing the reproductive output in population (Giesel 1972). However, further investigations are necessary to confirm the relation between sex-ratio and mating sys tem in A. pisonii as in other decapods.

ACKNOWLEDGMENTS
To the "Conselho Nacional de Desenvolvi mento Científico e Tecnológico -CNPq" that provided fellowship for the first author. To Adilson Fransozo from UNESP, Brazil for their suggestions. To the NEBECC memberS who helped us during the collections. !