Relative growth, sexual maturity and handedness in the ghost shrimp Callichirus major (Decapoda: Callianassidae) from the southwestern Atlantic

Summary: Biometric studies provide valuable information about changes associated with the growth and sexual maturity of living organisms. We analysed sexual dimorphism, allometric growth, sexual maturity and handedness in the ghost shrimp Callichirus major at Gonzaga beach, Brazil, where the catches of this species have been prohibited since 1992. To this end, a total of 544 individuals of C. major were collected during 12 months of sampling. Males were significantly smaller than females, denoting sexual dimorphism with respect to body size. The analysis of allometric growth between chelar carpus length and carapace length (CL) revealed a positive allometric relationship in juveniles of both sexes and adult males, but a negative allometry in adult females. Overall, our results showed the existence of two main growth phases related to sexual maturity, with a similar transition point for males (15.0 mm CL) and females (16.0 mm CL). Heterochely and homochely were registered in shrimp from both sexes, but in males heterochely occurred to a higher degree and was predominant (86.9%), whereas in females homochely was slightly more frequent (59.3%). The consequences of sexual dimorphism in terms of body size and chelipeds in the mating system of C. major are discussed in this study.


INTRODUCTION
In function of their feeding and behavioural interactions, crustaceans may have specialized claws as a first pair of pereopods (Mariappan et al. 2000).Hartnoll (1974Hartnoll ( , 2012) ) mentions a chelar difference (heterochely) in most decapod crustaceans, generally more pronounced in males, which use it in agonistic interactions related to territorial defence or during competition for sexual partners.During post-larval development, decapods can display body changes with the onset of the secondary sexual characters, generating two phases (juvenile and adult) better described by a power function (Y=aX b ) and separated by a critical moult (puberty moult) (Huxley 1950, Hartnoll 1974).According to Hartnoll (1982), when a dependent variable (e.g.claw size or abdominal article) is related to an independent variable (e.g.body size), growth rate can be obtained by allometric coefficient ('b' constant), generating an isometry (b=1) or allometry (positive b>1 and negative b<1).Such growth patterns are specific to each species and vary as a function of development phase of each sex (juvenile and adult), generating two discontinuous phases of growth and consequently two different levels of allometric coefficients that define the moment of the morphological sexual maturity in decapods (Pinheiro andFransozo 1998, Pardal-Souza andPinheiro 2013).
Heterochely, here defined as different-sized chelipeds, is a common attribute in most decapods (e.g.Aegla franca Schmitt, 1942: Bueno andShimizu 2009;Alpheus brasileiro Anker, 2012: Pescinelli et al. 2018;Platyxanthus crenulatus [Milne-Edwards, 1879]: Laitano et al. 2013), including members of the family Callianassidae Dana, 1852 (collectively known as ghost shrimps).Adult males of callianassids invariably develop one hypertrophied cheliped (heterochely), while females and juveniles have subequal or homochelous chelipeds (Hernáez 2018 and references therein).The handedness or laterality, i.e., the tendency for the major cheliped to appear on a particular side of the body, has been poorly studied in these species.A recent study in the callianassid shrimp Callichirus seilacheri (Bott, 1955) demonstrated that the larger chelipeds in adult males develop independently on either the right or left side of the body (Hernáez and João 2018).
The intertidal ghost shrimp Callichirus major (Say, 1818) is a conspicuous species on many sandy beaches of the western Atlantic (Felder 2001).Its geographical distribution covers from North Carolina, USA, to Santa Catarina, Brazil (Hay andShore 1918, Melo 1999).These animals build burrows in sediments, normally inhabited by only one shrimp (Rodrigues 1976).In the adult phase, the males of C. major are considered heterochelic due to an expressive claw difference in size/ shape (Rodrigues 1971, 1985, Manning and Felder 1986), but in juveniles and females the chelipeds are "subtly unequal" or considered homochelic (Rodrigues 1971).Chelar types in C. major have not been investigated and reported in detail.Rodrigues (1985) first examined the chelar growth pattern of C. major from the Bay of Santos, in the southeastern region of Brazil, based on the relationship between propodus length (PL) and carapace length (CL), finding a negative allometry in juveniles of both sexes and adult females, with a dimorphic pattern and positive allometry in adult males.Subsequently, Alves-Júnior et al. (2014) reported a positive allometry through this species's ontogeny with this same relationship, confirming a higher chelar growth in males than in females for a population of C. major from Recife, in the northeastern region of Brazil, a fact explained by the thermal regime promoted by latitudinal influence.Both of these studies considered the growth of a chelar article that reaches a smaller size than the carpus, a noteworthy characteristic in species of the genus Callichirus Stimpson, 1866 (Manning and Felder 1986, Hernáez et al. 2015, Hernáez and João 2018).Another recent contribution by Alves-Júnior et al. (2018) indicated that the whole cheliped in C. major grows at a higher rate than the carapace, especially for males.It is important to highlight that Hartnoll (1974Hartnoll ( , 1978) ) recommends the use of biometric variables resulting from a more prominent chelar article (dependent variable) to estimate the morphological maturity size in males, generally represented by chelar propodus variables.
Furthermore, considering the intense harvest suffered by C. major as bait for recreational fishing in several Brazilian regions (Rodrigues and Shimizu 1997, Souza and Borzone 2003, Hernáez et al. 2019), the population of this species often fails to maintain its natural structure and a better representation in size classes.We therefore estimated the chelar relative growth and morphological maturity sizes in C. major, using a more reliable biometric variable (carpus size) and based on specimens captured in a stable and well-balanced population protected by law (see Ordinance #850/1992) in Santos Municipality, Brazil.The patterns of handedness in this species were also determined in function of sex through the ontogeny.

Sampling and material processing
Specimens of C. major (Fig. 1A) were collected monthly (from March 2016 to February 2017) from the intertidal zone, at Gonzaga beach (23°58′13″S, 46°20′04″W), Santos (SP), southeast coast of Brazil (Fig. 1B).Then, ghost shrimps were randomly collected from burrows with a hand-made "yabbie" pump (diameter=77 mm, length=100 cm).Therefore, to obtain a representative monthly sample size we inspected all visible burrowing holes.After shrimp collection, each specimen was carefully rinsed with seawater, placed in a plastic bag and transported for further analysis in the laboratory.Part of the material analysed in this study was deposited in the Crustacean Biology Research Group (CRUSTA) collection, of UNESP IB/ CLP (CRUSTA 160058).
In the laboratory, each specimen was sexed according to the morphology of the first pair of pleopods (bi-segmented in males; tri-segmented in females; Rio 2018), following the methodology proposed by Hernáez et al. (2018) for other American species of Callichirus.The shrimps were measured under a Zeiss Stemi SV-6 stereomicroscope equipped with an image analysis system (Zeiss AxioCam MRc5) (Carl Zeiss, Oberkochen, Germany) to record the CL (from postorbital margin to posterior median region of carapace), right and left chelar carpus length (CCL, between the articulations of the merus and propodus), and chelar carpus height (CCH, between the dorsal and ventral margin, measured in the mid of the article) (Fig. 2).

Sexual dimorphism
The occurrence of sexual dimorphism in C. major was evaluated by comparing biometric variable values (CL, CCL and CCH) between sexes using a t-test (Sokal and Rohlf 2003), since distribution was normal and variances of the two groups being compared (α= 0.05) were homogenous.

Relative growth and morphological maturity
Carpus measures (CCL and CCH) were used as dependent variables, each one related to CL, which was considered an independent variable (body size).So, the relative growth and morphological maturity of C. major were estimated according to two biometric relationships (CCL vs. CL and CCH vs. CL), always considering the measurements of both chelipeds.Data obtained for each biometric relationship were fitted by a power function Y=aX b (Huxley 1950, Hartnoll 1974), followed by mathematical procedures indicated by Somerton (1980) and Somerton and Macintosh (1983), to estimate the size at onset of maturity in males and females, using the segmented package of the R software version 2.13.0 (R Core Team 2013).The constant "a" represents the intercept of the regression line on the y-axis, while the constant "b" represents the slope of the mathematic equation, and also the growth level of the dependent variable in relation to an independent one (isometry, b=1; negative allometry, b<1; positive allometry, b>1) (Hartnoll 1982).The fit of the regression line to data in each biometric relation was evaluated by a determination coefficient (R 2 ) based on logarithmic values of the variables.Student t-tests were employed to verify growth level, verifying the contrast of "b" from unit, and the Snedecor F-test was used to confirm one or two growth regression lines (Sokal and Rohlf 2003, α=0.05).Ghost shrimps with missing claws or limbs were excluded from the allometric analyses.

Handedness and chelar types
We explored whether the larger cheliped is carried at the right or left side of the body using a new methodology different from those commonly used in the literature (see Davanso et al. 2016 and references therein).For each specimen, a handedness ratio (HR) of the chelipeds was established on the basis of the variable CCL, using the division of the size of the right chelar carpus by the size of the left chelar carpus (HR=CCLR/CCLL).Specimens were classified as homochelous (HO, when HR=1) or heterochelous (HE, when HR≠1), the latter emphasizing the handedness: right (RHE) or left (LHE) heterochely.To minimize misclassification, homochelous specimens were covered by the confidence interval represented by 0.90≤HR≤1.10,which confirmed a small difference between the right and left carpus sizes.With this classification, heterochelic specimens were recognized as right (HR>1.10) and left (HR<0.90)heterochelic.The heterochely level of each specimen was obtained by subtracting the largest from the smallest CCL value and the result was expressed as a percentage.The proportion of homo and heterochelous specimens was contrasted between males and females through the ontogeny, revealed by chi-square test (χ 2 ) on statistical significance level of 5%.

Relative growth and morphological maturity
The chelar carpus relationships (CCL vs. CL and CCH vs. CL) showed some variations regarding the relative growth in different development phases for the two sexes (see juveniles and adults in Table 1), with two linear models better fitted in each case than in a single model.
A positive allometry was detected in the CCL vs. CL ratio through the ontogeny of males (Fig. 3A), with a higher slope in juveniles (b=2.02;θ=63.7°)than in adults (b=1.42;θ=54.8°),differing by 8.9° between these regression lines.A different pattern was obtained in the CCH vs. CL ratio of males (Fig. 4A), where juveniles also grew in positive allometry (b=1.92;θ=62.5°),but in the adult phase the growth was classified as allometric negative (b=0.77; θ=37.6°).In the latter case, the slope of log-regression lines was reduced by 24.9°, i.e. three times more than the biometric relationship CCL vs. CL.In females, a positive allometric growth of the CCL vs. CL ratio (b=1.22;θ=50.7°) was detected during the juvenile phase (Fig. 3B), changing to negative allometric growth in the adult phase (b=0.34;θ=18.8°), with a slope reduction by 31.9° when these log-regression lines were compared.This same allometric pattern was obtained for the CCH vs. CL ratio (Fig. 4B) in both juvenile (b=1.37;θ=53.9°) and adult females (b=0.16;θ=9.1°), with a deflection of 44.8° between regression slopes.
Summarizing, changes in regression slopes between development phases (see Table 1) were detected in chelar growth pattern when used invariably between length or height of carpus of C. major.Otherwise, in adult females these two carpal measurements showed a reduced growth pattern indicating b-values of 0.34 (CCL) and 0.16 (CCH), even with a relative growth more evident to CCL than to CCH.
For the ratios CCL vs. CL and CCH vs. CL the breaking points between regression lines (juvenile and adult) were located at 14.9 and 16.3 mm CL and at 15.0 and 16.2 mm CL for males and females, respectively.Synchronicity was observed among the estimated sizes, independent of the biometric variable of the chelar carpus (CCL or CCH), but more pronounced in CCH due to the expressive difference between slopes (b-values) of the regression lines (juvenile and adult).
C. major had the most evident heterochely from 10 mm CL, especially in males, with the carpus of the major cheliped showing an expressive change in the growth rate between 15 and 16 mm CL.It is possible that cheliped differentiation begins a little before the size of maturity, but it is still subtle, and exacerbated in males after the inflection point (puberty moult) that marks the ontogeny of this species.DISCUSSION Females of C. major attained a larger body size than males, thus confirming the general trend reported in other callianassid shrimps (e.g.Biffarius filholi [A. Milne-Edwards, 1878], Devine 1966;Callichirus garthi [Retamal, 1975], Hernáez and Wehrtmann 2007; Lepidophthalmus siriboia Felder andRodrigues, 1993, Rosa-Filho et al. 2013;for exceptions, Trypaea australiensis Dana, 1852, Hailstone andStephenson 1961).One of the few experimental studies about behaviour of ghost shrimps indicated that mating success does not depend on male size, because in these species the male does not guard the female during the sexual encounter (Somiya and Tamaki 2017).Therefore, there is no sexual selection for a large body in males of callianassid shrimps, as is likely to be the case in C. major.Moreover, small body size and one hypertrophied cheliped in males are obviously advantageous for reproduction in C. major given the considerable number of ovigerous females (>40%) observed by us during several sampling months.
Aggressive interactions are common in C. major when two male individuals encounter each other within a gallery, so the development of larger chelipeds in species of this genus might give comparative advantages during territory defense or also during competition for sexual partners (Rodrigues and Höld 1990).In males of C. major, the growth of the cheliped carpus showed positive allometry in individuals larger than 15 mm CL (Figs 3A, 4A), resulting in larger ghost shrimps likely having stronger chelipeds.However, relative growth of the cheliped length in males had a positive allometric pattern in both ontogenetic phases, independent of the article considered (carpus or propodus).This pattern differed from C. garthi, in which the PL grew more in adults (Hernáez and Wehrtmann 2007), a fact that was unfortunately not evaluated by us or other studies on C. major (Alves-Júnior et al. 2014, 2018, Rosa et al. 2018).In females of C. major, the relative growth pattern did not differ between biometric relationships involving the propodus or carpus of the cheliped as a dependent variable.This fact was also reported in Lepidophthalmus louisianensis (Schmitt, 1935) and Lepidophthalmus sinuensis Lemaitre and Rodrigues, 1991 (Table 3), and has been associated with the fact that female callianassids usually invest less energy in chelar growth than males (Hernáez and João 2018), because the reproductive success of females depends on reaching a larger body size and not larger chelipeds (see Bauer 2004).Biometric relationships involving variables measured from cheliped articles with higher growth can be used with success in maturity size estimates.This fact was verified in our study with chelar carpus variables of C. major, when compared with studies in which PL was used.A longer carpus than propodus seems to be a common character in heterochelic chelipeds of the Callichirus species previously studied (see Fig. 6).In contrast, other genera in Callianassidae have a chelar propodus that is more developed than the carpus (Manning and Felder 1991).Moreover, according to Hyzny and Muller (2010), even in other genera with an elongated carpus (e.g.Trypaea Dana, 1852 andPodocallichirus Sakai, 1999), this article has never been so developed as in the Callichirus species.
The inflection point between two straight lines represents the size of puberty moult (morphological maturity), from which juvenile and adult phases can be potentially separated (Hartnoll 1974(Hartnoll , 1978)).Our results revealed maturity sizes at 15 mm CL (males) and 16 mm CL (females), which are greater than those found for C. major by Rodrigues (1985), in which PL was used as a dependent variable (11 to 14 mm CL, respectively) and by Alves-Júnior et al. 2018 (11.2 mm for males and 11.04 mm for females, based on major cheliped propodus length).Alternatively, we detected that male heterochely of C. major started from 10 mm CL, reinforcing the information reported previously by these authors.Furthermore, the greater difference between b-values (see Table 3: A-J =-0.60 to -1.15) was obtained with the biometric relationships involving the carpus of the cheliped, independent of the variable considered (length or height).These values were negative and smaller than those found for other callianassid species (see Table 3), which can be related to the morphological structure of Callichirus chelipeds, as observed in Figure 6.Heterochelous males were common in C. major (87%) from Gonzaga beach, as occurs with other species of the genus Callichirus (Manning and Felder 1986, Hernáez and Wehrtmann 2007, Hernáez et al. 2015).On the other hand, most females of C. major were homochelous (59.3%).This information is also original for this species, because Manning andFelder (1986, 1991) indicated similar cheliped sizes in juveniles and adult females, the latter previously considered 'almost homochelous' or having a 'subtly unequal size'.Females of C. major had a greater mean size (CL) than males, but showed chelipeds with smaller sizes and slightly heterochelic.Furthermore, both sexes of C. major did not show laterality of the chelipeds in relation to the carpus length, which is a peculiar characteristic in Callianassidae species: e.g.Lepidophthalmus louisianensis (=Callianassa louisianensis), studied by Felder and Lovett (1989); Biffarius filholi studied by Berkenbusch and Rowden (1998); Pestarella tyrrhena (Petagna, 1792) (=Callianassa tyrrhena), according to Dworschak (1998); and Lepidophthalmus sinuensis, reported by Nates and Felder (1999).
According to Rodrigues and Höld (1990), the hypertrophied chelipeds in males of C. major are often used to defend galleries against invasion from other shrimps from the same or opposite sex.This information and the presence of a high frequency of heterochelous males in relation to heterochelous females suggest an intense male-male competition for sexual partners in C. major, and therefore offers indirect evidence of polygamy in this species.Alternatively, a low frequency of adult males in C. major were homochelous, which might be related to male feminization caused by hormonal production of the ovarian part of the testis.According to Souza et al. (2017), the ovarian section of the testis in adult males of C. major does not appear to be functional, because there are no ducts for releasing oocytes.However, the same authors mention the existence of mature oocytes in the male gonad, so it is totally possible that the ovarian section is producing hormones that feminize the individual.
Summarizing, in C. major the chelar growth differed according to sex and ontogenetic phases (juvenile and adults), changing with the onset of morphological maturity (15-16 mm CL), when an evident sexual dimorphism was detected (e.g. higher chelar size in males, mainly in adults).Analysis of biometric variables in callianassids, based on the carpus of the cheliped (length and height), can sometimes be more effective during the estimation of morphologic maturity size.Two morphotypes were described in this species according to their chelar sizes (homo-and heterochelous), an attribute observed in both sexes, although heterochely was more frequent (and evident) in adult males.The individuals of each sex did not show laterality of the major cheliped (handedness).

Fig. 1 .
Fig. 1. -Dorsal view of male and female individuals of the ghost shrimp C. major (A).Geographical position of the study area (B).Scale bar = 1 cm.

Fig. 3 .
Fig. 3. -Biometric relationship CCL vs. CL, regression analyses of both development phases in males (A) and females (B), and size of morphological maturity.CCL, carpus length of the major cheliped; CL, carapace length; light points, juveniles; dark points, adults.

Fig. 4 .
Fig. 4. -Biometric relationship CCH vs. CL, regression analyses of both development phases in males (A) and females (B), and size of morphological maturity.CCL, carpus length of the major cheliped; CL, carapace length; light points, juveniles; dark points, adults.