Freshwater Biology (\990) 23, 491-503

Growth and reproduction in the freshwater crah,Potamon fluviatile (Decapoda, Brachyura)

FIORENZA MICHELI, FRANCESCA GHERARDI and MARCO VANNINIDepartment of Animal Biology and Genetics, University of Florence, Italy

SUMMARY. 1. The patterns of relative and absolute growth and thereproductive biology of the freshwater crab Potamoti fluviatile were stud-ied in a natural population inhabiting a hill stream close to Florence, Italy,over an annual cycle. Periodical inspections of a stretch of the stream weremade and morphological and anatomical analysis carried out.

2. As in other decapods, the females are smaller than the males (they canhowever occasionally reach larger dimensions). This may be determinedby a lengthening of the moulting interval in the females, by their higherenergetic cost of reproduction (since reproduction occurs simultaneouslywith moulting, at least in this habitat) and/or by a higher mortality rateresulting from the risks associated with carrying eggs and vagility.

3. The analysis of relative growth in secondary sexual characters(abdomen width and major chela length) with respect to carapace length,shows that the pre-puberty and puberty ecdysis occur at 15 and 35 mmcarapace length respectively, which was also confirmed by the gonadweight, vas deferens in males, and the onset of vitcllogenesis in females.

4. A delay between anatomical and functional (i.e. the ability to copul-ate successfully) maturity was observed in the males. Being larger may bean advantage in intra-sexual competition for mating, but larger males,being more vagile. are also more likely to meet receptive females. Thefemales may mate before their puberty moult and store sperms in iheirseminal receptacles for when they attain full maturity; this could be adapt-ive since opportunities of encountering males are few and far between intheir adult phase, characterized by their vagile and amphibious habits.

Introduction

If the study of relative growth (that is the changeof shape as animals grow) throws light on the lifehistory of a given species, information on theabsolute growth patterns (determined by the

Correspondence: M. Vannini, Depanmenl of AnimalBiology and Genetics. University of Florence, V.Romana 17. 50125 Florence. Italy.

moult increment or growth factor and the inter-moult period or moult interval; Hartnoil, 1982)can undoubtedly heip analysis of the populationstructure. Though relative growth is obviouslynol an exclusive feature of crustaceans, the pre-sence of a rigid integument and thus the pos-sibility of accurate measurements, together withthe discontinuous growth pattern, make thisgroup an ideal object for allometric studies(Teissier. 1960; Hartnoil. 1982).

491

492 F. Micheli. F. Gherardi and M. Vannini

The most relevant changes in shape (apartfrom those connected with larval life) take placewith the attainment of sexual maturity, oftenaccomplished at one particular moult, thepuberty moult (Perez. 1928), after which chelae,abdomen and pleopods (the secondary sexualcharacters) are seen to follow differential growthrates (Hartnoli, 1978).

In some eases, predictions of the attainment ofsexual maturity made on the basis of mor-phometrical analysis have been confirmed byanatomical (Finney & Abeie, 1981)or histologi-cal (Haley. 1969. 1973) examinations. Only afew works examine growth patterns as well asreproductive cycles (Haley, 1969; Finney &Abele. 1981; Vannini & Gherardi, 1988).

Scattered references exist on populationstructure and allometry of the secondary sexualcharacters for river crabs, in Potamon gedro-sianum Alcock from Afghanistan (Schneider,1971), and Geotetphusa dehanni White fromJapan (Yamaguchi & Takamatsu. 1980). or tothe attainment of sexual maturity, inBarytetphusa guerini Milne Edwards fromIndia (Gangotri, Vasantha & Venkatachari,1978). Data on the population structure ofPotamon fltiviatite Herbst have already beenpublished (Gherardi, Guidi & Vannini, 1987).In the present work, the patterns of absolute andrelative growth are considered, together withgravimetric and histological analysis of thegonads, which allows identification of thepuberty moult. The interaction between moultand reproductive cycles and their effects on thepopulation structure are also discussed.

Materials and Methods

Potamon fluviatile is the only representative ofthe Potamoidea in Italy, occurring from Liguriaand Tuscany southwards, Sicily included(Gherardi ef at., 1987). Fieldwork was con-ducted along a hill stream near Florence (170 mabove sea level, mean width approximately 2 m,depth ranging from 10 to 90 cm, though greatlyvariable with the seasons).

In order to determine the timing of the mainevents of the life cycle, a 150 m stretch of thestream was periodically inspected from July toOctober 1986. No breeding females or moultinganimals (except some juveniles) were seen dur-ing all the other periods of the year.

During these inspections, the following mor-phometrical data were taken: carapace length(CL) and width, width of the sixth abdominalsegment, and length, width and height of chelae(Fig. 1). Measurements were also made of newlymoulted Individuals, including exuviae, whenthey were found near the moulted animal.Ovigerous females and mating pairs were alsorecorded. Other morphometrical data, i.e.abdomen and gonopod lengths, were taken fromanimals sacrificed for gonad examination.

About thirty-five specimens for dissectionwere collected from the stream once a monthduring 1986-87 and transferred to the labora-tory, where they were killed with chloroformwithin a few hours and weighed, The weight ofthe gonads from males and females, the colourof the ovaries and mean oocyte diameter (calcu-lated from the measurements of about ten ran-domly chosen oocytes under the binocularmicroscope) were recorded. Samples of ovariesand seminal receptacles were fixed in Camoy'sfluid and embedded in polystyrene resin. Sec-tions (7 fim thick) were stained using thehaemalum--eosin technique in order to check forthe presence of sperm in the seminal receptaclesor of yolk inside the oocytes. As a criterion ofanatomical sexual maturity in mates, the pre-sence of sperm in the vas deferens was adopted.Deferens tubules (from either freshly dissectedor fixed gonads) were dipped in distilled waterand broken with forceps, then examined underthe microscope.

Sex-ratio at birth was determined in fivebroods. About 50% of specimens from eachbrood were examined under a binocular micro-scope: buds of four pairs of pleopods are visiblein newly hatched females, whilst there are onlytwo in males.

The text gives mean values ± standard error.

ResultsAbsolute growth

Ecdysis was principally restricted to earlyautumn (Fig. 2); animals in the early instarsprobably moult more frequently (soft individ-uals smaller than 20 mm CL were found all yearround, especially in autumn and spring).

The correlation between per cent increment atmoult (i.e. the percentage by which the posl-moult length exceeds the pre-moult length;

Growth and reproduction in a freshwater crab 493

AL

C W

FIG. 1. A diagram of P. ftuviatite showing some of the measurements taken (CL-carapace length; CW=carapacewidth; AL=abdomen length; AW = abdomen width; PL=claw propodite length; PW=daw propodiic width;PH=claw propodite height).

%100-1

M -

to-

JULV AUGUST SEPTEMBER

FIG. 2. Time of moulting (per cent of moulting ani-mals found is transects along the stream from June toSeplember.

Mauchline, 1976) and CL was not quite signifi-cant (males; meanvalue=I6.1±1.5%;r=0.466,df=IO, ns; females; mean value=15.5±1.12%;/-=0.338. df=n, ns; males v. females; f=0.32,df=23. ns).

A significant difference existed between thesize of moulting and hard animals (Fig. 3), malesof the former more frequently being smallerthan 25 mm CL (/ = 9.757. df=2, /'<0.0I),whilst moulting females fell more in the range of25-35 mm CL (x^=^4.\21, df-2. /'<0.001).Moult frequency dropped dramatically after35 mm CL, with no difference between the

10-

25-3SSIZE RA^GE

•MALES84

FIG. 3. Size-frequency distribution of moulting malesand females.

sexes ix^=\.24, d f = l , ns). The difference be-tween sexes was not signficant for animalssmaller than 25 mm CL either (/ ' = O.I04.df= I.ns), although it was so in the size range 23-35mm (/ '=5.247, df- 1, P<0.05).

494 F. Micheti, F. Gherardi and M. Vannini

TABLE I. Paitcms of relative growth, /'=rcgrcssion coefficient (•/'<0.05; **/'<O.OI), (CL = carapacc length;CW^carapacf widlh: MPL = major claw propodite length; mPL-tninor claw propodile length; MPW = major clawpropodile width; mPW = minor claw propodite width; MPH=major claw propodile height; mPH — minor clawpropodite height; AL^-abdomen length; AW = abdomen width; BWt = body weight,)

Males Females

In (CL) v.n (CW)In (MPL)In (mPL)In (MPW)In (mPW)In (MPH)In (mPH)n (AL)

In AW)ln (BWt)

N i343229228231231231231

8206 (

16 :

.023

.262,176.364.205.314.140.570

).935J.836

(b*l)10.97*35.79*25,11*32,48*16.80*31.57*14.72*6.84*4.63*

38.22*

N30515315515515515415411

23515

i,025.217.152.411,179.354,11.014.928

i.693

(69*1)6.88* •

15.39*'4.09*"

17.73**7.39**

10.16**5.66**0.26

32.27**17.60"

0.372.90**0.811.951.021.361.435.70**

32.36"1.25

In (CL) v:MPL/mPL

N226

b0.085 16.35*

N153

b0,066 1.82 0.68

TABLE 2. Relative growth of pleopods

In (CL) v:In (female pleopods)1st cndopod.

cxopod.2nd endopod.

cxopod.3rd endopod

cxopod.4tl) endopod.

exopixl.

In (male pleopods)1st ploepod.2nd pleopod.

df

88887878

1915

b

1.3852.1121.6172.1821.5131.8061.5631.980

1.1610.897

(&#1)

3.I69**n.644**5.268«

11.491 ••2.952•2.650*3.545*"

IO.741'»

2.1011.253

The total number of moulting animals, com-pared with that of the non-moulting ones, didnot differ significantly between the sexes (males:27.38%; females: 3 1 % ; males v. females:f={).\A.6i=\. ns).

Adults do not moult every year: of the animalsmarked in autumn 1985, by gluing plastic tagsonto their carapaces (see Gherardi et at.. 1987)and captured the following year slill carryingtheir markers. 8% of males and 10% of femaleshad not moulted in autumn (all of them werelarger than 30 mm CL).

Relative growth

Carapace width, length, width and height ofboth chelae, degree of heterochely. abdomenlength and width, total body weight, and lengthof female pleopods all showed allometric growthrelative to CL (Tables 1 and 2). Bilogarithmictranformation was used (Huxley, 1932; Hartnoll,1974). and CL was referred to as the independentvariable (Gray & Ncwcombe, 1938).

Positive allometric growth of carapace widthversus CL was revealed with no difference

Growth and reproduction in a freshwater crab 495

TABLE 3. Growth rate of major claw length (r In of carapace length) in diflercm instars (•/'<0.05; • ' / '<0.0I)In (major claw length)

Juveniles(<I5 mm)

Sub-adults(15-35 mm)Adults(>35 mm)

Juv. V. Sub-ad.Sub-ad. V. Ad.

N

42

Males

N

53

159

b

1.075

b

1.231

1.541

Mates;

4.85*'9.31"*

(6^1)

2.27'

17.71'*

17.76* •

df93

210

Females

N

35

113

b

1.159

1.471

(ft^fel)

5.17**

7.56"

Females

1.814.46"

(cfi?)

2.45*

1.11

df75

146

between males and fctnales, and in the length ofthe minor chela, the width and height of bothchelae, and body weight (Table 1). On the con-trary, the length of the major chela differed sig-nificantly in its level of allometry in the sexes(Table 3), growing more rapidly in males(f=2.90. P<0.01). with no difference betweenthe growth rate for right- (over 90% of the entirepopulation; Gherardi et al., 1987) or left-handedindividuals (/=0.076. df= 148. ns).

The allometric growth of the major chela wasnot constant (Fig. 4 and Table 3): the slope of theregression line increases at 15 and 35 mm CL inmales, and at 3.* mm in females. The ccdysesoccurring at these sizes are assumed to corre-pond to the pre-puberty (otiiy shown in themales) and puberty moults, respectively, thusdemarcating three phases; juvenile (CL<15 mm), when individuals are still sexuallyundifferentiated; sub-adult (CL ^\5 mm and<35 mm), when sexual dimorphism becomesapparent: and adult (CL 3:35 mm), when theanimals attain sexual maturity. The abbrevia-tions J. SM. SF, AM and AF will be used in thetext to refer to juveniles, sub-adults and adults,males and females respectively.

The abdomen showed the most evidentgrowth differences between the sexes (Fig. 5 andTable 4). The female abdomen began to widenfaster at 15 mm CL. its growth rate slowingdown after 35 mm. In males, a variation in slopewas seen at 15 mm, with no difference in the

mm50-1

20-10-

20 30 50

20-

10-

15

n=153 Q

>

35

•

I 20 30 I15 35

CARAPACE LENGTH mm

I50

FIG. 4, Allometric growth of major claw length inmales (top) and females (bottom) versus carapacelength. The discontinuities of the curve mark the pre-puberlal (males, at \^ mm CL) and pubcrtal (malesand females, at 35 mm (CL) moults. Plotted onlogarithmic scale.

496 F. Micheli, F. Gherardi and M. Vannini

TABLE 4. Growth rate of abdomen width (v. In of carapace length) in different instars (•/ '<0.05: ' •P<0.01)

ln (abdomen width)

Juveniles(<15 mm)

Sub-adults(15-35 mm)Adults(>35 mm)

Juv. V. Sub-ad.Sub-ad. V. Ad.

N

27

Males

N

80

148

b

1.126

b

0,871

0.951

Males(2.02*1.13

0,90

(£.#1)

3.13'

1.05"

df105226

Females

N

90

145

b

2.058

1.363

f

2 0 . 1 8 "

3 . 1 6 "

Females

t7 .97"5.66*'

t

16.82"

3.54"

df115233

mm

zO 20-1

Z 10-1UlIa

= 497

10 15 30 35 50CARAPACE LENGTH mm

FIG. 5. Allometric growth of abdomen width versuscarapace length, Male and female curves separate incorrespondence to pre-pubertal moult (15 mm CL).while discontinuity in the female curve (top) marks thepubertal moult (35 mm CL). Plotted on logarithmicscale.

level of allometry between the S and A instars.In fetnales, the abdomen length exhibited a posi-tive allometry both in S and A instars (S:/=5.341, df=5. P<O.U1; A: r=4.988, df=55./'<0.001), though in the latter growth becamesignificantly slower (S v. A: /=2.654. df-55,/'<0.02). In males, abdomen length is isometri-cal in S. while in A it showed a negative allo-metry, but the change in slope was notsignificant (S: (=1.935, df=2. ns; A: (=5.156,df=42. P<0.001; S v. A: /=O.5l6. df=42. ns).

The growth of male gonopods (both first andsecond pairs) did not differ from isometry (Table

2). Gonopods of the first pair, which arepositively ailometric in immature crabs of otherspecies (Hartnoll, 1974), were isometric in bothS (r=1.317, df=2, ns) and A males (r=1.274,df=15, ns), and no change occurred at thepuberty moult (S v. A: r=0.346, df=15, ns).Female pleopod exopodites (considering theaverage length of the four pairs) grew faster,relative to CL, in S (/=8.823, df=4, P<0.001)and isometrically in A (r=0.905, df=l3, ns),with a significant difference between these twocategories (/=8.588.df= 13. P<O.(K)1).

With respect to abdomen length (as a struc-ture, the abdomen carries and covers pereopods),the exopodites showed a positive allometricgrowth in S (?=5.593. df=4, P<0.0\) and anisometric one in A (/= 1.506, df= 12, ns; S v. A:f=5.I47, df=12, P<0.00\). Pleopod etidopo-dites (mean values) grew isometricatty withrespect to abdomen length, with no change at thepuberty moult. Male first pleopods showed afairly positive allometry with respect to abdo-men length (( = 2.146. df - l9 , P<0.Q5), whilethe growth of the second male pleopods wasisomeirical (/=O.6I4. df= 15, ns).

Anatomical sexual mattirityGonad weight. Fig. 6 shows the net weight of

the reproductive tract of males and females as afunction of CL. In both sexes there was anabrupt change in the slope of the regression lineat about 35 mm CL (SM: r=0.894, fc=0.052,

Growth and reproduction in a freshwater crab 497

aio430

S

<oo

g1.0^

0.50-1

0.10-0.05

30 45 mm

CARAPACE LENGTHFIG. 6. Correlation belween body size (carapace length) and gonad weight (wet) in males (top) and females (bottom).

df=4, P<0.02: AM: r=0.861. &=0.26, df=14.P<0.001;SM v. AM: r=2.706. df=14,/'<0.02;SF: r=0.724, 6=0.013, df=3, ns; AF: r=0.643.6=0.088, df=13, P<0.01; SF v. AF: /=2.652,df=13, P<0.02). In S, gonad weight exhibited aslight positive allometry in males, while femaleovaries were inconspicuous until the onset ofsexual maturity. Weight increased rapidly inboth sexes in the larger classes.

Vas deferens. The anatomical subdivision pro-posed by various authors (Cronin, 1947; Ryan,1967; Hinsch & Walker. 1974} of vas deferensinto anterior, medial and posterior portions wasadopted. Classification was made on the basis ofthe external appearance and position of the def-erential tracts without histological examination.The presence of sperms was checked in the ante-rior and medial regions.

Examination of the vasa deferentia from six-

teen males (ranging from 25.4 to 50.5 mm CL)revealed the constant presence of sperms in ani-mals over 35 mm, while they were never foundin S individuals.

Viteilogenesis

At the onset of viteilogenesis, the originallywhite, thin ovaries acquire a creamy colour andstart to grow as a consequence of the depositionof yolk inside the oocytes. As this process con-tinues, the ovary turns yellow in colour and.finally, bright orange. Oocytes on the externalsurface of the ovary started to colour at about0.5 mm diameter. This agrees with the classifica-tion of vitellogenic stages proposed by Adiyodi(1968). In our histological sections of ovaries,minute eosinophilic granules were present at theperiphery of the oocytes over 0.5 mm in dia-

498 F. Micheli. F. Gherardi and M. Vannini

100-

5 0 -

0 - "

100-I

EGG-CARRViNG FEMALESn = 156

20 15 22 30 7 13 21

0 -

AUGUST SEPTEMBER OCTOBER

BROOD-CARRYING FEMALESn = 156

AUGUST SEPTEMBER OCTOBERFIG. 7. Time of spawning (lop) and hatching (bottom) (per cent of egg- and brood-carrying females, respectively,founded in transects along the stream from July to October),

meter and acidophily increased with cell size.Ovaries with oocytes larger than 0.5 mm onaverage were thus considered as vitellogenic.

All of the ten females measuring 26-35 mmCL examined had immature gonads, whilst onlyone of the sixty-six adults (ranging from 35 mmto 46 mm) had a non-vitellogenic ovary.

Functional sexual maturityOviposition and hatching. Egg-bearing

females were found in the stream from late Julyto early August (Fig. 7). Uke all freshwatercrabs studied to date (Koba. 1936; Schneider,1971; Minei, 1976; Yamaguchi & Takamatsu.1980; Pillai. 1982), and as already reported forthis species (Mercanti, 1885; Pace, Harris &Jac-carini. 1976), hatehing occurs in the pouch lim-ited by the mother's abdomen. Brood-carryingfemales were found in the second half of August

and the first half of September. The distancebetween the peaks of the diagram in Fig. 7 cor-responds to 1 month approximately but no directmeasurement of the incubation period wasmade. Pace et al. (1976) indicated a duration of45±2 days at 23°C for this species in Malta. Allbreeding females found at stream inspections(26.4% of total number of females) were in theinstar indicated here as A. and the size distribu-tions of breeding and non-breeding individuals

TABLE 5. Size frequencies of breeding and non-breeding females found along stream transects fromJuly to September 1986

Size range(CL, mm)<353=35 and <40•^40

No. of femalesBreeding0s

20

Non-breeding411225

Growth and reproduction in a freshwater crab 499

were significantly different (;f^=24.0, df=2,/ <(1.(IO1; Table5).

Although on the basis of morphometrical andanatomical analysis it could be inferred that ani-mals are already sexually mature at about35 mm CL, no breeding female under 38 mmwas ever found and larger ones were more com-mon (average CL of breeding females41.99±0.28mm, n=45).

The number of hatchlings for each female,counted in seven different broods, ranged from55 to 157 (average 103.7). No correlationseemed to exist between CL of these females andthe number of hatchlings (r=0.23. df =5, ns),i.e.fecundity does not seem to increase with size.The fecundity of twenty-four females was evalu-ated by classifying the numerical consistency of

TABLE 6. Female fecundity relative to size

Size range(CL. mm)38^242-46

Egg massesSmall36

Medium26

Large43

the egg masses in three categories, i.e. small,medium and large (Table 6). and this confirmedthe previotis results (Fisher test: ns).

Sex-ratio at birth (determined in 293 crablets)did not differ significantly from 1:1; 47.4%(x^=0.577, df= 1, ns) of the crablets were mate.

Mating

Fourteen mating pairs were observed innature, from May to October. In all cases bothpartners had hard integuments and were foundfree in the water, in the open; only in two caseswere they found at a burrow entrance. The part-ners remained motionless for a period varyingfrom 30 min to 21 h. with their ventral surfacesin contact, the male holding the female's

chilipeds with his, or with his first pereopods andgrasping her with his pereopods.

The most striking feature was the difference insize between the partners, the females alwaysbeing smaller than the males (females: CL from29.7to42 mm, average 35.74±(1.87 mm; males:CL from 41.2 to 50.5 mm, average45.37±0.76 mm), Size frequencies (Table 7) ofmating and non-mating animals (comparing datagathered over the same peHtxJ) were signifi-cantly different in boih sexes (males: x^= 19.624,df=l. /'<0.001; females: ;^=12.794, df=2.P<0.01), which suggests that the difference insize of mating males and females is not simply areflection of the size composition of thepopulation.

The mechanisms of sex-recognition are stillunknown. In two matings observed in the lab-oratory (male CL: 47.9 and 45.7 mm; femaleCL: 34.5 and 34,7 mm) the first phases showedall the signs of an aggressive interaction, with Ihemale grasping the female by thechelipeds. then,after a few minutes, turning her over with the aidof his pereopods and pushing her under himself,at the .same time inserting his abdomen tmder hispartner's. In both cases, mating lasted about 3 h.Minei (1976) gave a similar description of matingbehaviour in the river crab Geotelphusa dehaaniWhite.

Seminal receptacles from eighteen femaleswere examined histologically. Out of fourteenanimals with CL >35 mm, captured in allmonths of the year, only one had empty sper-mathecae. Two adult, post-moult ('soft')females also had masses of sperms in their recep-tacles, indicating that sperms are kept overecdysis. Sperms were also present in one YF(34 mm CL), while other three YFs had not yetcopulated.

Discussion

The analysis of the population structure (con-

TABLE 7. Size frequencies of mating and non-mating individuals during the breeding season(CL=carapace length)

Size range(CL. mm)<35>35 and <40>40 and <45>45

MalesMating0049

Non-mating328

197

FemalesMating4820

Non-mating3312340

500 F. Micheli. F. Gherardi and M. Vannini

ducted on the same population as the presentstudy; Gherardi et al.. 1987). showed that thesize classes were not equally represented in thesexes, individuals over 40 and under 30 mm CLmore often being males, whereas females wererepresented more in the35-4U mm CL range. Inother decapods too, especially brachyurans(Adams. Edwards & Emberton., 1985), femalesare smaller than males, as a consequence of anincrease in the intermoult interval, e.g. in Cancerpagurus Linnaeus (Bennet, 1974) and in Pachy-grapsus crassipes Randall (Hiatt, 1948). and/orof a reduction of the growth rate (as in the case ofCancer magister Dana. Cleaver, 1949). P. fluvia-tile does not exhibit any sexual difference incarapace growth rate, contrary to the freshwatercrab Paratelphusa hydrodromus Herbst fromIndia, where a differential growth rate betweensexes was detected during sexual maturity (Ku-nip& Adiyodi, 1981).

Though in most decapods the increment atmoult declines with size, some exceptions doexist, e.g. Carcinus maenas Linnaeus (Crothers,1967; Hogarth, 1975). C. mediterraneus Czer-niavsky (Veillet, 1945), Maja squinado Herbst(Carlisle, 1957) and Pisa tetraodon Pennant(Vemet-Cornubert. 1958), All of these have adefinite terminal anecdysis. which, however,must be excluded in our case because largefemales were also found, though only rarely (thebiggest animal found in the stream was in fact afemale measuring 51.9 mm CL).

A second possible mechanism limiting femalesize could be a lengthening of the moult intervalcompared with males, caused by the onset ofreproductive activity, which, at least inAnomura and Brachyura (Adiyodi, 1985). isrestricted to the intermoult period. Further-more, both moulting and reproduction entailconsiderable energy demands (Passano, 1960;Yamaguchi & Takamatsu, 1980; Adiyodi, 1985;Gherardi, Tarducci & Micheli, 1989). Since, inthe populations of P. fluviatile Herbst in temper-ate areas, the breeding and moulting seasonsalmost coincide, it is unlikely that an animalwould undergo both events in the same annualcycle. In terms of optimization strategies, thesize range of mature females could represent acompromise between the advantages to be hadfrom spawning every year, and from reaching alarger size before maturity, and thus higherfecundity (i.e. the number of eggs produced peranimal per breeding season) with the onset of

reproductive activity. However, while a positiverelation between female size and number of eggsspawned has been shown in many crustaceanspecies (e.g. Pseudocalanus minutus KreJyer,McLaren, 1965; Heterozitts rotundifrons MilneEdwards, Jones, 1978; Homarus americanusMilne Edwards, Aiken & Waddy, 1980; Eriphiasmithi MacLcay. Vannini & Gherardi. 1988).this was not the case in P. fluviatile, thoughovary weight does exhibit a positive allometrywith respect to body size.

An alternative hypothesis explaining thesmaller size of females could be their highermortality rate, caused by the high cost of eggproduction (especially if compared to the lowcost of spermatogenesis; Pillay & Nair, 1971),and/or by the Intense vagility exhibited duringthe pre-ovulatory period (Gherardi, Tarducci &Vannini, 1988), bringing exposure to risks ofboth predation and dehydration as a con-sequence. The evidence that the sex-ratio calcu-lated in this population (58.6% males) duringthe winter months (i.e. in the period of thelowest vagility in both sexes), differs signifi-cantly from 50% (Gherardi et al., 1987), con-trary to the sex-ratio at birth, could possiblysupport the hypothesis of a higher mortality rateamongst females.

During the breeding season the sex-ratiochanges, shifting in favour of the females in Sep-tember, i.e. when they carry the brood undertheir abdomens, so that their locomotor activityis limited and they crowd into the stream(Gherardi et al., 1988a). A similar seasonalvariation in sex-ratio has also been observed inGeotelphusa dehaani White (Yamaguchi &Takamatsu, 1980). The lowering of moult fre-quency with the onset of sexual maturity (over35 mm CL) is not surprising, because the samephenomenon is widely observed in all Crust-acea, and is probably connected to the energycost of reproduction (Hartnoll, 1982).

The different distribution of moulting animalsin the two sexes observed in smaller classes (inparticular for CL >25 and <35 mm, where therewere more moulting females than males) indi-cates that females grow faster than males as theyapproach the puberty moult, anticipating theonset of their reproductive activity. Theacceleration of moult frequency prior to sexualmaturity could explain the large number of adultfemales in the population (Gherardi ei ai,1987). The length and frequency of oogenic

Growth and reproduction in a freshwater crab 50!

cycles seem to be characteristic for each crusta-cean species in a given environment, becauseoviposition is timed by some environmentallycorrelated mechanism is such a way that it takesplace when conditions are likely to be favourablefor the release of the young. For instance, inParatelphusa hydrodromus Herbst, a freshwatercrab with a wide geographical distribution inIndia, oviposition occurs from March to Octoberin South Kerala, where rains are scattered, whilein North Kerala, where rains are more seasonal.it only occurs in March, so that release of theyoung coincides with the beginning of the mon-soon. In the first case, multiple spawning over asingle annual cycle seem to be the rule, while thenorthern populations spawn once only(Adiyodi, 1985).

The patterns of relative growth by chelae, het-erochely, and abdomen follow those alreadyshown in most species studied so far (Hartnoll.1974), and the adaptive value of the allometryexhibited by these organs has already beenlargely discussed (Hartnoll. 1974, 1982; Finney& Abeie, 1981; Vannini & Gherardi, 1988).

P. fluviatile males do not adhere to thegeneral branchyuran trend of positive allometryof the first pleopods versus carapace length inthe instars preceding puberty, as their growth isisometric relative to carapace and abdomenlengths in both S and A, According to Hartnol!(1974), allometric growth of male gonopodsbrings them to an operative size at puberty.Their growth is correlated witb the size of thegenital openings of the females. Large malegonopods would be disadvantageous, as theywould limit the size range of possible partners.In most species this is prevented by tbe negativeallometry exhibited by these structures in adultmales(Hartnoll. 1974). However, in P.fluviatitemating is accomplished only by large males withsmall females. This could justify the absence of apositive allometry in the male gonopods,because by the time a male begins its reproduc-tive activity its gonofwds are probably already ofthe effective size with no need for acceleratedgrowth.

Allometry of female pleopods is a feature notgenerally observed in brachyurans. Femalepleopod endopodites are slightly allometric rela-tive to carapace length, but isometric withabdomen length, which in turn is positively allo-metric with carapace length; endopodite growthcould simply follow abdomen growth. On the

contrary, exopodites grow faster relative both tocarapace and abdomen length prior to thepuberty moult, while they become isometric inthe adult phase. Female exopodites protect eggsand brood, preventing them from loss anddehydration. Apositiveallometry of these struc-tures before puberty would therefore grant anadequate protective structure as soon as thefemale is ready to spawn for the first time. Pro-tection of the brood is particularly important forthis species in relation to the long time the youngstay under the mother's abdomen and the risksarising from tbe environment (newly hatchedcrabs could be carried away by the current orbecome dehydrated).

No difference was observed between the sexesin the level of allometry of carapace width versusCL, which in many species is related to the ovarydevelopment (Hines, 1982; Vannini &Gherardi, 1988). However, in male /'. fluviatile,gonads are very voluminous, especially the def-erens portion, ranging from 1.5% to 5% of thetotal body wet weight in adult animals, while inthe females gonad weight only exceptionallyexceeds 2% (Micheli, 1988). The considerabledevelopment of the male reproductive system inthis species does not seem to be a general featureof freshwater brachyurans: in the river crabBarytelphusa guerini Milne Edwards, forexample, female gonads are heavier than thoseof males (Gangotri et al., 1978).

Both sexes of P. ftuviatile undergo the pubertymoult at approximately the same size (35 mmCL), but functional maturity (which in femalesmeans the capability of producing eggs, and inmales corresponds to the phase when they arefirst able to copulate; Hartnoll, 1969) is in bothcases delayed, though to a different extent. Thelong gap between the onset of sperm productionand their effective transfer observed also inPortunussanguinolentus Herbst (Ryan, 1967) andin the superfamily Oxyrhyncha (Hartnoll, 1963),might be adaptive either because larger maleshave an advantage in agonistic encounters withconspecifics (Vannini. Gherardi & Pirillo, 1983).or because large animals inspect wider areas andmeet more females; a positive correlation existsbetween locomotor activity and body size (Gher-ardi etai, 1988a).

A deeper knowledge of reproductivebehaviour, and sex-recognition mechanisms Inpartictilar, is of course necessary. In the case offemales, mating is usually coincident or even

502 F. Micheli, F. Gherardi and M. Vannini

precedent to the pubertal moult, so that spermsare kept in the seminal receptacles until ovula-tion - all adult females had sperms in their sper-mathecae. Storage of sj erms occurs in manycrustacean species (Sastry, 1983). InChionoccetes bairdi Rathbun, sperms stored inthe receptacles are still viable after 2 years (Paul,1984). In some cases, sperms received during asingle copulation may fertilize eggs for severalovipositions (Hartnoll, 1963; Ryan, 1967).

\x\ P.fluviatite, mating of pre-pubertal femalescould be advantageous, due to the vagile andlargely terrestrial habits characterizing tbeiradult phase, which reduce the possibility ofencountering males, and/or to the considerableoverlap of the mating and breeding season, sothat, by the time mating does occur, mostfemales over 38 mm CL already carry a brood.

Acknowledgments

Many thanks are due to Dr Carlo Calloni for hisassistance in the laboratory. The work was car-ried out with the financial support of M.P.I.('60%' and '40%' grants).

References

Adatns J,. Edwards A.J. & Emberton H. (1985) Sex-ual size dimorphism and assortative mating in iheobligate coral commensal Trapezia ferrugineaLatreille (Decapoda, Xanthidae). Crustaceana,48. ltiS-194,

Adiyodi R.G. (1968) On reproduction and moulting ofthe crab Paratetphitsa hydrodromus. PhysiologicatZootogy. 4U 204-209.

Adiyodi R.G. (1985) Reproduction and its control.The Biotogy of Crustacea, Vol. 8 (Ed. G. Abeic),pp. 147-215, Academic Press, New York.

Aiken D.B, & Waddy S.L. (1980) Reproductive biol-ogy. The Biotogv and Marmgement of Lobsters(Eds J. S. Cobb and B. F. Phillips), pp. 215-276.Academic Press, New York,

Bennet D.B. (1974) Growth of the edible crab (Cancerpagurus L.) of south-west England. Journal of theMarine Biological .Association. U.K. 54,803-823.

Carlisle D.B, (1957) On the hormonal inhibition ofmoulting in decapod Crustacea. 2. The lerminalanecdysis in crabs. Journal of the Martne Biatogicat.4ssociation. U.K. 36, 291-307.

Cleaver F.C. (1949) Preliminary results of the coasialcrab {Cancer magister) investigation, WashingtonState Department of Fisheries. Biology Repori.49A. 49-82.

Cronin L.E. (1947) Anatomy and histology of themale reproductive system of CalUnecies sapidusRathburn. Journat of Morphotogy. 81, 209-239.

Crothers, J.H. (1967) The biology of the shore crabCarcinus maenas (1..)- Fietd Studies. 2, 407-434.

Finney W.C, & Abele L G , (1981) AUometric vnri-ation and sexual maturity in the obligate coral com-mensal Trapezia ferruginea Latreille (Decapoda,Xanthidae), Crustaceana, 4\, 113-130,

Gangotri M.S,, Vasantha N. & Venkatachari S.A.T,(1978) Sexual maturity and breeding behaviour inthe freshwater crab. Barytetphusa guerini MilneEdwards. Proceedings of the Indian Academy ofScience. 87B, 195-201.

Gherardi F.. Guidi S. & Vannini M, (1987)Behavioural ecology of the freshwater crabPotamon Jtuviatile. preliminary observations.Investigacion Pesquera. 12, 222-234.

Gherardi F., Messana G.. Ugolini A. & Vannini M.(1988a) Smdies on the locomotor activity of thefreshwater crab, Potamon fiuviatile. Hydro-biotogia, 169, 241-250.

Gherardi F., Tarducci F. & Micheli F. (1989) Energymaximization and foraging strategies in PotamonJtmiattle. Freshwater Biotogy, 22, 233-245.

Gherardi F,. Tarducci F, & Vannini M. (1988) Loco-motor activity in the freshwater crab Potamonftuviatite: the analysis of temporal patterns byradio-telemetry. Ethotogy. 11, 300-316.

Gray E,H, & Newcombe C L . (1938) The relativegrowth of parts in the blue crab Cattinectes sapidus(Rathbum), Growth. 2, 235-246,

Haley S,R, (1969) Relative growth and sexual matu-rity of the Texas ghost crab, Ocypode quadrata(Fabr) (Brachyura. Ocypodidae). Crustaceana.17. 285-297.

Haley S.R, (1973) On the use of morpbometric data asa guide to reproductive maturity in the ghost crab,Ocypode ceratophtatmus (Pallas) (Brachyura,Ocypodidae). Pacific Science, 27, 350-362.

Hartnoll R.G. (1%3) The biology of manx spidercrabs. Proceedings of the Zootogicat Society ofLondon. 141,423-496.

Hartnoll R G , (1969) Mating in the Brachyura. Crust-aceana. 16. 161-181,

Hartnoll R.G. (1974) Variation in the growth patternbetween some secondary sexual characters m crabs(Decapoda. Brachyura). Crustaceana. 27. 131-136.

Hartnoll R.G, (1978) The determination of relativegrowth in Crustacea. Crustaceana. 34, 282-293.

Hartnoll R.G, (1982) Growth. The Biotogy of Crust-acea. 1. Embryotogy. Morptiotogy and Genetics(Ed. L. G. Abele), pp. 111-196. .Academic Press,New York.

Hiatt R.W, (1948) The biology of the lined shore crab,Pachvgrapsus crassipes Randall. Pacific Science, 2,135-213.

Hines A.H. (1982) Allometric constraints and variableof reproductive effort in brachyuran crabs. MarineBiotogy. 69. 309-320,

Hinsch G.W. & Walker M.H. (1974) The vas deferensof the spider crab Libinia emarginata. Journat ofMorphology. 143, 1-20.

Growth and reproduction in a freshwater crab 503

Hogarth P.J. (1975) Instar number and growth of juve-nile Carcinus maenas (L.) (Decapoda. Brachyura).Criutaceana. 29, 299-300.

Huxley J.S, (1932) Problems of Relative Growth.Methuen & Co., London.

Jones M.B. (1978) Aspects of the big-handed crabHeterozius rotundifrons (Decapoda: Brachyura).from Kaikoura, New Zealand. iVew Zealand Jour-nal of Zootogy. S, 783-794.

Koba K. (1936) Preliminar>' development of Geo-tetphusa dehaani (White), Proceeding.\ of the Impe-riat Academy (Tokvo). 12, 105-107.

Kurup K.N.P. & Adiyodi R.G. (1981) The program-ming of somatic growth and reproduction in thecrab, Paraielphusa hydrodromus (Herbst). Inter-national Journat of Invertebrate Reproduction. 3,27-39.

Mauchline J. (1976) The Hiatt growth diagram forCruslacea, Marine Biology. 35, 79-84,

McLaren LA. (1965) Some relationships belweentemperature and egg size, body size, developmentrate and fecundity of the Copepod Pseudocalanus.Limnotogy and Oceanography. iO, 528-538.

Mercanli F, (1S85) Sullo sviluppo postcmbrionaledelta Tetphusa fluviatitts Latr. Bollettino detlaSocietd Entomotogica Itatiana, 17, 209-216.

Micheli F. (1988) Bioiogia riproduttiva di Potamonfiuviatite Hcrbst (Decapoda, Brachyura). Gradua-tion thesis.

Minei H, (1976) Ecology of the freshwater crab Geo-telphusa dehaani. Anima, 41,1(H5 [in Japanese],

Pace F,, Harris R.R, & Jaccarini V, (1976) Theembryonic development of the mediterraneanfreshwater crab Potamon edutis (= PotamonfJuviatiie) (Cruslacea, Decapoda. Potamonidae).Journal of Zootogy, 180, 93-106.

Pas.sano L.M. (1960) Molting and its control. ThePhysiology of Crustacea, Vol, 1 (Ed, T. H. Water-man), pp. 473-536. Academic Press, New York.

Paul A,J, (1984) Mating frequency and viability ofsiored sperm in the tanner crab Chionoecetes bairdi(Decapoda, Ma]idas). Journal of Crustacean Bio-togy. 4. 375-381,

Perez C. (1928) Sur I'appareil d'accrochagc de I'abdo-men au thorax chez les Decapodes Brachyures.Comptes Rendus de I'Academie des Sciences deParis. 186, 461-463.

Pillai C.K. (1982) Studies on the breeding biology and

egg development in the freshwater crab Para-tetphusa hydrodromus (Herbst). Ph.D. disserta-tion. University of Madras, India.

Pillay, K,K,& Nair N,B,( 1971) The annual reproduc-tive cycles of Vca annutipes, Portuniu pelagicusand Metapenaeus affinis (Decapoda: Crustacea),from the South-Wcst coast of India. Marine Bio-logy, n , 152-166,

Ryan E.P, (1967) Structures and function o! thereproductive system of Ihe crab Portuniu san-guinotentus (Herbst) (Brachyura: Portunidac): themale system. Proceedings of the Marine BiologyAssociation of India. Symposium Scrioi, 2, 506-521,

Sastry A.N, (1983) Ecological aspects of reproduc-tion. The Biotogy of Crustacea. Vol. K (Ed. F J.Vernberg and W, B. Vcrnbcrg). pp. 179-270. Aca-demic Press, New York.

Schneider P. (1971) Zur Biologie der afghanishenRusskrabbe Potamon gedrosianum. I,Lcbenswcise, Verbreitung. Morphologie und sys-tematische Stellung. Bonner Zoologische Beitrage,22,305-321.

Teissier G. (1960) Relative growth. The Ptiy.'iiotogy ofCrustacea. Vol, 1 (Ed, T. H. Waterman), pp. 537-560, Academic Press, New York,

Vannini M, & Gherardi F. (1988) Studies on the peb-ble crab. Eriphia smithi MacLeay 1838(Xanthoidea Menippidae): patterns of relativegrowth and population structure. Tropical ZoO'logy, I, 203-216,

Vannini M., Gherardi F, & Pirillo M, (1983) Aggres-sive communication in Potamon fttmatiti- role ofsize, colour, and other visual clues. Crustaceana,45. 203-209.

Veillet A. (1945) Recherches sur Ic parasiiisme descrabes et des gasathecs par les rhizocephales el desepicarides. Annales de t'Institut Oceanographique.22, 193-341.

Vernel-ComuberlG. (1958) Biologie generalede Ptvatetraodon. Bulletin de I'lttstitut Oceanographique.1113, 1-52.

Yamaguchi T. & Takamatsu Y. (1980) Ecological andmorphological studies on the Japanese freshwatercrab Geotetphusa dehaani. Kuinamuto Journal ofScience. Biology, 15, 1-27,

(Manuscript accepted 12 Juty 1989)