Enhancement of the Response of Rock Crabs, Cancer irroratus, to Prey Odors following Feeding Experience

7 Pages·1999·3.2 MB·English

by  A Ristvey

Reference: Bio/. Bull. 197: 361-367. (December
Enhancement of the Response of Rock Crabs, Cancer
Odors
irroratuSy to Prey following Feeding Experience
ANDREW RISTVEY1 AND STEVE REBACH*
Department ofNatural Sciences. University ofMan-land Eastern Shore, Princess Anne, Man-land 21853
Abstract. The rock crab. Cancer irroratus Say, uses Search images, originally postulated for visual stimuli
chemically mediated learning in the search for food. Rock (Croze, 1970; Pietrewicz and Kamil, 1979; Lawrence,
crabs are opportunistic benthic predators and scavengers. 1985a, b) could also be associated with chemical stimuli
Observations indicate that although they eat a variety of (Atema et ai, 1980; Aterna and Derby, 1981; Derby and
items, they are more sensitive to, and prefer, odors offood Atema, 1981). In the aquatic environment, especially in the
items that they have been eating. We foundthat C. irroratus absence of light, chemical signals may be the best cues for
is more responsive to a familiar food source than to an information about the surrounding environment. Organic
unfamiliarone andcandistinguishbetween the odorsoftwo molecules are part of the ambient milieu, and an organism
different prey after being fed one species for an extended must sift out extraneous "noise" to find the information
time. Initial preferences for two mytilid bivalves, Mytiliis needed for foraging, predator avoidance, and mating (Zim-
edulis and Geukensia demissa, were determined in a Y- mer-Faust, 1991).
maze. Crabs were then fed only one of the mussel species Chemosensory cues play a major role in agonistic (Kara-
for 28 days and retested, using sequential and simultaneous vanich and Atema, 1998), sexual (Gleeson, 1980), host-
presentations, for their responses to familiar and unfamiliar finding (Atema and Derby, 1981), and foraging (Pearson
prey odors. Crabs increased their responses to familiar prey and Olla, 1977) behaviors in crustaceans. Experience influ-
odors, but not to unfamiliar odors. In foraging tests, crabs ences an animal's response to those cues. Derby and Atema
ate M. edulis more often regardless of the species to which 1981 demonstrated that after lobsters (Homarus america-
they had been familiarized. n(us) fe)d on a specific prey, theirsensitivity tothat prey odor
increased, and they developed a preference for that partic-
Introduction ular prey. The rocky shore gastropodNucella lamellosa can
A search image can be defined as a perceptual filtering dfliusecnrtismi(nMaatrekboeatnwdeePnalpmreerd,at1o9r9y1)a,nadndnotnhperenduadtiobrryanccrhabAeeof--
mechanism learned from experience. It may be only a lidiapapillosacandistinguish between odors fromalearned
tttaiirtvoatennesnai(tttiBtoooernnnytdriieomasnupnlcdrtaosRnvieilinmenyecc,nreetr1at9sai9eni1ntp)he.perrIcepneoypsftsocuirhabaaligrliaaincbttigyleirbotiefysh,tsaitbvciuismtourbtlehusiisesnlgdesecettldeieicvcs-e-- pmetroernyle.,asne1en9ms8i2ot)ni.eveYaetnlodlosfwpiefvcieinfoittchuenfraisph(oTsohsdiuobnrlnseusapfrtaeelyrbaafcneaeerdmeiosn)ngebsoenc(Hatahmlaelt
criminated by a predator, facilitating more efficient forag- prey foraperiodoftime, but lost theirsensitivity afterafew
ing. Search images increase accuracy and decrease response weeks without reinforcement(Atemaetai. 1980). In preda-
time because the predator requires less information about torily naive postlarval lobsters, responses to metabolites of
the prey and becomes more efficient in locating it (Law- Cancer irroratus and Mytiliis edulis (normal lobster prey)
were lower than in field-collected adult lobsters that may
rence, 1985a).
have had experience with those prey (Daniel and Bayer.
1987a). When naive lobsters were fed amphipods or clams,
Received 19 August 1998; accepted 2 September 1999. those fed amphipods developed stronger responses to am-
1 Present address: 10470 Longwoods Rd., Easton. Maryland 21601.
*Author to whom correspondence should be addressed. E-mail: phipod and not to clam metabolites, but those fed clams did
[email protected]. not develop strong responses to either prey (Daniel and
361
362 A. RISTVEY AND S. REBACH
Bayer, 1987b). In theory, search images need to be rein- The Y-maze
tfafhbooeirrMliaciragetidynsntygar(,enAnsdttgbeutupdhmtaialemafstaeeatyhbwiaalviviha,etaryvyt1eeo9wsf8it0ctet;edhhneAtteehtxfrepoeeeomdrrdaoileo(eannGnceodefnsdeacDranheorderncmb,hwoyi,r1tie9hmc81ae69tg)8phe.1ets)i..aovnaaWinlied-n aFYci-grCm.yrlaiaz1cb,esdRiecwvboeinardtceeahdi.tnteihs1net9get9d6o5)pi5.n50Aatoc9d2m6u0aoXlf-1th3ohe3efadmXsaaMlz2tae0swticaentrmteForla.tecwrxAoylpiaepcrriimepsslctaae(slstteoiieccf
examined search images and foraging behavior in the rock pump (Cole-Parmer #7553.20) delivered liquids through
carnadbMCv.tiihrtrso.rTathuesseprtewyoinggroounptsheofbmiuvaslsveelsmuoscsceulrsinGethuekernasnigae aptlatshtiecrtatuebionfgatbooaut2.05.3-3cm1 hmoilne~'i.n eAithceorncaernmtroaftitohnegYr-amdiaeznet
C was established, with the odor becoming more dilute at the
of irroratus. are frequently taken as prey items (Stehlik.
1993), and have similar physical characteristics. Mytilus is drain located at the base of the maze. In dye trials using
found on hard substrates in the tidal zone, and individuals methylene blue, the average dilution at the base ofthe maze
located close to shore are known to be eaten by rock crabs was determined by spectrophotometer to be 13.6% that of
(Drummond-Davis etui., 1982; Stehlik, 1993). Geukensia is the original concentration. These trials indicated that odor
found in soft substrates in tight clumps attached to marsh reached the crab within 3 min, with little mixing. In pre-
dgraatsasaerse, aavnadiliasblleesosnliinkenlaytetoprbeey eprnecfoeurnetnecreesdinbyC.aicrrraobr.atuNso. ultiemsinoafryittsriralesa,chcirnagbstrheesm.ponEdaecdhttoesotdolarstweidth1i0n maifnewdumriinn-g
Using effluents from these two species, we examined wthheicmhazteheaonbdsebrevhearviroercsoredxehdi,biftreod.m a blind, crab location in
changes in the responsiveness and sensitivity of crabs to
prey. Chemoreceptors on various body parts appear to in-
Mussel
fluence behaviors such as walking, searching, and dactyl effluent
grasping, and these actions are dependent upon the concen- Mussel effluent was produced daily by placing live mus-
tration of the stimulus (Derby and Atema, 1982). The po- sels (equivalent to 10 g soft tissue 1~') in seawater for 10 h
sition andmovementsofthese structures are good indicators (Derby and Atema, 1981). Mussels were checked weekly
of the sensitivity to food odors (Derby and Atema, 1981). for reproductive condition to ensure that effluents would be
We determined a baseline response for rock crabs, famil- consistent in character.
iarized them with a single prey, and retested them to deter-
mine ifthere were acquired orchanged responses tofamiliar Behaviors
and unfamiliar odors.
We adapted methods used by Derby and Atema (1981)
Materials and Methods for tests of chemoreceptive sensitivity and measured
changes in behavior in the Y-maze that reflected changes in
Rock crabs were collected by local watermen using traps sensitivity. An approach to the source of the effluent was
off the coast of Delaware and Maryland and by the inves- defined as a high-sensitivity behavior. Low sensitivity was
tigators at Chincoteague on Assateague Island, Virginia. characterized by the following behaviors:
Ribbed mussels. Geukensia demissa, between 2 and 6 cm.
were collected from salt marsh environments in Girdletree. I Chela raise claws lifted beyond normal position.
Maryland, and Chincoteague, Virginia. Blue mussels, Myti- 2. Antennule burst flicking rate increased suddenly.
lus edulis. between 2 and 6 cm, were collected on the rock 3. Antennule wipe antennules groomed, usually with
jetty at the Ocean City, Maryland inlet. third maxillipeds. May occur in bouts. Wipes occur-
ring within 5 s of each other were considered to be
Maintenance conditions one wipe.
A recirculating, biologically filtered saltwatersystem was 4. bMaaxcikllainpdedforwtahvweithtohuitrdtomuacxhiilnlgipoendes amnootvheedr.slowly
used for tests. The water temperature was 1 1 3C, the 5. Maxilliped wipe third maxillipeds rubbed against
sliaglhitn:it1y2 hwadsark3.2Ctroab3s5wpeprte, kaenpdt itnhe40p-1hottaonpkesri(5o0d Xwa2s5 c12m),h each other within a 5-s period.
6. Shift body position changed.
two pertank, with an acrylic plastic divider separating them 7. Body raise body raised up on dactyls.
to prevent aggression and to enable staggered feeding. They 8. Fanning rapid movement of second maxillipeds
were acclimated to laboratory conditions for 1-2 weeks and along with third maxillipeds opened widely to ex-
fweedreahdoiuesteodfinsqaunidis(oLloatleid^o80s-p1.)taenvkereyquoitphpeerddwaiyt.h Maupsosweelrs pose mouth parts.
filter. Fresh mussels were collected every week. The water An approach was scored when a crab crossed a line 8 cm
in the mussel tank was kept at 15" 2"C and the salinity at from the inlet flow at the end of an arm of the Y-maze
33 to 34 ppt. before a 10-min run was completed. Crabs began the ex-
EXPERIENCE AND RESPONSES TO PREY ODORS 363
periment at the base of the maze. If an approach did not the species handled first, eaten first, or rejected after being
occur during a test run. the low-sensitivity behaviors were handled were recorded. Mussel shells were marked with
used for scoring. In every run, each occurrence of a behav- small spots of white epoxy to make them easier to see
ior other than an approach was counted as one unit (Derby during videotape analysis.
and Atema. 1981). Totals for each crab were then averaged
to determine a mean frequency. The higher the value, the Analysis of data
more sensitive the crab was, or had become, to the prey
odor. Scores were obtained for each crab tested before and Initial scores for responses to seawater, the Mytilus and
after training was complete. Geukensia effluents, and the sequential presentation test
results were compared using the Friedman test (Systat 8.0,
Initial response to mussel effluent SPSS Inc., Chicago, Illinois). Responses to familiar and
unfamiliar odors for each familiarization group were com-
Sixteen crabs were fasted for 24 h and then tested. In pared with the Wilcoxon signed rank tests. (Systat 8.0).
control tests, seawater was used on both sides of the Y- Differences between means were determined with Bonfer-
maze. Each crab was allowed to acclimate in the maze for roni post hoc analysis (Systat 8.0). Mussel selection data
8-12 h and then tested with effluent and a seawatercontrol. from the foraging test was analyzed using a Wilcoxon
This was repeated for the other mussel species about 10 h signed rank test, a Mann-Whitney test with tied ranks, and
2x2
later. Initial response tests were completed within 24 days. a chi-square contingency table (Zar, 1984).
Familiarization with a specific mussel odor Results
After testing crabs for their initial response, training Behavioral obsen'ations
began. Eightcrabs were fed Geukensia and eightM\tilus for
a period of 28 days. Crabs were fed whole mussels ad Crabs responded toeffluents within 2 to 3 min. Those that
libitum during the training period. The average number of did not approach responded by displaying lower sensitivity
mussels eaten each day was recorded. behaviors. Typically, crabs flicked their antennules slowly
or intermittently, with occasional bursts, before odors
Response to odors in sequentialpresentation after orcecaucrhreeddtwhehme.nAthbeuresftf,luweintthremaaxcihleldiptheedcorraba.ntWeintnhuilne 5wimpiens,.
familiarization
chela waves and raises occurred, and crabs began to move.
Familiarized crabs were retested as in the first experi- Antennule flicks pointed in the direction of movement.
ment. The tests began with a post-familiarization seawater These behaviors continued until tests were concluded.
control using seawater in both arms of the Y-maze. Re- In a typical approach, the initial behavior was similar to
sponses to the two prey effluents (familiar and unfamiliar) thatofanon-approach. Atabout 5 min,crabsbeganwalking
were recorded based on sequential presentation: each prey towards the effluent. Upon reaching the end of the maze,
odor was tested against a seawater control. Approaches they often grabbed the inflow hole with their chelae. In
wererecorded when they occurred; ifno approachoccurred, some simultaneouspresentation tests, crabs entered one arm
low-sensitivity behaviors were scored. of the maze, turned back, and then proceeded down the
other side, through which the familiar effluent flowed.
Response to odors in simultaneous presentation after
familiarization Initial response to mussel effluent
Familiarized crabs were retested for preference between Figure 1 shows the responses of 16 crabs to mussel
the two effluents. Odors were presented simultaneously effluent before the crabs were familiarized with other mus-
without a seawater control. Distinguishing which odor elic- sel species. The Friedman test revealed no differences be-
ited heightened behavioral responses was not possible in tween the responses to the seawater control and to the
>
this test, so only approaches were scored. Mytilus or Geukensia effluents (P 0.05, Table I).
Influence ofexperience on mussel selection Familiarization with mussel odor
A foraging test was performed using live prey in 40-1 Familiarization periods began after initial responses were
tanks. Five equal-sized mussels of each species were ran- obtained. Crabs werefedtheirassigned speciesofmusselad
domly positioned in the tank and buried to about 667r of libitum. At first, crabs consumed 6 to 8 mussels a day,
their length in a calcite substrate to make them factually although within 4 weeks this decreased to 2 to 3 mussels a
cryptic. Crabs were allowed toforage for 12 h, andeach test day, especially those fed Geukensia. During familiarization,
was videotaped. The total number of mussels handled and crabs exhibited periods of increased activity. Usual behav-
364 A. RISTVEY AND S. REBACH
3.5 died during the familiarization period; another crab, fed
with blue mussel, stopped eating 3 weeks into the training
3.0 period. Activity scores from the remaining 12 crabs were
analyzed for increases in sensitivity to familiarprey odor. A
2.5 Friedman test determined that responses were significantly
<
different (P 0.01, Table I)topre-familiarization seawater,
post-familiarization seawater. unfamiliar mussel effluent,
2.0 H
and familiar effluent (Fig. 2). Bonferroni post hoc analysis
demonstrated that the response to familiar effluent was
s 1-5 H significantly greater than to all of the other water samples
<u
and effluents (P < 0.01). No other significant differences
1.0 were found (Table 1).
Figure 3 shows the responses of crabs familiarized with
0.5 Mytilus or with Geukensia to familiar and unfamiliarodors.
Wilcoxon signed ranks tests determined that responses dif-
0.0 fered to familiar and unfamiliar odors for both familiariza-
SeawaterControl Mytilus Odor Geukensia Odor tion groups (P < 0.05, Table 1). Bonferroni post hoc com-
parison determined that each familiarization group could
Figure 1. Comparison of mean (SEM) frequencies for the initial detect the difference between the odors that were familiar
tests for responses toodor. No differences were found between responses
to odors. and the odors that were unfamiliar to their training groups
(P < 0.01, Table I). There was nodifference in the response
of each of the familiarization groups in their ability to
iors consisted of climbing tank walls, eating, and resting distinguish their own familiar ortheir own unfamiliar odors
>
with intermittent antennule flicking. (P 0.05).
Response to odors in sequential presentation after Response to odors in simultaneous presentation after
familiarization familiarization
Two crabs, fed with blue mussel, made approaches to Of the 8 crabs familiarized with Mytilus, 4 approached
their familiar effluent. One crab, fed with ribbed mussel. familiar prey effluent; ofthe 8 familiarized with Geukensia,
Table I
Responses to effluents
Odors compared Method ot comparison P value
Initial odors Friedman statistic (Fig. 1) 16.16.16 >0.05
Pre-familiarization and post-familiarization seawater, familiar and unfamiliar Friedman statistic (Fig. 2) 12, 12, 12. 12 <0.01
mussel effluents
familiar vs. unfamiliareffluents Bonferroni test (Fig. 2) 2. 12 <0.01
familiar effluent vs. post-familiarization seawater Bonferroni test (Fig. 2) 2. 12 <0.01
familiar effluent r.v. pre-familiarization seawater Bonferroni test (Fig. 2) 2, 12 <0.01
unfamiliareffluent vs. pre-familiarization seawater Bonferroni test (Fig. 2) 2. 12 >0.05
unfamiliareffluent vs. post-familiarization seawater Bonferroni test (Fig. 2) 2. 12 >0.05
Familiar and unfamiliarodors-familiarized w/ Mvtiltis Wilcoxon signed ranks test (Fig. 3) 5 <0.05
Familiar and unfamiliarodors-familiarized w/Geitkcnsta Wilcoxon signed ranks test (Fig. 3) 7 <0.05
familiar r.v. unfamiliarodors-familiarized w/Mvtilus. Bonferroni test (Fig. 3) 5, 5 <O.OI
familiar vs. unfamiliarodors-familiarized vt/Gettki'iixia Bonferroni test (Fig.3) 7.7 <0.01
unfamiliarodor-familiarized vj/Gi'iikensia vs. unfamiliarodor-familiarized Bonferroni test (Fig. 3) 7. 5 >0.05
w/Mytilus
familiar odor-familiarized w/Geukensia r.v. familiar odor-familiarized w/Mytilus Bonferroni test (Fig. 3) 7. 5 >0.05
Mylilns eaten vs. Geukensia eaten Wilcoxon signed ranks test 14 <0.05
Mytilus rejected vs. Geukensia rejected Wilcoxon signed ranks test 9 >0.05
No. mussels eaten-familiarized on Mytilus or Geukensia Mann-Whitney tied ranks (Fig. 4) 7. 7 >0.05
No. mussels rejected-familiarized on Myiilitx or Geukcnsiu Mann-Whitney tied ranks (Fig. 4) 7. 7 >0.05
2X2
Species eaten vs. familiarization species chi-square (Fig. 4) 14 >0.05
2X2
Species rejected r.v. familiarization species chi-square (Fig. 4) 14 >0.05
EXPERIENCE AND RESPONSES TO PREY ODORS 365
14 T Low-sensitivity scoring could not be used because it would
not have been possible to determine which odor was influ-
encing the behavior.
Influence ofexperience on prey selection
8 Within 15 min ofbeing placed in the tanks, crabs probed
the calcite substrate with their dactyls, attempted to climb
tank walls, or walked around. During this period, all crabs
touched and moved both species of mussels.
More Mytilus (45) were eaten than Geukensia (19), but
about the same number of mussels of both species were
rejected (Mytilus. 13; Geukensia, 14). A Wilcoxon signed
H ranks test determined that, overall regardless of familiar-
Initial Post Unfamiliar FamiliarOdor ization group significantly more Mytilus were eaten than
Seawater Seawater Odor Geukensia (P < 0.05), but there was no difference in
Control Control numbers of Mytilus or Geukensia rejected (P > 0.05) (Ta-
Figure 2. Comparison of mean (SEM) frequenaes of initial and ble I).
post-familiarizationcontroltests,andtestsoffamiliarandunfamiliarodors Figure 4 compares the species ofprey eaten and rejected
after familiarization. No differences were found between responses to by crabs in the two familiarization groups. Crabs familiar
cswaeetarewedatsbeiyrgnaicfnoincatasrntotellrsyiskag)nr.edatuenrfatmhialniatro oudnofrasm,ilibaurt ordeosrpsonasensd ctoontfraomlislia(rasoidnodris- swiiat,haGnedukreenjsecitaedcu1mu1lMayttiivleulsy aatned211 1MyGteiulkuesnsainad.5CrGaebuskefna--
miliar with Mytilus ate 24 Mytilus and 14 Geukensia. and
rejected 2 Mytilus and 3 Geukensia. In all cases, Mytilus
1 approached unfamiliar effluent. These tests failed to yield was handled first and eaten first. Both groups of crabs
significant responses because only approaches were scored. handled about the same number ofmussels (43 for Mytilus-
The 1 1 crabs that did not approach during the simultaneous familiarizedcrabsand48 forGeukensia-famiYiarizedcrabs).
presentation exhibited increased sensitivity, as in the se- A Mann-Whitney test with tied ranks determined that
quential tests. These crabs began to search soon after the there was no significant difference in total mussels, regard-
odor reached them; they raised chela and walked upcurrent less of species, eaten (P > 0.05) or rejected (P > 0.05),
and downcurrent, but they failed to make an approach. between crabs familiarized withMytilus and crabs familiar-
ized with Geukensia (Table I).
18 - DFamw/ Mytilus
HFamw/Geukensia
16 30 i
14
25
QFamw/Mytilus
12
HFamw/ Geukensia
20-
10
8-
15
6
10
4. .
2 5
UnfamiliarOdor FamiliarOdor
Mytilus Eaten Geukensia Mytilus Geukensia
Figure 3. Comparison ofmean (SEM) frequencies between famil- Eaten Rejected Rejected
iarization groups. Nodifferenceswere found in unfamiliarodorresponses
between familiarization groups. Differences were found in familiar odor Figure 4. Comparison of lotal number of prey eaten and rejected by
responses. Abbreviations: Fam w/Mytilus = Familiarized withM. edulis; crabsfamiliarizedwilhMyiilusandGeukensia. Nodifferenceswerefound
Fam w/ Geukensia = Familiarized with G. demissa. in ratios. Abbreviations: Same as in Figure 3.
366 A RISTVEY AND S. REBACH
A chi-square analysis of a 2 X 2 contingency table g) perliter, andhandling and ingestionoffoodat milligrams
determined that the ratio ofMytilus to Geukensia eaten was (10"' g) per liter (McLeese, 1973; Mackie, 1973; Pearson
not different (P > 0.05), norwas there a difference between and Olla, 1977; Ache, 1982). The effluents used in this
>
the ratio of mussel species rejected (P 0.05) for each of study did not often direct the crabs' responses towards
the familiarization groups (Table I). Regardless ofthe mus- familiar effluents, but did arouse them. We may therefore
sel species that the crabs were familiar with, they ate and infer that these effluents had concentrations between 10~'
rejected the same proportion of mussel species. and 10"' g I"1.
The stomach contents ofrock crabs indicate that they are
Discussion opportunistic feeders (Drummond-Davis et ai, 1982). An
assortment of algae, polychaetes, gastropods, mussels, and
Our results show that, in the absence of recent experi- bits of hermit crabs and other crustaceans are typically
ence, Cancer irroratus did not strongly respond to the consumed. It is possible that nutritional needs were not
effluent of either Mytilus editlis or Geukensia demissa. being met by a diet restricted to a single food for an
Exposure to one prey type increased the sensitivity of the extended time, and the crabs may have lost interest in
crabs to that prey's odor, and responsiveness increased with familiarized prey. Again, feeding reinforcements were ab-
experience. The results of the sequential presentation tests sent and may have counteracted the effects of training.
indicated a significantly increased sensitivity towards famil- In the foraging tests, tactile and visual cues were intro-
iarized prey. Scores of low-sensitivity behaviors were duced by the use of living prey rather than effluents. Both
higher for familiar effluents than for unfamiliar effluents. groups of familiarized crabs ate and rejected similar num-
However, the simultaneous presentations could not distin- bers ofmussels, and there was no difference in the ratios of
guish the responses to the two odors because it was not Mytilus to Geukensia eaten and rejected. However, even
possible to determine which odor was eliciting the height- though crabs encountered both species of mussels, they
ened behaviors. handled Mvtilns first, ate Mytilus first, and consumed more
Crabs did not often approach the effluent source in either Mytilus than Geukensia regardless offamiliarization group.
the sequential or simultaneous tests. These results may be Since crabs walked over both species before selecting any
misleading since crabs did react to the odors. The use of prey, both species should have had an equal chance to be
prey effluents instead of live prey can influence observed handled. Geukensia has a heavier shell than Mytilus. possi-
behaviors, because the lack of reinforcement with actual bly making it more difficult toopen, but this did not account
prey may have been responsible for the observed decreases for the crabs' preference, because equal numbers of both
in response. species were rejected. Metabolites from both species were
It is also possible that the concentrations of stimulatory present in the test tank. As soon as a mussel was eaten,
compounds may have been below the thresholds necessary freshly killed preyodors would have filled the tank, possibly
to initiate search or approach. The amino acids glycine, decreasingthe importance ofodorinchoosingthe next prey,
taurine, glutamate, serine, and threonine have been found to and other cues may have become more important.
be the most stimulatory in feeding assays in several species Ifcrabs use more than one sensory cue in prey choice, a
of Cancer (Case, 1964; Allen et at., 1975). These amino hierarchy may exist forall sensory functions in determining
acids may have been present at low concentrations in test- prey selection. Maynard and Sallee (1970) found that che-
mussel metabolites. Palaemonetes pugio. a grass shrimp, motactile stimulation oflobster dactyls overrode antennular
specifically recognizes various foods by qualitative and stimulation. Our tests were run in the light to permit video-
quantitative differences in combinations of low molecular taping, so visual cues might have played a role in prey
weight substances intrinsic to those foods (Carr, 1978). selection. Geukensia was difficult to see against the mottled
Concentrations that elicited antennular responses may calcite, whereas the blue-black colored Mytilus contrasted
have been too low for activation ofapproaches and feeding well with the background. Arthropod compound eyes are
behaviors. Rebach etat. (1990) found antennular sensitivity adept at discerning contrasts (Evans. 1984). Mytilus may
for mussel extract in C. irroratus to be as low as 10""' g have been visually less cryptic and thus more susceptible to
I '. Pearson et at. (1979) found similar sensitivities for predation.
littleneck clam extract in C. magister. Both ofthose studies The crabs were more likely to have been in contact with
used tissue extracts, whereas this study used prey rinse Mytilus than with Geukensia before they were caught, and
(body odor) from intact animals. The threshold to elicit might have retained their sensitivity for that species. Alter-
feeding is 10s higher than the arousal threshold in rock natively, Mvtilim may have beeneasiertoopen, ormay have
crabs (Rebach et at.. 1990) and H)1" to l()17 times higher in simply tasted better than Geukensia. Lobsters are also able
blue crab (Cullinectes sapidus; Ache, 1982). Arousal to detect differences between two closely related mus-
thresholds are found at concentrations ofpicograms ( 10"i: sels in this case Mytilus and Modiolus and showed in-
g) per liter, search behaviorthresholds at micrograms (10~6 creased sensitivity with experience and training (Derby and
EXPERIENCE AND RESPONSES TO PREY ODORS 367
Aterna, 1981). Atema et al. (1980) found qualitative differ- Daniel, P. C.. and R. C. Bayer. 1987b. Development of chemicalh
ences in the amino acid content of live prey rinses. The mediated prey-search response in postlarval lobsters (Homarus ameri-
foraging study supported the differences found in sensitiv- canus) through feeding experience. J. Chem. Ecol. 13: 1217-1233.
ities between crabs familiarized with Mvtilus and crabs Derby, C. D.,andJ. Atema. 1981. Selective improvement in responses
to prey odors by the lobster, Homarus americanus. following feeding
familiarized with Geukensia. However, M\tilus appeared to experience. / Chem. Ecol. 7: 1073-1080.
be more attractive than Geukensia when crabs were given a Derby,C.D.,andJ.Atema. 1982. Chemosensitivityofthewalkinglegs
choice between live prey. of the lobster. Homarus americanus: neurophysiological response
The responses of crabs to mussel odors before and after spectrum and thresholds. J. Exp. Biol. 98: 303-315.
experience with those mussels indicated that familiarization Drummond-Davis, N. C.. K. H. Mann, and R. A. Pottle. 1982. Some
increased sensitivity towards a prey item. Whether this eCsatnicmeatreisrroofiatpuosp.uliantaiokneldpebnseidtyinaNnodvafeSecdoitniga.hCaabni.tsJ.inFisthh.eArio/ncakt.crSacbi..
resulted in the formation ofachemosensory search image or 39: 636-639.
a species-specific preference is not clear. However, recog- Evans,H. E. 1984. InsectBiology:A TextbookofEntomology. Addison-
nition and remembrance of familiar prey odors facilitates Wesley. Reading. Massachusetts. 436 pages.
the location of suitable prey in a benthic habitat where few Gendron, R. P. 1986. Searching forcryptic prey: evidence for optimal
other cues are available. search ratesandthe formationofsearch images inquail.Anim. Behav.
34: 898-912.
Gleeson, R. E. 1980. Pheromone communication in the reproductive
Acknowledgments behaviorofthebluecrab, Ctillinecte.tsapidits. Mar. Behav. Ph\siol. 7:
119-134.
We
thank D. French. C. Loshon, and V. Kennedy fortheir Hall, S. J., C. D. Todd, and A. D. Gordon. 1982. The influence of
advice and assistance, E. Layman, M. Ailes, and D. Birkett ingestive conditioning on the prey species selection inAeolicliapapil-
for their help and support. We also thank two anonymous losa (Mollusca: Nudibranchia). J. Client. Ecol. 51: 907-921.
reviewers for their helpful suggestions and thoughtful com- Karavanich, C., and J. Atema. 1998. Olfactory recognition of urine
pmfauerlntftiiaalrlllymy.enstAuoRpfpostrhutebemdMitSbtyeddNegpSarFretesGraotafnUttMhi#EsSRm.IaInT-uh8si7cs0r4ir0pe5ts4e.ianrTcphhairswtiaaisls Lawrmnsuieesgnn.nicalelBas,ehiLEan..v):diS.oosumh1rio9nr81ta53-na5tc.:eerm7f1Eil9vge-ihatd7rsen3in0bnc.eget.wfAeonreinsme.maarBclehehialmvo.absgt3ee3r:i.n9Hb2ol9amc-ak9rb3iu7rs.dsam(Tenrridciai-s
Contribution No. 30 from the Crustacean Research in Ecol- Lawrence,E.S. 1985b. Evidenceforsearchimageinblackbirds(Turdus
ogy and Mariculture (CREAM) Institute of the University menila L.): long-term learning. Anim. Behav. 33: 1301-1309.
of Maryland Eastern Shore. Mackie, A. 1973. The chemical basis of food detection in the lobster.
Hinnarusgainmarus. Mar. Biol. 21: 103-108.
Marko, P. B., and R. Palmer. 1991. Responses of a rocky shore
Literature Cited
gastropodtotheeffluentsofpredatory andnon-predatorycrabs: avoid-
Ache,B. 1982. Thermoreceptionandchemoreception. Pages369-398in ance and attraction. Biol. Bull. 181: 363-370.
TheBiologyofCrustacea. Vol. 3. H. L. Atwoodand D. C. Sandeman. Maynard, D. M.,and A.Sallee. 1970. Disturbanceoffeeding behavior
eds. Academic Press. New York. inthespiny lobster.Punulirtisargils, followingbilateralablationofthe
Allen, W., E. Frederick, and R. Wong. 1975. Experiments on the medulla tcniiinallis. Z. Vgl. Phvsiol. 66: 123-140.
developmentofartificial bait forthe dungeness crab. Cancermagister McLeese, D. 1973. Olfactory responses of lobsters (Homarus america-
(Dana). Sea Grant Rep. HSU-SG7. Humboldt Slate Univ.. Arcata, nus) to solution from prey species andto seawaterextracts andchem-
California. ical fractions of fish muscle and effects of antennule ablation. Mar.
Atema,J.,andC.Derby. 1981. Ethological evidenceforsearch images Behav. Phvsiol. 2: 237-249.
in predatory behavior. Pages 395-400 inAdvancedPhysiologicalSci- Pearson,W.H.,and B.L.Olla. 1977. Chemoreception inthebluecrab,
ences. Vol. 16. Sensory Functions. E. Graystyan and P. Molnar, eds. Ctillincctes stipidits. Biol. Bull. 153: 346-54.
Pergamon Press, New York. Pearson,W.H..P.C.Sugarman,D.L.Woodruff,andB.L.Olla. 1979.
Atema,J.. K. Holland,and W. Ikehara. 1980. Olfactory responses of Thresholds fordetection and feeding behavior in the Dungeness crab.
yellowfin tuna (Thunnus albacares) to prey odors: chemical search Cancermagister. J. Exp. Mar. Biol. Ecol. 39: 65-78.
image. J. Cliem. Ecol. 6: 457-465. Pietrewicz, A. T., and A. C. Kamil. 1979. Search image formation in
Bond, A. B., and D. A. Riley. 1991. Searching image in the pigeon: a the bluejay (Cyanocitta cristala). Science 204: 1332-1333.
test ofthree hypothetical mechanisms. Ethology87: 203-224. Rebach, S. 1996. The role of odor in prey recognition by rock crabs.
Carr, W. E. S. 1978. Chemoreception in the shrimp. Pulaemonetes Cancerirroratits Say. J. Chem. Ecol. 22: 2197-2207.
pugio: the role ofamino acids and betaine in elicitation ofa feeding Rebach,S.,D.P.French,F.C.VonStaden,M.Wilber,andV.E.Byrd.
response by extracts. Comp. Biochem. Phvsiol. 61: 127-131. 1990. Antennular sensitivity of the rock crab Cancer irroratits to
Case,J. 1964. Adequate stimulus ofcertain Cancerdactyl chemorecep- food substances. ./. Crustac. Biol. 10: 213-217.
tors. Am. Zool. 4: 276-277. Stehlik, L. L. 1993. Diets ofthe brachyurancrabs Cancerirroratits. C
Croze, H. J. 1970. Searching image in carrion crows. Z. Tierpiychol. borea/is, and Ovalipt.i ocellatus in the New York Bight. J. CniMm
Suppl. 5: 1-85. Biol. 13: 723-735.
Daniel, P. C.. and R. C. Bayer. 1987a. Attraction ofpredatorily naive Zar,J.H. 1984. Bit-statisticalAnalysis. PrenticeHall,EnglewoodCliffs,
postlarval lobster to extracts ofmetabolites ofcommon prey: Mvtilus New Jersey. 718 pages.
edulis,Myaarenaria. Cancerirroratits,andAsteriasvitlgaris.J. Chem. /Jmmer-Faiist.R.K. 1991. Chemical signal-to-noisedetectionbyspiny
Ecol. 13: 1201-1216. lobsters. Biol. Bull. 181: 419-426.