The effects of temperature on the growth of juvenile scleractinian corals
Tropical reef corals are well known for their sensitivity to rising temperature, yet surprisingly little is known of the mechanisms through which temperature acts on intact coral colonies. One such mechanism recently has been suggested by the association between the growth of juvenile corals and seawater temperature in the Caribbean, which suggests that temperature causes a transition between isometric and allometric growth scaling in warmer versus cooler years, respectively (Edmunds in Proc R Soc B 273:2275–2281, 2006). Here, this correlative association is tested experimentally for a cause-and-effect relationship. During April and May 2006, juvenile colonies (8–35 mm diameter) of massive Porites spp. from Moorea, French Polynesia, were incubated at warm (27.8°C) and cool (25.7°C) temperatures for 15 days, and their response assessed through the scaling of growth (change in weight) with colony size. The results reveal that the scaling of colony-specific growth (mg colony−1 day−1) was unaffected by temperature, although growth absolutely was greater at the cool compared to the warm temperature, regardless of colony size. This outcome was caused by contrasting scaling relationships for area-specific growth (mg cm−2 day−1) that were negatively allometric under warm conditions, but independent of size under cool conditions. In April 2007, a 22 days field experiment confirmed that the scaling of area-specific growth in juvenile Porites spp. is negatively allometric at a warm temperature of 29.5°C. Based on strong allometry for tissue thickness, biomass, and Symbiodinium density in freshly collected Porites spp., it is hypothesized that the temperature-dependency of growth scaling in these small corals is mediated by the interaction of temperature with biomass.
As tropical reef corals live near their upper thermal limit (Coles and Brown 2003) they are particularly susceptible to rising temperature (Berkelmans et al. 2004; Buddemeier et al. 2004). Therefore seawater warming is of great concern for this taxon (Hoegh-Guldberg 1999; Fitt et al. 2001), as evidenced since 1987 by the increasingly severe episodes of thermal bleaching (Fitt et al. 2001; Berkelmans et al. 2004) that have resulted in extensive coral mortality (Loya et al. 2001). During large-scale bleaching episodes, high temperature acts detrimentally to impair carbon fixation by the symbiotic algae (Dove and Hoegh-Guldberg 2006). However, there are also many other consequences of elevated temperature for corals including impaired calcification (Clausen and Roth 1975), elevated respiration (Coles and Jokiel 1977), reduced fecundity (Jokiel and Coles 1990), and perturbed reproductive synchronization (Babcock et al. 1986).
For most invertebrates, early life stages (hereafter described as juveniles) are more sensitive than adults to environmental stress (Hunt and Sheibling 1997), thereby creating the potential for population size to be influenced by the way in which juveniles respond to environmental signals. It is likely that the same is true for scleractinians (Edmunds 2005a), although the strength of the relationship between juvenile and adult population dynamics probably is weakened by the classically open population structure of this taxon (Caley et al. 1996), and the potential great age of colonies (Soong et al. 1999). Therefore, the dynamics of established scleractinian populations can be more strongly affected by the fate of adults than the success of juveniles (Hughes and Tanner 2000; Edmunds and Elahi 2007). Nevertheless, at some temporal scale the establishment of coral populations is dependent on juvenile colonies. As a result, the conditions affecting the success of juvenile corals should also affect the overall growth of the entire population. Although it is widely accepted that juvenile corals experience high mortality (Jackson 1977; Edmunds and Gates 2004), few studies have addressed post-settlement success for this taxon (e.g., Smith 1992; Dunstan and Johnson 1998). Critically, the important question of how a leading cause of coral death (i.e., rising temperature) might affect the success of juvenile corals remains virtually unaddressed, with pertinent empirical data available from only a handful of studies (Edmunds 2004, 2006). The empirical data of Edmunds (2004, 2006) originated from a decadal-scale study in St John, US Virgin Islands, and they reveal that warm conditions are associated with increased population densities and reduced growth (change in diameter) of juvenile corals. In warm years, the growth of juvenile corals scaled isometrically, but in cool years, it scaled allometrically [with scaling exponents >1 (Edmunds 2006)]. For juvenile corals, growth rates determine the duration of this important life stage, during which mortality rates are high (Jackson 1977; Edmunds and Gates 2004). Therefore, positive allometry for growth during cooler years has the potential to create a strong cohort of small colonies through enhanced growth, a shortening of the time spent in this life stage, and curtailed mortality (Edmunds 2006). This prediction is consistent with empirical data (Edmunds 2004). While collectively these trends from the shallow reefs of St John identify a potential mechanism through which changes in temperature could affect the post-settlement success of juvenile corals, they cannot establish a cause-and-effect relationship between temperature and growth scaling.
The objective of this study was to evaluate experimentally the effects of temperature on the growth of juvenile corals, and test the hypothesis that temperature causes a change in the scaling of growth. Although the empirical foundation for this study is based on the growth of juvenile corals as determined from changes in diameter (Edmunds 2004), the slow growth on this scale (Edmunds 2007) necessitated the use of an alternative technique for measuring growth. Therefore, growth was assessed from skeletal weight, which can be measured accurately over several days by buoyant weighing (Davies 1989). The biomass of freshly collected corals also was analyzed, as previous studies have shown that coral biomass can play an important role in the response to thermal stress (Loya et al. 2001; Grottoli et al. 2006), and in general, is quite labile in response to environmental conditions (Barnes and Lough 1992, 1999; Brown et al. 1999; Fitt et al. 2000).
Materials and methods
To test the effects of temperature on the scaling of growth, in April 2006 colonies of Porites spp. (Link, 1807) belonging to the massive morphological group were incubated at two different temperatures. Their response to these conditions was assessed as the change in skeletal weight as a function of colony size (diameter). To gain insight into the potential for biomass to modulate growth scaling, biomass was assessed by measuring tissue thickness, tissue weight, and the population density of Symbiodinium (Freudenthal, 1962) as a function of size. These tasks were accomplished using additional colonies of Porites spp. collected specifically for this purpose. A field experiment was also conducted in April 2007 in order to quantify the scaling of growth in juvenile Porites spp. at a naturally warm seawater temperature.
Juvenile colonies (≤40-mm diameter) of massive Porites spp. were selected for this study because they are common in shallow lagoons throughout the Indo-Pacific. In the lagoon of Moorea, massive Porites spp. are represented by two common species—P. lutea (Edwards and Haime, 1860) and P. lobata (Dana, 1846)—and two less common species—P. australiensis (Vaughan, 1918) and P. solida (Forskål, 1775)—all four of which are difficult to distinguish from one another. However, at 4 to 5-m depth on the western side of Avaroa Passe where the juvenile Porites spp. were collected, microscopic inspection of voucher specimens revealed that most of the adult Porites spp. either were P. lutea or P. lobata. Therefore, the adjacent juveniles probably were from the same species. The juvenile colonies were selected at random on 16 April 2006 from the population of small colonies that were approximately hemispherical in shape and largely free of conspicuous bioeroders. The freshly collected colonies were returned to the lab and placed into shaded tanks supplied with flowing seawater pumped from Cook’s Bay (≈28.0°C) prior to preparation for the manipulative experiment.
Temperature treatments were created in four 135 l acrylic tanks that were located outdoors and exposed to natural sunlight screened through a roof of neutral density mesh. This mesh reduced the ambient light to intensities (≈500 μmol photons m−2 s−1) within the range occurring at the depth (4–5 m) from which the corals were collected. A light sensor (Li-Cor LI 193SA) lowered to 4–5 m depth at the collection site recorded light intensities varying from ≤322 to 1,212 μmol photons m−2 s−1 (depending on weather) at about noon on multiple days in May. Clear plastic sheeting above the tanks prevented rain from mixing with the seawater. Each tank was operated with an independent, closed circuit of unfiltered seawater containing natural zooplankton and other particulate food. The seawater was collected from offshore and refreshed daily (20% turnover day−1), and the tanks each were equipped with a chiller, heater, and pump that was used to mix and aerate the seawater. Two tanks (the warm treatment) were maintained at ≈27.7°C, which is close to the values recorded in Avaroa Passe in April 2006 (≈28.5°C), and two at ≈2.0°C cooler (the cool treatment), which is close to the lowest temperature (25.2°C) in the Moorea lagoon during the winter (P.J. Edmunds, unpublished data). The mean temperatures in the two warm tanks were 27.6–27.8°C, and the mean temperature in both of the cool tanks was 25.7°C (SE values ≤ 0.1°C, n ≥ 112 measurements). To meet the needs of experimental objectives other than the one described here, one end of each tank was screened with neutral density mesh to create an additional treatment of light level. In the shaded end of each tank, the light levels at noon were reduced to ≈350 μmol photons m−2 s−1 compared to ≈500 μmol photons m−2 s−1 at the unshaded end. This additional treatment was not considered in the present analysis.
Following collection, the Porites spp. colonies were glued to plastic tiles using underwater epoxy (Z-Spar, A788) to make coral nubbins (Birkeland 1976), and were allowed to recover for 24 h in flowing seawater. The size of the corals was measured with calipers as the mean of the two major diameters of the basal portion of the colonies (±0.1 mm), and the weight was recorded by buoyant weighing (±1 mg) (Davies 1989). Twenty corals were allocated randomly to each of the tanks (40 temperature−1), with 10 placed randomly at each end of each tank. After 15 days, the corals again were buoyant weighed, and their surface areas determined using aluminum foil (Marsh 1970). Growth rates were calculated as an increase in dry weight, which was calculated from the change in buoyant weight assuming a constant skeletal density (see Davies 1989), and standardized by colony (mg colony−1 day−1), and area (mg cm−2 day−1).
To gain insight into the role of biomass in the scaling relationships of small corals (Edmunds 2006), additional colonies of Porites spp. were collected from the same location to test for an association between tissue traits and colony size. Experimental logistics prevented the measurements of these variables on the corals incubated in the temperature treatments. The freshly collected corals were fixed in 5% formalin in seawater and processed for tissue thickness, tissue weight, and the population density of Symbiodinium. After measuring colony size (diameter and area as described above), the formalin-fixed corals were decalcified outdoors in 10% HCl in freshwater until the tissue remained as a limestone-free tunic. The thickness of the decalcified tissue was measured using a dissecting microscope and an eyepiece micrometer after bisecting the tissue with a scalpel. The thickness at the cut surface (normal to the outer layer) was determined at 4–5 locations selected at random, and the average of these values used to characterize the tissue thickness (mm) of each colony. After measuring the thickness, the formalin-fixed tissue was homogenized with an ultrasonic dismembrator (Fisher 15-338-550) fitted with a 3.2-mm diameter probe (Fisher 15-338-67), in order to quantify the Symbiodinium. Preliminary trials showed that ultrasonic treatments of short duration and low power were sufficient to disrupt the animal cells without damaging algal cells. The Symbiodinium in the slurry were counted using a haemocytometer (5 replicate counts) and the density expressed as cells cm−2. Finally, biomass (mg cm−2) was determined by drying an aliquot of the slurry at 60°C until a constant weight was obtained.
In April 2007, the mean seawater temperature at 5 m depth in Avaroe Passe was 29.5°C (SE < 0.1°C, n = 22 days), and this created the opportunity for an in situ experiment to determine the scaling of growth in juvenile Porites spp. under warm conditions. Although it was not possible to carry out a similar experiment under cool conditions (i.e., in the Austral winter), quantification of growth scaling under naturally warm conditions proved helpful in interpreting the results from the experiment completed in 2006. For the in situ experiment, juvenile colonies of massive Porites spp. again were collected from 4–5 m depth in Avaroe Passe, returned to the lab, prepared as nubbins, and buoyant weighed (all as described above). These corals were returned to the collection site on 10 April 2007, and left to grow until 3 May 2007 (22 days) when they were recovered for the measurement of the buoyant weight and tissue area (by aluminum foil). The change in buoyant weigh was converted to dry weight, and the growth rates (mg cm−2 day−1) used in the analysis of the scaling of growth on size (diameter).
The corals analyzed in the present study were also used in a subsequent experiment (P.J. Edmunds, in preparation), in which the replicate tanks were treated as a random factor nested within temperatures. To meet the objectives of the present analysis, the coral colonies were treated as statistical replicates for the temperature contrast, based on the rationale that: (1) the aggregate coral biomass was small relative to the size of the tanks, and therefore together with the 20% day−1 water change, it was unlikely that the colonies affected one another, and (2) the difference in light levels between ends of the tanks did not have a biologically significant effect on growth. These assertions are supported by a comparison of growth (mg cm−2 day−1) first, between tanks within temperatures, and second, between each end of the tanks. The first analysis was completed using one-way ANOVAs, which revealed no difference between tanks in either the cool (F = 0.515, df = 1,30, P = 0.479) or the warm (F = 1.116, df = 1,29, P = 0.299) treatments. The second analysis was completed using a three-way ANOVA that included the effects of shading, temperature, and tank (nested within temperature), which revealed that shading had no effect on growth (F = 0.105, df = 1,57, P = 0.747); none of the other factors were significant (P > 0.504). Thus, corals were pooled between tanks within temperatures, and between ends within each tank, for the purpose of exploring growth scaling as a function of temperature. To test for an effect of temperature on growth scaling, the warm treatment served as an experimental control in that it was created to simulate ambient conditions in a contrast with conditions that were cool relative to ambient seawater.
For the manipulative experiment completed in 2006, the scaling of growth was investigated by regressing growth on size on double logarithmic plots, with separate analyses completed for the growth increment (mg colony−1 day−1) and the growth normalized to area (mg cm−2 day−1). In these relationships, the slope of the regression lines provides the scaling exponent (given the notation “b”). The significance of the relationships between growth and size were tested with Model I regressions and, where these were significant for both temperature treatments, the slopes and the elevations of the regression lines were compared with ANCOVA. For significant relationships, best-fit lines were fitted by Model II (Reduced Major Axis [RMA]) techniques, because both variables were measured with error (Sokal and Rohlf 1995). Standard errors of these slopes were calculated by Model I regression techniques. The same techniques were also used to analyze the scaling of biomass traits for the corals collected in 2006, and in 2007 for the corals used in the growth experiment completed in situ.
All statistical analyses were completed with Systat 9.2 running in a Windows operating system. The statistical assumptions of ANOVA and ANCOVA were tested through graphical analyses of the residual sum-of-squares that detected departures from normality and homoscedasticity.
The incubation experiment was completed between 17 April 2006 and 1 May 2006, which was a period characterized by cloudy conditions, heavy rain, and low surface light levels. Overall, mean light levels in the tanks between 10:00 and 14:00 hours varied between 291 ± 26 and 719 ± 25 μmol photons m−2 s−1 (±SE, n = 6 measurements over 5 days), and these low irradiances resulted in slow growth rates.
ANCOVA comparing the scaling relations for growth increment (mg colony−1 d−1) versus size (diameter, mm) between temperatures for juvenile Porites spp
For tropical reef corals, rising temperature brings the possibility of death (Hoegh-Guldberg 1999), but it can also bring more subtle effects such as differential survivorship during bleaching (Loya et al. 2001), increased rates of calcification (Buddemeier and Kinzie 1976), and perhaps enhanced growth and population density of juvenile colonies (Edmunds 2004). The study of the subtle effects of temperature was the objective of this study, specifically by testing the hypothesis that rising temperature alters the scaling of growth in juvenile colonies of massive Porites spp. The most noteworthy outcome of this analysis is the demonstration through manipulative experimentation that growth scaling is negatively allometric at 27.8°C, but independent of size at 25.7°C. Growing corals in situ during an unusually warm period, which also resulted in negative growth allometry, supported the hypothesized role of temperature in causing this response. Together, the present findings are biologically significant because they establish a cause-and-effect relationship between temperature and the scaling of growth in juvenile corals. This outcome supports a mechanism that was hypothesized to account for a similar association first identified through a mensurative approach (sensu Hurlbert 1984) applied over 9 years on a Caribbean reef (Edmunds 2004, 2006). Although there is uncertainty regarding the means by which temperature creates this outcome, the positive allometry of biomass in freshly collected Porites spp. suggests that the effects may be caused by the quantity of coral tissue.
In this study, juvenile Porites spp. grew at rates within the lower range of values reported for other corals in shallow water (Lough and Barnes 2000; Edmunds 2005b). Moreover, the broad effects of temperature, notably with a slight enhancement of growth at the cool compared to the warm temperature (Fig. 1a, b; Table 1), are consistent with the inverse temperature dependency of coral calcification that occurs beyond the thermal optimum (Buddemeier and Kinzie 1976; Reynaud-Vagany et al. 1999; Edmunds 2005b). The thermal optimum for calcification has been quantified in only a few tropical corals, but the data available suggests it occurs between 26 and 28°C (Buddemeier and Kinzie 1976; Reynaud-Vagany et al. 1999; Edmunds 2005b). Therefore it is conceivable that the cool treatment in the present study was close to the optimum for calcification in massive Porites spp., such that growth at a higher temperature was reduced. Regardless of this putative mechanism, the slight difference in growth rates between temperatures probably also reflects the interaction of temperature and colony size that depressed the growth of larger colonies at the warm temperature.
The statistical procedure required to test the central hypothesis of this study is a contrast of growth scaling at different temperatures. Ideally, this would be accomplished through a comparison of slopes (i.e., the scaling exponents) of regression lines on double logarithmic plots of final versus initial diameter. This approach would be consistent in analytical style with many scaling studies (Gould 1966; Schmidt-Nielsen 1989) as well as the mensurative analysis that provided the context for the present study (Edmunds 2006). Unfortunately, the use of diameter as a dependent variable in an analysis of short-term coral growth is precluded by the small changes that would be expected over this time (Edmunds 2007). Skeletal weight provides a tractable alternative to diameter that can be measured with milligram-resolution (Davies 1989) over days to weeks (Edmunds 2005b). Theoretically, growth on a mass basis should also scale isometrically for corals that conform to the classic constraints of a colonial, modular design (Jackson 1979; Sebens 1987; Hughes 2005). Empirical evidence reveals however, that scaling in colonial corals is not always isometric (Patterson 1992; Vollmer and Edmunds 2000; Edmunds 2006; but see Elahi and Edmunds 2007a), with a leading explanation being that the size of modules (i.e., polyps) varies (Vollmer and Edmunds 2000; Edmunds 2006), and therefore is not as conserved as was once assumed (Jackson 1979; Patterson 1992).
In the present study most of the Porites spp. increased in weight, but the increments were relatively small and variable compared to the results of experiments that have used similar approaches with corals from equivalent depths (≈6 m) (Edmunds 2005b). It is likely that these discrepancies resulted from using at least two species in this study (P. lobata and P. lutea), as well as inclement weather that reduced irradiances, and therefore light-dependent growth (Barnes and Chalker 1990). While the broad conclusion from this experimental work is the same as that from a mensurative analysis on a Caribbean reef (Edmunds 2006), there are subtle differences between the studies. Namely in the present study, warm water created a growth disadvantage (i.e., negative allometry) to becoming larger, whereas previously warm water was associated with isometry rather than positive allometry (Edmunds 2006). It is likely however, that these dissimilar trends are the product of a common mechanism operating during a short exposure to constant temperature (this study), or a year-long exposure to seasonally varying conditions (Edmunds 2006). Thus for example, the negative allometry recorded here might occur in situ only during the warmest period of a year, which together with isometry or positive allometry during the cooler months, could sum to produce isometric growth over the whole year. Cooler years therefore might be characterized overall by positive growth allometry because there is no growth disadvantage to being larger during cooler summers. Alternatively, over a whole year there may be time for changes in skeletal porosity to play an important role in translating mass deposition of aragonite (as recorded here) into size (as recorded in Edmunds 2006), thereby creating apparently dissimilar responses of growth scaling to temperature. Clearly, to formalize the role played by temperature in mediating the scaling of growth in juvenile corals, a hypothesis is required to explain how scaling—that typically is constrained tightly by organism design (Gould 1966)—can be plastic for growth in this system.
The hypothesis most consistent with the present results posits that the effects of temperature on growth scaling are mediated by biomass. In this hypothesis, bigger juveniles have thicker tissue than smaller juveniles, and therefore, the stimulatory effects of high temperature on physiological processes is likely to cause rate-limitation in larger colonies due to slow internal transport of metabolites. As a result of rate-limitation, growth would be impaired and growth scaling would become negatively allometric. Apart from the scaling of growth reported here, support for this hypothesis comes from three sources: (1) the analysis of biomass scaling in Porites, (2) previous studies of variation in biomass of reef corals, and (3) evidence from reef corals of the limitation of physiological processes as a result of slow metabolite transport.
First, the present results show that the thickness of the tissue (i.e., biomass per area), and the population density of Symbiodinium, all increase with size, at least in freshly collected corals. Although the thickness of coral tissues can be altered through the absorption of water, notably into the gastrovascular cavity, this mechanism probably was not important in the present study. The chief arguments against this possibility arise from the anastomosing of the thick tissue through the highly porous skeleton (i.e., it was not simply “inflated”), and its high biomass content (i.e., thick tissue contained more than just additional water). For Porites spp., while the relationship between tissue thickness and polyp dimension is made unclear by the intertwining of the tissue with the perforate skeleton (Barnes and Lough 1992), it would be surprising if polyp size did not also increase with biomass as it does in the sympatric coral Montastrea curta (P. J. Edmunds, unpublished data). The possibility of variable polyp dimensions (specifically contracted length) is important, because it demonstrates that the assumption of conserved module size for colonial modular taxa (Jackson 1979; Hughes 2005) is incorrect for some juvenile corals. With the potential for variable polyp size, a mechanism is created through which scaling potentially can be altered through changes in body design (Schmidt-Nielsen 1989). Unfortunately, it is not possible to be certain of the extent to which this played a role in the present results, as the scaling of biomass in the corals exhibiting changes in growth scaling was inferred from freshly collected corals rather from those exposed to the treatments.
Second, there is strong evidence that the biomass of reef corals is plastic. For example, coral biomass varies among seasons (Brown et al. 1999, Fitt et al. 2000), and locations differing in environmental stress (Barnes and Lough 1992, 1999). It is also affected by feeding regimes (Anthony et al. 2002), and the sizes of both colonies (Vollmer and Edmunds 2000) and polyps (Elahi and Edmunds 2007b). To a large extent, such variation reflects the consequences of the net energy balance and strategies of energy allocation, such that biomass increases when an energy surplus is realized (Anthony et al. 2002) and nutrients are available to support protein synthesis (Davies 1984). In contrast, biomass can decline when energy demands exceed supply. For instance, tissues reserves can fuel coral growth when food supplies are limiting (Grottoli et al. 2006), and be used to meet the energy demands for reproduction (Szmant and Gassman, 1990). In addition to serving as a plastic food reserve, the three-dimensional structure created by coral biomass can also create beneficial microhabitats in which Symbiodinium can be accommodated (Enriquez et al. 2005) and gametes stored (Hall and Hughes 1996). Overall, there is sufficient evidence to conclude that variation in biomass is a general characteristic of tropical scleractinians with potentially strong selective value.
Third, it is likely that tissue thickness has consequences for the rate of transport of metabolites across this layer. Such transport could be accomplished by diffusion (Patterson 1992), active transport (Gattuso et al. 1999), or flagellum-mediated flow within the gastrovascular cavity (Gladfelter 1983). Thus, the rate of transport across this layer will depend on how the transport mechanisms are affected by physical conditions, as well as the thickness of the layer. In the case of elevated temperature, the effects on metabolite transport initially is likely to be stimulatory, at least based on what is known from other eukaryotic systems for the effects of temperature on diffusion (Raven and Geider 1988), active transport (Raven and Geider 1988; Hochachka and Somero 2002), and the beating of flagella (Alavi and Cosson 2005). Because such mechanisms play important roles in the translocation of metabolites across coral colonies (Oren et al. 2001), it is conceivable that the effects of temperature on these processes might be sufficient to affect scaling, as do changes in fluid-mediated modular integration in a colonial ascidian (Nakaya et al. 2003). However, rising temperature also increase rates of physiological functions up to their thermal optima (Hochachka and Somero 2002), and this increases the requirements for metabolites, particularly in corals with greater biomass. With increased demand for metabolites, there is the potential to outstrip the ability to satisfy them, thereby creating rate-limitation of physiological functions that covary with temperature.
For tropical reef corals, respiration, photosynthesis, and calcification provide good examples of the stimulatory effects of temperature, with all increasing with temperature until their respective thermal optima are reached (Buddemeier and Kinzie 1976; Coles and Jokiel 1977; Iglesias-Prieto et al. 1992). Empirical data describing these effects still are limited for many functions, but it is not unreasonable to anticipate that an increase from 26 to 28°C (close to the present treatments levels) would elevate dark respiration by 22% (Edmunds 2005a) and photosynthesis by ≈20% (Iglesias-Prieto et al. 1992). The effects of such a rise in temperature on calcification are less clear, because a range of 26–28°C is likely to include the thermal optimum for this process (Buddemeier and Kinzie 1976). Nevertheless, a temperature coefficient of 8–9%/°C is not unreasonable for coral calcification (Buddemeier and Kinzie 1976), with the direction of the effect depending on whether the temperature changes are above, or below, the thermal optimum. Such changes in physiological rates create a context in which metabolite transport could become rate limiting. Interestingly, rate limitation already has been described for physiological processes in corals, for instance, when dark respiration at low flow speeds is impeded by the delivery of oxygen (Patterson et al. 1991). In the light, rate limited removal of oxygen from coral tissue can lead to oxygen accumulation and impaired photosynthesis by the symbiotic Symbiodinium (Finneli et al. 2006).
In summary, an attractive hypothesis to explain the present results [as well as those from the field (Edmunds 2006)] is that the thickening of tissues with increasing size for juvenile corals causes the supply of metabolites to become limiting at high temperature. Clearly, there are critical aspects of the hypothesis that require attention, notably to explore the taxonomic and biogeographic generality of the temperature-growth scaling phenomenon, and to test experimentally for rate-limitation of physiological processes through restricted metabolite transport. Potentially this might be accomplished through the measurement of the growth of juvenile corals under differing metabolite concentrations, such as could be created easily for oxygen through the use of hyperoxic and hypoxic treatments.
This research was supported by grant OCE 04-17412 from the National Science Foundation, and gifts from the Gordon and Betty Moore Foundation; it was completed under a research permit issued by the French Polynesian Ministry of Research. I am grateful to N. Davies and the staff of the U.C. Berkeley, Richard B. Gump South Pacific Research Station for making my visits to Moorea productive and enjoyable, M Murray for outstanding field support, and my graduate assistants for help in multiple aspects of this project. Three anonymous reviewers together with N. Muehllehner and H. Putnam provided valuable comments that improved an earlier draft of this paper. This is a contribution of the Moorea Coral Reef (MCR) LTER Site, and is contribution number 185 of the Marine Biology Program of California State University, Northridge.
- Anthony KRN, Connolly SR, Willis BL (2002) Comparative analysis of energy allocation to tissue and skeletal growth in corals. Limnol Oceangr 47:1417–1429Google Scholar
- Barnes DJ, Chalker BE (1990) Calcification and photosynthesis in reef-building corals and algae. In: Dubinsky Z (ed) Ecosystems of the world, vol 25. coral reefs. Elsevier, New York, pp 109–131Google Scholar
- Birkeland C (1976) An experimental method of studying corals during early stages of growth. Micronesica 12:319–322Google Scholar
- Buddemeier RW, Kleypas JA, Aronson RB (2004) Coral reefs and global climate change. Pew Center on Global Climate Change ReportGoogle Scholar
- Buddemeier RW, Kinzie RA (1976) Coral growth. Ocean Mar Biol Ann Rev 14:183–225Google Scholar
- Davies PS (1984) The role of zooxanthellae in the nutritional energy requirements of Pocillopora eydouxi. Coral Reefs 2:181–186Google Scholar
- Dove SG, Hoegh-Guldberg O (2006) The cell physiology of coral bleaching. In: Phinney JT, Hoegh-Guldberg O, Kleypas J, Skirving W, Strong A (eds) Coral reefs and climate change: science and management. American Geophysical Union, Washington, pp 55–71Google Scholar
- Edmunds PJ (2007) Evidence for a decadal-scale decline in the growth rates of juvenile scleractinian corals. Mar Ecol Prog Ser (in press)Google Scholar
- Elahi R, Edmunds PJ (2007b) Determinate growth and the scaling of photosynthetic energy intake in the solitary coral Fungia concinna (Verrill). J Exp Mar Biol Ecol (in press)Google Scholar
- Gattuso JP, Allemand D, Frankignoulle M (1999) Photosynthesis and calcification at cellular, organismal and community levels in coral reefs: a review on interactions and control by carbonate chemistry. Am Zool 39:160–183Google Scholar
- Hochchka PW, Somero GN (2002) Biochemical adaptations. Oxford University Press, OxfordGoogle Scholar
- Hughes RN (2005) Lessons in modularity: the evolutionary ecology of colonial invertebrates. Sci Marina 69:169–179Google Scholar
- Jackson JBC (1979) Morphological strategies of sessile animals. In: Larwood G, Rosen BR (eds) Biology and systematics of colonial organisms. Academic Press, London, pp 499–555Google Scholar
- Oren U, Beneyahu Y, Lubinevsky H, Loya Y (2001) Colony integration during regeneration in the stony coral Favia favus. Ecology 82:802–813Google Scholar
- Patterson MR, Sebens KP, Olson RR (1991) In situ measurements of flow effects on primary production and dark respiration in reef corals. Limnol Oceanogr 36:936–948Google Scholar
- Schmidt-Nielsen K (1989) Scaling: why is animal size so important? Cambridge University Press, CambridgeGoogle Scholar
- Sebens KP (1987) Coelenterata. In: Pandian TP, Vernberg FJ (eds) Animal energetics. Academic Press, New York, pp 55–120Google Scholar
- Smith SR (1992) Patterns of coral recruitment and post-settlement mortality on Bermuda’s reefs: comparisons to Caribbean and Pacific reefs. Am Zool 32:663–673Google Scholar
- Sokal RR, Rohlf FJ (1995) Biometry. Freeman and Company, New YorkGoogle Scholar