INTRODUCTION
Studies of animal populations are essential to understand the dynamics of ecosystems. Generally, population analyses have focused on descriptions of density, size-structure, sex-ratio and breeding periods that can be compared to other populations of the same species, genus or other taxonomic levels (Litulo, Reference Litulo2005a). Although hermit crabs are not likely to be of commercial importance, they contribute to the mass of planktonic organisms in coastal waters that help to feed the young of commercial species (Squires et al., Reference Squires, Ennis and Dawe2001), and they are an important part of the trophic chain, mainly where their populations are largest.
In crustaceans, the reproductive cycle is influenced by biotic and abiotic factors that rule the larval, gonadal and embryonic development (Mantelatto & Garcia, Reference Mantelatto and Garcia1999; Bertini et al., Reference Bertini, Fransozo and Braga2004). The breeding season may vary between populations in response to the influence of environmental factors in a given area (Litulo, Reference Litulo2005b). Sastry (Reference Sastry and Watermam1983) proposed two reproductive patterns for crustaceans: (1) continuous reproduction, when ovigerous females occur uniformly throughout the year; or (2) seasonal reproduction, with peaks in some months, with ovigerous females occurring only in some months of the year.
In hermit crabs, four aspects of reproduction have been studied: (1) frequency of ovigerous females by months or seasons (Manjón-Cabeza & Garcia-Raso, Reference Manjón-Cabeza and Garcia-Raso1995; Martinelli et al., Reference Martinelli, Mantelatto and Fransozo2002; Litulo Reference Litulo2005a, Reference Litulob); (2) analysis of fecundity (Mantelatto & Garcia, Reference Mantelatto and Garcia1999; Turra & Leite, Reference Turra and Leite1999; Negreiros-Fransozo et al., Reference Negreiros-Fransozo and Fransozo1992; Litulo, Reference Litulo2005c); (3) macroscopic analysis of the gonads (only by Bertini et al., Reference Bertini, Fransozo and Braga2004); and (4) histology of gonads (Manjón-Cabeza & García-Raso, Reference Manjón-Cabeza and Garcia-Raso2000a, Reference Manjón-Cabeza and Garcia-Rasob).
The hermit crab Clibanarius vittatus (Bosc, 1802) is the most abundant in the intertidal zone of the São Vicente Estuary. According to Melo (Reference Melo1999), this species lives in estuarine areas, coral reefs and sandy substrates, in shallow waters up to 22 m deep. Studies have treated its population biology and reproduction (Kicher, Reference Kicher1967); larval and juvenile development (Young & Hazlett, Reference Young and Hazlett1978; Brossi-Garcia, Reference Brossi-Garcia1988, respectively); and other reproduction and population analysis (Fotheringham, Reference Fotheringham1975; Lowery & Nelson, Reference Lowery and Nelson1988; Reigada & Santos, Reference Reigada and Santos1997; Turra & Leite, Reference Turra and Leite2000, Reference Turra and Leite2001; and Hazlett et al., Reference Hazlett, Rittschof and Bach2005). In these studies, the reproductive period was determined by the frequency of ovigerous females during the year, a common method used to study the reproduction of pleocyemata crustaceans.
This study aimed to analyse the population structure (sizes of sexes, seasonal size–frequency distribution and sex-ratio) of the hermit crab C. vittatus in the estuarine region of São Vicente, state of São Paulo, Brazil, and the reproductive period based on the occurrence of ovigerous females during the seasons, compared to the frequency of gonad maturation stages.
MATERIALS AND METHODS
The study was conducted in the intertidal zone of Pescadores Beach (23°58′21″S–46°23′35″W: Figure 1), located on the estuarine channel of São Vicente (São Paulo), Brazil, which shelters a diverse marine fauna, including intertidal molluscs, macroalgae and benthic crustaceans (Sant'Anna et al., Reference Sant'Anna, Zangrande, Reigada and Pinheiro2006).
Fig. 1. Estuary of São Vicente (São Paulo), Brazil, showing the location of Pescadores Beach (setae).
The hermit crabs were captured by hand during low tide, from May 2001 to April 2003. Monthly captures were made by two people, with a standard capture effort of 10 minutes. Water temperature (°C) was measured monthly. The hermit crabs were stored frozen. For analysis, the specimens were defrosted and removed from their gastropod shells, for identification according to Melo (Reference Melo1999). The carapace shield length (CSL) was recorded from the tip of the rostrum to the midpoint of the cervical groove, using a Vernier caliper (0.01 mm). For the size–frequency distribution analysis, size-class intervals of 1 mm of CSL were used. The size intervals of each class were determined according to the mathematical formula of Sturges (Reference Sturges1926). After identification and biometric analysis, the sex of each individual was determined from the position of the gonopores and the presence of eggs, and four population categories were established: (1) males, with gonopores on the basis of the fifth pereopods; (2) non-ovigerous females, with gonopores on the basis of the third pereopods; (3) intersex individuals, with gonopores on the bases of both the third and fifth pereopods; and (4) ovigerous females. Individuals of uncertain sex and intersex individuals were not included in the statistical analyses of population and reproduction.
All the individuals were dissected to observe the stage of gonad maturation, and were classified according to three gonad stages determined for males (immature or rudimentary gonads, which are undifferentiated or thin and translucent with two small white ducts; developing, with an initial winding, pale or white filament; and developed, with a strongly coiled testis with the dominant visible organ pale white and a white vas deferens), and females (immature or rudimentary, with ovaries as two fine, wine-coloured filaments; developing, with ovarian maturation beginning and represented by two wine-coloured filaments, occupying a large volume in the abdomen; and developed, when ovaries are dark-wine-coloured and fill the abdomen). The reproductive period was determined by the percentage of individuals of the same sex with developed gonads and the monthly incidence of ovigerous females during the course of the sampling period.
The normality of the size–frequency distribution was tested by the Kolmogorov–Smirnov (KS) test for each population category, and the size of individuals was compared by the Kruskal–Wallis test. The monthly and overall sex ratios (M:F) were tested with the Chi-square test (χ2). The relationship of sex-ratio to size-class was fitted only for classes with more than ten individuals, as recommended by Wenner (Reference Wenner1972). The correlation of water temperature (°C) and the incidence of individuals with developed gonads were tested by Spearman's correlation. A significance level of 5% was adopted for all statistical analyses (Sokal & Rohlf, Reference Sokal and Rohlf1995).
RESULTS
A total of 2554 hermit crabs were captured in the intertidal zone of Pescadores Beach: 701 males, 1741 non-ovigerous females, 48 ovigerous females and 64 intersex individuals. Male individuals were significantly larger than females (P < 0.05), but intersex individuals were the same size as males (Table 1).
Table 1. Number of individuals and size-range (carapace shield length) of the hermit crabs captured on Pescadores Beach. N, number of individuals; Min, minimum; Max, maximum; x mean; SD, standard deviation; M, males; F, non-ovigerous females; OF, ovigerous females; INT, intersex individuals.
Means followed by the same lower-case letter did not show significant differences (P > 0.05).
The overall size–frequency distribution of the population differed from normality (KS = 0.0755; P < 0.01), with a unimodal distribution. The size–frequency distribution of the males was represented by a platykurtic bell-shaped curve that differed from the leptokurtic bell-shaped curve for females (Figure 2). Although the size–frequency distribution of the population did not show a normal distribution, this was not true for all sexes; the size–frequency distribution for males differed from normality (KS = 0.0791; P < 0.01), but the size–frequency distributions of non-ovigerous females and ovigerous females showed significant normal distributions (KS = 0.03; P > 0.05 and KS = 0.1079; P > 0.05). In the same figure, we can observe that the smaller and intermediate classes were composed mainly of females (modal size 6.5–7.5 mm CSL), and the larger classes only of males (modal size 9.5–10.5 mm CSL).
Fig. 2. Size–frequency distribution of individuals of Clibanarius vittatus caught on Pescadores Beach, São Vicente Estuary (São Paulo), Brazil. CSL, carapace shield length.
Fig. 3. Seasonal size–frequency distribution of individuals of Clibanarius vittatus captured on Pescadores Beach, São Vicente Estuary, Brazil. CSL, carapace shield length.
Fig. 4. Percentage (fitted with a mobile mean) of males by size-classes, of Clibanarius vittatus caught on Pescadores Beach, São Vicente Estuary, Brazil. CSL, carapace shield length; horizontal line (- - -) represents the expected sex-ratio (1:1).
The seasonal size–frequency distribution of the population (Figure 3) showed unimodal and bimodal distributions during the seasons. Few juveniles were recorded in this location, and females were more abundant than males during the two years of the study. Ovigerous females occurred in all seasons of the year, but in lower frequency in the winter. The abundance of females by month is shown in Table 2. The sex ratio (M:F), was nearly always skewed toward females, except in two months of the study period, the overall sex-ratio (0.39:1; M:F) differed significantly from 1:1 (χ2 = 475.399; P < 0.001). The sex-ratio related to size, showed an anomalous curve (Figure 4), with females extremely abundant in intermediate size-classes, and males numerically superior in the larger ones.
Table 2. Sex-ratio by month obtained for the population of hermit crabs Clibanarius vittatus on Pescadores Beach, São Vicente (São Paulo), Brazil.
ns, non-significant; * P < 0.05.
The percentage of individuals with developed gonads was larger in the warmer months of the year (Figure 5). The percentage of ovigerous females (Figure 6) evidenced a seasonal reproductive pattern at this location, with more intensive reproductive activity in the warmer months. This is supported by a positive correlation between water temperature and females with developed gonads (rs = 0.61; P = 0.0017). Dissections of the intersex individuals revealed male gonads, with one paired coiled whitish duct.
Fig. 5. Seasonal percentage of males (A) and females (B) of Clibanarius vittatus, with rudimentary (RU), developing (DV) and developed (DE) gonads during the study period on Pescadores Beach, São Vicente Estuary, Brazil.
Fig. 6. Seasonal abundance of ovigerous females of Clibanarius vittatus caught during the study period on Pescadores Beach, São Vicente Estuary, Brazil.
DISCUSSION
The sexual dimorphism in relation to size observed in the present study is a common aspect of hermit-crab biology, as suggested by several previous reports (Fransozo & Mantelatto, Reference Fransozo and Mantelatto1998; Manjón-Cabeza & García-Razo, 1998; Turra & Leite, Reference Turra and Leite2000; Mantelatto & Martinelli, Reference Mantelatto and Martinelli2001; Branco et al., Reference Branco, Turra and Souto2002; Bertini et al., Reference Bertini, Fransozo and Braga2004; Macpherson & Raventos, Reference Macpherson and Raventos2004; Litulo, Reference Litulo2005a, Reference Lituloc; Litulo & Tudge, Reference Litulo and Tudge2005; Mantelatto et al., Reference Mantelatto, Christofoletti and Valenti2005). According to Abrams (Reference Abrams, Chelazzi and Vannini1988), three factors may influence sexual dimorphism in hermit crabs: (1) the difference of energy available for growth, with less somatic growth in females due to greater expenditure of energy in egg production; (2) the larger reproductive effort exhibited by males, which are able to copulate with more than one female; and (3) the larger dimensions of males, to optimize fertilization of the females and win intraspecific fights.
In this population of C. vittatus in São Vicente Estuary, the differential growth rate between the sexes (as proposed by Abrams, Reference Abrams, Chelazzi and Vannini1988) suggests an important factor for the sexual dimorphism. According to the growth analysis of this same population by Sant'Anna et al. (Reference Sant'Anna, Christofoletti, Zangrande and Reigada2008), males have a larger constant growth than do females (males, K = 0.51; and females, K = 0.40), and the different growth periods of the sexes reduce intraspecific competition for gastropod shells. A slower growth rate of females would be a consequence of utilization of relatively small shells in nature, as suggested by Litulo (Reference Litulo2005 a, Reference Lituloc) for Dardanus deformis (H. Milne-Edwards, 1836) and Clibanarius longitarsus (De Haan, 1849), respectively.
A few ovigerous females and juveniles were captured during the present study, a result similar to that obtained by Litulo (Reference Litulo2005a) for D. deformis in Maputo Bay, Mozambique. The lower abundance of juveniles suggests that the recruitment area is different from the area inhabited by adults, possibly another area with more protection and specialized or different resources for the juveniles. Although there is a good supply of smaller shells of Littorina sp., suitable for juveniles, at Pescadores Beach (personal observation), it is possible that C. vittatus ovigerous females migrate to the entrance of the estuary, where their larvae have access to optimum salinities for the first juveniles (25 to 35‰, according to Young & Hazlett, Reference Young and Hazlett1978), avoiding the wide salinity variations in the study area (20 to 35‰, according to Sant'Anna et al., Reference Sant'Anna, Zangrande, Reigada and Pinheiro2006). Thus, when the juvenile hermit crabs return inside the estuary they already measure several millimetres in size. A similar absence of juvenile individuals was noted by Turra & Leite (Reference Turra and Leite2000), who studied three sympatric species of hermit crabs of the genus Clibanarius, on another part of the Brazilian coast. As estimated by these authors, the recruitment period (December to March, according to Sant'Anna et al., Reference Sant'Anna, Christofoletti, Zangrande and Reigada2008) was equivalent to those months when greater reproductive activity was observed in this population.
The sex-ratio of the C. vittatus population in the present study was skewed toward females, with a difference from the expected proportion (1:1). The same has been recorded for various populations of different hermit-crab species around the world (Manjón-Cabeza & García-Raso, Reference Manjón-Cabeza and García-Raso1998; Benvenuto & Gherardi, Reference Benvenuto and Gherardi2001; Macpherson & Raventos, Reference Macpherson and Raventos2004; Litulo, Reference Litulo2005c; Litulo & Tudge, Reference Litulo and Tudge2005) and in two other populations of C. vittatus (Lowery & Nelson, Reference Lowery and Nelson1988; Turra & Leite, Reference Turra and Leite2000). The same pattern was obtained for overall sex-ratio, and as a function of size, showed an anomalous curve pattern according to the classification proposed by Wenner (Reference Wenner1972). This author suggested four explanations for this pattern of the sex-ratio: (1) differences between the sexes in longevity and time for growth; (2) differential migration; (3) different mortality or growth rates between sexes; and (4) sex reversal. In C. vittatus, the most probable causes are the difference in the growth rates between the sexes, with an influence of longevity (Sant'Anna et al., Reference Sant'Anna, Christofoletti, Zangrande and Reigada2008) or habitat partitioning as suggested by Turra & Leite (Reference Turra and Leite2000). Sex reversal occurs in a low percentage of C. vittatus populations (see present study and Turra & Leite, Reference Turra and Leite2000), representing a third and minor cause of the sex-ratio pattern observed, and there still is no explanation for the mechanism of sex reversal in hermit crabs (Turra & Leite, Reference Turra and Leite2000; Turra, Reference Turra2004).
Hermit crabs may have continuous or seasonal reproduction (Manjón-Cabeza & Garcia-Raso, Reference Manjón-Cabeza and García-Raso1998; Bertini & Fransozo, Reference Bertini, Fransozo, Klein von and Schram2000; Turra & Leite, Reference Turra and Leite2000). In this study, the C. vittatus population showed a seasonal reproductive pattern with some large peaks of individuals with developed gonads in some warmer months (seasonal–continuous pattern, according to Pinheiro & Fransozo, Reference Pinheiro and Fransozo2002). These data corroborate those obtained for another C. vittatus population by Turra & Leite (Reference Turra and Leite2000). Seasonal reproduction is associated with environments characterized by wide variations in temperature or food availability (Giese, Reference Giese1959). Temperature may act as a metabolic, biochemical and hormonal modulator, triggering the mechanisms of ecdysis, mating and gonad development. However, the availability of and/or competition for shells may influence reproductive activity and cause displacement of reproductive peaks and recruitment periods in coexisting populations (Turra & Leite, Reference Turra and Leite2000; Wada et al., Reference Wada, Kitaoka and Goshima2000).
Finally, Turra & Leite (Reference Turra and Leite2000) suggested that the occurrence of populations with seasonal reproductive patterns in the tropics and with continuous reproduction in temperate areas may be based on populations' evolutionary histories, although local factors should be considered. But a statistical analysis of a more exhaustive data set is required to elucidate any patterns in breeding peaks across latitudes (in both hemispheres), and comparative breeding records for congeners found over a wide range of latitudes should provide the best picture (Litulo & Tudge, Reference Litulo and Tudge2005).
ACKNOWLEDGEMENTS
Thanks to the biologist Cilene Mariane Zangrande for help in sampling and laboratory analyses in part of this study and to Dr Janet Reid for final English revision of the manuscript.
