Coral Reefs (2008) 27:389–393 DOI 10.1007/s00338-007-0346-3

NOTE

Long-term dynamics of the brown macroalga variegata on deep reefs in Curaçao

M. M. Nugues · R. P. M. Bak

Received: 28 March 2007 / Accepted: 4 December 2007 / Published online: 18 December 2007 © Springer-Verlag 2007

Abstract Lobophora variegata occurs in the eulittoral have declined, while, in many areas, the cover and biomass zone and in deep water on coral reefs in Curaçao. An analysis of macroalgae have increased (Hughes et al. 2003). The of the long-term (1979–2006) changes in the vertical distri- causes of these changes remain highly controversial. Coral bution of the macroalga in permanent quadrats indicated a bleaching, diseases, nutrient enrichment and overWshing signiWcant increase in cover of the deepwater community. In appear to be major factors (Precht and Aronson 2006). 1998, Lobophora covered 1 and 5% of the quadrats at 20 and However, studies over broad spatial and temporal gradients 30 m, respectively. By 2006, these values had risen to 25 and are lacking (Hughes et al. 2003). In particular, deep 18%, precipitating a shift in abundance of corals and macro- (>20 m) coral reef communities have been neglected from algae at both depths. This increase coincided with losses in the majority of long-term studies documenting shifts to coral cover, possibly linked to bleaching, disease and storm- algal dominance. These communities have been considered related mortality in deep water plating Agaricia corals. In to be relatively stable due to their seemingly constant phys- contrast, macroalgae remained relatively rare (<6% cover) on ical conditions and protection against many of the anthro- shallower (10 m) and deeper (40 m) reefs despite declines in pogenic stressors impacting shallow reefs (Bak et al. 2005). coral cover also occurring at these depths, illustrating the The brown foliose macroalga Lobophora variegata depth-dependent dynamics of coral reefs. Several hypotheses (J. V. Lamouroux) Womersley ex E. C. Oliveira is distributed are suggested to explain these changes. worldwide in tropical and warm temperate seas. This mac- roalga has increased on many degraded reefs (e.g., Mumby Keywords Coral mortality · Coral bleaching · Deep coral et al. 2005), and poses a threat to coral populations by reef · Herbivory · Macroalgae · Phase shifts overgrowing adult colonies, reducing growth rates and inhibiting recruitment (Nugues and Bak 2006). In 1975, Lobophora showed a discontinuous vertical distribution Introduction pattern, forming two distinct belts: one in the eulittoral community and one in deep water on coral reefs in Cura- Coral reefs have experienced widespread ecological çao, Netherlands Antilles (van den Hoek et al. 1978). In changes during the last decades. Coral cover and diversity this study, long-term (1973/1979–2006) changes in the ver- tical distribution of the deep water Lobophora population and other benthic components were studied using a series Communicated by P. J. Mumby. of photographs of permanent quadrats distributed along a broad (10–40 m) depth gradient in Curaçao. M. M. Nugues (&) · R. P. M. Bak Royal Netherlands Institute for Sea Research (NIOZ), P.O. Box 59, 1790 AB Den Burg, Texel, The Netherlands e-mail: [email protected] Materials and methods

R. P. M. Bak University of Amsterdam, IBED, The abundance of L. variegata and other benthic organisms P.O. Box 94766, 1090 GT Amsterdam, The Netherlands was surveyed from permanent quadrats established in 1973 13 390 Coral Reefs (2008) 27:389–393 and repeatedly photographed until 2006. The quadrats were root transformed prior to analysis, and residual plots were arranged along three vertical transects: two located at checked for violations of model assumptions. All analyses Carmabi Buoy One (transects I and II) and one at Carmabi were performed with Brodgar version 2.5.2. Buoy Two (transect III) (for exact transect location and site descriptions see Nugues and Bak 2006). Each transect con- sisted of four 3 £ 3-m quadrats positioned at 10, 20, 30 and Results and discussion 40 m depths, respectively. The quadrats were permanently marked at each corner with stainless steel bars or plastic Lobophora was not recorded in the permanent quadrats Xoats. At each time interval, photographs were taken of between 1979 and 1992 (Fig. 1). This absence was proba- each complete quadrat and supplemented with a series of bly related to the rarity and small patch size of the macro- detailed (approximately 2 m2) overlapping photographs, alga, rather than it being completely absent from the except in 1973. Photographs were taken using a Nikonos V quadrats. In 1975, the deepwater population formed a dis- underwater camera Wtted with a 15-mm lens and Nikon tinct belt between 30 and 38 m along two 5 m wide vertical SB104 Xash. Black and white 400 ASA Wlms were used transects located within 200 m of the present quadrats (van until 1997 and subsequently replaced by color Wlms. den Hoek et al. 1978 but no quantitative data were High quality scans were made from complete quadrat reported). Furthermore, at the time of the die-oV of the her- photographs taken at eight 3- to 6-year intervals: December bivorous in 1983 and 1 year 1973, January 1979, September 1983, August 1989, Octo- later in 1984, Lobophora covered 0.2 and 2%, respectively, ber 1992, December 1998, May 2002 and February 2006. of the reef substratum in randomly placed 1 m2 permanent Digital Wles were imported into Photoshop 7. An 11 £ 11 quadrats at 27 m at Carmabi Buoy One (de Ruyter van unit uniform grid was superimposed over the image, and Steveninck and Bak 1986). the shape of the original image was adjusted using the free Lobophora increased signiWcantly in abundance between transform function until the quadrat matched the grid. 1992 and 2006 at 20 and 30 m (20 m: F1,8 = 26.42, Benthic organisms under each of the 100 points where grid- P <0.001; 30m: F1,8 = 43.79, P < 0.001). Absent in 1992, lines intersected were recorded in seven functional categories: the macroalga averaged 1 and 5% cover at 20 and 30 m, (a) hard coral (scleractinians), (b) sand, (c) other inverte- respectively, in 1998 and rose rapidly thereafter. By 2006, brates (gorgonian, sponge), (d) a category combining crus- these values had increased to 25 and 18%. It is unclear tose coralline algae, algal turfs (Wlamentous and Xeshy whether such expansion was the result of vegetative propa- algae <2 cm in size) and bare space (hereafter abbreviated gation or spore dispersal. In Curaçao, Lobophora coloniza- CTB), (e) L. variegata, (f) other macroalgae (e.g., Halimeda, tion seems to occur primarily by vegetative propagation, Dictyota, Sargassum), and (g) unidentiWable (i.e., due to but reproductive plants are present all year around (de shading or presence of mobile organisms). IdentiWcation Ruyter van Steveninck and Breeman 1987a, personal under each point was crosschecked with the detailed photo- observations). graphs. Since those were not available in 1973, only the Lobophora can exhibit seasonal Xuctuations in abun- cover of hard coral and sand was recorded for this year. dance. On the temperate North Carolina continental shelf, Crustose coralline algae, algal turfs and bare space were peak abundance of the alga occurs during periods of warm recorded together because they could not be reliably distin- temperatures (Peckol and Searles 1984). In tropical loca- guished from the photographs. This grouping is similar to tions, contrasting eVects of temperature have been reported Aronson and Precht (2000), but also includes small-sized for Dictyota, another genus of the same family (Dictyota- (<2 cm) Xeshy macroalgae, such as short and sparse indi- ceae; Mumby et al. 2005 and references therein). However, vidual Dictyota branches tightly adhering to the substratum. de Ruyter van Steveninck and Breeman (1987b) found no Only patches of several Lobophora blades could be reliably seasonal Xuctuation in Lobophora cover in Curaçao identiWed on the photographs, therefore, in cases where the throughout 1983–1985; therefore, it is unlikely that diVer- macroalga was not recorded, it could have still been present ences in sampling times could explain such large increase as sparsely distributed blades or small sized patches. in abundance from 1992 to 2006. Because the same quadrats were repeatedly measured An examination of the temporal patterns of change sug- over time, data were analyzed using repeated measures gests that the Diadema die-oV in 1983 had a minor impact ANOVA (Quinn and Keough 2002). Quadrat was consid- on deep water Lobophora. This result agrees with de ered as a random factor, and year as a Wxed factor. Separate Ruyter van Steveninck and Bak (1986) who found that the analyses were done for each depth at two time intervals: macroalga covered only 2% of the reef substratum at 27 m 1973–1992 and 1992–2006. Principle components analysis 1 year after a die-oV (see above), and could be related to the was also used to describe main trends of variation in ben- low pre-mortality densities of Diadema below 20 m, e.g., thic community at each depth. Data were arcsine square 0.4 urchins m¡2 at 27 m at Carmabi Buoy One (de Ruyter 13 Coral Reefs (2008) 27:389–393 391

᭣ Fig. 1 Temporal changes (mean § SE, n = 3) in the abundance of Lobophora variegata, hard coral, other macroalgae, other invertebrates and crustose coralline algae, algal turf and bare space together (CTB) between 1973/1979 and 2006 at each depth. Only percent cover of hard coral was recorded in 1973. Arrows at top of graph indicate timing of mass mortality of the sea urchin Diadema antillarum (I), intense bleaching period in deep water agaricids (II), and Hurricane Lenny (III)

van Steveninck and Bak 1986). In contrast, the increase in Lobophora coincided with signiWcant losses in coral and CTB at 20 and 30 m between 1992 and 2006 (20 m: coral—

F1,8 =7.80, P =0.02; CTB—F1,8 = 13.60, P =0.006; 30m: coral—F1,8 = 12.34, P = 0.008; CTB—F1,8 = 6.68, P = 0.03), resulting in a shift in abundance of corals and macroalgae. Increase in macroalgal abundance has been widely reported following coral mortality due to coral bleaching, diseases or hurricanes (Precht and Aronson 2006). In Glov- ers Reef, Belize, Lobophora increased in cover following the impact of hurricane Mitch in 1998 (Mumby et al. 2005). In Curaçao and Bonaire, coral cover on the deep reef (30–40 m) has remained relatively constant (until recently), because of positive growth by the dominant coral genus Agaricia, mainly Agaricia lamarcki and Agaricia graha- mae (Bak et al. 2005). Lobophora appears unable to directly overgrow and kill the living tissue of A. lamarcki (Nugues and Bak 2006). Therefore, it is likely that the increase in Lobophora was primarily a consequence of coral mortality rather than a cause. Coral mortality may have opened up new substrate to algal colonization, result- ing in reduced grazing pressure without even necessitating a change in abundance (Williams et al. 2001). The decrease in herbivory may have allowed the macroalga to increase in abundance until a new equilibrium between herbivory and algal production was reached. Deep water bleaching of plating Agaricia colonies was observed between 1996 and 1998 (Bak et al. 2005). In 1998, bleaching was particularly intense aVecting all plat- ing colonies below 20 m in the permanent quadrats. Mortal- ity in these colonies was observed in repeated photographs in 1999 and 2002 (Fig. 2), with several colonies showing additional signs of disease (personal observations). Subse- quently, sedimentation generated by hurricane Lenny in 1999 may have aVected plating agaricids. In Bonaire, Wne sediments generated by the hurricane killed nearly 40% of corals with Xattened colony shape versus less than 5% of non-Xat and hemispherical on aVected reefs below 30 m (Bak et al. 2005). The increase in Lobophora cover exceeded the loss of coral cover at both 20 and 30 m and was associated with a decline in CTB. This contradicts the assumption of Williams et al. (2001) in which, above a threshold of reef substratum available for algal growth (ca. 50–65%), any

13 392 Coral Reefs (2008) 27:389–393

conspicuous above 20 m (de Ruyter van Steveninck and Bak 1986; Fig. 1). Changes on shallow reefs may have impacted grazing pressure on deepwater algae. In Hawaii, the persistence of the macroalga Dictyosphaeria cavernosa on reef slopes was positively correlated with the coverage of non-native algae on reef Xats that were preferred by Wshes in palatability tests (Stimson et al. 2001). Finally, deep reef communities can experience a high variability in temperature, dissolved nutrients and sus- pended particle concentrations due to internal wave activity (Leichter et al. 2006). Temperature records at Carmabi Buoy One suggest that deep reefs are regularly exposed to cold-water inXux (Bak et al. 2005). Coincidentally, periods of reduced temperatures have been linked to deepwater bleaching in agaricids in Bonaire (Kobluk and Lysenko 1994). Fluctuations of internal waves may have aVected coral communities by inXuencing the concentrations of nutrients available to algal growth, as well as causing coral bleaching and diseases on deep reefs. Shifts in coral and macroalgal abundances did not occur at 10 and 40 m despite a signiWcant decline in coral cover at

10 m between 1973 and 1992 (F1,11 =20.02, P <0.001). This resulted in distinct trajectories of change in benthic Fig. 2 a Bleached colonies of Agaricia lamarcki (or Agaricia graha- mae) in December 1998 at 30 m depth (transect II). b Same colonies in community across depths (Fig. 3). The cover of other inver- November 1999, showing loss in coral cover and colonization of dead tebrates and unidentiWable remained below 6% at any depth areas by Lobophora variegata (arrows) and year. Sand cover showed no variation apart from a sig- niWcant increase at 40 m between 1992 and 2006 loss in coral cover should translate in a similar gain in mac- (F1,8 =5.76, P = 0.043). At 40 m, hard substrata are domi- roalgal cover. may avoid areas occupied by nated by sediment-laden algal turfs which may be unsuit- Lobophora once the macroalga become dominant. Larger able for Lobophora to recruit and/or grow on (personal Lobophora patches also have more numerous and larger observations). Further studies are required to understand blades and subsequently greater reproductive capacity via why macroalgae, and particularly Lobophora, are not yet dispersal of spores and vegetative propagation which may dominating Curaçao’s reefs at 10 and 40 m. necessitate a higher grazing pressure to control. Both mech- anisms raise the possibility of positive feedback in which coral mortality could lead to disproportionate increases in macroalgae. Alternatively, several hypotheses beside coral mortality deserve consideration. First, Curaçao’s reefs are considered as moderately overWshed, but no long-term data exists on herbivorous Wsh biomass which may have declined during the study period, resulting in reduced grazing pressure on Lobophora. Additionally, following the Diadema die-oV, herbivorous Wshes became the major grazers on coral reefs. Unlike sea urchin grazing, Wsh grazing may result in a sig- niWcant part of Lobophora blades being untouched (de Ruyter van Steveninck and Breeman 1987a). These rem- nants could continue to grow. Moreover, parrotWsh have been observed to reject Lobophora fragments after grazing (H. Zomer, personal observations), but the survivorship of Fig. 3 Principal components analysis (PCA) of changes in benthic these fragments has never been studied. composition inside permanent quadrats between 1979 and 2006 at each Secondly, CTB (mainly Wlamentous algae) rose in abun- depth. Symbols represent averages of three quadrats. Inset shows dance following Diadema die-oV, and macroalgae became eigenvectors of each benthic variable 13 Coral Reefs (2008) 27:389–393 393

Acknowledgments Authors thank the CARMABI foundation and Leichter JJ, Helmuth B, Fisher AM (2006) Variation beneath the personnel for continuous support. Authors are very grateful to surface: quantifying complex thermal environments on coral reefs G. Nieuwland, M. Vermeij and E. Meesters for help in the Weld and to in the Caribbean, Bahamas and Florida. J Mar Res 64:563–588 E. de Ruyter van Steveninck and A. Szmant for discussions. Mumby PJ, Foster NL, Glynn PW, Fahy EA (2005) Patch dynamics of coral reef macroalgae under chronic and acute disturbance. Coral Reefs 24:681–692 Nugues MM, Bak RPM (2006) DiVerential competitive abilities References between Caribbean coral species and a brown alga: a year of exper- iments and a long-term perspective. Mar Ecol Prog Ser 315:75–86 Aronson RB, Precht WF (2000) Herbivory and algal dynamics on the Peckol P, Searles RB (1984) Temporal and spatial patterns of growth coral reef at Discovery Bay, Jamaica. Limnol Oceanogr 45:251– and survival of invertebrate and algal populations of a North 255 Carolina continental shelf community. Estuar Coast Shelf Sci Bak RPM, Nieuwland G, Meesters EH (2005) Coral reef crisis in deep 18:133–143 and shallow reefs: 30 years of constancy and change in reefs of Precht WF, Aronson RB (2006) Death and resurrection of Caribbean Curaçao and Bonaire. Coral Reefs 24:475–479 reefs: a palaeoecological perspective. In: Côté I, Reynolds J (eds) de Ruyter van Steveninck ED, Bak RPM (1986) Changes in abundance Coral reef conservation. Cambridge University Press, Cambridge, of coral-reef bottom components related to mass mortality of the pp 40–77 sea urchin Diadema antillarum. Mar Ecol Prog Ser 34:87–94 Quinn GP, Keough MJ (2002) Experimental design and data analysis de Ruyter van Steveninck ED, Breeman AM (1987a) Deep water veg- for biologists. Cambridge University Press, Cambridge etations of Lobophora variegata (Phaeophyceae) in the coral reef Stimson J, Larned ST, Conklin E (2001) EVects of herbivory, nutrient of Curaçao: inXuence of grazing and dispersal on distribution levels, and introduced algae on the distribution and abundance of patterns. Mar Ecol Prog Ser 38:241–250 the invasive macroalgae Dictyosphaeria cavernosa in Kanoehe de Ruyter van Steveninck ED, Breeman AM (1987b) Deep water veg- Bay, Hawaii. Coral Reefs 19:343–357 etations of Lobophora variegata (Phaeophyceae) in the coral reef van den Hoek C, Breeman AM, Bak RPM, van Buurt G (1978) The of Curaçao: population dynamics in relation to mass mortality of distribution of algae, corals and gorgonians in relation to depth, the sea urchin Diadema antillarum. Mar Ecol Prog Ser 36:81–90 light attenuation, water movement and grazing pressure in the Hughes TP, Baird AH, Bellwood DR, Card M, Connolly SR, Folke C, fringing coral reef of Curaçao, Netherlands Antilles. Aquat Bot Grosberg R, Hoegh-Guldberg O, Jackson JBC, Kleypas J, Lough 5:1–46 JM, Marshall P, Nyström M, Palumbi SR, PandolW JM, Rosen B, Williams ID, Polunin NVC, Hendrick VJ (2001) Limits to grazing by Roughgarden J (2003) Climate change, human impacts, and the herbivorous Wshes and the impact of low coral cover on macroal- resilience of coral reefs. Science 301:929–933 gal abundance on a coral reef in Belize. Mar Ecol Prog Ser Kobluk DR, Lysenko MA (1994) ‘‘Ring’’ bleaching in southern 222:187–196 Caribbean Agaricia agaricites during a rapid water cooling. Bull Mar Sci 54:142–150

13