sign in sign up
<<
Home , Limnology

FAU Institutional Repository

http://purl.fcla.edu/fau/fauir

This paper was submitted by the faculty of FAU’s Harbor Branch Oceanographic Institute.

Notice: ©1999 American Society of and , Inc. This manuscript is an author version with the final publication available at http://aslo.org/lo/ and may be cited as: Lapointe, B. E. (1999). Simultaneous top‐down and bottom‐up forces control macroalgal blooms on coral reefs (Reply to the comment by Hughes et al.). Limnology and Oceanography, 44(6), 1586‐1592.

1586 Comnment

. 1996. Demographicapproaches to communitydynamics: MORRISON, D. 1988. Comparingfish and urchingrazing in shallow A coral reefexample. 77: 2256-2260. and deepercoral reefalgal communities.Ecology 69: 1367- , B. D. KELLER, J. B. C. JACKSON, AND M. J.BOYLE. 1985. 1382. Mass mortalityof theechinoid Diadeina antillarumin Jamaica. MUNRO, J. L. 1983. Caribbeancoral reeffisheries. ICLARM Stud. Bull. Mar. Sci. 36: 377-384. Rev. 7: 1-276. , D. C. REED, AND M. J. BOYLE. 1987. Herbivoryon coral OGDEN, J. C., R. A. BROWN, AND N. SALESKY. 1973. Grazingby reefs:Community structure following mass-mortality of sea ur- the echinoidDiadema antillarum.Formation of halos around chins.J. Exp. Mar. Biol. Ecol. 113: 39-59. West-Indianpatch reefs. Science 182: 715-717. JANSSON, M., H. OLSSON, AND K. PATTERSON. 1988. Phosphatases: SAMMARCO, P W, J. S. LEVINTON, AND J. C. OGDEN. 1974. Grazing Origin,characteristics and functionin . and controlof coralreef community structure by Diadeina an- 170: 157-175. tillarumPhilippi (Echinodermata: Echinoidea): A preliminary LAPOINTE, B. E. 1997. Nutrientthresholds for and study.J. Mar. Res. 32: 47-53. on coral reefsin Jamaicaand southeast macroalgalblooms STENECK, R. S. 1988. Herbivoryon coral reefs:A synthesis.Proc. Florida.Limnol. Oceanogr. 42: 1119-1131. 6thInt. Symp.1: 37-49. LEE, T. N., M. E. CLARKE, E. WILLIAMS, A. M. SZMANT, AND T . 1993. Is herbivoreloss moredamaging to reefsthan hur- BERGER. 1994. Evolutionof theTortugas Gyre and itsinfluence ricanes?Case studiesfrom two Caribbean reef systems (1978- on recruitmentin the Florida Keys. Bull. Mar. Sci. 54: 621- 1988), p. 220-226. In R. Ginsburg Proceedingsof a col- 646. [ed.], loquiumon global aspectsof coral reefs:Health, hazards and LEICHTER, J. J., S. R. WING, S. L. MILLER, AND M. W. DENNY. 1996. Pulseddelivery of subthermoclinewater to Conch Reef, history.Univ. of Miami. Florida Keys by internaltidal bores. Limnol. Oceanogr.41: SZMANT, A. M. 1997. Nutrienteffects on coral reefs:A hypothesis 1490-1501. on the impoltanceof topographicand trophiccomplexity on LESSIOS, H. A. 1988. Mass mortalityof Diadeina antillarumin the nutrientdynamics. Proc. 8th Int. Coral Reef Symp.2: 1527- Caribbean:What have we learned?Annu. Rev. Ecol. Syst.19: 1532. 371-393. , AND A. FORRESTER. 1996. Watercolumn and sedimentni- LEWIS, S. 1986. The role of herbivorousfishes in the organization trogenand phosphorusdistribution patterns in the Florida of a Caribbeanreef community. Ecol. Monogr.56: 183-200. Keys, and potentialrelationships to past and presentcoral reef LIDDELL, W D., AND S. L. OHLHORST. 1986. Changesin thebenthic development.Coral Reefs 15: 21-41. communitycomposition following the mass moralityof Dia- VAN DEN HOEK, C., A. M. BREEMAN, R. P M. BAK, AND G. VAN dema antillarumat Jamaica.J. Exp. Mar. Biol. Ecol. 95: 271- BUURT. 1978. The distributionof , coralsand gorgonians 278. in relationto depth,light attenuation, movement and , AND . 1992. Ten yearsof distributionand change grazingpressure in the fringingcoral reefof Curacao, Neth- on a Jamaicanfringing reef. Proc. 7th Int. Coral Reef Symp., erlandsAntilles. Aquat. Bot. 5: 1-46. Guam 1: 144-150. WOODLEY,J. D. 1999. Diadema exertstop-down control of mac- MCFARLANE, A. H. 1991. The mariculturepotential of Gracilaria roalgaeon Jamaicancoral reefs.Coral Reefs 18: In press. species (Rhodophyta)in Jamaicannitrate-enriched back-reef habitats:Growth, uptake and elementalcomposition. 16 Ms. thesis,Univ. of Miami. Received. April 1998 MILLHAM, N. P., AND B. L. HOWES. 1994. Nutrientbalance of a Accepted. 14 September 1998 shallow coastal embayment:I. Patternsof groundwaterdis- Amended. 6 April 1999 charge.Mar. Ecol. Prog.Ser. 112: 155-167.

Li,niniol.Oceaniogr., 44(6), 1999, 1586-1592 @ 1999,by theAmiei-icaii Society of Limnologyand Oceainography,Ilc.

Simultaneoustop-down and bottom-upforces control macroalgal blooms on coral reefs (Reply to the commentby Hughes et al.)

In a recentarticle (Lapointe 1997), I reporteda studyof chemicalassay datathat I presenteddid notsupport my con- macroalgalblooms on coral reefsin Jamaicaand southeast clusions.I considerthese topics in the above orderand sug- Floridathat I hypothesizedwere related to simultaneousbot- gest thatnone of theirarguments accurately accounts for or tom-up(nutrient enrichment) and top-down(grazing) con- refutesthe data, interpretation,or conclusionsin Lapointe trols (relative-dominancemodel, RDM,- Littlerand Littler (1997) or themacroalgal bloom dynamics that have occurred 1984). Hughes et al. (1999) argued that (1) an exclusive in thetwo studyareas in question.Nonetheless, I am grateful grazinghypothesis is a more parsimoniousexplanation for forthis timely and importantexchange, as it will hopefully these blooms,(2) the nutrientthreshold concept I used to providea more refinedand clear understandingof the po- calibratethe nutrient dimension of the RDM was not valid, tentialcause(s) leadingto the demise of coral reefecosys- and (3) the nutrientconcentration, physiological, and bio- tems. Comment 1587

GRAZING LOSSES (75% AT DAY 80) trientcontrols was presentedin my conclusion: "The im- portanceof top-downcontrol of macroalgalstandings crops is much more compellingwhen combinedwith the syner- NUTRIENT-ENHANCED(650% AT DAY 80) LiIi GROWTH gisticeffects of bottom-upcontrol" (Lapointe 1997), an ex- 80_ planationthat I believe to be generaland robust. In contrastto my interpretation,Hughes (1994) and ~7O Hugheset al. (1999) arguedthat macroalgal blooms on reefs in Jamaicaand southFlorida are controlledexclusively by 3 6 0- SRP decliningherbivory from overfishing and the widespread ma- ( 0 die-offof the echinoidDiademna. Exclusive controlof E 5 0 -ENRICHED /, croalgalblooms by decreasedherbivory is not supportedby grazer-reductionstudies, e.g. Sammarco (1983), Lewis m 40 (1986), Carpenter(1988), and others,which reported rela- 2 0--~ ~ ~ ~~~~~~o----- tivelyminor expansion of algal turfs(< 2 cm tall, Lewis 30 CONTROL t> -- -e----*-- --e-- - 1986) and not macroalgalblooms. Data from Carpenter (1988) showedthat the increase of algal turfbiomass within 1 week of the sea urchinmass mortalitiesin St. Croix cor- 10- respondedwith an increasein turfheight from 1.0 to 2.9 mm and a 61% decrease of algal productivityper unitbio- mass, not the "2-fold increase in algal "as 10 2 0 3 0 4 0 5 0 6 0 7 0 8 0 claimedby Hugheset al. (1999). Lewis (1986) reportedonly DAYS a modestincrease in algal turfs(28% over 10 weeks) and no significantincrease in macroalgae(Halimeda, Turbinar- Fig. 1. Cumulativebiomass productionof S. polyceratiunin ia) followingexclusion of grazingfishes on theBelize - DIN-rich(>1 ,uM) watersof the Florida Keys withand without rier Reef. The RDM readily,and explicitly,predicts such (control)experimental SRP enrichment(data fromLapointe 1989). Also shownare the comparativecumulative grazing losses calcu- increasesin algal turf,not macroalgae, as a resultof reduced lated fromthe data of Lewis (1986). Note thatcumulative grazing grazingwhere nutrientavailability is low. The distinction losses are comparableto productiononly in the control betweenalgal turfsand macroalgaeis important,and be- culturesthat lack SRP enrichment;SRP-enrichment shifts macroal- cause Hughes et al. did not make thisdistinction, their ar- gal growthrates from linear to exponential,and the resultingbio- gumentas to the causes of "algal blooms" is confounded, mass increasecannot be controlledby grazers. at best. The RDM readilypredicts that decreased grazing on reefs withhigh nutrient availability, such as thereefs on Jamaica's (1) Top-downand bottom-upcontrols northcoast, will resultin dominanceby macroalgaerather thanby algal turfs.This is supportedby studiesthat have Hugheset al. (1999) impliedthat I discountedthe role of demonstrateddramatic increases in thegrowth of germlings herbivoryas a factorcontrolling macroalgal abundance on and frondosemacroalgae with increased nutrient availability, coral reefs.However, I explicitlyrecognized the contempo- as well as case studiesof eutrophicationon coralreefs where raneouseffects of both grazingand nutrientavailability as macroalgae overgrewreef corals (referencesin Lapointe potentialcontrollers of macroalgalblooms on coralreefs (see 1997). The relativeimportance of nutrientenrichment vs. Fig. 1, RDM, and referencesin Lapointe 1997). While the grazingon macroalgalstanding crops can be illustratedby term"macroalgal bloom" has been used loosely and is not comparingcumulative biomass of Sargassumpolyceratium definedin priorstudies, I definethis termto indicatean with and withoutenrichment of the growth-limitingnutri- increase in abundance, above normal background levels, of ent-soluble reactivephosphorus (SRP)-vs. grazinglosses a species in a geographic area or particularcoral reefhabitat. based on the "high grazing"data of Lewis (1986) (Fig. 1). My articlefocused on the dynamicsof the nutrientdimen- Hugheset al. (1999) citedthe work of Liddell and Olhorst sion of the RDM, a subject thathas been largely unexplored (1992) as supportfor their exclusive grazing hypothesis for on coral reefscompared to the wealthof studiesthat have Jamaicanreefs in statingthat "immediatelyfollowing the examinedherbivory but have largelyignored nutrient avail- mass mortalitiesof sea urchinsin 1983/84,algal biomass ability.For example,none of the "substantialnumber of increased."However, the data of Liddell and Olhorstdo not experimentalstudies" of Diadema removalcited by Hughes supportthat claim and actuallyshow the reefsat 15 m off et al. measurednutrient concentrations, so the claim that DiscoveryBay began theirshift from corals to macroalgae theseexperimental sites had "the same nutrientstatus" can- followingHurricane Allen in 1980, not followingthe Dia- not be made. In his review of the Diadema die-off,Lessios dema die-offin 1983-1984. Liddell and Olhorst(1992) re- (1988) did not considerthe possible role of nutrientavail- portedthat "it is of interestto note thatincreases in non- ability in explainingthe variable patternof macroalgal crustosealgae actuallybegan severalyears prior to themass blooms in the Caribbean,despite the knownimportance of mortalityof Diademna"and thatthis trendwas also "mir- nutrientsto growthand reproductionof macroalgaeand the roredby shallowerreef sites on Zingorroreef and otherJa- feedingpreferences and growthof marineherbivores (Matt- maican localities." Reefs in the relativelyoligotrophic wa- son 1980). My recognitionof simultaneousgrazing and nu- ters of the Cayman Islands and otherremote, low nutrient 1588 Comment

;Mi.R. F%., T

.4*<

Fig. 2. Bloom(up to 1.5 m deep, -3-ha area) of macroalgae(C linumand Lyngbyasp.) overgrowingfore reef communities on fringingreefs in theNegril Marine Park offshore Orange Bay,Jamaica, August 1998, following the onset of the wet season. During the past decade, macroalgalblooms like this have become increasingly common on Jamaican fringing reefs adjacent to nutrientinputs from sewage or agriculturalrunoff. Note the browsing sea urchinTripneustes ventricosusthat is normallyfound in morenutrient-rich meadows in theback reef along Jamaica'snorth coast. The recent expanded distribution ofthis urchin has followed the development of macroalgalblooms, suggesting cascading trophic responses to nutrientenrichment on thesefor- merlycoral-dominated fore reef communities. areas of the Caribbeandid not experiencethe macroalgal gal blooms on coral reefsin the widerCaribbean area. The blooms and coral loss seen in Jamaicafollowing Diadema nutrientthreshold concept offers a functionalframework for die-off.Liddell and Olhorst(1992) concludedthat a com- researchand managementof oligotrophicaquatic ecosys- binationof reduced grazing and eutrophicationwas most temsand is being used in restorationefforts for South Flor- likely fuelingthe increasedmacroalgal cover on shallow ida's Evergladesthat are affectedby agriculturalrunoff (see reefs in Jamaica,consistent with my previousconclusion McCormicket al. 1996 for alterationof naturalperiphyton (Lapointe 1997). That view is furthersupported by Goreau communities). (1992), who reportedthat "algal overgrowthof corals is a The nutrientthreshold model is supportedby studiesthat seriouslyescalating problem in mostJamaican reefs ... due have shownthe relationship between nutrient uptake/growth to a combinationof reducedherbivory and to increasedfer- of macroalgaevs. concentrationof the limitingnutrient to tilizationby nutrientsderived from inappropriate land and generallyfollow a rectangularhyperbolic function (e.g., Mi- sewage managementpractices" (Fig. 2). chaelis-Mentenkinetics). Continuous-culture laboratory ex- perimentsand detailedfield studies have quantifiedthis re- (2) The nutrientthreshold concept lationshipfor DIN-limitedgrowth, collectively showing a high affinityof macroalgaefor DIN, withgrowth rates be- My studywas not "a testof thehypothesis that reefs ... coming maximal (i.e., exponential)at very low DIN con- had exceeded a thresholdlevel of eutrophication"as stated centrationsin the rangeof -0.5-1.0 uM (Fig. 3). Because by Hugheset al. (1999) butrather, a fieldtest of thenutrient water-columnnutrient concentrations represent the net resid- thresholdmodel formacroalgal blooms. The onsetof eutro- ual sum of internalnutrient cycling, algal assimilation,and phicationand alterationof microbialpopulations, endolithic externalinputs, they offer the mostdirect method to assess algae, and microfilamentousalgal turfswould likelyoccur nutrientsufficiency for growth of macroalgae.Inasmuch as at nutrientconcentrations lower than those of macroalgal algal nutrientuptake/growth rates are concentration-depen- blooms,the latter reflecting a moreadvanced stage of eutro- dent,a nutrientthreshold model based on nutrientconcen- phication.I specificallyused this concept to calibratethe trations(rather than on nutrientfluxes) is not onlyvalid but nutrientdimension of the RDM and quantifythe approxi- is likelythe best index of nutrientstatus on a reefrelative matewater-column concentrations of dissolvedinorganic ni- to macroalgalgrowth demands. For example,despite overall trogen(DIN) and SRP thatcan initiateand sustainmacroal- high DIN fluxesassociated withhigh flow rate waterson Comment 1589

o Gracilaria tikvahias creasingDIN concentration.Thus, it appearsthat this level E Neoagardhiella balleyl of nitrogenousfertility may represent a moregeneral thresh- A Ulva fasciata old fora broadrange of algal-mediatedbenthic and pelagic S Dictyosphaeria cavernosa biologicalprocesses in themarine environment. * Macrocystis pyrifera Hugheset al. (1999) discountedthe nutrient threshold con- cept by statingthat "it is now widelyaccepted that (coral) 10 10- ? ?~~~~' [ C m== ------Sr - - E reefsare notlimited to low nutrientenvironments" and cited wl Lmax ~ -. NUTRIENT reviewsby Hatcher(1997) and Szmant(1997) forsupporting I1_ Vwf//2tNUTRIENT ".,SATURATION INHIBITION evidence.Hatcher (1997) presentsno nutrientdata to support thatclaim and, to thecontrary, notes that "models that predict theresponse of reefproduction processes to local and global 0 s 0-- ~~MAXIMUMGROWTH RESPONSE increasesin nutrientsupply are the mosturgent requirement of ecosystemscience." Szmant(1997) does notpresent a re- view of nutrienteffects on coral reefs,either, but rather, pro-

-LJ poses "a hypothesison the importanceof topographicand trophiccomplexity to reefnutrient dynamics." The few nu- 0.5 1.0 1.5 2.0 2.5 3.0 trientconcentration data presentedare not representativeof publishedvalues forthe locations; Szmant gives a value of 2 DIN, [tM ,uM NO- forreefs on the northernGreat Barrier Reef, but Fig. 3. Relationshipof relativegrowth rates (g: llnax) in a va- thatvalue exceedsthe median DIN concentrationreported for rietyof macroalgaeto increasingDIN concentration.Data forthe the outercoral-dominated reefs by more thanone orderof tropicalmacroalgae Gracilaria foliifera and Neoagardhiellabaileyi magnitude(0.13 ,uM DIN, Furnaset al. 1997). are fromDeBoer et al. (1978), Ulvafasciatadata are fromLapointe Hugheset al. (1999) also citedGlynn (1993) as evidence and Tenore(1979), and Dictyosphaeriacavernosa data are from Larned and Stimpson(1996). Also shown are data for the giant that"coral reefsthrive in upwellingareas." However,Glynn kelp,Macrocystis pyrifera, from a temperateupwelling (1993) concludedotherwise-"coral reefs are generallyabsent in California(Zimmerman and Kremer1982). These data showthat fromthe Arabian Sea coast of southernOman, most likely a the growthresponse region for DIN-limited macroalgal growth is resultof monsoon-inducedupwelling" and noted that "nutrient betweenundetectable (zero growth)and 0.5-1.0 ,iM DIN. At DIN pulsesthat accompany promote the growth of ben- concentrations> -1 ,iM, growthrates either become DIN-saturat- thicalgae whichmay interferewith coral growthand survi- ed or decreasedue to epiphyticfouling or toxiceffects. vorshipthrough increased competition." Hubbard (1997) noted thatreferences to coralreefs preferring areas of upwelling,or othersources of ,are a misconception. oligotrophicreefs with very low DIN concentrations,such habitatswould remainDIN-limited because the low DIN (3) Nutrient and concentrationswould restrictnutrient uptake and growthof concentration,physiological, macroalgae.Hence, when average DIN concentrationsare biochemical data below -0.5 ,uM on coral reefs(Fig. 3), growthrates of ma- croalgae are stronglylimited by DIN irrespectiveof flow I did notuse "a regressionwith only two points"to infer rate; underthese conditions,increased DIN concentrations temporaltrends in nutrientconcentrations at DiscoveryBay can lead to increased productivity(i.e., Pnmax}a) and growth, as assertedby Hughes et al.(1999) but rather,a timeseries initiatingthe macroalgalbloom, providingthat irradiance, comprisedof threeindependent studies spanning from 1979 temperature,and otherfactors are favorable. to 1989-the period duringwhich the macroalgalblooms The validityof a nutrientthreshold model forcoral reefs developedat thislocation. Hughes et al. citedD'Elia et al. was demonstratedby Bell (1992) in his criticalanalysis of (1981), the firststudy, as evidence of "an exponentialde- eutrophicationon reefsin Kaneohe Bay, Hawaii; Barbados; cline in surfacewater nutrientsand salinitywithin a few and theGreat Barrier Reef .Despite thesebroad geo- metersof the springsdue to dilution."However, D'Elia et graphicdifferences, increased algal abundanceand coralreef al. actuallyreported the relationshipbetween and declinewas apparentat averageconcentrations of -1.0 JIM NO- as linearor near-linear,with the mixing zone extend- DIN and 0.1-0.2 ,uM SRP. Very similar DIN and SRP ing all the way fromthe shorelinesprings to the forereef thresholdswere reported for macroalgal overgrowth of coral > 400 m fromshore, consistentwith my data fromJuly and seagrasshabitats along naturalnutrient gradients on the 1987. Macfarlane's(1991) thesisdata, collected between No- Belize BarrierReef (Lapointeet al. 1992). These thresholds vember1987 and June1989 in the same area, showedthat are furthersupported by a wide varietyof macroalgalblooms while reducedsalinity surface of the back reefhad reportedfor coral reefregions globally (see 3 in La- NO- concentrationsaveraging -13.2 ,uM,concentrations of pointe 1997), includingmacroalgal-dominated high-latitude highersalinity water at depthaveraged >2.0 ,uM-still two- reefs at the HoutmanAbrolhos Islands, WesternAustralia fold over the DIN threshold.The NOQ concentrationsI re- (Crosslandet al. 1984). Yentschand Phinney(1989) showed portedfor Discovery Bay (four samples per station,not that 1.0 ,uM DIN (as nitrateN) representsa value above "two" as claimedby Hughes et al.) were bracketedwithin whichthe dimensionless f-ratio and opticalproperties of the the rangeof NO- concentrationsreported by bothD'Elia et oceans (i.e., a*440) become saturatedwith respectto in- al. (1981) and Macfarlane(1991) and were not "atypical" 1590 Comment

as claimed.D'Elia et al. (1981) concluded"there appears to forereef. In Florida,the averageC: P (480) and N: P (37) be a hydrographicmechanism for transporting NO--rich wa- ratios of C. isthmocladumwere considerablylower than ter to the shallow reef," a statementinconsistent with those in Jamaicaand, in the case of the C: P ratio,lower Hughes et al.'s claim that "it is not knownhow much of thanmarine plants in general,reflecting a tendencytoward these . .. nutrientswere available to ... the fore reef." Tis- DIN limitation.DIN limitationof C. isthmocladumwas fur- sue analysisof themacroalga Lobophora variegata suggests thersupported by the low mean seawaterDIN: SRP ratio thatgroundwater NO- enrichesthe forereef at Discovery (8.4) on theFloridian reefs. Bay to depthsof at least 24 m (Lapointeet al. 1997). Hughes et al. (1999) assertedthat "no Codium blooms Contraryto thearguments of Hugheset al. (1999), a com- weredocumented during the summer [my] samples were col- parisonof thethree nutrient data sets does suggestthat SRP lected" in Florida.However, as I noted(Lapointe 1997), the concentrationsincreased at DiscoveryBay duringthe 1980s. initialmassive C. isthmocladumblooms thataffected deep D'Elia et al. (1981) foundno significantcorrelation between reefs(25-45 m) in southernPalm Beach Countyin the vi- salinityand SRP (average = -0. 15 ,uM) and concludedthat cinityof sewage outfallsand agriculturaldischarges were the groundwaterdischarges were "devoid of P." However, mostlyunattached plants and were physicallydisturbed by my SRP data in summer1987 were significantlyand nega- HurricaneAndrew in August1992. Beginningin 1994,when tivelycorrelated with salinity(Lapointe 1997), suggesting my samplingbegan, C. isthmocladumnpopulations spread to thatgroundwaters were a sourceof SRP at the timeof my northernPalm Beach Countywhere, for the firsttime, ex- sampling.Similarly, Macfarlane's (1991) more extensive tensiveattached populations began growingon reefswhere SRP measurementsfrom the same area, over a 19-month they had not been reportedpreviously (constitutinga period,were significantlyand negativelycorrelated with sa- "bloom" by definition).This northernexpansion was doc- linitybut averaged0.32 + 0.19 ,uM (n > 18); thisvalue is umentedand reportedin a Seagrant-fundedresearch project quitehigh for coral reefsand abouttwofold higher than the (Lapointeand Hanisak 1997) and thelocal media. Since the mean value of D'Elia et al. (1981). MaximumSRP concen- developmentof annualCodiumn blooms, these reefs in north- trationsreported by Macfarlane(1991)- .01 ,uM in May ernPalm Beach Countyhave been overgrownby extensive 1989-were the highestrecorded of all threestudies and populationsof Caulerpa verticillata,an attachedmacroalga occurredin the same timeframe when the blooms of Chae- typicallyfound in relativelynutrient-rich habitats. tomorphalinum and S. polyceratiumexpanded offshore from I was also criticizedfor not consideringupwelling as a restrictedareas of theback reef(not from "around the grot- nutrientsource to the Codiumblooms and forthe methods tos," as attributedto me). Because of theadditional evidence and interpretationregarding my 85N data. I specificallynot- I presentedthat SRP was a primarynutrient limiting pro- ed that"natural upwelling of deep offshorewater potentially ductivityof thesetwo macroalgaeat thissite, this trend of contributesnutrient inputs to theFlorida studyarea" in my increasingSRP could explainthe development of macroalgal discussion. However,none of the three studies cited by blooms in the late 1980s, irrespectiveof decreasedgrazing Hughes et al. providedany nutrientor physicaldata forthe by Diadema. My conclusionis supportedby observationsat Palm Beach Countystudy area, so theirclaim that "up- Dragon Bay in Portland,Jamaica, where SRP-rich sewage welling is a major source of nutrientsto these offshore and laundrydetergents from a resortled to thedevelopment reefs"is unsupported.The referenceto Leichteret al. (1996) of similarC. linumblooms in theadjacent bay; diversionof as evidenceof upwellingis incorrect,as thatstudy reported theSRP by thehotel managers led to a "rapidand dramatic the phenomenonof internaltidal bores, not upwelling.The improvementin the ecological conditionof the bay," and claim that85N of "upwelledwater from the North Atlantic "the weedy algae nearlyvanished" (Goreau et al. 1997). can be in the +10 to + 12%orange" is also unsubstantiated The evidencedoes not supportthe statementsof Hughes by eitherdata or citationsto otherwork. During my studies et al. (1999) thatthe high C: N: P ratiosof Jamaicanma- at JupiterLedge in summer1995, intrusionsof cold, NO- croalgae "do not indicatenutrient enrichment" or thatthe richupwelled water were not apparent in thestudy area, and "high C: P ratios from Florida . . . are more suggestive of DIN measurementsat thissite showed higher concentrations limitationby P thanN." My reportedaverage C: N ratioof of NH+ comparedto NOQ (a low f-ratio),evidence against theJamaican macroalgae (22) was veryclose to themedian the upwellinghypothesis. Even if episodic upwellingdoes (20) for benthicmarine plants in general (Atkinsonand occur at JupiterLedge, such transientnutrient inputs have Smith1983), supportingthe view thatthis reef experiences not historicallyinitiated or sustainedblooms of C. isthmno- relativelyhigh N availability.However, the C: P (973) and cladum(Lapointe 1997). N: P (45) ratiosof the Jamaicanmacroalgae were consid- The correlationbetween increased summer rainfall and erablyhigher than the median for marineplants (700 and increased85N to values between+ 10 and + 12%oin C. isth- 35, respectively),supporting the conclusionof SRP limita- mocladum(the methodsfor collection of tissue,drying, and tion.These highC: P and N: P ratiosreflected the high DIN: 85N analysis were describedin the methods)during the SRP ratioof the groundwaterinputs (-85: 1), furtherevi- summerbloom, togetherwith high near-bottomconcentra- dence that these macroalgae were enriched by tionsof NH+ and the supportingreferences I provided,col- -bornenutrients. The importanceof nutrientde- lectivelyprovided evidence thatwastewater-enriched DIN liveryby groundwaterswas also supportedby a significant pools may be contributingto growthof C. isthmocladum linearregression of N: P ratiosin themacroalgae vs. salinity (Lapointe 1997). The 85N values I reportedfor C. isthmo- (F = 10.36,P = 0.032), rangingfrom high N: P ratios(48- cladumduring summer 1995 are typicalof secondarilytreat- 57.1) aroundthe grottosto lowervalues (39.3-39.8) on the ed wastewater(greater than + 10%o;Lindau et al. 1989) and Comment 1591 are similarto values measureddowngradient of septictank- fact,my interpretationdid acknowledgenutrient limitation, contaminatedgroundwaters in the adjacentTown of Jupiter albeitonly "weak, N limitationof productivity"of C. isth- (greaterthan +7.3%o, unpubl.data). Approximately400 mil- mocladumncompared to more significantSRP limitationof lion gallonsper day of DIN-richsecondarily treated waste- C. linurn.My interpretationwas based on a two-wayAN- waterare disposed of by Class I injectionwells in south OVA, whichpartitioned the overallvariability within these Florida,which includes wells withinmiles of mystudy area. experimentaldesigns by assessmentof the sums of squares The 85N values I reportedfor C. isthmocladumare substan- and degrees of freedom,not just a visual comparisonof tiallyhigher than those reported for tropical macroalgae that treatmentmeans, which can be misleading.The bioassayre- rely on N fixation(+0.5%o) and are comparableto values sultsare entirelyconsistent with the other corroborating ev- formacroalgae from sewage-polluted sites such as Boston idence forthe two studysites thatincluded seawater DIN: Harbor (+7%o, France et al. 1998). SRP ratios,APA assays, and tissueC: N: P ratios.Primary The claim of Hughes et al. (1999) that "no evidenceis SRP limitationof C. linumat DiscoveryBay is consistent providedto supportthis assertion" regarding the relationship withthe longerterm macroalgal growth studies of Macfar- betweenhigh alkaline phosphatase activity (APA) and SRP- lane (1991), who reportedonly significantSRP enrichment limitedproductivity of theJamaican macroalgae is inconsis- effects,concluding that "growth of Gracilaria in nitrateen- tentwith the data I presented.That statementignores the riched habitats at Discovery Bay . . . is phosphate limited." significantdecreases in mean APA (n = 4) with SRP en- In summary,the multiple lines of evidencethat I advanced richmentof C. linum(F = 9.7, P = 0.05) and S. polycer- in Lapointe(1997) supportedmy conclusionsthat nutrients, atium(F = 23.7, P = 0.008) I reported(Lapointe 1997), in additionto decreasedherbivory, exert a significantand which confirmedSRP limitationin these two species. sometimesprimary forcing mechanism underpinning ma- Hughes et al. claimedthat "high levels of APA have been croalgalblooms on reefsin Jamaicaand southeastFlorida. shownto be inducedby low levels of P04 . . . but not of N The evidenceincluded a large body of publisheddata and enrichment."However, that opinion does notconsider stud- informationthat collectively showed (1) macroalgalblooms ies withCladophora prolifera, where NOQ enrichmentsig- on coral reefsare generallycontrolled by a complexinter- nificantlyincreased APA (P < 0.01; Lapointeand O'Connell actionof simultaneousbottom-up and top-downcontrols, (2) 1989) to the highlevels characteristicof the Jamaicanma- nutrientuptake/growth studies, togetherwith ecosystem croalgae.The highAPA levels of the Jamaicanmacroalgae studiesof eutrophicationalong nutrientgradients, support contrastsharply with the low APA levels of C. isthmnocladum thenutrient threshold concept for macroalgal blooms, (3) my fromFlorida, supporting SRP limitationin the formerbut data, and thoseof others,have demonstratedsignificant nu- not in the latter.The suggestionthat the APA should be trientlimitation of macroalgaeon coral reefs,and (4) ma- reportedas ",umolP" is incorrect(it could be expressedas croalgalblooms, such as thoseI described,are a logical and ,umolSRP or ,umolP03- but not the elementP) and irrel- predictableresponse to the widelyrecognized phenomenon evantbecause theAPA assay was used as a relativemeasure of increasednutrient inputs (especially ) to coastal of SRP limitation(i.e., "unitsml-"'). watersthat have occurredat local, regional,and globalscales I also replyto the criticismsof the nutrient-enrichmentover the past severaldecades (Vitouseket al. 1998). productivitybioassays with C. linumand Codium isthmo- Brian E. Lapointe cladum.Hughes et al. (1999) assertedthat the nutrient con- centrationsI used in the nighttimepulses (whichwere ad- Division of MarineSciences ministeredover severaldays, notjust one )exceeded HarborBranch Oceanographic Institution, Inc. the ambientnutrient concentrations at the two sites by a 5600 U.S. Highway1 factorof "10-fold" and were thereforenot "ecologically FortPierce, Florida 34946 relevant."The NO concentration(160 ,uM) used at Dis- coveryBay was about sixfoldhigher than the highestam- References bientDIN concentration(28 ,uM) measuredin surfacewa- ATKINSON, M. J., AND S. V. SMITH. 1983. C: N: P ratiosof benthic and I used a similarenrichment in where ters, factor Florida, marineplants. Limnol. Oceanogr. 28: 568-574. lowerconcentrations of NH4 (20 ,uM) wereused because of BELL, P. R. F 1992. Eutrophicationand coralreefs: Some examples lower ambientDIN concentrationson those reefs. The in theGreat Barrier Reef lagoon. WaterRes. 26: 553-568. "pulsed" nutrientbioassays were never intended to represent CARPENTER, R. C. 1988. Mass mortalityof a Caribbeansea urchin: ambientnutrient concentrations at thesites but rather, to sat- Immediateeffects on communitymetabolism and otherherbi- urate the nutrientuptake mechanismsof macroalgae and vores.Proc. Natl. Acad. Sci. USA 85: 511-514. elicit physiologicalresponses as a functionof antecedent CROSSLAND, C. J., B. G. HATCHER, M. J. ATKINSON, AND S. V. levels of DIN, SRP,or bothnutrients. Previous work reported SMITH. 1984. Dissolved nutrientsof a high-latitudecoral reef, by myselfand othershas been publishedin peer-reviewed HoutmanAbrolhos Islands, Western Australia. Mar. Ecol. Prog. scientificjournals and has demonstratedthat this short-term Ser. 14: 159-163. DEBOER, H. GUIGLI, T ISREAL, AND C. E D'ELIA. 1978. methodis a reliable assay fornutrient-limited productivity J. A., J. L. Nutlitionalstudies of twored algae. I. Growthrate as a function and reflectslonger term growth rates of macroalgae. of nitrogensource and concentration.J. Phycol.14: 261-266. Hughes et al. (1999) statedthat "the mostparsimonious D'ELIA, C. F, K. L. WEBB, AND J. W. PORTER. 1981. -rich interpretationof the resultsis thatinteractive effects of N groundwaterinputs to DiscoveryBay, Jamaica:A significant and P were greaterthan those of P alone, which suggests sourceof N to local reefs?Bull. Mar. Sci. 31: 903-910. nutrientlimitation, contrary to Lapointe'sinterpretation." In FRANCE, R., J. HOLMQUIST, M. CHANDLER, AND A. CATTANEO. 1592 Comment

1998. &5N evidencefor nitrogen fixation associated with ma- the coral reefchlorophyte Dictyosphaeria cavernosa, and the croalgaefrom a seagrass-mangrove-coralreef ecosystem. Mar. effectof exposureto sedimentderived nitrogen on growth. Ecol. Prog.Ser. 167: 297-299. Mar. Ecol. Prog. Ser. 145: 95-108. FURNAS,M., A. MITCHELL, AND M. SKUZA. 1997. Shelf-scaleni- LEICHTER, J. J., S. R. WING, S. L. MILLER, AND M. W. DENNY. trogenand phosphorusbudgets for the centralGreat Barrier 1996. Pulseddelivery of sub-thermoclinewater to ConchReef, Reef (16-19?S). Proc. 8thInt. Coral Reef Symp. 1: 809-814. Florida Keys by internaltidal bores. Limnol. Oceanogr.41: GLYNN, P W. 1993. Monsoonalupwelling and episodicAcanthaster 1490-1501. predationas possiblecontrols of coralreef distribution and com- LESSIOS,H. A. 1988. Mass mortalityof Diadema antillarumin the munitystructure in Oman,Indian Ocean. AtollRes. Bull. 379. Caribbean:What have we learned?Annu. Rev. Ecol. Syst.19: GOREAU, T. J. 1992. Bleachingand reefcommunity change in Ja- 371-393. maica: 1951-1991. Am. Zool. 32: 683-695. LEWIS,S. 1986. The role of herbivorousfishes in the organization , L. DALEY, S. CIAPPARA, J. BROWN, S. BOURKE, AND K. of a Caribbeanreef community. Ecol. Monogr.56: 183-200. THACKER. 1997. Community-basedwhole-watershed and LIDDELL,W. D., AND S. L. OLHORST. 1992. Ten yearsof distribution coastalzone managementin Jamaica.Proc. 8thInt. Coral Reef and change on a Jamaicanfringing reef. Proc. 7th Int. Coral Symp.2: 2093-2096. Reef Symp.1: 144-150. LINDAU, C. W., R. D. DELAUNE, W. H. PATRICK,JR., AND E. N. HATCHER, B. G. 1997. Coral reefecosystems: How muchgreater is the whole thanthe sum of theparts? Coral Reefs (suppl.) 16: LAMBREMONT.1989. Assessment of stable isotopes in finger- printingsurface water inorganic nitrogen sources. Water Air S77-S91. Pollut.48: 489-496. HUBBARD, D. K. 1997. Reefsas dynamicsystems, p. 43-67. In C. LITTLER, M. M., AND D. S. LITTLER. 1984. Models of tropicalreef Birkeland[ed.], Life and death of coralreefs. Chapman and Hall. biogenesis:The contributionof algae. Prog. Phycol. Res. 3: T. P. and HUGHES, 1994. Catastrophes,phase-shifts, large-scaledeg- 323-364. radationof a Caribbeancoral reef.Science 265: 1547-1551. MACFARLANE,A. H. 1991. The mariculturepotential of Gracilaria , A. M. SZMANT, R. STENECK, R. CARPENTER, AND S. MILL- species (Rhodophyta)in Jamaicannitrate-enriched back-reef ER. 1999. Algal blooms on coral reefs:What are the causes? habitats:Growth, nutrient uptake and elementalcomposition. Limnol.Oceanogr. 44: 1583-1586. M.S. thesis,Univ. of Miami. LAPOTh4TE,B. E. 1989.Macroalgal production and nutrientrelations in MATTSON,W. J. 1980. Herbivoryin relationto plantnitrogen con- oligotrophicareas of FloridaBay. Bull. Mar.Sci. 44: 312-323. tent.Annu. Rev. Ecol. Syst.11: 119-161. . 1997. Nutrientthresholds for bottom-upcontrol of ma- MCCORMICK, P V., P. S. RAWLIK,K. LURDING,E. P. SMITH, AND F croalgalblooms on coral reefsin Jamaicaand southeastFlor- H. SKLAR.1996. Periphyton-waterquality relationships along ida. Limnol.Oceanogr. 42: 1119-1131. a nutrientgradient in thenorthern Florida Everglades. J. North , AND M. D. HANISAK. 1997. Algal bloomsin coastalwaters: Am. Benthol.Soc. 15: 433-449. Eutrophicationon coral reefsof southeastFlorida. Florida Sea SAMMARCO,P. W. 1983. Effectsof fishgrazing and damselfishter- GrantProj. R/C-E-34. ritorialityon coral reef algae. I. Algal communitystructure. , M. M. LITTLER,AND D. S. LITTLER. 1992. Modificationof Mar. Ecol. Prog. Ser. 13: 1-14. benthiccommunity structure by naturaleutrophication: The Be- SZMANT,A. M. 1997. Nutrienteffects on coralreefs: A hypothesis lize BarrierReef. Proc. 7thInt. Coral Reef Symp.1: 323-334. on the importanceof topographicand trophiccomplexity to 5 , AND . 1997.Macroalgal overgrowth of fring- reef nutrientdynamics. Proc. 8th Int. Coral Reef Symp. 1: ingcoral reefs at DiscoveryBay, Jamaica: Bottom-up versus top- 1527-1532. downcontrol. Proc. 8th Int. Coral ReefSymp. 1: 927-932. VITOUSEK,P. M., AND OTHERS. 1997. Human alterationof the global , AND J.D. O'CONNELL. 1989. Nutrient-enhancedgrowth of nitrogencycle: Sourcesand consequences.Ecol. Appl. 7: 737- Cladophoraprolifera in HarringtonSound, Bermuda:Eutro- 750. phicationof a confined,-limited . YENTSCH,C. S., AND D. A. PHINNEY.1989. A bridge between ocean EstuarineCoastal ShelfSci. 28: 347-360. opticsand microbialecology. Limnol. Oceanogr. 34: 1694-1705. ZIMMERMAN,R. C., AND J. N. KREMER. 1984. Episodic nutrient 5 AND K. R. TENORE. 1981. Experimentaloutdoor studies withUlva fasciata Delile. I. Interactionof lightand nitrogen supplyto a kelp forestecosystem in southernCalifornia. J. Mar. Res. 42: 591-604. on nutrientuptake, growth, and biochemicalcomposition. J. Exp. Mar. Biol. Ecol. 53: 135-152. Received: 25 January 1999 LARNED, S. T., AND J. STIMSON. 1996. Nitrogen-limitedgrowth in Accepted: 13 May 1999