Antipredator Defenses in Tropical Pacific Soft Corals (Coelenterata: Alcyonacea). I. Sclerites as Defenses Against Generalist Carnivorous Fishes

10 Pages·1992·3.9 MB·English

by  K L Van Alstyne

Reference: Biol. Bui! 182: 231-240. (April, 1992)
Antipredator Defenses in Tropical Pacific Soft Corals
(Coelenterata: Alcyonacea). I. Sclerites as Defenses
Against Generalist Carnivorous Fishes
KATHRYN L. VAN ALSTYNE1 CHAD R. WYLIE,
VALERIE J. PAUL, AND K,AREN MEYER
University ofGuam Marine Laboratory, UOG Station, Mangilao, Guam 96923
Abstract. Calcified sclerites are common in many in- functions at particular locations as well as constraints
vertebrate species and are frequently used as taxonomic upon their use or production.
indicators; however, little isknown about the function of
sclerites. To determine whether sclerites could function Introduction
as antipredator defenses, we conducted field assays in Softcoralsare frequently aconspicuous component of
which sclerites from the Indo-Pacific softcoralsSimdaria shallow, Indo-Pacific tropical reef communities despite
maxima, S.polydactyla, and 5. sp. were incorporated into the abundance of carnivorous fishes. For example, on
an artificial diet and offered to a natural assemblage of Guam, soft corals have been reported to provide ~95%
fishes in the field. Sclerites from both the tips and bases ofthe total livinganimal cover on some reefs(Wylie and
ofall three species ofSinularia reduced feeding by a nat- Paul, 1989). On NewGuinean reefs, softcoralsconstitute
ural assemblageofgeneralist carnivorousfishesoffGuam approximately 50% ofthe living coverbetween depths of
by 27-44%; however, sclerites from the bases ofthe col- and 5 m (Tursch and Tursch, 1982).
onies were 18-51% more deterrent than tip sclerites. The The persistence ofsoft corals and gorgonians in areas
greatereffectiveness ofsclerites from the bases ofthe col- withhighlevelsofpredation haspreviouslybeenattributed
onieswas largely attributable to their high concentrations. to their production ofpredator-deterrent secondary me-
Scleritesin thetipsofthecoloniesoccurred in mean con- tabolites (Coll et ai, 1983; LaBarre et al., 1986; Pawlik
centrations from 24 to 58% by dry weight and were gen- et ai, 1987; Wylie and Paul, 1989); but, soft corals and
mm
erally lessthan 0.5 in length. Sclerites in the bases of gorgonians also produce sclerites or spicules that could
the colonies were larger and occurred in average concen- potentially serve as antipredator defenses (Harvell and
trations of82-88%. Simdaria sclerites were increasingly Suchanek, 1987; Sammarco et al., 1987; Harvell et al.,
effective as feeding deterrents with increasing concentra- 1988;HarvellandFenical, 1989). Mineral-hardened spic-
tion at concentrations less than 30-50% by dry weight. ulesare common within a numberofinvertebrategroups
The effectiveness of sclerites as deterrents leveled offat including sponges, cnidarians, platyhelminth worms,
higher concentrations. Sclerite morphology was also im- mollusks, echinoderms, and ascidians (Kingsley, 1984).
portant in determining antipredator activity. Although However, despite the widespread occurrence ofsclerites
sclerites can play a role in predator deterrence, they can within marine invertebrates, little is known about the
also function in the structural support ofcolonies. Thus, function ofthese structures, in particular, their ability to
the sizes, shapes, and abundances ofsclerites at different deter feeding by potential predators.
locations within colonies will be determined by their Only in recent studies has the role ofsclerites as anti-
predatordefensesbeen explored. When incorporated into
an artificial diet, sclerites ofthegorgonian seawhip Pseu-
R'ePcreeisevnetdA1dd1rJeuslsy:1D9e9p1a;ratcmceenptteodf2B3iolJoagnyu,arKyen1y9o92n.College,Gambier. dopterogorgoria acerosa deterred feeding by carnivorous
Ohio43022. fishes in field assays in Belize (Harvell et al.. 1988). Scler-
231
232 K. L. VAN ALSTYNE ET AL.
itesofthe Caribbean gorgonian Gorgonia ventalinadeter
feeding by natural assemblages offishes in the field and
by the gorgonian specialist Cyphomagibbosum (Van Al-
styne and Paul, in press). Toxicity of soft corals to the
mosquito fish Gambmia affiniswas negatively correlated Tips
with the degree ofarmament ofthe polyps in the Neph-
thedidae and negatively correlated with the degree of
mineralization ofthe coenenchyme in 14 species ofSin-
itlaria (Sammarco et a!., 1987). Alcyonarian sclerites are
extremely variable in size and morphology and are fre-
quently used as taxonomic characters (e.g., Bayer, 1956,
1961; Bayerelai. 1983). Seasonal fluctuationsin collagen
levelsin the organic matrix ofscleritesindicate that these Base
structures are dynamic and may undergo seasonal cycles
ofdemineralization andremineralization (Kingsleyetai.
1990).
Inthis paper, we provide direct experimental evidence Figure2. SamplinglocationswithinanindividualSinulariacolony.
thatsclerites from thealcyonian softcoralsSinulariaspp. Threetransectsofsixsampleseachweremadefromthetipsofthecolony
serve as antipredator defenses to generalist carnivorous tothebaseofthecolony. Transects ran from themostdistaltipsofthe
fishes in the field. Soft corals ofthe genus Sinularia are colonies(location 1)tothepointofattachmentatthebaseofthecolony
widely distributed throughout the Indo-Pacific region (location 6).
(Verseveldt, 1977, 1980), and are an important compo-
nentofshallowcoral reefcommunitiesonGuam (Gawel,
Materials and Methods
1977). Sinulariaspp. generally havealobate morphology
with an upper portion that contains polyps, and a lower Studvsites
sterile stalk (Verseveldt, 1977). The surface ofthe lobes
c(Voenrtsaeivnesldstm,all198c0l)u.b-sWheapeadlsosclderoicteus,mernotds,diafnfderesnpciensdleisn lecStiendulfarroima amapxaitmcha,reSe.fpionlyCdoacctoyslaL.agaonodnS,.Gspu.awme,reUcSolA-
sclerite morphologies and densities within soft coral col- a(Fnidg.V1a).nTAhlisstyrneeef(h1a9s87b)eeanndprWeyvliioeusalnyddePsacurlib(e1d98b9)y.PIatuils
onies and explore the consequences ofthese differences a small patch reef, ~35 m X 50 m, that is composed
in deterring feeding by fishes. mainly ofdead Acropom. Soft corals comprise ~ 10% of
the cover ofthe reef(Wylie, 1988).
All feeding experiments took place on Fingers Reefin
Apra Harbor, Guam. The reefat Fingers (Fig. 1) is com-
posed ofa diverse assemblage ofscleractinian corals, as
wellasmany unidentifiedspeciesofSinularia. Numerous
species of herbivorous and carnivorous fish inhabit this
reef. Thefishesthatwereobservedtofeedon ropesduring
feedingexperimentsincludedsergeant majors(Abudefduf
spp.), damselfish (Amblyglyphidodon curacao), wrasses
(Cheilinusfasciatm, Gomphosus varius, Halichoeres tri-
macitlalus), and triggerfish (Balistapus undulatiis).
Quantification ofsclerite concentrations
Five soft corals each ofSinularia maxima, S. polydac-
tyla, and 5. sp. were collected from Cocos Lagoon. Six
^Cocos samplesof~2 cm (height) X 1 cm (length) x 1 cm (width)
were removed from each ofthree "transects" ofeach col-
Lagoon
ony. The transects ran from the most distal part of the
Figure 1. Locations ofthe Fingers Reefand Cocos Lagoon study tips to the point ofattachment at the base (Fig. 2). These
sitesonGuam. samples areas will be referred to as locations #1 (most
STRUCTURAL DEFENSES IN SOFT CORALS 233
distal tip sample) through #6 (base sample closest to the Two-choicefeeding experiments
attachment point ofthe colony). Eighteen samples were
taken from each ofthe 15 colonies. Sclerites concentra- Sclerites from soft coral tips and bases were incorpo-
tions from each sample were quantified as described in rated intoan artificial diet consistingof2.5 gcarrageenan
Van Alstyne and Paul (in press). (Sigma #C-1013), 4 g ofparaffin wax, and 70 ml water,
in Amneaannalsycsliesriotefvcaornicaenncterawtaisonusseadmtoontgestspfeocriedsi,ffaermeoncnegs rheaagteeednainnmaixmtiucrreowwaasveheoatveedn,f~o5r075gso.fAsfqtueirdthhoemwoagxe/ncaatre-
individuals within a species, among transects within an (250ml water:500gsquid)andscleriteswereadded. Con-
individual,andamonglocations. Amixed modelANOVA centrations ofsclerites used in these feeding experiments
was conducted with the factors being (1) species, (2) in- aorfecolniscteendtrinatTiaobnlsemeI.asTuherseed ivnalbuaessesaraendwittihpisnoftShienurlaanrgieas
dividuals nested within species, (3) transects nested within
dtiinupdcsitvetioddutwahilest,hbaaasnnedsS(P4o)SflStoPhceCatci+oolnsostnaiatelisso.tnigcTathlhepeatacrnkaaanlgsyees.citsswfarsomcotnh-e ssw1tpiecpte.mlhoipunXotttah1sitccsloemsrsithtuseedqseyu.r.asTrCtheohsenatttdrhiwaoeetltrswweecarocsenostmi~hpseton1esdcpeomdoufrohtiefhgdeah.ia7nrPttXriofiio7scrtigaatriloinddltieheostesf
Measurement ofsderites addition ofthe diet, black plastic o-rings (3/8 in OD, '/ in
ID) were placed in each cube in the potato sheer. After
Sclerites from the samples taken above were used to the diet had gelled, it was removed from the mold and
obtain measurements of sclerite sizes within colonies. attached to ropes to be placed out in the field. The diet
Measurements of maximum sclerite length were made was presented to fishes on 50 cm long pieces of3 strand,
from 100 to 200 sclerites from each sample using a video 'A in yellow polypropylene rope in which four 3-cm long
image analysis system. All ofthe sclerites in at least two safety pins were attached equidistantly along the top 30-
randomly selected fields from each ofthree subsamples 40 cm ofthe rope (Fig. 3).
from each sample were measured to the nearest jum at The ropeswere placed on the reefin pairsby attaching
magnifications ranging from 60X to 500X using a JAVA them to pieces of coral. All experiments were done on
computer program. Fingers Reefin Apra Harbor, Guam, at adepth of5-8 m.
Sclerite masses were estimated by using a log-log Pairs of ropes were placed on the reef and removed
regression of sclerite masses to lengths. To obtain the when at least four ofthe eight pieces ofartificial diet on
regression, 15-20 sclerites from each species were mea- the two ropes had been completely consumed. Each ex-
suredtothenearest^mwithavideoimageanalysissystem periment consisted of 17-21 pairs ofropes. Differences
and then weighed to the nearest 0.01 mg. The regression in consumption were compared with a Wilcoxon Signed
equations for each species were: S. maxima: mass(g) Ranks test (a = 0.05).
= e-2685+3.10-ln[length(^m)] (n = 2Q, r = 0.696, P< 0.05), S. Feeding experiments at Fingers Reefconsisted ofthe
p=ol0v.d8a0c2t,vlPa:<m0a.s0s5)(,g)5.=spe:-m|a8s8s0+(2g)15*='nlele-n2st0h0("0m+)2]41("nlnl=|i:nP20h(,''mr)2) fraotlelsowoinngcpoanitrreodlexdpieetrimveernstuss: d(i1e)tcocmopnatraiinsionngs oefitfheeerditnigp
(n = 15, r2 = 0.960, P < 0.05). sclerites orbase scleritesat natural concentrations (Table
I), (2) comparisons of feeding rates on control diet and
Extraction ofscleritesforfeeding assays dietcontainingeithertip orbase scleritesat 10%, 25%, or
50% bydryweight, and(3)comparisonsofdietcontaining
cieAspwperroexicmoaltleelctyed15frinodmivCioducaolss eLaacghooonf.3Tsiosfstuceosraaltstphee- tipscleritesatnaturalconcentrationswithdietcontaining
base sclerites at natural concentrations. Each ofthe ex-
bases ofthe soft corals were separated from the tips and
lobes (sensu Wylie and Paul, 1989). The polyp-bearing, periments described above was conducted with all three
top 10-15 cm ofthe colonies will hereafter be referred to species ofSinularia.
as the tips (Fig. 2). The pieces from all individuals ofa
species were combined and extracted twice in 1:1 meth- TableI
ylene chloride:methanol. The organic extracts were re-
moved, and the remaining soft coral pieces were dried, Concentrations(aspercentofdryweight) ofscleritesused
then digested in bleach until no organic material re- infeedingassays
mained. The sclerites were rinsed five to eight times in Tips Bases
tapwater, thenoncewithacetone, anddried. Examination
ofsclerites using light microscopy and scanning electron Sinn/aria maxima
microscopy showed no signs ofsclerite deterioration due
to the extraction process.
234 K. L. VAN ALSTYNE ET AL.
presented in each replicate (n = 19 or 20). Ropes were
removedandthenumberofpiecesofeachdietremaining
on each rope were counted after fishes had consumed at
least halfofthe sixteen pieces ofdiet in the set. Statistical
analyses were conducted with a two-way Kruskal-Wallis
test. Factors used in the analysis were sclerite type and
sclerite concentration.
Results
Distribution andabundance ofsclerites
Dry weight concentrations ofsclerites differed among
thethreespeciesofSinulariaandamonglocationswithin
thecolonies(Table II). In general, sclerite concentrations
(as%ofdryweight)increased fromthetipsofthecolonies
to their bases (Fig. 4). Mean concentrations at the tips of
the colonies ranged from 24% in S. polydactyla to 58%
in 5. sp. Mean concentrations of sclerites in the most
basal portion ofthe coloniesranged from 82% inS. max-
ima to 88% in S. sp.
The size distributions ofsclerites also showed consid-
erable intra-colony variation. Almost all ofthe sclerites
in the most distal tips ofthe colonies ofthe three species
mm
ofsoft coralswere less than 0.5 in length (Figs. 5-7).
However, inS. sp., the majorityofthe massofthesclerites
in the most distal tips were comprised of sclerites that
mm
were greater than 0.5 long; the largest fraction by
mm
masswas 1.0-1.5 long(Fig. 5). Inthe mostdistal tips
Figure 3. Rope used in a feeding assay. The assay consisted ofat- ofS. polydactyla, approximately 55% ofthe sclerite mass
staafcehtiyngpiann(Bo)-.riTnhge(Csa)feetmybpeidndsewderweitahtitnacahepdietcoeao0f.a5rtmifipciiaelcedioeft3(-Es)trtaonda wascomprised ofscleriteslessthan0.5 mmin length(Fig.
polypropylenerope(F)andwerebuoyedwith small piecesofneoprene 6); and, in S. maxima, all ofthe mamsms ofthe sclerites in
(D).Theropeswereattached in thefieldtopiecesofcoral(G). Ten-cm the most distal tipswas in the <0.5 category (Fig. 7).
piecesoflabelledsurveyor'stape(A)wereusedtodistinguishropescon- Larger sclerites increased in abundance in the more
tainingdifferenttypesofdiet. A singlerope held onlyonetypeofdiet. basal portionsofthecoloniesofallthreesoftcoral species
(Figs. 5-7). Even in the most basal samples, sclerites less
mm
than 0.5 long were numerically abundant; however,
Four-choicefeeding experiments
they made up only a small fraction ofthe total sclerite
To determine the relative effects of sclerite morphol-
ogiesandconcentrationson feedingdeterrence, fisheswere
offeredartificialdietscontainingtwotypesofsclerites(tip Table II
and base) at two different concentrations (tip concentra- Ana/vsisofvarianceofscleriteconcentrations inSinulariaspp.
tions and base concentrations). Thus, fishes were offered
tip sclerites at natural tip concentrations, tip sclerites at Source df SS MS
natural base concentrations, base sclerites at natural tip
concentrations, and base scleritesat natural baseconcen- Species
trations. Separate experiments were conducted for each
ofthe three Sinularia species.
Artificial diets were attached to polypropylene ropes
and offered to fishes in the field in the same manner as
described forthetwo-choiceexperiments, with theexcep-
tion that a set offour ropes were offered instead ofa pair
ofropes. Each ofthe fourropescontained one ofthe four
diets described above; each ofthe four types ofdiet was
STRUCTURAL DEFENSES IN SOFT CORALS 235
Sclerite Concentrations significantly (P< 0.05) deterrent to fishesat Fingers Reef
100 (Fig. 8). Theaddition ofsclerites from thetipsofSinularia
spp. reduced feeding by 33-44%, whereas sclerites from
80 the bases ofthe colonies reduced feeding by 27-33%.
60J SINULARIA SP.
40
Oo
20-
Tips Base
S max sclentes S poly sdemes-*-S sp sclenies
Figure4. Scleriteconcentrations(as%ofdryweight)withincolonies
ofSinn/aria spp. Samples were obtained from five colonies ofeach of
thethreespecies.Thesamplescorrespondwithlocations 1 (tips)through
6(base)shown in Figure 2. Horizontal barsrepresent 1 SD.
mass. In the most basal samples, the largest fractions of
mm
sclentes by mass were in the 1.0-1.5 size class in S.
maxima (Fig. 7) and in the 1.5-2.0 mm size class in S.
polydactyla (Fig. 6) and S. sp. (Fig. 5). Thus, in all three
species ofSinularia examined in this study, sclentes in-
creased in both size and concentration from the tips of
the colonies to the bases.
Sclerites in S. maxima ranged in length from 0.06 to
5.0 mm. The largestsclentesin S. maximawereelongated
spindleswith complex tubercles(sensu Bayeretal., 1983).
mm
The small sclerites (<0.5 long) were rods with vol-
cano-like processes and wart clubs (sensu Bayer et al.,
1983). Wart clubs were proportionately more abundant
in thetipsthanthebasesofS. maxima. Thelargesclerites
mm
(>0.5 long) of S. polydactyla were comprised pri-
marily of spindles with complex tubercles. Most of the
spindleswerestraight; however,afewwerebent. Thesmall
scleriteswereeitherthorn clubsorrodswith volcano-like
processes. Sclerites from S. polydactyla ranged in length
from 0.08 to 5.0 mm. In S. sp., the larger sclerites were
comprised ofspindleswith complex tubercles, thorn stars,
and thorn scales. The spindles were frequently bent and
occasionally bifurcated on one end. The tubercles on the
spindles ofS. sp. were smaller than those on 5. maxima
and S. polydactyla. The smaller sclerites were rods and
thorn clubs. Sclerites from 5. sp. ranged in length from
0.08 to 4.6 mm.
Effectiveness ofsclerites asfeeding deterrents
At natural concentrations, all ofthe sclerites from the
bases and the tips ofall three species of Sinularia were
236 K. L. VAN ALSTYNE ET AL.
SINULARIA POLYDACTYLA
STRUCTURAL DEFENSES IN SOFT CORALS 237
100 1.4
80
60-
c 40-
o
20-
10 30 39 SO
Tips Base Tips Base Tips Base [Sclerites](%d.w.)
S.m^axHimCaontrol S.polydactSycllaentesAddedS.sp. S. maxima S. polydactyla S.sp.
| | Figure 10. FeedingdeterrenceofSinulariatipscleritestowardsgen-
Figure 8. Mean amount consumed by generalist fishes at Fingers eralistcarnivorousfishesatdifferentscleriteconcentrations. Pointsrep-
Reefofartificialdietwithscleritesfrom Sinn/ariaspp. (whitebars)and resent the mean feeding rate on artificial diet with sclerites relative to
withoutsclentes(darkbars). Numbersinparenthesesatthebasesofthe controldietlackingsclerites.Eachvaluerepresentstheresultsofasingle
barsarePvaluesfromWilcoxonSignedRankstestsforpairedcomparisons. two-waychoiceexperiment.Circledpointsdenoteexperimentsinwhich
scleritesweresignificantlydeterrent(P<0.05).
to artificial diet lacking sclerites (Fig. 9); artificial diets
containingtipscleritesfrom S. maximaandS. sp. at 10% greater than 25% by dry weight; S. sp. base sclerites were
by dry weight were consumed at a greater rate than the notsignificantlydeterrent(P>0.05)at 50% bydryweight,
control diets (Fig. 10). The relative consumption rate of but were deterrent at 82% by dry weight.
artificial dietswith scleritesrelativetocontrol dietslacking
scleritesdecreased with increasing sclerite concentration. Intracolony differences in sclerite effectiveness
All ofthe sclerites except those from the bases of5. sp.
were significantly deterrent (P < 0.05) at concentrations When directly tested in paired feeding experiments,
sclerites from thebases ofthecolonieswere moreeffective
feeding deterrents than sclerites from the tips ofthe col-
BASES onies (Fig. 1 1). The amount offeeding on artificial diets
SinulanaSclentes
100
20 40 60 80 100
[Sclerites](%d.w.)
S. maxima S polydactyla S.sp. S maxima S polydactyla S.sp.
m
Figure9. FeedingdeterrenceofSinulariabasesclentestowardsgen- ITips Base
eralistcarnivorousfishesatdifferentsclenteconcentrations. Pointsrep-
resent the mean feeding rate on artificial diet with sclerites relative to Figure11. Meanamountconsumedbygeneralistcarnivorousfishes
controldietlackingsclerites. Eachvaluerepresentstheresultsofasingle ofartificialdietcontainingtip(darkbars)andbase(whitebars)sclerites
two-waychoiceexperiment.Circledpointsdenoteexperimentsinwhich in paired feedingexperiments(n = 19 to 20). Numbers in parentheses
scleritesweresignificantlydeterrent(P<0.05). arePvaluesfrom Wilcoxon Signed Rankstestsforpairedcomparisons.
238 K. L. VAN ALSTYNE ET AL.
containing sclerites from the bases of the colonies was 100 Sclerites
18% (in S- sp.) to 51% (in S. maxima) lower than the 1
consumption ofdiets containing tip sclerites. 80
Because both the morphologies and concentrations of I
sclerites differbetween the bases and tips ofthe colonies,
E 60
differencesinthedeterrenceoftipandbasescleritescould
result fromdifferences in thesizesand shapesofthescler-
ites or from differences in their amounts. To sort out the 40
relativecontributionsofsclerite morphologyandconcen-
tration to predator deterrence, experiments were con- 20
ducted in which fishes were offered four choices offoods
simultaneously: diet containing tip sclerites at tip con-
centrations, diet containing tip sclerites at base concen- S. maxima S. polydactyla S.sp.
trations, diet containing base sclerites at tip concentra-
tions, and diet containing base sclerites at base concen- <jj Tips. Low | | Tips, High [ | Base, Low jggj Base, High
trations. Theseexperimentsdemonstrated thattheability
Figure 12. Results offour-way feeding experiments. Vertical bars
ofscleritesto deter feedingwas dependent upon both the represent mean feeding rateson tiporbasescleritesat natural tip(low)
morphology of the sclerites and their concentrations; or natural base (high) concentrations. Vertical lines represent 1 SE.
however, concentration is a more important factor than Statistical analysesare presented in Table III.
morphology in Sinularia spp. (Table III). Feeding rates
were higher on artificial diets containing sclerites at the
lowertip concentrations than the higher base concentra- 1988) and Gorgonia ventalina (Van Alstyne and Paul, in
tions, regardless ofsclerite type (Fig. 12). In S. maxima, press) have also demonstrated that sclerites can deter
the smallertip sclerites were less effective deterrentsthan feedingin the field. However, notall alcyonarian sclerites
the larger base sclerites at high concentrations, but were are deterrent at natural concentrations. For example,
poonllyydascltiyglhatl,ytliepssscelfefreictteisveweartelmoowrceonecfefnetctriavteiodnest.erIrnentS.s astclcerointceesnftrroamtitohneswohfit~e4w5hi%p,bJyitdurcyeewleaisgph.t,,naatsuirmailllyaroccocnu-r
than base sclerites at both high and low concentrations. centration to those found in Sinularia: but, Junceela
In S. sp., tip and base sclerites were equally effective at sclerites are not deterrent towards carnivorous fishes at
low concentrations; however, at high concentrations, tip naturaldryweightconcentrations(Pauland VanAlstyne,
sclerites were less effective deterrents. in prep.).
Invertebrate spicules and sclerites vary widely in size,
shape, andconcentration. Wehavedemonstratedthat, in
Discussion
these three species of Sinularia, concentration is more
This and other studies have clearly demonstrated the important than morphology in determiningthe ability of
potential foradefensiveroleforalcyonarian sclerites. We sclerites to deter the generalist predators ofGuam. Sin-
have shown that sclerites from three species ofSinularia ularia sclerites were increasingly effective as deterrents
in Guam, when incorporated into an artificial diet, re- until concentrations ofabout 30-50% by dry weight, at
duced feeding by generalist carnivorous fishes in the field which point deterrence leveled off(Figs. 9, 10). The level
(Fig. 4). Similar studies with sclerites from the Caribbean at which maximum effectiveness was reached was deter-
gorgonians Pseudoptcrogorgia acerosa (Harvell et at.. mined by the sizes and shapes ofthe sclerites. The differ-
Table III
Resultsof'hw-mn- Kruskal-Wallistestsondatafromfourchoicefeedingexperiments
Location ofsclerites Concentration ofsclerites Location* concentration
Species
Sinularia maxima
Dataarepresented in Figure 12.
STRUCTURAL DEFENSES IN SOFT CORALS 239
ences in deterrence ofsclerites from the tips and bases of Paul, 1989; Van Alstyne et at., in prep.) supports the hy-
thecolonies (Fig. 1 1) was primarily a result ofconcentra- pothesisthatscleriteconcentrationsinthetipsofcolonies
tion differences of sclerites in these two regions (Table are under functional constraints.
III), not ofmorphological differencesbetween tipandbase
sclerites. Acknowledgments
Although differences in thesizesand shapesofsclerites
between thetipsand the basesofthecolonies had a lesser The authors are grateful to D. Carandang-Liberty, K.
impact on the relative effectiveness ofstructural defenses Foster, K. Kuetzing, H. Sanger, and K. Sonada for the
than scleriteconcentration, sclerite morphology isstill an many hours spent attaching o-rings onto safety pins. We
importantdeterminant offeedingdeterrence. Differences are also grateful to these individuals and the attendees of
in sclerite morphologies between the tips and bases of5. the UOG POETS club for spending many more hours
polydactylacoloniessignificantly affected feedingby fishes taking the o-rings offsafety pins. We are indebted to D.
offGuam (Table III). Two components ofsclerite mor- Carandang-Liberty, K. Foster, B. Irish, L. Meyer, H. San-
phology may influence their effectiveness as feeding de- ger, and K. Sonada for their assistance with the field ex-
terrents: sizeandshape. The effectsofsclerite morphology periments. We also thank S. Murray of the California
may be more important when making interspecific com- State University at Fullerton foruse ofhis image analysis
parisons of deterrence. For example, sclerites from system,J. SmithofCSUFforadviceonstatisticalanalyses,
Junceelasp. do notdeterfeedingby fishesat Fingers Reef and A.O.D. Willows and the staffofthe Friday Harbor
(Paul and Van Alstyne, in prep.). TheseJunceela sclerites Laboratories forthe use ofthe scanning electron micros-
are similar in size and concentration to many of the copy facilities. This manuscript greatly benefitted from
smaller Sinularia sclerites, but, they differ in shape (Paul thecommentsoftwoanonymous reviewers. This research
and Van Alstyne, in prep.). Sclerite shape probably has a wasfundedbyNIHgrantGM 38624 to VJPandaGuyer
significant effect on function, particularly in determining postdoctoral fellowship from theUniversity ofWisconsin
the ability ofthe sclerite to deter potential predators. to KLV. This is contribution number 31 1 ofthe Univer-
Although our experiments have demonstrated that sity ofGuam Marine Laboratory.
Sinularia sclerites can deter feeding by fishes, predator
deterrence may not be the primary function of these Literature Cited
structures. Spicules from marine invertebrates also play
a role in structural support for colonies, increasing the Bayer, F. M. 1956. Octocorallia. Pp. F166-F231 in Treatise on In-
stiffness of connective tissues by acting like reinforcing vertebratePaleontology, PanF, Coelenterata, R.C. Moore,ed.Uni-
versityofKansasPress, Lawrence, KS.
fibers(Koehl, 1982). Smallscleritestendtoincreasecolony Bayer,F.M. 1961. TheshallowwaterOctocoralliaoftheWestIndian
stiffness more than large sclerites; however, stiffness, like region.StudiesontheFaunaofCuracaoandOtherCaribbeanIslands
predator deterrence, increases with increasing sclerite 12: 1-373.
concentration (Koehl, 1982). Further work is needed to Bayer, F. M., M.Grasshoff, and J. Verseveldt. 1983. Illustrated Tri-
clarify the structural and defensive functions ofinverte- lingualGlossaryofMorphologicalandAnatomicalTermsAppliedto
Octocorallia. E. J. Brill. Leiden. 75 pp.
brate spicules. Coll, J. C., D. M. Tapiolas, B. F. Bowden, L. Webb, and H. Marsh.
The differences in sclerite morphologies and concen- 1983. Transformation of soft coral (Coelenterata: Octocorallia)
trations within Sinularia colonies may reflect differences terpenesby Ovula ovum (Mollusca: Prosobranchia). Mar Biol. 74:
in the function of sclerites in different locations within 35-40.
Gawel, M. J. 1977. The common shallow-water soft corals (Al-
the colony ordifferences in constraints upon sclerite use. cyonacea) ofGuam. M.Sc. Thesis, University ofGuam. Agana.
Scleritesare mostconcentrated in the basesofthecolonies Guam. 201 pp.
(Fig. 4), making the bases less susceptible to attack than Harvell, C. D., and W. Fenical. 1989. Chemical and structural
the tips (Fig. 12, 13, Table III). However, it is the tips of defensesofCaribbean gorgonians(Pseudopterogorgia spp.):in-
the colonies that receive the majority ofattacks by pred- tracolony localization ofdefense. Limnol. Oceanogr. 34: 380-
asctloerrsit(eWsylinietahendcoPlauoln,y 1t9i8p9s).maTyherelfalcekctofcohnisgthralienvteslsoonf Harvs3et8lr7lu.,ctuCr.alD.d,efWen.seFsenoifcCaal,riabnbdeaCn.goHr.goGnrieaennse.(Ps1e9u8d8o.pteCrhoegomrigciaalspapn.d)
sclerite use. The presence oflarge quantities ofsclerites I:Developmentofan //;situfeedingassay.Mar Ecol. Prog. Ser. 49:
may interfere with the functioningofthesoftcoral polyps, 287-294.
which are found only in the tips of the colonies. Alter- Harvell,C.D.,andT.H.Suchanek. 1987. Partialpredationontropical
natively, sclerites in the bases ofthe colonies may serve gorgonians by Cyphoma gibbosum (Gastropoda). Mar. Ecol. Prog.
Ser 38: 37-44.
primarily as structural support rather than predator de- Kingsley,R.J. 1984. Spiculeformationintheinvertebrateswithspecial
terrents. Thepresenceofhighconcentrationsofpredator- referencetothegorgonianLeptogorgiavirgiilata.Am.Zool.24:883-
deterrent extracts in the tips ofthe colonies (Wylie and 891.
240 K. L. VAN ALSTYNE ET AL.
Kingsley, R. J., M. Tsuzaki, N. Watabe, and G. L. Mechanic. Tursch, B.,andA.Tursch. 1982. Thesoftcoralcommunityonashel-
1990. Collagen inthespiculeorganicmatrixofthegorgonianLep- tered reefquadrat at LaingIsland (Papua NewGuinea). Mar. Biol.
togorgia virgiilat/t. Biol. Bull. 179: 207-213. 68:321-332.
Koehl, M. A. R. J982. Mechanical design ofspicule-reinforced con- Van Alstyne, K. L.,andV.J. Paul. 1992. Chemical andstructural
nectivetissue.J. Exp. Biol 98: 239-268. antipredatordeterrentsin thesea fan Gorgonia ventalina:effects
LaBarre,S.,J.C.Coll.andP.W.Sammarco. 1986. Defensivestrategies against generalist and specialist predators. Coral Reefs In
ofsoftcorals(Coelenterata: Octocorallia)oftheGreat BarrierReef. press.
II: The relationship between toxicity and feeding deterrence. Biol. Verseveldt,J. 1977. Octocoralliafrom variouslocalitiesinthePacific
Bull. 171:625-636. Ocean. Zoo/. Verhand. Leiden 150: 1-52.
Paul,V.J.,andK.L.VanAlstyne.1987. Chemicaldefenseandchem- Verseveldt,J. 1980. A revisionofthegenusSinularia May(Octocor-
ical variation in some tropical Pacific species ofHalimeda (Hali- allia, Alcyonacea). Zoo/. Verh. Leiden 178: 1-165.
medaceae;Chlorophyta). CoralReefs6: 263-269. Wylie, C. R. 1988. Chemical defenses in three species ofSinularia
Pawlik,J.R.,M.T.Burch,andW.Fenical. 1987. Patternsofchemical (Coelenterata, Alcyonacea): effects against the predator Cheatodon
defenseamongCaribbeangorgoniancorals:apreliminarysurvey.J. unimaculalus (Perciformes). M.Sc. Thesis, University of Guam,
Exp. Mar. Biol. Ecol. 108: 55-66. Mangilao, Guam. 38 pp.
Sammarco,P.W.,S.LaBarre,andJ.C.Coll.1987. Defensivestrategies Wylie,C. R.,andV.J.Paul. 1989. Chemicaldefensesinthreespecies
ofsoftcorals(Coelenterata: Octocorallia)oftheGreat Barrier Reef: of Sinularia (Coelenterata, Alcyonacea): effects against generalist
III. The relationship between icthyotoxicity and morphology. Oec- predators and the butterflyfish Chaetodon unimaculalus Bloch. J.
0/ogza74:93-101. Exp. Mar. Biol. Ecol. 129: 141-160.