Sponge distribution and coral reef community structure off Mactan Island, Cebu, Philippines

6 Pages·1999·1.6 MB·

by  G J Bakus

SPONGE DISTRIBUTION AND CORAL REEF COMMUNITY STRUCTURE OFF
MACTAN ISLAND, CEBU, PHILIPPINES
G.J. BAKUS AND G.K.NISHIYAMA
Bakus, G.J. &Nishiyama, G.K. 19990630: Spongedistributionandcoral reefcommunity
structure offMactan Island, Cebu, Philippines. Memoirs ofthe QueenslandMuseum 44:
45-50. Brisbane. ISSN 0079-8835.
Coralreefcommunitystructurewasstudiedduring 1994-95atMactanIsland,offCebuCity,
Cebu,Philippines.Threetransectlinesperpendiculartotheshoreweresurveyedfromdepths
of 7-32m. Transect slack line distances were 55-68m long. Live hard coral represented
29-42%(mean=36%)ofcategoriesinterceptedandsponges 1-5%(mean=3%),representing
the two most abundant groups ofbenthic organisms. All remaining benthic taxa together
comprised only an average of 1% of the intercept distance. The number of sponges
intercepted along each line ranged from 4-19 (mean=12). Approximate sponge densities
from lineinterceptdatarangedfrom I/20m2to l/250m2(mean=l/40m2)andweretypically
large specimens. Sponge densities offMactan Island were considerably lowerand species
richness much higher than that of the Caribbean. A transition frequency matrix was
calculated for all line intercepts and a test for a Markov chain was conducted. The most
frequently encountered sequence was coral rubble followed by live hard coral (11%
sequencefrequency). Livehardcoralfollowedbyspongeswas5%andspongesfollowedby
livehardcoral was 5%. Inasimilarstudyofthebenthoswithfive line interceptsparallel to
theshoreatdepthsof7-12m,themostfrequentlyencounteredsequencewaslivehardcoral
tospongeandspongetolivehardcoral(36%).Spongetospongetransitionsrepresented4%.
None ofthe sequences were significant at P=0.05; i.e. the succession ofsubstratum types
were independentofeach other, supportingthenull hypothesis. O Porifera, Philippines,
coralreef, communityecology, benthic, densities, Markov Chains, transects.
G.J. Bakus (email: [email protected]) & G.K. Nishiyama, DepartmentofBiological
Sciences, UniversityofSouthernCalifornia, LosAngeles, CA 90089-0371 USA; 4January
1999.
ospue„Frcig.ierel.osdu.sopt,fuh.das.vpeesonbogeneesnspowonenargoeeisnfgm.ousAtinlhndecet„Po.h1i9.bl9..ei4p.ptSio.enxevisecr.batylo o?fr*fr^t£,h„e*x,?Tgakm*m«bVu&l1i,7?Rm«es£o?r&tQ, Pa'tt^y.PM..Paciltfn.aalensIs(,lF«a.ingdr.ee(ja/I)n')
fishesandhardcoralsand researchonthe effects (1 17 N, 124 E). This site is located approx-
ofthree species of allomone-secreting subtidal imate|y 400m N ot the University ofSan Carlos
sponges on adjacent corals are continuuing Manbago Marine Station. The Hilutangan
(Nishiyama, 1999, this volume). Sponges and Channel betweenMactanandOlangoIslandsisa
other marine taxa ofthe Philippines are poorly 300mdeeptrench(vonBodungenetal., 1985).A
known. Gomez (1980) lists 16 publications on gently sloping reefextends to a depth of10m on
marine sponges forthe Philippines and fewhave each side ofthe channel then typically, abruptly
appeared since that time (e.g., Raymundo & plunges steeply downward. The slope off the
Harper, 1995). No quantitative studies on Tambuli resort is considerably less steep.
spongeshavebeencarried outinthe Philippines, ^Oceanographic characteristics ofchannel waters
osbojefcatriveass oits tkhinsoswtnudyThweerree:for1)e,totdheetemramjionre (l99f1oau,mbj) ainndvolnlanBood&unDgaeclneset(a1L99(41)9.g5) Anon.
sponge distribution and abundance and compare v
it with similar data from other tropical regions, Three line intercepts perpendicular to the
and 2) to characterise coral reef community shoreline were surveved 10m apart from 7-32m
structureasabasicframeworkforourcontinuing depth with slack line distances of55-68m. Five
research on toxic sponges. additional line intercepts were surveyed parallel
MATERIALS AND METHODS totheshoreatdepthsof7,8,10,11 and 12m,with
slack line distances of 100m, excepting the 12m
Coral reef community structure and sponge line,whichwas 70m in slacklength (slacklength
distributions were studied between 1994-1996 partially follows the reefcontour and the tape is
46 MEMOIRS OF THE QUEENSLAND MUSEUM
10°20' reef community structure (after Licuanan &
—7'"' Bakus, 1992), so far as is known. There do not
IVlactan ^<$ appear to be null models developed for
island-/
Markov-type chain analyses dealing with
sequences ofsubstratum types wherethe species
)• V ^1> / { : are unidentified (Gotelli & Graves, 1996).
Approximate spongedensitieswerecalculated
using a modified Strong's equation (Cox, 1996),
as follows:
10°00' w D=(2 l/M)(A/T)
123°53' 124°04' where D=density of sponges (no./m M=
);
maximum animal or plant orthogonal width (m)
10*18'" ofan organism interceptingthe transect line; A=
Mactan island unit area (lm2); T=total transect length (m).
RESULTS
Univ. ofSanCarios
Marine Station Line intercept or line transect surveys are
recognisedasoneofthebestmethodsofstudying
Maribago
study site coral reeforganisms (e.g. Lova, 1978; Marsh et
10°17' al., 1984; Reichelt et ah, 1986). We compared
different sampling methods inthe field in Kenya
Agus ( and found that line intercepts gave us density
/ estimates closest to those of actual counts. For
example,thedensities(no./m2)ofPadina(?)ona
reef near Mombasa in 1997 and Ipomoea (I)
^Ns? leaves on a flat berm at Malindi in 1998 were as
follows: actual count (P-10, 1-14), point-center
10°16' quarter (P-2, 1-32), stratified random sampling
123°59' 124°00' 124°01 (P-20, 1-19) and line intercept (P-16, 1-14). The
FIG. 1. Mactan Island, Cebu, Philippines, showing major advantage of using intercept width
studysite. measurements for density estimates is that the
technique is about an order ofmagnitude more
easy to deploy formeasurements; see Reichelt et rapid than are direct counts or quadrat studies of
al., 1986, for methodology on reeftransects). organisms underwater. One advantage of using
transects both perpendicular and parallel to the
Organisms and substrata encountered along shore is that allelochemical effects can be
transects were recorded as line intercept
directional (i.e. directional currents) and using
distances. Most organisms were individuals or only one transect orientation might overlook
single colonies rather than conspecific patches.
transitions resulting from such chemicals.
Toxic sponges reported by Nishyamand Bakus
(1999, this volume) were most abundant at TRANSECTS PERPENDICULAR TO
depths of 7-12m. The number of transitions SHORELINE. Live hard coral represented
between sequences of organisms and substrata 29-42% (mean=36%) of categories intercepted
was tallied. The null hypothesis was that the and sponges 1-5% (mean=3%), representing the
sequence of occurrences of benthic organisms two most abundant groups ofbenthic organisms
and substratum types along transect lines (depth (Table 1). The standard errors (SE) of the
gradients)arerandom.Theresultsweretested for percentage intercept ofcommon categories were
a Markov chain (i.e. tendency ofone thing to be relatively small whereas those of the least
followed by another, as in repeated textile abundant categories were considerable, due
patterns) by the non-parametric chi-square mostly to small numbers. Obviously, increased
statistic. Markovchainanalysesofsequences are sample size would improve the results. All
commonlyusedinstratigraphy(Davis, 1986)and remaining large benthic taxa togethercomprised
in the prediction ofplant succession after fires only an average of 1% ofthe intercept distance.
(Isagi & Nakagoshi, 1990). The present study is The number of sponges along each intercept
the second use ofMarkov chainanalysis in coral ranged from 4-19 (mean=12). Sponge
MACTAN ISLAND SPONGES 47
TABLE 1.MactanIslandtransects,perpendiculartoshoreline(7-32mdepthregime),showingpercentageoftotal
interceptdistance(^seagrass was avoided afterthe firsttransect).
Category Line 1 Line2 Line3 Mean%±SE
LiveCoral 35 42 29 36±4
Sand 19 24 32 25±4
CoralRubhle 25 22 22 23i 1
DeadCoral 7 6 15 9±3
Seagrass1 9 - . 3±3
Sponges 3 5 I 3±I
OtherOrganisms : 1 1 1 ±0.03
%
Total 100 100 100 100
SlackLineDistance(m) 55 62 68 62
No.SpongesMeasured 14 19 4 12
TotalSpongeWidth(cm) 111 323 33 156
SpongeDensity(No./m') 0.02 0.05 0.004 0.025
SpongeDensity
Converted(No./nr) 1/50 1/20 1/250 1/40
approximate density from line intercept data percentagebetween sponges and live hardcorals
ranged from I/20m2 to l/250m2 (mean=l/40m2) was 10%fordepthsof7-32mand36%fordepths
and were typically large specimens (Table 1). of7-12m. The dominance ofsponges (afterhard
The most frequently encountered sequence was corals) is typical ofthe Caribbean but sponge
coral rubble followed by live hard cora! (11% abundance can be considerably greater in the
sequence frequency)(Table2A). Livehardcoral Caribbean than in the Philippines (Table 3). For
followed by sponges was 5% and sponges example,Zea(1994)foundthatspongesofSanta
followedby livehard coral was 5%. None ofthe Marta, Colombia, were an average ofabout four
sequences were significantat P=0.05 (calculated times as abundant as the sponges of Mactan
X"=0.19; DF=24); i.e. the succession of organ- Island. The species richness and irregular
isms and substrata were independent of each distribution of individual sponges at Mactan
other. This supports the null hypothesis. Island, however, were considerably greaterthan
at Santa Marta. Species richness can be high in
TRANSECTS PARALLEL TO SHORELINE.
the Caribbean. Alcolado (1979) reported 47
The sponge percentage of the total transitions species of sponges at depths of 11-16m in
increases rapidly from 7-10m depth, then Havana, Cuba. This rough inverse relationship
changes rapidly again between 11-12m depth between species richness and abundance was
(Table 2B). The most frequently encountered expected (Krebs, 1994).
sequence was live hard coral to sponge and
sponge to live hard coral (total combined=36%) As in the Caribbean, the sponges of Mactan
(Table 2C). Sponge to sponge transitions Islandrepresentthe secondmostimportanttaxon
represented 4%. Sponges were most frequently of benthic invertebrates after hard corals.
associated with live coral, dead coral and coral Although there is no universally recognised
rubble. Again, the sequence of organisms and system of classifying coral reefs based on
substrata were independent of each other at percentage cover, one frequently used rule of
P=0.05 (calculated x2=^\2.9\ DF-25). This thumb would consider that a 36% coverage (the
supports the null hypothesis. MactanIslandstudysite)representsamoderately
rich coral reef
DISCUSSION The reefs at Mactan Island are probablynot as
rich as they were 20 years ago because ofdense
Beyondtheshallowwaterseagrasscommunity development ofcoastal resorts and concomitant
(0-8m depth; principally Thalassia hemprichii), anthropogenic effects (principally sewage
sponges represented the second most dominant effluents).The 1997-98ElNinomayhavecaused
invertebratetaxon(afterhardcorals),comprising the death ofmany hard corals andthe disappear-
1-5% (mean=3%) of intercept distances (depth anceofnumeroustoxicspongesinourstudyarea
7-32m) and 5-28% oftransitions. The transition (Nishiyama, unpublished data).
48 MEMOIRS OF THE QUEENSLAND MUSEUM
TABLE 2. Mactan Island transects. A, transects increase in sponges. The marked increase in
perpendicularto shoreline, running from onshore to sponge transitions at a depth of 10-12m was
offshore, showing transition frequencies similar to that recorded by Alcolado (1979) for
(,=unbroken dead coral substratum). B, transects sponges at depths of 1 l-16m offHavana, Cuba
parallel to shoreline, showing sponge transitions by (Table 3). Presumably this depth is optimal
dtreapntshec(t2w=h1ic0h0mwassla7c0kmlleonnggt)h.eCx,cterpatnsfeocrts1p2armaldleelpttho bpheoctaousysnetheitsipsirnogvisdyemsbiosnutfsfiacnidenits blieglhotw ftohre
shoreline, showingspongetransition frequencies.
depth of major destruction by hurricanes and
%Totaltransition typhoons.
A. Category
frequencies
None ofthe transitional sequence patterns of
CoralRubbletoLiveCoral 11
organisms or substrata at Mactan Island were
LiveCoraltoCoralRubble *> statistically significant, thus the null hypothesis
LiveCoraltoSand 8 is supported. This was because species richness
SandtoLiveCoral 8 wasveryhighandindividualdistributionshighly
LiveCoraltoLiveCoral 8 irregular. Tropical island forests show the same
DeadCoral'toLiveCoral 6 patterns (e.g. Fiji Islands, Bakus, pers. obs.).
LiveCoraltoSponges 5 Significant sequential patterns of the benthos
SpongestoLiveCoral 5 would be expected to occur only at larger scales
LiveCoraltoDeadCoral 5 (e.g. community scale changes s&uch as trans-
itionsfromcoraltoseagrass)(Isagi Nakagoshi,
Others(numeroustransitioncategories) 35
1990).
Total 100
Bradbury& Young(1983)concludedthathard
coraldistributionsontheGreatBarrierReefwere
B. Totalno. Sponge%of Mean random. This was based on the fact that only 5
Depth(m) transoitfions" trantsoitatlions No.sponges (iSB) species pairs out of 545 species pairs were not
7 256 5 6 random. They concluded that coral interactions
on a small scale do not produce random neigh-
8 371 11 16
bours. They also concluded that random
10 426 15 32 27±7
neighbours express the combined workings of
11 396 16 31 many effects but did not specify what those
12 165 28 48 factors were. Licuanan & Bakus (1992) found
random distributions of Philippine reef
C. Category Percentageoftotal organismsatadepth of12m atPuertoGaierabut
transitions a significant non-random unidirectional pattern
LiveCoraltoSponge 18 at 6m depth, possibly the result ofstrong long-
SpongetoLiveCoral 18 shore currents. The possible reasons for the
SpongetoDeadCoral 12 random distributions of benthic organisms are
CoralRubbletoSponge 10 complex and include high species richness
SandtoSponge 8 (Bakus & Ormsby, 1994), fish predation and
grazing on benthic organisms (Bakus, 1964,
SpongetoSand 8
1967, 1969), predation on planktonic stages of
DeadCoraltoSponge 6 benthic species (Gili & Coma, 1998), perhaps
SpongetoCoralRubble 6 allelochemical defenses (Bakus et al., 1986,
SpongetoAlgae 5 1989/90) and other factors such as currents
SpongetoSponge 4 (Licuanan& Bakus, 1992). Adiscussionofthese
AlgaetoSponge 3 topics is beyond the scope ofthe present paper.
Others 2
Total 100 ACKNOWLEDGEMENTS
The somewhat abrupt change in sponge per- We greatly appreciate the help ofFilapina B.
Sotto, Head,MarineBiology Section, University
centage oftotal transitions (Table 2B) between ofSan Carlos, Cebu City, wrho made our studies
depths of7-10m was the result ofthe disappear- possible. Thanks to the following people who
ance ofseagrass (i.e. it is the edge ofthe shallow aided us in the laboratory and field: Mike Cusi,
water seagrass community) with a concomitant Jason Young, Benny Pangatungan, Jimmy and
MACTAN ISLAND SPONGES 49
TABLE3. Comparisonsofspongedensities(aspercentageofinterceptdistance,TD)between Indo-Pacificand
Caribbean localities (?=densities could notbedeterminedfrom the published data).
Location Depth(m) Method Density(range) Density(mean) Sourceofdata
Indo-Pacific
1-5%ID 3%ID
Cebu,Philippines 7-32 Lineintercept 1/50-1/250m2 1/40m" thisstudy
CFloirnadlerSseaR,eAeufstflraatlsi,a 1-4 Quadrats 7.3%ID 9 WiDlakniinelsoent(al1.98in7)
Caribbean
SantaMarta,
Colomia 17-22 Chaintransects 5-24%ID 13%ID Zea(1994)
RPaorqku,eVseNnaetziuoenlaal 1-35 Quadrats 7 1/2m2 Alvarezetal.(1990)
BiscayneNational
Park,FloridaKeys, 1-18 Quadrats 3-18/m2 8/m2 Schmahl(1990)
USA
1-17 Quadrats 0-lVnr 6/m2
Havana,Cuba AIcoIado(1979)
11-16 Quadrats 15/nr 9
Tonette de la Victoria, Chona Sister, Jonathan Allan Hancock Foundation Occasional Papers
Apurado, Frances Friere, Terence Dacles, Dave (27): 1-29.
Valles, Kathrin Meyer and Cynthia Nishiyama. 1967. The feeding habits of fishes and primary
Jason Young kindly sent us information on and production at Eniwetok, Marshall Islands.
figures of Mactan Island, and Chona Sister Micronesica 3: 135-149.
constructedthefinalfiguresofMactanIslandand 1969. Energetics and feeding in shallow marine
waters. International Revue of General and
profile ofthe study site. Fortheir reviews ofthe
mOcahnauvsiclrliop,tWwilefrtehdaonkLiCcuhaonnaan,SiPsatuelr,SDaommmianrgcoo BAKUESx,peGriJm.en&taOlRZMooSlBoYg,y4B:.271599-43.69C.omparison of
speciesrichnessanddominanceinnorthernPapua
and two anonymous reviewers. Rob van Soest New Guinea and Northern Gulf ofAlaska. Pp.
and John Hooper assisted with sponge 169-174.InSoest.R.W.M. Van, Kempen,T.M.G.
identifications. van& Braekman,J.C.(eds)Sponges inTimeand
Space. (Balkema: Rotterdam).
LITERATURE CITED BAKUS, G.J., TARGETT, N.M. & SCHULTE, B.
1986. Chemical ecology ofmarineorganisms: an
ALCOLADO, P.M. 1979. Ecological Structure ofthe overview. Journal of Chemical Ecology 12(5):
sponge fauna in a reef profile of Cuba. Pp. 951-987.
297-302. In Levi, C. & Boury-Esnault, N. (eds) BAKUS, G.J., STEBBINS, T.D., LLOYD, K.,
Sponge Biology. Editions du Centre National de CHANEY. H., WRIGHT, M.T., BAKUS, G.J.,
ALVAlRaERZe,cheBr.,chDeISAcZie,ntiMf.iCq.ue&(29L1A)U(CGNHRLSI:NP,ariRs.).A. PAODNAZMICNZIE,SKIC,. ME.ID&. CCR,EWHSU,NTP.E1R9,89/9L0..,
1990.Thespongefaunaonafringingcoralreefin Unpalatabilityanddominanceoncoralreefsinthe
Venezuela, I: composition, distribution, and Fiji and Maldive Islands. Indian Journal of
abundance. Pp.358-366. InRutzler, K.(ed.)New Zoology andAquatic Biology 1&2(1): 34-53.
Perspectives in Sponge Biology. (Smithsonian BODUNGEN, R. VON, BALZER, W., BOLTER, M.,
Institution Press: Washington DC). GR^vF, G., LIEBEZEIT, G. & POLLEHNE, F.
ANONYMOUS 1991a. PacificCebu Report. Baseline 1985. Chemical and biological investigations of
MarineResourceAssessment. Final Report,July. the pelagic system of the Hilutangan Channel
Pp. 1-56. (University of San Carlos, Marine (Cebu, Philippines). The Philippine Scientist 22:
Biology Section: Cebu, Philippines). 4-24.
1991b. Environmental Impact Study of proposed BRADBURY, R.H. & YOUNG, P.C. 1983. Coral
reclamations in the Municipalities of Lapulapa interactionsandcommunitystructure:ananalysis
City,CordovaandConsolacion.Part2.Cordova. ofspatialpattern.MarineEcology ProgressSeries
AreaResearchandTrainingCenter. Department 11:265-271.
of Marine Biology, Water Resources Center. COX, G. W. 1996. Laboratory Manual of General
(University of San Carlos, Marine Biology Ecology. 7th Edition. (William C. Brown:
Section: Cebu, Philippines). Dubuque).
BAKUS, GJ. 1964. The effects of fish grazing on DAVIS, J.C. 1986. Statistics and Data Analysis in
invertebrateevolution in shallowtropical waters. Geology. (John Wiley& Sons: NewYork).
50 MEMOIRS OF THE QUEENSLAND MUSEUM
GIL1, J-M. & COMA, R. 1998. Benthic suspension LOYA, Y. 1978. Plotless and transect methods. Pp.
&
feeders: their paramount role in littoral marine 197-218.InStoddart,D.R. Johannes,R.E.(eds)
food webs. TREE 13(8): 316-321. Coralreefs:researchmethods.(UNESCO:Paris).
GOMEZ, E.D. 1980. SectionK, Porifera. Pp. 35-36. In MARSH,M.,BRADBURY,R.H.& REICHELT, R.E.
984.Determinationofthephysicalparametersof
BibliographyofPhilippinemarinescience- 1978. 1
coral distributions using line transect data. Coral
(Filipinas Foundation, Inc. and Marine Sciences
Reefs2: 175-180.
Center, University ofthe Philippines: Manila). RAYMUNDO, L.J.H. & HARPER, M.K. 1995. Notes
GOTELLI,N.J. & GRAVES, G.R. 1996. Null Models onthe ecologyand distribution ofsponge genera
in Ecology. (Smithsonian Institution Press: (Porifera: Demospongiae) ofthecentral Visayas,
Washington, DC). Philippines. Pp. 39-52. In Proceedings ofthe 3rd
National Symposium on Marine Science.
II \NO,A.S.& DACLES,T.P.U. 1994.Environmental Philippine Scientist, Special Issue (1995).
kPpe.y p1a5-r1a8m.eteInrs,Sclhirghatm,mt,emWpe.ra&turDey,anDd.Ts.ali(neidtsy.) REICHELT, R.E., LOYA, Y. & BRADBURY, R.H.
1986. Patterns in the use of space by benthic
sfErnuiemnrgmgiyengrfrleoecfwo.uaRrenssdee,anurt4crhiAePnpatpreierlx,cFhtIoanMg4FeSMiCnaoyan,storro1pti9ic9ua4m! SCHMcBAaormHrmLiue,nriRtGei.eefPs..Coo1nr9a9l0t.wRoeCeocfomsrma5ul(2n)ri:ete7yf3s-8so0tf.ructthuereGraenadt
(University of San Carlos, Marine Biology
ecologyofspongesassociatedwithfoursouthern
Section: Cebu, Philippines)
Florida coral reefs. Pp. 376-383. In Rutzler, K.
ISAGI, Y. & NAKAGOSH1, N. 1990. A Markov (ed.) New Perspectives in Sponge Biology.
approach for describing post-fire succession of (SmithsonianInstitutionPress:WashingtonDC).
vegetation. Ecological Research5: 163-171. WILKINSON,C.R. 1987. Productivityandabundance
KREBS, C.J. 1994. Ecology. 4th Edition. (Harper oflargespongepopulationsonFlindersReefHats,
Collins College Publishers: NewYork). ZEA,CSo.ra1l99S4e.a.PaCtotrearnlsRoefecfosra5:l a1n83d-s1p88o.ngeabundance
LICUANAN,W.Y.&BAKUS,G.J. 1992.Coralspatial instressedcoral reefsatSanta Marta, Colombian
distributions: the ghost of competition past Caribbean. Pp. 257-264. In Soest, R.W.M., Van
roused? Pp. 545-549. In Proceedings ofthe 7th Kempen, T.M.G. & Braekman. J.C. (eds)
International Coral Reef Symposium. Vol. 1 Sponges in Time and Space. (Balkema:
(University ofGuam: Agana, Guam). Rotterdam).
SPONGES, INDICATORS OF MARINE responses to sediment disruption from trawling and
ENVIRONMENTAL HEALTH. Memoirs of the sand mining, and responses to water quality change
QueenslandMuseum44: 50. 1999:-Thereisanurgent duringalgal bloom events.
need for marine ecosystem indicators to facilitate Marine environmental indicators are likely to take
management aimed at either ameliorating impacts or the form of well-defined ecotypes described by
guiding sustainable utilisation of marine resources. characterising species presence. These species have
We propose that qualitative and quantitative knownrangesoftolerancetoenvironmental variables
examination of marine benthic communities will Suchaslight,current, foodsupply,turbidity,BOD,and
provide robust indication of responses to short and sediment regime. They are by their very nature,
long term environmental conditions, and further relevantataregionallevelandwillbesetinthecontext
suggest that information exists which permits the ofabiogeographicclassificationforanycoastorshelf.
creation ofa hierarchy ofindicators for establishing Theycanbefurtherrefinedby interrogationofmodels
ecosystemhealthinaregionalcontext.Theseareinthe relating population structure of key species to
form ofidentifiable marine community assemblages, biological and physical attributes oftheenvironment.
togetherwith biomass and growth indicesdetermined O Porifera, growth, morphology, indicators,
from morphological parameters associated with the environmental health, marine resources, benthic
characterisingspecies foreachassemblage. Examples communities.
are provided to demonstrate the sensitivity of such
indicatorsbyfocusingonspongecharacterisedcomm- CM Battershill* (email: c.battershill@/iiwa.cri.nz) & R.
unities. The composition ofassemblages and popul- Abraham, National Institute ofWater andAtmospheric
ationstatisticsofkeyspeciesreflectecosystemdisturb- Research, P.O. Box J4-901, Kilbirnie, Wellington, New
ances following catastrophic sediment deposition Zealand; *Presentaddress:AustralianInstituteof'Marine
followingcyclones,and inresponsetomorerecentand Science, PMB3, TownsvilleMC, Old, 4SI0,Australia; 1
relatively short-term impacts. The latter include June 199S.