Sexual Dimorphism and Reproductive Status of the Red Swamp Crayfish Procambarus clarkii

© 2020 Academia Sinica, Taiwan Open Access Sexual Dimorphism and Reproductive Status of the Red Swamp Crayfish Procambarus clarkii Katsuyuki Hamasaki*, Naoko Osabe, Sota Nishimoto, Shigeki Dan, and Shuichi Kitada Department of Marine Biosciences, Tokyo University of Marine Science and Technology, Konan, Minato, Tokyo 108-8477, Japan. *Correspondence: E-mail: [email protected] E-mail: [email protected] (Osabe); [email protected] (Nishimoto); [email protected] (Dan); [email protected] (Kitada) Received 15 December 2018 / Accepted 12 February 2020 / Published 12 March 2020 Communicated by Benny K.K. Chan The red swamp crayfish Procambarus clarkii is native to northeastern Mexico and the south-central USA, and it has expanded its distribution worldwide and negatively impacted the ecosystems in the invaded regions. The dynamics of the P. clarkii populations have been studied as the basis for the development of effective control measures against this invasive alien species. Adult males of P. clarkii exhibit a cyclical dimorphism between two sexual morphotypes; reproductive form I has large chelae and non-reproductive form II has small chelae. However, whether P. clarkii females have two sexual morphotypes and exhibit form alternation has not been resolved, and little is known about the degree of intra-sexual dimorphism of the chelae, even among males. We employed allometric growth analysis on the chelae dimensions of P. clarkii females and males that were collected from a small pond in Yokohama, Japan. Our analysis demonstrated the existence of form I, which has larger chelae, and form II, which has smaller chelae, in P. clarkii females and highlighted the intraand inter-sexual dimorphisms in the chelae of this species. The reproductive cycle of the population was successfully traced by the reproductive status of P. clarkii based on the occurrence patterns of each sexual morphotype; the form I crayfish occurred throughout the sampling period from April to December, while the occurrences of form I females and males were highly correlated, peaking in October. Our results suggested that alternation of sexual forms occurs in P. clarkii females. The ability to discriminate between the sexual morphotypes based on chelae allometric growth would allow us to evaluate the female reproductive status more easily and precisely in invasive P. clarkii populations. Key words: Form alternation, Reproductive ecology, Allometric growth, Sexual maturity, Invasive alien species. Citation: Hamasaki K, Osabe N, Nishimoto S, Dan S, Kitada S. 2020. Sexual dimorphism and reproductive status of the red swamp crayfish Procambarus clarkia. Zool Stud 59:7. doi:10.6620/ZS.2020.59-07. BACKGROUND Red swamp crayfish Procambarus clarkii (Girard 1852) (Decapoda; Cambaridae) is native to northeastern Mexico and south-central USA (Hobbs 1972); it inhabits various freshwater environments, including swamps and marshes that are periodically flooded and drained (Huner and Barr 1991). This species has been introduced in several states in the continental USA and many other countries in Asia, Africa and Europe for aquaculture purposes (Hobbs et al. 1989; Loureiro et al. 2015), and now its aquaculture industry is growing in the USA, China and Spain (Souty-Grosset et al. 2016). Procambarus clarkii has also been introduced outside its native range as prey for aquaculture organisms such as bullfrogs (Sako 1987; Kawai and Kobayashi 2005). Procambarus clarkii exhibits rapid growth rates, early maturation, year-round egg production, and extended maternal care in which hatchlings and juveniles are attached to the mother’s pleon (Suko Zoological Studies 59: 7 (2020) doi:10.6620/ZS.2020.59-07 1 © 2020 Academia Sinica, Taiwan 1953 1956 1961; Huner and Barr 1991; Paglianti and Gherardi 2004; Scalici and Gherardi 2007). Thus, the life history characteristics of P. clarkii may have a high potential for rapid increases in population size in new available habitats; indeed, P. clarkii successfully established self-sustaining populations after escaping into freshwater bodies from aquaculture ponds in much of the introduced range (Loureiro et al. 2015; SoutyGrosset et al. 2016). This crayfish species has also become a popular ornamental animal and has expanded its populations through releases of pets from aquariums in some regions (Chucholl 2011 2013; Souty-Grosset et al. 2016). Procambarus clarkii exerts negative impacts on the ecosystems of newly colonized environments through the consumption of aquatic plants and algae and predation on several aquatic species, including amphibians, molluscs, and macroinvertebrates, leading loss of biodiversity (Souty-Grosset et al. 2016). Additionally, native crayfish populations have been seriously damaged due to the crayfish plague caused by the parasitic oomycete derived from vector P. clarkii (Souty-Grosset et al. 2016; Martín-Torrijos et al. 2018). In regions invaded by P. clarkii, control measures such as trapping; biocontrol with indigenous fish predators; sterile male release techniques, which use sterile but sexually active males; and the use of pheromones as bait for traps have been applied to manage populations of this invasive alien species (Aquiloni et al. 2009 2010; Aquiloni and Gherardi 2010; Gherardi et al. 2011; Loureiro et al. 2015). To manage the populations of biological resources, information on population structure and dynamics is crucial; therefore, population ecology, such as growth and reproduction, has been extensively studied for P. clarkii populations in introduced regions (Scallici and Gherardi 2007; Alcorlo et al. 2008; Anastácio et al. 2009; Chucholl 2011). After reaching sexual maturity, crayfish males that belong to the family Cambaridae, which originated in North America, exhibit a cyclical dimorphism between the reproductive phenotype (form I) and the nonreproductive phenotype (form II); after breeding season, form I males moult to form II and then, upon the return of the breeding season, form II males moult and return to form I (Scudamore 1948; Stein 1976; Taketomi et al. 1990; Payne 1996; McLay and van den Brink 2016; Kawai 2017). Form I males have longer and wider chelae, hooks on the ischia of the third and fourth pereiopods for holding females during copulation, and more calcified copulatory pleopods (gonopods); form II males lack these morphological characters (Suko 1953; Stein et al. 1977; Taketomi et al. 1990 1996; Huner and Barr 1991; McLay and van den Brink 2016; Kawai 2017). In contrast to males, the form alternation of cambarid females has only been reported for some species in the genus Orconectes Cope 1872 (Wetzel 2002; Wetzel et al. 2005; Buřič et al. 2010a) and for the species Cambarus elkensis Jezerinac and Stocker 1993 (Jones and Eversole 2011). The classification of these Orconectes species has been updated (Crandall and De Grave 2017), and here we use the reclassified genus name Faxonius Ortmann 1905. In these Faxonius and Cambarus species, form I females exhibit larger chelae and wider pleons than the form II females. In P. clarkii, form alternation is evident in males (Taketomi et al. 1990), but it has been assumed that females do not exhibit form alternation (Oluoch 1990; Loureiro et al. 2015). Consequently, the reproductive phase of males has been assigned by identifying either sexual form I or II, whereas that of females is assigned by examining gonad maturation; active glair glands of the ventral pleon, uropods and telson; and the eggs, juveniles and egg-remains on the pleopods (Scallici and Gherardi 2007; Alcorlo et al. 2008; Anastácio et al. 2009; Chucholl 2011). Suko (1953), however, documented the presence of different sexual morphotypes with longer (form I) or shorter (form II) chelae and the occurrence of form alternation in P. clarkii females, although he did not show the actual measurements of the body parts. Thus, previous reports of intra-sexual dimorphism in P. clarkii females have been contradictory. If P. clarkii females exhibit different sexual forms, their reproductive status would be more easily assessed based on the chela morphology. The present study therefore aimed to elucidate the existence of different sexual forms in P. clarkii females by allometric growth analyses of some body parts with reproductive information, such as the gonad maturity condition and the presence of eggs and juveniles on the pleopods, and to highlight the intraand inter-sexual dimorphisms in P. clarkii. The reproductive status of a P. clarkii population was then evaluated by the occurrence of sexually active form I females and males. MATERIALS AND METHODS Crayfish samples In Japan, Procambarus clarkii was imported from the USA as bait for culturing bullfrogs in 1927, and approximately 20 individuals were stocked into a pond in Kamakura, Kanagawa Prefecture (Sako 1987); since then, this species has expanded its range to cover the entire country (Kawai and Kobayashi 2005; Kawai 2017). Procambarus clarkii is listed in the “100 of Japan’s Worst Invasive Alien Species” (Ecological Society of Japan 2002), and removal campaigns using page 2 of 14Zoological Studies 59: 7 (2020) © 2020 Academia Sinica, Taiwan fishing gear such as traps and nets have been practised in an effort to eradicate the populations (Nakata 2018). In the present study, we used P. clarkii samples that were collected during the period from April to December 2011 through removal campaigns at the public natural park in Yokohama, Kanagawa Prefecture, Japan (35°32'22–23"N, 139°34'44–46"E). The crayfish were eliminated from a pond (~1900 m2, maximum depth ~1.5 m) with fishing gears such as bait traps (mesh size, 9–13 mm) and scoop nets (mesh size, 4 mm). The number of days that were taken to collect the crayfish and the number of crayfish collected are summarized for each month in table 1. The total numbers of female and male specimens were 566 and 476, respectively. Crayfish measurements The crayfish samples were brought to the laboratory and stocked in the refrigerator at 4°C. The samples were sexed based on the morphology of the pleopods (Suko 1953) (see Fig. S1A, B). Several body parts of the intact specimens were then measured to the nearest 0.01 mm as follows (see Fig. S1C–E): for both sexes, the postorbital carapace length (POCL, from the edge of the eye socket to the posterior margin of the carapace) (Fig. S1C), and the propodus width (the widest part of the propodus) and length (from the tip of the propodus to the carpal joint) of the right and/or left chelipeds were measured (Fig. S1D); for males, the first gonopod length (from the base to the tip of the right gonopod) was also measured (Fig. S1E), while for the females, the pleon width (the widest part of the second pleonite) was assessed (Fig. S1C). Males were assigned as either form I or form II morphotypes based on the presence of hooks on the ischia of the third and fourth pereiopods and calcified whitish gonopods (see Fig. S2) with developed and cornified apical lobe (Taketomi et al. 1990 1996; Kawai 2017). Females were checked for eggs or juveniles on their pleopods. All male and female specimens were dissected, and the gonads were removed and weighed to the nearest 1 mg. Some minute gonads could not be removed from the specimens (36 females and 6 males). The gonadosomatic index (GSI), an effective tool for representing gonad maturity condition in crayfish species, was used and calculated as [gonad weight / body weight] × 100 (Alcorlo et al. 2008; Chucholl 2011; McLay and van den Brink 2016; Jin et al. 2019). In the present study, there were specimens that lost both chelipeds (13 females and 7 males) or one cheliped (54 females and 34 males), and form I crayfish with larger chelipeds may be heavier than same-sized form II, leading to bias in the calculations of the weight-based GSI values. Therefore, we calculated the GSI as [gonad weight / POCL3] × 1000. Alcorlo et al. (2008) reported that oocyte diameter increased with increasing GSI values in P. clarkii females. To examine the oocyte development using GSI values, the gonads of 14 females, with GSI values from 0.9 to 13.6, were preserved in 10% formalin solution. The maximum diameters of oocytes were then measured to the nearest 1 μm using a Nikon stereomicroscope (MZ-800; Nikon Corp., Tokyo, Japan) equipped with a digital camera and an image analysing system (Nikon Digital Sight and NIS-Elements software). The number of oocytes measured for each gonad specimen ranged from 153 to 291, and frequency distributions of the diameters of oocytes were illustrated for respective gonads. The carapace length (CL, from the tip of the rostrum to the posterior margin of the carapace) or the total length (TL, from the tip of the rostrum to the posterior margin of the telson) are often measured as body size of P. clarkii (Fig. S1C). To compare results of the present study, which employed POCL, to other studies that employed the CL or TL measurements, Table 1. Collection records of red swamp crayfish Procambarus clarkii from a small pond in Yokohama, Japan, in 2011 Month Number of days collecting crayfish Number of crayfish Female Female (ND) Male April 243 226 May 3 56 3 54 June 5 117 2 105 July 4 130 2 122 August 230 0 15 September 4 97 154 October 2 28 1 34 November 4 34 046 December 318 2 20 ND: The sexual maturity of some females could not be determined because of the loss of both chelipeds. page 3 of 14Zoological Studies 59: 7 (2020) © 2020 Academia Sinica, Taiwan the POCL, CL and TL were measured for additional specimens (50 males and 50 females) collected in October and November 2019 from the same pond as the 2011 collections. The linear regression equations used to convert POCL to CL and TL values were as follows (Fig. S3): CL = -0.5871 + 1.321 POCL (R2 = 0.9832, F1, 98 = 5726, P < 0.0001); and TL = 3.6814 + 2.4523 POCL (R2 = 0.9757, F1, 98 = 3935, P < 0.0001). In the present study, the statistical analyses were performed with R statistical software (R3.5.2; R Core Team 2018) at a 5% significance level. Intraand inter-sexual dimorphisms The intraand inter-sexual dimorphisms were evaluated based on an allometric growth analysis. The relative growth of the body dimensions to a reference dimension (POCL) was examined using the following allometric growth equation (Huxley 1932): y = axb, where x is POCL, y is the measurement for another body-part (chela propodus width and length, pleon width, and gonopod length), b is the allometric growth coefficient, and a is the initial growth constant. The relative growth patterns were defined as follows: b > 1 indicates positive allometric growth, or faster growth of y than x; b = 1 indicates isometric growth, or the same growth rate for y and x; and b < 1 indicates negative allometric growth, or slower growth of y than x. The parameters were estimated by applying a general linear model (lm function) to the log-transformed equation lny = lna + blnx. The right and left chelae were basically symmetrical, and some specimens appeared to have regenerating smaller right or left chelipeds, as shown in figure S4. Therefore, we selected the larger measurement between the right and left chelae as chela dimensions. Suko (1953) stated that in P. clarkii females, the reproductive form I has longer chelae than those of form II of the same body size. In the present study, two different morphotypes with larger or smaller chelae were distinguished by the allometric growth equations with the aid of the information about reproductive status, such as the gonad maturity condition represented by the GSI values and the presence of eggs and juveniles on the pleopods. The GSI values of the P. clarkii females increased with oocyte development, and the oocyte diameter particularly increased after the gonad reached the GSI value of 3, as shown in figure S5. Therefore, we considered females with GSI values > 3 to be candidates for the reproductive morphotype (form I). Additionally, one ovigerous female and five females carrying juveniles were found in the collected specimens. The chela propodus widths and lengths were plotted against the POCL in these potentially or actively reproductive females, with the exception of one ovigerous female that had lost both chelipeds, and are shown in figure S6. The allometric growth equations between the POCL and chelae dimensions were estimated as follows: chela propodus width, y = 0.0725x1.4043 (statistics between lny and lnx: n = 41, R2 = 0.8756, F1, 39 = 274.5, P < 0.0001); and chela propodus length, y = 0.2408x1.3752 (statistics between lny and lnx: n = 41, R2 = 0.9373, F1, 39 = 582.7, P < 0.0001). Our female samples appeared to have two morphological groups with relatively large or small chelae, and the chela dimensions of the potentially or actively reproductive females were scattered within the larger chela group (see the RESULTS section). Here, the above mentioned allometric growth curves of these potentially or actively reproductive females decreased in parallel with decreasing intercept values, so that the lowest data plots for these females just matched the adjusted allometric growth curves (see Fig. S6) (chela propodus width, y = 0.0633x1.4043; chela propodus length, y = 0.2184x1.3752); these adjusted allometric growth curves successfully discriminated between the larger and smaller chelae groups (see the RESULTS section). We assigned a female as form I when either the chela propodus width or chela propodus length were beyond the discriminant value calculated by substituting its POCL into the adjusted allometric growth equation or form II when both the chela propodus width and chela propodus length were below the discriminant values. To statistically infer the intraand inter-sexual dimorphisms in females and males, we applied three models that included a continuous explanatory variable (POCL) and a categorical explanatory variable (sexual morphotype (MT), form I or II; Sex, female or male) with the lm function as follows: model 1, lny ~ lnPOCL + MT or Sex + lnPOCL × MT or Sex; model 2, lny ~ lnPOCL + MT or Sex; and model 3, lny ~ lnPOCL, where y is the measurement for another body-part; then, we selected the best model with the lowest AIC value (Akaike 1973; Burnham and Anderson 2002). The models showed whether the allometric growth equation had different intercepts and slopes (model 1), different intercepts and the same slope (model 2), or the same intercept and slope (model 3) between the sexual morphotypes or sexes. When model 1 or 2 was selected, intraand inter-sexual dimorphisms are detected. The allometric growth analysis was also applied to the gonad weight and GSI values in each sex. Reproductive status To statistically evaluate the seasonal changes in the reproductive status of the P. clarkii population, a generalized linear model (GLM) (glm function with a Gaussian error distribution) was performed to evaluate page 4 of 14Zoological Studies 59: 7 (2020) © 2020 Academia Sinica, Taiwan the differences in GSI values (response variable) among the months (April–December) and the sexual morphotypes (form I and II) (categorical explanatory variables) of each sex. Next, the numbers of form I and II crayfish after reaching the size of onset of sexual maturity (females, 22 mm POCL; males, 21 m POCL) (see the RESULTS section) were summarized for the different body size groups (22–25.99 mm (female), 21–25.99 mm (male), 26–29.99 mm, 30–33.99 mm, 34–37.99 mm, 38–41.99 mm (both sexes) POCL) and months (April–December) (see Table S1 for the number of crayfish and figures S7 and S8 for the size frequency distributions of crayfish). Then, differences in the proportions form I (response variable) among the different body size groups and months (categorical explanatory variables) were evaluated for each sex using the GLM analysis (glm function with a quasibinomial family (logit link), taking into account the overdispersion of the data). All females with ≥ 22 mm POCL were form I in October, and all females and males were form I in the body size class with ≥ 38 mm POCL; these data were excluded from the analyses because reliable coefficients could not be estimated by the GLM analysis. The statistical significance of the explanatory variables in the GLM analysis was evaluated with an F-test using the Anova function (Type II) (Fox and Weisberg 2011). Pearson’s product moment correlation coefficient (r) was used with a t-test to evaluate the relationship between the overall monthly proportions of form I females and males. The sex ratio was calculated as the [number of males / number of total crayfish], and a binomial test was used to test the null hypothesis for the sex ratio in each month (H0; sex ratio = 0.5). RESULTS Intraand inter-sexual dimorphisms The females could be divided into two morphotypes with relatively large or small chelae by the discriminant allometric growth equations between the POCL and chela propodus width or length that were derived from the potentially reproductive and sexually active females (Fig. 1A, B). These females with larger or smaller chela should correspond to form I or II because potentially or actively reproductive females were included in the group with larger chela. Form I occurred in females that had grown to a POCL of 22 mm or greater. Among the three log-transformed allometric growth equation models applied for the chela dimensions, model 1 and model 2 were selected as the best (Table 2), indicating the existence of intrasexual dimorphism in the chelae of P. clarkii females. Model 1 was also the best for pleon width (Table 2) as the differences in the intercept and slope were small between the morphotypes (Table 2, Fig. 1C). Males were assigned as either form I or II based on their external reproductive characteristics, and form I occurred in males that had grown to POCL of 21 mm or larger (Fig. 2A, B). The form I and II males had relatively large and small chelae, respectively. As seen in the females, the allometric growth analyses detected intra-sexual dimorphism among the males, and there was distinct dimorphism in the chelae (Table 2, Fig. 2A, B) but small dimorphism in the first gonopod (Table 2, Fig. 2C). The chelae showed positive allometric growth (b = ~1.2–1.5) regardless of the sexual morphotype in both sexes, and the pleon widths and gonopod lengths showed approximately isometric growth (b = ~0.9–1.1) (Table 2). When the allometric growth models were tested for the chela dimensions between the sexes in each sexual morphotype, model 1 and model 2 were selected as the best (Table 3), indicating the existence of intersexual dimorphism in the chelae of P. clarkii. In the comparison of the chela dimensions among crayfish with the same POCL, the form I males were found to have wider and longer chelae than the form I females, the form I females had wider chelae than the form II males, and the form II males showed wider and longer chelae compared with the form II females after reaching the size at onset of sexual maturity (~21–22 mm POCL) (Fig. 3). The allometric growth of the gonad weight and GSI were best described by model 1 for both sexes (Table 2); however, these values varied widely in form I females (Fig. 4). In males, the gonad weight tended to increase with increasing body size, whereas the GSI values tended to decrease in larger males (Fig. 5). The gonad weight and GSI values of the males tended to be larger in form I than those in form II (Table 2), but the scatter plots of these values against the POCL largely overlapping between the form I and II males. Reproductive status The GSI values significantly varied among the months and sexual morphotypes for both females (month, F = 9.5279, d.f. = 8, P < 0.0001; morphotype, F = 157.77, d.f. = 1, P < 0.0001) and males (month, F = 26.697, d.f. = 8, P < 0.0001; morphotype, F = 28.316, d.f. = 1, P < 0.0001). The form I females showed higher GSI values than the form II females, and the form I females with higher GSI values were found in April and September (Fig. 6A). One ovigerous female (23.2 mm POCL) and five females carrying juveniles page 5 of 14Zoological Studies 59: 7 (2020) © 2020 Academia Sinica, Taiwan (26.8–37.7 mm POCL) were found in the specimens collected in April 24, 2011. The differences in the GSI values between form I and II males were small, and the GSI values tended to slightly decrease or vary less from April to July, then increase until September and decrease again until December (Fig. 6B). The proportions of form I crayfish significantly varied among the months and body size groups in the females (month, F = 3.1070, d.f. = 7, P = 0.0248; body size, F = 23.653, d.f. = 3, P < 0.0001) and males (month, F = 6.6543, d.f. = 8, P = 0.0003; body size, F = 12.0611, d.f. = 3, P < 0.0001). The proportions of form I crayfish tended to increase with increasing body size, and the largest group (≥ 38 mm POCL) comprised all form I females and males (Fig. 7). The monthly proportions of form I females and males were significantly correlated (r = 0.7743, t = 3.2369, d.f. = 7, P = 0.0143) (Fig. 8). The overall proportions of form I females and males fluctuated around 50% from April to September, increasing to 83–100% in October. The proportion of form I was higher in males (76%) than in females (50%) in November and decreased to the same level in both sexes in December (Fig. 8). The sex ratio fluctuated around 0.5, but significant unbalanced sex ratios that favoured the females and males were detected in April (P = 0.0161) and November (P = 0.0275), respectively (Fig. 8). The sex ratio appeared to be male biased in December, but this was not statistically significant due Table 2. Intra-sexual dimorphism evaluated with a general linear model for some body parts of red swamp crayfish Procambarus clarkii. Three models were applied: model 1, lny ~ lnPOCL + MT + lnPOCL × MT; model 2, lny ~ lnPOCL + MT; and model 3, lny ~ lnPOCL, where POCL is the postorbital carapace length and MT is the sexual morphotype (form I, reproductive type; form II, non-reproductive type). The coefficient estimates of the categorical explanatory variable including the MT were outputted for form I representing the change in the response variable relative to the baseline category (form II) Coefficient estimates Sex Response variable Model Intercept lnPOCL MT-Form I lnPOCL × MT-Form I AIC R2F df1df2P Female Chela propodus width 1-3.3995 1.5293 0.6040 -0.0737 -981.9 0.9486 3374 3 548 < 0.0001 2-3.3622 1.5175 0.3561 - -982.7 0.9485 5059 2549 < 0.0001 3 -4.6273 1.9418 - - -384.9 0.8474 3055 1 550 < 0.0001 Chela propodus length 1-1.2164 1.2489 -0.3876 0.1787 -1476.8 0.9642 4921 3 548 < 0.0001 2-1.3066 1.2775 0.2135 - -1461.0 0.9630 7151 2549 < 0.0001 3 -2.0651 1.5319 - - -919.0 0.9010 5004 1 550 < 0.0001 Pleon width 1-0.7177 1.0039 0.2013 -0.0573 -1988.9 0.9645 4976 3 549 < 0.0001 2-0.6882 0.9945 0.0087 - -1986.2 0.9642 7411 2 550 < 0.0001 3 -0.7190 1.0049 - - -1984.2 0.9640 14740 1 551 < 0.0001 Gonad weight 1-8.8117 3.4061 10.6352 -2.7842 1230.3 0.6235 283.7 3 514 < 0.0001 2-6.9706 2.8283 1.2888 - 1254.8 0.6037 392.3 2 515 < 0.0001 3 -12.1032 4.5324 - - 1426.9 0.4454 414.4 1516 < 0.0001 GSI 1-1.9039 0.4061 10.6352 -2.7842 1230.3 0.3782 104.2 3 514 < 0.0001 2-0.0628 -0.1717 1.2888 - 1254.8 0.3455 135.9 2 515 < 0.0001 3 -5.1955 1.5324 - - 1426.9 0.0841 47.37 1516 < 0.0001 Male Chela propodus width 1-3.9124 1.7184 1.2432 -0.2287 -618.8 0.9364 2283 3 465 < 0.0001 2-3.7314 1.6619 0.4815 - -611.7 0.9352 3361 2466 < 0.0001 3 -5.2405 2.1807 - - 20.0 0.7496 1398 1467 < 0.0001 Chela propodus length 1-1.8619 1.4885 0.5511 -0.0526 -921.2 0.9526 3116 3 465 < 0.0001 2-1.8203 1.4755 0.3760 - -922.2 0.9525 4674 2466 < 0.0001 3 -2.9988 1.8806 - - -231.0 0.7918 1776 1467 < 0.0001 First gonopod length 1-1.3167 1.0637 0.6268 -0.1782 -1341.8 0.9263 1977 3 472 < 0.0001 2-1.1779 1.0203 0.0334 - -1319.2 0.9224 2811 2 473 < 0.0001 3 -1.2825 1.0563 - - -1293.7 0.9178 5291 1474 < 0.0001 Gonad weight 1-5.4075 2.3708 2.7552 -0.6876 540.5 0.6405 276.8 3 466 < 0.0001 2-4.8662 2.2017 0.4652 - 545.5 0.6351 406.4 2467 < 0.0001 3 -6.3169 2.7008 - - 639.9 0.5520 576.6 1468 < 0.0001 GSI 11.5002 -0.6292 2.7552 -0.6876 540.5 0.2095 41.18 3 466 < 0.0001 22.0415 -0.7983 0.4652 - 545.5 0.1977 57.52 2467 < 0.0001 3 0.5908 -0.2992 - - 639.9 0.0149 7.076 1468 0.0081 The bold AIC value is the lowest among the three models for each body part. page 6 of 14Zoological Studies 59: 7 (2020) © 2020 Academia Sinica, Taiwan to the small sample size (n = 8, P = 0.7266). DISCUSSION Intraand inter-sexual dimorphisms Males of the American cambarid crayfish occur as either sexually active form I or sexually inactive form II morphotypes based on the presence of hooks on the ischia of the third and fourth pereiopods and more calcified rigid gonopods (Suko 1953; Stein et al. 1977; Taketomi et al. 1990 1996; Huner and Barr 1991; McLay and van den Brink 2016; Kawai 2017); form I males have longer and wider chelae than form II males (Suko 1953; Stein et al. 1977; Huner and Barr 1991; McLay and van den Brink 2016; Buřič et al. 2010b). Sexually active females (form I) of the American cambarids have been identified based on the presence of developed glair glands, oocytes in the gonopores, and/or eggs and juveniles on the pleopods (Wetzel 2002; Buřič et al. 2010a; Jones and Eversole 2011), the architecture of annulus ventralis (Wetzel et al. 2005; Jones and Eversole 2011) and the occurrence of mating with 0 2 4 6 8 10 12 14 10 15 20 25 30 35 40 45 Propodus width (mm) (A) Form I Form II Females with GSI values > 3 Females carrying juveniles 0 5 10 15 20 25 30 35 40 45 10 15 20 25 30 35 40 45 Propodus length (mm) (B) 0 5 10 15 20 25 10 15 20 25 30 35 40 45 Pleon width (mm) Postorbital carapace length (mm) (C) y= 0.0633x 1.4043 y= 0.2184x 1.3752 Fig. 1. Growth of the chela propodus width (A), chela propodus length (B) and pleon width (C) relative to the postorbital carapace length in female red swamp crayfish Procambarus clarkii. Data are shown for the different sexual morphotypes, reproductive form I and non-reproductive form II, the potentially reproductive females with GSI values > 3 and the reproductive females carrying juveniles. Allometric growth curves discriminating the different sexual morphotypes are shown for chela propodus width and length. See figure S6 for discriminant functions. Fig. 2. Growth in the chela propodus width (A), chela propodus length (B) and first gonopod length (C) relative to the postorbital carapace length in male red swamp crayfish Procambarus clarkii. Data are shown for the different sexual morphotypes, reproductive form I and non-reproductive form II. 0 2 4 6 8 10 12 14 16 18 20 10 15 20 25 30 35 40 45 Propodus width (mm) (A) Form I Form II 0 10 20 30 40 50 60 10 15 20 25 30 35 40 45 Propodus length (mm) (B) 0 2 4 6 8 10 12 14 16 10 15 20 25 30 35 40 45 First gonopod length (mm) Postorbital carapace length (mm) (C) page 7 of 14Zoological Studies 59: 7 (2020) © 2020 Academia Sinica, Taiwan Table 3. Inter-sexual dimorphism evaluated with a general linear model for the chelae of red swamp crayfish Procambarus clarkii. Three models were applied: model 1, lny ~ lnPOCL + Sex + lnPOCL × Sex; model 2, lny ~ lnPOCL + Sex; and model 3, lny ~ lnPOCL, where POCL is the postorbital carapace length and Sex is female or male. The coefficient estimates of the categorical explanatory variable including the Sex is outputted for male representing the change in the response variable relative to the baseline category (female) Coefficient estimates Form Response variable Model Intercept lnPOCL Sex-Male lnPOCL × Sex-Male AIC R2F df1df2P IChela propodus width 1-2.7954 1.4556 0.1263 0.0342 -814.2 0.8994 1025 3 344 < 0.0001 2-2.8620 1.4751 0.2421 - -815.8 0.8993 1540 2345 < 0.0001 3 -2.4061 1.3777 - - -368.9 0.6341 599.6 1346 < 0.0001 Chela propodus length 1-1.6040 1.4276 0.2932 0.0083 -1049.0 0.9528 2316 3 344 < 0.0001 2-1.6202 1.4324 0.3214 - -1051.0 0.9528 3483 2345 < 0.0001 3 -1.0149 1.3030 - - -247.4 0.5223 378.3 1346 < 0.0001 II Chela propodus width 1-3.3995 1.5293 -0.5129 0.1892 -871.4 0.8662 1444 3 669 < 0.0001 2-3.6608 1.6119 0.0894 - -859.6 0.8635 2119 2670 < 0.0001 3 -3.6928 1.6340 - - -785.2 0.8471 3716 1671 < 0.0001 Chela propodus length 1-1.2164 1.2489 -0.6455 0.2396 -1381.3 0.9107 2273 3 669 < 0.0001 2-1.5473 1.3536 0.1173 - -1337.3 0.9043 3167 2670 < 0.0001 3 -1.5893 1.3826 - - -1103.1 0.8641 4266 1671 < 0.0001 The bold AIC value is the lowest among the three models for each body part. Fig. 3. Inter-sexual dimorphism of the chela propodus width (A) and chela propodus length (B) in red swamp crayfish Procambarus clarkii. Allometric growth curves were calculated based on the equations estimated for females and males in the respective sexual morphotypes shown in table 3. Fig. 4. Relationships between the postorbital carapace length and gonad weight (A) and the gonadosomatic index (GSI) (B) in female red swamp crayfish Procambarus clarkii. Data are shown for the different sexual morphotypes, reproductive form I and nonreproductive form II. 0 2 4 6 8 10 12 14 16 18 20 10 15 20 25 30 35 40 45 Propodus width (mm) (A) Females: form I Females: form II Males: form I Males: form II 0 10 20 30 40 50 60 10 15 20 25 30 35 40 45 Propodus length (mm) Postorbital carapace length (mm) (B) 0 100 200 300 400 500 10 15 20 25 30 35 40 45 Gonad weight (mg) (A) Form I Form II 0 5 10 15 20 25 10 15 20 25 30 35 40 45 GSI Postorbital carapace length (mm) (B) page 8 of 14Zoological Studies 59: 7 (2020) © 2020 Academia Sinica, Taiwan males (Wetzel 2002; Buřič et al. 2010a); then, the body dimensions such as chela size and pleon width were compared between form I and form II females. Wetzel et al. (2005) documented that form I and form II females of Faxonius pardalotus (Wetzel, Poly and Fetzner 2005) exhibited different morphologies of annulus ventralis: form I females have a corneous sternum and strongly convoluted structures, whereas those of form II females are less cornified and convoluted. Kawai (2017) found these two characters in the annulus ventralis of P. ckarkii and suggested the existence of form I and form II morphotypes in P. clarkii females. In the present study, based on the documentation by Suko (1953) that different sexual morphotypes with longer (form I) or shorter (form II) chelae were found in P. clarkii females, we attempted to elucidate the existence of different sexual forms in P. clarkii females by allometric growth analyses of some body parts with reproductive information, such as the gonad maturity condition and the presence of eggs and juveniles on the pleopods. Females could be assigned into two morphotypes with relatively large or small chelae, corresponding to form I or II because potentially or actively reproductive females were included in the morphotype with larger chela. We also identified the sexual morphotypes of males based on the external morphological characteristics, and conducted allometric growth analyses of some body parts. Our results confirm the existence of different sexual morphotypes in P. clarkii females and highlight the intraand inter-sexual dimorphisms in the chelae of this species (Figs. 1–3); the chelae were largest in the form I males, followed by the form I females and form II males, and were the smallest in the form II females (Fig. 3). To further elucidate the reproductive characteristics of form I and form II females of P. clarkii, the architecture of annulus ventralis should be examined in relation to the sexual morphotypes revealed by the present study and confirm the occurrence of mating of these females with form I males. Suko (1953) illustrated that the relative growth patterns between the TL and chela propodus length of P. clarkii females and males collected from Urawa (35°52'N, 139°35'E), Saitama Prefecture, Japan are similar to those listed in figure 3 of the present study, while the actual measurements and allometric growth coefficients are not shown. Suko (1953) stated that the form I females and males occurred from 56 mm TL (21.3 mm POCL) and 54 mm TL (20.5 mm POCL), respectively, which corresponded to the size at onset of sexual maturity of the males (21 mm POCL) and females (22 mm POCL) in the present study. To evaluate the reproductive potential of the decapod crustacean populations, the size at which 50% of the animals reached maturity was evaluated previously (Pescinelli et al. 2016; Waller et al. 2019). In P. clarkii, form alternation is evident in males (Taketomi et al. 1990), and it might also occur in females as discussed below, so that in our crayfish specimens, the form II crayfish might include the subadults that have never moulted to the sexually active form I and adults that have moulted and returned to the sexually inactive form II. To determine the size at which 50% of the crayfish are mature in P. clarkii populations, the characteristics of the subadults and form II adults need to be clarified. Stein (1976) suggested that chelae of crayfish may be used for (1) prey capture and manipulation, (2) defence against predators, (3) interand intraspecific interactions, and (4) reproductive activities. Faxonius propinquus (Girard 1852), which have large chelae, are less vulnerable to predation by fish and superior competitors in intraspecific interactions (Stein 1976). Ueno and Nagayama (2015) reported that large P. clarkii with smaller chelae were beaten by small ones with larger chelae. Thus, selection might favour large chelae in P. clarkii, resulting in the positive allometric growth of the chelae regardless of sex or the sexual morphotype, as demonstrated by our allometric growth analyses. The observation that form I cambarid males have large chelae suggests that the chelae of sexually 0 10 20 30 40 50 60 10 15 20 25 30 35 40 45 Gonad weight (mg) (A) Form I Form II 0 0.5 1 1.5 2 2.5 3 10 15 20 25 30 35 40 45 GSI Postorbital carapace length (mm) (B) Fig. 5. Relationships between the postorbital carapace length and gonad weight (A) and the gonadosomatic index (GSI) (B) in male red swamp crayfish Procambarus clarkii. Data are shown for the different sexual morphotypes, reproductive form I and non-reproductive form II. page 9 of 14Zoological Studies 59: 7 (2020)