Induction of defenses and within-alga variation of palatability in two brown algae from the northern

14 Pages·2005·0.17 MB·English

JournalofExperimentalMarineBiologyandEcology325(2005)214–227
www.elsevier.com/locate/jembe
Induction of defenses and within-alga variation of palatability in
two brown algae from the northern-central coast of Chile:
Effects of mesograzers and UV radiation
Erasmo C. Macayaa, Eva Rotha¨uslerb, Martin Thiela,*,
Markus Molisc,1, Martin Wahlc
aFacultadCienciasdelMar,UniversidadCato´licadelNorte,Larrondo1281,Coquimbo,Chile
bInstitutfu¨rAquatischeO¨kologie,Albert-Einstein-Str.3,18057Rostock,Germany
cLeibnizInstituteofMarineSciences,MarineEcology–MarineZoology,Du¨sternbrookerWeg20,24105Kiel,Germany
Received10November2004;receivedinrevisedform5May2005;accepted9May2005
Abstract
Macroalgaepossessdifferentdefensemechanismsinresponsetoherbivory.Somespeciesproduceanti-herbivoresecondary
metabolites, but production of these substances can be costly. Therefore, algae may produce defensive metabolites only in
response to herbivory (inducible defense) or defend particular parts of the alga differentially (within-alga variation). In the
present study, we examined whether two species of brown algae from the SE-Pacific show evidence of inducible chemical
defense(non-polar compounds)or within-alga variation of defense, whichwe estimated in form of palatability of differently
treated algae to amphipod grazers (with live algae and agar-based food containing non-polar algal extracts). In Glossophora
kunthii(C.Agardh)J.Agardh,weobservedanincreaseinpalatabilityafteralgaewereacclimatedfor12dayswithoutgrazers.
Subsequent addition of grazers for 12 days then resulted in a reduction of palatability indicating the existence of inducible
defense.Afterremovalofgrazersfor12days,theseinducedeffectsagaindisappeared.ThereactionofG.kunthiiwastriggered
evenbythe mere presenceof grazers, whichsuggeststhat thisalga canrespondtowaterborne cuesbyreducingpalatability.
Effectswereonlyfoundforagar-basedfoodcontainingnon-polarextracts,butnotforlivealgae,suggestingthatsomepartsof
thealgaeareundefended.Oursecondexperimentonwithin-algavariationconfirmedthatonlyapical(growthregion)andbasal
parts(neartheholdfastregion)ofG.kunthii aredefendedagainstherbivores.Forthesecondspecies,Macrocystisintegrifolia
Bory,thefirstexperimentrevealednoinductionofdefense,whilethesecondexperimentonwithin-algavariationshowedthat
amphipods avoided basal parts and in particular stipes of M. integrifolia but only in live algae. Although both studied algal
species differed substantially in their defensive strategies, their reaction was independent of the presence or absence of UV
* Correspondingauthor.Tel.:[phone];fax:[phone].
E-mailaddress: [email protected](M.Thiel).
1 Marine Biological Station Helgoland, Foundation Alfred Wegener Institute for Polar and Marine Research, Kurpromenade 201, 27498
Helgoland,Germany.
0022-0981/$-seefrontmatterD2005ElsevierB.V.Allrightsreserved.
doi:10.1016/j.jembe.2005.05.004
E.C.Macayaetal./J.Exp.Mar.Biol.Ecol.325(2005)214–227 215
radiation.Thus,itappearsthatUVeffectsplayonlyaminorroleinanti-herbivoredefense,whichisinaccordancewithmost
previous studies.
D2005Elsevier B.V. All rightsreserved.
Keywords:Chile;Defenses;Herbivory;Induction;Macroalgae;UVradiation
1. Introduction 1979). Differential distribution of defensive com-
pounds has also been reported for some kelp species
Herbivory influences marine benthic community (Steinberg, 1984; Tugwell and Branch, 1989), sipho-
structure (Lubchenco and Gaines, 1981) and can be nous green seaweeds (Hay et al., 1988; Paul and Van
intense in both tropical (Carpenter, 1986; Hay, 1997) Alstyne, 1988), antarctic species of Desmarestia
andtemperatehabitats(BreenandMann,1976;Schei- (Fairhead et al., 2005) and rockweeds (Tuomi et al.,
blingetal.,1999;Gagnonetal.,2004).Inresponseto 1989;Paviaetal.,2002;Tothetal.,2005).Croninand
herbivory, macroalgae have evolved a variety of Hay (1996b) reported differences in palatability be-
mechanisms including escape in time or space (Lub- tween different parts of the brown algae Dictyota
chenco and Gaines, 1981), tolerating (Hay and Feni- ciliolata. Within-alga variation should be expected
cal, 1988) or deterring herbivores (e.g., Hay and to be most pronounced in species with different
Fenical, 1988). Several modes are known by which degrees of differentiation and translocation of nutri-
algae deter consumers (Cronin, 2001), of which mor- ents. Large kelps such as, for example, Macrocystis
phological and chemical traits in anti-herbivore spp. have internal transport systems (Raven, 2003)
defenses are the most commonly employed mechan- andhighdegreeoftissuedifferentiationwithstructur-
isms (Hay, 1996; Cronin, 2001). altissuessuchastheholdfastandstipes,ononehand
Although difficult to assess, the production of and, on the other hand, blade regions that are mainly
defenses might be costly (Hay and Fenical, 1988; active in nutrient acquisition and photosynthesis
Cronin, 2001) if defenses use resources that could (North, 1994). While blades of these large kelps are
havebeenallocatedtogrowthorreproduction(Cronin readilyconsumedbygrazers andareconsideredtobe
andHay,1996a;Hay,1997;DuffyandHay,2001).In largely undefended (Steinberg, 1985; Winter and
this sense, constitutive, i.e., permanent, defenses re- Estes, 1992), there exists evidence that structural tis-
quire expenditure of resources even when consumers sues contain higher concentrations of chemical com-
are absent and the benefits of protection are not poundsthanothertissues(TugwellandBranch,1989;
realized (Cronin and Hay, 1996a). In contrast, induc- Van Alstyne et al., 1999). Fairhead et al. (2005) also
ible defenses allow costs of defenses to be deferred revealed recently that the primary stem of the brown
until enemies have been detected, at which time the algae Desmarestia anceps resisted amphipod grazing
costs can be offset by the benefits of protection (Har- by toughness and deterrent chemistry.
vell, 1990; Cronin and Hay, 1996a). At present, inducible chemical defenses have been
Besides temporal variation in defense levels, de- documentedforover100terrestrialplantspecies(Kar-
fensive traits are assumed to vary within plants (Zan- banandBaldwin,1997),butthereareonlyfewreports
gerl and Rutledge, 1996). According to optimal on herbivore-induced defenses in seaweeds (e.g., Van
defense theory, valuable tissues (e.g., growth meris- Alstyne, 1988; Cronin and Hay, 1996a; Pavia and
tems)athighriskofgrazerattacksshouldbeprotected Toth, 2000; Sotka et al., 2002; Rohde et al., 2004;
by strong constitutive defenses, while tissues that are Weidneretal.,2004).Inasuiteofstudies,algaewere
lessvaluableorlesslikelytobeattackedshouldhave artificially wounded (dclippingT) instead of using nat-
lowerlevelsofconstitutivedefensesand/orbecapable ural herbivores (e.g., Pavia et al., 1997). Lack of
of inducing defenses following attack. First evidence herbivore-specific cues, e.g., grazer-associated micro-
of within-plant variation in concentrations of second- organims, are seen as one explanation for missing
ary metabolites came from terrestrial plants (McKey, effects in assessing the induction of defenses in most
216 E.C.Macayaetal./J.Exp.Mar.Biol.Ecol.325(2005)214–227
clippingexperiments(Hay,1996andreferencesthere- reduced when grazing ceased, (4) whether UV radia-
in). Experiments exposing natural grazers to algae tionaltersmacroalgalresponsesinthiscontextand(5)
revealed exclusively inducible defenses for brown if there exists within-alga variation in palatability.
seaweeds. Furthermore, waterborne cues released by
directly grazed Ascophyllum nodosum individuals in-
duced chemical defenses in ungrazed adjacent con- 2. Materials and methods
specifics(TothandPavia,2000).However,itremains
unclearwhetherwaterbornechemicalsoriginatedfrom 2.1. Sampling sites of algae and grazers
the grazer, the algae, or both.
Algal secondary metabolites also have other func- WecollectedbothalgaeinthevicinityofCoquimbo
tions besides deterrence of herbivores. These com- (Chile), Glossophora kunthii from the intertidal zone
pounds can inhibit settlement or development of of Totoralillo (30803VS; 70838VW) and Macrocystis
foulingorganisms(Schmittetal.,1995,1998),micro- integrifolia from subtidal habitats of Islas Damas
bial films (Hay, 1996) and furthermore provide pro- (29813VS; 71831VW). For all experiments, the amphi-
tection against damaging effects of UV radiation pod Parhyalella ruffoi (Lazo-Wasem and Gable) was
(Pavia et al., 1997; Stachowicz and Lindquist, 1997; used as grazer after confirming in preliminary studies
SwansonandDruehl,2002).Paviaetal.(1997)found thatthisamphipodconsumedawidevarietyofmacro-
that the brown seaweed Ascophyllum nodosum in- algae, including G. kunthii and M. integrifolia
creased the concentration of phlorotannins under en- (0.012F0.005 g ind(cid:1)1 day(cid:1)1 and 0.041F0.020 g
hanced UV-B levels. On the other hand, Cronin and ind(cid:1)1 day(cid:1)1, respectively). All amphipods were only
Hay(1996c)suggestedthatinDictyotaciliolata,UV- used once, either in grazing treatments or in feeding
stress may reduce chemical defenses and increase assays.
algal susceptibility to herbivores. Thus, UVradiation
may also affect alga–herbivore interactions by influ- 2.2. Experiment I: effects of grazing, grazer presence
encing concentrations of secondary metabolites, and and UVR on palatability
given globally changing UV-regimes, it appears im-
portant to include this factor in studies testing induc- 2.2.1. Experimental design and set-up
ible defenses of shallow-water algae. We designed two three-phased, two-factorial
Finally,moststudiesoninducibledefensesfocused experiments to test the effects of amphipod grazing
only on factors that initiate anti-herbivore responses. (4 levels, fixed), UV radiation (2 levels, fixed) and
However, if induction of anti-herbivore defenses is a block (4 levels, random) on palatability levels of the
mechanismtoreducecosts,thendefensesshouldalso brown algae Glossophora kunthii (entire individuals)
decreasewhengrazingpressurediminishes.Neverthe- and Macrocystis integrifolia (apical tissues), respec-
less, to date, experimental evidence on testing the tively. Using a randomized block design, each treat-
decrease of anti-herbivore defense in absence (or at ment combination of the fully crossed fixed factors
lowlevels)ofgrazingisnotavailable(forexceptions, was replicated 5 times, where each combination was
see Hemmi et al., 2004; Rohde et al., 2004; Weidner represented once within each of the four blocks, and
et al., 2004; Ceh et al., in press). the fifth replicate was assigned randomly to one of
Thepresentstudyassessed(i)theeffectofgrazing these four blocks.
on inducible defenses and (ii) the palatability of dif- Induction experiments were set-up in an outdoor
ferent algal parts of two common brown macroalgae, flow-through aquaria system. Filtered seawater (10
Glossophora kunthii and Macrocystis integrifolia, Am cotton cartridge) was pumped from the shallow
from the northern-central coast of Chile. The specific subtidal of Bah´ıa La Herradura (29858V S; 71821V W)
aimsofthisstudyweretotest(1) whethergrazingby into four plastic reservoirs (70 l), supplying each EU
amphipods can induce defenses in both species of (=transparent plastic aquarium, 10(cid:2)19(cid:2)13 cm, 1.5-
macroalgae, (2) whether waterborne cues from adja- l volume) via flow-regulated hoses individually at a
cent grazed conspecifics or from non-feeding amphi- rate of 0.1 l h(cid:1)1 with seawater. The EUs were dis-
pods can induce defenses, (3) whether defense is tributed among 4 tables (=blocks) that were covered
E.C.Macayaetal./J.Exp.Mar.Biol.Ecol.325(2005)214–227 217
with plastic shading cloth toprotectalgae from direct adding 15 amphipods to the upstream compartment
sunlight.Duringthelateafternoon,theaverage(FSD) togetherwithonenon-targetpieceofthesamespecies
water temperatures in the aquaria increased to 17.57 as the target alga, (3) waterborne cues from non-
8C(F0.48),slightlyaboveambientwatertemperature grazing amphipods by introducing 15 amphipods
in Bah´ıa La Herradura (16.72 8CF0.67), but well without alga in the upstream compartment, and (4)
within the normal growth range of G. kunthii (Hoff- controls where neither amphipods nor non-target
mann and Malbran, 1989) and Macrocystis (North, pieces of algae were introduced. At the end of the
1994). treatment phase, all amphipods and non-target algal
To restrict movements of grazers (P. ruffoi), each pieces were removed from the aquaria to give algal
EU was divided midway into an up- and a down- pieces during the recovery phase a chance to lower
stream compartment by a rigid green plastic mesh (1 any anti-herbivore defensespreviously induced in the
mm mesh size). Prior to the experiments, all algae treatment phase. Using cut-off filters, we tested for
were carefully cleaned from grazers and epiphytes, UV radiation (UVR) effects by manipulating irradi-
dried with absorbent paper, and weighed. Between ance regimes during the treatment and the recovery
January and March 2003, we ran the first induction phase.Toblockradiationb400nmcompletely(PAR),
experiment using the tips (top 25 cm) of M. integri- half of all EUs were covered with 4 mm thick sheets
folia, whereas entire individuals of G. kunthii were of Makrolon (long life plus 293, Ro¨hm, Germany),
used in the second induction experiment, conducted while the other half of the EUs were covered with 3
between March and April 2003. Both experiments mm thick Perspex sheets (GS 2648 Ro¨hm, Germany)
were separated into an acclimation, a treatment, and allowing complete transmission of ambient solar irra-
arecoveryphase.Eachphaselasted12or14daysfor diance (PAR+UV) (see Molis and Wahl, 2004 for
G. kunthii and M. integrifolia, respectively. The ob- details on optical properties of cut-off filters). Proce-
jective of the acclimation phase was to reduce the dural controls from unfiltered EUs were compared
probability of past consumption history to decrease with algae from the Perspex-covered EUs. In most
variation among replicates. At the beginning of the comparisons, amphipod consumption rates were not
acclimationphase, we storedfive algalpiecesat(cid:1)40 significantly different between Perspex-filtered and
8Ctoconservetheirnaturaldefenselevels.Moreover, unfiltered individuals of both brown algae, meaning
4 algal pieces (=target algae, used later in feeding that no filter artifacts were found herein, with the
assays) were placed in the downstream compartment single exception of Macrocystis integrifolia in the
of each EU, to pass through the acclimation phase. treatment phase (paired t-test: t=3.22, p=0.032).
This was also done with 5 additional algal pieces,
placed separately in five aquaria, which were then 2.3. Experiment II: within-alga variation in
used to assess the acclimated state of defense at the palatability
end of the acclimation phase after transferring them
for 48 h to (cid:1)40 8C. Subsequently, the palatability of 2.3.1. Experimental design and set-up
agar-basedfoodcontainingextractsofacclimatedand Ten individuals of G. kunthii and eight of M.
non-acclimatedalgaewascomparedinfeedingassays integrifolia were collected in October 2003 in order
(see subsection Feeding assays). Both acclimated and to examine whether amphipod grazers preferentially
non-acclimated algae werefrozen priortotheirusein consumedifferentpartsofthealgae.Apical,medium,
feeding assays to avoid confounding effects of freez- and basal portions of G. kunthii, and apical, medium,
ing non-acclimated algae alone. At the beginning of basalblades,andstipeportionsofM.integrifoliawere
the following treatment phase, algae were exposed to offered to the amphipod P. ruffoi. Individuals of G.
different combinations of UV radiation and grazing kunthii weredissectedintotheapicalparts(thetop2–
treatments to allow them to produce anti-herbivore 3cm)withthelargeapicalcell,themediumpartswith
defenses. Grazing treatments tested the effects of (1) a dense cover of ligulae (causing an irregular surface
directgrazerexposurebyadding15amphipodstothe texture—Hoffman and Santelices, 1997), and the
target algae in the downstream compartment, (2) wa- basal parts immediately above the small holdfasts.
terborne cues from nearby grazed conspecifics by For the morphologically more differentiated alga M.
218 E.C.Macayaetal./J.Exp.Mar.Biol.Ecol.325(2005)214–227
integrifolia, we dissected blades from the apical re- used in the first experiment, because it is considered
gion (the top 3–5 cm), blades from the medium part unlikely that amphipods in their natural habitat chose
(middle of the alga), basal blades without sporangial among several algal individuals before starting to
tissues (the blades that are located at the base of the feed. This will in particular be true for P. ruffoi,
alga)andthestipes directlyabovetheholdfasts.Sim- which lives in association with drift algae (personal
ilar amounts of pieces from the respective algal observation). However, P. ruffoi probably chose
regions were used in choice and no-choice feeding among different tissue types within one algal individ-
assays (see next subsection). ual before a meal, which is why in the second exper-
iment, we conducted choice assays in addition to no-
2.4. Feeding assays choice assayswhen assessingwithin-algavariationof
palatability. Choice and no-choice feeding assays
Theaimoffeedingassayswastoexaminewhether were conducted in Petri dishes (30 ml) using 6 and
anti-herbivore defenses were induced and reduced in 4 amphipods per assay, respectively.
theinductionexperimentsbyassessingandcomparing Asresponsevariableforassessmentofpalatability,
palatability levels separately for live algae and agar- we used for live algae wet mass of consumed algal
based food containing non-polar algal extracts after tissue [g 3 day(cid:1)1] and the number of consumed
each phase of the induction experiment. Significantly squares for agar-based food (see next subsection for
higherconsumptionratesoftheamphipodP.ruffoi on details).Wetmassofeachalgawasdeterminedtothe
live control algae compared to treated algae would nearest milligrams, using an analytical balance (Den-
indicateanincreaseinmorphologicaland/orchemical verInstrument100AF0.2mg),afterblottingthealga
defenses, while the same result for agar-based food for30swithabsorbentpapertoremoveexcesswater.
confines anti-herbivore defenses to non-polar algal Inordertoaccountfornon-grazingrelatedchangesin
chemistry. live algae, we calculated real consumption according
All feeding assays lasted 3 days and were con- to Cronin and Hay (1996a) as C =T (C/C)(cid:1)T
real i i f f
ducted in a culture room under constant temperature where T and T correspond to initial and final wet
i f
(15 8CF18), using a 12-h photoperiod of 40F10 massofthealgausedinthefeedingassay,respective-
Amol m(cid:1)2 s(cid:1)1 (fluorescent lamp, 40 W, Phillips, ly,andC andC toinitial andfinalmassofacontrol
i f
Brazil). Feeding assays with live algae were per- algae not used in the feeding assay, respectively.
formedintransparent(volume:1l)plasticcontainers,
while those with agar-based food containing non- 2.5. Production of agar-based food containing
polar algal extracts used Petri dishes (volume: 30 non-polar extracts
ml). In both types of assays, water was exchanged
twice daily. Prior to feeding assays, amphipods were Single algal pieces, stemming from the induction
offered a diverse algal diet, because starvation may experiments,wereplacedfor48hindichloromethane
alterthefeedingbehaviorofsomeherbivores(Cronin (DCM)afterblottingthealgadry.Weused2mlDCM
and Hay, 1996c). per 1 g algal wet mass. DCM extracts the non-polar
After the acclimation phase, one piece of agar- algal chemistry, which had been shown for several
based food containing non-polar extracts of acclimat- algae species to possess effective anti-herbivore com-
ed and one piece of agar-based food from non-accli- pounds (Steinberg and Van Altena, 1992; Paul et al.,
mated algae were simultaneously incubated with 6 2001; Steinberg et al., 2001). Herein, we focused on
amphipods in choice feeding assays. Miniscule inci- the non-polar compounds because Macrocystis is
sions were cut into food items to distinguish treat- known to be poor in phenolic compounds (Steinberg,
ments. At the end of both, the treatment and the 1985), and the strong emphasis in linking some polar
recovery phase, two pieces of algae were removed compounds(e.g.,phlorotannin)withchemicaldefense
from each aquarium and used in no-choice feeding in algae seems unjustified by present experimental
assaysofeitherlivealgae(exposureto30amphipods) evidence (Jormalainen et al., 2001, 2003, 2005; Deal
oragar-basedfoodcontainingnon-polaralgalextracts et al., 2003; Hemmi et al., 2004; Kubanek et al.,
(exposure to 4 amphipods). No-choice assays were 2004). Recently, Taylor et al. (2002) revealed that
E.C.Macayaetal./J.Exp.Mar.Biol.Ecol.325(2005)214–227 219
non-polarextractsofstipes inthebrownalga Sargas- 3. Results
sum filipendula had a deterrent effect on the amphi-
pod Amphitoe longimana, and similar results were 3.1. Experiment I: effects of grazing, grazer presence
found by Fairhead et al. (2005) in Desmarestia and UVR on palatability
anceps. In experiments similar to ours, Ceh et al. (in
press) recently demonstrated induction of chemical NosignificantblockeffectsweredetectedinGlos-
defense with non-polar compounds in two species of sophora kunthii after the treatment phase with live
brown algae. For G. kunthii, several non-polar com- algae and agar-based food (ANOVA: F=3.19, p=
pounds such as pachydictyol A and dictyodal had 0.084 and F=0.93, p=0.471, respectively) and after
been described previously (see Rivera et al., 1987; the recovery phase (ANOVA: F=0.51, p=0.683 and
deNysetal.,1993),whichareknowntobedeterrents F=1.37, p=0.320, respectively). Similar results were
to amphipods, sea urchins and fishes (Hay et al., foundforM.integrifoliaafterthetreatmentphasewith
1987; Pereira et al., 1994; Cronin and Hay, 1996a; live algae and agar-based food (ANOVA: F=0.96,
Cronin et al., 1997; Barbosa et al., 2004). p=0.458 and F=0.10, p=0.958, respectively) and
In order to obtain comparable concentrations of aftertherecoveryphase(ANOVA:F=3.06,p=0.090
extracted chemistry between the original alga and andF=0.76,p=0.545,respectively).
agar-based food items, we used the ratio wet mass:
dry mass of U. lactuca (3:1) and mixed extracts with 3.1.1. Glossophora kunthii
an amount of dry U. lactuca powder representing Consumption rates of P. ruffoi on agar-based food
one-third of algal wet mass from which extracts were containingnon-polarextractsofacclimatedalgaewere
made. Following the methodology of (Hay et al., higherthanonextractsofnon-acclimatedalgae(paired
1994), blended U. lactuca powder was incorporated t-test:t=4.80,p=0.001,Fig.1I).Therewasnosigni-
into agar, poured into a mould lying over a fly mesh ficant UVR(cid:2)grazing interaction for live algae at the
(mesh size 1 mm2), and allowed to harden prior to endofthetreatmentandtherecoveryphase(Table1),
excision of a 200 mm2 quadratic section, which was indicatingthateffectsofgrazingtreatmentswereinde-
offered as agar-based food containing non-polar algal pendent from UVR-regimes throughout the experi-
extracts to amphipods in feeding assays. Consump- ment. There were no significant UVR-effects on
tion rates of these food items were determined by consumption rates neither after the treatment phase
counting emptied 1 mm2 fly mesh squares with a nor after the recovery phase (Table 1). No grazing
stereomicroscope. effects were found in feeding assays with live algae,
neitherafterthetreatmentphasenoraftertherecovery
2.6. Statistical analysis phase (Table 1). However, using agar-based food,
strong grazing effects were detected at the end of the
Choice feeding assays after the acclimation phase treatmentphase.Amphipodconsumptionratesonagar-
(induction experiment) and from within-algal palat- based food containing non-polar extracts of control
abilityexperimentswereevaluatedwithpairedt-tests. algae were significantly higher than on extracts from
Consumption rates from no-choice feeding assays G.kunthiiindividualsexposedtodirectgrazing,nearby
from within-alga palatability experiments were ana- grazed conspecifics, and non-grazing consumers
lyzedwitha1-wayANOVA.Consumptionratesfrom (Table 1, Fig. 1II(B)). At the end of the recovery
no-choice feeding assays after the treatment and re- phase,neitherlivealgaenoragar-basedfoodwassig-
covery phases (induction experiment) were analyzed nificantlydifferentamonggrazingtreatments(Table1,
by a mixed model 3-way ANOVA. Since no block Fig. 1III(A) and III(B)). There was no significant in-
effects(mainandinteractions)werefoundata=0.25, teraction between UVR and grazing treatment effects
datafrom allblockswerepooled andreanalyzed bya (Table1).
2-way ANOVA. Tukey HSD was used for post-hoc
comparisons of significant ANOVA results. Homoge- 3.1.2. Macrocystis integrifolia
neity of variances was confirmed with Cochran’s test After the acclimation phase, amphipods P. ruffoi
for all data. showed no significant feeding preferences for either
220 E.C.Macayaetal./J.Exp.Mar.Biol.Ecol.325(2005)214–227
I ACCLIMATION PHASE II TREATMENT PHASE III RECOVERY PHASE
-UV -UV
(A) UV (A) UV
med 100 b d (g) 23..50 d (g) 23..50
u 80 a me me
ns u 2.0 u 2.0
ares co 6400 nt cons 11..50 nt cons 11..50
qu 20 ou 0.5 ou 0.5
S m m
N 0 A 0.0 A 0.0
Non
Acclimated
Acclimated b b
d 100 (B) d 100 (B)
e e
m m
u 80 a a a a a a u 80
s s
n n
o 60 o 60
c c
s s
e 40 e 40
ar ar
u u
q 20 q 20
S S
N 0 N 0
Direct Indirect Grazer Direct Indirect Grazer
Control Control
Grazing Grazing Presence Grazing Grazing Presence
Fig.1.MeanconsumptionofGlossophorakunthii bytheamphipodParhyalellaruffoi afterthethreeexperimentalphasesI–III.(A)Feeding
assayswithintactalgalpieces,(B)feedingassayswithagar-basedfoodcontainingnon-polaralgalextracts;differentlettersabovebarsindicate
significantdifferencesinconsumption,absenceofletterabovethebarsindicatesnosignificantdifferences.Errorbarsrepresent +1S.D.(n=5
replicatesforeachtreatment).
acclimated or non-acclimated food items (paired t- M. integrifolia as well as agar-based food containing
test: t=(cid:1)0.80, pN0.05, Fig. 2I). Regardless of the the non-polar extracts of this species (Table 1, Fig.
UVR-regime, grazing treatments were not signifi- 2II and III). UVR-effects as well as UVR(cid:2)grazing
cantly different at the end of both, the treatment interactions of amphipod consumption rates on M.
and the recovery phase in feeding assays using live integrifolia were missing, both after the treatment
Table1
Resultsof2-wayANOVAs,testingforthecombinedandinteractiveeffectsofUVRandgrazing,onamphipodconsumptionrates,obtainedin3-
daylongno-choicefeedingassaysattheendofthetreatmentandtherecoveryphaseoftheinductionexperiment
Glossophorakunthii Macrocystisintegrifolia
Live Extracts Live Extracts
df F p F p F p F p
Treatmentphase
Grazing(G) 3 1.87 0.155 42.68 b0.001 1.49 0.236 2.09 0.121
UV 1 1.38 0.250 0.70 0.497 0.66 0.425 0.1 0.757
G*UV 3 0.57 0.634 0.81 0.408 0.14 0.936 1.95 0.141
Residual 32
Recovery phase
Grazing(G) 3 0.44 0.728 0.48 0.699 0.21 0.89 0.83 0.487
UV 1 4.58 0.050 1.19 0.285 0.66 0.423 1.12 0.299
G*UV 3 0.78 0.513 1.12 0.357 1.01 0.403 1.18 0.332
Residual 32
Analysesofpooleddataafterconfirminglackofblockeffectsata=0.25with3-waymixedmodelANOVA(resultsnotpresented).Extracts—
agar-basedfoodcontainingnon-polaralgalextracts(6amphipodsperassay),live—intactalgalpieces(30amphipodsperassay).
E.C.Macayaetal./J.Exp.Mar.Biol.Ecol.325(2005)214–227 221
I ACCLIMATION PHASE II TREATMENT PHASE III RECOVERY PHASE
-UV -UV
(A) UV (A) UV
med 100 d (g) 1.5 d (g) 1.5
nsu 80 ume 1.0 ume 1.0
co 60 ns ns
ares 40 nt co 0.5 nt co 0.5
qu 20 ou ou
S m m
N 0 A 0.0 A 0.0
Non
Acclimated Acclimated
d 80 (B) d 80 (B)
e e
m m
su 60 su 60
n n
o o
s c 40 s c 40
e e
uar 20 uar 20
q q
S S
N 0 N 0
Direct Indirect Grazer Direct Indirect Grazer
Control Control
Grazing Grazing Presence Grazing Grazing Presence
Fig.2.MeanconsumptionofMacrocystisintegrifoliabytheamphipodParhyalellaruffoiafterthethreeexperimentalphasesI–III.(A)Feeding
assayswithintactalgalpieces,(B)feedingassayswithagar-basedfoodcontainingnon-polaralgalextracts;differentlettersabovebarsindicate
significantdifferencesinconsumption,absenceofletterabovethebarsindicatesnosignificantdifferences.Errorbarsrepresent +1S.D.(n=5
replicatesforeachtreatment).
and the recovery phase (Table 1, Fig. 2II(B) and basal and apical parts, respectively, while no signifi-
Fig. 2III(B)). cant difference was detected between consumption
rates of basal and apical parts of live G. kunthii
3.2.Within-algavariationinpalatabilityofG.kunthii (Table 2, Fig. 3(A)). The same pattern with smaller
and M. integrifolia effects was observed in feeding assays using agar-
based foodcontaining non-polaralgal extracts. Medi-
Feeding assays with different parts of G. kunthii um parts were preferred over apical and basal parts,
indicatedaclearpreferenceofamphipodsformedium respectively (Table 2, Fig. 3(B)). This tendency was
parts of the algae. Using live algae, amphipods con- similar in no-choice assays, where amphipods con-
sumed medium parts 8.4 and 27.3 times more than sumed significantly more of the medium parts, both
Table2
Resultsofpairedt-tests,comparingamphipodconsumptionratesfromchoicefeedingassaysbetweenthree(Glossophorakunthii)andfour
(Macrocystisintegrifolia)differenttissuetypes
Glossophorakunthii Macrocystisintegrifolia
Live Extracts Live Extracts
t p t p t p t p
9 9 7 7
Apicalvs.medium (cid:1)5.72 b0.001 (cid:1)7.71 b0.001 3.37 b0.001 2.13 0.059
Apicalvs.stipe 11.62 b0.001 0.99 0.339
Apicalvs.basal (cid:1)2.06 0.069 1.64 0.136 2.44 0.054 2.03 0.061
Mediumvs.stipe 14.93 b0.001 2.29 0.056
Medium vs. basal (cid:1)11.19 b0.001 (cid:1)9.13 b0.001 2.36 0.051 0.73 0.423
Basalvs.stipe (cid:1)6.89 b0.001 0.64 0.540
Extracts—agar-basedfoodcontainingnon-polaralgalextracts,live—intactalgalpieces,subscript—degreesoffreedom.
222 E.C.Macayaetal./J.Exp.Mar.Biol.Ecol.325(2005)214–227
Glossophora kunthii Macrocystis integrifolia
A C
*** n.s. *** *** *** n.s. *** n.s. ***
g) 0.07 0.16
d (
e
m 0.05 0.12
u
s
n
o 0.08
c 0.03
nt
u 0.04
o
m 0.01
A
0.00 0.00
AP MED AP BAS BAS MED APMED APSTI APBASMEDSTI MEDBAS STIBAS
B D
ed 40 *** n.s. *** 40 n.s. n.s. n.s. n.s. n.s. n.s.
m
u
s 30 30
n
o
c
s 20 20
e
r
a
qu 10 10
S
N
0 0
AP MED AP BAS BAS MED APMED APSTI APBAS MEDSTI MEDBAS STIBAS
Fig.3.MeanconsumptionofdifferentpartsofGlossophorakunthii andMacrocystisintegrifolia bytheamphipodParhyalellaruffoi.(A–C)
Feedingassaywithintactalgalpieces,(B–D)feedingassaywithagar-basedfoodcontainingnon-polaralgalextracts.***Pb0.001(two-tailed
pairedt-tests).n.s—Nosignificantdifferences,AP—apical;MED—medium;STI—stipe;BAS—basal.Errorbarsrepresent+1S.D.(n=10for
Glossophorakunthii andn=8forMacrocystisintegrifolia).
in the feeding assays with live intact algae (1-way for any of the tissue comparisons (Table 2). Similar
ANOVA; F=23.41, pb0.001) as well as in those resultswereobtainedintheno-choiceassayswithlive
with agar-based food (1-way ANOVA; F=5.82, algae, where consumption of the stipe parts was sig-
pb0.01). nificantly lower than that of other parts (1-way
As a general trend, palatability levels in live M. ANOVA; F=7.99, pb0.001). Amphipods also con-
integrifolia declinedfromtoppartsofthealgaetothe sumed less ofthe agar-basedfood made with extracts
bottom parts (Fig. 3(C)). The live apical parts were from the lower region of the alga (1-way ANOVA;
preferredovermedium,basal,andstipeportions(Fig. F=18.28, pb0.001).
3(C)), with significant differences between apical vs.
medium and apical vs. stipe parts (Table 2). Further-
more, there was a very strong difference in consump- 4. Discussion
tion rates between medium parts and stipes, but no
significant difference in consumption rates between We observed strong effects of amphipod grazing
medium and basal portions of M. integrifolia (Table (at least in Glossophora kunthii) but no UVR-effects
2). Basal parts of this brown seaweed were signifi- on algal palatability. Grazing effects were dependent
cantly preferred over stipes (Table 2, Fig. 3(C)). The on algal species, but independent of UVR levels and
general pattern of tissue palatability in live M. integ- diminished within 14 days of amphipod removal. We
rifolia individuals was confirmed when using agar- observed strong differences in palatability levels
based food containing its non-polar extracts (Fig. amongtissuesinfield-grownindividualsofG.kunthii
3(D)), but no significant differences were detected and Macrocystis integrifolia.
E.C.Macayaetal./J.Exp.Mar.Biol.Ecol.325(2005)214–227 223
4.1. Inducible defense in Glossophora kunthii phyllum nodosum (Pavia et al., 2002; Toth et al.,
2005).
The feeding assays for G. kunthii revealed induc- Inducibledefenseinresponsetodirectgrazinghad
ible defense in response to direct grazing, indirect beenreportedforseveralspeciesofbrownmacroalgae.
grazing, and even non-grazing amphipods (Parhya- For example, Cronin and Hay (1996a) reported that
lella ruffoi). However, effects were only significant Dictyota menstrualis responded with increased con-
with agar-based food, which suggests that defense is centrations of diterpenoid compounds to direct graz-
basedonnon-polaralgalcompounds.Thediscrepancy ing.Severalotherstudiesalsodemonstratedenhanced
between the results with live algae and those with secondary metabolite production of macroalgae in re-
agar-based food appears inconsistent at first view, sponsetodirectgrazing(PaviaandBrock,2000;Pavia
because if chemical defense is ecologically signifi- andToth,2000;Sotkaetal.,2002;Rohdeetal.,2004;
cant, then one should expect similar results for both Weidneretal.,2004;Cehetal.,inpress).Furthermore,
live algae and for agar-based food. The results from grazing on adjacent conspecifics induced anti-herbi-
the second experiment offer an explanation for this vore defense in G. kunthii, which deterred further
apparent discrepancy. Medium parts of the algae are feeding by the amphipod P. ruffoi. Palatability of the
consumed significantly more than apical and basal brown alga Ascophyllum nodosum (Toth and Pavia,
parts suggesting that chemical defense (constitutive 2000)andFucusvesiculosus (Rohdeetal.,2004)also
and/or inductive) is absent or low in these medium decreased in response to waterborne signals from
partsofthealga.Inthefeedingassayswithlivealgae, neighboring grazed algae. The decline of palatability
grazers probably consumed the undefended medium due to waterborne cues could be caused either by
partsofthethallus,andconsequently,wedidnotfind chemicalsreleasedfromamphipods,woundedconspe-
significantdifferencesbetweenthegrazingtreatments cificalgae,orboth.Thefactthatreducedconsumption
and the control. Thus, the induction effect that we was also found in feeding assays with G. kunthii
observed in G. kunthii was limited to the defense of maintained in mere presence of grazers suggests that
apicalandbasaltissues.Possibly,theagar-basedfood waterborne signals are released from the consumers
contained a blend of non-polar extracts from induced themselves. Algae that are able to assess and respond
and non-induced tissues, which in sum could have to waterborne cues indicating an imminent risk of
been effective to deter amphipods. Poore (1994) also herbivore attack may have a selective advantage
foundwithin-algavariationinsecondarychemistryof (Toth and Pavia, 2000). Glossophora kunthii grows
Zonariaangustata (Dyctotales)occurringoveravery on cold-temperate shores and the dense monospecific
small spatial scale (a few millimeters). Their small stands of this alga are often inhabited by amphipods
size and ability to feed selectively at similar scales (Gelcich, 1999; Palma and Ojeda, 2002). These gra-
may allow amphipods to exploit small regions of the zersusuallyremovemodestamountsofalgalbiomass
algal tissues with low levels of secondary chemicals (Carpenter, 1986), which has been viewed as an in-
thatareunavailabletolargeherbivores(Poore,1994). trinsicelementofinducingdefensesinseaweeds(Kar-
Anti-herbivore defense of tissues that are respon- ban et al., 1999; Cronin, 2001). Consumption by
sible for the production of new cells has also been grazers with patchy spatial and/or variable temporal
reasoned as a useful strategy to allocate resources for distribution patterns, whose individuals remove only
defending tissues, which are of superior importance modest amounts of algal biomass, e.g., amphipods,
for algal fitness in the brown alga Sargassum filipen- may represent reliable cues for the induction of anti-
dula (Taylor et al., 2002). Similarly, the green alga herbivoredefense(Hay,1996;PaviaandToth,2000).
Halimeda macroloba, particularly defended new seg-
ments that are non-calcified (Paul and Van Alstyne, 4.2. No inducible defense in Macrocystis integrifolia
1988), and Tuomi et al. (1989) reported phenols ac-
cumulating especially in vegetative apical parts of For the giant kelp M. integrifolia, no significant
Fucus vesiculosus. Basal parts also were defended differencesweredetectedinpalatabilitybetweencon-
(result from live intact algae and agar-based food) in trolandgrazer-treatedalgaepieces,neitherwithagar-
Sargassumfilipendula (Tayloretal.,2002)andAsco- based food nor with live apical tips. This absence of