ELSEVIERJournal of Experimental Marine Biology and Ecology

193 (1995) 197-223

JOURNAL OFEXPERIMENTALMARINE BIOLOGYAND ECOLOGY

The ecological consequences of limb damage and loss Indecapod crustaceans: a review and prospectus

Francis Juanes 3 ,*, L. David Smithb

a Department of Forestry and Wildlife Management, University of Massachusetts, Amherst,MA 01003-4210, USA

bSmithsonian Environmental Research Center, P.O. Box 28, Edgewater, MD 21037, USA

Abstract

Autotomy, the reflex severance of an appendage, is considered an adaptation to avoidpredators and limit wounds. While an autotomy response may provide immediate survivalbenefits, the loss of one or more appendages can result in long-term functional andenergetic costs. In this paper, we present an overview of the incidence of limb damage andloss in decapod crustaceans; review the literature on the ecological consequences of suchinjury; and suggest areas for future research, A survey of limb damage and loss in fieldpopulations showed consistently high incidences of injury in 14 reviewed species. Typically,chelipeds were the limb type lost most often and injuries were distributed symmetrically.No consistent correlation existed between injury frequency and body size among species. Ingeneral, the frequency of injury was independent of sex and moult stage. Fishery practiceswere responsible for substantial limb loss in some commercial species. In terms of energeticcosts, experiments demonstrated that limb injury could reduce growth increment and affectintennoult duration. Functionally, limb damage was capable of reducing foraging efficiencyand mating success, and increasing vulnerability to intra- and interspecific attack. Themagnitude of these effects depended on the type and number of limbs lost. Given theprevalence of injury in decapod crustacean populations, the costs involved, and theecological importance of many crustacean species, nonlethal injury has the potential toaffect population dynamics and community processes. Convincing evidence of autotomy'seffects beyond the level of the individual, however, is, at present, lacking. Future workshould redress this shortcoming. In addition, comparative studies are needed on decapodspecies from different habitats and with different lifestyles before generalizations can bemade about the costs and benefits of autotomy.

Keywords: Autotomy; Decapod crustacean; Functional and energetic costs; Injury fre­quency; Limb damage; Limb loss; Regeneration; Survival benefits

* Corresponding author.

0022-0981195/$09.50 © 1995 Elsevier Science BV All rights reservedSSDI0022-0981(95)00118-2

198 F Ii/anel, LD. Smith! 1. Ex!), /·vfar. BIOi. Eeol. J'J3 (1995) 197-·223

1. Introduction

Autotomy is a reflexive response to injury or its threat that results in the castingoff of an appendage at a predetermined breakage plane (Wood & Wood, 1932;Hopkins, 1993). Tail autotomy by vertebrates (lepidosaurans, salamanders, androdents) (Arnold, 1988) and limb autotomy in asteroids, ophiuroids (Mauzey etal., 1968; Lawrence, 1992) and spiders (Formanowicz, 1990) is usually a responseto predator attack that facilitates escape but one that also carries ecological costs(Arnold, 1988; Lawrence & Larrain, 1994). Many decapod crustaceans are capableof autotomizing limbs; evidence of the response can be found in primitive fossilcrabs from the Cretaceous (Wood & Wood, 1932). Fredericq published adescription of the crustacean autotomy mechanism in 1882, and since that time,most of the research has focused on the physiology and anatomy of limb loss andhow it affects moulting and regeneration (see reviews by Wood & Wood, 1932;Bliss, 1960; Needham, 1965; Goss, 1969; McVean, 1975, 1982; Skinner, 1985;Hopkins, 1988). In contrast, relatively few studies have examined the ecologicalcauses and effects of such trauma in decapod crustaceans.

The benefits of autotomy are thought to include predator avoidance and woundlimitation (Bliss, 1960; McVean, 1982). The response likely serves as a final meansof preventing subjugation by a predator after other avoidance measures havefailed (crypsis, fleeing; Endler, 1986). For example, Lawton (1989) noted thatautotomy in the anomuran Porcellana platycheles was effective in preventingcapture by the brachyuran Cancer pagurus. Autotomy may be particularlyimportant in aquatic environments, where waterborne chemical cues can alertpredators (e.g. Zimmer-Faust, 1989) to a wounded animal's predicament. Injury,however, is not requisite to autotomy. Robinson et al. (1970) demonstrated thatsome species of tropical terrestrial crabs will grasp attacking predators with theircheliped(s) and actively autotomize the limb(s).

Benefits of autotomy will be partially offset if functional and energetic costs areincurred. These potential costs include reduced growth, lowered foraging ef­ficiency and mating success, and increased vulnerability to intra- and interspecificattack (e.g. Kuris & Mager, 1975; Sekkelsten, 1988; Smith & Hines, 1991a;Davenport et al., 1992; Smith, 1992, 1995). The effects of limb loss on growth arepotentially important because differences in relative body size have been shown tobe critical to the outcome of various ecological interactions such as predator-preyencounters and intraspecific competition (O'Neill & Cobb, 1979; Hyatt, 1983;Garvey et al., 1994). Body size can also strongly determine the ability to attractmates (e.g. Salmon, 1983; Sekkelsten, 1988; Reid et al., 1994; van der Meeren,1994) and the level of reproductive output (Hines, 1982). The degree to whichindividual performance is handicapped will vary with the type and number ofmissing or regenerating limbs (Smith & Hines 1991a, Smith 1992, 1995) and withthe length of time the animal must function without its services. In turn, the extentto which populations are affected by autotomy will depend on the frequency andtype of limb loss in the population and the cumulative effect these injuries haveon survival and reproduction (Harris, 1989). Many decapod crustaceans are

F. .Iuane.~. L.D. Smith I .I. Exp. Mar. Bioi. f:'col. 193 (1995) 197-223 199

important predators, grazers, and bioturbators in aquatic communities (e.g.Abrahamsson, 1966; Bertness, 1985; Coen, 1988; Hines et a1. 1990). If injury iscommon in such species and costs are significant, autotomy could have profounddirect and indirect effects in the community. Commercially, knowledge of theextent of injury in fishery populations (including injury due to the fishing processitself) and the effects of injury on growth and reproduction should be incorpo­rated into population growth models to produce improved stock assessments. Inthis paper, we present an overview of the incidence of limb damage and loss infield populations; review the literature on the ecological effects of limb loss anddamage; suggest important individual, population and community effects; andfinally give a prospectus in which we outline specific areas where more work isneeded.

2. Prevalence of limb loss in natural populations

Table I summarizes a literature survey of the incidence of injury in naturalpopulations of decapod crustaceans. A total of 14 species in 27 studies arerepresented, of these, four were commercially important (Cancer magister,Callinectes sapidus, Paralithodes camtschatica, Homarus americanus). Injury wasmeasured in a variety of ways (e.g. missing limbs only, missing and regeneratinglimbs, missing chelipeds only), but for all species, injury was consistently high.Various studies noted lower frequencies of regenerating than missing limbs,suggesting that animals may experience increased mortality following injury(McVean & Findlay, 1979; Shirley & Shirley, 1988). Such comparisons are difficultto interpret, however, because: (l) animals that have nearly regenerated a limb ofnormal length are easily missed (instances of reversal of cheliped asymmetrybeing an exception), and (2) the relative frequency of injured animals will increasein populations dominated by older individuals and in species with a terminalecdysis.

Patterns of limb loss were remarkably consistent within and among species andprobably reflect both the limb function and the behavioral response of the animalto the injury-causing agent. In species where it was examined, single limb loss wasthe most common injury; chelipeds were the limb type lost most often (exceptionsare c..Yrtograpsus angu[atus, Paralithodes camrschatica, and Chionoecetes bairdiwhere walking legs were most often autotomized); and injuries were typicallydistributed symmetrically (when exceptions occurred, injuries were all biased tothe right side). These patterns would be expected if: (1) animals seek to minimizethe number of limbs last; (2) autotomy's effectiveness as an escape mechanismdiffers along the anterior-posterior axis; and (3) prey movement and predatorattack are random 1.n direction. In any encounter with a predator, decapodcrustaceans should aUempt to escape with the loss of, at most, a single limb. Infact, multiple limb loss is relatively rare, yet it occurs more frequently than wouldbe predicted by chance. A number of authors (e.g. Needham, 1953; Easton, 1972;McVean, 1976; Shirley & Shirley, 1988) have suggested that prior limb loss

t...l

8

Table 1Incidence of limb damage and loss in natural populations of decapod Crustaceans

Species Size (mm)' Sex ratio h Samplmg time No. of % Injury Methods Most Symmetry Size Sex Rderence(or other slJrling) animals used to injured in limb correlation~ bias h :-r,

examined' measure limb' loss! :::-.. J 1;mJury ::Callcer magis/er 159-169 65:35 May-Nov. '84-'85 878 25,0 '""E C Y 0 0 Shirley & ShIrley. 1988 '"Callcer magis/er 62- 197 89:11 Dec. '71-Feb. '73 3085 66.0 B C y Durkw e! aL 1984 :-Callcer magis/er Jan. -Sept. '47 3886 15.8 Mm Cleaver. 1949 '::J

Jan. Sept. '48 5383 20.1~Cmcer mag/s/er (offshore) Nov. '47-Jan, '50 4092 23.1 M Waldron, 1958

(bays) Nov. '47-Jan. '50 2157 37,8 -Callcer magis/er '70-'72 ]7,0 Ames. pers. comm. In DurkinCallccr magister >iOO Females May '86 181 12.0 (60)J E C Y Juancs & H~rtwick, 1990

~

Mated males 227 8,0 (9) t"'1

Unmaled males 500 22.0(60) -8Callcer pagurUJ Females Summer '72 938 4,8-9.9 B C y (I Bennett. 1973 ~

t::Males 1482 8.1-132 :"

Cal/incurs sap/dILl 12:48:40' 67:33 Julv-Nov. '86 1400 24.8 B C Y 0 Sm!lh. j 990b t::::20:26:54 8U9 May-Nov. 'S7 622 18,8 Y 0 ~

Callmenes Jlipidus 21:38:4l' 83:17 May-Nov. '88 649 175 B C N (r) 0 IJ Smith & Hines. 1991b t!'J3218:50 76.24 May-Nov. '89 692 17.8 Y 0 "-

2.-125:74 40:60 Aug.-Sept. '89 679 19,1 Y IJ

'-012:88 5150 Oct. '89 103 38.8 Y 0 0 ~

~,

169::75 26:74 Oct '89 224 330 Y (I 0 .......'-

17:17:66 61:39 May '89 204 319 N (f) M 'C~

25:5:70 32:68 May'89 201 343 Y 0 0 ~15:28:57 37:63 May '89 260 26.5 Y 0 0 '-

'CPurfwlllJ peiagicJis 60-148 3763 Jan.-Dec. '90 951 10,8 B C y 0 f) Shield\. unpub!. data "-J

CarcinllS maenaJ 20-34,9 Males June-Aug. '85 173 17 C Sekkel,ten, 1988 I"r·,...J35-449 onlv 144 3.5 ""45-549 JSO 10.055-64.9 237 16,065-79.9 134 17.9

Careinlls maenas 11-57 Females Summer '89 and '91- '93 1116 7.9 C '\ (r) 01+" M Abello et al.. 199410-75 Males 2457 12.5

Careinus maenas 5-80.1 Females Sept. '73 633 40.0-55.0 B C M' McYean. 1976males 390 270-53.0

~Carcinus maenas 8-88.1 Females Feb. '76-Jan. '77 1917 165 B C + 0 McYean & Fmdlay, 1979 '-,

Males 1421 20.0 :;;:;::;

Cancer maenas 3-38.1 234 26.5 B C N (r) i\eedham. 1953;:,

""Poreellana platyeheles 115 28,8 B C Y

.Co

/l!ceom pubcr <40 49:51 May '86-April '87 448 23.1 M C y 0 Norman & Jones, 1991 t-+ ;....

40-80 60:40 1079 285 R .'-'

CyrlograpsllS angula/us <15 62:38-41:59 Aug. '86-May '87 2243 50.0 B 3.4 F Spivak & Politis, 1989c",

~.15-24.9 62:38-41 :59 60.0 ;i.25-34.9 38:72-94:6 80.0>35 38:72-94:6 70.0 :--

Pachygrapsus craSS/pes >lO()(J 30.0 B Hiatt. 1948 ~Ch/onoreelcs bairdi Juvenile males 202 34.6 B WL Y 811 M Edwards, 1972

...,',:::

Adult males 196 43.0 2;:Juvenile females 67 34.0 .,Adult females 90 23.3 .'"

Parahllwdes (am/seha/lea Juvenile males Sept.-Oct. '69 25.6 PWL Edwards. 1971I;;:

291 B Y 0~Adult males 372 14.7

Juvenile females 228 293 l"r1

Adult females 487 16.6 ~Paralithodes camlsehalica 8.5-30.2 CL May-June '61 371 57.0 B Kurata, 1963

,\0Paralilhodl's camcJchatica 106-191 CL Males April-June '62 33744 153 B PWL N*(r) 0 Niwa & Kurata, 1964 w

82-176 CL Females 1783 19.5--.......

Panulirus argus 15-101 CL FEb. '76-Dec. '77 7643 40.3 I Davis, 1981 ~

Nephrops norvegicus Females Sept. '68 ]()() 41.0 0 C Chapman & Rice, 1971 ~

Males 1()() 62.0 'C".J

Humarl(~ americanus <63.5 CL Spring '71-'72 12344 30-18.0 C Scarral, 1973 I~"

63.5-80.9 4266 5.0-24.0 ~,J',,-,

>81 411 0.0-11.5

HOmaflt.5 americanus 47320 5.0-12.0 C Wilder. pers. camm. in Scarrat

Homarus amencanus 36-101 CL (research) Jan.-Dec. '69-'74 6417 21.0 C Krouse, 1976

81->121 (commercial) 68'-'74 20226 6.5

N0-

c

c

C

Mostinjuredlimb'

Symmetryin limblossf

Si7ecorrelation"

Sexbias h

F

Reference

Briggs & Mu~hacke. 1979

Briggs & Mushacke, 1980

Mnriyasu et a!.. unpuh!. data

B Estrella & Armstrong, 19')4---

" All mea,urcmenls arc for carapace WIdth (except, where noted, I.or carapace length, Cl) and are in mm.h Sex rallO is shown as M:F., When only one number appears it represents total for study,j M = mlSSmg limbs: R regeneratlJlg limbs: B = both missing and regenerating limbs: D ~ damag~d limbs i IC worn nr brok<:n dactyls): E = missmg. regenerating and damagcd limbs: C = mis'ing andregenerating chelae: I = missing limbs and antennae and bodily damage., C = chdipeds: R = random: 3.4 = 3 and 4th walking legs: PWl = posterior walking legs: Wl = walking legs.! Y= yes: N = No; r = biased toward right side: N* = not symmetrical for chelae but symmetrical for walking legs., + ~. Injurv positively correlated wlth size: - = injun' negatively correlated with size. {] = no size correlation: i "maXimum injun at intermediate siDe'.11 M :;; male bias~ F = ~'C.ma\c bias: CJ ;;..c ro !ji:}$.

'Size ratli" ISML. S c <61. ~1=61 ~ i1O: l= >110)Numbers m parentheses refer to proportion 01. animals with worn claws.

, Sex bias was only present til animals from subtidal habitats. Intertidal animals showed no sex hias in the proportion of injured animals./>, negative correlation m females and a positive corrclotion in males.

m Note that for columns d, g~k the information in the first line is valid lilr oil components of the study unle" otherwise noted. Blank spaces in ht first line of these columns means tliat this aspect was notinvestlgated in the studv." No correlatIon til green crab phenutype and a positive correlation in red crob phenotype.

F. Juanes, LD. Smith / J. Exp. Mar. BioI. Ecol. 193 (1995) 197-223 203

increases an animal's vulnerability to subsequent injury. An alternative explana­tion is that multiple limbs are lost in a single attack event (Smith, 1990a). Thelatter interpretation is supported by observations that multiple limb loss is mostoften associated with sequential limb loss (i.e. losses tend to occur on the sameside and in adjacent limbs) (Spivak & Politis, 1989; Norman & Jones, 1991). Mostbrachyurans respond to threats with outstretched claws (Schone, 1968; Robinsonet aI., 1970; Jachowski, 1974). While this behavior makes anterior limbs par­ticularly vulnerable to injury, the chances of surviving an anticipated attack arerelatively high. In contrast, surprise attack from the rear may often prove fatal;consequently, damage to posterior limbs will be observed infrequently in popula­tions. The predominance of symmetrical injury is not surprising. Althoughasymmetrical limb loss might be expected in crabs that move predominantly inone direction (Needham, 1953), in most populations, directionality of movementis probably random (e.g. Callinectes sapidus, Smith, 1995). Furthermore, predatorswould not be expected to attack one side preferentially.

The correlation between injury frequency and body size is not consistent amongspecies and can vary within a species over temporal and geographic scales (Smith& Hines, 1991b). Three species generally experienced increased frequency ofinjury as body size increased (Callinectes sapidll.S, Care/nus maenas and Necorapuber); two species showed no size correlcition (Cancer pagurus and Portunuspelagicus); one species had maximum levels of injury at intermediate sizes(Cyrtograpsus angulatus); and three species showed a negative correlation(Paralithodes camtschatica, Chionoecetes bairdi and Homarus americanus). Apositive correlation between injury and body size may result from decreasedpredator efficiency as .xey size increases (Smith 1990a) and an accumulation ofdamage in older aninrals due to longer intermoult periods, reduced regenerationpotential (Hamilton et aL, 1976; Spivak & Politis, 1989; Smith & Hines, 1991b)and attainment of terminal moult (Sekkelsten, 1988)" The incidence of injury canalso be a function of moult stage (Juanes & Hartwick, 1990) and season (Shirley& Shirley, 1988). Limb loss is generally independent of sex, although exceptions,which may be habitat-specific, occur. For example, McVean (1976) noted thatthere was no difference between the sexes in the incidence of autotomy inCarcinus maenas found in the intertidal zone. However in sublittoral habitats,males showed signifkm tly greater incidences of autotomy than females.

Another type of limb dam1ge, chelae tooth wear, is prevalent in the two specieswhere it has been documented (Menippe mercenariu, Bender, 1971; Cancermagister, Juanes & Hartwick, 19(0). Chelne tooth wear results from feedingactivity (as oppo~ed to inter- or intfE'Specitic aggression) and can range from slightto extreme wear where balf or n~.ore of the tooth vo;'urnes have disappeared(Juanes, 1992).

Humans can be responsihle for substantial limb loss in some commerciallyfished species, either through intentional harvesting of claws (e.g. Menippemercenaria, Savage & Sullivan, 1978) or as a resutt of incidental damageassociated with fishing gear (e.g. trawls, Reilly, 1983) or culling of undesirableindividuals (Bennett, 1973; Kennelly et al., 1990). For example, Krouse (1976)

204 F juanes, L..). Smith; J. Fxp. /vlllr. did. Ecoi. 193 (l,'95) ])7-.223

showed a direct relationship between fishing mtensity of lobsters (Homarusamericanus) and the incidence of culls (i.e. individuals missing or regeneratingchelae). Similarly, Scarratt (1973) has suggested that claw loss in commerciallycaught lobster may be due not only to the handling and moving of fishing gear butalso to the local (Prince Edward Island, Canada) practice of harvesting Irish mossby rakes. Sublegally sized spanner crabs (Ranina ranina) can lose on average 3-4dactyli and 7 limbs while being disentangled from the commercial tangle-nets(Kennelly et al., 1990). The potential for extensive injury as a consequence offishing (both directly and indirectly) suggests that fishing seasons should be timedwith life history events so as to avoid fishing during those times when animals areparticularly susceptible to injury. For example, Shirley & Shirley (1988) foundthat appendage injuries increased 157% from July to August, a period ofsimultaneous moulting, mating and fishing.

3. Ecological consequences of limb damage

Consistently high levels of injury in natural populations of decapod crustaceanssuggest important fitness benefits to the autotomy response, but also raise thequestion of potential costs of limb damage and loss. Functional costs are likely,otherwise, animals would not invest the energy necessary to regenerate themissing structure (Goss, 1969). While these costs may be ameliorated in crusta­ceans because of limb redundancy (Reichman, 1984), loss of more specializedlimbs (e.g. dimorphic chelipeds, swimming legs) should still impose significanthandicap. Below we review the consequences of limb damage on feeding, growth,regeneration, reproduction, competitive ability, predator avoidance, and survival.Much of this work was performed by manipulaIing levels of damage and loss inlaboratory situations although a few studies also performed experiments in fieldsituations. Studies have typically focused on the effects of damage to chelipeds(versus walking or swimming legs), because of their functional importance andtheir relatively large biomass.

3.1. Feeding

Chelipeds are used by decapod crustaceans to capture, manipulate, and subdueprey (e.g. Vermeij, 1982; ap Rheinallt & Hughes, 1985; Lawton, 1989). Theirdamage or loss could have profound effects on foraging efficiency, yet few studieshave tested for foraging costs. Juanes & Hartwick (1990) demonstrated thatDungeness crabs, Cancer magister, with damaged (broken dactylus or propodus)or worn chelae were unable to feed successfully on a hard-shelled bivalve,Protothaca staminea. Crabs with worn chelae teeth had significantly longerhandling times than uninjured individuals feeding on similarly-sized prey. Blue

F. Juane.\'. L.D. Smith / 1. Exp. Mar. Bioi. Leol. 193 (1995) 197-223 205

crabs, Callinectes .'lapidus, missing both chelipeds had significantly lower feedingrates and ate smaller sized soft-shell clams (Mya arenaria) than intact crabs orcrabs missing one cheliped (Smith & Hines, 1991a). The absence of one cheliped,however, did not significantly affect blue crab feeding rate or size selectivity onthis lightly armored bivalve. Compensatory use of walking legs in the absence ofchelipeds has been observed in Callinectes sapidus (Smith & Hines, 1991 a), stonecrabs, Menippe mercenatia (Savage & Sullivan, 1978) and lined shore crabs,Pachygrapslls crassipes (Hiatt, 1948).

Reduced feeding efficiency may persist throughout the regenerative process.Elner (1980) showed that shore crabs, Carcinus maenas, with smaller-than­average (i.e. presumab\y regeneratirg) chelae had a lower energy intake rate than"normal" crabs because they chose smaller sLc:td mussels. Brock & Smith (unpub!.data) recorded significantly lower maximum crushing forces in regenerating clawsthan contralateral norma:1 claws of Cancer productus even after two instars hadpassed since autotomy. In addition, when adjusted for propodus length, maximumcrushing forces delivered by the regenerating and opposite "normal" claws ofinjured crabs were significantly less than those produced by normal claws of intactcrabs. This overall weakening of chelipeds in injured crabs might be expected ifexercise influences claw size (lnG crushing force (Smith & Palmer, 1994) and crabswere selecting smaller prey.

In heterochelous species, loss of a major cheliped could pennanently impactforaging success if dimorphism is not re-established. Reversal of handedness is anoften noted consequence of autot]my ceria in ~,pedes of heteroche1ous de­capods (Przibram 19::q· Hamilton et a1., J976; Cheung, 1976; Vermeij, 1977;Barnwell, 1982; Abello ~t aL,l990; Norman 8;: Jones, 1991). Following loss of themajor cheEped, the existing minor claw transfomls into a major claw and theautotomized limb is replaced by a miner claw. The reversal process generallytakes two to three moults to complete and intermediate stages resemble homoch­elaus animals with two minor chelae (Przibram 1931; Lewis, 1969; Hamilton eta1., 1976; Savage & Sulliv;:m, 1978; Abby-Kalio & Warner, 1989; Norman & Jones,1991). If chelotomy of 3. major claw OCC'lrs in older individuals (even olderjuveniles), claw dimorphi~m may never be recovered (Smith, 1990b) Smaller clawsizes and th~ absence specializc:c crushing dentition in hOffiocheloUls individualscould reduce foraging effkiency. Even if claw rever:;al j~ accomplished, regener­ated major claws may he 'less effective as foraging:nstrun:,ents. Govind & Blundon(1985) showed that left crushers:)f blue Cf'?bs were weaker and had a smaIlermechanical advantage than similarly-sized right Gllshers. Furthermore, if right­handedness is, in fact,on adapt'1tion to T113ximize }C'ding efficiency of dextrallycoiled marine gastropod" (Ng & Tan, 1985), left-handed crabs would be at adisadvantage when trying to cut spirally into tbe shelL

Lastly injury to chelipeds may lead to a f,hift to aH':rnative prey. For example,chelotomized stene crRbs alter tb~ir diet to consume soft invertebrates. detritus,algae and grasses (B".tHiCL ~ 97n and blue manna crabs (Portunuspelagicus) switched tc ,~ ':cer!JivO[vlS diet f;-nffi the narc typical cclrnivorous diet

206 F. JUaIlCS, L.D. Smith 1. Fxp Mar. Bio!.':col. !93 (1;95) 197-223

(Edgar, 1990). These changes in diet (either to different prey species or smallersizes of the same species) could have direct effects OJ] gW1Nth and regeneration.

3.2. Growth and regeneration

In decapod crustaceans, growth and limb regeneration are intimately linked tothe moulting cycle. Several reviews (e.g. Bliss, 1960; Skinner, 1985) have summa­rized an extensive literature on the physiological effects of limb loss on growth,regeneration and moultin.g frequency; in this section, we win discuss the ecologicalramifications.

Limb loss can affect crustacean growth, and potentially fitness, by reducing sizeincrease at the moult, altering the timing of ecdysis, and impairing foragingefficiency (see above). Limb loss is known to reduce moult increment (e.g. percentincrease in carapace width at the moult) in various decapod species (Hiatt, 1948;Hughes & Matthiessen, 1962; Bennet, 1973; Kuris & Mager, 1975; Savage &Sullivan, 1978; Davis, 1981; Hopkins, 1982; Ary et at, 1987; Smith, 1990b; Cheng& Chang, 1993; Moriyasu et aI., unpubl. data). Because chelipeds comprise asubstantial portion of the total body weight in many decapods (e.g. 20% inCarcinus maenas and Liocarcinus holsatus; Lee & Seed, 1992; 50% in Menippemen'enaria, Simonson & Steele, 1981), their replacement should divert thegreatest amount of energy from growth relative to other limb types. It issomewhat surprising, then, that the loss of one cheliped had no effect on sizeincrease at the next moult in Hemigrapsus oregonensis (Kuris & Mager, 1975),Cancer pagurus (Bennett, 1973) or Callinectes sapidus (Smith, 1990b). A likelyexplanation is that although significant biomass is lost following chelotomy,individuals reduce their total energetic burden by replacing limb mass incremen­tally over several moults. Because the majority of injured crabs in Bennett's(1973) and Smith's (1990b) study populations were missing a single cheliped, bothauthors concluded that autotomy had little overall effect on crab populationgrowth. Foraging costs in these laboratory experiments, however, were negligible(crabs were given abundant, easily consumed food); such may not be the case inthe wild. Using mark-recapture techniques in the field, Davis (1981) recordedsignificant reductions in growth in spiny lobster, Panulirus argus, as a result ofinjury (usually 1-2 antennae and 1-2 legs). He estimated an additional 33 wk forinjured juveniles to reach legal harvestable size compared to uninjured juveniles.

Reductions in growth increments become evident as more limbs are lost(Bennett, 1973; Chittleborough, 1975; Kuris & Mager, 1975; Hopkins, 1982; Smith,1990b, but see Davis, 1981). Skinner (1985) termed this phenomenon "regenera­tive load", where the size increase at the moult will be devalued by the extent ofregeneration required. She proposed that the maximum amount of tissue a crabcould regenerate during a single moult cycle was equivalent to 12-15% of themetabolically active weight of the premoult animal. Thus, lobster moult incre­ments are reduced 30-40% in autotomized animals that regenerated limbs, but

F ii/anes, LD. Smith / 1. Exp. A1ar. Bioi. Ecol. 193 (1995) 197-223 207

are only reduced 5-10% in autotomized animals that do not regenerate limbs(Cheng & Chang, 1993). Ecological costs of smaller body size are well known formany organisms (Werner & Gilliam, 1984). If smaller animals are forced toreduce activity levels to avoid predators (Smith, 1995), abandon shelter to largerconspecifics (O'Neill & Cobb, 1979,) or eat less valuable prey (Bender, 1971), thecumulative effects of the initial injury on growtb could be profound.

Functional costs associated with limb autotomy can be ameliorated if limbs areregenerated quickly. The rate of replacement depends on the moulting frequencyand the limb length regenerated at each ecdysis. Autotomy can either shorten orprolong the instar period (Hiatt, 1948; Cleaver, 1949; Skinner & Graham, 1970;Stoffel & HUbschman, 1974; Kuris & Mager) 1975; Fingerman & Fingerman, 1976;Davis, 1981; Hopkins, 19HZ; Skinner, 1985; Spivak, 1990; Cheng & Chang, 1993)depending 0 11 the timing of the injury with respect to the moult cycle (Bliss, 1960;Savage & Sullivan, 1978; Skinner, 1985; Spiv:]k & Politi~" 1989) and number andtypes of limbs lost (Skinner & Graham, 1970; FingcIman & Fingerman, 1976).Such variation in moulting times may be detrimental. if, as Reaka (1976) hassuggested, moult synchronization i:; a strategy to avoid cannibalism. Weis (1976,1977) sh.owed that conspecifk pres.ence ran retard or stop limb regeneration infiddler crabs and suggest~d thaI these responses may be a mechanism to reducevulnerability to conspccifcs, In contrast, I\-loriyasD et a1. (unpubl. data) found thatin an estuarine lobster population, two distinct moulting f'eaks are obvious. In thefirst season (l\fTay-June:' loth norm"l and injmed 10bsV::rs moulted; :n the secondseason (August--S!3ptcl'fb~r),moc;t!y daw-mi~sing individuals r(81-83%) moulted.This second :season may res11lt from delays in modting c<'.med by breakage late inthe previous intermou1t period.

The ability and length cf time required to reg::J.eratf: ('l missiDg Umb to normalsize can vary dramafc211y h~tween life·hi::'ory stpges and species. In general,regeneratio'l. poteHtia1. rnd ~. L1CCCSS decreases '\llith siz~ (313,0) of the individual andstops when th:: c;nimaI reaches ~erminal 2:necdysis (HaTtnoU. 1965; Cheung, 1973;Miller & Watson, 1976; Brock &. Smith, unrubl. dat2l). iFnr example, small bluecrabs are c.hle to regenerate 2l1rno~t. 90% of the normal limb length in the firstpost-autotom:' moult ?x'd 'Jearly 100% of the length regenerated after the secondmoult (Smith, 1990b) r~ f'::HlS~ blue i~r3bs rnoult":.,/p,ry 3 to 4 wk, completeregeneration can OCC:.'f i'1 q fiing11: season. \12 t me ferDal:' hhe crabs, in contrast,are in terminal ane':dy~js an,e{ C3n no longer replace missing limbs. King crabs(Paralithodes ramtschc recover subs':anb'.lJy less limb length at each moultand regeneration of a nmma! kngth limb l':2'i1 t3.ke r.l.--7 instars. V!hil~ th1S processcan he accomplished "/ithin a yep:- juveniles, ct is estim2'.ed to take four to 7 yrin 1"')U116 3iults (Niw21 lr I(WJt3, lC'Ik,d; Edwards, 1977~), !\ variety of other speciesthaI have ">';;~n exa[);"jned huv,,; f'~gC'::-.'~ration ;13'.:':;s ir;term ~diate to these extremes(Emmel, 1907; Hi3rt, 1948: MacGj~iti~: & MacGinitie. lJLi9; Ski:'].ner & Graham,1972; Berl1nett, 197:3; Miller & Wa'soo, 1976; S~]Vage &. Sullivan, 1978; Brock &Smith, unpu~':'L d3t2\

Limb t:ir~S that ?.TC ('r;tk,,1 to av,;,r(\Hfit"'es~ Plight he eX0ected to regenerate

208 F. Juanes, L.,). SlliUl / J. Mar. Bioi. Ecol. 1)3 (i;95) 197-223

more quickly than les3 important appendages, and, in ~(WU:; species, regenerativeallocation varies with limb type ana sex. For example, Kurata (1963) showed thatfor Paralithodes camtsc}zarica , the recovery rate at the first post-regenerationalmoult was greater in chelipeds than in the other walking kgs. Similarly, Hopkins(1985) found that multiple autotomy affected growth and regeneration differentlyin Uca pugilator depending on whether the large male cheliped was left intact orautotomized. Mohrherr (1987), working on fiddler crabs (U. lactea and U.chlorophthalmus), found that the rate of regeneration was sex-biased. In males,accelerated regeneration would result if either of the (;helipeds was autotomized;whereas, in females, both chelipeds had to be aJtotomit:eJ for a similar effect tooccur. He suggested tha~ for males, loss of a minor chdip~;j would severely impairfeeding, while loss of a major cheliped would stro,1g1y handicap mating success.Finally, cheliped reversal foHowing loss of the maj or chr'N may bE; an adaptationfor recovering a fun-size crushing appendage in the least ",mount of time possible(Przibram, 1931).

Regeneration of parts of limbs (e.g. a dactylus of a cheliped) has not beenexplored, but presumably these can rapidly return to normal size in the next moult(Edwards, 1972). Typically, damage to the dactyl or propodus tips does notstimulate autotomy of the entire limb. This lack of response may be adaptive,particularly if the potential for sucb breakage during foraging is high. Althoughforaging costs exist (Juanes & Hartwick, 1990), the missing tips can be replacedmuch more rapidly than entire limbs. Various reports of abnormal appendages(Butler, 1956; Shelton et aL, 1981; Carvacho, 1988; Juanes & Hartwick, 1990)however, suggest that partial limb loss (or perhaps limb JOSS without autotomy)may lead to permanent damage. Moreover, Hopkins (1993) has shown thatregeneration of limbs in crustaceans is most efficient when it foIlows autotomy.

3.3. Reproduction

Limb loss could be detrimental both to individual fitness and to populationgrowth if it significantly impaired reproductive success and was common in thepopulation. Limb loss mi.ght decrease an individual's ability to attract mates, lowercompetitive performance, physically hinder copulation, or reduce fecundity(Smith, 1992), yet few studies have addres~;ed these possible costs. In manydecapod crustaceans, chelae size is criticaIly important to mating success (Stein,1976; Salmon, 1983; Christy, 1987; Snedden, 1990; Garvey & Stein, 1993; Claxtonet aI., 1994), therefore their loss would be predicted to be especialIy costly.Among field-collected Carcinus maenas, mating males have larger-than-averagechelae compared to the entire male population (Lee & Seed, 1992), and limbdamage has a negative effect on male pairing success (Sekkelsten,1988; Abello eta1., 1994). Sekkelsten (1988) noted that medium-sized damaged males were ableto carry a female in precopula to the same extent as normal males but tended tolose their mates before copulation. Abello et a1. (1994) found that the proportionof male crabs with missing chelae found in mating pairs, both in precopula

F. Juanes. LD. Smith I 1. Exp. Mar. BioI. Ecol. 193 (1995) 197-223 209

(15.9%) and copula (12.9%), was much lower than that found in the adultunpaired (24.8%) population. I.P. Smith (1990) noted that injured Necora pubercrabs were less active than uninjured individuals and were never found in sexualactivity (pre-copulatory, copulatory or post-copulatory pairing). Similarly, few(51176) mature male tanner crabs, Chionoecetes bairdi, with regenerated clawswere found grasping mature females (Stevens et aI., 1993).

In one of the few experimental studies to date, Smith (1992) showed thatguarding male blue crabs missing one or both chelipeds were at a significantdisadvantage in trying to prevent displacem;:;nt by similarly sized intact males. Incontrast, intact males in precopula were able to ward off both injured and intactintruders. This matinglai'dicap Vias Dot eviden: in CalLnectessapidus males inthe field (SrTith, 19911. Th":;re, L)aired and unpaired males did not differ in thefrequency of limb injmy. These differences hetween exp;;rimental and field resultsmay have reflected: (1) greater !)pportunity for handkapped Buarding males toescape in Ih::- field; (2) a, knv inddenc:: of autotomy in the study population; or (3)confounding effects of 'male brxly size during ccmpetiticr.. If body size is the chiefdeterminant of male mClting succe~:s in this :~pecies (Smith, 1992; unpebL data),then limb lms will be important in competidv,:, bouts where the sizediffeJence bet\veen oppJnent<; is smal1. In a similar experiment using shore crabs,AbeHo et <11. (1994) fonf:d that chela loss was a handir<lp for a male crab wheneither ~orrpetin.j for];" defendirg a paired pn~molllt female, The authorsestimated har.,':1;(np 10 be equ~va1eTY :'0 a re::iaC:':1on si7,e of 7-8 mmcarapace 'width to the si7rA th~ oYDpetitciL H("d~ver, in C0ntra~',t to bluecrabs, fi~ld·cone':ted shore CT':1ibs exhihit 2. rnating dic8P due to limb loss (seeabove; Sekkelsten,198F: Abeno r;t aL, 19~Mi,

The effect of autotomy on fernale reproductive pr;rformance has receivedremarkably litIe altenticn, given fIC potentic;J for dam?ge to the female duringcourtship and trade-oft:" thct might exist betwe~n fecllndi.ty and regeneration. Instaged competitions in pon1B., prepHb~rtal femr,k bh;e sufferer.! significantlyhigher limr Im~;!:hafL m;;les, mo~;~ probabl;r as 2. re",ilh c{ cJrnbsJ between theguarding mId intrudin§Ct1<Jles (~}mith, J9'·;7,) In F:~ld, p()jred prepubertalfemale~ ShYNr:c1 a t: ilk~cy tov.;;mi highc' frequency of limb loss than didunpaired;:,rqmb::rLll P~ecq:mlatory interacticm may alw he c'esponsiblefor limb darn8I;e in 5')xies of genu:; (Hubc!l' 1985), WI1fm singleffi,dcs and ff'mr,J.;->,S 'tJc:r~ r:rp,entr:l to ead'. other, injuries were relatively low (6%of trials). 1njuri es were -rDre fHl).l',:nt in Tale (tYVO fernales and one male) andfemale (tuvo maIer: ard C'1e hma]I;) mate choice en)cr;ments :md reflected the, ,higher levels of :;)gue~:}';ml in thef:·<; trials. Ynterestingly, more injuries occurred tothe sex being chosen tl.;?P to the "choosing se"'(" (ie. to kmaJ.es in the mnle-choicetrials and j,,) m3Je~; ir thf' fem(l\e,r;hoici~ tri~ls\ PeD}'I:d',;c·;,ve t:rar.k-oft; betweenfecundity 2"10 f':c:en :1fC we;,1 kl0wn h li~'.''1''ds f1nd calamanders (e.g. Smyth,197 4: Mai0nn::. 1977 & Fit:,;)')trick, 19£[n hut, data exist for crusta-ce'U1S. NCr'r-n "': JClre~ (993) <;hr)wed that field-cC'ller:tf'd female velvet swim­ming (r3d)'~ ("/ecC)yc Tubor) with missing h?d s!1'<Jller hrood ei:r:es than

210 F. il.ann, L1 .. S:;'itJ: i J. i:xp. Ajar. Bit,/, l~(ol. i~}; 119';5) 197-.:'23

uninjured females. Although the mechanism is Unknc.JiWL, the authors suggestedthat ene;:gy available fOf reproducti,;)n was shur:ted to glc,~th and r~pair.

3.4. Intraspecific aggression and competitive ability

Limb loss also has the potential to alttI the ou.L:on~e of cor;lpetItlOn forresources. Reduced ability to obtain or deferld sheller ix)uld result in mortality ifpredation intensity wer~ high, or lower reprodu.ctive fiiuess if mabng oppor­tunities were lost. Among stcmawpods (whicll are similar \.0 decapods in generalbody plan), loss of onc or both raptorial appendages severely limited fightingability, so that injured animals were less able to defend cavities or evict intactintruders (Berzins & Caldwell, 1983). Similarly, cheliped loss reduced competitiveability for refuge in alpheid shrimp (Conover & Miller, 1978), lobsters (O'Neill &Cobb, 1979), and hCITllic crabs (Neil, 1985). Blue crab m.ales missing Olie or bothchelipeds were hampered in their ability to guard female .• from intact intrudersand were unable to displace intact males (Smith, 1992). It should be noted thatthese agonisitic interactions are not likely to be a major source of limb loss. Incrustaceans, most intraspecific competitive inter~lctions are highly ritualized anddo not result in autotomy (e.g. Hiatt, 1948; lachowski, 1974; Sinclair, 1977; Hyatt& Salmon, 1978).

3.5. Predator avoidance

If the primary function of autotomy is to prevent subjugation of crustacean preyby a predator, then prior limb loss should increase prey vUlnerability in sub­sequent attacks by impairing defensive capacity or escape ability (Bildstein et aI.,1989; Davenport et aI., 1992; Smith, 1995). The net effect of these handicaps onsurvival will vary with the extent and type of injury, escape mode, and thebehavioral response to such injury. Chelipeds are important defensive weapons(e.g. Robinson et aI., 1970; Stein, 1977; Lawton, 1989), and crabs missing claws aremore vulnerable to vertebrate and invertebrate predatOfs than are intact crabs.For example, predator) diamondback terrapins (lHalaclemys terrapin) showedstrong preference for shore crabs without chdipeds (DaH~nport et aI., 1992), andibises (Eudocimus albus) selectively ingested female and declawed male crabsmore often than intact male sand fiddler crabs (u. pugilator) (Bi1dstein et aI.,1989). Using large (10 m 2

) field enclosures, Smith (1995) showed that chelipedloss increased juvenile blue crab vulnerability to cannibalistic adults, hut onlywhen escape was restricted by a tether. Untethered, injured animals experiencedsignificantly less mortality, illustrating the importance of escape in this species.Surprisingly, severely injured juveniles (missing both chelipeds, one walking andone swimming leg) suffered relatively little mortality, suggesting that injury mayalso alter activity patterns so as to decrease vulnerability. While reduced activitycould improve chances of surviving until limbs regenerate, such behavior mightalso decrease feeding rates or mating success (Sih, 1982, 1992). Limbs other thanchelipeds may also be critical for predator avoidance. For example, loss of a single

F JUanfS, L.D. Smith! 1. Exp. Mar. Biof Ecol. 193 (1995) 197-223 211

swimming leg or asymmetrical limb loss can reduce escape speed (Spirito, 1972;Smith, 1995).

If prey vulnerability varies with ;arey size, then the relative costs and benefits ofautotomy will change as the animal grows. For example, small Callinectes sapidusjuveniles «61 mm carapace width) tethered in the open field suffered similar,high levels of mortality 2nd injury regardless of the type (chelipeds, walking orswimming legs) or number (0, 1,2, or 4) of missing limbs (Smith 1990a, 1995). Incontrast, larger juveniles (61-110 mm CW) experienced lower mortality rates, andtheir survival depended em the number of prior limbs lost. Severely injured largerjuveniles had significantly higher mortality than similarly sized intact crabs orthose missi:lg one or two limbs. Adult intermoult blue crabs (>110 mm CW) werenearly invulnerable to attack in the Rhode River, Maryland. For small crabs, then,benefits of ;mtotomy (incre<lsed ~urvivorship:1 weL': relc.i tj\'ely high, 'J1hile costs ofprior limb loss (incTca:ccc' vaJnenbUity) v/ere low. fer crabs nearing a size refugefrom predation, minor limb Joss 'Nag still useful and carried no penalty, but severelimb 'coss imposed a high cost A simi'ar ~:ize-dependentpattern was observed ingreen cr31b~;, Carcin.'ls r/{icnas, exposed to diamondbar,},( terrapins (Davenport etaI., 1992). Small crabs 00-25 mm eVil) viere h.ighly vulnerable to attack inaquaria (>80°1.) cafel', <20% Croprr~d); r1ccEvm crabs (30-50 mm CW) were lessvulnerable «20% eaten. >60% cropped); and large CTi1tn (52--75 mm CW) wererelatively invulnerable ro attack (O%eateo i 10% cropped).

3.6. Survival

Ultimately, we wcuIcl liJre to know ?lh;~lter (HJ~otomy affects sL'xvivaJ. incru:;tzrceans, but it is to assessmorta1.ity in reld or reliably extrapolateestimates from laborrtnry experiments. VariollsswthJIs have not.cd bigh mortalityof autotomized individuab (e.g. Simonson) 1985). Figi.d & Miller (1995) showedthat in the laboratory, ·::helotomized cray5sh have: signifIcantly tower "mrvival thanintactcnilrrals and t'n+ these surrillal ra!e~ were den:;ity·dependenL Cleaver(1949) usirg data obtained from tag recoverie~; fmm Dungeness crab fisheryestimated that animals missing one claw suffered 20-30% higher mortality thannormal crahs, and crabs with one claw and one walking leg missing experienced40-50% greater illorrality th(ln normal Cf3bs. The ~tonf.' crab fishery in SouthFlorid3 is ';oique in t~wl (ely claw:; 8re har'IPsted deckwed aricI'.als areretmned to the w3ter to Tegenera1e new dmvs (Lindherg & fVf3',n:baU, 1984). Inone exper;,cmert. thow~~. stC'o.e cr?b:; sllff~r·;~d high m()r~ality after c1a'N removal;47% of crahs exrerie··;in~~ couhle che1o~omy (Fed, re2S 28% with single clawsremoved di;d, uS'JaIlv w',thin 14 h of ded;r,ling (Davis ct. aI., 1978). The h~gh

mortalities e;cpeicnccG by d"m3~~d individuals of ttis species may be due, inpart, to the: di:;propnrhnnatel:" 1m,?'? size of s~orlf; crab dauvs, but also to whetherclawf, ?ire bmken along the natura! fracture ':'lane (Sim:lnsoll & Hochberg, 1986).Labcr2torll~xper:mentshne shown that the width of the wound following clawremoval cDrrelates signi6cantly wtth SU:viV;l; (Davis ct a1., 1978). Based on thisrebtionshir and comm~rci?l c3trhe~, the aHtl'ors esim:1ted mortaHy rates from

212 ii )uanes, LJ) Smith, .,. Lxp. Mar, BLiJI. Eenl. 19,) (]"J',l) l'S'1-,"Z,

23--51% following claw :,'ewGvd, TLese: resulLs, c;xLb{nCG wah the 10\/ regenera­tion frequencies lin field pOpU,2clCI1S (Savage ;,;( al., Li7~i) and LIte Ic,ct (hal clawsare not :ikdy to regcl1c;rak CLick \.0 norme,," :si2:c anlrnal~ rca;:;h theirterminal moult (Cheung, 1973), suggest that declawing is of limited value as asustainable fishery management technique (but see Sullivan, 1979).

Survival can be affected by partial breakage as well as limb loss. The spannercrab, Ranina ranina, the:. only br23:hyuran reported to be incapabie of autotomy,showed very high mmtality (60-100%) after losing one or more limbs andmoderate mortahty (50%;' in arjmals losing one or rlime dactyli (Onizuka; 1972;Kennely et al., 1990). U~ub damage :)I loss may also increase the suscepLibility todisease. For example, fvl1oriYZl.SLl ct aL (unpnbL data) have reported that claw lossin lobsters leads to inCJea5ed:.:lisea~;e:.In their work on lohsters, the) noted that 15animals missing claws WEre in(c~;kd '\vhh an uHideDtified ciliate species and diedwithin 10 weeks in captivity.

3.7. Population and communi!} effecls

Most experiments 011 autotomy have focused on tbe consequences (primarilycosts rather than benefits) of the response to the mdividual. Predicting whateffects autotomy might have on the population and community has largely been amatter of extrapolation and supposition. Theoretical models suggest that nonleth­al injury could result m population stabilization if (l) injury rates \','ere density­dependent, and (2) injmy signifIcantly reduced surviva: Of reproduction (Harris,1989). This regulatory effect requires close coupling between predator and preyabundances; such couplting is most likely to occur if predator and prey areconspecifics. Figiel & Miller (1995) have shown that lirnb damage is related todensity of conspecifics in crayfish and that mortality is dcm;ity-depcndent amonginjured animals. Similarly, unsuccessful predation by conspecifics is thought to bethe primary source of injury to blue crabs in the Rhode River, Maryland, and thefrequency of limb loss there correlated positively with annual blue crab abun­dances between 1986 and 1989 (Smith & Hines, 1991b). The critical question interms of population effects, however.. is whether common forms of injury meet thesecond of Harris's (1S\~9) cri1eria; i.e. do these injuries significantly depresssurvival or reproduction? Based on the most complete data Set available for asingle species, Smith (1995) concluded that the costs of single cheliped loss (themost common injury) to blue crab growth (Smith, 1990b), foraging (Smith &Hines, 1991a), mating success (Smith, 1992), and predator avoidance (Smith,1990a; 1995), were sufficiently small as to not affect the population. It is possible,however, that minor costs could become major ones during years of higherconspecific abundance, lower prey densities, or other unfavorable conditions.Undoubtedly, the severity of autotomy's effect on fitness will be species-specific.Many of the studies reviewed here suggest that limb loss can affect survival andreproduction, but comparative work is needed. Discerning the effects of injury oncommunity processes will be particularly challenging. One of the most promising

F. ./uanes. LD. Smith! .!. Exp. Mar. BioI. Leo! 793 (1995) 197-223 213

areas for study will be to examine direct and indirect effects of reduced foragingefficiency on prey popu~ations.

4. Prospectus

Unsuccessful predati:m is sufficiently common in the animal kingdom (Vermeij,1982) that nonlethal injury is likely in the lifetime of any animal. Most ecologicalexperiments, however, ignore wounded or injured individuals (Juanes, 1992).Given the prevalence of injury in decapod crustacean populations. the potentialCOStS involved, and the ecologlCal importance of many species in the community, itis surprising how little attention has been focused on the ecological. ramificationsof injury. In part, this is because the issue is complex, the incidence and effects ofautotomy will be influenced by myriad factors, including phylogenetic history,for8;ging ,'1'... ,1 locomotory modes, habitat type, nature of injury-causing agent,predator and prey densities, repair rates, size structure of population, and sexratio. As n rule, mmT comparative work is needEd on decapod species fromdifferent h2bitlt, (e.g t.errestrial VS. ma6r.e) and wit\- different lifestyles (e.g.mobile VS. ;,ec1er:tary) bef:>Tf; Wf; can w~m~rali:[c about the:-osts aId benefits ofautotomy. Beloyv, we pIes~nt seY'3raI arem we c:e need of siudy,

4.1, Causal agents

Autotomy is presumed to be a defense against predators (Robinson et al.,1970), but tnis function bas rarely been observed or tested experimentalliy. Suchinformation is important if we are to explain patterns of limb loss m individualsand understand the generaL diicacy of the response m predator-prey encounters.Despite artifacts associ2cled with tethers ana enclosures (Peterson & Black, 1994),their use, in conJunctwn with a dme-Iapse underwater video system, may be thebest way to identify sources or injury Hi mocLie species in subtIdal systems.Subtidal species that maintain [erritories, and certam intertidal, semi-terrestrial,and '(errcstr~3; f>pecies, however, might be monitored for injurious events withoutresorting to unnatural restraining devices. Once principal injury-causing agentshave beer- identified, videa sys>:mE mr.:.y be used to study the kinematics of attackand aV'Jic'ar.ce in det?.!L It :s impcrtant~o ;-ealize that the degree of Testraintutil];~ed in sll1ch?xperimenls will influence interpTi:tsJiom; at the !Jtility and cost ofautotomy For '~':;l!npk cndo:mres (or win iacn:2se enccnmterrate:, pn:dalCc" ;CVi;rmcs for Pj":~Y e~;,~ape. Th:seexpcn:::ncnh, ir- effect, :111011 s';queuce (Eridler,1985), one m tbe UYi captm:c if) whichautc·tc-my i~ ,nosi like~y Ie el:1p)'jcd. In SJc!l ,::,ECS, i,:JtoL-:orny v/ill appear toha'.e la;-gf' 'mp~vaI rCI;!?fil'; and p:ior los:: will have high costs En eGntrsM, ifexp,:-rimen t.~ are condlCtcd in large cndosmcs2nd prey are free to mO'/e, preyrna:' utilize avoidan::e nlc:h.Jn;sms other th8f\ autot:o",l)' (eg. swimminr., burrow-

214 F llwnes, LD. Smii/z j I Lep. ,Aar, ,'1ioi Lcul 1>:1 (J 19/-223

ing). Und2r 'd.lese cin':UBEi,.1.1lCCS, fll iur' loss may s~sl!l to impu~,e .:itlle cost(S '., Q9~'HUtl'l, _/ )).

Identification of princtiplc injury· causing agents and their size-and dcnsity­dependent aHacl< cfficie,Jcies are necessary if injury frequencies are to be used asan index of predation pressure (e.g, BaHinger, 1979; Aronson, 1987; McCallum etaI., 1989). A number of "uthurs (ScL'Jener, iCr;'9; Ja.<..sic ~[ Fuemes, 1980; Wilson,1991) have indicated th3.[ inj Uty frquency f1,;:;') be a poeI' predictor predationintertSlty, and that additonal data 0n injury rates, pred2l(Jr dficiency, and preysurvivorship are needed. Injury frequency lr.ay prove fIJOSC useful as a measure ofpredation intensity in simple precatorprey :.;yskrHs in which densitie~ correlatedirectly 1vith injury ane: indirecUy wrvival (e.g. if the principal predators areconspecifics;Morin, 19155; 'Petranka, LBt;; \':.111 13 uskirk & Smir:1, 1(91).

42. Behavioural decision"

Behavioural decisions of all <mirna! can be strongly Influenced by its perceivedrisk of danger (see review, Lima & Dill, 1990). If prior limb loss increases ananimal's percept~on of its risk, then autotomy could indirectly impact foragingtime and efficiency, habitat use, and mating behaviour. To understand fully theimpact of autotomy on fitness, then, experiments mW;I201''fAparC behaviours ofinjured and imact crustaceans in presence and absence of predators. We alsohave little knowledge of the decision-making fHocess that determines when preyshould or shoukj not auto[(J.nize d limb (e.g. Easton, 1']72; Weis, 19Ti). Does thisvary with th.e threat imposfcd by il1e predate'l (e.g. is autowmy used more readilyfor more dangerous pn.:dators) or alternativciy as [he level of prey vulnerabilityvaries (e.g. size, injury, or moult status)?

4.3. Life-history variation

Arnold (1988) has argued that the value of autotomy depends on its cost:benefitratio and will be selectively advantageous only when benefits exceed costs. If themain benefit is escape from certain death then almost any cost can be borneprovided the animal survives to reproduce. Nevertheless, if vulnerability topredation is a function of body sIze 01 ontogenetic sll'.fts Il1 habitat or resourceuse, the costbenefit ratio is likely to change during the cmmlal s lifetime. Does theresponse diminish or disappear with age or vary with changes in habitat andresources or must an autotomy response always be maintained for use nearvulnerable moulting events? Are animals with terminal moults more reluctant toshed limbs than those that continue to moult throughout life?

4.4. Functional costs

More comparative studies on functional costs in decapod crustacean species areneeded. Future work should attempt to determine: (1) whether individual costsexist, (2) how long costs persist, and (3) whether these costs are sufficiently great

F. luanes, L,D. Smith J. Lxv. Mar. Bioi. Leol 193 (l995) 197--223 215

as to affect processes at the population and community levels. Undoubtedly, thecosts (and benefits) of autotomy will vary with important external factors (e.g.food availability, predator density, shelter). Whenever possible, studies shouldmanipulate one or more of these variables to determine the conditions in whichautotomy will be most useful or costly.

Foraging costs: Juanes & Hartwick (1990) suggested that dactyl breakage andclaw tooth wear were direct consequences of feeding on hard-shelled molluscanprey. Durophagous crabs are able to crush large clams by repeatedly loading themand eventually c3using lovl'-cycle fatigu:: (Bouldi'1g & Lc.Barbera, 1986). Similarfatigue damage, however, may occur in crab claws and ultimately lead to clawbreakage. This risk of damage may be; partially rc:;poJ1fli'bc for the obsen1cd trendof .:)fefcrencc for small· sized 11'01luscan prey by decapod crustacean:; (Juanes,1992). Do::,; foraging hehaviou'" lead to limb loss, pcrtia l claw ~)Teakage or wear inother species? How doe" limb ioss or the threat of limb damage affect preychoice? Prey diet in c:-m:tac:-ans ofter changes during ontogeny (e.g. Stevens et al.,1982; Edgar, 1(90):, is limb loss mwe costly to foraging efficiency 2it certain ofthese life stages? Impo:-tantly, does reduction i11 foraging efficiency :lffeet growth,or are injur,~d animab able to switch to 2\hernatiyc prey until regeneration iscompleted? Finally, ?re fJraging costs suJfkientl;7great and injuries commonenough to affect prey popl11ation dynamir:::;?

Mating: Auto·.cmy [1ukl hnve its gre'31 ::.t irr:lx1rt 'J~ indi'\idu<} fitness andpopu.lation dynamics if effort is comprnmi~;ed, yet thi~ ofres'~~arch rcmz1r:s larJr:ly m:;-xplor;;d, f'rachym2T, i~r3bs exhibit i>. div 3rsity ofmating ,'?3sociations f8p.gi.ng from female· or ;-esnnrce·:::cntered comgetition toones based pure:y on ·~nC)IH'fer HJte (Christ.y, 1987\. The cost of :mtotomy mightbe expected toincre21c;e as the intensity of sexlJitl sel'~cljon jn~rer!f,e:s, bl't perhapsto differe'll. degre,~s de:: ending rm the tYI)(~ '':if For e'Gunple, dlelipedloss in 2. tcnitorial ht::-Ollif:r (e.g. Uca) may hs:.re much h;.p,her reproductive coststhan ,in a mobile species that Interc~pts vJ:l"1der:ng fe-nales (e.g Hemigrt:p.nls).Does prior limb loss affect male competition for females or male/female choice?Is iimb Joss a ke~1 f:tetr·r to InPti'1g 'mccess')~ arc other characteristics (e.g. bodysize, bu-rrry.>l Ct'Rlity) tn0re~litica1" Wt.'J:t jf 3r! ,·-"o.rgRtic trad'f' off:;; existbetwee- ;efrJ cl CornprP', tiw <;'udies te:.:,;ng for trade-offsbetv!cen dnse1y rclatr:d sl=)!;;C':fsNilh and wi'.hwt mOillt~ might p""oveillumjnT~ing B~! extrapol;>ht: the type and fre(ju(f!c.y of injury h a population tolost fecundity and JTI?tl:IT7 op:,ortllni'ies, 'Nf' car f~'V,tirnote" of its potc:ntialimpact or pfJrm!afi8n

4.5. Tpmpc

DC\ta 'JI1 cat':eflJ.~ of {e.g limh f;'pe (mel r'umher, siz~ s':x biases)in study pr::rulatioDs arc f':".,;enti" ';f i:xp,,>,iiy·':.onts de";r:n(';r to examin~ C0f>'S are tobe eco!<=-g'r,:;iPy r:lev?L T::J ardir.; :)11. :'-ecorjs Of krrmc-,,1 and rc:,g;anhic varhflionin iniur;! h:>'1ten:::y c"'~ neces~ar;' to de'(:rT"lire ho'l; sekctivr: hrces r~'ight varyover ~ur:h sea lese For f'xample, axe th~re ~;pecifk ?Te2S where damage occurs or

where damaged allimais aggr~gate? Results of several smdies suggest that smallestuaries may be places where damag..:d animals come LO moult, perhaps attractedby lower predation ris~:. (ShiLey ct aL, 1990) or higher tempenjurcsllowersalinitIes (wbich can facilitate moulting and regeneration, deFur et a1., 1(88). DoesauLYcom.:r accurately reflect ;'H:da<:wn pressure d\U time: or space (e.g. intertidalor latitudinal gradient~;)? Can iniu~y frequencies be correLated with dominantpredator densIties? If cdJmibatsm is a primary source of nonlethal (and lethal)injury in many crab populations, then tile potential for density-dependentregulation cx~sts (Hani~;, 1989). Good estimates sc.lfvivarship are needed.Mark-recapture techniques will bc most effcc:lve Lll mere :,c:oeniary species or incommercially impGftaJJt species.

4.6. Modelling

Model~i are needed tu examine the dfcus of autOl\Jlny un indiVidual behaviourand popuKation dynamics and to generate testable hypotheses. For example,future work should explore an animal's decision to autotomize a limb using adynamic programming approach, one which considers the current physiologicalstate of the animal (Mangel & Clark, 1988). These models can also help elucidatethe deci~ion to accelerate or delay the next rnoult dep,::;ndi:1g on the physiology,age, predation dsk, and reproducti/e status of the animaL To allow a betterunderstanding of populaticJI: dynamics, mojd~; should incorporate size-dependentinjury frequency and p~,patr rattS and COStS of injury to survi'\lal, growth, andreproduction. Such madd" would be parti\~ul&rly useful for commerciany-im­portant species where lirnb 10s5 c£on have considerable impaCl:s en recruitment tothe fishery (e.g. Davis, 1981).

4.7. Conclusion

The phenomen;:m of autotomy in decapod crustaceans provides an outstandingexperimental system for studying physiological, behavioural and ecological trade­offs. Experimental removal of different types and numbers d limbs can be used tomodify an animal's phy:;ical status ,md, presumably, dCClsion-rnaking processes.Assessment of injury patterns within and between species may provide evidenceof how selectlve factors influence an an.imal's behaViour and life history. If costs oflimb loss are significant and the frequency of injury high in a population, thennonlethal injury could have profound effects on the population and community.

Acknowledgements

We thank R. Hughes and R. Seed for the opportunity to present and publishthis work as part of the decapod workshop. Travel funds were provided by aUniversity of Massachusetts faculty travel grant (to FJ). We also thank l.P. Smithand c.P. Norman for sharing unpublished data and long lists of references, and M.

F. Juanes, LD. Smith I 1. txp. Mar. BioI. Em!. 793 (7995) 197-223 217

Moriyasu, C. Figiel, and J. Shields for sen.ding !Us unpublished data: and manu­scripts. Two reviewers made helpful suggestions on an earlier version of themanuscript. J. Myers and J. Smith were very hospitable hosts during the timespent writing this paper in Vancouver.

References

Abby-Kalio. N.J. & G.F. IVarn':L 1989. Hetsro.:hdy and hadeclness in tr,e shore crab Carcinusmaenas (L.) (Crustacea: Brachyura). Zool. J. Linn. Soc.. Vol. 96, pp. 19-26.

Abello, P.. l.R. Pcrti~rra & n G R~id, J990. Sexu"l dimorphism, n~lat've growth 2nd handedness inLiocarcinus depurator and Maeropipus 'uherculatus (BrachYlJf3: Portunidae). Sci. Mar" Vol. 54, pp.195-202.

Abello, P, C:.G. Warril.an, D.G. ;{eHj & lC:. Naylor, 19{,,4. Chela ;(;SS in the si10re crab Carcinus maenas(Crustacea: Brachyura) and it~. effect 011 ;-nating success. Mar. Bioi., Vol. 12], pp. 247-252.

Abrahamsson, S.A.A., ] 966. Dynamics of an isolated population of the crayfish Astacus astacus Linne.Oikos, Vol. 17, pp. 96-107.

ap Rheinallt, T. & R.N. Hughes, 1985. Handling mcthods used by the velvet swimming crabLiocarcl!lllS puber when feeding on molluscs and shore crabs. Mar. Ecol. Prog. Ser., VoL 25, pp.63-70.

Arnold, A., 1988. Cn:dal aut "(Y:Jy ,is a deL'n:;e. In BiD I,'gy'he R~.t:tili(!, ~~?l. 2,' Ecology B, editedby C. Gans &. R.B. Hue). A.H.. Liss, New York. pp. 235-273.

Aronson, R.B., 1<.187. Pred;1t'('fj or fmsil ar.d ,'ccenl orhiuroids. Vol. 13. pp. 187-192.Ary, RD.. CK. Bartc:lI & ~lr.A Poirri",· 19R7. The effects of chelotGmy on mobng'Il the blue crab,

CLillinectes sapidus. J. Shellf. Res., VoL 6, pp. 103-108.Ballinger, R.E" 1979, In'.l'l5,Jc;c· fic V31iatio,1 i" demog,a,Jfly and FL h';tory of th<; lizard. ,S:elOpoTUS

jarrovi, ,!lone; an alti'i.'din;j; g::ldie:t in w'lth~3stcm AriZOiJa. Fco!ogy, Vol. 60, pp. 901-909.Barnwell, F.B., 1982. The pr~vaJence of male right-ha"ldcdncf''' in the Jndo-West Pacific fiddler crabs

Uca vacans and U. letragonol1 (Decapoda: Ocypodidac). 1. Crust. BioI., VoL 2, pp. 70-83.Bender, E.S., 197] Studies 1;/ the life history of ihe Hone crab. Merippe mercenaria (Say), in i11e Cedar

K,'y area. h1:Sc. thesis, University of F1or id3, Gaine,ville, Florid", 110 pp.Bennett, D.B., 1973. The effect of limb loss and regeneration on the growth of the edible crab, Cancer

pagurus, L. J. Exp. Mlr. F!iol. f::'cn! Vol. ! -'. r:) 4<;' -53.Bertn{~ss. M.D., 198';. Fiddler 'crab r::gu!,'tlcn of abcl:iffora production OIl a New England

saltmarsh Smleg)!, Vol. 66. fT 104/--10.":'<:.Berzins, LK. [':{ R L ('aldw,~'1 1983 The effect of 0'1 t01:; hch':vicr of the <tomai'opod,

Gmodaetylu:; nr('dmi (M8f.iHi"g}. Mer. Bfh<lv. Phylio!., Vol. 10, pp. 83-96.Bilc1stcin, V.L.. S.G. 1'·1'(:1:>:-"121: & I Ihs"in nW) ",eX_HI 6m'lrphism in :;.md

fiddler cnb,; fica jJltgil '1/W: ,]i f f,:"cTli,',l vIlInerabili'.y· 10 avian p,,'~d:!ljGIl. l.nim. Behlu., Vel. 37, pp.133-139

Bliss, D.E., 1960. Autotomy and rcgenemtion. Tn, The pJ"Y.I'iolcgy of Crustacea, Va!. 1. cdited by T.H.Wz,terrnan, Academic Pre';'; New York, [Jp. 561-58°.

Bouldir;g, E.G. &. M. LaBarhcrz:, 1986. Fatigu', d:rTi"ge' repee:!c>:) I"'ld'ng enables crabs to open largerbiva]o;e:;, BioI. Ball Woods }-foil' Mass, VoL 171. pp. <538-547 .

Briggs. PT. & :~·.M Musha;;!::' 1979 The; Ar:i'ricfI lolYl:~r in westcPl Long Island Sound. N. Y. PishGame J. VC L 26. pp 59- W,.

Brigg<. PT. & EM M:'shar:l;c. 198(' The: i\:nerican lotTiter ?nQ t'il: p0t fish"ry in the im;hore watersoff the S'lUt:1 ,hmc ()f to::g IS'3111, New yc,ri' .Y !'i.<lI '.::1''''' Vol. 27 pro )')6-17g

Butkr, T.H ]956. A boon.,..,,) cF 're P,Tific edible crab, CarL'a I'wgistcr Dana. Province ofBritish enlumbia.' Provi'1 r ;eI ,H''',Hllm 1>'0 },·t for f('(' .!'U1T I Q"'6 Pc'. (,7-69.

Carvacha. A. 1988. On ,O'l''C i' J)nnr mi1!ities 'n lh,} chdae of (:oP1'ncrcictl cI'2hs (GC!".1S Cancer) fromBaja California, Mexico. Calit: Fish Game. Vol. 74. rr. 172 174.

218 r. JUlines, Ii) ,'mili, 1. Ltp. Mur. Bioi. Ecol. J9.: (i

Chapman, CJ. & A.L Rice, 197L Some direct observations on the ecology and behaviour of theNorway lobster Nephrops l1orvegiclls. A1ar BioI., VoL 10, pp. 321--32lJ.

Cheng, J.H. & E.S. Chang. I;"93. Determinants of po~tmolt sile in the American lobster (Homarusamencanus). L D, is the cntlC<Ji stage. Can . .I. Fish. Aqua!. Sci .. VoL 50, pp. 2106-2111.

Cheung, T.S., 1973. Experim.ents on the simultaneous regeneration of claws in the aged male stonecrab, Menippe mercenaria (Say), with speci,;! reference to the termmal moll. HI/ii. Insl. Zool., Acad.Sin., Vol. 12, pp. 1-11

Cheung, T.S., 1976. A biostatistical :,tudy of the functional conslslcncy in the reversed claws of theadult male stone crab. Menippe mcrcl'Ilaria (Say). Crustaeeana, Vol. 3L pp. 137-144.

Chittleborough, R.G., 1975. Environmental factors affecting growth and survival of juvemle westernrock 10bHers Panufirus iongipe; (Milne·F:dwards)). II liSt. 1. Mar. Freshwater Res., VoL 26, pp,177-196.

Christy, J,H., 1987. Competitive mating, mate ChOICC and mating associations of brachyuran crabs.Bull. Mar. Sci., VoL 41, pp. 177·-191.

Claxton, WT, C.K. Clovind & K\V Elner, 1904 Chela function, morphometric maturity. and themating ernbracr; in male S,.10W crac, Cnioflo/,cetes 0,01110. Can. j. Fish. Aquat. SCI., Vol. 51. pp.IlJO~l1iK

Cleaver, F.e., 1949. Preliminary results of the coastal crah (Cal'lar magister) investigation. Wash. Dep.Fish., Bioi. Rep. No. 49A, no. 18:2.

Coen, LD., 1988. Herbivory by crab:; ami control 01 algal epiblonts on Caribbean host corals.Oecologia, VoL 75, pp. 190-203.

Conover, M.K & D.E. Miller, 1978. The importance of the large chela in the territorial and pairingbehaviour of the snapping shrimp, Alpheus heterochaelis. Mar. Behav. Physiol., Vol. 5, pp. 185-192,

Davenport, J., M. Spikes, S.M. Tll<'rnton & B.O. Kelly, 1992. Crab-eating in the diamondback terrapinMalaclemys terrapin: dealing With dangerous prey. J Mar. Bioi. Assoc. U.K., Vol. 72, pp. 835-848.

Davis, G.E., 1981. Effects of injury on spiny lobster. Parwlirus ar!!lIS. and implications for tlsherymanagement. Fish. Bull., Vol. 78, pp. 979-984.

Davis, G.E., D.S. Baughman, .D. Chapman, D. MacArthur & A.C. Pierce, 1978. Mortality associatedwith deelawing stone erabs, Menippe mercenaria. South Florida Research Center Report, Vol. T-522,23 pp.

deFur, P.L, D. !'-iusbaumcr & R.i. Lewis, 1988. Physiological aspects of molting in blue crabs from thetidal fresh-water Potomac River., Virginia. 1. ems/. Bioi., Vo!. 8, pp. 12-19.

Dial, B.E. & L.c. Fitzpatrick, 1981. The energetic costs of taii autotomy to reproduction in the lizardColeonyx brevis (Sauria: Gekkoniuae). Oecologia, Vol. 5 .. pp. 3\0 317,

Durkin, 1.'1" K.D. Buchanan & T.H. Blahm, 1984. Dungeness crab leg loss in the Columhia Riverestuary. Mar. Fish. Rev., Vol. 46, pp. 22-24.

Easton, D.M., 1972. Autotomy of walking leg~ m the Pacilic shore crab Hemigrapslls OI'egoncnsis. Mar.Behav. PhysioJ., Vol. L pp. 209-2,7.

Edgar, G.1 .. 1990. Predator-prey interactions in seagrass beds. II. Distribution and diet of the hluemanna crab POrlw/Us peiagiClis Linnaeus at ('Ii;'! Head, Western Australia. 1. Exp. Mar. Bioi. £col.,Vol. 139, pp. 23-:12-

Edwards, J .S., 1972. Limb ios~j and regeneration in two crabs: the ,.;ing crab Parall/hodes camtschaticaand the tanner crab Chiollol'cetes bairdi. Acta Zoo!.. Vol. 53. pp. 10'1112.

Elner, R.W" llJ80. The influence of temperature, sex and chela size in lhe foraging strategy of the shoreerab, Careinus maenas (L.). Afar. Beha.'. Phvsiol .. Vol. 7. pp. 1.5-24.

Emmel, Y.E., 1907. Regeneration and the question 01 asymmetry 1I1 the big claws uf the lobster.Science, VoL 26. pp. 83-87.

Endler. J.A., 1986. Defence against predators. In, Predator-prey rdatiollShips: perspectives andapproaches from [he study of lower vertebrates, edited hy M.E. Feder & G.Y. Lauder, University ofChicago Press, Chicago, pp. 109 134.

Estrella, 13.T. & M.P. Armstrong, 1994. Massachusetts coastal commercial lohster trap samplingprogram May-November, 1993. Mass Div. Mar. Fish., 30 pp.

Figicl, CR. & G.L. Miller. 1995. The frequency of chelae autotomy and its influence on the growth andsurvival of crayfish Procamollfl{s cll1rkii (Girard, 1852) (Decapoda. Cambaridae). Crustoceana. Vol.61', pp. 472-3S3.

F Juanes, L.D. Smith / J. Exp. Mar. Bioi. f~col. 193 (1995) 197-223 219

Fingerman, S,w. & M. Finger-man, 1976. Effects of time of year and limb removal on rates of ecdysis of

eyed and cyestalkless ftddler crabs, Uca pugilator. Mar. Bioi., Vol. 37, pp. 357-362.

Fonnanowicz, D.K, IlJ90. The antiprcdator efficacy of spider leg autotomy. Anim, Schav., VoL 40, pp,400-409

Fredericq, L., 1882, Amputation des pattcs par mouvcment rerlexe chez les crabs. Archiv. BioI., VoL 3,pp. 235-240.

Garvey, J.E & R.A. Stein, 1993. Evaluating how chela size influences the invasion potential of anintroduced crayfish (Oreor!cetes mslieus). Am. MidI. Nat., Vol. 129, pro 172-181.

Garvey, J.E., R.A. Stein & H.M. Thomas, ] l)lJ4. Assessing how fish predation and interspecific preycompetition influence a crayfish assemhlage. Ecology, VoL 75, P[J. 532-547,

Goss, R.J' j 1969. Principles of regene;ation. Academic Press, London, 2fS7 pp.Govind, C.K. & J .A. B1undon, 1985. Form and function of the asymmetric chelae in hlue crabs with

normal and reversed handedness, Biol. BIA!!. Woods Hole Mass" Vo,. 168, pp. 321-331.

Hamilton, P.V., R:C. Nishimoto &. J.G. Halusky, 1976, Cheliped latcrality in Callinectes sapidus(Crustacea: ?o[(uniJae), 1];oi. f,'ull, Woods Hole Mass., V& 150, pp. 393-401.

Harris, R.N., 1989. Nonlethal injury to organisms as a mechanism of population regulation. Am. Nat.,VoL 134, pp. 835-847,

HartnoH, R.G., 1965. The biology of spider crabs: a comparison of British and Jamaican species,Crustaeeana, Vol. 9, pp. 1-· I\),

Hiatt, R.W., 1948. The biology of the lined shore crab, Paehygrapsils cnssipcs RandalL Pac. Sci., Vol. 2,pp. 135-213.

Hines, A.H., 1982. Allometric constraints and variables of repwductive effort in brachyuran crabs.Mar. Bioi., Vol. 69, pp. 309-320.

Hines, A.H., A,M. Haddon & LA. Weichert, 1990, Guild stmcture and foraging impact of hlue crabsand epibenthic fish in J subcstuary <If Cbe.,apeake Bay. Ajar. C.COI. Prog. SET., Vol. 67, pp. 105-126.

Hopkins, P,M., 1982. Growth and l'egenaalioil patterns in tile fic}olcr clab, Uca pugilator. BioI. Bull.Woods Hole Mass., Vol. 16:3, pp, 301-319.

Hopkins, P.M., 1985. Regeneration and relative growth in the liddltr crab. In, Factors ill adult growth,Crustacean issues 3, edited by A.M. Wenner, Balkema, Rotterdam, The Netherlands, pp. 265-275.

Hopkins, P.M., 1988. Control of regeneration in Crustacea. In, Endocrinology of selected invertebratetypes, edited by H. Laufer de R.G,H. Downer, A,R, Liss, Inc., New Ykork, pp. 327-340.

Hopkins, P,M., 1993. Regency. alion of walkmg legs In the fiddkr crab Uca pllgiiator. Am, Zool., VoL 33,pp. 348-356

Huber, M.E.. 1985. Nonrandom n,ating with respect to mate s<zc in tLe crab Trapezia (Brachyura,Xanthidae). /'vlar. Behav. Physiol., Vol. 12, pp, 19-32.

Hughes, ,LT, &. G.e. IVlatlLiessen, i962. Observations on the bioi:>gy of the Arnerican lobster,Homarus americwiUs. L:WllOI. Oawwgr., ",,\. 7, pp. 1:,14--·',.21.

Hyatt, G.W., 198,), QllaliU.li ',Ie and \ e dirnellsi,m~; uf (rLstac~an aggressioil, In, Studies inadaptaticn: the behavior oj nigher CnlS!t.JLea, edited by S. Rebach & DW. Dunham, J. Wiley & Sons,New Ycrk, pp. 113·· ;:,2.

Hyal~, G.\V. & M. Salmon, 19/8, COI"bat if. fidditr era::" uca r~giiator and U. p1igrux: a quantitativeanalysis. Behaviour, Vol. 65, pp. ] 82-211,

lachowski, R.L., ]~174. Agunistic b:c!la"ior Df the blu~ ([<jb, Callin~c{::s sapidus Rathbun. Behaviour,Vol. 50, pp. 232-2:'3

Jaksie, EM, & E.R. Fuentes, 1980. Correlales of tail losses in twelve species of Liolaemus lizards. J.Hcrpetol. Vol. 14. pp. 1]"7- ,'I.

luan:::~;, 1-'., 1992. Why c!,) decapod crustaceans prekl small-sized molluscan prey'! Mar. 1:'col. Prog. Ser.,Vol. 87, ;:p. 239-2t9.

luane, .. r .& ".n J !:!rlwie,., 990. :,II.C ,,!c:ion ,n Dur:zer.css crabs: the effect of claw damage.Ecology, Vol. 7;, Pi' 7,,:' 751-)

Kennelly,:i,J D. 'Nat:,;ns " ,; R. ! l)':'O. Mo,·t,. . cli,c:l[l]cd Sjian:.o.:r eLihs Rallina ranina(Li::1,n<-letr):;: a ta1.gk ~nct Eshc;ry-laborato:..y and ileld experirncnts. J. l:.xp ..Mat>. £Jiol. Ecol.~ Vol.1'10, pp .'1'"'-41\.

220 F. J:ar.",. L.D.YMir/i / I FxfJ. /\1ar. Bioi. Eml. 193 (/095) N72..?3

Kl()u:;c. 1.S, 1'nt,. Inc:dence oj Cll;' lor ,',JeTS, l!oliZllrw (iniNicarlLlI, in c{Jiamercial and research catches

uH tk :C,1ainc ,:o~st. nsh. nuli, Vol. 74, fir 7)9-77";.KLti"la, H., 191J:,. Limb 10"s and 1ecu' cry iE lhe YOUi.g kJllg -':fdb, {'ld'alitllOdl?s camtschaticu. Bul!,

!fukkaido Reg. }-',sh. Res. L,;b .. V,J!. 26, pp. 75··g0.

Kuris, A.:\1. & :\1. N,agsL 1975. Erred o!' limh regener::tion on size inuL'asc at molt for the shore crabs

Hemigrapsw oregonetlsis cud Oli.ISlpl., . .I. Frp. Zoo!.. Vol. l'n, pp. 3'11--360.Lawrence, J.lVi., ;992. Arm JOss unO. regeneratwn in Asteroidea (Echinodermata). In, Echinoderm

I<'.l('arc·h 1991, cditeo L scakt:l·L;i,CI &: C Canica\li, Balkema, !i..olteruam, 'fhe Netherlands, pp.39--52.

La" rence, I.M. 8:: A. Larram. 199·J.. 'rlle CO'it o' autotlHl1Y iI' the starfish SliclUister strillius. Mar. Ecol.hog. Sa., VI;L 109, PI' .'111 ~:;l3

Lav,tol1, P., ] 9t\9. Predatory interaction between the orachyuran crab Cancer pagurils and decapodcrustaceCJ1 prey. Mar. Eeu!. Prog. 5'(-,... Vol. 52, pp. 169--179.

Lee, S'y' & R. Seed, 1992. Ecological implications of cheliped size in cabs: some data from Cardl/us

/lli/OWS and Liceareinus h,/isa;co. Mar. Fcc!. Prog..)'cr" \lul. 84, pp. 15/-160.

Lewis, J,E., 1969. R'.:versal of ,,;ymmelry of chelae in Calappa Weber 1975 (Decapoda: Osystomata).Pn;c. Bioi. Soc Wllsh., Vol. 82, pp. 63···g0.

Lima, S.L. & L.M. Dill, Ilj9G. Beha Vlim\! decision made under ih" risk of predation: a review andprospectus. ('an. 1. Zool., Vol. c;:;. pp. 619-640.

Lindberg. W.J. & MJ. Marshall, 1%4. Species profriles: lilt hiStories and environmental requirements of

coastal fishes and imertehn.tes (:;outh Florida) stone crab. U.S r'ish Wild!. Servo FWS/OBS-82/

J1.21. U.S. Army Corps of E'ngdleer::, IH. EL-82--4, 17 Pl'.

MacGinitie, G.t: & N. MacGinitie, 1949. Nalural hiSlorv of marim animals. McGraw· Hill, New York,473 pp.

Maiorana, Vc., j CJ77. Tail auhr,omy, functional conl1icts and their reSOlution by a salamander. Nawre,Vol. 26'1, pp. 533--535

Mangel, M. & c.\-V. Clark, 19';,:j. D)lwtriie muddling in hehaviOral ('cology. Princeton University Press,

Princeton, New Jersey, .lOB 11p.

Mauzey, K.P., C. Birkeland & P.K. Dayton, ! 968. Feeding behavior of asteroids and escape responses

of their prey in the Pugct Sound region. Ecology, Vol. 49, pp. 60:;·619,McCallum, H.I., R. Endean & /\.1v1. Cameron.. 1989. Sublethal damage to Amtll}wster planei as an

index of predation pressure. iv/al'. Ecol. Frog. Sa., Vol. :;6, pp. 2q 36

McVean, A., 1975. Autotomy. COlf/p. Biochem. Physiol.. Vol. 51A, pp. 497--505.McVean, A., 1976. The incidence of aUiotomy in CarcLii/lS /IIi/elias (1..). 1. Erp. /',1ar. BiD!. Ecol., Vol.

24, pp. 177-187.McVean, A., 1982. Autotomy. In, The Biology of Crustacea, Vol . .J, edited by D.E. Bliss, Academic

Press, New York, pp. j07-132.McVean, A. & I. FincllilY. I '-J7lJ. Thl' incidence or autotomy in an estuarine population of till' crab

Careinus nlllenas. J. Afar. Biot. Assos. U.K., vol. 59. pp. .1,41-354.Miller, R.J. & .!. Watson, 1976. Growth per moll and limb regeneration in the spider crab. Chio!loecetes

opilio, J. Fish. Re;. Board CUll., Vol. :B, pp. 1644-1649.Mohrherr, C.J .. i 98-/. Cheliped autotomy and limb regeneration in the east African fiddler crabs Uea

laetca (De Haan) and U. chlowplu!ial/llils (H. Milne Edwards). 1. Cmst. Uiol .. Vol. 7, Pl'. 102307.Mo~in, P..!., 19R5. Predation intensity, prey survival and injury frequency in an amphibian predator­

prey intcracticn. Copeia, Vol. 1')85, pp.Needham, A .E., 195.l. The incidence and adaptive value of autotomy and regeneration in Crustacea.

Fmc. Zool. Soc. i.ond.. Vol. 12l, pp. I II 122.

Needham. A.I: .. 196'1. Regeneration in thl' Arthropoda and it:.; endocrine control. In, RegeneratIOn i/1(lnimals lInd related prohlel/ls, editccJ hl' V Kior(sis & 11./\.1.. Trampusch, Plsevier, Amsterdam, pr.2iD323.

Neil. S.J .. JCJi:l5. Size asscssment anJ cues: studieS of hermit crab contests. Bcha\'iollr. Vol. 92, pp,22--31'.

Ng, P.KL. & IW.H. Tan. 198'1. "Right handedness" in hcterocheloLJs calappoid and xantlloidcrabs-suggestion for a functional advantage. Crustace(lna, VoL 49. pp. 9t\ 1DO.

F. Juanes. L.D Smith I J. Exp. Mar. Bioi. Ecol. 193 (/995) 197-223 221

Niwa, K. & H. Kurata, 1964. Limb loss and regeneration in the adult king crab Paralithodescamtschatica. Bull. Hokkaido Ref{. Fish. Res. Lan., Vol. 28, pp. 51-55.

Norman, c.P. & M.B. Jones, 1991. Limb loss and its effect on handedness and growth in the velvetswimming crah Necora puner (Brachyura: Portunidae). 1. Nat. Hist .. Vol. 25, pp. 639-645.

Norman, c.P. & M.B. lones, 1993. Reproductive ecology of the velvet swimming crab. Necora puber(Brachyura: Portunidae). at Plymouth . .I. Mar. Bioi. Assoc. U.K.. Vol. 73, pp. 379-389.

O'Neill, 0.1. & 1.S. Cobb, 1979. Some factors influencing the outcome of shelter competition inlobsters (llomarus americanus). Mar. Behav. Physiol., Vol. 6, pp. 33-45.

Onizuka, E.W.. 1972. Management and development investigations of the kona cran. Ranina ranina(Unnaeus). Div. Fish Game. Dept. Land Nat. Res. Rep., Honolulu, Hawaii, 28 pp.

Peterson, C.H. & R. Black. 1994. An experimentalisl"s challenge: when artifacts of interventioninteract with treatments. Mar. Em/. Prof{. Ser., Vol. Ill, pp. 289--297.

Petranka, J .W., 1989. Densi! y'-dependent growth and survival of larval Arnnystoma: evidence from

whole-pond manipulations Ecolof{y, Vol. 7(\ pp. 1752·-1767Przibram, H., 1931. Connecting I::/ws in IJnimal morphology. University of London Press, London, 62

pp.Reaka, M.L, 1976 l.unar and tidal period~city of molting and reproduction in stomatopod Crustacea:

a selfish here! hypothesis. Bio!. Bull. Woods Hole Mass., Vol. 1SO, pp. 468-490.Reichman, 0..1, 1984. Evolution of regeneration capabilities. Am Nat .. Vol. 123, pp. 757.--763.Reid, D.G., P. Abello, c.G. Warman & E. Nnylor, 1994. Sizi:-related mating success in the shore crab

Carcillus Inaenas (CrustaceC': Brachyura). J. Zoo!., Vol. 232, pp. 397-..407.Reilly, P.N., 1983. Effects of commercial trawling on Dungeness crab survival. Calif. Dept. Fish Game,

Fish. Bull.. Vol. 172, pp. 165-169.Robinson, M.H., A.G Abe:e & B. Robinson, 1970. A.ttack autotomy: a defence against predators.

Science, Vol. 169, pp. 300- 30'.Salmon, M., 1983. Courtship. mating systems, and sexual selection in Decapods. In, Studies in

adaptation: the behavior of higher Crustacea. edited by S. RebN:h & D.W. Dunham. J. Wiley & Sons,New York, pp, 143--170.

Savage, T. & J .R. Sullivan, 197B. Growth ,"'1d regcner"'t;ol'l of '.be stonc crab, Menippe mercenaria. Fla.Mar. Res. Pl..bl., No. 32, 2? pp.

Savage, T., J.R. Su;!ivan & C.F. Kalman, ]975. An analysis of ~;tone crab landings on Florida's westccast, with a brief synopsi,; of the fishery. Fla. Mar. Res. Pub!., Nc. 13, 37 pp.

Scarrat, DJ., 1973. Claw Joss and other wounds in commercie.!Jy caught lobsters (Homarus ameri­canus). J. Fish. Res. Board Can., Vo!. 30, pp. 1370-1373.

Schoener, T.W, 1979. Inferring thE'; properties of predation and O1.herinjury-producing agents frominjury frequencies. Ec%gy, Vo!' 60, pp. 11. lO-1 115.

Schone. H., 1968. Agcnist:c and sexual displays in aquatic and semi-terrestrial brachyuran crabs. Am.Zool., Vol. 8, pp. 6L;1-6S4

Sekkelsten, G.!., 1988. Effc;t of handicap on mating succes'i in mille shore crabs Carriflils maenas.Oikos, Vol. 51, pp. 131--134.

Shelton, P.M.J.. P.R. Truhy & KG J. Shelton, ]981. Nat IHally occurring abnornlil!ilie,. (Bruchdrcifach­bildungen) in the chelae Gr three species of Crustacea (Deeapoda) and a possible explanation. J.Embryol. Exp. Mr>rplJ.. Vol. 63, pp. 285-3('r,4.

Shirley, S.M. & TC Shirky, 1988. Appendage injmy in Dlln~;el1ess crabs. Cancer magister, insoutheastern Alaska.. hsh. BI/II" Vol. 86, pp. 156·- I60

Shirley, M.A, A.H. Hines & T.G. Wolco\!., 1990. Adar'live signifIcance of habitat. selection hy moltingadult hlue crabs Cal.'inecte'; sapidus (Rathbun) within a subcstuary of central Chesapeake Bay. 1.Exp. Mar. Bioi. Ecol., Vol 1,10, pp. 107-119.

Sih. A" 1981. Foragir'g str,l\e,ies and the avoidance of pr'odation by an aquatic insect Notonectaholfmarmi. Eco[oiJ." Vol 63. pp. 71\6-796.

Sih, A .. 1992. Prey L'ncerta;nt:; ard the balancing of antiT'Tccbtor and feeding needs. Am. Nat., Vol. 139,PfJ 1052-1069.

Simonson, .J [ .. , j 985. Revenal Of handedness. growth, and claw st.ridu!atcry patterns in the: stone crabM('f,~ippe mercenar;" (Say) (Crustacea: Xanthidae). f. Crus!. Bin!, Vol. :i. po. 281-293,

222 F jlilmn. L.D. lim/ih i J Fxp.Har IJiol Fco!. 191 (1995) 197-223

Simenson, J.L & RJ. Hoch b<:rg, 199" Effects of air exposlire ,md (:Iaw breaks on s"rviv'll of stone

cnbs Mempf!c merCf'/Wr/ii. FwlS AI!! Fhh. Soc., VOl. 115. f'P. 47 1··,rn.Sirw.1Dson, J.L & P. Steele. 1981. Cheliped asymmetry in the stone cpb. Menippe f/u'rcenarill, with

notes on claw n;versa1 and ref,crC'iltion. No,.,h. Cuii' Sci, VoL S. fJP ')1-.:\(l.

Sinclair, M.E.. 1977. Agonistic behaviour of the stone crab Menippe mercel/aria (Say). Anim. Bellav.,VeL 25. PI'. 1.93-207.

Skinner. D.M., 19115. MO:ling :lnd l!~gereration In, Biologv of Crusll!Cea, Vol. 9. edited by D.E. Bliss &L.H. Mantel, Ac·ldcmi.c Press, Nc\\! York. PI'. 43·14h.

SkiIlner, DM. & D.E Grahan, 1970. \1011ing in land crahs: stimulation hy leg removal. Science, VoL169, Pl'. 383-3S5

Skinner, JJ.!Y1. 8~ [).E CirahJiTl, I (.~72. LfY-;S !in-,bs as d qirrlldus to ecdysis in Brachyura (true crabs).BioI. fit/Ii. \Vooo's flole iH.·lss., Vol 143, Pl'. 222-233.

Smith, r P., i 990. The agoni.,ric hchaviour of Ihe velvel swinming crlib, L!::"arciIlIlS Imber (L)(Brachvlll'i1, Por:ul1idIlC). 1'1".1) dissertation, University Of Glasgow, Glasgow 252 PP

Smith. LI\ 199ih. The frcollcnc'.' (i"d ecological consei/uences of lim/J UllOloriH' iii the hlue crab,

Cdlinccles .Iapillus Rathbull. Ph.D. dissertation, University of Maryl?nd, College Park, Maryland,259 pp.

Smith, L.D, I990h. Patterns of limb )O's in the blue crab, Callinecles sapidus Rathbun, and the effects

of autotomy on growth. Buff Mar. Sci., Vol 46, pp. 2.1·36.

Smith, L.D, 1992. The impact of limb ;"utotomy on male competition in blue crabs Callinecles sapidusRathbun. Oecologia, Vol. R9, pp. 494501.

Smith, L.D" 1995. Effects of limb autotomy and tethcring on juvenile blue erab survival fromcannibali·;m. M,lT. Eco!. Prog. Ser., Vol 116, pr. 65-74.

Smith, LD. & AI-!. Hines, 199] a, Th~ effect of cheliped loss on blue crab Ca/lil/nles .Iapidus Rathbun

foragifig nte en sofHhellcd clams Mya arenarill L 1. Exp. Mllr. Bio!. Ecol .. Vol. ]51, pr. 245 .. 256.

Smith, LD. & A. H. Hir:es, 1991 b. i\ uturomy in blue crab (CaIlineC/fs .\IIpidus Rathbun) populations:

geographic, lempora', and ('neogenetic Yari;:tion. BioI. Bull. Wool's flo/e~v!!I\s Vol. 1180, pro41 h-43 I.

Smith, LD. & t-\.R. Palmer.. 1':JSl4. EJTects of rranipulated dict on siZe: ;lccI performance of brachyuran

crab claws, Science, VoL 2f>'f, pp, 710 712.

Smyth, M., 1974. Changes in fat stores of the skinks Morelhia IJou!angeri aacl flclIllcrgis peronii

(Lacertilia), AilS!. J. Zooi" VoL?2, pp, 135·-145.

Snedden, W.A., 1990. Determinan is of m ale mating SUCCCSli in the ternpemtc crayfish OrconeclesrLlsticus: chela size and sperm competition. Behaviour, Vol. 115, pp, lOO ! 13.

Spirito, c.P.. 1972. An anaIysi; of swimming behavior in the pori unid crab Carlinecles sapidus. Mar.Behav. Physiol., Vol. I, pp. 2(11-276.

Spivak, E.D.. .990. Limn regeneration in a comillon Sout h·American littoral crah CVr!ograpsus

IF/gillallis. 1. Nili. lh\!., VOL 24, Pl'. 393-402.Spivak, ED. & M.A. Politis, 19K9. High incidence OJ limb autntom) in a crab population from a

coastal I"goon I!l thc provine..; of kuenos ;\i:'cs, Aq:.cntina. Cali. J. /,xil., VoL 07, f)P. 1':J76·-1985.

Stein, R,A., ]lI76, Sexual dimorphism in crayfish chelae: functional significance linked to reproductive

activities, Can. J. Zool, Vol. 54, pp 220 227.

Stein, R.A., 1977. Selective predation, optimal foraging, and the prcdator,prey interaoion hetween fish

and crayfish. Ecologv. Vol. 'i8, Pl'. 12'1,7-1253.

Stl,;vens, B,G" D.A, Armstrong & R. Cusimano.. IlJ~2. Feeding habits 01 the DUl1gencss crab Calleermagisler as determined by the index of relative importance. Mar. BioI., Vol. 72. pp. l'l,5-145.

Stevens. B.()., W.E. Donaldson, .lA. Jlaaga & J.E. Munk, 1993. Morphometry and maturity of paired

tanner crabs, Chiolloec('/es hairdi. from shallow- and deepwater environments. Call. J. Fish, Aquat,Sci .. Vol. 50 pp I.'i04-151 (J.

Stoffel. LA. & J.lI. IIubsehman. J974. Limh loss and the molt c"cle in Ihe freshwater shrimp,I'alacmunelcs klldillketl.li\. Bioi. Bill/. Woods !lole ,",1al.\. Vol. 147 pp. 20'l, 212.

Sullivan, J.R, 1979. The stone crah, Mcnippl' mercellaria, in the SOUthwest Florida tisher\!. Fla. Mar.Res. Pubt., No. 3f], .37 Pl'.

F. Juanes, LD. Smith / J. F:xp. ,'Har. Bioi. Eeu!. 193 (1995) 197-223 223

Van Buskirk, J. & D.C. Smith, 1991. Density-dependent population regulation in a salamander.Ecology, Vol. 72, pp. 1747-1756.

van del' Meeren, G.1., I. 994. Sex- and size-dependent mating tactics in a natural population of shorecrabs Carcinu.I' maena.l'. J. Amm. Ecol., Vol. 63, pp. 307-314.

Vermeil, G.J., 1977 Patterns in crab claw size: the geography of crushing. SYSI. Lool., Vol. 26, pp.138-151.

Vermeij. G.J., 1982. Unsuccessful predation and evolution. Am. Nal., Vol. 120. pp. 701-720.Waldron, K.D., 195R. The fishery and biology of the Dungeness crab (Cancer magister Dana) In

Oregon waters. Fish Comm. Oreg. COn/rib., Vol. 24. pp. 1-43.Weis. J .S .. 1976. Effects of environmental factors on regeneration and molting in fiddler crabs. Bio/.

Bull. Woooc!s Hole Mass .. Vol. 150, PI'. 152·-162.Weis. 1.S., 1977. Limb regeneration in fiddler crabs: species diiTercnccs and effects of methylmercury.

Bioi. Bull Woods !io!e Mas,., Vol. 152, pp. 263-274.Werner. E.E. & .I.F. Gilliam, 1984. The ontogenetic niche and species interactions in size-structured

populat'ons. Ann. Rev. E'col. Sysl .. Vol. 15. pp. 393-·425.Wilson, B.S., 1991. Latitudinal variation in activity sea:-.on mortality rates of the lizard Ula slans­

buriana. Ecol. MOl1ogr .. Vol. 61. pp. 393-414.Wood, rD. & H.E. Wood. 1932. Autotomy in Decapod Crustacea. J. Exp. Zoo/ .. Vol. 62. pp. 1-55.Zimmer-Faust, R.K., 1989. The relationship between chemoreception and foraging behavior in

crustaceans. Limnol. Oceanogr .. Vol. 34, pro 1367···1374.