Pseudogenes Contribute to the Extreme Diversity of Nuclear RibosomalDNA in the Hard Coral Acropora

Luis M. Marquez,*1 David J. Miller,� Jason B. MacKenzie,*� and Madeleine J. H. van Oppen§*Biochemistry and Molecular Biology, �Comparative Genomics Centre, and �Marine Biology and Aquaculture, James CookUniversity, Townsville, Queensland, Australia; and §Australian Institute of Marine Science, Townsville, Queensland, Australia

One characteristic of Indo-Pacific Acropora spp. is extremely high diversity in rDNA sequences at both the species andindividual levels. In order to test the hypothesis that pseudogenes may contribute to this diversity, three kinds of analyseswere conducted. First, for A. millepora (the species containing the most diverse suite of rDNA types), RT-PCR was usedto determine which 5.8S rDNA types are expressed. Second, as previous studies have indicated that interspecifichybridization has occurred in the genus Acropora and silencing of rDNA loci via nucleolar dominance has been shown insome cases to involve methylation, patterns of variation were examined at methylation-susceptible sites. Third, patternsof substitution at conserved sites (including those that are likely to contribute to secondary structure in rRNA) in the 5.8SrDNA were examined. These analyses consistently indicated that one rDNA sequence type present in a broad range ofIndo-Pacific Acropora species is likely to consist predominantly of pseudogenes. Patterns of variation also suggest thatspecies may differ with respect to which rDNA sequence types have been silenced and which are active. Thesepseudogenes are likely to have arisen as a consequence of the introduction of highly divergent rDNA types into singlegenomes by interspecific hybridization events, and we attribute the extreme rDNA diversity characteristic of manyAcropora species to both the independent evolution of these silenced rDNA types and to the suppressive effects of highsequence diversity on homogenization processes acting on functional loci.

Introduction

To date, the most frequently used DNA marker tostudy reef coral evolution is the internal transcribed spacer(ITS1-5.8S-ITS2) region of the nuclear ribosomal RNA(rRNA) transcription unit (Hunter, Morden, and Smith1997; Lopez and Knowlton 1997; Odorico and Miller1997; Medina, Weil, and Szmant 1999; van Oppen et al.2000, 2002; Diekmann et al. 2001; Rodriguez-Lanetty andHoegh-Guldberg 2002). rDNA constitutes a multigenefamily of tandem repeats, which evolve in a concertedfashion (Arnheim et al. 1980) through unequal crossingover and gene conversion (Dover 1982). As a result ofconcerted evolution, all copies of rDNA families aregenerally rapidly homogenized within individuals andspecies, but interspecific divergence can be high (Hillis andDixon 1991).

The presence of highly divergent rRNA types withina single genome can, however, result in homogenizationproceeding extremely slowly (Modrich and Lahue 1996),particularly if these are on different chromosomes(Arnheim et al. 1980). Hence, ancestral polymorphism atrDNA loci can be maintained for long periods of timerelative to single-copy loci. Another possible consequenceof combining diverse rDNA copies during hybridization issilencing of some loci. Failure of one of the parents’chromosomes to organize nucleoli in an interspecifichybrid may lead to nucleolar dominance—the selectivesilencing of the corresponding rRNA genes (Honjo andReeder 1973; Durica and Krider 1977) by mechanisms thatinvolve chromatin modification (Chen, Comai, and Pikaard

1998; Frieman et al. 1999; Pikaard 2000a, 2000b; Muir,Fleming, and Schlotterer 2001). In these cases, silenced

rDNA loci can then evolve independently as pseudo-genes (although concerted evolution is maintained withineach locus), potentially complicating phylogenetic anal-

yses. It is believed that this has happened in the hybridiz-ing oak species Quercus petrea and Q. robur (Muir,Fleming, and Schlotterer 2001). These two species hybrid-

ize frequently and share three divergent rDNA types, twoof which are believed to be pseudogenes whose origin

predates the species divergence (Muir, Fleming, andSchlotterer 2001).

Extremely high levels of variability in the ITS1-5.8S-

ITS2 region have been observed within and among severalspecies of the hard coral Acropora (Odorico and Miller1997; vanOppen et al. 2002). InAcropora, up to nine rDNAtypes can be present in a single colony (Odorico and Miller1997; van Oppen et al. 2002), and distinct ITS types areoften shared between species. These patterns are broadly

consistent with the occurrence of interspecific hybridiza-tion, although alternative explanations such as incomplete

lineage sorting cannot yet be completely ruled out (vanOppen et al. 2000, 2001, 2002).

To examine the possibility that some rDNA copies in

Acropora are nonfunctional and constitute pseudogenes,three kinds of analyses were conducted. First, we used RT-PCR to determine which 5.8S rDNA types are expressed in

A. millepora, the species containing the most diverse suiteof rDNA types. Second, we examined patterns of methyl-ation in the 5.8S rDNA that may indicate silencing caused

by nucleolar dominance. Third, we examined the patterns ofsubstitution at conserved sites in the 5.8S rDNA, looking formutations that may disrupt the secondary structure and

functionality of the rRNA. These studies indicate thatpseudogenes are likely to constitute at least one of the major

sequence types present in a broad range of Indo-PacificAcropora species.

1 Present address: Centro de Ecologıa, Instituto Venezolano deInvestigaciones Cientıficas, Caracas, Venezuela.

Key words: ribosomal DNA, pseudogenes, ITS, reticulate evolu-tion, coral, Acropora.

E-mail: m.vanoppen@aims.gov.au.

Mol. Biol. Evol. 20(7):1077–1086. 2003DOI: 10.1093/molbev/msg122� 2003 by the Society for Molecular Biology and Evolution. ISSN: 0737-4038

1077

by guest on June 9, 2013http://m

be.oxfordjournals.org/D

ownloaded from

Materials and MethodsSample Collection, PCR, Cloning, and Sequencing

In addition to 171 published sequences (Odorico andMiller 1997; van Oppen et al. 2000, 2002), 243 new se-quences were obtained from 73 Acropora colonies (table 1)(GenBank accession numbers AF538378 toAF538600 andAF540600 to AF540620). Tissue samples were collected bysnapping off small branches (2 to 5 cm) from individualcolonies and stored in 70% to 90% EtOH. DNA extraction,PCR, cloning, and sequencing procedures followed vanOppen, Willis, and Miller (1999) and van Oppen et al.(2002).

Examination of Expressed rDNA Types

Total RNA was extracted from three A. milleporaindividuals collected at Magnetic Island in the centralGreat Barrier Reef using the Choczynski and Sacchi(1987) method. The RNA was DNase-treated several timesuntil no DNA contamination could be detected using PCRamplification of scnDNA Acropora loci. RT PCR wasperformed on 1.3 lg of RNA using the Amersham FirstStrand cDNA kit and the primer 5.8S_RP1 (59-GACRS-

CCTYRCTCGCCTCG-39). One ll of undiluted cDNA/RNA hybrid was used in the subsequent PCR reactions toamplify the partial 5.8S region (;120 bp in length) usingthe primers 5.8S_FP1 (59-GGCTYGSGYATCGATGAA-GA-39) and 5.8S_RP1. PCR products were run on a 1.5%TAE-agarose gel stained with EtBr and excised from thegel. Excised fragments were purified by spinning the gelslice through Whatman 1 filter paper. The spun-throughfragments were cloned into pGEM-T (Promega) followingthe manufacturer’s protocol. Ten, nine, and five clones ofeach of three individuals were sequenced as before.

Alignment

Sequences were aligned manually using Sequencher3.0 (Gene Codes Corporation). The gene/spacer bound-aries were identified using Odorico and Miller’s (1997)alignment of Acropora rDNA sequences. The 5.8S regionwas easily alignable, but even following Odorico andMiller’s (1997) alignment, it was impossible to align theITS1 region objectively, and this region was therefore notused in subsequent analyses. However, the ITS2 regionwas relatively easy to align in a subset of 185 sequences

Table 1Samples and rDNA Sequence Codes

Species Group Species Collection Location Sequence Codes

hyacinthus Acropora hyacinthus Coral Bay, WA 1.1, 1.2, 1.3, 1.4, 1.5, 2.1, 2.2, 2.3, 2.4, 8.2Heron Is., GBR 14.1, 14.2, 14.3, 14.4, 14.7Lizard Is., GBR 16.3, 16.4, 16.5, 16.9, 18.5Britomart, GBR 23.2, 23.3, 23.4, 23.5, 24.1, 24.2, 24.3, 24.4, 28.4Dungeness, TS 28.1, 28.2, 30, 30.2, 31.1, 31.2, 31.5, 32, 32.1, 32.2, 32.4, 32.5

A. cytherea Dampier, WA 1.2, 1.3, 1.5, 2.2, 2.3, 2.5Bundegi, WA 11.2, 11.3Heron Is., GBR 13, 13.3, 13.5Lizard Is., GBR 6.2, 6.3Britomart, GBR 25.1, 25.3, 25.5, 26.1, 26.2, 26.3Trunk, GBR 34.2Dungeness, TS 20.1, 20.2, 20.3, 21.1, 21.2, 22.2

selago A. tenuis Dungeness, TS 18, 18.1, 18.3, 18.4, 18.5, 19.2Cumberland, TS 71.1, 71.2, 71.5, 72.2, 74.2Trunk, GBR 1.1, 1.3, 2.1, 2.4, 2.5, 3.1Coral Bay, WA 3.4

aspera A. spicifera Coral Bay, WA 9.1, 80.1Bundegi, WA 12.1, 12.4, 75.1, 75.2, 75.4, 76.1, 76.2, 76.3, 76.4,

78.1, 78.4, 79.1, 79.2, 79.3, 79.4, 79.5A. aspera Coral Bay, WA 84.1, 84.2, 84.3, 84.4, 84.5, 85.1, 85.2, 85.3, 85.4,

85.5, 86.1, 86.2, 86.3, 86.4, 87.2, 87.3, 87.4, 87.5A. millepora Coral Bay, WA 71.1, 71.2, 71.3, 71.4, 71.5, 72.1, 72.2, 72.3, 72.4,

72.5, 73.1, 73.2, 73.3, 73.4, 73.5, 74.1, 74.2, 74.3,74.4, 74.5, 75.1, 75.2, 75.3, 75.4

Bundegi, WA 75.1, 75.2, 75.3, 75.4, 75.5, 75.6, 75.7, 75.8, 75.9,75.10, 76.1, 76.2, 76.3, 76.4, 76.5, 77.1, 77.2, 77.3,77.4, 78.1, 78.2, 78.3, 78.4, 78.5

A. pulchra Bundegi, WA 79.1, 79.2, 79.3, 79.4, 79.5, 80.1, 80.2, 80.3, 80.4,80.5, 81.1, 81.3, 81.4, 81.5, 82.1, 82.2, 82.3, 82.4,82.5, 83.1, 83.2, 83.3, 83.4, 83.5, 88.1, 88.2, 88.3,88.5, 89.1, 89.3, 89.5, 90.1, 90.2, 90.3, 90.4, 90.5,91.1, 91.2, 91.3, 91.4

valida A. cerealis Trunk, GBR 113.2, 113.3, 114.1, 114.2, 114.3, 114.4, 117.3, 117.5A. longicyathus Trunk, GBR 84.1, 84.2, 84.3, 84.4, 85.2, 85.3, 85.4, 85.5, 86.1, 86.2A. gemmifera Trunk, GBR 93.2, 93.3, 93.4, 93.5, 94.3, 94.5, 96.2, 96.5A. cuneata Britomart, GBR 48.1, 48.3, 48.4, 48.5, 50.1, 50.2, 50.4, 50.6

NOTE.—The number after the dot for each sequence indicates the clone sequenced. GBR, Great Barrier Reef; TS, Torres Strait; and WA, Western Australia.

1078 Marquez et al.

by guest on June 9, 2013http://m

be.oxfordjournals.org/D

ownloaded from

corresponding to the largest clade in the 5.8S phylogeny(clade IVB, see below). Phylogenetic analysis of ITS2 wasonly performed on this subset of sequences.

Phylogenetic Analyses

Due to the difficulty in aligning the ITS regions (asdiscussed above), only the 5.8S gene was used for thephylogenetic analyses across the entire data set. The dataset was reduced from 414 to 160 sequences, becauseidentical sequences (table 2) were excluded from theanalysis. A maximum-likelihood phylogeny was con-structed using the HKY85 (Hasegawa, Kishino, and Yano1985) substitution model in MOLPHY 2.3 (Adachi andHasegawa 1996), which estimates approximate bootstrapprobabilities (Felsenstein 1985) among trees by the RELLmethod (Kishino, Miyata, and Hasegawa 1990; Hasegawaand Kishino 1994). The tree was rooted using Acropora(Isopora) cuneata as an outgroup. The subgenus A. Isopora

forms a sister group to A. Acropora in both cladisticanalyses using morphology (Wallace 1999) and sequenceanalysis of cytochrome b and NADH dehydrogenasesubunit 2 (van Oppen, Willis, and Miller 1999; Fukami,Omori, and Hatta 2000). Analysis of molecular variance(AMOVA) (Excoffier, Smouse, and Quattro 1992) wasused to partition the genetic variance among species andamong phylogenetic clades using the Kimura two-param-eter distance (1980) in ARLEQUIN 2.0 (Schneider, Roessli,and Excoffier 2000). A Permutation Tail Probability (PTP)test was performed to examine whether a strong phyloge-netic signal is present in the 5.8S data set by generating100,000 random trees and comparing the lengths of thesewith a consensus tree obtained using parsimony analysis inPAUP* 4.0 beta (Swofford 1999). The phylogenetic signalwas also assessed by evaluating the skewness in thedistribution of the length of another 100,000 randomlygenerated trees (i.e., a strong phylogenetic signal is charac-terized by a left skew in the tree length distribution, caused

Table 2Identical 5.8S Sequences in the Original Data Set of 416 Acropora Sequences

Sequence in Tree Identical Sequences

aspera 98.4 aspera 53.4, 84.1, 84.2, 84.4, 84.5, 85.1, 85.2, 85.3, 85.4, 85.5, 86.2, 86.3,86.4, 87.3, 87.4, 87.5, 92.2, 92.4, 93.3, 93.5, 94.1, 94.2, 94.5, 96.3, 96.5,98.2, 98.3, 98.5, 99.3, 100.2, 101.3, 103.3, 103.5, 104.7, 104.8, florida 103.2,pulchra 30.1, 83.1, 83.3, 83.4, 83.5

aspera 87.2 aspera 53.7, 53.8aspera 101.4 aspera 92.1, 92.3aspera 93.1 aspera 93.4aspera 96.2 aspera 96.4pulchra 82.5 pulchra 13.2, 22.2, 26.1, 80.1, 80.2, 80.4, 81.1, 81.5, 82.1, 82.2, 88.1, 88.2,

88.3, 89.1, 89.3, 89.5, 91.1, 91.3, spathulata 68.2, 68.5, 68.9, papillare 48.1,48.2, 49.1, 49.2, 49.4, 60.1, 60.5

pulchra 80.5 pulchra 88.4pulchra 22.1 pulchra 13.1, 81.3, 88.5, 89.4, 91.2, 91.4,pulchra 26.2 millepora 90.3spathulata 38.1 spathulata 38.10spathulata 48.6 aspera 97.5millepora 71.2 millepora 75.1, 75.2, 75.3, 75.5, 79.4millepora 73.2 millepora 73.1, 73.3millepora 77.3 millepora 72.1, 72.2millepora 19.2 millepora 21.2hyacinthus DOAht.1 millepora 2, 9.2, 9.3, 9.4, 11, 12.1, 12.8, 17.2, 19, 19.3, 20.1, 21, 21.1, 21.6,

22, 24, 25.1, 25.2, 26.6, 29.2, 29.3, 29.5, 34.1, 50.4, 55.5, 56.1, 56.2, 71.1,72.3, 72.4, 72.5, 74.1, 74.3, 74.4, 75.6, 75.7, 75.8, 75.9, 75.10, 76.1, 76.2,76.4, 76.5, 77.1, 77.2, 77.4, 78.1, 78.4, 78.5, 90.5, cytherea 1, DO3, DO4,DO8, 6.2, 6.3, 11.2, 11.3, 13, 13.3, 13.5, 18.2, 18.5, 20.2, 22.2, 25.3, 25.5,26.1, 26.2, 26.3, 34.2, hyacinthus DO2, DO5, DO9, 1.1, 1.3, 1.4, 1.5, 2.1,2.4, 8.2, 14.1, 14.2, 14.3, 14.7, 16.2, 16.4,16.9, 23.2, 23.3, 23.4, 23.5, 24.1,24.2, 24.3, 28.1, 28.2, 30, 30.2, 31.1, 31.2, 32, spicifera 9.1, 12.1, 12.4, 75.1,75.2, 75.4, 76.3, 76.4, 78.1, 78.3, 79.3, 80.1, spathulata 1.6, 8.3, 11.1, 11.4,11.5, 12.9, 16, 17.4, 24.1, 24.2, 24.3, 28.3, 28.8, 32.5, 32.8, 50.1, 51.8, 54.1,57.1, 57.2, 58.2, aspera 97.1, 97.3, papillare 55.1, 55.2, 62.2, 62.3, pulchra79.1, valida 17.3, 17.8

hyacinthus 1.2 WA hyacinthus 18.4cytherea DOAcy.2 cytherea 17.2, 17.4cytherea 2.2 WA millepora 90.1, spicifera 79.2tenuis sperm 3 tenuis sperm 5tenuis 4 cytherea WA 1.2, 1.5, tenuis sperm 1, sperm 2, eggs3, eggs 5, 2.1, 3.1, 18,

18.1, 71.1, 71.5, hyacinthus 31.5, 32.4tenuis 72.2 tenuis 1tenuis 74.2 cytherea 21.2, spicifera 76.2cuneata 48.3 cuneata 48.5, 50.2, 50.4, 50.6longicyathus 84.3 longicyathus 84.1longicyathus 85.3 longicyathus 85.5, 86.1gemmifera 93.2 gemmifera 93.3 longicyathus 85.4cerealis 113.2 cerealis 114.2, 114.4, 117.5

Extreme nrDNA Diversity in Acropora 1079

by guest on June 9, 2013http://m

be.oxfordjournals.org/D

ownloaded from

by several trees having a reduced length due to phylogeneticsignal in addition to those being short by chance).

Relative-Rate Test

To examine the evolutionary rate constancy amongclades in the 5.8S phylogeny, a two-cluster relative-ratetest (Takezaki, Rzhetsky, and Nei 1995) was implementedin Phyltest 2.0 (Kumar 1996) using the Kimura two-parameter distance. A representative of each clade washaphazardly chosen (from fig. 1 and supplement 1 in theonline Supplementary Material) and used in the analysis.The test was only performed for the 5.8S region.

Methylation-Related Substitutions

For representatives of each clade in molecular phy-logenies (and using A. cuneata as a reference for theancestral state), the number of cytosine sites showing atleast one deamination-like substitution and those that didnot contain such substitutions were compared with those ofthe most basal clade in the phylogeny (Muir, Fleming, andSchlotterer 2001) using a 23 2 contingency table. Giventhe small sample sizes, the asymptotic property of a v2

distribution cannot be assumed. Instead, we performed anexact test that uses the random permutation procedure ofRoff and Bentzen (1989). In this procedure, a contingencyv2 statistic is calculated and the probability of observing theexact test statistic or larger is generated using a randompermutation procedure that maintains the marginals butsimulates the null hypothesis of no association. Therandom permutation is implemented in Chiperm 1.2(Chiperm is available at the Crandall lab Web site athttp://bioag.byu.edu/zoology/crandall_labprograms.htm).

Results

The complete sequence data set consists of 414sequences of ITS1-5.8S-ITS2 from a range of Acroporaand one Isopora species. Many 5.8S sequences areidentical (table 2), whereas ITS sequences are generallyvery different and sometimes almost impossible to align.Nevertheless, 54 ITS1 sequences are not unique. Most ofthese are observed only twice, but four sequences areobserved 11, seven, six, and four times, respectively.

Another five ITS1 sequences are found three times eachwithin this data set. The maximum length of the sequences(table 3) is slightly larger than those reported previously(i.e., 166 and 200 for 5.8S and ITS2, respectively, ascompared with 112 and 158 reported by Odorico andMiller [1997]), due to the inclusion of the Isopora se-quences that contain inserts not present in the subgenusAcropora. The base composition of both regions (table 3)does not differ significantly among species, among sub-genera, or in comparison with those previously reported(i.e., Odorico and Miller 1997).

Levels of genetic diversity in the 5.8S gene within thesubgenus Acropora are extremely high, with a maximumcorrected distance between A. cerealis and A. gemmiferaof 20.3% (table 3). The distance between subgenera isslightly higher, with a maximum of 22.8% between A.cuneata and A. gemmifera. The alignment of representa-tives from observed rDNA types is attached as supplement1 in the online Supplementary Material. ITS2 distanceestimates are very high (table 3), and the difficultyencountered in aligning these sequences suggests thatthey are saturated, although estimates of transition/trans-version ratios do not indicate saturation (0.95 and 1.01 forITS1 and ITS2, respectively).

Phylogenetic Analysis

The 5.8S alignment of 160 sequences used in thephylogenetic analysis consists of 166 positions of which58 are constant, 33 are variable but parsimony un-informative, and 75 are parsimony informative. The phy-logenetic signal in this alignment is high, as indicatedby permutation tests (length of most parsimonious tree is227 steps, shortest length of randomly generated tree is606, P ¼ 0.01), respectively, and a significantly left-skewed tree length distribution, which was tested bygenerating 100,000 random trees (g1¼�0.419) (Hillis andHuelsenbeck 1992).

Four major clades are resolved in the 5.8S phylog-eny (fig. 1), corresponding to clearly distinct ITS2 types(see supplement 1 in the online Supplementary Material).The base of the tree forms a trichotomy leading to thefollowing clusters: clade I, clades II and III, and the majorclade IV. Clade I consists only of two A. millepora and eightA. cerealis sequences (fig. 1 and table 2); note that theonly other A. cerealis sequence obtained falls into cladeIVC, whereas A. millepora sequences are scattered through-out the tree. As in the case of scnDNA and mtDNA data(van Oppen et al. 2001), sequences from the Caribbeanspecies constitute a clade (II) that is well separatedfrom all Indo-Pacific species. The composition of cladeIII is also broadly consistent with mtDNA and scnDNAtrees, except that a few A. spicifera, A. cytherea, and A.hyacinthus sequences cluster in clade III based on 5.8Ssequences, whereas the corresponding clades in scnDNAand mtDNA trees contain no sequences from thesespecies (Marquez et al. 2002b). The major derivedclade IV contains sequences from a wide range of speciesand shows clear substructuring. Most A. aspera and all A.florida sequences fall in clade IVB, which is again con-sistent with previous analyses based on scnDNA and

Table 3Length, Mean Base Compositions (%), and Ranges ofKimura Two-Parameter Pairwise Sequence Distances (%)for the 5.8S and the ITS Regions of Acropora

Region Length A C G TDistanceRange

ITS1 70–94 bp 31.58 18.76 22.53 27.13 n.d.(4.85) (4.62) (1.98) (3.16)

5.8S 166 bp 20.73 28.83 28.32 22.12 0.00–20.33(0.80) (1.26) (0.94) (1.31) (22.77)

ITS2 200 bp 22.87 23.85 28.00 25.28 0.00–46.64(2.73) (2.70) (2.16) (2.44) (59.48)

NOTE.—For the base composition, standard deviations are given in

parentheses. Maximum distance between Isopora and Acropora are also given

between parentheses. n.d. ¼ not determined.

1080 Marquez et al.

by guest on June 9, 2013http://m

be.oxfordjournals.org/D

ownloaded from

mtDNA (van Oppen et al. 2001; Marquez et al. 2002). Insome cases, clones from single individuals occur incompletely different clades (e.g., A. spicifera 79 subcladeIIIA and IVB). The AMOVA results are concordant withthis in that the variance in the data is higher betweenclades than between species but at the same time is veryhigh within species (table 4).

A phylogenetic analysis (not shown) of the largestsubclade IVB was also performed based only on the ITS2

region, which was unambiguously alignable, but found noevidence for clustering based on either species identity orgeographic localities.

Expressed 5.8S Sequences

To examine whether all or only a subset of rDNAtypes are expressed in Acropora, RT PCR was used todetermine 5.8S rRNA sequences from three colonies of A.

FIG. 1.—Rooted maximum-likelihood tree of the 5.8S in Acropora. Bootstrap values are above the branches (1,000 replicates). Major clades andsubclades are indicated. Sample codes are as in table 1 in this manuscript and in Odorico and Miller (1997), van Oppen et al. (2000), and van Oppenet al. (2002). A dot and a number after the sample code indicate the clone that was sequenced. Sample names for the A. hyacinthus and A. aspera groupsindicate geographic origin: GBR, Central Great Barrier Reef; TS, Torres Strait; and WA, Western Australia.

Extreme nrDNA Diversity in Acropora 1081

by guest on June 9, 2013http://m

be.oxfordjournals.org/D

ownloaded from

millepora, the species that contains the most diverse suiteof rDNA types. Twenty-two of the 24 cDNA sequencesobtained were identical; the remaining two sequences hadsingle substitutions at different positions, which are likelyto represent PCR errors. The maximum number ofsubstitutions observed in this 81-bp region of the 5.8Sgene (;120 bp minus the primers), among all pairs ofsequences in the A. aspera group (of which A. millepora isa member), is five. These expressed sequences areidentical to the corresponding 81 bp of the sequences inclades IVA and IVB and some sequences in IVE, theydiffer by one substitution from the sequences in clade I andsome sequences in clade IVE, by two to four substitutionsfrom sequences in clade IVD, and up to five substitutionsfrom those in clade IVC. Hence, it appears that in A.millepora, only rDNA genes falling into clades IVA and B(and possibly IVE) are likely to be functional.

Relative-Rate Test

Rate constancy may differ between different groupsof species, as they are not necessarily subjected to thesame evolutionary constraints, whereas substitution ratesare expected not to differ significantly within species. Withrate constancy rejected at the 5% level (table 5), therelative-rate test indicates that some clades and subcladesin the 5.8S phylogeny have different rates of evolution.Clade I evolves more slowly than clades II, III, and IV, andclades II and III have a faster substitution rate comparedwith clade IV (table 5). Moreover, subclade IVB evolvessignificantly slower than clades IVC and IVD, althoughthe latter two clades do not differ from one another. Notethat sequences of some species (e.g., A. millepora and A.pulchra) are spread across subclades that exhibit differentsubstitution rates (table 5).

Methylation Mutation Analysis

Gene silencing in animals frequently involves DNAmethylation at cytosine residues in CpG motifs. Becausethe resulting methylcytosine residues are prone tospontaneous deamination (resulting in mutation to thy-mine), silenced chromatin typically drifts towards higher(A þ T) content relative to active regions. Methylation hasbeen implicated in nucleolar dominance phenomena, andthe resulting rDNA pseudogenes therefore typically showpatterns of deamination-like substitutions (CfiT and

GfiA) at methylation sensitive sites relative to their activecounterparts. Therefore, following Muir et al. (2001) andusing A. cuneata as a reference for the ancestral state, wecounted the number of cytosine sites (only CpG, becauseanimals are not known to methylate CpNpG) showing atleast one deamination-like substitution and those thatdid not contain any of such substitutions for each clade.Table 6 shows that the number of deamination-like substi-tutions at methylation sites in the 5.8S region (in compari-son to the number of mutations at nonmethylation sites)is not significantly higher in any clade or subclade com-pared with clade I, which is considered the ancestral state.However, the number of deamination-like substitutions atmethylation sites per sequence in subclade IVC is surpris-ingly high (20 in 14 sequences) and significantly largerthan in subclade IVB (10 in 57 sequences), in relation tothe number of substitutions in nonmethylation sites (4 and7 respectively) (v2 ¼ 29.7, P ¼ 0.001).

Analysis of Conserved Sites and Secondary Structure

Pseudogenes are expected to mutate freely because,unlike their active counterparts, they are not under selectiveconstraints. Therefore, mutations are expected to occureven at otherwise conserved positions. At active rDNAloci, some of these constraints result from the secondarystructure of the resulting rRNA; mutation at sites criticalfor secondary structure is either selected against or accom-panied by compensatory mutations at complementary sites.Conversely, in rDNA pseudogenes, selection is notimposed against mutations affecting the secondary struc-ture of the rRNA, and noncompensatory mutations aretolerated.

Based on an alignment of 30 5.8S rDNA sequencesfrom a broad range of eukaryotes including fungi, plants,invertebrates, and vertebrates, a total of 45 conserved siteswere identified (the alignment is attached as supplement 2in the online Supplementary Material). Patterns of muta-tion at these conserved sites were examined in sequencesrepresentative of each clade and subclade in the Acroporaphylogeny. Although differences in at least one of thosesites were present in each of the Acropora (sub)clades, thenumber of such differences was very small in all cases(table 7). In terms of substitutions at these conserved sites,

Table 4Results of AMOVA Between Acropora Species andAcropora Clades in the Maximum-Likelihood Phylogeny

Source ofVariation df

Sum ofSquares

VarianceComponents

Percentageof Variation P

Species

Among 14 696.127 4.18531 Va 56.11 ,0.00001Within 161 527.084 3.27381 Vb 43.89Total 175 1223.210 7.45913

Clades

Among 11 626.129 4.07751 Va 74.13 ,0.00001Within 165 234.816 1.42313 Vb 25.87Total 176 860.944 5.50063

Table 5Relative-Rate Test Results for Acropora 5.8S Clades inMaximum-Likelihood Phylogeny

OutgroupCluster

ACluster

B bA bB d z-score

Isopora II þ III þ IV I 0.0478 0.0026 0.0451 4.295*I IV II þ III 0.0296 0.0777 0.0480 2.582*I III II 0.0818 0.0488 0.0330 1.135II þ III IVB–IVE IVA 0.0263 0.0042 0.0221 1.937IVA IVB þ IVC

þ IVDIVE 0.0362 0.0007 0.0355 2.787*

IVE IVC þ IVD IVB 0.0384 0.0067 0.0318 2.402*IVB IVC IVD 0.0201 0.0338 �0.0137 0.694

NOTE.—bA and bB are average number of substitutions per site; d ¼ bA � bB.

z-scores are the values from the standardized normal distribution (i.e., 1.96 is the

value for a ¼ 0.05). *P , 0.05.

1082 Marquez et al.

by guest on June 9, 2013http://m

be.oxfordjournals.org/D

ownloaded from

the highest number (five) was observed in the A. cerealis114.1 sequence, a member of clade IVC.

A broad range of metazoan 5.8S sequences, contain-ing annotations of the secondary structure derived bycomparative sequence analysis, was obtained from theEuropean Large Subunit Ribosomal RNA Database(Wuyts et al. 2001) (http://rrna.uia.ac.be/lsu). Representa-tives from each clade and subclade in the Acroporaphylogeny were aligned with these data to identify non-compensatory mutations that may disrupt the formation ofhelices. This approach allowed the identification of threesequences in clade IVC containing noncompensatorymutations that potentially disrupt helices B6 and B9(fig. 2). In these cases, maintenance of secondary structurewould require the pairing of two guanine residues, whichis energetically unfavorable.

DiscussionCharacteristics of the rDNA Region in Acropora

Our results confirm the notion that the rDNA ITSregions of Acropora, including the subgenus Isopora, areamong the shortest in corals (Takabayashi et al. 1998) andeukaryotes in general (Odorico and Miller 1997). To ourknowledge, only the green alga Halimeda has shorter ITSregions (42 to 77 bp and 149 to 161 bp for ITS1 and ITS2,respectively [Kooistra, Coppejans, and Payri 2002]). Thelevels of variability in the 5.8S and ITS2 regions of Acro-pora are the highest reported for any coral, at the level ofboth the species and the individual. The complete ITS1-5.8S-ITS2 region exhibits only 2% variation amongmorphospecies in the Montastrea annularis complex(Lopez and Knowlton 1997; Medina, Weil, and Szmant1999) and only up to 4.9 % between species of the genusMadracis (Diekmann et al. 2001). Takabayashi et al.(1998) reported only ITS1 variation, which was 15% inGoniopora tenuidens, 2% in Heliofungia actiniformis, 11%in A. longicyathus, and 31% in Stylophora pistillata. InPlesiastrea versipora, a species with an extreme latitudinaldistribution range (408 N to 408 S) and accordingly ex-pected to be genetically variable, populations across thewhole range showed sequence variation of just 1.53% 60.43 and 3.15% 6 2.24 in ITS1 and ITS2, respectively(Rodriguez-Lanetty and Hoegh-Guldberg 2001). Maxi-mum pairwise distances in the ITS1-5.8S-ITS2 regionbetween Porites species are approximately 11% (Hunter,Morden, and Smith 1997). Despite the 5.8S subunit being

a coding region, we estimated corrected genetic diver-gences of approximately 20% within the subgenus Acro-pora, which is five times higher than the estimate ofOdorico and Miller (1997) using a smaller sample size.These high levels of variability prompted us to questionwhether some of these 5.8S genes may have lost theirfunction and are evolving as pseudogenes (see below).

Phylogeny

The topologies of all Acropora phylogenies obtainedso far using scnDNA loci, mtDNA, and rDNA haveseveral features in common, including the derived anddistinct A. aspera clade with A. florida as a sister taxon,and the basal A. tenuis–A. longicyathus clade (van Oppenet al. 2001; Marquez et al. 2002b). The 5.8S phylogenyresembles that of the mtDNA in that the Caribbean taxa areamong the basal groups. However, in contrast to theseother cases, the base of the 5.8S tree is unresolved andforms a trichotomy. Moreover, unlike phylogenies basedon the Pax-C intron, A. aspera and A. tenuis are not strictlymonophyletic in the 5.8S tree, with a small number ofsequences from each species occurring in clades IVA andIVB. Several other species that are nonmonophyletic basedon the 5.8S gene show the same pattern in the nuclearintron and mtDNA phylogenies. These patterns of speciespolyphyly and paraphyly may either result from recentspeciation coupled with incomplete lineage sorting or fromintrogressive hybridization. Low rDNA diversity both inthe three Caribbean Acropora species (van Oppen et al.2000) and in the genetically distinct brooding speciesAcropora (Isopora) cuneata (Ayre, Veron, and Dufty1991) (fig. 1) relative to most Indo-Pacific members of thisgenus, together with the fact that genetically distinctspecies tend to differ in the timing of release of theirgametes, supports the hybridization hypothesis and isdiscussed in detail elsewhere (van Oppen et al. 2000,2001, 2002; Marquez et al. 2002a, 2002b).

Although incongruences exist between the 5.8Sphylogeny and the nuclear intron and mtDNA phylogenies(van Oppen et al. 2001), the 5.8S gene provides a high

Table 6Deamination-Like Susbtitutions in 68 Methylation and 21Nonmethylation Sites of the 5.8S in Acropora

CladeMethylation

SitesOtherSites

v2 Values Obtained inComparison with Clade I P

I 3 0II 3 0 NA NAIII 5 1 0.563 1.000IVA 2 0 NA NAIVB 9 5 1.518 0.512IVC 10 4 0.852 0.569IVD 3 1 0.875 1.000IVE 2 1 1.200 1.000

Table 7Number of Mutations in 45 Conserved Positionsof the 5.8S

Clades or Introgressed Sequences Number of Mutations

Outgroup (cuneata 48.3) 4Clade I (cerealis 113.5 millepora 38.1) 1Clade II (prolifera 22.4) 3Clade III (longicyathus84.4 and tenuis19.2) 2Clade IVA (gemmifera 93.5 gemmifera 96.2) 2Clade IVB (cytherea 20.1) 2Clade IVB (hyacinthus 28.4) 3Clade IVC (spicifera 78.4) 3Clade IVD (aspera 93.3) 3Clade IVE (pulchra 82.5) 1cerealis 114.1 5pulchra 83.2 3longicyathus 85.4 1hyacinthus 32.5 1spicifera 79.5 1

NOTE.—Representative sequences used are shown in parentheses.

Extreme nrDNA Diversity in Acropora 1083

by guest on June 9, 2013http://m

be.oxfordjournals.org/D

ownloaded from

phylogenetic signal. Nevertheless, true phylogenetic re-lationships between specimens can be confounded by thepresence of pseudogenes, and care needs to be taken ininterpreting relationships from ribosomal DNA alone.

Identification of Possible Pseudogenes

Substitution rates vary between some clades thatdiffer in species composition, possibly as a consequence ofheterogeneity in the predominant evolutionary forcesoperating on different species. Subclade IVC is composedmainly of sequences from colonies (including A. mille-pora) that are also represented in other clades and showsan accelerated evolutionary rate compared with the cladeto which the expressed A. millepora rDNA sequencesbelong (subclade IVB). Although the total numbers ofsubstitutions per sequence at conserved sites are notatypically high, sequences in subclade IVC typically showlarge numbers of substitutions at potential methylationsites relative to other clades. Furthermore, A. cerealis114.1, a member of subclade IVC, does show the highestnumber of mutations and the 5.8S gene sequences fromclade IVC are characterized by noncompensatory muta-tions that can potentially disrupt the secondary structure ofthe large subunit of the rRNA. These results support thenotion that at least some rDNA types (those of clade IVC)in Acropora are likely to be nonfunctional.

Neither the Caribbean Acropora nor the A. floridaspecies is represented in subclade IVB, which contains theexpressed A. millepora sequences. Similarly, very few ofthe A. aspera and A. tenuis sequences fall into thissubclade. This suggests that rDNA types that have beensilenced in some species may not necessarily have beensilenced in others. Independent hybridization events couldlead to different patterns of nucleolar dominance, possiblyexplaining this observation. Species such as A. tenuis andA. florida may have been reproductively isolated fromtheir congeners for sufficient time to allow homogeniza-tion of rDNA units to approach completion, and in the caseof A. tenuis, the small number of sequences outside cladeII may reflect incomplete lineage sorting. Whether this islikely also in the case of A. aspera is unclear, as A. asperaappears to have a semipermeable species boundary with

other members of its species group (van Oppen et al.2002).

Origin of Extreme rDNA Diversity in Acropora

How did the unprecedented rDNA diversity withinindividual Acropora colonies and species originate?Interspecific hybridization may have combined divergentrDNA copies within single genomes. As the mismatchrepair machinery seems sensitive to high numbers ofmismatches, large sequence divergences in the spacerregions may have suppressed recombination across theentire rDNA array, impeding concerted evolution (Petitet al. 1991; Muir, Fleming, and Schlotterer 2001) andcausing the persistence of divergent rDNA types withinindividuals. In contrast, when parental species have verysimilar sequence composition, such as Armeria villosa andArmeria colorata (Plumbaginaceae), where the ITS2region differs in only six out of 245 sites, homogenizationcan occur in only two generations (Fuertes Aguilar,Rosello, and Nieto Feliner 1999). Asexual reproductionmay also limit the effectiveness of concerted evolution tohomogenize divergent rDNA copies because in asexualtaxa, gene conversion and crossing over are restricted tomitotic divisions (Pringle, Moncalvo, and Vilgalys 2000).This may be the case in A. millepora, which has very highrates of asexual reproduction (Ayre and Hughes 2000)and exhibits the most extreme pattern of rDNA diversity ofall species examined (fig. 2). Some of the rDNAtypes brought together by hybridization may have beensilenced by nucleolar dominance, causing them to evolveas pseudogenes and sequence diversity to further increase.Alternatively, hybridization may have caused chromosom-al rearrangements (Rieseberg, van Fossen, and Desrochers1995), which relocated rDNA copies to different chromo-somal positions, also reducing the homogenizing action ofconcerted evolution (Arnheim et al. 1980; Muir, Fleming,and Schlotterer 2001). In situ hybridization is required todetermine whether the divergent sequences in Acroporaare present in more than one nucleolus organizer region(NOR).

In conclusion, it is likely that the clade IVC rDNAtype has been silenced by nucleolar dominance after

FIG. 2.—Folding of 5.8S rRNA sequences from A. hyacinthus DOAht.1 (from clade IVB in fig. 2) and A. cerealis 114.1 (from clade IVC in fig. 1)following Odorico and Miller (1997). Helices numbers after Wuyts (2001). The arrows indicate noncompensatory mutations.

1084 Marquez et al.

by guest on June 9, 2013http://m

be.oxfordjournals.org/D

ownloaded from

a hybridization event and has since evolved neutrally asa pseudogene. These nonfunctional rDNA types areexpected to show Mendelian inheritance and are subjectto concerted evolution. Hence, the presence of pseudo-genes does not affect inferences from these data aboutinterspecific hybridization. However, inclusion of func-tional and nonfunctional copies in the same analysis mayaffect the resulting phylogenetic hypotheses, as mutationpatterns are likely to differ between functional andnonfunctional copies. The data also imply that there maybe differences between species with respect to whichrDNA types are expressed and which are silent. Expres-sion studies across a range of species are required toaddress this issue.

Acknowledgments

We thank Danielle de Jong and Lesa Peplow forobtaining the expressed A. millepora sequences. We arealso thankful to Bette Willis and Jackie Wolstenholme fortheir help collecting and identifying samples. This projectwas supported by the Australian Research Council, theAustralian Institute of Marine Science, and James CookUniversity. L.M.M. acknowledges receipt of a scholarshipfrom the Venezuelan Council for Scientific Research.

Literature Cited

Adachi, J., and M. Hasegawa. 1996. MOLPHY version 2.3:program for molecular phylogenetics based on maximumlikelihood. Comput. Sci. Monogr. 28:1–150.

Arnheim, N., M. Krystal, R. Shmickel, G. Wilson, O. Ryder, andE. Zimmer. 1980. Molecular evidence for genetic exchangesamong ribosomal genes on nonhomologous chromosomes inman and apes. Proc. Natl. Acad. Sci. USA 77:7323–7327.

Ayre, D. J., and T. P. Hughes. 2000. Genotypic diversity andgene flow in brooding and spawning corals along the GreatBarrier Reef, Australia. Evolution 54:1590–1605.

Ayre, D. J., J. E. N. Veron, and S. L. Dufty. 1991. The coralsAcropora palifera and Acropora cuneata are genetically andecologically distinct. Coral Reefs 10:13–18.

Chen, Z. J., L. Comai, and C. S. Pikaard. 1998. Gene dosage andstochastic effects determine the severity and direction ofuniparental ribosomal RNA gene silencing (nucleolar domi-nance) in Arabidopsis allopolyploids. Proc. Natl. Acad. Sci.USA 95:14891–14896.

Choczynski, P., and N. Sacchi. 1987. Single-step method of RNAisolation by acid guanidinium thioxyanate-phenol-chloroformextraction. Anal. Biochem. 162:156–159.

Diekmann, O. E., R. P. M. Bak, W. T. Stam, and J. L. Olsen.2001. Molecular genetic evidence for probable reticulatespeciation in the coral genus Madracis from a Caribbeanfringing reef slope. Mar. Biol. 139:221–233.

Dover, G. 1982. Molecular drive: a cohesive mode of speciesevolution. Nature. 299:111–117.

Durica, D. S., and H. M. Krider. 1977. Studies on the ribosomalRNA cistrons in interspecific Drosophila hybrids. I. Nucleolardominance. Dev. Biol. 59:62–74.

Excoffier, L., P. E. Smouse, and J. M. Quattro. 1992. Analysis ofmolecular variance inferred from metric distances amongDNA haplotypes: application to human mitochondrial DNArestriction data. Genetics 131:479–491.

Felsenstein, J. 1985. Confidence limits on phylogenies: anapproach using the bootstrap. Evolution 39:783–791.

Frieman, M., Z. J. Chen, J. S. Vasquez, L. A. Shen, and C. S.Pikaard. 1999. RNA polymerase I transcription in a Brassicainterspecific hybrid and its progenitors: tests of transcriptionfactor involvement in nucleolar dominance. Genetics 152:451–460.

Fuertes Aguilar, J., J. A. Rosello, and G. Nieto Feliner. 1999.Nuclear ribosomal DNA (nrDNA) concerted evolution innatural and artificial hybrids of Armeria (Plumbaginaceae).Mol. Ecol. 8:1341–1346.

Fukami, H., M. Omori, and M. Hatta. 2000. Phylogeneticrelationships in the coral family Acroporidae, reassessed byinference from mitochondrial genes. Zool. Sci. 17:689–696.

Hasegawa, M., and H. Kishino. 1994. Accuracies of the simplemethods for estimating the bootstrap probability of a maxi-mum likelihood tree. Mol. Biol. Evol. 11:142–145.

Hasegawa, M., H. Kishino, and T. A. Yano. 1985. Dating of thehuman-ape splitting by a molecular clock of mitochondrialDNA. J. Mol. Evol. 22:160–174.

Hillis, D. M., and M. T. Dixon. 1991. Ribosomal DNA:molecular evolution and phylogenetic inference. Q. Rev.Biol. 66:411–453.

Hillis, D. M., and J. P. Huelsenbeck. 1992. Signal, noise, andreliability in molecular phylogenetic analyses. J. Hered. 83:189–192.

Honjo, T., and R. H. Reeder. 1973. Preferential transcription ofXenopus laevis ribosomal RNA in interspecific hybridsbetween Xenopus laevis and Xenopus mulleri. J. Mol. Biol.80:217–228.

Hunter, C. L., C. W. Morden, and C. M. Smith. 1997. The utilityof ITS sequences assessing relationships among zooxanthellaeand corals. Proc. 8th Int. Coral Reef Symp. 2:1599–1602.

Kishino, H., T. Miyata, and M. Hasegawa. 1990. Maximumlikelihood inference of protein phylogeny, and the origin ofchloroplasts. J. Mol. Evol. 31:151–160.

Kooistra, W. H. C. F., E. G. G. Coppejans, and C. Payri. 2002.Molecular systematics, historical ecology, and phylogeog-raphy of Halimeda. Mol. Phyl. Evol. 24:121–138.

Kumar, S. 1996. PHYLTEST: phylogenetic hypothesis testing.Version 2.0. Pennsylvania State University, University Park.

Lopez, J. V., and N. Knowlton. 1997. Description of Montastreacoral sibling species with multiple genetic loci. Proc. 8th Int.Coral Reef Symp. 2:1613–1618.

Marquez, L. M., M. J. H. van Oppen, B. L. Willis, and D. J.Miller. 2002a. Sympatric populations of the highly cross-fertile coral species Acropora hyacinthus and Acroporacytherea are genetically distinct. Proc. R. Soc. Lond. Ser. B.269:1289–1294.

Marquez, L. M., M. J. H. van Oppen, B. L. Willis, A. Reyes, andD. J. Miller. 2002b. The highly cross-fertile coral species,Acropora hyacinthus and Acropora cytherea, constitute sta-tistically distinguishable lineages. Mol. Ecol. 11:1339–1349.

Medina, M., E. Weil, and A. M. Szmant. 1999. Examinationof the Montastrea annularis species complex (Cnidaria:Scleractinia) using ITS and COI sequences. Mar. Biotech. 1:89–97.

Modrich, P, and R. Lahue. 1996. Mismatch repair in replicationfidelity, genetic recombination and cancer biology. Ann. Rev.Biochem. 65:101–133.

Muir, G, C. C. Fleming, and C. Schlotterer. 2001. Threedivergent rDNA clusters predate the species divergence inQuercus petrea (Matt.) Liebl. and Quercus robur L. Mol.Biol. Evol. 18:112–119.

Odorico, D. M., and D. J. Miller. 1997. Variation in theribosomal internal transcribed spacers and 5.8S rDNA amongfive species of Acropora (Cnidaria; Scleractinia): patterns ofvariation consistent with reticulate evolution. Mol. Biol. Evol.14:465–473.

Extreme nrDNA Diversity in Acropora 1085

by guest on June 9, 2013http://m

be.oxfordjournals.org/D

ownloaded from

Petit, M. A., J. Dimpfl, M. Radman, and H. Echols. 1991. Controlof large chromosomal duplications in Escherichia coli by themismatch repair system. Genetics 129:327–332.

Pikaard, C. S. 2000a. The epigenetics of nucleolar dominance.Trends Genet. 166:495–500.

———. 2000b. Nucleolar dominance: uniparental gene silencingon a multi-megabase scale in genetic hybrids. Plant Mol. Biol.43:163–177.

Pringle, A., J. M. Moncalvo, and R. Vilgalys. 2000. High levelsof variation in ribosomal DNA sequences within and amongspores of natural populations of the arbuscular mycorrhizalfungus Acaulospora colossica. Mycologia 92:259–268.

Rieseberg, L. H., C. van Fossen, and A. Desrochers. 1995.Hybrid speciation accompanied by genomic reorganization inwild sun-flowers. Nature 375:313–316.

Rodriguez-Lanetty, M., and O. Hoegh-Guldberg. 2002. Thephylogeography and connectivity of the latitudinally wide-spread scleractinian coral Plesiastrea versipora in the WesternPacific. Mol. Ecol. 11:1177–1189.

Roff, D. A., and P. Bentzen. 1989. The statistical analysis ofmitochondrial DNA polymorphisms: v2 and the problem ofsmall samples. Mol. Biol. Evol. 6:539–545.

Schneider, S., D. Roessli, and L. Excoffier. 2000. ARLEQUINver.2.000: a software for population genetics data analysis.University of Geneva, Geneva, Switzerland.

Swofford, D. L. 1999. PAUP*: phylogenetic analysis usingparsimony (*and other methods). Version 4.0. SinauerAssociates, Sunderland, Mass.

Takabayashi, M., D. Carted, S. Ward, and O. Hoegh-Guldberg.1998. Inter- and intra-specific variability in ribosomal DNAsequence in the internal transcribed spacer region of corals. Pp237–244 in J. G. Greenwod and N. J. Hall, eds. Proceedingsof the Australian Coral Reef Society 75th Anniversary

Conference, Heron Island October 1997. School of MarineScience, The University of Queensland, Brisbane, Australia.

Takezaki, N., A. Rzhetsky, and M. Nei. 1995. Phylogenetic testof the molecular clock and linearized trees. Mol. Biol. Evol.12:823–833.

van Oppen, M. J. H., B. J. McDonald, B. L. Willis, and D. J.Miller. 2001. The evolutionary history of the coral genusAcropora (Scleractinia, Cnidaria) based on a mitochondrialand a nuclear marker: reticulation, incomplete lineage sorting,or morphological convergence? Mol. Biol. Evol. 18:1315–1329.

van Oppen, M. J. H., B. L. Willis, and D. J. Miller. 1999.Atypically low rate of cytochrome b evolution in thescleractinian coral genus Acropora. Proc. R. Soc. Lond. BBiol. Sci. 266:179–183.

van Oppen, M. J. H., B. L. Willis, T. van Rheede, and D. J.Miller. 2002. Spawning times, reproductive compatibilitiesand genetic structuring in the Acropora aspera group:evidence for natural hybridization and semi-permeable speciesboundaries in corals. Mol. Ecol. 11:1363–1376.

van Oppen, M. J. H., B. L. Willis, H. W. J. A. van Vugt, and D. J.Miller. 2000. Examination of species boundaries in theAcropora cervicornis group (Scleractinia, Cnidaria) usingnuclear DNA sequence analyses. Mol. Ecol. 9:1363–1373.

Wallace, C. C. 1999. Staghorn corals of the world: a revision ofthe genus Acropora. CSIRO Publishing, Collingwood,Australia.

Wuyts, J., P. De Rijk, Y. Van de Peer, T. Winkelmans, and R. DeWachter. 2001. The European large subunit ribosomal RNAdatabase. Nucleic Acids. Res. 29:175–177.

Axel Mayer, Associate Editor

Accepted March 10, 2003

1086 Marquez et al.

by guest on June 9, 2013http://m

be.oxfordjournals.org/D

ownloaded from