CALIFORNIA STATE UNIVERSITY, NORTHRIDGE
EFFECTS OF AGE AND SIZE ON THE GROWTH AND PHYSIOLOGY OF SCLERACTINIAN CORALS
A thesis submitted in partial fulfillment of the requirements For the degree of Master of Science in Biology
By
Robin Elahi
December 2005
The thesis of Robin Elahi is approved:
______Robert C. Carpenter, Ph.D. Date
______Steven R. Dudgeon, Ph.D. Date
______Brian S. T. Helmuth, Ph.D. Date
______Peter J. Edmunds, Ph.D., Chair Date
California State University, Northridge
ii Acknowledgements
The completion of this thesis would have been impossible without the support of many people, and to all I am grateful, including numerous individuals not specifically acknowledged. First, I would like to thank my primary advisor, Dr. Pete Edmunds, for the opportunity to join the Polyp Lab. I am a better scientist, and person, as a direct result of the countless enjoyable hours we shared in the classroom, field and laboratory. The development and execution of this project benefited greatly from the many discussions I had with my committee members, Drs. Bob Carpenter, Steve Dudgeon, and Brian Helmuth. Thoughtful insight was provided by Drs. Ruth Gates, Dave Gray, Fritz Hertel, and Paul Wilson on several occasions. Many people assisted with fieldwork, especially Christina Buck and Mike Murray, as well as Laurie Allen-Requa, Danny Green, Joshua Idjadi, Sarah Lee, Mairead Maheigan, and students of the Threes Seas XX and XXI program. The staff at the Discovery Bay Marine Laboratory and the Richard Gump Station was very hospitable and helpful. The Northeastern University Three Seas Program spearheaded my graduate career, supported my research efforts extensively in Jamaica and Moorea, and I thank Dr. Sal Genovese for his help over the past three years. This research was additionally funded by P.A.D.I Project Aware, CSUN University Corporation (#620200), CSUN Associated Students, CSUN Graduate Studies, Research & International Programs, Sea Grant Program of the University of Puerto Rico (#R-101-2-02), NSF-LTREB (DEB 03443570), and NSF-LTER (NSF-OCE 041412). Vickie Everhart, Linda Gharakhanian, and Cherie Hawthorne in the CSUN Biology Department made my life easier by resolving many bureaucratic issues. My CSUN peers reminded me there was more to life than research; in particular I thank Kylla Benes, Peter Holmquist, Becca Kordas, Pavel Lieb, Mairead Maheigan, Kathy Morrow, and Mike Murray for their friendship. Finally, I am grateful to my parents for unconditionally supporting my endeavors. I thank my mother for instilling in me a sense of adventure, and my father for pushing me academically. Without their strong influence my life would be far less interesting.
iii TABLE OF CONTENTS
Signature page ii
Acknowledgements iii
Abstract v
Chapter 1 - General introduction 1
Chapter 2 - The effects of age and size on the physiology of the branching coral, Madracis mirabilis
I. Introduction 7 II. Methods 10 III. Results 15 IV. Discussion 17 V. Tables and figures 23
Chapter 3 - The consequences of fission in the massive coral Siderastrea siderea: growth rates of small colonies and clonal input to population structure
I. Introduction 28 II. Methods 31 III. Results 35 IV. Discussion 37 V. Tables and figures 40
Chapter 4 - Energetic constraints on indeterminate growth in the solitary coral, Fungia concinna
I. Introduction 44 II. Methods 49 III. Results 60 IV. Discussion 65 V. Tables and figures 70
Chapter 5 - Concluding remarks 79
Literature Cited 83
iv ABSTRACT
EFFECTS OF AGE AND SIZE ON THE GROWTH AND PHYSIOLOGY OF SCLERACTINIAN CORALS
By
Robin Elahi
Master of Science in Biology
The growth of modular organisms is achieved by the asexual iteration of conserved units, and the biological implications of this type of growth are vast. One direct consequence of modularity is the potential for exponential growth through asexual reproduction and dispersal, thereby removing the genotype from the physiological constraints of senescence and permitting it to become virtually immortal. However, senescence at the level of individual modules may still exist. Scleractinian corals are an excellent model system to test for effects of age and size because colonies often experience fission, fusion, and fragmentation, thereby decoupling the relationship between age and size. Understanding how fission and fragmentation affect coral growth is timely because the likelihood of partial mortality and fission will increase due to global degradation of coral reefs, resulting in large numbers of small, yet old, colonies. In order to test the effects of age and size on growth in corals, two approaches were taken. First, age and size were manipulated experimentally by breaking branches of the coral, Madracis mirabilis, into young and old fragments, and growth subsequently was quantified as calcification rate. Growth scaled isometrically in both age groups, and although scaling exponents were statistically indistinguishable among ages, young fragments calcified faster than old fragments. In other words, the effect of age was absolute and independent of size. The second approach involved a mensurative analysis of the massive coral, Siderastrea siderea. This species often undergoes fission to produce small daughter colonies that are old. The growth of similarly sized sexual recruits (young) and daughter colonies (old) was monitored for a year, and these two colony types exhibited significant differences in lateral extension. Furthermore, age affected the scaling of calcification so that young corals grew disproportionately faster than old corals over the smallest size range. Together, the experiments with M. mirabilis and S. siderea demonstrate that age significantly affects coral growth, and suggest that the rapid growth of juvenile corals can be attributed to their young age, rather than their small size. Although most scleractinian corals are modular, certain species are not, and thus the relationship between age and size typically is predictable. In contrast to colonial corals that grow indeterminately and are typically only limited by space, solitary species are likely to exhibit an upper maximum size that may result from energetic constraints. An energetic model originally developed for anemones was modified in order to test the hypothesis that energetic constraints limit the maximum size of the solitary coral Fungia concinna. The model assumed that photosynthesis was the primary source of energetic intake and metabolic cost was quantified as aerobic respiration. The scaling exponent on
v mass was higher for energetic intake than metabolic cost, allowing large individuals to maintain an energetic surplus over the size range studied, even when the energy required for daily host tissue and symbiont growth was incorporated into the model. Therefore, it appears that growth in F. concinna is not limited energetically. Instead, mechanical constraints on locomotion may set the maximum size of this solitary coral. The results of these three studies demonstrate clearly that age and size separately affect the physiology of solitary and modular corals, but also highlight the potential for interactive effects of these two demographic parameters that likely are under strong selective pressure.
vi CHAPTER 1
General introduction
Organisms that grow by the repeated iteration of multicellular parts, or modules, are defined as modular (Harper 1977). In contrast, solitary organisms grow by increasing the size of a single unit, which typically begins as a zygote that first undergoes development. A variety of organisms display modularity (i.e., clonality), including algae, higher plants, fungi, cnidarians, bryozoans and ascidians. The biological implications of modularity are vast, and have been reviewed by Jackson et al. (1985) and by many authors in an issue of the Philosophical Transactions of the Royal Society of London B devoted solely to the growth and form of modular organisms (Harper et al. 1986). Some of the recent work on modular invertebrates has focused on the morphological integration of complex colonial forms (Gateño & Rinkevich 2003, Nakaya et al. 2003, Sánchez
2003, Kaandorp et al. 2005), allorecognition responses of self and non-self clones
(Chadwick-Furman & Weissman 2003, Cadavid 2004), and the constraints on modular growth (Kim & Lasker 1998, Tanner 1999, 2002).
One direct consequence of modularity is the potential for genetic individuals
(hereafter referred to as genets) to grow exponentially through asexual reproduction, whereas solitary organisms are typically subject to allometric constraints on size (Gould
1966, Schmidt-Nielsen 1984). As a result, clonal organisms can become locally abundant in both terrestrial and marine habitats. Modular design also allows the genet to fragment into physically unattached, viable individuals of the same genotype (hereafter referred to as ramets). This can provide the genet with asexual dispersal opportunities,
1 and can spread the risk of mortality over a large area, as observed for clonal plants
(Harper 1985) and many benthic invertebrates (Jackson 1985). As a result, the clonal propagation of ramets can theoretically confer “immortality” upon a genet (Jackson &
Coates 1986). In other words, genets may not experience an increase in mortality rates with advancing age (Caswell 1985), which is the original definition of senescence
(Medawar 1952). In accordance with the physiological nature of this thesis, an alternative, but not mutually exclusive, definition of senescence will be used: the decline in organismal performance and fitness with increasing age (Hughes & Reynolds 2005).
Although genet senescence may be avoided by the asexual iteration of new modules, there is considerable evidence for individual modules to experience physiological deterioration with age. Most modular organisms display a pattern of iteration that resembles a branching structure (Harper 1985), such that distal modules towards the ends of branches are younger than the proximal modules. Comparisons of distal and proximal modules are thus frequently considered as age contrasts. For example, photosynthesis and calcification rates of the green alga Halimeda decrease with age, i.e., they are highest at tips and lowest at bases in segmented branches (Borowitzka
& Larkum 1976). The proximal shoots and leaves of clonal plants senesce and are shed regularly (Watt 1947, Leopold 1961, Watkinson & White 1986). Old zooids of the bryozoan Steginoporella exhibit decreased feeding rates and slower regeneration of broken colony margins than young zooids (Palumbi & Jackson 1983). Tissue regeneration decreases with age in scleractinian corals (Meesters & Bak 1995).
Senescence in fungi results in the cessation of vegetative growth rates, and appears to be related to errors in cytoplasmic protein synthesis (Holliday 1969). Based on the evidence
2 from various diverse phyla, senescence appears to be a universal biological phenomenon for the modules of clonal organisms.
Scleractinian corals are an excellent model system to test the effects of age on modular organisms because the age of a colony often is decoupled from its size due to fission, fusion and fragmentation (Hughes & Jackson 1980). Although the size dependence of coral growth, survival, fecundity and physiological traits has been relatively well-studied (Hughes 1984, Hughes & Jackson 1985, Jokiel & Morrissey 1986,
Kinzie & Sarmiento 1986, Babcock 1991, Barnes & Lough 1992, Soong 1993, Kim &
Lasker 1998, Yap et al. 1998, Hughes & Tanner 2000, Vollmer & Edmunds 2000,
Edmunds & Gates 2004), the effects of age have received less attention (but see Kojis &
Quinn 1985, Rinkevich & Loya 1986, Hughes & Connell 1987, Babcock 1991, Meesters
& Bak 1995).
Despite the lack of study, the relative effects of age and size on fundamental aspects of coral biology are relevant ecologically because coral populations are comprised of sexual recruits and asexually derived fragments that vary greatly in both age and size. Furthermore, the likelihood of coral partial mortality, fission and fragmentation will probably increase, because corals and coral reefs are undergoing degradation at global scales due to natural and anthropogenic disturbances (Hughes et al.
2003, Bellwood et al. 2004). In support of this prediction, rates of fission for the
Caribbean reef building coral, Montastraea annularis, increased between 1977 and 1993 in Jamaica, creating large numbers of small, yet old, colonies (Hughes & Tanner 2000).
The probability of fission also has increased for M. annularis in St. John over the past 10 years (Edmunds & Elahi in review), which is concordant with the regional scale of coral
3 decline (Gardner et al. 2003). Given their role as the ecosystem engineers of coral reefs
(Jones et al. 1997), there is a clear need to understand better the relative effects of age and size on the growth of corals.
Although the majority of scleractinian corals are modular, there are certain solitary representatives. Some of the most ecologically important species are corals within the family Fungiidae. Fungiids are common on many Indo-Pacific and Red Sea reefs (Wells 1966), and even can build enduring reef structures (Littler et al. 1997).
Further, fungiids are unusual because most are free-living as adults (Veron 2000), and many are mobile, particularly representatives of the genus Fungia (Hubbard 1972,
Chadwick 1988, Chadwick-Furman & Loya 1992). Mobility may play an important role in the selection of favorable habitats because zooxanthellate fungiids exhibit positive phototaxis (Yamashiro & Nishihara 1995), which is an advantageous behavior because zooxanthellate corals receive photosynthetically fixed carbon from their symbiotic dinoflagellates (Muscatine & Hand 1958, Muscatine et al. 1981).
Solitary fungiids fundamentally are different from modular corals because they appear to exhibit indeterminate growth through the increase of polyp size rather than by modular iteration. Therefore, size typically is a reliable predictor of age (Chadwick-
Furman et al. 2000) and the demographic complications pertaining to colonial scleractinians (Hughes & Jackson 1980) largely are avoided. However, the contrast to modular corals allows various predictions to be made about the growth and physiology of solitary corals. The characteristic of increasing polyp size suggests that larger polyps possess a competitive advantage over smaller polyps, because selection should favor coloniality otherwise (Sebens 1987b). For example, large polyps of the solitary anemone
4 Anthopleura xanthogrammica expand their prey size range by consuming larger mussels as they grow, and thus obtain an energetic advantage (Sebens 1981). Additionally, the growth of fungiids probably is not limited by competition for space (Jackson 1977) due to their mobility (Chadwick 1988), but instead may be constrained by the weight of their skeleton (Chadwick-Furman & Loya 1992), or by energetics (Sebens 1982).
A useful method of quantifying the effects of size on physiological traits is logarithmic linear regression, in which size is the independent variable and the physiological trait is the dependent variable. Depending on the slope of the regression, traits can be considered to scale isometrically (b =1), or allometrically (b ≠ 1) [Schmidt-
Nielson 1984]. Isometry indicates a directly proportional relationship between the two variables, whereas allometry demonstrates that the physiological trait changes disproportionately with size. This approach is commonly referred to as a scaling relationship, and it has enjoyed a long history of investigation, dating back to 1897
(Gayon 2000). At least for unitary organisms, most traits scale allometrically with size due to geometric or mechanical constraints (Schmidt-Nielsen 1984, Brown et al. 2000).
For example, metabolism typically scales allometrically as a result of surface to volume constraints (West et al. 1999). In contrast, clonal organisms are thought to escape the constraints of allometry because module size is conserved and growth is achieved by asexual iteration (Jackson et al. 1985). Although rare, empirical tests of isometry in clonal organisms have yielded contrasting results, and probably are dependent on the specific organism (Hughes & Hughes 1986, Sebens 1987a, Vollmer & Edmunds 2000).
In this thesis, I address the effects of age and size on the growth and physiology of three scleractinian corals. In Chapter 2, manipulative experiments in Discovery Bay,
5 Jamaica are used to test the hypothesis that age affects calcification rates in the branching coral Madracis mirabilis. In Chapter 3, mensurative surveys test the hypothesis that young sexual recruits grow faster than old fragments of similar sizes, and describe the contribution of ramets to the population structure of the massive coral, Siderastrea siderea, in Great Lameshur Bay, St. John. Finally, a test of the hypothesis that growth of the solitary coral, Fungia concinna in Moorea, French Polynesia, is limited by energetic constraints is presented in Chapter 4.
6 CHAPTER 2
The effects of age and size on the physiology of the branching coral, Madracis mirabilis
Introduction
The modular design of many benthic marine invertebrates, including cnidarians, bryozoans and ascidians, has profound implications for their biology (Jackson et al.
1985). In particular, the ecological success of modular invertebrates on subtidal hard substrata is related directly to their ability to spread across the benthos by the rapid, asexual iteration of modules (e.g., polyps, zooids), which together comprise individual colonies of physiologically integrated units. One direct consequence of this asexual iteration is a difference in age among modules of a colony, with the youngest tissue located at sites of active module proliferation, which typically occurs at peripheral margins (Boardman & Cheetham 1973). Thus while a colony has an absolute age defined by the elapsed time since larval settlement, it consists of a spatial mosaic of modules differing in age as defined by the elapsed time since their asexual origin.
To date, the effects of age on the biology of modular invertebrates largely have been overlooked, partially because individual genotypes have the ability to persist indefinitely through asexual propagation, and thus are regarded as “immortal” (Jackson &
Coates 1986). In addition, colony age is difficult to determine because colony fission, fusion, and fragmentation can uncouple the relationship between age and size (Hughes &
Jackson 1980), and therefore age can only be determined reliably by tracking individual recruits through time. In spite of these issues, several studies have demonstrated significant physiological effects of age in modular organisms. For example, old zooids in
7 the bryozoan Steginoporella exhibit decreased feeding and regenerative abilities relative to young zooids (Palumbi & Jackson 1983), and in scleractinian corals, age has significant effects on reproduction (Kojis & Quinn 1985) and the regeneration of tissue lesions (Meesters & Bak 1995). These studies suggest that modular invertebrates cannot escape the nearly universal physiological constraints set by advancing age as shown by other multicellular animals (Hughes & Reynolds 2005).
Although the effects of age on modular invertebrates are unclear, the effects of the environment in driving phenotypic plasticity are well-documented (Kaandorp 1999,
Anthony & Hoegh-Guldberg 2003), and such plasticity is widespread among sessile organisms (Schlichting & Pigliucci 1998). Much of the research attention accorded to phenotypic plasticity has focused on the interaction between genotypes and the environment, where a single genotype displays a gradient of phenotypes in response to an environmental gradient (i.e., a norm of reaction, sensu Schlichting & Pigliucci 1998).
Typically, reaction norms have been examined during adult stages (Schlichting 1986), but recent studies highlight the potential importance of ontogenetic shifts in the type and extent of plasticity (Pigliucci & Schlichting 1995, Arnqvist & Johansson 1998, Ostrowski et al. 2002). In other words, the phenotypic response of organisms to their environment may vary with age.
The objective of the present study was to examine how age affects the phenotypic expression of calcification in scleractinian corals. Due to the difficulties associated with determining the age of coral colonies (Hughes & Jackson 1980), we exploited the physiological age gradient created along the branches of a coral by apical growth to create a relative age contrast between proximal (older) and distal (younger) portions of
8 branches. Specifically, we tested whether: (1) calcification is altered by tissue age, and
(2) corals of varying age exploit plasticity to differing degrees by altering calcification rates under new environmental conditions. Calcification is the physiological basis of skeletal growth in scleractinian corals (Gattuso et al. 1999), and is an indirect measure of fitness due to strong inverse size-dependent mortality (Hughes & Jackson 1985) and positive size-dependent fecundity (Soong 1993). To achieve these objectives, two factorial transplant experiments were conducted. The first experiment was designed to separate and test the effects of age, size, and genotype on the plasticity of calcification.
The second experiment tested solely the effects of age and size on the plasticity of calcification, and also tested for intrinsic differences in Symbiodinium density and tissue biomass between young and old fragments. If such differences are present, they could potentially provide important insight into the mechanistic basis for a physiological age contrast.
9 Methods
The Caribbean scleractinian coral Madracis mirabilis (Duchassaing and
Michelotti) was selected as a model system because it is abundant in a variety of environments (Bruno & Edmunds 1997), and because it has a branching morphology that facilitates an experimental analysis of age. Branches were collected from aggregates
(i.e., genets; defined as an aggregation of clonal branches connected by a common skeleton but not by tissue) separated by >15 m distance to increase the probability of sampling genetically distinct individuals (Bruno & Edmunds 1997). Aggregates were sampled from 13 to 18-m depth at Columbus Park, Jamaica, where branches of M. mirabilis are long, slender and widely-spaced (Bruno & Edmunds 1997, Sebens et al.
1997), and therefore well-suited to preparing experimental units (fragments) varying in age depending on the distance from the growing tip.
Because M. mirabilis grows primarily by apical extension (Bruno & Edmunds
1997), the youngest tissue is located at the distal ends of branches, and the proximal portions of branches are covered by relatively older tissue. Thus, by breaking branches in two, physiologically young (i.e., distal) and old (i.e., proximal) fragments can be created.
Based on skeletal extension rates of 2.2 ± 0.5 cm yr-1 (mean ± SE; n = 12) from 10-m depth at Columbus Park (calculated from Fig. 6 in Bruno & Edmunds 1997), young fragments 2.5 cm in length ranged in age from several weeks at their tip to 1.3 years old at their base, and old fragments 2.5 cm in length ranged in age from 1.3 years at their tip to 2.6 years old at their base. These age estimates likely are conservative because coral growth rates decrease with depth (Huston 1985) and the branches in this study were collected at depths greater than those used by Bruno and Edmunds (1997). The curved
10 tips of distal fragments were removed so that neither old nor young fragments possessed tissue-covered tips. Each fragment was epoxied (Z-spar® Splash Zone A-788) upright to a plastic tile (4 x 4 cm), which then was transplanted to a platform made of cement blocks at one of two depths (described below) on the same reef from which they were collected.
To consider fragments statistically independent, tiles were spaced so that fragments were
5 cm apart and exceeded the branch spacing found naturally in aggregations at 10-m depth at Columbus Park (~ 1 cm; Bruno & Edmunds 1997) and at 20-m depth on the forereef at Discovery Bay (1.7-1.9 cm; Sebens et al. 1997).
To test the hypothesis that the plasticity of calcification rate varied with coral age
(ontogenetic plasticity, sensu Pigliucci & Schlichting 1995), the young and old fragments were transplanted to the Columbus Park reef and distributed randomly to one site at the collection depth (17 m), or to a second site in shallower water (9 m). We designed the
14-week experiment to be analyzed with a factorial three-way analysis of covariance testing the main effects of age (fixed), depth (fixed), and genet (random) using surface area as the covariate. Subsequently, a similar 3-week experiment was designed to test solely the fixed effects of age and depth using surface area as the covariate. Significant first-order interactions between age and genet with depth would represent ontogenetic and phenotypic plasticity, respectively, indicating that corals varying in age and genotype exhibit dissimilar responses to changing environmental conditions. Scaling relationships between calcification rate and coral surface area also were quantified for corals in each treatment in order to test the hypothesis that calcification scales isometrically, and that this relationship does not vary with the age of the coral. Because calcification may be mass-transfer limited (Dennison & Barnes 1988) and is a surface area phenomenon
11 (Gattuso et al. 1999), we expected it to scale isometrically with surface area. If, however, we observed an allometric scaling of calcification with surface area, then this would indicate that small fragments calcify at disproportionately faster rates than larger fragments, regardless of age. The surface area of the skeleton covered by tissue was estimated geometrically from the area of a hollow cylinder having dimensions equal to the length and mean diameter of each fragment. Preliminary analyses demonstrated that geometric estimates of surface area were indistinguishable statistically (paired t-test, t = 0.551, df = 23, p > 0.5) from areas estimated using the foil wrapping method (Marsh
1970).
Calcification rate (mg day-1) was quantified by determining skeletal weight at the beginning and end of each transplant period using the buoyant weight method (Davies
1989). Each fragment was weighed (± 1 mg) in seawater after removing epiphytes, and the change in buoyant weight converted to dry skeletal weight using equations from
Davies (1989), assuming the density of aragonite to be 2.93 mg cm-3.
The effects of age, depth, genet, and size on calcification were tested during a fourteen-week period spanning the winter and spring of 2004. Branches of M. mirabilis were collected from each of three aggregates at 13 to 17-m depth, and 7 – 8 fragments of each aggregate and age group were weighed and transplanted to one shallow (9 m) and one deep site (17 m) on 5 March 2004. The collection, weighing and transplanting procedure was accomplished within 48 hours and thereafter the corals were left undisturbed until retrieval on 13 June 2004 when they again were buoyant weighed. A second transplant experiment was conducted for three weeks in the summer of 2004 to ensure that the outcome of the age contrast was not unique to the three aggregates
12 originally sampled, and to test for intrinsic physiological differences between young and old fragments. For the 3-week experiment, three young and three old fragments were generated from each of 12 aggregates of M. mirabilis collected from 15 to 18-m depth at
Columbus Park reef. After buoyant weighing, 18 fragments of each age class were allocated haphazardly (without regard to genet) to shallow (9 m) and deep (17 m) transplant treatments on 13 June 2004; they were retrieved and reweighed on 3 July 2004.
To determine if the age groups exhibited differences in tissue content and population densities of symbionts, 12 fragments of each age class were sacrificed prior to the start of the 3-week experiment. Coral tissue was removed from the skeletons using a
Waterpik® filled with 0.45 µm filtered seawater (Johannes & Wiebe 1970) and used for the determination of Symbiodinium density and protein biomass. Symbiodinium densities were quantified for each age class (n = 12) using a hemocytometer (10 replicate counts/sample). Protein was solubilized from aliquots of coral slurry (1.8 ml) by incubating with 1M NaOH at 50°C for five hours, followed by the addition of 1M HCl to neutralize the pH of the slurry. Duplicate subsamples (0.8 ml) of the incubated slurry were assayed for protein using the Coomassie Brilliant Blue dye binding procedure
(Bradford 1976) with standards prepared from bovine serum albumin in filtered seawater.
For the 14-week experiment, a three-way analysis of covariance (ANCOVA) was used to test for the effects of age (A), depth (D) and genet (G) using the log of the surface area (SA) as the covariate and the log of the calcification rate as the dependent variable.
For the 3-week experiment, a two-way ANCOVA tested the fixed effects of age and depth using the log of the surface area as the covariate. The mean square and degrees of freedom of statistical interactions (e.g., A x D x G, G x SA) were pooled into the error
13 term when the interactive effects were not significant statistically (p > 0.25; Quinn &
Keough 2002). Type I sum of squares was used in all analyses.
Logarithmic linear regressions using a measure of body size as the independent variable and a physiological trait as the dependent variable are recognized commonly as scaling relationships (Schmidt-Nielsen 1984), which can be described as isometric or allometric depending on the slope of the linear regression (i.e., the scaling exponent).
Reduced major axis (RMA) regression was used to calculate the scaling exponent (b) for each regression because surface area was a random variable potentially estimated with error (Quinn & Keough 2002). The scaling exponent for each treatment was tested against the null hypothesis that b = 1 (isometry) using a t-test (Sokal & Rohlf 1995).
Standard errors for the regression slopes were taken from ordinary least squares (OLS) analyses, because the variance of OLS and RMA estimators are identical to the third significant digit (McCardle 1988). Because RMA ANCOVA techniques are not available, ordinary least squares ANCOVA was used to test for the main and interactive effects (Sokal & Rohlf 1995). Assumptions for parametric testing were met by graphical inspection of residuals (Quinn & Keough 2002). The hypothesis that young and old fragments regenerated new tissue covered branch tips with equal frequency was tested using a 2 x 2 contingency table. A t-test was used to determine if Symbiodinium densities
(cells cm-2) and protein concentrations (µg cm-2) differed between the two age classes.
All statistical analyses were completed using JMP 5.01 for Macintosh.
14 Results
Coral fragments used in the 14-week experiment ranged in initial mean size
(surface area) from 7.0 to 8.8 cm2 (Table 2.1). Therefore, the exact age of fragments could not be fixed due to size variation, but a relative two-fold difference in physiological age remained consistent between the tissue of young and old fragments. During the experiment, these fragments calcified at a rate of 0.79-1.50 mg cm-2 day-1 (Table 2.1), and two individuals died. After 14 weeks, 35 (79.5%) young fragments possessed tissue covered branch tips, in contrast to nine (20.5%) old fragments, which is a greater number than chance alone would predict (Table 2.2).
Calcification scaled isometrically with surface area (b = 1) for all treatments
(Table 2.3), and scaling exponents were indistinguishable statistically among ages (Table
2.4). Young fragments calcified faster (15-30%) than old fragments as shown by the difference in elevation of the regression lines, but this difference was only marginally significant (Fig. 2.1a, Tables 2.1 & 2.4). Therefore, it appears that the effect of age was absolute and independent of size. The effect of depth on calcification also was significant for both age classes (Table 2.4), with fragments at 9-m depth calcifying 46-
65% faster than fragments at 17-m depth (Table 2.1). However, the test for ontogenetic plasticity, the age x depth interaction, was not significant (Table 2.4), therefore young and old corals expressed a similar response of calcification to depth treatments. The effect of genet and all associated interactions were not significant (Table 2.4), indicating that genotypes exhibited similar calcification responses to the experimental treatments.
Fragments used in the 3-week experiment ranged in initial mean surface area from
5.7 to 6.7 cm2, and calcified 1.81 to 2.47 mg cm-2 day-1 (Table 2.1). All fragments
15 survived the transplant experiment, but none extended new tissue over the broken surface of the branch tips. Calcification again scaled isometrically for all treatments (Table 2.3).
Relative to the 14-week experiment, the magnitude of the age effect (i.e., the difference between young and old normalized mean calcification rates) increased by 113% and
230% for corals at the shallow and deep depths, respectively (Table 2.1). However, the depth effect no longer was significant (Table 2.4), with corals calcifying at similar rates in shallow and deep environments (Fig. 2.1b, Table 2.1). The age x depth interaction for ontogenetic plasticity was not significant (Table 2.4).
Tissue characteristics did not vary with age in the fragments initially sacrificed from the 3-week experiment. Symbiodinium population densities were similar for both age classes (df = 22, t = 0.22, p = 0.83), with young and old fragments harboring 1.60 ±
0.14 and 1.61 ± 0.14 x 106 cells cm-2 (mean ± SE), respectively. Young fragments contained 153 ± 9 µg protein cm-2 (mean ± SE), which is indistinguishable statistically
(df = 22, t = 0.17, p = 0.87) from 156 ± 11 µg protein cm-2 (mean ± SE) contained by old fragments.
16 Discussion
This study demonstrates that an approximately two-fold difference in relative age has measurable effects on calcification in the scleractinian coral, Madracis mirabilis. To my knowledge, only two other experimental studies have demonstrated that age has biological consequences for scleractinian corals, specifically with respect to reproduction
(Kojis & Quinn 1985) and tissue regeneration (Meesters & Bak 1995). In addition, calcification rates of Stylophora pistillata decrease prior to the appearance of tissue mortality, thus indirectly supporting senescence, or the deterioration of physiological function with age (Rinkevich & Loya 1986). Notably, in Acropora palmata, the ability to regenerate tissue lesions decreased exponentially during the first 2-3 years of polyp life
(Meesters & Bak 1995), an age range equivalent to the relative difference in tissue age observed to affect calcification in the present study. Despite the cnidarian potential for cell renewal and exchange through gastrovascular connections between polyps (Crowell
1953, Martínez 1998, Müller et al. 2004), young polyps of branching scleractinian corals appear to be distinct physiologically from older polyps, exhibiting faster calcification and regeneration rates.
Calcification rates also varied along the distal 10 mm of branches of A. cervicornis, but these differences were attributed to translocation of organic compounds and chlorophyll content, rather than polyp age (Pearse & Muscatine 1971). Rapid calcification in apical polyps of A. cervicornis, which harbor very few symbiotic algae, is stimulated by photosynthesis occurring farther down the branch (Pearse & Muscatine
1971). Beyond the apical tip, overall calcification rates are much lower, but increase towards the base, correlating positively with chlorophyll content (Pearse & Muscatine
17 1971). In contrast, the branch tips of M. mirabilis do not exhibit the striking pale coloration of A. cervicornis tips (pers. obs.), and furthermore, the tips of the experimental fragments in this study were removed. Consequently, the age effect in this study cannot be explained by coincidental physiological differences arising from a branching morphology, at least based on the similarity of Symbiodinium population densities and protein biomass in fragments from both age classes. Furthermore, it is unlikely that the present results can be attributed to multiple taxa of Symbiodinium, because M. mirabilis harbors only clade B13 (Diekmann et al. 2003), even though a high diversity of symbiotic coral dinoflagellates occurs within the Caribbean (Baker 2003). The similar symbiont density and protein content of young and old fragments suggest that extrinsic factors, such as reduced light or flow, did not effect differences along branch axes within the aggregates sampled in this study, probably because the branches of aggregates at 17-m depth are widely-spaced. In contrast, localized differences of Symbiodinium population density and chlorophyll concentration along the branches of tightly-spaced colonies of
Agaricia tenuifolia were thought to be photoacclimatory responses to light gradients
(Helmuth et al. 1997).
The effect of age on calcification rates was absolute and independent of size in both experiments with M. mirabilis, because calcification increased proportionately with surface area for all fragments. The deposition of calcium carbonate is a surface area process in scleractinian corals, with calcium being taken up at the seawater-tissue interface, and aragonite being deposited at the two-dimensional interface of the aboral ectoderm and skeletal surface (Gattuso et al. 1999), and therefore it is not surprising that calcification scaled isometrically for M. mirabilis. Although modular organization
18 theoretically alleviates the surface to volume constraints which limit the growth of unitary organisms (Hughes & Cancino 1985), allometric limitations on resource capture can constrain isometric growth in modular organisms (Kim & Lasker 1998). Such limitation probably was avoided in the present study by spacing the experimental fragments of M. mirabilis far apart from one another, so that they probably did not engage in interference competition for light or particle flux. In contrast to our fragments, resource capture in naturally occurring aggregates of M. mirabilis likely is allometric because interior polyps of branching aggregates experience decreased light, flow and particle flux as a result of self-shading (Helmuth et al. 1997, Sebens et al. 1997).
Therefore, the observation of isometric scaling of calcification in this study may apply only to individual branches (e.g., recruits, fragments) of M. mirabilis, rather than entire aggregates. Additionally, the thickness of the boundary layer will differ with branch size and could lead to allometry if mass transfer is altered, but such an effect was not apparent in the present study.
Although young fragments calcified faster than old fragments, there was no interactive effect of age and depth on calcification rates. Factorial analysis of variance commonly has been used to test for the plastic responses of several genotypes in a series of environments (Schlichting 1986). The goal of this study was to investigate an age x environment interaction, to determine if plasticity varied ontogenetically. In the 14-week experiment, both young and old fragments increased their calcification rates when transplanted to a shallower depth, whereas in the 3-week experiment all corals calcified similarly at both depths. Therefore, there was no age related difference in plasticity, even though a relative age difference of one to two years was enough to detect a difference in
19 calcification rates between age classes. Although M. mirabilis exhibits genotype x environment effects with respect to morphological traits (Bruno & Edmunds 1997), calcification rates of genets appear to be independent of the environment (Bruno &
Edmunds 1997, this study), age, and size (this study). The lack of genotype effects in the present study accentuate the importance of age and size in affecting the calcification rates of M. mirabilis, and suggest that genets are under significant pressure to maintain similar growth rates in different environments.
This study demonstrates that old fragments of M. mirabilis calcify at slower rates than young fragments, but the causal basis for this age effect is unclear. The symbiosis between the coral and dinoflagellate complicates interpretation of this age effect. For example, the symbiont may experience some negative consequence of increasing age, or the communication between the host and symbiont may degrade with age. The lack of studies regarding either of these hypotheses precludes distinction between the two partners. However, with respect to the cnidarian host, differences in gastrovascular transport between modules of the young and old fragments of M. mirabilis may have contributed to the observed physiological differences. In support of this hypothesis, variation in gastrovascular pumping (Blackstone 1996) may mediate age-dependent differences in growth and competitive ability of two hydrozoan species (Van Winkle &
Blackstone 2002), but the applicability of this observation to a scleractinian is uncertain.
An alternative explanation relates to senescence of the coral (Rinkevich & Loya
1986, Meesters & Bak 1995). Previous authors (Palumbi & Jackson 1983, Meesters &
Bak 1995) have noted that senescence in old, proximal modules of clonal organisms is widespread phylogenetically, including algae (Borowitzka & Larkum 1976), plants (Watt
20 1947, Leopold 1961), fungi (Holliday 1969), and several metazoan lineages (Campbell
1968, Ryland 1979, Sabbadin 1979, Rinkevich et al. 1992). This observation suggests that an intrinsic cellular process, common to all taxa, explains senescence in modular organisms. Oxidative stress has been proposed to explain senescence in unitary organisms, based on studies of model organisms including nematodes (Caenorhabditis elegans), insects (Drosophila melanogaster), and rodents (Sohal et al. 2002, Hughes &
Reynolds 2005), and in scleractinian corals oxidative damage is well-known from studies of its role in the process of bleaching (Lesser 1997, Downs et al. 2002). It is tempting to suggest that at least some modules of scleractinian corals cannot escape the physiological constraints of age observed in unitary organisms (Downs et al. 2002), even though all cnidarians possess a reservoir of totipotent interstitial cells available for cellular turnover
(Müller et al. 2004) and some hydrozoans exhibit periodic cellular regeneration (Crowell
1953, Martínez 1998). Experiments explicitly addressing the oxidative stress hypothesis for senescence in corals are necessary, preferably using individuals whose absolute age can be measured accurately (e.g., sexual recruits).
Although the physiological basis of the age effect in M. mirabilis remains unclear, senescence of proximal modules in M. mirabilis and other scleractinian corals potentially is adaptive, and may play a role in controlling morphology. Slow tissue regeneration rates at the bases of A. palmata branches may facilitate bioerosion and subsequent fragmentation of branches large enough to enhance survivorship (Meesters & Bak 1995).
Likewise, the colonization of branch bases of M. mirabilis by boring sponges and algae
(Bruno & Edmunds 1997) may contribute to the generation of fragments typically larger than 5 cm (Fig. 1; Bruno 1998), which appears to be a threshold size for increased
21 survivorship (Fig. 6; Bruno 1998). The present study provides indirect evidence for this possibility, because very few of the old fragments of M. mirabilis extended new tissue over their broken tips during the 14-week experiment, and filamentous algae frequently settled onto this bare space. In contrast, 80% of the young fragments were able to form new tissue-covered tips. Although tissue-covered branch tips clearly are not a requirement for calcification in M. mirabilis, the apparent inability of old fragments to extend tissue over broken tips suggests that the age of tissue is an intrinsic influence on the apical growth observed in this species. In support of this hypothesis, the skeletal growth and form of Pocillopora damicornis branch tips are thought to be controlled by an intrinsic cycle in the overlying tissues (LeTissier 1988). Although traditionally overlooked, the age of coral tissue appears to have considerable consequences for the biology of at least M. mirabilis, and probably other coral species with developmental gradients associated with a branching morphology (Soong & Lang 1992). Furthermore, the physiological patterns observed in this study are consistent with the senescence of proximal modules observed in many other clonal organisms.
22 Table 2.1. Surface area normalized calcification rates and sizes for fragments of Madracis mirabilis by age and depth treatments. Results from the 14-week and the 3- week experiment are presented; all values are means ± SE.
14-week 3-week Cal ci fi cat i o n Su rfa ce area Calcification Surface area N N Treatment (mg cm-2 day-1) (cm2) (mg cm-2 day-1) (cm2)
Young/Shallow 23 1.50 ± 0.03 8.8 ± 0.3 18 2.47 ± 0.16 6.4 ± 0.5 Old/Shallow 22 1.30 ± 0.05 7.5 ± 0.6 18 1.81 ± 0.09 6.7 ± 0.5 Young/Deep 21 1.03 ± 0.04 7.7 ± 0.4 18 2.42 ± 0.12 6.0 ± 0.4 Old/Deep 23 0.79 ± 0.03 7.0 ± 0.4 18 1.91 ± 0.10 5.7 ± 0.3
23 Table 2.2. Contingency table for the number of fragments from each age class that extended new tissue to cover branch tips during the 14-week experiment. Tips were overgrown significantly more frequently in young fragments than old fragments (G = 31.11, df = 1, p < 0.001).
Covered tips Uncovered tips Young 35 10 Old 9 35
24 Table 2.3. Reduced major axis regression analyses used to estimate scaling relationships between log calcification (mg day-1) and log surface area (cm2) for fragments of Madracis mirabilis by age and depth treatments. Results presented for a) the 14-week, and b) the 3-week experiment. A t-test was used to test for departures from the null hypothesis of isometry (b = 1).
Slope (b) r2 df t p
a) 14-week Young/Shallow 1.01 0.76 21 0.09 >0.5 Old/Shallow 1.14 0.82 20 1.25 >0.2 Young/Deep 0.98 0.55 19 0.10 >0.5 Old/Deep 1.19 0.69 21 1.29 >0.2
b) 3-week Young/Shallow 0.90 0.24 17 0.43 >0.5 Old/Shallow 0.88 0.59 17 0.84 >0.5 Young/Deep 1.10 0.41 17 0.34 >0.5 Old/Deep 1.32 0.36 17 1.22 >0.2
25 Table 2.4. Results of a three-way analysis of covariance testing the effects of age (fixed), depth (fixed), and genet (random) on calcification rates (log mg day-1) in Madracis mirabilis during a) the 14-week experiment, and a two-way analysis of covariance testing the fixed effects of age and depth on calcification rates in Madracis mirabilis during b) the 3-week experiment. Mean squares and degrees of freedom of the interactions between main effects (age, depth, genet) and the covariate (log surface area) were pooled into the error term when p > 0.25.
Source df MS F p
a) 14-week Age (A) 1 0.2044 20.37 0.061 Depth (D) 1 0.8788 73.56 0.018 Genet (G) 2 0.0168 1.05 0.474 Log Surface Area (SA) 1 0.4554 97.64 <0.001 A x D 1 0.0076 1.62 0.207 A x G 2 0.0092 1.98 0.146 G x D 2 0.0111 2.37 0.100 A x SA 1 0.0108 2.32 0.132 D x SA 1 0.0072 1.53 0.220 Residual 76 0.0047
b) 3-week Age (A) 1 0.2503 25.70 <0.001 Depth (D) 1 0.0014 0.14 0.709 A x D 1 0.0002 0.02 0.890 Log Surface Area 1 0.4112 42.20 <0.001 Residual 66 0.0086
26
1.4 1.4 A B
Young/Shallow ) 1.2 1.2 Young/Shallow -1 Young/Deep
1.0 Old/Shallow 1.0 Old/Shallow
Young/Deep 0.8 0.8
Old/Deep 0.6 0.6 Young/Shallow Log calcification (mg day Old/Deep Old/Shallow Young/Deep Old/Deep 0.4 0.4 0.4 0.6 0.8 1.0 1.2 0.4 0.6 0.8 1.0 1.2 Log surface area (cm2) Log surface area (cm2)
Figure 2.1. Log calcification rate (mg day-1) versus log surface area (cm2) of young and old fragments of Madracis mirabilis in shallow (9 m) and deep (17 m) environments at Columbus Park, Jamaica. Two experiments were completed, during a) a 14-week period (6 March – 13 June 2004), and b) a 3-week period (13 June – 3 July 2004).
27 CHAPTER 3
The consequences of fission in the massive coral Siderastrea siderea: growth rates of small colonies and clonal input to population structure
Introduction
Scleractinian corals are the ecosystem engineers of coral reefs (Jones et al. 1997) and their growth is critical in maintaining the structural complexity of tropical reef habitat. Renewal of the complex reef framework is currently threatened because scleractinian corals are succumbing to bleaching events (Hoegh-Guldberg 1999), diseases
(Harvell et al. 1999), and a variety of other natural and anthropogenic disturbances
(Hughes et al. 2003, Pandolfi et al. 2003). In addition to killing colonies outright, these stresses can cause partial mortality of tissue leading to colonial fission (upon an intact skeleton) and fragmentation of the calcareous skeleton (and overlying tissue), because the modular organization of colonies provides physiological independence to each polyp
(Hughes & Jackson 1980). Consequently, coral populations typically consist of a mosaic of intact, sexually derived colonies and asexually derived fragments and daughter colonies.
The processes of fission, fusion and fragmentation have important consequences for the population dynamics of corals (Highsmith 1982, Hughes & Jackson 1985,
Babcock 1991). For example, some coral species rely on fragmentation as a method of asexual reproduction and dispersal (Highsmith 1982), but fission instead can have negative consequences. The population decline of three dominant coral species in
Jamaica between 1977 and 1993 was characterized by the mortality and fission of large colonies, resulting in large numbers of small colonies (Hughes & Tanner 2000). A direct
28 consequence of coral fission and fragmentation is the decoupling of the relationship between age and size, such that small fragments can be older than similarly-sized sexual recruits (Hughes & Jackson 1980). Fission has negative implications for coral growth rates, because the growth of corals is dependent on both their age and size (Hughes &
Connell 1987, Babcock 1991). The frequency with which coral colonies experience partial mortality and fission likely will increase, given the continuing degradation of coral reefs (Bellwood et al. 2004, Pandolfi et al. 2005). Therefore, environmental stress will affect coral populations through not only complete and immediate mortality, but also by affecting overall population growth rates through the creation of many small, yet old, colonies.
The primary goal of this study was to test the hypothesis that age affects growth in small corals (≤50 mm). Although the size range of corals was limited, it permitted an investigation of how age affects the scaling of growth. Scaling relationships examine the effect of body size on physiological traits using logarithmic linear regression (Schmidt-
Nielsen 1984), and depending on the slope of the regression (i.e., the scaling exponent), the relationship can be isometric (b = 1) or allometric (b ≠ 1). Physiological traits, such as metabolism, typically scale allometrically with body size in unitary organisms due to some physical limitation, including surface area to volume ratios (Gould 1966) and mechanical constraints (McMahon 1973). However, the modular organization and essentially two-dimensional structure of the soma in ascidian, bryozoan, and cnidarian colonies is thought to provide freedom from allometric constraints (Jackson et al. 1985).
In reality, scaling in modular organisms can be isometric or allometric, and probably is dependent on a variety of factors, including the trait in question, the organism itself, and
29 its ontogenetic stage (Hughes & Hughes 1986, Sebens 1987b, Vollmer 1999, Vollmer &
Edmunds 2000, Nakaya et al. 2003). In particular, the ontogenetic stage plays a critical role in affecting the scaling of metabolism and growth in several metazoans (Zeuthen
1953, Riisgård 1998). The juveniles of at least one scleractinian coral display allometric rates of growth, such that smaller individuals grow disproportionately faster than larger juveniles (Vollmer 1999), and may reflect a selective pressure to escape the mortality risks associated with small size (Jackson 1977). However, it is unclear whether rapid growth in juvenile corals is a function of their young age or small size. In order to distinguish between these two competing hypotheses, I examined the scaling of growth in young sexual recruits and old fragments. In addition, the relative contribution of fragments to a population of coral colonies was quantified to provide an ecological context for the growth hypothesis. The study was completed in Great Lameshur Bay, St.
John, which is located within the U.S. Virgin Islands National Park. The habitat is relatively protected from local anthropogenic disturbances such as overfishing and pollution, but coral populations still are susceptible to regional increases in seawater temperature (Edmunds 2004) and natural disturbances such as hurricanes (Witman 1992).
Therefore, the results of this study can be interpreted within the context of regional scale influences on the population dynamics of corals.
30 Methods
Study site and organism
The study was completed along the coast of Great Lameshur Bay, St. John, U.S.
Virgin Islands in 3 – 5 m depth. The habitat is dominated by slabs of granite rock covered with turf algae and small mounding coral colonies, as well as occasional large colonies (>20 cm) of Siderastrea siderea, Montastraea annularis and Diploria spp.
(Edmunds 2002). S. siderea was chosen as the study species because it is an ideal model system for studying the relative effects of tissue age and colony size on coral growth.
This species attains large sizes (>1 m) and possesses a massive skeletal morphology
(Foster 1979, Soong 1993). The hemispherically domed skeleton remains intact following tissue death, and thus retains evidence of partial mortality and fission (Babcock
1991). Therefore, comparisons of age can be made between intact colonies and similarly sized colonies that are older due to fission. At the study site, large colonies of S. siderea often exhibited partial tissue mortality, resulting in smaller colonies separated by turf algae and the bare skeleton of the parent colony. Hereafter, such colonies created by fission of the parent colony will be referred to as ramets (after Babcock 1991), and the parent colonies will be referred to as parents. Intact colonies and parents were assumed to be derived sexually, and thus collectively will be considered genets. Although it is possible that S. siderea larvae recruit to surfaces of dead skeleton on parent colonies
(Edmunds 2000b), failure to distinguish between sexual recruits and ramets is considered to be minimal given the distinctive growth form of this species (Babcock 1991).
31 Growth monitoring
In order to test the hypothesis that age affects the growth of small (6 – 50 mm in diameter) colonies of S. siderea, the growth of sexual recruits (i.e., young genets) and old ramets was measured from August 2004 to August 2005. Measurements (±0.1 mm) included the height of the colony, the greatest basal diameter, and the diameter perpendicular to the base, in order to account for both lateral and vertical growth. In
August 2004, six and four parents that had undergone fission were selected on the west and east coasts of Lameshur Bay, respectively. One to five ramets (6 – 50 mm) were tagged and measured on each parent. Only approximately circular/oval ramets that appeared healthy (e.g., no evidence of bleaching or tissue injury) were selected in order to approximate the shape of recruits. Within an approximately 5-m radius of the parent, one to five recruits (10 – 43 mm) also were tagged and measured. A total of 30 recruits and 30 ramets were tagged, and in August 2005, the colonies were relocated and measured if alive.
Population census
In order to determine the relative abundance and size distribution of genets and ramets, five haphazard belt transect censuses were conducted along the east and west coasts of Great Lameshur Bay in August 2005. All colonies within a 10 x 2 m area were measured (mean basal diameter; ±1 mm) and identified as genets or ramets.
32 Statistics
The mean growth in diameter and height of recruits and ramets was compared using a t-test. Because multiple ramets were tagged from a single S. siderea parent genet, they cannot be considered statistically independent due to their common genotype and close spatial proximity. Therefore, the growth of ramets from each genet was averaged to obtain an independent statistical replicate for the t-test. Similarly, because the recruits were tagged nearby parents, their growth also was averaged to obtain a statistical replicate. Although this method is conservative statistically, it should be noted that the mean growth rates are similar and statistical outcomes are identical if each colony is treated as an independent replicate.
In order to determine how the scaling of linear growth varied by colony type
(recruit vs. ramet), data were analyzed using analysis of covariance (ANCOVA) with the log final diameter (mm) as the dependent variable and the log initial diameter (mm) as the covariate. Logarithmic linear regressions using a measure of body size as the independent variable and a physiological trait as the dependent variable are commonly recognized as scaling relationships (Schmidt-Nielsen 1984), which can be described as isometric or allometric depending on the slope of the linear regression (i.e., the scaling exponent). Reduced major axis (RMA) regression was used to calculate the scaling exponent (b) for the regression of log final diameter against log initial diameter because diameter was a random variable potentially estimated with error (Quinn & Keough 2002).
The standard errors for the regression slopes were taken from ordinary least squares
(OLS) regression analyses, because the variance of OLS and RMA estimators are identical to the third significant digit (McCardle 1988). The scaling exponent for each
33 colony type was tested against the null hypothesis that b = 1 (isometry) using a t-test
(Sokal & Rohlf 1995). All analyses were completed using JMP 5.01 for Macintosh; residuals were inspected graphically for normality and homoscedasticity.
To test the hypothesis that the size-frequency distributions of genets and ramets differed significantly from each other, a Kolmogorov-Smirnov test statistic was calculated as described by Sokal and Rohlf (1995).
34 Results
Growth monitoring
Of the 60 colonies tagged in August 2004, 26 sexual recruits and 29 ramets were relocated in August 2005. Three recruits and three ramets died, and one ramet fused with the parent. The initial diameter of the recruits was 24.9 ± 1.8 mm (mean ± SE, n = 23), thus recruit age averaged 5.4 ± 0.4 years (±SE) assuming a constant annual diameter growth rate of 4.6 mm yr-1 (see below). Although the initial diameter of ramets
(26.0 ± 1.9 mm; mean ± SE, n = 25) was statistically similar to the initial diameter of recruits (t = 0.17, df = 46, p = 0.68), the diameter of the parent from which ramets originated was 50.1 ± 4.9 cm (mean ± SE, n = 10). Consequently, the age of ramets averaged 108.9 ± 10.7 years (±SE), assuming a constant growth rate of 4.6 mm yr-1. The initial height of recruits (5.9 ± 0.8 mm) was statistically indistinguishable from the height of ramets (6.5 ± 1.0 mm) [t = 0.17, df = 46, p = 0.68]. Therefore, although the two colony types were of similar size, sexual recruits were an order of magnitude younger than ramets derived from the fission of parent colonies.
Between August 2004 and August 2005, sexual recruits grew 4.6 ± 0.8 mm
(mean ± SE, n = 23), while ramets grew 1.5 ± 0.9 mm (mean ± SE, n = 25). This 3.1-fold difference was significant (t = 7.2, df = 16, p = 0.02). However, both colony types exhibited statistically similar increases in height (t = 0.2, df = 16, p = 0.66) [Fig. 3.1].
Final diameter significantly varied by colony type (i.e., recruit vs. ramet)
[Table 3.1]. However, the interaction between colony type and the covariate (log initial diameter) was significant (Table 3.1), indicating that the scaling exponents for recruits and ramets are not equal. The slope of the recruit regression is 0.89 ± 0.07
35 (±SE; r2 = 0.87, n = 25), while the slope of the ramet regression is 1.08 ± 0.06
(±SE; r2 = 0.92, n = 23) [Fig. 3.2]. Although the two slopes were significantly different from one another, neither the regression for recruits (t = 1.69, df = 21, p = 0.11) nor the regression for ramets (t = 1.26, df = 23, p = 0.22) was indistinguishable statistically from
1, indicating isometric scaling.
Population census
A total of 176 Siderastrea siderea genets were found and measured in ten 20-m2 belt transects; 153 (87%) of these were intact (0.77 ± 0.17 intact colonies m-2) while 23
(13%) had undergone fission (0.12 ± 0.02 parents m-2). The mean diameter of parents was 28.5 ± 3.4 cm (± SE, n = 23). Assuming a constant annual diameter growth rate of
4.6 mm yr-1 (see Fig. 3.2), the mean age of parents was at least 62.0 ± 7.4 years when fission occurred. Each parent possessed 4.8 ± 0.6 ramets, for a total of 111 measured ramets (0.56 ± 0.11 ramets m-2). In other words, 42% of the live S. siderea colonies were ramets, and the ratio of ramets to genets was 264:176, or 1.5 (after Babcock 1991). All values are means ± SE.
The size-frequency distributions of genets and ramets were indistinguishable statistically from one other (K-S test, D = 0.09, D0.05 = 0.16, p > 0.05); both were positively skewed (Fig. 3.3), and the median diameter was 3.9 cm for both genets and ramets. A total of 89 (58%) genets were ≤4 cm, and can be defined as juvenile corals
(Bak & Engel 1979, Edmunds 2000a); 58 ramets (52%) measured ≤4 cm.
36 Discussion
Age significantly affected the lateral growth of small Siderastrea siderea colonies. Although it was not possible to measure precisely the ages of the sexual recruits and ramets monitored in this study, the two colony types were estimated to vary by approximately an order of magnitude. The age difference also was manifest in the scaling of growth. The scaling exponent was significantly lower for sexual recruits than ramets, indicating that the growth of the smallest corals in this study depended upon their age. Although the scaling exponent for the recruit regression was indistinguishable statistically from isometry (i.e., b = 1), this outcome potentially is a result of the small sample size, because in a study of juvenile S. siderea in Jamaica, the increase in diameter over 8 months was allometric (b = 0.87 ± 0.02) [Vollmer 1999]. Considerable variation in the age of small coral colonies has been documented previously (Hughes & Jackson
1985, Hughes & Connell 1987, Babcock 1991), and whether corals grow, shrink or die is not independent of their age (Hughes & Connell 1987, Babcock 1991). In addition, age has significant effects on reproduction in Goniastrea favulus (Kojis & Quinn 1985), tissue regeneration in Acropora palmata (Meesters & Bak 1995), and on calcification in
Madracis mirabilis (Chapter 2).
The two-fold difference in lateral growth rates between sexual recruits and ramets suggests that fission is not an adaptive life history strategy for S. siderea, unless ramets allocated energy towards reproduction rather than growth (see below). Unlike fragmentation, which allows for dispersal and spreads the risk of genet mortality for many branching corals (Highsmith 1982), fission probably has negative consequences for the population dynamics of S. siderea, and three other massive coral species, G. aspera,
37 G. favulus, and Platygyra sinensis (Babcock 1991). Nearly half of the colonies of S. siderea observed in the present study were ramets, indicating that partial mortality and fission commonly affects the population in Great Lameshur Bay. Not surprisingly, ramets typically were found on large colonies of S. siderea, because the risk of partial injury increases with coral size (Hughes & Jackson 1985, Babcock 1991). The clonal input to the population of S. siderea in the present study, measured as the ratio of ramets to genets (1.5), is slightly higher than the mean clonal input of 1.30 ± 0.07 (±SE; n = 18) for three massive species between 1982 and 1984 (from Table 5 in Babcock 1991). This difference suggests that the colonies of S. siderea in Great Lameshur Bay have experienced greater stress in their recent history than the three species on the Great
Barrier Reef twenty years ago. This is contrary to what one might expect, because
Babcock (1991) studied corals on a reef flat, which were exposed to extreme physical stresses associated with intertidal environments (Raffaelli & Hawkins 1996). Coral populations in Great Lameshur Bay are relatively protected from local anthropogenic disturbances due to its inclusion within an internationally recognized marine protected area, and thus coral fission and partial mortality likely were caused by natural disturbances and/or regional scale processes.
The proximal physiological causes responsible for the different growth rates exhibited by young recruits and old ramets are unclear. The ramets may have been allocating resources to tissue or skeletal regeneration (Bak 1983), but no evidence for this hypothesis was observed from visual inspection of the colonies. Instead, ontogenetic differences between sexual recruits and ramets may have influenced the observed variation in growth rates. Sexual recruits must attain some minimum size before
38 becoming reproductively active (Soong 1993), and the low scaling exponent of growth for recruits relative to ramets may have evolved as a response to this selective pressure.
Alternatively, slow growth of old ramets may reflect a gradual deterioration of physiological functions with age, i.e., senescence. Although the genets of corals and other modular organisms are theoretically “immortal” (Jackson & Coates 1986), there is evidence for module senescence in corals (Meesters & Bak 1995, Chapter 2), bryozoans
(Palumbi & Jackson 1983), algae (Borowitzka & Larkum 1976), plants (Watt 1947,
Leopold 1961), and fungi (Holliday 1969). Finally, it is possible that older ramets were investing energy into reproduction rather than growth. The colony size at maturation for
S. siderea ranges from 5.9 to 11.4 cm (Soong 1993), which is slightly larger than the largest ramets used in the present study, but due to their old age, some ramets may have been capable of producing gametes at a smaller size than is typical for sexual recruits
(Kojis & Quinn 1985). This hypothesis is easily testable, and the outcome would provide insight into the consequences of fission for the demographics of S. siderea. Although the mechanism is unclear, the main finding of this study, that fission of S. siderea creates ramets that exhibit depressed growth rates, has implications for the population dynamics of this common reef-building species in the Caribbean.
39 Table 3.1. Results of a one-way analysis of covariance testing the fixed effect of colony type (recruit vs. ramet) on log final diameter (mm) using log initial diameter (mm) as the covariate.
Source df MS F p Type of colony (T) 1 0.0252 7.30 <0.01 Log initial diameter (D) 1 1.3037 377.73 <0.001 T x D 1 0.0164 4.75 <0.05 Residual 44 0.0035
40 6
Recruit Ramet 5
4
3
Growth (mm) 2
1
0 Diameter Height
Figure 3.1. Linear increase (mean ± SE) in diameter and height of Siderastrea siderea recruits and ramets from August 2004 – August 2005 in Lameshur Bay, St. John. Recruits grew significantly more than ramets with respect to diameter (t = 7.20, df = 16, p = 0.02), but the height increase was similar statistically for both colony types (t = 0.20, df = 16, p = 0.66).
41 1.8
1.6
1.4
1.2
1 Log final diameter (mm)
0.8 Recruit Ramet 0.6 0.6 0.8 1 1.2 1.4 1.6 1.8
Log initial diameter (mm)
Figure 3.2. Growth of Siderastrea siderea sexual recruits (10 – 43 mm) and small ramets (6 – 50 mm) plotted as log final diameter against log initial diameter. Although the slopes of both regressions are indistinguishable statistically from 1 (see Results), the slope of the regression for recruits (b = 0.83 ± 0.07; ± SE, n = 23) is significantly lower than the slope of the regression ramets (b = 1.08 ± 0.07; ± SE, n = 25) [Table 3.1].
42
A B 60 60
50 50
40 40
30 30 # of genets
# of ramets 20 20
10 10
0 0 0 4 8 12 16 20 24 28 32 36 40 44 48 52 56 0 4 8 12 16 20 24 28 32 36 40 44 48 52 56 Diameter (cm) Diameter (cm)
Figure 3.3. Size-frequency distribution of (A) genets (n = 176) and (B) ramets (n = 111) of Siderastrea siderea in Lameshur Bay, St. John. Genets depicted in A are either intact (black bars; n = 153) or are parent colonies that have undergone fission (white bars; n = 23). Ramets in B were derived from the parent colonies in A. The size-frequency distributions were indistinguishable statistically from one another (see Results).
43 CHAPTER 4
Energetic constraints on indeterminate growth in the solitary coral, Fungia concinna
Introduction
Although age is profoundly important in affecting organismal physiology
(Chapters 2,3), the size of an individual arguably is as important in influencing the physiological traits, ecological relationships and evolutionary success of organisms
(Schmidt-Nielsen 1984). From the smallest bacterium to the largest sequoia tree, life on earth spans approximately 21 orders of magnitude in mass (Brown et al. 2000). For many marine invertebrates, the maximum size an individual attains primarily is dependent upon extrinsic environmental factors, rather than genetic composition (Sebens
1987b). In other words, animals that do not display a finite adult size exhibit indeterminate growth (Sebens 1987b). For example, intertidal invertebrates tend to be larger in sheltered coves than those inhabiting wave-exposed sites, which likely reflects the risk of dislodgement due to wave stress (Denny et al. 1985). Alternatively, energetic costs may limit the size of benthic marine organisms (Sebens 2002). Sea anemones occupying the lower intertidal zone are larger than those inhabiting the upper intertidal, potentially a result of increased feeding time and decreased aerial exposure (Sebens
1982). Using an energetic approach based on habitat suitability, the optimal and maximum size of passive suspension feeders (e.g., anemones) was modeled by Sebens
(1982, 2002). In its simplest form, Sebens’ model derived the optimal size of an individual by maximizing the difference between energetic intake and metabolic cost, while the maximum size was attained when the difference between intake and cost was
44 zero (Figure 4.1). Therefore, optimally-sized individuals have the greatest energetic surplus available for growth and reproduction. In addition, Sebens (1982) considered seasonal energetic budgets and the spatial constraints on gonad production associated with reproductive allocation. For the purposes of this study, these finer aspects of the model will not be considered.
Sebens (1982) empirically tested his model using Anthopleura xanthogrammica, a large solitary anemone found in rocky intertidal and subtidal habitats of western North
America (Sebens 1981). Intake was modeled with the primary assumption that prey capture rates depended on tentacular surface area, increasing as the 0.67 power of mass for animals that grow isometrically. For A. xanthogrammica, prey capture was a linear function of surface area, with a scaling exponent of 0.7 on mass (Sebens 1981). The second assumption of the model was that the scaling exponent (on mass) always is higher for the energetic cost function than for the intake function. Consequently, an energetic surplus would exist for a finite size range, but eventually would decrease to zero at the maximum size, as depicted in Figure 4.1. Experimentally derived cost and intake exponents for A. xanthogrammica were 1.1 and 0.7, respectively (Sebens 1982), thus supporting the model. In addition, the observed optimal and maximum sizes of anemones in the field fit the predictions of the model well (Sebens 1982). Inherent in the model is the implication that a solitary anthozoan possesses the potential for an energetic advantage over modular forms because a large polyp size allows capture of larger, albeit less frequent, prey (Sebens 1979). Anthopleura xanthogrammica, other anemones, and solitary corals are thought to exhibit this type of growth form (Sebens 1982, 2002). In
45 contrast, environments with small, common prey (e.g., microcrustacea, photons) should favor colonial anthozoans such as scleractinian corals (Sebens 1979).
Although Sebens’ energetic model was tested using a suspension-feeding anemone, it is applicable to indeterminately growing animals in general, despite potentially contrasting strategies of energy acquisition. For example, the model was applied successfully to the herbivorous chiton Chiton pelliserpentis and the actively filtering mussel Mytilus edulis (Sebens 1987b). However, to the best of my knowledge, the model has not been tested in any other anthozoans, many of which contain endosymbiotic dinoflagellates and rely on these autotrophic symbionts (Symbiodinium spp.) for energy (Sebens 1987a), especially tropical reef corals (Muscatine 1990). The contribution of energy in the form of photosynthetically fixed carbon from the dinoflagellate symbiont to the coral host has been well-studied (Muscatine et al. 1981,
Davies 1984), and for many shallow water coral species autotrophy can fulfill the daily energetic needs of the holosymbiont (Edmunds & Davies 1986, Davies 1991). However, the majority of scleractinian corals are colonial, and the availability of bare substratum likely determines the upper size limit of colonies (Jackson 1977).
Corals in the genus Fungia (Scleractinia: Fungiidae) are atypical scleractinian corals because they are large, solitary polyps, and thus provide a unique comparison to A. xanthogrammica with respect to Sebens’ energetic model. The Fungia polyp is supported by a calcareous skeleton, which has consequences for the mobility of this unattached scleractinian. The weight of large Fungia polyps may impede locomotion and the ability to turn upright or excavate from burial (Goreau & Yonge 1968, Hubbard 1972,
Jokiel & Cowdin 1976, Fisk 1983, Chadwick 1988, Chadwick-Furman & Loya 1992),
46 and thus negatively affect phototaxis (Yamashiro & Nishihara 1995, Goffredo &
Chadwick-Furman 2000), an energetically advantageous behavior for Fungia.
Photosynthetic input can supply 70% of the carbon required for animal respiration in F. scutaria (Muscatine et al. 1981), and autotrophy alone is sufficient for growth in this species (Franzisket 1969). Therefore, it appears that heterotrophy contributes less to the energetic requirements of Fungia than to A. xanthogrammica. Although Sebens’ model predicts that solitary polyps should possess an advantage over colonial forms because they are capable of expanding their prey size range (e.g., A. xanthogrammica), it is uncertain whether the size of prey (if available) scales with the size of the predator because Fungia typically possess diminutive tentacles relative to their large polyp size
(Wells 1966, Veron 2000).
The present study was designed to test two hypotheses derived from Sebens’ energetic model using the solitary coral F. concinna. First, do energetic constraints limit the size attained by F. concinna? Secondly, do large F. concinna possess an energetic advantage relative to small individuals? In this analysis, energetic cost was measured as aerobic respiration of the holosymbiont, and energetic intake was assumed primarily to originate from carbon fixation by symbiotic algae. Whole organism metabolic rates typically are limited by rates of uptake across surfaces (West et al. 1999), and for symbiotic corals which harbor an autotrophic partner, respiratory and photosynthetic rates will be limited by the scaling relationships of surface area against physiological traits. For example, coral respiration varies with the amount of metabolically active tissue biomass (Anthony & Hoegh-Guldberg 2003), and photosynthesis is affected by the density of algal symbionts and their chlorophyll content (Porter et al. 1989, Warner et al.
47 1999). Therefore, in the present study the scaling of surface area against tissue biomass,
Symbiodinium population density, and chlorophyll a content were quantified to gain insight into the energetic scaling relationships for Fungia. Inherent in the application of an energetic model to coral size constraints, such as the one developed here, is the implication that growth rates eventually decline at some size, which has been demonstrated in many corals, including Fungia (Hughes & Jackson 1985, Rinkevich &
Loya 1986, Babcock 1991, Chadwick-Furman et al. 2000). Calcification is the basis for skeletal growth in scleractinian corals, thus calcification rates were examined for corals of various sizes. Calcification rates also were used to estimate the cost associated with growth to clarify the energetic surplus available to F. concinna for other important investments, such as reproduction and locomotion. Because the weight of the calcareous skeleton imposes mechanical limitations on mobility (Chadwick 1988, Chadwick-Furman
& Loya 1992, Yamashiro & Nishihara 1995), the scaling of skeletal weight was investigated as an alternative constraint on growth in Fungia.
48 Methods
It was necessary to select a relatively small solitary fungiid species for this study because the equipment available for lab analyses (described below) could not accommodate corals >78 cm2 in planar area. Fungia concinna was chosen to test the energetic constraints on growth because it attains the smallest maximum size of congenerics (F. scutaria, F. paumotensis, F. repanda, F. fungites) present on the north coast of Moorea (17º 28.528' S, 149º 49.970' W), French Polynesia (Veron 2000; pers. obs.), where this study was completed. The habitat studied is typical of forereef slopes on Moorea’s north coast (Adjeroud et al., 2002) with scleractinian corals as the dominant space occupiers, especially the genera Acropora, Pocillopora, Montipora and Porites
(pers. obs.).
Fungia concinna were sampled from a forereef buttress at 15 – 16 m depth for the scaling analyses required for the energetic model. First, to address the two main hypotheses derived from Sebens’ model (described above) the scaling of energetic cost and intake was quantified as aerobic respiration and photosynthesis, respectively. To investigate the role of surface area constraints on physiological correlates of respiration and photosynthesis, the scaling of symbiont density, chlorophyll a content, and tissue biomass against surface area is described. In addition, the energetic cost associated with growth was quantified using the physiological traits necessary to construct a growth budget, namely, calcification rates, Symbiodinium population density, and tissue biomass
(Davies 1984, Edmunds & Davies 1986). Lastly, at the study site, light levels were measured for the calculation of daily energetic intake from photosynthesis, and the
49 population size structure of F. concinna was quantified to provide a context for the energetic model.
Respiration and photosynthesis
The effects of flow on mass transfer-dependent processes such as respiration and photosynthesis have been well-studied in benthic marine organisms (Carpenter et al.
1991, Patterson et al. 1991, Sebens et al. 2003). For example, in low flow environments mass transfer may be limited, thus potentially reducing the metabolic rates of corals
(Bruno & Edmunds 1998). The nature of the flow environment around an organism can be characterized using the Reynolds number, Re (Denny 1988), a dimensionless ratio of the inertial and viscous forces acting upon an organism:
Re = (UL)ν-1
where U is the velocity of the fluid relative to the organism, L is a characteristic linear dimension of the organism, and ν is the kinematic viscosity of the fluid. Inertial forces are related to the product of the flow speed and the organism size, thus organisms of varying size experience different flow environments in the same flow speed. Because the present study addressed the effects of coral size on energetic cost (respiration) and intake
(photosynthesis), the metabolic trials were designed to create a uniform flow environment for corals of varying size by adjusting the absolute flow speeds.
Two recirculating chambers differing in size were used to accommodate corals ranging in planar surface area from 8.8 – 67.2 cm2. The small chamber (working section
50 = 8.1 x 6.1 x 7.0 cm; length x width x height) accommodated corals 8.8 – 20.8 cm2
(n = 4), and the large chamber (working section = 12.0 x 10.0 x 6.0 cm) accommodated corals 32.7 – 67.2 cm2 (n = 5). Flow was created in the respirometry chambers using a submersible pump (Rule 350 gph) operated at a variable input voltage (1 – 12 V DC).
Oxygen was measured every 15 s using a fiber-optic oxygen electrode (FOXY-R, Ocean
Optics, Fl) placed above the coral in the working section of the chamber. The electrode was calibrated daily using a chemical zero solution (sodium sulfite and sodium tetraborate) and air-saturated seawater obtained by bubbling seawater with a pump and airstone.
The nature of the flow environment was estimated for the working section of each chamber by photographing Artemia (brine shrimp) cysts (Helmuth & Sebens 1993). A coral mimic made of modeling clay was placed in each chamber to approximate the flow environment with a coral present. Flow straighteners delivered unidirectional flow 5 – 15 mm above the coral mimics, and flow speeds in this area were estimated over a range of input voltages. Regression equations of mean flow speed against input voltage for each chamber were used to determine the specific flow speed necessary to achieve a similar Re for each coral based on its mean diameter (small chamber, n = 16, r2 = 0.91; large chamber n = 21, r2 = 0.92). Thus, large corals were exposed to slower flow speeds than small corals. Mean diameter along the axis of flow is an appropriate measurement of length for use in the calculation of Re for F. concinna, because the height of individuals increases proportionately with diameter as individuals grow (see below). In order to expose corals to flow speeds typical for 15-m depth benthos (Helmuth & Sebens 1993,
Sebens et al. 2003), speeds ranged from 1.8 – 5.0 cm s-1 to attain an Re of 1950 for the
51 corals used (ν = 8.6 x 10-3 cm2 s-1; Table 3.1). To further minimize the possibility of a mass transfer artifact arising from the interaction of size and flow, preliminary trials were conducted to determine if flow speeds creating Re ≈ 1950 were limiting mass flux.
Twelve corals (13.9 – 67.9 cm2 in surface area) each were subjected to a respiration trial at a net flow of Re ≈ 1950, and a net flow of Re ≈ 3900 by doubling the flow speed.
Oxygen consumption per coral was not significantly different for the two flow speeds
(paired t-test, t = 0.85, df = 11, p = 0.41).
Corals for oxygen flux trials were collected from the study site and immediately transported to flow-through aquaria at UC Berkeley’s Gump Research Station in Cook’s
Bay. To avoid the effects of light history on metabolic rates (Edmunds & Davies 1988), corals were dark-acclimated for 18 – 24 hrs prior to metabolic trials. To minimize photoacclimation to the laboratory light levels, all trials were completed within 48 hrs of collection. All trials were completed between 1200 – 1800 hrs. Chambers were immersed in 35 L of filtered seawater (5 µm) to maintain the temperature of the seawater inside the chamber within 0.5 ºC of ambient seawater temperature during trials; ambient sea surface temperature in Cook’s Bay ranged between 28.2 – 29.8ºC. Respiration trials for each coral were first conducted in the dark, and photosynthesis subsequently was measured at six photon flux densities (62 ± 3, 87 ± 4, 116 ± 5, 155 ± 3, 244 ± 8, 386 ± 16
µmol photons m-2 s-1; mean ± SE; n = 8 light level-1) provided by a metal halide lamp
(Lithonia, 70 watts) in a random sequence. Light levels were established by adjusting the height of the lamp, and photon flux density (PFD) was measured with a LiCor 193-SA spherical quantum sensor connected to a LiCor Li-1000 datalogger. PFD was recorded every 15 seconds for eight minutes, and the mean value for each minute served as the
52 statistical replicate for the mean light level at each height. Net photosynthetic rates at the two highest PFD levels (244 and 386 µmol photons m-2 s-1) were omitted from the calculation of the photosynthesis – PFD regression because in situ corals did not experience mean PFD levels > 150 µmol photons m-2 s-1 (described below). Trials lasted
8-12 minutes, and the chamber was flushed with fresh seawater after each trial to reestablish ~100% oxygen saturation. Trials were conducted between 93 and 118% oxygen saturation, and all oxygen flux rates were linear within this range. Corals were acclimated to each PFD for 15 minutes prior to each trial and controls (without corals) were run daily in the dark. Preliminary analyses showed that oxygen flux in controls in maximum light (386 µmol photons cm-2 s-1) did not differ significantly from that in controls in dark (paired t-test, t = 1.17, df = 6, p = 0.29), and therefore controls were not conducted in light. Salinity was assumed to be 35 ‰ and values for oxygen solubility in seawater were taken from Ramsing and Gunderson at http://www.unisense.com (Riley &
Skirrow 1975). Oxygen flux rates were normalized to the planar surface area (cm2) of the polyp oral surface only (Masuda et al. 1993), because preliminary analyses indicated that the majority of host tissue (69%) and Symbiodinium (84%) are located on the oral surface.
Zooxanthellae, chlorophyll and biomass
The scaling relationships between the size of F. concinna and their respiration and photosynthesis may be explained in part by the scaling of host and symbiont physiological parameters, thus Symbiodinium densities, chlorophyll a content and host tissue biomass were quantified. After oxygen flux trials, corals were kept in flow-
53 through aquaria for no more than one week before coral tissue was removed from the skeleton using a Waterpik® (Johannes & Wiebe 1970) filled with filtered seawater (0.45
µm; FSW), and the resulting slurry was homogenized. To determine how Symbiodinium population densities varied with coral size, a 1-ml aliquot of the homogenate was centrifuged (13,000 rpm) and the resulting pellet was resuspended in 90 µl FSW. Ten replicate counts of Symbiodinium were made from the diluted homogenate using a hemocytometer. To assess the scaling of chlorophyll a content, pigment concentrations were estimated using an acetone extraction method (Jeffrey & Humphrey 1975) within three weeks of freezing (-4°C) the tissue slurry. For each sample, 1.6 ml of the homogenate was centrifuged (13,000 rpm), and the resulting pellet resuspended in 100% acetone to extract pigments in the dark at 4°C for 24 hrs, and samples were read at 630 and 663 nm against acetone blanks.
The corals used in oxygen flux trials were used for symbiont characteristics, thus
12 newly-collected F. concinna (6.2 – 69.4 cm2) were used to assess how tissue biomass varies with size. Corals were fixed in 5% formalin for 24 hrs, then decalcified in 10%
HCl for 24 – 48 hrs, depending on the size of the coral. The resulting tissue tunics then were dried at 60 ºC for 7 days prior to weighing (Edmunds & Davies 1986).
Calcification and skeletal characteristics
The goal of this study was to test the hypothesis that energetics constrain the size of an indeterminately growing scleractinian coral. Implicit in this hypothesis is the assumption that growth rates vary with size, with an eventual decline in growth at some finite size. The scaling of calcification was quantified to test this assumption. The
54 skeletal weights of 28 corals were measured using the buoyant weight method (Davies
1989) at the beginning and end of a 25-day period at the study site (10 February – 6
March). Corals (6.2 – 95.9 cm2) were collected from the study site and tagged using small plastic squares (0.5 cm2) that were epoxied (Z-spar® Splash Zone A-788) to the aboral surface. A preliminary experiment demonstrated that the tags did not significantly affect coral calcification rate (ANCOVA, F1, 17 = 1.43, p = 0.25). Each coral was weighed (±1 mg) in seawater, and the change in buoyant weight converted to dry skeletal weight using equations from Davies (1989), assuming the density of aragonite to be 2.93 mg cm-3.
If the skeletal weight of F. concinna increases disproportionately as individuals grow, large adults might suffer from impeded locomotion and excavation from burial, which may act as another potential constraint on growth (aside from energetics). The scaling of surface area against skeletal weight was tested against the null hypothesis of isometry using the initial weight data from the calcification experiment (n = 28).
The diameter and height of F. concinna increase as individuals grow, and both linear dimensions may affect the flow environment experienced by corals. To ensure that mean diameter is an appropriate measure of the characteristic length for calculation of Re the scaling relationship between septa height and diameter was quantified. The height of septae determine the projected height of the coral above the substratum, and it was quantified by inserting the thin metal end of a caliper into the mouth of each coral, and measuring the distance between the bottom of the oral cavity (i.e., columella) and the top of the septae. The height of septae and mean diameter was quantified (±1 mm) for 42 individuals at the study site.
55 Field photon flux density and energy budget calculations
In order to calculate an energy budget for F. concinna based on daily productivity, the PFD at the study site was quantified using a LiCor LI-193SA spherical quantum sensor deployed in situ at 15-m depth from 11:48 am to 12:01 pm on 27
February 2005, a day characterized by a clear sky with no clouds present. PFD was measured every 5 seconds and the average reading for each minute was used as a statistical replicate for calculating the percent transmittance of surface PFD to 15 m.
Immediately prior to measuring in situ PFD, surface measurements were taken between
11:22 am and 11:31 am. To characterize daily light conditions at the study site, surface
PFD was logged (LiCor Li-1000) continuously from 1 March – 13 March. PFD was measured every minute, and the mean PFD over half-hour intervals was converted to in situ PFD using the mean percentage of light transmitted to 15-m depth.
The in situ PFD on one sunny day (13 March) was used to quantify net oxygen flux rates for daily coral energy budgets. The mean in situ PFD did not exceed 150 µmol photons m-2 s-1 at the collection site (Figure 4.2), therefore net photosynthetic rates were calculated from the mean PFD in half hour intervals using only the initial slope (0 – 155
µmol photons m-2 s-1) of each photosynthesis-PFD curve. The linear regression of the net photosynthetic rates against the five lowest PFDs (0 – 155 µmol photons m-2 s-1) provided a better fit over ecologically relevant light levels than the hyperbolic tangent curve over all seven light levels. Net photosynthetic rates for each half-hour interval were converted to energy (J) by assuming that 6 moles of oxygen resulted in the formation of 1 mole of glucose, with an energy equivalent of 2817 kJ (Lehninger 1973).
The resulting values were summed over 14.5 hrs of daytime to give the integrated net
56 energetic intake. Respiratory oxygen consumption was converted to energetic cost using
-1 the oxy-joule equivalent of 19.63 J ml O2 (Elliot & Davison 1975), assuming that lipid is the main respiratory substrate (Patton et al. 1977). The daily net energetic intake was added to the energetic cost over 14.5 hrs to give the daily gross energetic intake.
In addition to quantifying cost and intake for Sebens’ energetic model, the energetic cost of growth was calculated using empirical scaling relationships for calcification, host biomass and symbiont density. The calculation consisted of the cost associated with new host tissue and algal division, while the cost of calcification and assembly were considered to be included in the cost of aerobic respiration. It was assumed that skeletal growth (i.e., calcification) was accompanied by a concomitant increase of tissue (Davies 1991); tissue growth therefore was calculated using a regression between biomass and skeletal weight (y = 48.99x, r2 = 0.94). Tissue biomass was converted to energy using a tissue energy-content of 16.18 J mg-1 dry tissue
(Edmunds & Davies 1986). Likewise, it was assumed that Symbiodinium divided at rates necessary to colonize new tissue. The number of new Symbiodinium mg-1 dry tissue was calculated using a regression between symbiont density and biomass
(y = 51850.71x, r2 = 0.78), and this value was converted to energy using a factor of 11.69
J x 106 Symbiodinium cells (Edmunds & Davies 1986). The summed cost of respiration, tissue growth, and symbiont growth hereafter will be referred to as total cost.
Population structure
In order to quantify the population density and size distribution of F. concinna, the longest and shortest diameters (±1 mm) were measured for every individual within 12
57 randomly located 2 x 1 m areas along the 15-m depth contour until >200 individuals were measured. Surface area was used as a simple measure of size instead of converting to biomass, because a conversion would have required a substantial extrapolation beyond the limits of the empirical relationship between area and biomass. Fungia concinna are approximately circular, thus the planar surface area of tissue was estimated as a circle
(A = πr2).
Statistics
Logarithmic linear regressions using a measure of body size as the independent variable and a physiological trait as the dependent variable commonly are recognized as scaling relationships (Schmidt-Nielsen 1984), which can be described as isometric or allometric depending on the slope of the linear regression (i.e., the scaling exponent).
Reduced major axis (RMA) regression was used to calculate the scaling exponent (b) for each regression because surface area was a random variable potentially estimated with error (Quinn & Keough 2002). The standard errors for the regression slopes were taken from ordinary least squares (OLS) regression analyses, because the variance of OLS and
RMA estimators are identical to the third significant digit (McCardle 1988). The scaling exponent for each treatment was tested against the null hypothesis that b = 1 (isometry) using a t-test (Sokal & Rohlf 1995). The scaling exponents for the energetic cost and intake regressions also were compared using a t-test that incorporated variance estimates from both regressions. Because the biomass of corals could not be determined for individuals whose tissue was removed for zooxanthellae and chlorophyll analysis, the least squares linear regression between surface area (cm2) and biomass (mg) was used to
58 convert coral surface area into biomass for the energetic cost and intake regressions.
Residuals were graphically inspected to ensure homoscedasticity and normality for regression analyses (Sokal & Rohlf 1995).
59 Results
Respiration and photosynthesis
-2 -1 Mean coral respiration was -16.7 ± 2.3 nmol O2 cm min (±SE, n = 9), and net photosynthesis at the maximum experimental PFD (386 µmol photons m-2 s-1) was 49.8 ±
-2 -1 5.2 nmol O2 cm min (mean ± SE, n = 9; Figure 4.2 inset). Photosynthetic rates were a linear function of PFD over the range of ecologically relevant light levels
(0 – 155 µmol photons m-2 s-1), and photosynthesis did not saturate at the maximum in situ PFD. Therefore, a hyperbolic tangent function was not used to calculate coral energy budgets (see below), although it typically is used to describe characteristics of the photosynthesis-PFD curve across a large range of light levels (Chalker 1981).
Zooxanthellae, chlorophyll and biomass
The scaling of Symbiodinium population density, chlorophyll a content, and tissue biomass was examined in order to gain insight into the physiological correlates of photosynthesis and respiration. Individual corals possessed 10.5 ± 1.4 x 105 symbionts cm-2 (mean ± SE; n = 12), and the symbionts contained 10.7 ± 1.2 pg chlorophyll a cell-1
(mean ± SE; n = 12). Symbiodinium population densities (Figure 4.3A) and chlorophyll a content (Figure 4.3B) increased isometrically with surface area. In contrast, tissue biomass (Figure 4.3C) increased allometrically with surface area, ranging between 8.3 mg cm-2 for the smallest individual (6.2 cm2) to 46.6 mg cm-2 for the largest individual
(69.4 cm2), averaging 18.6 ± 3.0 mg cm-2 (±SE; n = 12). Therefore, large corals have disproportionately more biomass per unit area than small corals, even though
Symbiodinium population densities and their chlorophyll a content remain similar for
60 corals of different sizes. The physiological scaling relationships are summarized in Table
4.2.
Calcification and skeletal characteristics
Twenty-seven of the 28 tagged F. concinna were recovered at the end of the calcification period. Two of these were upside down, and thus excluded from the analysis, and an additional coral was a statistical outlier due to high leverage (studentized residual = 3.4) and also was removed from the analysis. Calcification scaled allometrically with respect to tissue biomass (Figure 4.4, Table 4.2), indicating that calcification rates declined with size. For example, the five smallest corals (9.5 ± 1.5 cm2; mean ± SE) calcified at 0.27 ± 0.05 mg mg-1 dry tissue day-1 (mean ± SE), while the five largest corals (73.5 ± 5.6 cm2; mean ± SE) calcified at 0.05 ± 0.01 mg mg-1 dry tissue day-1 (mean ± SE). The initial weight of the 28 corals ranged from 2.3 g to 197.2 g for corals 6.2 to 95.9 cm2 in size. Skeletal weight scaled allometrically with surface area
(Figure 4.3D), demonstrating that large corals are disproportionately heavy.
Septa height ranged from 3 to 17 mm in corals with mean diameters of 25.5 and
136.5 mm, respectively. The height of septae scaled isometrically with the mean diameter of corals (Figure 4.5), indicating that the height of F. concinna remains proportional to its diameter as it grows. Therefore, diameter is an appropriate measure of length for use in the quantification of Re. The physiological scaling relationships are summarized in Table 4.2.
61 Field photon flux density and energy budget calculations
From 1 March – 13 March 2005, the mean integrated PFD reaching 15-m depth on the forereef was 4.28 ± 0.27 mol photons m-2 day-1 (± SE, n = 13) for a day lasting
14.5 hrs. On one cloudless day (13 March), the integrated PFD was 5.50 mol photons m-2 day-1. PFD as a function of the time of day is displayed in Figure 3.2. The PFDs experienced by F. concinna at 15-m depth are below the level necessary to achieve saturated photosynthetic rates, as illustrated by the net photosynthetic rates at the maximum PFD provided (Figure 4.2 inset). Thus, the daily productivity was calculated using only the initial slope of the P-I curve.
The daily energetic cost associated with aerobic respiration ranged from -181.7 J
(for a coral 8.8 cm2 in size) to -770.8 J (for a coral 67.2 cm2 in size), while the net energetic intake from photosynthesis ranged from -9.3 J (for a coral 8.8 cm2 in size) to
567.2 J (for a coral 67.2 cm2 in size). Assuming that respiratory cost in the light is equivalent to that in the dark, the gross daily intake ranged from 88.2 J (for a coral 8.8 cm2 in size) to 984.5 J (for a coral 67.2 cm2 in size). For the same size range, the daily energetic cost associated with tissue growth and Symbiodinium replication ranged from
-22.2 J to -98.3 J, and -0.8 J to 3.7 J, respectively. Summing the costs represented by respiration, tissue growth and symbiont division results in a total cost of -203.7 J and
-872.8 J for a coral 8.8 cm2 and 67.2 cm2 in size, respectively.
For the purpose of statistical comparison, the double logarithmic plot of energetic cost and intake against tissue biomass is depicted in Figure 4.6. Energetic cost and intake both scaled allometrically with tissue biomass (Table 4.2). The scaling exponent of cost
62 is significantly lower than intake (t = 2.44, df = 14, p < 0.05), in contrast to the expectation that the exponent of cost would be higher than the exponent of intake.
Daily energetic cost (respiration), total cost (respiration + growth), and intake
(gross photosynthesis) are plotted against the dry tissue biomass (g) of individual F. concinna in Figure 3.8 inset; all three regressions are significant (all r2 > 0.82, F > 28, df
= 7, p < 0.01). The cost and intake regression lines cross at 188 mg, or a coral size of
16.6 cm2 (Fig. 4.7 inset), indicating the size at which individuals gain an energetic surplus that apparently increases for the remainder of the size range studied. Below this size, metabolic cost is greater than gross photosynthetic intake. Upon reaching a size of
23.3 cm2, F. concinna attain an energetic surplus beyond the total cost associated with respiration and growth (Fig. 4.7 inset), such that corals 30, 45, and 60 cm2 in size have
33, 108, and 182 J of surplus energy, respectively.
Population structure
A total of 210 F. concinna at a population density of 9.1 ± 1.2 individuals m-2
(mean ± SE; n = 12) were measured at the study site. The smallest and largest corals were 0.2 and 140.0 cm2, respectively; the median size was 27.3 cm2. A histogram of the population size structure is presented in Figure 4.7. Thirty-eight of these individuals still were attached to the substratum (i.e., acanthocaulus stage), and did not exceed a size of
11.6 cm2. The greatest number of individuals (69) for a given size class were <10 cm2, which corresponds to a diameter of 36 mm. Although only 32 individuals in the population (17%) exceeded the size of the largest coral used in the metabolic trials (67.2
63 cm2), the results of the energetic analyses must be interpreted with caution because the entire size range of corals could not be tested with the available equipment.
64 Discussion
Energetic cost and intake clearly change at different rates with respect to size in
Fungia concinna. On a cloudless day at 15-m depth, F. concinna larger than 16.6 cm2 in surface area attain an energetic surplus (sensu Sebens 1982) that is enhanced with a further increase in size. In light of the prevailing modular growth form of most anthozoans, large solitary polyps are expected to possess a competitive advantage over small solitary polyps (Sebens 1982), and this hypothesis is supported with respect to energetics in F. concinna. Initial empirical support for this hypothesis was drawn from the ability of the solitary anemone Anthopleura xanthogrammica to increase its prey size range as it grew larger (Sebens 1981). In contrast, this study demonstrates that autotrophy alone in F. concinna lends an energetic advantage to larger individuals. The size advantage for Fungia is not limited to energetics; large F. scutaria are better inter- specific competitors than small conspecifics by using nematocyst-laden mucus to induce tissue damage on encroaching colonial corals (Chadwick 1988). It seems reasonable to suggest that once Fungia become too large to escape competitive interactions effectively
(Chadwick 1988), they must aggressively defend their position by allocating part of their energetic surplus towards mucus production.
The apparent lack of energetic constraints on the maximum size of F. concinna must be interpreted cautiously because the entire size range of corals was not studied.
However, a critical assumption in Sebens’ energetic model, that the scaling exponent on mass is greater for the cost function than the intake function (Sebens 1982), was not supported by the present data. Beyond a critical size, the energetic surplus increased indefinitely (Fig. 4.7 inset), rather than reach a maximum and then decline (Fig. 4.1). For
65 passive suspension feeders, Sebens’ (1979) assumption almost certainly is met because the surface area of the feeding structure cannot increase indefinitely. Perhaps the ideal scleractinian comparison to A. xanthogrammica for Sebens’ model would be the fungiid
Heliofungia actiniformis. This solitary species has tentacles that can reach lengths greater than 40 mm (Wells 1966), and thus the potential for the scaling of prey capture size exists (Sebens 1981).
Instead, the goal of this study was to apply Sebens’ model to an anthozoan that relies heavily on autotrophy. Fungia scutaria can obtain the majority of carbon required for animal respiration from photosynthesis (Muscatine et al. 1981), and continue growth without heterotrophic input (Franzisket 1969). For F. concinna, prey were assumed to be photons instead of mussels, and the feeding structure was considered to be photosynthetic machinery rather than tentacular surface area. In this case, the feeding structure essentially is defined as Symbiodinium population density and chlorophyll a content, and both of these photosynthetic traits increased isometrically with surface area. In contrast, tissue biomass increased disproportionately faster than surface area, suggesting that surface constraints limited respiratory cost, which is thought to be the case for most unitary animals (Schmidt-Nielsen 1984). Alternatively, the presence of thick tissue that was relatively inactive metabolically may have contributed to the lower mass specific energetic cost in large corals. In support of this hypothesis, deep tissue in corals with perforate skeletons (e.g., Porites) is thought to contribute to low respiration rates relative to corals that have a thin layer of tissue lying on top of an imperforate skeleton (e.g.,
Montipora, Pocillopora) [Davies 1991].
66 Sebens’ energetic surplus for animals that grow indeterminately, defined as the difference between the energetic intake of an organism and its metabolic maintenance cost, can be used for a variety of functions, especially growth or reproduction. To further define the energetic surplus for F. concinna, the cost of growth was calculated using an energy budget approach previously applied to other scleractinian corals, which has demonstrated that many shallow water coral species can be fully autotrophic with respect to energetic requirements (Edmunds & Davies 1986, Davies 1991). When the investment associated with tissue and symbiont growth was incorporated into Sebens’ energetic model, large corals still maintained an increasing energetic surplus. The persistence of this surplus, despite the inclusion of growth costs into the model, reinforces the hypothesis that energy was not setting the upper size limit of F. concinna in the present study.
An energetic surplus may be especially important for unattached, mobile corals.
The fitness benefits associated with locomotion (Chadwick-Furman & Loya 1992,
Yamashiro & Nishihara 1995, Goffredo & Chadwick-Furman 2000), excavation from burial (Goreau & Yonge 1968, Hubbard 1972), and sediment rejection (Fisk 1983,
Stafford-Smith 1993) likely are higher for Fungia than other scleractinian corals due to their mobility and frequent inhabitation of sandy substratum (Fisk 1983, Chadwick-
Furman & Loya 1992). Once these costs are met, extra energetic resources presumably can be allocated to reproduction. Fungia are capable of sexual and asexual reproduction
(Krupp 1983, Kramarsky-Winter & Loya 1996, 1998), but asexual budding generally is considered a stress response to partial tissue death and skeletal breakage (Chadwick &
Loya 1990, Jokiel et al. 1993, Kramarsky-Winter & Loya 1996). No corals in this study
67 were observed to be budding, and thus energetic costs associated with asexual reproduction are considered to be negligible.
Although photosynthetic input alone met respiratory and growth demands for the majority of the size range studied, the smallest corals demonstrated an energy deficit. For these small corals, heterotrophic sources of energy, such as zooplankton, dissolved organic material or bacteria may determine their survival, but were not quantified in this study. Assuming that size-specific mortality rates are reflected in the size-frequency distribution of the population by the relative number of individuals in each size class
(Fadlallah 1983), the smallest F. concinna (<10 cm2) experience intense mortality pressure, as do F. granulosa (Chadwick-Furman et al. 2000). Therefore, the energetic deficit in small F. concinna may reflect the well-known selective pressure for rapid growth, in order to escape the mortality risks associated with small size (Jackson 1977).
Indeed, the smallest corals exhibited disproportionately high calcification rates, and growth declines considerably in larger corals. Similarly, linear growth rates of F. granulosa decline over the entire size range (Chadwick-Furman et al. 2000).
The relatively low calcification rates of large F. concinna suggest that the maximum individual size reflects a real constraint on growth. Furthermore, Fungia and other free-living species exhibit upper size limits (Chadwick-Furman and Loya 1992,
Chadwick-Furman et al. 2000 and references therein), reflecting the commonality of determinate growth in unattached, mobile corals. The results of this study suggest that maximum F. concinna size is not constrained by energetics, especially because heterotrophic input was not quantified and thus energetic intake was likely underestimated. Energy in the form of translocated carbon from the symbiont often
68 comprises the majority necessary for the daily needs of colonial scleractinians (Muscatine et al. 1981, Davies 1991), and excess carbon can be stored as lipid reserves (Crossland et al. 1980, Harland et al. 1992), which may explain the disproportionate amount of tissue biomass in large F. concinna. Instead, alternative hypotheses for the limitations on growth should be considered. For example, elements derived from heterotrophy, including nitrogen and phosphorus, may limit coral growth in oligotrophic reef environments (Johannes et al. 1970, Muscatine & Porter 1977). Alternatively, mechanical constraints may limit size in free-living Fungia. Natural selection may have inhibited indeterminate growth as an adaptation to prevent sinking in soft substrata
(Chadwick-Furman & Loya 1992). In support of this hypothesis, very few F. concinna in the present study exceeded sizes of 100 cm2, and these largest corals were particularly susceptible to burial due to their disproportionately heavy weight. Large skeletons likely impede locomotion and excavation activities in Fungia, both of which are advantageous behaviors for an unattached coral. Finally, space may be limiting the upper size of F. concinna, because the percent cover of scleractinians at the study site is high (>60%; pers. obs.). Additional energetic analyses using the largest corals in the population are necessary to confirm the lack of energetic constraints on size in F. concinna, and long- term field manipulations may be used to distinguish between the alternative competing hypotheses for maximum size.
69 Table 4.1. Sizes of Fungia concinna used in oxygen flux trials. Each individual coral was subjected to a flow speed necessary to achieve a uniform Reynolds environment (Re ≈ 1950). The Reynolds number was calculated using mean diameter as the characteristic length (see text).
Mean diameter (cm) Surface area (cm2) Chamber Flow speed (cm s-1) 3.35 8.8 Small 5.0 4.20 13.9 Small 4.0 5.15 20.8 Small 3.2 5.15 20.8 Small 3.2 6.45 32.7 Large 2.6 6.85 36.9 Large 2.4 7.70 46.6 Large 2.2 8.05 50.9 Large 2.1 9.25 67.2 Large 1.8
70 Table 4.2. Summary of Model II regression analyses and t-tests testing the scaling relationships (Ho: b = 1) in Fungia concinna. Significant departures from isometry are in bold. All variables were log transformed.
Independent (x) Dependent (y) n r2 b ± SE t p Surface area Symbiodinium 12 0.88 0.972 ± 0.104 0.27 0.792 Surface area Chlorophyll a 12 0.93 0.948 ± 0.081 0.64 0.538 Surface area Tissue biomass 12 0.97 1.588 ± 0.087 6.73 <0.001 Surface area Skeletal weight 28 0.98 1.526 ± 0.039 13.41 <0.001 Tissue biomass Calcification 24 0.73 0.592 ± 0.064 6.33 <0.001 Diameter Septa 42 0.88 0.934 ± 0.052 1.27 0.212 Tissue biomass Cost 9 0.70 0.457 ± 0.095 5.74 <0.001 Tissue biomass Intake 9 0.90 0.729 ± 0.085 3.17 0.016
71
Figure 4.1. Hypothetical energy intake and cost as a function of individual mass (after Sebens 1982). Es, the energetic surplus, is the difference between the intake and cost curves; Es is greatest at Mopt, the optimal size for the individual. The intake and cost curves cross at Mmax, the maximum size possible. The coefficients ai and ac are normalization constants; bi and bc are scaling exponents.
72 200 1 March - 13 March 13 March
) -1 s 150 -2
100 60
40
mol photons m ! 20
50 Oxygen flux 0 PFD ( -20 0 100 200 300 400 PFD 0
600 1200 1800
Time (hrs)
Figure 4.2. Photosynthetically active radiation (PFD) reaching the forereef at 15-m depth on the north coast of Moorea, French Polynesia. Closed circles are means ± SE over a 13-day summer period (1 March – 13 March), and open circles are PFD values for one cloudless summer day (13 March). Inset graph displays Fungia concinna net oxygen flux -2 -1 -2 -1 rates (nmol O2 cm min ) plotted against PFD (µmol photons m s ); values are means ± SE (n = 9). The solid regression line is plotted for the range of PFDs experienced by F. concinna at 15-m depth, and represents the initial slope of the photosynthesis-PFD curve. The dashed curve is a 2nd order polynomial function plotted for illustration of the entire curve. A linear regression provided a better fit over an ecologically relevant range of light (0 – 155 µmol photons m-2 s-1) than the hyperbolic tangent curve because photosynthetic rates did not saturate at the maximum PFD provided during trials.
73
8.5 4 A C
3.5 8 (# cells) 3 7.5 2.5
7 Symbiodinium 2 Log tissue biomass (mg)
Log 6.5 1.5 0.5 1 1.5 2 0 0.5 1 1.5 2 2.5
3.5 2.5 B D
g) 2 ! 3 ( a 1.5 2.5 1
2 0.5
Log chlorophyll Log skeletal weight (g)
1.5 0 0.5 1 1.5 2 0 0.5 1 1.5 2 2.5 Log surface area (cm2) Log surface area (cm2)
Figure 4.3. The linear relationship of (A) log Symbiodinium count (n = 12), (B) log chlorophyll a content (n =12), (C) log tissue biomass (n = 12), and (D) log skeletal weight (n = 28) against log surface area for Fungia concinna. The scaling exponents of Symbiodinium number and chlorophyll content on surface area are indistinguishable statistically from 1 (isometry), whereas the scaling exponents of tissue biomass and skeletal weight deviate significantly from 1 (allometry).
74 2.5 ) -1
2
1.5
1 Log calcification (mg day
0.5
1.5 2 2.5 3 3.5 4
Log tissue biomass (mg)
Figure 4.4. Log calcification rate plotted against log tissue biomass for 24 individuals of Fungia concinna. The slope of the Model II regression deviates significantly from 1, indicating allometric scaling.
75 1.4
1.2
1
0.8 Log septa (mm)
0.6
0.4 1.4 1.6 1.8 2 2.2 Log diameter (mm)
Figure 4.5. Log septa height plotted against log diameter for 42 individuals of Fungia concinna. The slope of the Model II regression is not distinguishable statistically from 1, indicating isometric scaling.
76 3.2 Cost Intake
2.8
2.4
Log daily energy (J) 2
1.6 2 2.4 2.8 3.2
Log tissue biomass (mg)
Figure 4.6. Log daily energetic cost and intake plotted against log tissue biomass for 9 individuals of Fungia concinna. The slopes of both Model II regressions deviate significantly from 1, indicating allometric scaling (Table 3.2). The slope of the cost regression is significantly lower (bc = 0.457 ± 0.095) than the slope of the intake regression (bi = 0.729 ± 0.085) [t = 2.44, df = 14, p < 0.05].
77
70 1000 Cost Total cost 60 Intake
50 500
40 Daily energy
Count 30 0 0 0.5 1 1.5 2 Biomass 20
10
0 0 20 40 60 80 100 120 140 2 Surface area (cm )
Figure 4.7. Size-frequency distribution of Fungia concinna (n = 210) at 15-m depth on the forereef of the north coast of Moorea, French Polynesia. Inset graph displays daily energetic (J) cost (r2 = 0.82), total cost (r2 = 0.87), and intake (r2 = 0.96) plotted against tissue biomass (g) for nine individuals. Cost refers to the metabolic cost measured as aerobic respiration, total cost is the sum of metabolic cost plus the energetic costs associated with daily growth, and intake is the energetic input from gross photosynthesis (see Methods for further details). Black arrows indicate the size (188 mg of tissue; 16.6 cm2 in surface area) at which the cost regression intercepts the intake regression, and the gray arrows indicate the size (449 mg of tissue; 23.3 cm2 in surface area) at which the total cost regression intercepts the intake regression.
78 CHAPTER 5
Concluding remarks
In this thesis, I set out to examine the effects of age and size on the growth and physiology of scleractinian corals. At least for one branching species and one massive species, older individuals have depressed growth rates relative to younger conspecifics
(Chapters 2, 3). Age also may affect Fungia, but in these solitary corals age typically is correlated with size and therefore could not be isolated from size effects. The most parsimonious hypothesis to explain the decline in growth with age is the cellular senescence of modules, which has been suggested previously for other modular organisms (Chapters 2, 3). Notably, age also affected the scaling of growth in
Siderastrea siderea and Madracis mirabilis (albeit not significantly in this species).
Size-specific calcification rates were higher for younger corals than for older corals in both species. Although the effect is small, it suggests that mortality risk may have selected for faster growth in young corals because they are necessarily small. The pressure for rapid growth decreases with size, and old corals that fragment to a smaller size probably cannot revert ontogenetically to a juvenile condition.
Scaling analyses were useful in comparing the effects of size on various physiological traits. A major consideration not previously discussed is how the metric of size chosen as the independent variable affects the scaling analysis. Traditionally, scaling relationships have been quantified with respect to organismal biomass, but other measures of size may be used if biologically appropriate (Schmidt-Nielsen 1984). For example, the relevant scale for tail beat frequency in fish was body length, rather than
79 mass (Schmidt-Nielsen 1975). In the case of modular invertebrates, such as some ascidians, bryozoans, and cnidarians, surface area is a biologically appropriate measure of size because their soma is effectively a two-dimensional layer. The surface area of a benthic invertebrate is important ecologically because it is directly related to the amount of space it occupies, and physiologically, diffusion through a boundary layer across a surface limits the mass transfer of substances in sessile aquatic organisms (Patterson
1992). However, the use of surface area, rather than mass, as an independent variable in scaling relationships may be misleading if one assumes that tissue biomass and surface area are related directly (i.e., isometry). For example, M. mirabilis displayed isometric scaling of calcification when plotted against surface area (Chapter 2). In contrast, calcification in F. concinna was allometric (Chapter 4). It is tempting to suggest that the difference is attributable to the modular nature of M. mirabilis, but if calcification instead is plotted against biomass (measured as protein content), the scaling exponent becomes allometric (data not shown). Conversely, if the independent variable in the F. concinna calcification regression is switched to surface area, the scaling exponent becomes isometric (data not shown), further supporting the hypothesis that skeletal deposition is an area related process (Chapter 2).
The root cause of this isometric-allometric switching of growth is the functional relationship between tissue biomass and surface area. In both F. concinna and M. mirabilis (data not shown) surface area is an allometric function of biomass, indicating that tissue increases disproportionately fast as these corals grow. Preliminary data (not shown) on calcification in S. siderea also support the dependency of growth scaling on the choice of surface area or biomass. Vollmer and Edmunds (2000) suggested that the
80 allometric scaling of respiration in juveniles of S. siderea was related to the allometric scaling of surface area against biomass. Vollmer and Edmunds (2000) also reanalyzed data from Jokiel and Morrisey (1986), and found both respiration and surface area to scale allometrically with biomass in Pocillopora damicornis. Similarly, the scaling exponents of respiration and photosynthesis decrease when regressed against biomass in
F. concinna (data not shown).
Clearly, the nature of the scaling relationship is dependent on the metric of size chosen. Although this may not seem that surprising, it highlights the need to validate the assumption of the “two-dimensional” structure of modular organisms. Evidence from four stony corals with very different morphologies (F. concinna, chapter 4; M. mirabilis, unpublished data; S. siderea and P. damicornis, Vollmer & Edmunds 2000) suggests that the allometric scaling of tissue biomass is common in the Order Scleractinia. However, this may not be the case for other modular cnidarians, ascidians or bryozoans, but is probably species dependent. In particular, the bryozoan Electra pilosa exhibited isometric scaling of respiration (Hughes & Hughes 1986), and was suggested to be due, in part, to a constant surface to volume ratio. Furthermore, the scaling exponent may depend on the level of module integration. For example, the ascidian Botrylloides simodensis retained surface to volume isometry, but the scaling of respiration switched between allometry and isometry (Nakaya et al. 2003). Respiration scaled allometrically when the zooids were interconnected, which typifies the ordinary state of ascidian organization (Nakaya et al. 2003). Respiration switched to isometry when the parent zooids underwent degeneration and gave rise to daughter zooids, during which the vascular system was inactive and being reconstructed (takeover state). These results
81 suggest that metabolic allometry was derived from mutual interaction between the zooids, but during takeover state, the colony is effectively an aggregation of independent units, thereby achieving metabolic isometry. In conclusion, the nature of scaling relationships clearly is context dependent, despite the theoretical expectation of isometry for modular organisms. The results of the present study suggest that age may also play a role in affecting the scaling of growth in scleractinian corals. Future work examining the effects of age and size on the physiology of modular organisms should provide further insight into the relative importance of each of these factors.
82 Literature Cited
Anthony KRN, Hoegh-Guldberg O (2003) Variation in coral photosynthesis, respiration and growth characteristics in contrasting light microhabitats: an analogue to plants in forest gaps and understoreys? Funct. Ecol. 17:246-259
Arnqvist G, Johansson F (1998) Ontogenetic reaction norms of predator-induced defensive morphology in dragonfly larvae. Ecology 79:1847-1858
Babcock RC (1991) Comparative demography of three species of scleractinian corals using age- and size-dependent classifications. Ecol. Mono. 61:225-244
Bak RPM (1983) Neoplasia, regeneration and growth in the reef-building coral Acropora palmata. Mar. Biol. 77:221-227
Bak RPM, Engel MS (1979) Distribution, abundance and survival of juvenile hermatypic corals (Scleractinia) and the importance of life history strategies in the parent community. Mar. Biol. 54:341-352
Baker AC (2003) Flexibility and specificity in coral-algal symbiosis: diversity, ecology, and biogeography of Symbiodinium. Annu. Rev. Ecol. Syst. 34:661-689
Barnes DJ, Lough JM (1992) Systematic variations in the depth of skeleton occupied by coral tissue in massive colonies of Porites from the Great Barrier Reef. J. Exp. Mar. Biol. Ecol. 159:113-128
Bellwood DR, Hughes TP, Folke C, Nyström M (2004) Confronting the coral reef crisis. Nature 429:827-833
Blackstone NW (1996) Gastrovascular flow and colony development in two colonial hydroids. Biol. Bull. 190:56-68
Boardman RS, Cheetham AH (1973) Degrees of colony dominance in Stenolaemate and Gymnolaemate Bryozoa. In: Boardman RS, Cheetham AH, Oliver WA (eds) Animal Colonies: development and function through time. Dowden, Hutchinson & Ross, Stroudsburg, p 121-220
Borowitzka MA, Larkum AWD (1976) Calcification in the green alga Halimeda. J. Exp. Bot. 27:864-878
83 Bradford MM (1976) A rapid and sensitive method for the quantification of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72:248-254
Brown JH, West GB, Enquist BJ (2000) Scaling in biology: patterns, processes, causes and consequences. In: Brown JH, Enquist BJ (eds) Scaling in Biology. Oxford University Press, New York
Bruno JF (1998) Fragmentation in Madracis mirabilis (Duchassaing and Michelotti): how common is size-specific fragmentation survivorship in corals? J. Exp. Mar. Biol. Ecol. 230:169-181
Bruno JF, Edmunds PJ (1997) Clonal variation for phenotypic plasticity in the coral Madracis mirabilis. Ecology 78:2177-2190
Bruno JF, Edmunds PJ (1998) Metabolic consequences of phenotypic plasticity in the coral Madracis mirabilis (Duchassaing and Michelotti): the effect of morphology and water flow on aggregate respiration. J. Exp. Mar. Biol. Ecol. 229:187-195
Cadavid LF (2004) Self-discrimination in colonial invertebrates: genetic control of allorecognition in the hydroid Hydractinia. Dev. Comp. Immunol. 28:871-879
Campbell RD (1968) Colony growth and pattern in the two-tentacled hydroid, Proboscidactyla flavicirrata. Biol. Bull. 135:96-104
Carpenter RC, Hackney JM, Adey WH (1991) Measurements of primary productivity and nitrogenase activity of coral reef algae in a chamber incorporating oscillatory flow. Limn. Oceanog. 36:40-49
Caswell H (1985) The evolutionary demography of clonal reproduction. In: Jackson JBC, Buss LW, Cook RE (eds) Population Biology and Evolution of Clonal Organisms. Yale University Press, New Haven
Chadwick NE (1988) Competition and locomotion in a free-living fungiid coral. J. Exp. Mar. Biol. Ecol. 123:189-200
Chadwick NE, Loya Y (1990) Regeneration after experimental breakage in the solitary reef coral Fungia granulosa Klunzinger, 1879. J. Exp. Mar. Biol. Ecol. 142:221- 234
84 Chadwick-Furman NE, Goffredo S, Loya Y (2000) Growth and population dynamic model of the reef coral Fungia granulosa Klunzinger, 1879 at Eilat, northern Red Sea. J. Exp. Mar. Biol. Ecol. 249:199-218
Chadwick-Furman NE, Loya Y (1992) Migration, habitat use, and competition among mobile corals (Scleractinia: Fungiidae) in the Gulf of Eilat, Red Sea. Mar. Biol. 114:617-623
Chadwick-Furman NE, Weissman IL (2003) Effects of allogeneic contact on life-history traits of the colonial ascidian Botryllus schlosseri in Monterey Bay. Biol. Bull. 205:133-143
Chalker BE (1981) Simulating light-saturation curves for photosynthesis and calcification by reef-building corals. Mar. Biol. 63:135-141
Crossland CJ, Barnes DJ, Borowitzka MA (1980) Diurnal lipid and mucus production in the staghorn coral Acropora acuminata. Mar. Biol. 60:81-90
Crowell S (1953) The regression-replacement cycle of hydranths of Obelia and Campanularia. Physiol. Zool. 26:319-327
Davies PS (1984) The role of zooxanthellae in the nutritional energy requirements of Pocillopora eydouxi. Coral Reefs 2:181-186
Davies PS (1989) Short-term growth measurements of corals using an accurate buoyant weight technique. Mar. Biol. 101:389-395
Davies PS (1991) Effect of daylight variations on the energy budgets of shallow-water corals. Mar. Biol. 108:137-144
Dennison WC, Barnes DJ (1988) Effect of water motion on coral photosynthesis and calcification. J. Exp. Mar. Biol. Ecol. 115:62-77
Denny MW (1988) Biology and the Mechanics of the Wave-swept Environment, Vol. Princeton University Press, Princeton
Denny MW, Daniel TL, Koehl MAR (1985) Mechanical limits to size in wave-swept organisms. Ecol. Mono. 55:69-102
85 Diekmann OE, Olsen JL, Stam WT, Bak RPM (2003) Genetic variation within Symbiodinium clade B from the coral genus Madracis in the Caribbean (Netherlands Antilles). Coral Reefs 22:29-33
Downs CA, Fauth JE, Halas JC, Dustan P, Bemiss J, Woodley CM (2002) Oxidative stress and seasonal coral bleaching. Free Radic. Biol. Med. 33:533-543
Edmunds PJ (2000a) Patterns in the distribution of juvenile corals and coral reef community structure in St. John, U.S. Virgin Islands. Mar. Ecol. Prog. Ser. 202:113-124
Edmunds PJ (2000b) Recruitment of scleractinians onto the skeletons of corals killed by black band disease. Coral Reefs 19:69-74
Edmunds PJ (2002) Long-term dynamics of coral reefs in St. John, U.S. Virgin Islands. Coral Reefs 21:357-367
Edmunds PJ (2004) Juvenile coral population dynamics track rising seawater temperature on a Caribbean reef. Mar. Ecol. Prog. Ser. 269:111-119
Edmunds PJ, Davies PS (1986) An energy budget for Porites porites (Scleractinia). Mar. Biol. 92:339-347
Edmunds PJ, Davies PS (1988) Post-illumination stimulation of respiration rate in the coral Porites porites. Coral Reefs 7:7-9
Edmunds PJ, Gates RD (2004) Size-dependent differences in the photophysiology of the reef coral Porites astreoides. Biol. Bull. 206:61-64
Elliot JM, Davison W (1975) Energy equivalents of oxygen consumption in animal energetics. Oecologia 19:195-201
Fadlallah YH (1983) Population dynamics and life history of a solitary coral, Balanophyllia elegans, from Central California. Oecologia 58:200-207
Fisk DA (1983) Free-living corals: distributions according to plant cover, sediments, hydrodynamics, depth and biological factors. Mar. Biol. 74:287-294
86 Foster AB (1979) Phenotypic plasticity in the reef corals Montastraea annularis (Ellis & Solander) and Siderastrea siderea (Ellis & Solander). J. Exp. Mar. Biol. Ecol. 39:25-54
Franzisket L (1969) Reef corals can live as autotrophs. Die Naturwiss. 56:144
Gardner TA, Cote IM, Gill JA, Grant A, Watkinson AR (2003) Long-term region-wide declines in Caribbean corals. Science 301:958-960
Gateño D, Rinkevich B (2003) Coral polyp budding is probably promoted by a canalized ratio of two morphometric fields. Mar. Biol. 142:971-973
Gattuso J-P, 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. 39:160-183
Gayon J (2000) History of the concept of allometry. Am. Zool. 40:748-758
Goffredo S, Chadwick-Furman NE (2000) Abundance and distribution of mushroom corals (Scleractinia: Fungiidae) on a coral reef at Eilat, Northern Red Sea. Bull. Mar. Sci. 66:241-254
Goreau TF, Yonge CM (1968) Coral community on muddy sand. Nature 217:421-423
Gould SJ (1966) Allometry and size in ontogeny and phylogeny. Biol. Rev. Camb. Philos. Soc. 41:587-640
Harland AD, Davies PS, Fixter LM (1992) Lipid content of some Caribbean corals in relation to depth and light. Mar. Biol. 113:357-361
Harper JL (1977) Population biology of plants, Vol. Academic Press, London
Harper JL (1985) Modules, branches, and the capture of resources. In: Jackson JBC, Buss LW, Cook RE (eds) Population Biology and Evolution of Clonal Organisms. Yale University Press, New Haven
Harper JL, Rosen BR, White J (1986) Preface to the growth and form of modular organisms. Philos. Trans. R. Soc. Lond. B. Biol. Sci. 313:3-5
87 Harvell CD, Kim K, Burkholder JM, Colwell RR, Epstein PR, Grimes DJ, Hofmann EE, Lipp EK, Osterhaus ADME, Overstreet RM, Porter JW, Smith GW, Vasta GR (1999) Emerging marine diseases - climate links and anthropogenic factors. Science 285:1505-1510
Helmuth B, Sebens KP (1993) The influence of colony morphology and orientation to flow on particle capture by the scleractinian coral Agaricia agaricites (Linnaeus). J. Exp. Mar. Biol. Ecol. 165:251-278
Helmuth BST, Timmerman BEH, Sebens KP (1997) Interplay of host morphology and symbiont microhabitat in coral aggregations. Mar. Biol. 130:1-10
Highsmith RC (1982) Reproduction by fragmentation in corals. Mar. Ecol. Prog. Ser. 7:207-226
Hoegh-Guldberg O (1999) Climate change, coral bleaching and the future of the world's coral reefs. Mar. Fres. Res. 50:839-866
Holliday R (1969) Errors in protein synthesis and clonal senescence in fungi. Nature 221:1224-1228
Hubbard JAE (1972) Diaseris distorta, an "acrobatic" coral. Nature 236:457-459
Hughes DJ, Hughes RN (1986) Metabolic implication of modularity: studies on the respiration and growth of Electra pilosa. Philos. Trans. R. Soc. Lond. B. Biol. Sci. 313:23-29
Hughes KA, Reynolds RM (2005) Evolutionary and mechanistic theories of aging. Annu. Rev. Entomol. 50:421-445
Hughes RN, Cancino JM (1985) An ecological overview of cloning in Metazoa. In: Jackson JBC, Buss LW, Cook RE (eds) Population Biology and Evolution of Clonal Organisms. Yale University Press, New Haven
Hughes TP (1984) Population dynamics based on individual size rather than age: a general model with a reef coral example. Am. Nat. 123:778-795
88 Hughes TP, Baird AH, Bellwood DR, Card M, Connolly SR, Folke C, Grosberg R, Hoegh-Guldberg O, Jackson JBC, Kleypas JA, Lough JM, Nystrom M, Palumbi SR, Pandolfi JM, Rosen B, Roughgarden J (2003) Climate change, human impacts, and the resilience of coral reefs. Science 301:929-933
Hughes TP, Connell JH (1987) Population dynamics based on size or age? A reef coral analysis. Am. Nat. 129:818-829
Hughes TP, Jackson JBC (1980) Do corals lie about their age? Some demographic consequences of partial mortality, fission, and fusion. Science 209:713-715
Hughes TP, Jackson JBC (1985) Population dynamics and life histories of foliaceous corals. Ecol. Mono. 55:141-166
Hughes TP, Tanner JE (2000) Recruitment failure, life histories, and long-term decline of Caribbean corals. Ecology 81:2250-2263
Huston M (1985) Variation in coral growth rates with depth at Discovery Bay, Jamaica. Coral Reefs 4:19-25
Jackson JBC (1977) Competition on marine hard substrata: the adaptive significance of solitary and colonial strategies. Am. Nat. 111:743-767
Jackson JBC (1985) Distribution and ecology of clonal and aclonal benthic invertebrates. In: Jackson JBC, Buss LW, Cook RE (eds) Population Biology and Evolution of Clonal Organisms. Yale University Press, New Haven
Jackson JBC, Buss LW, Cook RE (eds) (1985) Population Biology and Evolution of Clonal Organisms, Vol. Yale University Press, New Haven
Jackson JBC, Coates AG (1986) Life cycles and evolution of clonal (modular) animals. Philos. Trans. R. Soc. Lond. B. Biol. Sci. 313:7-22
Jeffrey SW, Humphrey GF (1975) New spectrophotometric equations for determining chlorophylls a, b, c1 and c2 in higher plants, algae and natural phytoplankton. Biochem. Physiol. Pflanz. 167:191-194
89 Johannes RE, Coles SL, Kuenzel NT (1970) The role of zooplankton in the nutrition of some scleractinian corals. Limn. Oceanog. 15:822-824
Johannes RE, Wiebe WJ (1970) Method for determination of coral tissue biomass and composition. Limn. Oceanog. 15:822-824
Jokiel PL, Cowdin HP (1976) Hydromechanical adaptation in the solitary free-living coral Fungia scutaria. Nature 262:212-213
Jokiel PL, Hunter CL, Taguchi S, Watari L (1993) Ecological impact of a fresh water "reef-kill" in Kaneohe Bay, Oahu, Hawaii. Coral Reefs 12:177-184
Jokiel PL, Morrissey JM (1986) Influence of size on primary production in the reef coral Pocillopora damicornis and the macroalga Acanthophora spicifica. Mar. Biol. 91:15-26
Jones CG, Lawton JH, Shachak M (1997) Positive and negative effects of organisms as physical ecosystem engineers. Ecology 78:1946-1957
Kaandorp JA (1999) Morphological analysis of growth forms of branching sessile organisms along environmental gradients. Mar. Biol. 134:295-306
Kaandorp JA, Sloot PMA, Merks RMH, Bak RPM, Vermeij MJA, Maier C (2005) Morphogenesis of the branching reef coral Madracis mirabilis. Proc. R. Soc. Lond. B. Biol. Sci. 272:127-133
Kim K, Lasker HR (1998) Allometry of resource capture in colonial cnidarians and constraints of allometric growth. Funct. Ecol. 12:646-654
Kinzie RA, Sarmiento T (1986) Linear extension rate is independent of colony size in the coral Pocillopora damicornis. Coral Reefs 4:177-181
Kojis BL, Quinn NJ (1985) Puberty in Goniastrea favulus age or size limited? Proceedings of the Fifth International Coral Reef Congress, Tahiti, p 289-293
Kramarsky-Winter E, Loya Y (1996) Regeneration versus budding in fungiid corals: a trade-off. Mar. Ecol. Prog. Ser. 134:179-185
90 Kramarsky-Winter E, Loya Y (1998) Reproductive strategies of two fungiid corals from the northern Red Sea: environmental constraints? Mar. Ecol. Prog. Ser. 174:175- 182
Krupp DA (1983) Sexual reproduction and early development of the solitary coral Fungia scutaria (Anthozoa: Scleractinia). Coral Reefs 2:159-164
Lehninger AL (1973) Bioenergetics, Vol. Benjamin, Elmsford
Leopold AC (1961) Senescence in plant development. Science 134:1727-1732
Lesser MP (1997) Oxidative stress causes coral bleaching during exposure to elevated temperatures. Coral Reefs 16:187-192
LeTissier MDA (1988) The growth and formation of branch tips of Pocillopora damicornis (Linnaeus). J. Exp. Mar. Biol. Ecol. 124:115-131
Littler MM, Littler DS, Brooks BL, Koven JF (1997) A unique coral reef formation discovered on the Great Astrolabe Reef, Fiji. Coral Reefs 16:51-54
Marsh JA (1970) Primary productivity of reef-building calcareous red algae. Ecology 51:255-263
Martínez DE (1998) Mortality patterns suggest lack of senescence in hydra. Exp. Gerontol. 33:217-225
Masuda K, Goto M, Maruyama T, Miyachi S (1993) Adaptation of solitary corals and their zooxanthellae to low light and UV radiation. Mar. Biol. 117:685-691
McCardle BH (1988) The structural relationship: regression in biology. Can. J. Zool. 66:2329-2339
McMahon T (1973) Size and shape in biology. Science 179:1201-1204
Medawar PB (1952) An unsolved problem in biology, Vol. H.K. Lewis, London
91 Meesters EH, Bak RPM (1995) Age-related deterioration of a physiological function in the branching coral Acropora palmata. Mar. Ecol. Prog. Ser. 121:203-209
Muscatine L (1990) The role of symbiotic algae in carbon and energy flux in reef corals. In: Dubinsky Z (ed) Ecosystems of the World: Coral Reefs, Vol. 25. El Sevier, Amsterdam, p 75-87
Muscatine L, Hand C (1958) Direct evidence for the transfer of materials from symbiotic algae to the tissues of a coelenterate. Proc. Natl. Acad. Sci. U. S. A. 44:1259-1263
Muscatine L, McCloskey LR, Marian RE (1981) Estimating the daily contribution of carbon from zooxanthellae to coral animal respiration. Limn. Oceanog. 26:601- 611
Muscatine L, Porter JW (1977) Reef corals: mutualistic symbiosis adapted to nutrient poor environments. Bioscience 27:454-460
Müller WA, Teo R, Frank U (2004) Totipotent migratory stem cells in a hydroid. Dev. Biol. 275:215-224
Nakaya F, Saito Y, Motokawa T (2003) Switching of metabolic-rate scaling between allometry and isometry in colonial ascidians. Proc. R. Soc. Lond. B. Biol. Sci. 270:1105-1113
Ostrowski M-F, Jarne P, Berticat O, David P (2002) Ontogonetic reaction norm for binary traits: the timing of phallus development in the snail Bulinus truncatus. Heredity 88:342-348
Palumbi SR, Jackson JBC (1983) Aging in modular organisms: ecology of zooid senescence in Steginoporella sp. (Bryozoa: Cheilostomata). Biol. Bull. 164:267- 278
Pandolfi JM, Bradbury RH, Sala E, Hughes TP, Bjorndal KA, Cooke RG, McArdle D, McClenachan L, Newman MJH, Paredes G, Warner RR, Jackson JBC (2003) Global trajectories of the long-term decline of coral reef ecosystems. Science 301:955-958
92 Pandolfi JM, Jackson JBC, Baron N, Bradbury RH, Guzman HM, Hughes TP, Kappel CV, Micheli F, Ogden JC, Possingham HP, Sala E (2005) Are U.S. coral reefs on the slippery slope to slime? Science 307:1725-1726
Patterson MR (1992) A mass transfer explanation of metabolic scaling in some aquatic invertebrates and algae. Science 255:1421-1423
Patterson MR, Sebens KP, Olson RR (1991) In situ measurements of flow effects on primary production and dark respiration in reef corals. Limn. Oceanog. 36:936- 948
Patton JS, Abraham S, Benson AA (1977) Lipogenesis in the intact coral Pocillopora capitata and its isolated zooxanthellae: evidence for a light-driven cycle between symbiont and host. Mar. Biol. 44:235-247
Pearse VB, Muscatine L (1971) Role of symbiotic algae (zooxanthellae) in coral calcification. Biol. Bull. 141:350-363
Pigliucci M, Schlichting CD (1995) Ontogonetic reaction norms in Lobelia siphlitica (Lobeliaceae): response to shading. Ecology 76:2134-2144
Porter JW, Fitt WK, Spero HJ, Rogers CS, White MW (1989) Bleaching in reef corals: physiological and stable isotopic responses. Proc. Natl. Acad. Sci. U. S. A. 86:9342-9346
Quinn GP, Keough MJ (2002) Experimental Design and Data Analysis for Biologists, Vol. Cambridge University Press, Cambridge
Raffaelli D, Hawkins S (1996) Intertidal Ecology, Vol. Chapman & Hall, London
Riisgård HU (1998) No foundation of a "3/4 power scaling law" for respiration in biology. Ecol. Lett. 1:71-73
Riley JP, Skirrow G (1975) Chemical Oceanography, Vol 4. Academic Press, New York
Rinkevich B, Lauzon RJ, Brown BWM, Weissman IL (1992) Evidence for a programmed life span in a colonial protochordate. Proc. Natl. Acad. Sci. U. S. A. 89:3546-3550
93 Rinkevich B, Loya Y (1986) Senescence and dying signals in a reef building coral. Experientia 42:320-322
Ryland JS (1979) Structural and physiological aspects of coloniality in Bryozoa. In: Larwood G, Rosen BR (eds) Biology and Systematics of Colonial Organisms. Academic Press, New York, p 211-242
Sabbadin A (1979) Colonial structure and genetic patterns in ascidians. In: Larwood G, Rosen BR (eds) Biology and Systematics of Colonial Organisms. Academic Press, New York, p 433-444
Schlichting CD (1986) The evolution of phenotypic plasticity in plants. Annu. Rev. Ecol. Syst. 17:667-693
Schlichting CD, Pigliucci M (1998) Phenotypic evolution: A reaction norm perspective, Vol. Sinauer Associates, Inc, Sunderland
Schmidt-Nielsen K (1975) Scaling in biology: the consequences of size. J. Exp. Zool. 194:287-308
Schmidt-Nielsen K (1984) Scaling. Why is animal size so important? Vol. Cambridge University Press, Cambridge
Sebens KP (1979) The energetics of asexual reproduction and colony formation in benthic marine invertebrates. Am. Zool. 19:683-697
Sebens KP (1981) The allometry of feeding, energetics, and body size in three sea anemone species. Biol. Bull. 161:152-171
Sebens KP (1982) The limits to indeterminate growth: an optimal size model applied to passive suspension feeders. Ecology 63:209-222
Sebens KP (1987a) Coelenterata. In: Pandian TJ, Vernberg FJ (eds) Animal Energetics, Vol. 1. Academic Press, San Diego, p 55-120
Sebens KP (1987b) The ecology of indeterminate growth in animals. Annu. Rev. Ecol. Syst. 18:371-407
94 Sebens KP (2002) Energetic constraints, size gradients, and size limits in benthic marine invertebrates. Integ. Compar. Biol. 42:853-861
Sebens KP, Helmuth B, Carrington E, Agius B (2003) Effects of water flow on growth and energetics of the scleractinian coral Agaricia tenuifolia in Belize. Coral Reefs 22:35-47
Sebens KP, Witting J, Helmuth B (1997) Effects of water flow and branch spacing on particle capture by the reef coral Madracis mirabilis (Duchassaing and Michelotti). J. Exp. Mar. Biol. Ecol. 211:1-28
Sohal RS, Mockett RJ, Orr WC (2002) Mechanisms of aging: an appraisal of the oxidative stress hypothesis. Free Radic. Biol. Med. 33:575-586
Sokal RR, Rohlf FJ (1995) Biometry, Vol 3. W. H. Freeman and Co., New York
Soong K (1993) Colony size as a species character in massive reef corals. Coral Reefs 12:77-83
Soong K, Lang JC (1992) Reproductive integration in reef corals. Biol. Bull. 183:418- 431
Stafford-Smith MG (1993) Sediment-rejection efficiency of 22 species of Australian scleractinian corals. Mar. Biol. 115:229-243
Sánchez JA (2003) Patterns of morphological integration in marine modular organisms: supra-module organization in branching octocoral colonies. Proc. R. Soc. Lond. B. Biol. Sci. 270:2039-2044
Tanner JE (1999) Density-dependent population dynamics in clonal organisms: a modelling approach. J. Anim. Ecol. 68:390-399
Tanner JE (2002) Consequences of density-dependent heterotrophic feeding for a partial autotroph. Mar. Ecol. Prog. Ser. 227:293-304
Van Winkle DH, Blackstone NW (2002) Variation in growth and competitive ability between sexually and clonally produced hydroids. Biol. Bull. 202:156-165
95 Veron JEN (2000) Corals of the World, Vol. Australian Institute of Marine Science, Townsville
Vollmer SV (1999) Physiological scaling in small scleractinians. M.S. thesis, California State University, Northridge
Vollmer SV, Edmunds PJ (2000) Allometric scaling in small colonies of the scleractinian coral Siderastrea siderea (Ellis and Solander). Biol. Bull. 199:21-28
Warner ME, Fitt WK, Schmidt GW (1999) Damage to photosystem II in symbiotic dinoflagellates: a determinant of coral bleaching. Proc. Natl. Acad. Sci. U. S. A. 96:8007-8012
Watkinson AR, White J (1986) Some life-history consequences of modular construction in plants. Philos. Trans. R. Soc. Lond. B. Biol. Sci. 313:31-51
Watt AS (1947) Contribution to the ecology of bracken. IV. The structure of the community. New Phytol. 46:97-121
Wells JW (1966) Evolutionary development in the scleractinian family Fungiidae. Symposia of the Zoological Society of London 16:223-246
West GB, Brown JH, Enquist BJ (1999) The fourth dimension of life: fractal geometry and allometric scaling of organisms. Science 284:1677-1679
Witman JD (1992) Physical disturbance and community structure of exposed and protected reefs: a case study from St. John, U.S. Virgin Islands. Am. Zool. 32:641-654
Yamashiro H, Nishihara M (1995) Phototaxis in Fungiidae corals (Scleractinia). Mar. Biol. 124:461-464
Yap HT, Alvarez RM, Custodio HM, Dizon RM (1998) Physiological and ecological aspects of coral transplantation. J. Exp. Mar. Biol. Ecol. 229:69-84
Zeuthen E (1953) Oxygen uptake as related to body size in organisms. Q. Rev. Biol. 28:1-12
96