Journal of Experimental Marine Biology and Ecology, 227

LJournal of Experimental Marine Biology and Ecology,227 (1998) 187–201

Extended parental care in marine amphipods. I. Juvenilesurvival without parents

*Martin ThielDarling Marine Center, University of Maine, Walpole, ME 04573, USA

Received 23 June 1997; received in revised form 3 November 1997; accepted 1 December 1997

Abstract

Extended parental care is found among a variety of marine peracarid species includingendobenthic and epibenthic amphipods. After hatching from the female’s brood pouch, smalljuveniles enjoy parental care for time periods of several days up to several weeks. This studyexamined whether small juvenile amphipods can survive without parental care to determine if thisreproductive strategy is obligate or facultative. Juveniles of the burrow-dwelling speciesLeptocheirus pinguis (Stimpson 1853) and Casco bigelowi (Blake 1929) survived well inpredator-free trays, indicating that extended parental care in these species is a facultativereproductive strategy — juveniles can survive without their parents. Medium-sized juveniles of theepibenthic, suspension-feeding species Dyopedos monacanthus (Metzger 1875) survived relativelywell in predator-free traps, but small juvenile D. monacanthus did not. This low survival rate ofsmall D. monacanthus, even in predator-free trays, demonstrates that extended parental care isobligate for the survival of early juvenile stages in this epibenthic amphipod. When exposed topredators, juveniles of all three amphipod species were susceptible to predation, and reacted byemigration. Most juvenile L. pinguis and C. bigelowi emigrated or disappeared from sand shrimpCrangon septemspinosa treatments, but many remained in hermit crab Pagurus longicarpustreatments. Almost all juvenile D. monacanthus emigrated from both sand shrimp and hermit crabtreatments, indicating that this epibenthic amphipod species is very susceptible to epibenthicpredation. Survival of juvenile D. monacanthus was lower than that of L. pinguis and C. bigelowiboth in predator-free and in predator trays. These results suggest that juveniles of this and otherepibenthic peracarids depend on maternal care to a higher degree than juveniles of endobenthicburrow-dwelling species. 1998 Elsevier Science B.V. All rights reserved.

Keywords: Reproduction; Extended parental care; Amphipoda; Soft-bottom; Predation

*Present address: Smithsonian Marine Station, 5612 Old Dixie Highway, Fort Pierce, FL 34946, USA. E-mail:[email protected]

0022-0981/98/$ – see front matter 1998 Elsevier Science B.V. All rights reserved.PI I : S0022-0981( 97 )00268-2

188 M. Thiel / J. Exp. Mar. Biol. Ecol. 227 (1998) 187 –201

1. Introduction

Extended parental care is a common reproductive strategy among terrestrial inverte-brates (Wilson, 1975; Clutton-Brock, 1991), but at present only few marine invertebrateshave been found to engage in this reproductive strategy. Parents provide a suitableoffspring habitat (e.g. burying beetles Nicrophorus sp. — Trumbo, 1992; or marineisopods Sphaeroma terebrans — Messana et al., 1994), they gather food resources fortheir developing offspring (e.g. most hymenopterans — Wilson, 1975; Heinrich, 1979;some caprellid amphipods — Aoki and Kikuchi, 1991), and they guard and defend theiroffspring (e.g. the green lynx spider Peucetia viridans — Fink, 1986; the pine engraverbeetle Ips pini — Reid and Roitberg, 1994; or caprellid and podocerid amphipods —Stephan, 1980; Mattson and Cedhagen, 1989; Aoki and Kikuchi, 1991). In many cases,developing offspring cannot survive without parental care or at least its survival chancesseverely decrease when deprived of parental care (Clutton-Brock, 1991 and referencescited therein). Extended parental care is thus an obligate reproductive strategy in thesespecies: juveniles cannot survive without parental assistance. If parental care improvesthe chance of survival for the offspring, but is not essential, then it should be considereda facultative reproductive strategy. Examples for the latter are species where parentalforaging provides additional food resources for growing offspring thus enhancing theirsurvival chances (e.g. Brown and Formanowicz, 1995; Lubin and Henschel, 1996).

Among marine invertebrates, extended parental care is an uncommon reproductivestrategy. In many species, developing eggs are provided with some food resources, butthen grow and develop without further parental care. Diverse forms of parental care arefound in a few species of almost any marine invertebrate phylum, but it is only inperacarid crustaceans (amphipods, isopods, cumaceans, etc.) that in all species embryosand larvae develop in the female brood pouch from which late larval stages or smalljuveniles emerge. In most peracarid species, juveniles leave their mother immediatelyafter hatching from the brood pouch, and no further parental care for the young offspringis provided. There are, however, several endo- and epibenthic peracarid species thatengage in extended parental care. In these species, juveniles remain with their motherafter hatching from the brood pouch. In endobenthic species, juveniles live in themother’s burrow for time periods of several days (Thamdrup, 1935; Goodhart, 1939;

¨Watkin, 1947; Buckle-Ramirez, 1965; Johnson and Attramadal, 1982; Hassack andHoldich, 1987), or several weeks (Shillaker and Moore, 1987a,b; Thiel et al., 1997). Inepibenthic species, juveniles either cling to the mother (Aoki and Kikuchi, 1991;Svavarsson and Davidsdottir, 1995; Thiel, 1997a) or to structures built by the mother(Mattson and Cedhagen, 1989). During the period of extended parental care, juvenilescan grow to more than half adult size (Conlan and Chess, 1992; Thiel, 1997b; Thiel etal., 1997).

Leptocheirus pinguis (Stimpson 1853) and Casco bigelowi (Blake 1929) are commonendobenthic amphipods that inhabit burrows in muddy sediments along the Atlanticcoast of North America (Bousfield, 1973; Dickinson et al., 1980; Dickinson and Wigley,1981; Wildish, 1980; Michael, 1987). L. pinguis is a filter-feeder living in U-shapedburrows, and C. bigelowi is a deposit-feeding species dwelling in irregular burrows. L.pinguis is an important prey for a variety of fish (Wigley and Theroux, 1965; Tyler,

M. Thiel / J. Exp. Mar. Biol. Ecol. 227 (1998) 187 –201 189

1971; Hacunda, 1981; Mahon and Neilson, 1987). In both species, burrows of adultsextend 5–10 cm below the sediment surface (Thiel et al., 1997). Dyopedos monacanthus(Metzger 1875) is an epibenthic amphipod that constructs mud whips (Mattson andCedhagen, 1989) which it utilizes as vantage points for suspension-feeding (sensu Mooreand Earll, 1985). Female D. monacanthus host their offspring on these mud whips forseveral weeks (Thiel, 1997c). Despite substantial differences in their general biology, allthree amphipod species, L. pinguis, C. bigelowi, and D. monacanthus, engage in aprogressive form of extended parental care.

In an earlier study (Thiel et al., 1997), we hypothesized that in endobenthic speciesextended parental care provides shelter from predation, whereas in epibenthic species itis primarily a mechanism to lift small juveniles into water layers with sufficient supplyof suspended food. All three amphipod species reach high abundances in soft sedimentsjust at and below mean low water (MLW) (Thiel, 1997b), where predator abundancesare high. Juveniles of these amphipods might be particularly susceptible to predation,and extended parental care could be a successful strategy to protect small juveniles fromepibenthic predators such as fish, shrimp or conspecifics.

Not all juveniles attain the same age and size during extended parental care. Someindividuals of a clutch leave their mothers a few days after birth whereas others remainfor weeks or even months (Thiel, 1997c). The question arises whether juveniles thatleave their mothers at an early age will have the same survival chances as those that staylonger — i.e. is extended parental care an obligate or a facultative reproductive strategyin these amphipods?

This study examined the survival of juvenile amphipods Leptocheirus pinguis, Cascobigelowi and Dyopedos monacanthus that were separated from their mothers. The majorgoal was to examine the survival chances of these orphan juveniles both under protectedconditions (i.e. free from predation), and under predation pressure.

2. Materials and methods

Experiments were conducted in the flowing seawater facility of the Darling MarineCenter, located on the Damariscotta River Estuary, Maine (lat. 438509 N, long. 698379

W). Seawater is continuously pumped from a depth of about 2 m below MLW, andsalinity varies between 29 and 31 ppt, which is similar to conditions that amphipods

2encounter in the field. Small trays (900 cm ) were established in the laboratory and filledwith mud from an adjacent mud flat (a 10-cm thick mud layer for L. pinguis and C.bigelowi and a 3-cm thick layer for D. monacanthus). A constant flow of seawaterthrough trays was maintained for the duration of the experiment. Seawater leaving traysthrough outflow pipes passed over a screen trapping emigrating juveniles. All emigrantswere thus retained in these traps where they could be collected and counted.

Juvenile amphipods for the experiments were collected from sediments at and belowMLW in the Damariscotta River estuary, and after separation from their mothersimmediately introduced to the trays. For L. pinguis and C. bigelowi, burrows of parentalfemales were excavated and all juveniles collected. For D. monacanthus, juveniles weretaken from mud whips of parental females. A second experiment was conducted for D.


monacanthus in which juveniles were taken from a culture tank. Some of these juvenileswere still living on the maternal mud whips while others had already built their ownwhips. Juveniles were sorted and equally divided among the different trays where theywere allowed to establish burrows or mud whips. During an establishment period of 1–2weeks, all juveniles that left trays and were caught in the traps were reintroduced to theirtrays every morning. During a pre-experimental period, 5–10 days up to the com-mencement of the actual experiment, all juveniles that were caught in traps werecounted, but then reintroduced to their trays. Following the introduction of predators totrays, all emigrating juveniles were collected from the traps and preserved. At the end ofthe experiment (7–11 days after introduction of predators), the entire content of eachtray was sieved over a 500-mm-mesh and preserved. All amphipods were sorted fromthese samples, counted and measured along their dorsal surface from the rostrum to thetelson using a computer-based image analysis system.

Two epibenthic predators that are common in the habitat of the amphipods wereselected for experiments, the sand shrimp Crangon septemspinosa, and the hermit crabPagurus longicarpus. Ten sand shrimp were introduced to each of three shrimp trays,three hermit crabs were introduced to each of three hermit crab trays, and three trayswere left predator-free as controls. During predation experiments, all trays (includingcontrol trays) were covered with a 1-mm-screen to prevent shrimp or hermit crabs fromleaving trays. The percentage of juveniles surviving as residents and emigrants wascalculated from the number of juveniles initially introduced to each tray. To test whetherpercentages of survivors differed significantly between treatments (sand shrimp, hermitcrabs and controls), percentage values were arcsin-transformed and tested by ANOVA(n 5 3 replicate trays for each treatment, df 5 2) followed by Fisher’s PLSD post-hoctest.

3. Results

3.1. Emigration of juveniles

Juveniles of all three amphipod species responded to addition of predators byemigrating from trays (Fig. 1). Numbers of Leptocheirus pinguis emigrating from trayswere relatively low in the 6 days preceding the introduction of predators (Fig. 1A), andnumbers of juveniles emigrating from control trays remained low for the entire durationof the experiment (Fig. 1A). Immediately following the introduction of sand shrimp to

21the three shrimp trays, large numbers of juveniles emigrated (35.3610.9 S.E. tray atday 1 — Fig. 1A). In hermit crab trays, about four juvenile L. pinguis emigrated fromeach tray immediately after introduction of hermit crabs, but during the subsequent days,

21 21numbers of emigrants dropped again to less than one amphipod tray day . For Cascobigelowi, relatively large numbers of juveniles emigrated during days preceding theaddition of predators, but emigration decreased until day 7 in both experimental andcontrol trays (Fig. 1B). A slight increase of C. bigelowi emigrants was noted in predatoradditions at days following predator introduction (Fig. 1B), yet this emigration responsewas not as pronounced as for L. pinguis juveniles (compare Fig. 1A). Relatively large


21 21Fig. 1. Mean number (6S.E.) of emigrating juveniles tray day from control, hermit crab and sand shrimptrays for (A) Leptocheirus pinguis, (B) Casco bigelowi, (C) small Dyopedos monacanthus and (D) medium-sized D. monacanthus; time period when predators were present in trays indicated by light shading; trap fromeach tray was monitored every morning; during the pre-experimental period, emigrating juveniles werecounted but then reintroduced to their tray; during the experimental period, all emigrants were collected andpreserved; n 5 3 trays for each treatment.

numbers of small Dyopedos monacanthus that were separated from their mothersemigrated during the pre-experimental period (Fig. 1C). No strong increase in numbersof small D. monacanthus emigrants was observed after addition of predators, probablydue to the low number of small orphan juveniles remaining at that time. Medium-sizedjuvenile D. monacanthus that had already built their own mud whips immediatelyconstructed new whips after introduction to trays, and few individuals emigrated duringthe pre-experimental period (Fig. 1D). Few medium-sized D. monacanthus ( , 1


21 21juvenile tray day during most days) emigrated from control trays over the entireperiod of the experiment (Fig. 1D). Immediately following introduction of predators,almost all medium-sized D. monacanthus emigrated from predator additions (33.367.96

21 21 21 21S.E. emigrants tray day from shrimp trays; 43.0614.36 S.E. emigrants tray dayfrom hermit crab trays) (Fig. 1D).

3.2. Survival of juvenile amphipods

Many juvenile amphipods survived as residents in predator-free control trays, whereasin sand shrimp trays few or no juveniles survived as residents but rather emigratedduring experiments (Fig. 2). More than 90% of juvenile L. pinguis that were introducedto control and hermit crab trays survived the experiments (Fig. 2A). A relatively largepercentage (63.6%) of juvenile L. pinguis survived in sand shrimp trays, but less than10% of these survivors remained as residents, while the majority (50%) emigratedduring the experimental period (Fig. 2A). While the percentages of L. pinguis survivorsdid not differ significantly among different treatments (ANOVA, P . 0.05), significantly

Fig. 2. Mean percentage of juveniles surviving in control, hermit crab and sand shrimp trays; for (A)Leptocheirus pinguis, (B) Casco bigelowi, (C) small Dyopedos monacanthus and (D) medium-sized D.monacanthus; calculated as percentage of potential survivors; n 5 75 juvenile Leptocheirus pinguis, n 5 90Casco bigelowi, n 5 55 small Dyopedos monacanthus and n 5 94 medium-sized D. monacanthus were initiallyintroduced to each tray in the respective experiments; lines indicate significant differences for: top lines —total survivors ( 5 residents 1 emigrants); middle lines — emigrants and bottom lines — residents (ANOVAwith n 5 3 replicate trays for each treatment, df 5 2; followed by post-hoc Fisher’s PLSD-test; P , 0.05).


higher percentages emigrated from shrimp trays than from hermit crab and control trays(ANOVA P , 0.05). In total, 77% of juvenile C. bigelowi survived in control trays(about 65% as residents), while 58.15% survived in hermit crab additions (37.8% asresidents) (Fig. 2B). No juvenile C. bigelowi remained in shrimp trays after theexperiment, but 30.7% escaped predation by emigrating from trays (Fig. 2B). Thepercentage of survivors and of residents differed significantly among all treatments(ANOVA, P , 0.05), and a significantly higher percentage of juveniles emigrated fromshrimp trays than from control trays (ANOVA, P , 0.05). Of the small D. monacanthusthat were separated from their mothers, less than 12% survived ( 5 emigrants 1

residents) in all three types of trays (predator additions and controls). Only 6% of thesesmall D. monacanthus remained in control trays at the end of the experiment (Fig. 2C).In this same experiment, no significant differences in survivorship or emigration of smallD. monacanthus among different treatments were found (ANOVA, P . 0.05), while thefew small D. monacanthus remaining in control trays were significantly different fromhermit crab and sand shrimp trays where no residents remained (ANOVA, P , 0.05). Inthe experiment with medium-sized D. monacanthus, many juveniles survived in controltrays ( . 50% as residents) (Fig. 2D). No medium-sized D. monacanthus remained inshrimp trays at the end of the experiment but 37.2% survived as emigrants. Only onemedium-sized D. monacanthus remained in one hermit crab tray at the end of theexperiment, but 48.9% emigrated after introduction of hermit crabs (Fig. 2D). Thepercentages of total survivors were not significantly different in the three differenttreatments (ANOVA, P . 0.05), but significantly higher percentages of juvenilesemigrated from shrimp and hermit crab trays than from control trays (ANOVA,P , 0.05).

3.3. Sizes of juvenile amphipods surviving in trays

For all experiments, the majority of juveniles originally introduced to trays were ofsizes that are typically found in burrows or on mud whips of parental females (Fig. 3).

Most juvenile L. pinguis originally introduced to trays were between 3 and 7 mm insize, sizes that are commonly found in burrows of parental females (Thiel, 1997d). Incontrol and in hermit crab trays many of these juveniles survived and grew to sizes . 5mm (Fig. 3A). In sand shrimp additions most of the few remaining resident L. pinguiswere . 7 mm (Fig. 3A). The sizes of juvenile C. bigelowi introduced to trays are typicalfor those of juveniles found in the burrows of parental females (Thiel et al., 1997). Themajority of juvenile C. bigelowi surviving as residents in control and hermit crab trayswere , 12 mm, and thus of a size that usually still enjoys parental care. The juvenilesemigrating from control and hermit crab trays were mostly , 10 mm, but all sizes ofjuvenile C. bigelowi emigrated from shrimp trays (Fig. 3B). In the experiment withsmall D. monacanthus juveniles, all juveniles initially introduced to trays were , 3 mm(Fig. 3C). Very few of these juveniles survived in either control or predator additiontrays (Fig. 3C). In the experiment with medium-sized D. monacanthus, 23.4% ofjuveniles that were initially introduced to trays were . 2.4 mm, and thus had alreadyreached sizes at which they normally have left their mother’s mud whip (Thiel, 1997c)(Fig. 3D). All medium-sized juveniles that remained as residents in control trays were


Fig. 3. Size frequency distribution of juvenile (A) Leptocheirus pinguis, (B) Casco bigelowi, (C) small juvenileDyopedos monacanthus and (D) medium-sized juvenile D. monacanthus surviving as residents (upper darkcolumns) and emigrants (lower light columns) in the different treatments; shown are resident and emigrantjuveniles surviving in sand shrimp trays, hermit crab trays, control trays and juveniles initially introduced tothe trays; size range of juveniles that are commonly found to leave their mothers are highlighted by lightshading; all juveniles from the three replicate trays for each treatment were pooled; n 5 number of totalsurvivors (residents 1 emigrants) from three replicate trays represents 100% for each treatment; n 5 number ofi

initially introduced juveniles per tray represents 100%.


. 1.4 mm. All size ranges of medium-sized D. monacanthus juveniles emigrated fromboth sand shrimp and hermit crab additions (Fig. 3D). Thus, in this experiment withmedium-sized D. monacanthus, sizes that are typically found on their own mud whips inthe field (Thiel, 1997c) emigrated after addition of predators.

4. Discussion

Juvenile amphipods Leptocheirus pinguis and Casco bigelowi that had been separatedfrom their mothers survived in large numbers in predator-free trays where theyimmediately had established their own burrows. In contrast, almost no small juvenileDyopedos monacanthus separated from their mothers survived even when kept inpredator-free environments. This indicates that juvenile D. monacanthus depend onmaternal care to a higher degree than juvenile L. pinguis and C. bigelowi. Orphanjuveniles of all three species were very susceptible to sand shrimp predation, indicatingthat predation strongly affects juvenile survival in these amphipods with extendedparental care.

4.1. Survival of juvenile amphipods without parental care

Several endo- and epibenthic amphipod species provide extended parental care fortheir developing offspring (Aoki, 1997; Thiel et al., 1997), yet it is not well knownwhether small juveniles of these species can survive without parents.

In the majority of marine amphipods and other peracarids, juveniles leave theirparents immediately after hatching from the female’s brood pouch (Wildish, 1982; seeSainte-Marie (1991) for recent overview). Survival of freshly hatched juveniles isinfluenced by the availability of habitat and resources (e.g. Shillaker and Moore, 1987a),by competition and predation (e.g. Nelson, 1979; Grant, 1981; Wilson, 1989, 1991), andby abiotic factors such as temperature, salinity and disturbance (Wildish, 1982; Wilsonand Parker, 1996). Staying with parents might increase the survival chances of juvenileamphipods because parents can provide resources (see e.g. Richter, 1978a,b; Laval,1980), protect their offspring from competition and predator encounters (Aoki andKikuchi, 1991), and mediate the impact of abiotic factors. Observations on parent-offspring interactions by Richter (1978a), (1978b), Lim and Alexander (1986) and Aokiand Kikuchi (1991) suggest improved offspring survival resulting from extendedparental care, but it was only recently that Aoki (1997) demonstrated experimentally thatjuveniles may benefit from parental care.

The present study demonstrates that juveniles of the burrow-dwelling species L.pinguis and C. bigelowi survive relatively well without parental care when kept inpredator-free environments, while small juveniles of the epibenthic species D. monacan-thus were less competent under such conditions. In most burrow-dwelling amphipods,small juveniles are thought to establish their own burrows immediately after hatchingfrom their mothers’ brood pouches (e.g. Wildish, 1984; Shillaker and Moore, 1987a,b;Dauvin, 1989; Wilson, 1989; Sudo and Azeta, 1996), yet in some species, juvenilesremain in the maternal burrows for several days (Thamdrup, 1935; Goodhart, 1939;


Watkin, 1947; Shillaker and Moore, 1987b). None of the authors that found juvenileamphipods remaining in maternal burrows suggests that this reproductive behavior isobligate. Reports of juveniles that feed on their mothers food (Shillaker and Moore,1987b) imply some positive effects on the growth potential of juveniles that remain inmaternal burrows versus those that leave at an early age and size, however thesepotential benefits require further investigation. In these (and possibly other) burrow-living amphipods extended parental care can be considered a facultative reproductivestrategy: small juveniles are able to survive on their own.

Small juvenile Dyopedos monacanthus that were separated from their mothers did notsurvive whereas medium-sized juveniles, many of which had already built their ownmud whips, did. Low survival rates of small juvenile D. porrecta and D. monacanthusthat were separated from their mothers were also observed by Stephan (1980). All thesepodocerid amphipod species construct mud whips that are used as ‘vantage points’ forsuspension-feeding (sensu Moore and Earll, 1985) and vigorously defended againstconspecifics (Stephan, 1980; Thiel, 1997c). Suspension-feeding amphipods extend theirsetose antennae into the water current, and at frequent intervals ingest suspendedparticles gathered by the antennae. Small juvenile D. monacanthus cling to the upperparts of their mothers mud whips, i.e. high in the benthic boundary layer where thefeeding environment may be optimal for them (Thiel, 1997c). The low survival rate ofsmall juveniles in the present study indicate that in these podocerid amphipods extendedparental care is an obligate reproductive strategy: the early and small juvenile stagesdepend on maternal care to survive. Small podocerid juveniles may not be able toconstruct and successfully defend large enough mud whips that are suitable for efficientsuspension-feeding.

A similar form of extended parental care, where small juveniles are lifted above thesubstratum by their mothers, can be seen in several caprellid amphipods (Harrison, 1940;Lim and Alexander, 1986; Aoki and Kikuchi, 1991; Aoki, 1997; Thiel, 1997a).Interestingly, small orphan juveniles of caprellid species with extended parental care hadrelatively high survival rates in predator-free environments (Aoki, 1997). Containersused in Aoki’s (Aoki, 1997) study provided large areas of suitable clinging substratumfor small caprellid juveniles, while the podocerid juveniles in this study were forced toconstruct their own mud whips. Caprellid amphipods usually inhabit preexistingsubstrata such as macroalgae and hydroids, whereas most podocerid amphipodsconstruct their own substrate structures in the form of mud whips. Therefore, juvenilepodocerids may depend on their mothers (as constructors of a mud whip) to a largerdegree than juvenile caprellids. These observations suggest that the availability ofsuitable clinging substratum influences the survival chances of small juveniles inepibenthic peracarids with extended parental care.

4.2. Effects of predators on the survival of juvenile amphipods

The addition of predators to the trays caused decreased survival of juveniles in allthree amphipod species. These results confirm earlier studies that demonstrated theimpact of epibenthic predators on soft-bottom dwelling macro- and meiofauna (Reise,1985; Mattila et al., 1990). Juvenile L. pinguis and C. bigelowi were less susceptible to


epibenthic predators than juvenile D. monacanthus, probably due to the endobenthiclife-style of the former two species.

In both L. pinguis and C. bigelowi a positive relationship exists between the size ofamphipods and the depth of their burrows (Thiel, 1997e,f). Large individuals can builddeeper burrows in which they are relatively safe from epibenthic predators such asshrimp. Small juvenile L. pinguis and C. bigelowi, that dwell in shallow burrows close tothe sediment surface were susceptible to shrimp predation whereas a high percentage oflarge adults that build deep burrows survived, even under high predation pressure (Thiel,1997e). Many burrow-dwelling amphipods are sensitive to epibenthic predators, butoften large individuals are preferably taken (due to higher profitability) by predatorswhile small individuals are ignored (see Wilson (1989) and references cited therein).Compared to many other burrow-dwelling amphipods (compare Sainte-Marie, 1991)adults of both L. pinguis and C. bigelowi are relatively large (18–20 mm and 20–26mm, respectively), and therefore they are probably relatively safe from epibenthicpredation in their deep burrows, whereas juveniles in shallow burrows are not.

Sand shrimp also had a very strong effect on the survival of juvenile D. monacanthus.This epibenthic amphipod is particularly susceptible to epibenthic predators and allindividuals quickly disappeared from the trays. Caine (1979) and Aoki (1988) noticedan immediate disappearance of epibenthic caprellid amphipods in field surveys followingthe appearance of demersal fish species. Similarly, Nelson (1979) observed that ineelgrass beds, epibenthic amphipods are least abundant during the summer and fall whenpredators are most active. The fact that only one individual D. monacanthus remained inthe trays with hermit crabs underscores their susceptibility to epibenthic predation.Stephan (1980) noted caprellids Phtisica marina, several polychaete species and evenophiurids Ophiura albida as important predators of Dyopedos sp., indicating thatpodocerid amphipods are susceptible to a wide variety of epibenthic predators.

In summary, juvenile L. pinguis and C. bigelowi are only susceptible to specificpredators (here: sand shrimp), whereas juvenile D. monacanthus appear susceptible to asuite of epibenthic predators (here: sand shrimp and hermit crabs).

4.3. Reaction of juvenile amphipods to epibenthic predators

All three amphipod species reacted to the addition of common estuarine predators byemigration from the trays. This escape-migration has proven a successful strategy forpotential survival of at least 30% of all juveniles initially introduced to trays (except inthe experiments with small juvenile D. monacanthus, see above). The relatively weakemigration-response of juvenile C. bigelowi after the introduction of sand shrimp (seeFig. 1B) is probably responsible for their relatively low survival (only about 30% of alljuveniles) compared to more than 60% survival in L. pinguis (compare Fig. 2).

Escape-migration as a response to predators is relatively common for a variety of¨soft-bottom dwelling amphipods (Ambrose, 1984; Ronn et al., 1988; Thiel and Reise,

1993) and other soft-bottom macro- and meiofauna (Armonies, 1989, 1994). Theseescape-migrations are induced by immediate predator encounters, and therefore mightoccur during times unfavorable for survival of drifting amphipods. Amphipods that haveescaped from their burrows or their mud whips in response to predators are drifting in


the water column where they can become easy prey to visual predators. Furthermore,these escapees are faced with finding a new habitat and constructing a new dwelling.Thus, escape-migration via the water column can only be viewed as a last andpotentially costly resort. The ability to avoid predators and subsequent escape-migrationsmight significantly enhance the survival chances of soft-bottom infauna. Extendedparental care in burrow-dwelling amphipods may be an effective strategy to avoidepibenthic predators.

5. Conclusions

The results of this study suggest that extended parental care is a facultativereproductive strategy in endobenthic marine amphipods while it is obligate for smalljuveniles in epibenthic podocerid amphipods but apparently not in epibenthic caprellidamphipods (Aoki, 1997). Although in L. pinguis and C. bigelowi extended parental careis not considered an obligate reproductive strategy, it might be highly essential for thesurvival of offspring in environments with high predation pressure, such as shallowcoastal waters. Extended parental care is obligate for small juvenile D. monacanthus.Following the establishment of their own mud whips, all size classes in this species(including adults) remain susceptible to epibenthic predators which is probably thereason why its major reproductive period is during winter /early spring when predatordensities and activities on estuarine soft-bottoms are low. Small juvenile L. pinguis andC. bigelowi, that are usually found in their mothers burrows, are rarely found in theirown burrows in the field (Thiel et al., 1997; Thiel, 1997f), and small and medium-sizedjuvenile D. monacanthus are often found drifting in the water column (Thiel, 1997c).These observations indicate that in their natural habitat where predation pressure is high(e.g. Reise, 1985; Thiel, 1997f), predation may substantially affect the survival of thesejuvenile amphipods in the field. In a future study, it will be examined whether parentalprotection can indeed positively affect juvenile survival in an environment with highpredation pressure.

Acknowledgements

I thank K. Reise, I. Voparil and L. Watling for comments on the experimental designand the manuscript. Two anonymous reviewers and J. Dineen gave many helpfulcomments on the final manuscript. A very special thanks goes to S. Zimsen and L.Stearns who ‘saved’ the Casco-experiment while my inflamed appendix was removed. Iwas supported by a graduate fellowship from the Center for Marine Studies at theUniversity of Maine. Additional support was received from the Association of GraduateStudents at the University of Maine, SigmaXi, the American Museum of Natural Historyand an Urda McNaughton scholarship.


References

Ambrose, Jr. W.G., 1984. Increased emigration of the amphipod Rhepoxynius abronius (Barnard) and thepolychaete Nephtys caeca (Fabricius) in the presence of invertebrate predators. J. Exp. Mar. Biol. Ecol. 80,67–75.

Aoki, M., 1988. Factors affecting population fluctuations of caprellid amphipods inhabiting Sargassum patensbed (preliminary report). Bull. Jpn. Assoc. Benthol. 32, 42–49.

Aoki, M., 1997. Comparative study of mother-young association in caprellid amphipods: is maternal careeffective? J. Crust. Biol. 17, 447–458.

Aoki, M., Kikuchi, T., 1991. Two types of maternal care for juveniles observed in Caprella monoceros Mayer,1890 and Caprella decipiens Mayer, 1890 (Amphipoda: Caprellidae). Hydrobiologia 223, 229–237.

Armonies, W., 1989. Meiofaunal emergence from intertidal sediment measured in the field: significant¨contribution to nocturnal planktonic biomass in shallow waters. Helgolander Meeresunters. 43, 29–43.

¨Armonies, W., 1994. Drifting meio- and macrobenthic invertebrates on tidal flats in Konigshafen: a review.¨Helgolander Meeresunters. 48, 299–320.

Bousfield, E.L., 1973. Shallow-water gammaridean Amphipoda of New England. Cornell University Press,Ithaca, New York.

Brown, C.A., Formanowicz, Jr. D.R., 1995.Variation in reproductive investment among and within populationsof the scorpion Centruroides vittatus. Oecologia 103, 140–147.

¨ ¨ ¨Buckle-Ramirez, L.F., 1965. Untersuchungen uber die Biologie von Heterotanais oerstedi Kroyer (Crustacea¨Tanaidacea). Zeitschrift. Morph. Okol. Tiere 55, 714–782.

Caine, E.A., 1979. Population structure of two species of caprellid amphipods (Crustacea). J. Exp. Mar. Biol.Ecol. 40, 103–114.

Clutton-Brock, T.H., 1991. The Evolution of Parental Care. Princeton University Press, NJ.Conlan, K.E., Chess, J.R., 1992. Phylogeny and ecology of a kelp-boring amphipod, Peramphithoe

stypotrupetes, new species (Corophioidea: Amphithoidae). J. Crust. Biol. 12, 410–422.Dauvin, J.C., 1989. Life cycle, dynamics and productivity of Crustacea-Amphipoda from the western English

Channel. 5. Ampelisca sarsi Chevreux. J. Exp. Mar. Biol. Ecol. 128, 31–56.Dickinson, J.J., Wigley, R.L., 1981. Distribution of Gammaridean amphipoda (Crustacea) on Georges Bank.

US Dep. Commerce, NOAA Tech Rep NMFS SSRF-746.Dickinson, J.J., Wigley, R.L., Brodeur, R.D., Brown-Leger, S., 1980. Distribution of Gammaridean amphipoda

(Crustacea) in the Middle Atlantic Bight region. US Dep. Commerce, NOAA Tech Rep NMFS SSRF-741.Fink, L.S., 1986. Costs and benefits of maternal behaviour in the green lynx spider (Oxyopidae, Peucetia

viridans). Anim. Behav. 34, 1051–1060.Goodhart, C.B., 1939. Notes on the bionomics of the tube-building amphipod, Leptocheirus pilosus Zaddach.

J. Mar. Biol. Assoc. UK 23, 311–325.Grant, J., 1981. Dynamics of competition among estuarine sand-burrowing amphipods. J. Exp. Mar. Biol. Ecol.

49, 255–265.Hacunda, J.S., 1981. Trophic relationships among demersal fishes in a coastal area in the Gulf of Maine. Fish.

Bull. 79, 775–788.Harrison, R.J., 1940. On the biology of the Caprellidae. Growth and moulting of Pseudoprotella phasma

Montagu. J. Mar. Biol. Assoc. UK 24, 483–493.Hassack, E., Holdich, D.M., 1987. The tubicolous habit amongst the Tanaidacea (Crustacea, Peracarida) with

particular reference to deep-sea species. Zool. Scripta 16, 223–233.Heinrich, B., 1979. Bumblebee Economics. Harvard University Press, Cambridge.Johnson, S.B., Attramadal, W.G., 1982. Reproductive behaviour and larval development of Tanais cavolinii

(Crustacea: Tanaidacea). Mar. Biol. 71, 11–16.Laval, P., 1980. Hyperiid amphipods as crustacean parasitoids associated with gelatinous zooplankton.

Oceanogr. Mar. Biol. Annu. Rev. 18, 11–56.Lim, S.T.A., Alexander, C.G., 1986. Reproductive behaviour of the caprellid amphipod, Caprella scaura

typica, Mayer 1890. Mar. Behav. Phys. 12, 217–230.Lubin, Y., Henschel, J., 1996. The influence of food supply on foraging behaviour in a desert spider. Oecologia

105, 64–73.


Mahon, R., Neilson, J.D., 1987. Diet changes in Scotian Shelf haddock during the pelagic and demersal phasesof the first year of life. Mar. Ecol. Prog. Ser. 37, 123–130.

´Mattila, J., Olafsson, E.B., Johansson, A., 1990. Predation effects of Crangon crangon on benthic infauna onshallow sandy bottoms — an experimental study from southern Sweden. In: Barnes, M., Gibson, R.N.(Eds.), Trophic Relationships in the Marine Environment. University Press, Aberdeen, pp. 503–516.

Mattson, S., Cedhagen, T., 1989. Aspects of the behaviour and ecology of Dyopedos monacanthus (Metzger)and D. porrectus Bate, with comparative notes on Dulichia tuberculata Boeck (Crustacea: Amphipoda:Podoceridae). J. Exp. Mar. Biol. Ecol. 127, 253–272.

Messana, G., Bartolucci, V., Mwaluma, J., Osore, M., 1994. Preliminary observations on parental care inSphaeroma terebrans Bate 1866 (Isopoda, Sphaeromatidae), a mangrove wood borer from Kenya. Ethol.Ecol. Evol. 3 (Special issue), 125–129.

Michael, A.D., 1987. Variability of the benthic fauna, I. The New England outer continental shelfenvironmental benchmark program, 1977. In: Backus, R.H., Bourne, D.W. (Eds.), Georges Bank. MITPress, Cambridge, pp. 296–302.

Moore, P.G., Earll, R., 1985. Sediment ‘whips’: amphipod artifacts from the rocky sublittoral in Britain. J. Exp.Mar. Biol. Ecol. 90, 165–170.

Nelson, W.G., 1979. Experimental studies of selective predation on amphipods: consequences for amphipoddistribution and abundance. J. Exp. Mar. Biol. Ecol. 38, 225–245.

Reid, M.L., Roitberg, B.D., 1994. Benefits or prolonged male residence with mates and brood in pineengravers (Coleoptera: Scolytidae). Oikos 70, 140–148.

Reise, K., 1985. Tidal Flat Ecology. Ecological Studies 54. Springer-Verlag, Berlin.Richter, G., 1978. Beobachtungen zu Entwicklung und Verhalten von Phronima sedentaria (Forskal),

(Amphipoda). Senckenbergiana maritima 10, 229–242.Richter, G., 1978. Einige Beobachtungen zur Lebensweise des Flohkrebses Siphonoecetes della-vallei. Natur.

Mus. 108, 259–266.¨Ronn, C., Bonsdorff, E., Nelson, W., 1988. Predation as a mechanism of interference within infauna in shallow

¨water soft-bottoms; examples with an infauna predator, Nereis diversicolor O.F. Muller. J. Exp. Mar. Biol.Ecol. 116, 143–157.

Sainte-Marie, B., 1991. A review of the reproductive bionomics of aquatic gammaridean amphipods: variationof life history traits with latitude, depth, salinity and superfamily. Hydrobiologia 223, 189–227.

Shillaker, R.O., Moore, P.G., 1987. The feeding habits of the amphipods Lembos websteri Bate and Corophiumbonnellii Milne Edwards. J. Exp. Mar. Biol. Ecol. 110, 93–112.

Shillaker, R.O., Moore, P.G., 1987. The biology of brooding in the amphipods Lembos websteri Bate andCorophium bonnellii Milne Edwards. J. Exp. Mar. Biol. Ecol. 110, 113–132.

Stephan, H., 1980. Lebensweise, Biologie und Ethologie eines sozial lebenden Amphipoden (Dulichiaporrecta, Dulichia monacantha und Dulichia falcata — Crustacea, Malacostraca). Unpublished Ph.D.thesis, University of Kiel, Kiel, 326 p.

Sudo, H., Azeta, M., 1996. Life history and production of the amphipod Byblis japonicus Dahl (Gammaridea:Ampeliscidae) in a warm temperate zone habitat, Shijiki Bay. Jpn. J. Exp. Mar. Biol. Ecol. 198, 203–222.

Svavarsson, J., Davidsdottir, B., 1995. Cibicides spp. (Protozoa, Foraminifera) as epizoites on the Arcticantenna-brooding Arcturus baffini (Crustacea Isopoda, Valvifera). Polar Biol. 15, 569–574.

¨¨Thamdrup, H.M., 1935. Beitrage zur Okologie der Wattenfauna auf experimenteller Grundlage. Meddr.Kommn. Danm. Fisk.-og. Havunders. (Ser: Fiskeri) 10, 1–125.

Thiel, M., 1997a. Another caprellid amphipod with extended parental care — Aeginina longicornis. J. Crust.Biol. 17, 275–278.

Thiel, M., 1997b. Extended parental care in a high food environment — ‘Babies don’t go in the mud’. In:Hawkins, L.E., Hutchinson, S., Jensen, A.C., William, J.A., Sheader, M. (Eds.), Responses of MarineOrganisms to their Environment. Proceedings of the 30th European Marine Biology Symposium,Southampton, September 1995. Southampton Oceanography Centre, Southampton, 163–170.

Thiel, M., 1997c. Reproductive biology of an epibenthic amphipod with extended parental care, Dyopedosmonacanthus. J. Mar. Biol. Assoc. UK 77, 1059–1072.

Thiel, M., 1997d. Reproductive biology of a filter-feeding amphipod, Leptocheirus pinguis, with extendedparental care. Mar. Biol. 130, 249–258.


Thiel, M., 1997e. Epibenthic predation on marine soft-bottoms — being small and how to get away with it.Proceedings of the 31st European Marine Biology Symposium, St. Petersburg, Russia, September 1996.Hydrobiologia 355, 11–19.

Thiel, M., 1997f. Extended Parental Care in Estuarine Amphipods. Unpublished Ph.D. thesis, University ofMaine, Orono, 188 p.

Thiel, M., Reise, K., 1993. Interaction of nemertines and their prey on tidal flats. Neth. J. Sea Res. 31,163–172.

Thiel, M., Sampson, S., Watling, L., 1997. Extended parental care in two endobenthic amphipods. J. Nat.History 31, 713–725.

Trumbo, S.T., 1992. Monogamy to communal breeding: exploitation of a broad resource base by buryingbeetles (Nicrophorus). Ecol. Entom. 17, 289–298.

Tyler, A.V., 1971. Monthly changes in stomach contents of demersal fishes in Passamaquoddy Bay, N.B. Fish.Res. Bd. Canada. Tech. Rep. No. 288.

Watkin, E.E., 1947. The yearly life cycle of the amphipod Corophium volutator. J. Anim. Ecol. 10, 77–93.Wigley, R.L., Theroux, R.B., 1965. Seasonal food habits of Highlands Ground haddock. Trans. Am. Fish. Soc.

94, 243–251.Wildish, D.J., 1980. Reproductive bionomics of two sublittoral amphipods in a Bay of Fundy estuary. Int. J.

Invert. Reprod. 2, 311–320.Wildish, D.J., 1982. Evolutionary ecology of reproduction in gammaridean Amphipoda. Int. J. Invert. Reprod.

5, 1–19.Wildish, D.J., 1984. Secondary production of four sublittoral, soft-sediment amphipod populations in the Bay

of Fundy. Can. J. Zool. 623, 1027–1033.Wilson, E.O., 1975. Sociobiology: the new synthesis. Belknap Press, Cambridge.Wilson, Jr. W.H., 1989. Predation and the mediation of intraspecific competition in an infaunal community in

the Bay of Fundy. J. Exp. Mar. Biol. Ecol. 132, 221–245.Wilson, W.H., Jr., 1991. The importance of epibenthic predation and ice disturbance in a Bay of Fundy

mudflat. Ophelia (Suppl. 5) 507–514.Wilson, Jr. W.H., Parker, K., 1996. The life history of the amphipod, Corophium volutator: the effects of

temperature and shorebird predation. J. Exp. Mar. Biol. Ecol. 196, 239–250.

Documents

Journal of Experimental Marine Biology and Ecology, 227