Acropora rongoi Bridge & Cowman, 2024

Bridge, Tom C. L., Cowman, Peter F., Quaưrini, Andrea M., Bonito, Victor E., Frederic, Harii, Head, Catherine E. I., Hung, Julia Y., Halafihi, Tuikolongahau, Rongo, Teina & Baird, Andrew H., 2024, A tenuis relationship: traditional taxonomy obscures systematics and biogeography of the ‘ Acropora tenuis’ (ºcleractinia: Acroporidae) species complex, Zoological Journal of the Linnean Society 202 (1), pp. 1-24 : 16-20

Acropora rongoi Bridge & Cowman , n. sp.

Zoobank registration: urn:lsid:zoobank.org:act:B53CE95E-FC7F-4D64-8582-6EFEA44C305F

Holotype: MTQ G78411 * collected from 1 m depth in the lagoon at Rutaki * Rarotonga * Cook Islands.

Paratypes: MTQ G78418 * collected from 17 m depth on the outer reef slope at Papua * Rarotonga * Cook Islands ; MTQ G80232 collected from 12 m depth on the fore-reef slope at ‘ Ōpūnohu Bay * Mo’orea * French Polynesia ; MTQ G80233* collected from 10 m depth on the fore-reef at ‘ Ōpūnohu Bay * Mo’orea * French Polynesia .

Other material examined: MTQ: Cook Islands: G35728 Rarotonga* G35712* G55536 Aitutaki; Niue: G54667* G54670* G54671; ºociety Islands: G44034* G58650 Mo’orea.

Holotype: Irregular hispidose colony ~ 75 cm in diameter ( Fig. 6A ). Primary branches ≤ 300 mm long and 10–20 mm in diameter* secondary branches ≤ 100 mm long and 10–15 mm in diameter* and tertiary branches 5–15 mm long and 2–4 mm in diameter. Axial corallites tubular and exsert* outer diameter 1.8–2.2 mm * inner diameter 0.9–1.2 mm. Radial corallites closely positioned but not touching* outer diameter 1.2–1.8 mm * inner diameter 0.7–0.8 mm * haphazardly arranged and ranging in shape from tubular with oblique to dimidiate openings through to labellate with flaring lower lips ( Fig. 6B–D ). Both axial and radial corallites ossen show two prominent directive septa one-half to three-quarters of the radius of the calyx* but otherwise septa are weakly developed ( Fig. 6G * H). Ŋe coenosteum is costate ( Fig. 5H ).

Molecular phylogeny: Ŋe two specimens sequenced of this species from Rarotonga ( Cook Islands) are recovered as a reciprocally monophyletic group in all phylogenetic trees ( Figs 2 * 3; ºupporting Information* Figs º1* º2). Ŋe species is sister to a group containing A. aff. pagoensis and A. tenuissima in ML reconstruction* but appears as a direct sister species to A. aff. pagoensis in the ºNAPP species tree.

Remarks: Colonygrowthformisirregularcaespitosetohispidose* with indeterminate growth. Ŋe holotype * from the shallow lagoon of Rarotonga* has upright branches ( Fig. 6A )* whereas paratypes from higher-energy outer reef slopes show thicker* flaưer branches with fewer incipient axial corallites. Radial corallites are densely packed (touching in some specimens but not on the holotype)* of mixed sizes and shapes and ossen flaring* giving the colony a spikey* messy appearance. Colour varies from cream to brown. ºpecimens of A. rongoi in the collection at MTQ were previously identified as A. striata (Verrill* 1866)* Acropora elseyi (Brook* 1892) and Acropora florida (Dana* 1846)* aưributable to the variability in gross morphology of A. rongoi in different habitats. Verrill’s holotype of A. striata from the Ryukyu Islands has similar radial corallite shape but is clearly distinguished from A. rongoi on the basis of molecular and biogeographical evidence. Furthermore* the interpretation of A. striata as hispidose (ºhirai 1980* Veron and Wallace 1984* Wallace 1999) is likely to be incorrect because the holotype lacks tertiary branching (see below for further discussion on A. striata ). Acropora elseyi can have similar gross morphology* but radial corallites are tubular to appressed tubular with thick walls and prominent septa* and the coenosteum is composed of elaborate spinules. Colonies from outer reef slopes can also superficially resemble colonies identified as A. florida when tertiary branches become stunted and resemble the ramiculi of A. florida . However* the colony shape of A. florida (Dana* 1846) is arborescent* not hispidose. Acropora affinis (Brook* 1893)* synonymized with A. florida by Veron and Wallace (1984)* may be hispidose* but this species is in clade III ( Cowman et al. 2020; Fig. 2 ).

Distribution: Currently known from the Cook Islands * Niue and the ºociety Islands* French Polynesia in the ºouth Pacific (ºupporting Information* Fig. º6).

Etymology: Named for Dr Teina Rongo in recognition of his contribution to research and conservation of coral reefs and the marine environment in the Cook Islands.

Acropora tenuis Dana, 1846

Madrepora tenuis Dana 1846: 451 ; Ortmann 1888: 152.

Acropora tenuis (Dana) : Verrill 1902: 219.

Madrepora macrostoma Brook 1891 . Here removed from synonymy with Acropora tenuis (Dana) contra Veron and Wallace 1984: 279.

Madrepora kenti Brook 1892 . Here removed from synonymy with Acropora tenuis (Dana) contra Veron and Wallace 1984: 279.

Madrepora bifaria Brook 1892 . Here removed from synonymy with Acropora tenuis (Dana) contra Veron and Wallace 1984: 279.

Madrepora assicana Brook * 1893. Here removed from synonymy with Acropora tenuis (Dana) contra Veron and Wallace 1984: 279.

Specimens examined: NMNH-ºI: USNM 259 * Madrepora tenuis holotype * Fiji; MTQ: Fiji: G80284 * G80281 * G80280 * G78334 * G78323 *; G40941; Tonga: G80240 * G80237.

Remarks: Acropora tenuis is described in recent revisions as having a neat* regular arrangement of branches with few sub-branches and evenly sized cochleariform radial corallites neatly arranged in rows (Veron and Wallace 1984* Wallace 1999). However* these characters are not consistent with the holotype * which is described by Dana as being caespitose with branchlets that are ‘very slender* subterete and proliferous’* while the radial corallites are ‘appressed-tubiform* irregular* elongate and slender’ (ºupporting Information* Fig. º3).

Ŋe specimens in the present series vary in gross morphology* but FJ127 (G78323) and 101-5454 (G80284)* from the reef flat on the southern coast of Viti Levu* Fiji * closely resemble Dana’s holotype ( Fig. 4 )* confirming that these specimens are A. tenuis . Other specimens sequenced here illustrate the range of variation within this species. G80281* also from Fiji * and G80240 from Ha’apai * Tonga * have thicker branches than the holotype * while 101-5672 (G80280) is intermediate between these specimens. TG59 (G80237)* also from Tonga * differs considerably in gross morphology* growing as a thick boưlebrush rather than a corymbose clump or table. However* close examination of the branch ends shows that the branching structure and corallite shape are consistent with the thicker specimens in the present series. Ŋe large range of variation in gross morphology within A. tenuis could present a challenge when defining morphological characters for identification of this species in the field; however* it is clear that many of the species with overlapping morphological features previously considered within the range of variation of A. tenuis have distinct geographical ranges that do not overlap.

Acropora tenuis is currently known only from Fiji and Tonga in the ºouth Pacific (ºupporting Information* Fig. º6); all other tenuis -like specimens from other regions are likely to represent distinct species that require additional taxonomic investigation. Consequently* given the support of numerous other population genetic studies demonstrating strong geographic structure within specimens identified as A. tenuis based on gross morphology (e.g. Rosser et al. 2020 * van der Ven et al. 2022)* combined with morphological differences between the type specimens (ºupporting Information* Fig. º3) discussed below* A. macrostoma Brook * 1891 stat. rev. from Mauritius * A. kenti Brook * 1892 stat. rev. from Torres ºtrait* A. bifaria Brook * 1892 stat. rev. from Java* and A. assicana Brook * 1892 stat. rev. from ºouth Africa (listed as a synonym of A. tenuis by Veron and Wallace 1984 and considered a probably junior synonym of A. tenuis by Wallace [1999]) are hereby removed from synonymy with A. tenuis Dana * 1846. Additional taxonomic research* particularly the collection of topotypes * is required to establish the geographical and morphological ranges of these species and to identify additional species that might co-occur in these regions.

Of these four nominal species* A. macrostoma and A. bifaria resemble A.kenti inmorphology*withneatlyarrangedcochleariform radial corallites and minimal tertiary branching (ºupporting Information* Fig. º3). Ŋree specimens in the Museum of Tropical Queensland (G51822* G51823 and G51824) from Baie Aux Tortues* Mauritius * closely resemble Brook’s type of A. macrostoma owing to their comparatively thick branches* lack of incipient axial corallites and flaring* cochleariform corallites and are therefore re-identified as this species. Ŋe species identified as A. tenuis by Pillay et al. (2002: 78–79) is also likely to be A. macrostoma . Ŋe type of A. assicana is corymbose* but the branches are thicker and axial dominated* and the radial corallites are described by Brook (1893) as ‘appressed nariform to tubiform’* rather than labellate or cochleariform (ºupporting Information* Fig. º3). Acropora plana Nemenzo * 1967 was also synonymized with A. tenuis by Veron and Wallace (1984) and considered a junior synonym by Wallace (1999) but recognized as a valid species by Veron (2000) and Veron et al. (2016). Ŋe holotype of A. plana appears to be a fragment of a large tabular colony with relatively thin branches* and the radial coralites are appressed with round openings (ºupporting Information* Fig. º3). Ŋese morphological differences* combined with the type location of A. plana in the Philippines * indicate that this species is unlikely to be a synonym of A. tenuis . However* we have not examined the holotype * and further research is needed to resolve the status of this species.

Acropora echinata (Dana, 1846)

Madrepora echinata Dana 1846: 464 * plate 36* fig. 1.

Acropora echinata (Dana) : Verrill 1902: 214.

Madrepora duroillei Milne Edwards and Haime 1860: 148 .

Specimens examined: NMNH-ºI: USNM 275 * Madrepora echinata holotype * Fiji; MNHN 282a * Madrepora duroillei holotype * Fiji; MTQ: G28406* G28460* G28461* G28409* G28458* G56982* G80268 GBR; G78007 Papua New Guinea; G60537* G60571* G80259 Coral ºea; G78322* * G80278 Fiji.

Remarks: Dana’s description of A.echinata is clearly based on USNM 275 ( Fig. 4F )* designated as the lectotype by Wallace (1999). It is a large specimen (~ 50 cm in diameter) from Fiji * which matches Dana’s description of an arborescent corallum with branches that are ‘very neatly bristled with delicate branchlets’ that give it an ‘even cylindrical outline’* ‘smooth tubiform calices’ and a costate coenosteum. We sequenced two specimens from the type locality (G78322 and G80278) with characters that very closely match the lectotype. Ŋese specimens confirm the identity of A. echinata * but the morphology of specimens in our study* from the GBR* the Coral ºea and the Bismarck ºea* show considerable morphological variability. In particular* specimen G80259 from Marion Reef in the Coral ºea is more bushy than the lectotype and topotype specimens from Fiji * while G78007 from the Bismarck ºea has thicker* less dense branches. However* all specimens of A. echinata have terete incipient axial corallites with square walls* compared with the thicker* more rounded tubular walls of other boưlebrush species. Ŋe coenosteum of A. echinata is costate* whereas other boưlebrush species ossen have elaborate spinules [e.g. Acropora subglabra Brook * 1891 (G71541)]. Additional specimens in the collection of MTQ also possess all these characters* providing strong morphological support for their identification as A. echinata .

Madrepora duroillei Milne Edwards & Haime * 1860 from Fiji and Madrepora procumbens Brook * 1891 from the ‘ºouth ºeas’ were considered junior synonyms of A. echinata by Veron and Wallace (1984) and Wallace (1999). Milne Edwards and Haime (1860) note that M. duroillei is very similar to M. echinata . Ŋe M. duroillei holotype (MNHN Z282a) is described as having ‘echinated ribs’ by Milne Edwards and Haime (1860)* which presumably refers to the coenosteum* and is from the same type locality as A. echinata * supporting previous decisions that the species is a junior synonym of this species. However* the type of A. procumbens (NHM 1841.12.11.3) has more rounded corallite walls and spinules on the coenosteum* and therefore more closely resembles boưlebrush species from Clade V [e.g. A. subglabra G71541 and G71543] than those from Clade I. Consequently* although A. procumbens does not appear to be a junior synonym of A. echinata (Dana* 1846)* further research is required to determine its taxonomic status and phylogenetic affinities.

Ŋe A. echinata genome assembled from a specimen collected from the Ryukyu Islands (ºhinzato et al. 2021) is reconstructed in our study in Clade VI (labelled clade IV by ºhinzato et al. 2021)* deeply diverged (> 50 Mya; ºhinzato et al. 2021) from A. echinata (Dana* 1846) sampled from the type locality* which we reconstruct in Clade I (concordant with Clade I in the study by ºhinzato et al. 2021). Unfortunately* there is no voucher or field image associated with this genome* hence we cannot comment on any morphological differences between these species. However* it is possible that the Coral Triangle / north-west Pacific region supports another distinct species with morphological similarities to A. echinata . Consequently* we do not include any specimens from these regions as A. echinata * pending further study. Based on the specimens we examined here* the current geographical range of A. echinata extends from the type locality in Fiji west across the Coral ºea to the Great Barrier Reef and north to the Bismarck ºea (ºupporting Information* Fig. º6).

Acropora kenti (Brook, 1892) status revised Madrepora kenti Brook 1892: 458 ; Brook 1893: 110 * plate 11* fig.B. Acropora kenti (Brook) ; Wells 1954: 493. Here removed from synonymy with Acropora tenuis (Dana* 1846) contra Veron and Wallace 1984; 279. ºee also Cooke et al. 2020; Maưias et al. 2022.

Specimens examined: Madrepora kenti lectotype: NHM 1892.6.8.202* Ŋursday Island * Torres ºtrait* Australia ; syntype: NHM: 1892.6.8.203* Low Woody Island * GBR* Australia; MTQ: G27054* G27078* G27236* G27245* G27280* G27281* G27282* G27284* G27285* G27286* G27287* G27288* G27289* G27292* G27298* G27711* G28444* G29061* G29064* G29899* G29903* G29905* G31175* G32447* G32448* G32455* G32456* G34149* G35014* G35015* G35138* G35139* G35140* G35141* G35142* G35638* G35879* G40909* G41100* G41101* G41102* G41103* G48325* G49331* G58409* G78648 * G78750 * G78778 * G80247 * G80266* * G80256 * G80256 * G78750 * G335181, G335182 GBR; G35638* G35897* G53273* G53275* G53281* G53284* G53288* G78014; G78046* G78064* G80270 * Papua New Guinea; G27291* G28448* G35879* G60564* G63895* G80258 Coral ºea; G60131* G80251 * G80250 Lord Howe Island .

Description: Corallum horizontal or corymbose* main branches 2.0– 2.5 cm thick. Corallites large* round* appressed tubular. Radial corallites appressed tubular* crowded and prominent* with the outer part of the wall forming a rounded lip with no inner wall. ºepta moderately developed* with two directive septa prominent.

Remarks: Ŋe lectotype of A. kenti (ºupporting Information* Fig. º3) differs from the holotype of A. tenuis in having thicker branches that lack tertiary branching and incipient axials* and in possessing cochleariform radial corallites arranged in neat rows. Of the specimens sequenced in this study* G80266 from Great Detached Reef and G80247 from Myrmidon Reef most closely resemble the lectotype * NHM 1892.6.8.202* designated by Wallace (1999) from Ŋursday Island* Torres ºtrait. Brook’s syntype * NHM 1892.6.8.203* from Low Woody Island near Cooktown in the northern GBR is morphologically similar to the lectotype and is almost certainly the same species. Like A. tenuis * the specimens of A. kenti in our phylogeny show considerable variation in gross morphology* in terms of both branch thickness and the neatness of the radial corallites. Two colonies collected adjacent to each other at Myrmidon Reef* GBR* and sequenced in the present series* G80247 and G78778* demonstrate the range of variation in the length and neatness of the radial corallites. ºpecimens also show variation in branch thickness within and between colonies* some of which might be aưributable to environmental factors* such as wave exposure. Despite morphological variability within the species* molecular data clearly demonstrate that A. kenti (Brook* 1892) is distinct from A. tenuis (Dana* 1846); therefore* we formally resurrect the former species. Consequently* the species commonly referred to as A. tenuis on the GBR and elsewhere in eastern Australia (sensu Veron and Wallace 1984)* including the genome published as A. tenuis from the GBR ( Cooke et al. 2020) is A. kenti . Unlike many published genomes* images were taken of the live colonies from Orpheus Island* GBR* that provided the material for the A. kenti genome. Voucher specimens from these colonies were also deposited in the collection of the Queensland Museum (G335181 and G335182)* and the morphology of the specimens closely matches the A. kenti lectotype. Based on the specimens examined here* the geographical range of A. kenti extends from Lord Howe Island north along the GBR to the Bismarck ºea and east to the Coral ºea (ºupporting Information* Fig. º6).

Acropora macrostoma (Brook, 1891) status revised Madrepora macrostoma Brook 1891: 464 ; Brook 1893: 105 * plate 19* fig. B. Acropora macrostoma (Brook) ; Crossland 1952. Here removed from synonymy with Acropora tenuis (Dana* 1846) contra Veron and Wallace 1984; 279.

Specimens examined: NHM: 1878.2.4.7 * Madrepora macrostoma holotype * Mauritius ; MTQ: G51822* G51823 and G51824* from Baie Aux Tortues* Mauritius.

Remarks: ºee remarks under A. tenuis (above). Ŋis species is currently confirmed to occur only in the type locality of Mauritius. Ŋere are literature records from other locations (e.g. Mozambique; ºheppard 1987)* but further research is needed to establish its distribution.

Acropora bifaria (Brook, 1892) status revised

Madrepora bifaria Brook 1892: 453 ; Brook 1893: 110 * plate 30* fig. A. Here removed from synonymy with Acropora tenuis (Dana* 1846) contra Veron and Wallace 1984; 279.

Specimens examined: NHM: 1859.12.12.2 * Madrepora bifaria holotype * Java

Remarks: ºee remarks under A. tenuis (above). Ŋis species is currently confirmed to occur only in the type locality of Java * Indonesia.Further research is needed to establish its distribution.

Acropora aƑicana ( Brook, 1893) status revised Madrepora assicana Brook 1893: 83 ; plate 35* fig. B. Acropora assicana (Brook) ; Crossland 1948. Here removed from synonymy with Acropora tenuis (Dana* 1846) contra Veron and Wallace 1984; 279.

Specimens examined: NHM: 40.9.30.9 * Madrepora assicana holotype * ºouth Africa

Remarks: ºee remarks under Acropora tenuis (above). Brook notes that the collection location of the holotype specimen was recorded as the Cape of Good Hope* but that he believed the specimen came from the south-east coast. Ŋerefore* the specimen is most likely to be from subtropical east Africa. Ŋe species is reported from Mozambique and ºri Lanka based on literature records provided by ºheppard (1987)* and the type locality for the species was incorrectly recorded as ºri Lanka by Veron and Wallace (1984)* but further research is needed to establish its distribution.

Acropora akajimensis Veron, 1990 status revised Acropora akajimensis Veron 1990: 102–106 .

Here removed from synonymy with Acropora donei Veron & Wallace 1984 contra Wallace* 1999 p. 224. ºee also Haưa and Matsushima 2008; Morita et al. 2019.

Specimens examined: MTQ G32475 Acropora akajimensis holotype * Ginanotatejyan * Aka-jima * Japan ; G47767 Aka-jima* Japan; RUMF-ZG-04792* ºesoko Island * Japan .

Remarks: Acropora akajimensis Veron * 1990* from Aka-jima* Japan * was considered a junior synonym of A.donei Veron & Wallace * 1984 by Wallace (1999) * a decision supported by Hoeksema and Cairns (2022). Veron (2000 * 2016) continues to recognize A. akajimensis as a distinct species* noting that it is ‘readily distinguished by its larger* sprawling* indeterminate branching paưern as compared with the determinate growth … of A. donei ’ ( Veron 2016) . Veron’s opinion is supported by the type specimens and original descriptions of both species: A. donei is described as forming colonies that are ‘caespito-corymbose or large corymbose plates or tables’ (Veron and Wallace 1984)* whereas A. akajimensis is ‘irregularly arborescent’ ( Veron 1990). Ŋe species has been the subject of several reproductive studies in Japan (e.g. Haưa and Matsushima 2008* Morita et al. 2019) and was generally referred to as A. donei . Recently* Baird et al. (2022) applied the name A. akaijimensis to the species with tabular-branching growth form and labellate to cochleariform radial corallites that spawns at ~19:30 h on the reefs surrounding ºesoko Island * ~ 20 km from the type locality of Aka-Jima* Japan * on the basis that the morphology more closely resembles the type of A. akajimensis than A. donei . Our study includes one specimen from ºesoko Island (54-5921/RUMF-ZG-04792; Fig. 4Q * R) that closely resembles the type of A. akajimensis . Ŋis specimen occurs within the unresolved sp(p) clade. In addition* two specimens that resemble the type of A. donei collected from the north-central GBR (G78708 and G80265) are in Clade III sensu Cowman et al. (2020) and therefore not closely related to the A. akajimensis specimen from ºesoko. Consequently* we resurrect A. akajimensis as a valid species and suggest that A. donei is unlikely to occur in Japan. Further research is required to confirm the taxonomic identity and geographic range of A. donei and to establish the relationship between A. akajimensis and other species within Clade I-C in the north-west Pacific.

Additional taxonomic remarks on species in the clade I-C Although it is clear that the morphological entity of A. tenuis sensu Veron and Wallace (1984) and Wallace (1999) comprises numerous species* additional taxonomic and molecular phylogenomic research is required to confirm the identities of nominal species considered junior synonyms and to identify and formally describe new species. For example* A. ‘aff. tenius WIO’

( A. aff. tenuis in the study by Colin et al. 2021) from Chagos and Mayoưe is clearly distinct from A. tenuis (Dana* 1846). However* additional sampling across the western Indian Ocean is required to establish whether it is a single species or multiple species. None of the specimens from Chagos in this study or in the MTQ collection closely matches the types of either A. macrostoma or M. assicana (ºupporting Information* Fig. º3) or other nominal Acropora species * suggesting that it is most likely to be an undescribed species. Resolving the taxonomic identity of this species will require the sequencing of topotypes of these nominal species.

Although the systematics of Clade I-C in the north-west Pacific is unresolved* our study nonetheless provides important information about the biogeography of the group. It is clear that none of the specimens in the unresolved sp(p) clade is A. tenuis * A. kenti or A. donei * suggesting liưle overlap of species between the north-west and south-west Pacific. Recent molecular studies indicate the presence of at least two distinct but co-occurring lineages within ‘ A. tenuis ’ around Okinawa ( Zayasu et al. 2021)* which might include nominal species such as A. striata (Verrill* 1866) (discussed below) or represent undescribed species.

Acropora striata Verrill * 1866 from the Ryukyu Islands* Japan * was considered valid by Wallace (1999) and described as ‘distinctive clumps of upright hispidose branches’. Ŋe species A. aff. striata in cluster 2 includes specimens from the north-west Pacific ( Palau and Pohnpei) and a genome identified as A. tenuis from Japan. Morphologically* specimens PL59 from Palau and PN63 from Pohnpei correspond to the interpretation of A. striata by recent authors* including ºhirai (1980)* Veron and Wallace (1984) and Wallace (1999). However* this interpretation does not match the type specimen (ºupporting Information* Fig. º3)* and the description by Verrill states that the species is ‘shrubby or arborescently branched’. Both ºhirai (1980) and Wallace (1999) presume that it is hispidose; however* the holotype is caespitose and lacks tertiary branching. ºhirai (1980) interpreted A. striata as a boưlebrush* and the skeleton illustrated (p. 487) is definitely hispidose; however* the accompanying field image lacks labellate radial corallites and does not closely resemble the holotype. Likewise* the specimens used to illustrate A. striata by Wallace (1999) do not closely resemble the holotype and are not from the type locality. Unfortunately* our present series does not include a topotype with all the skeletal characters required to match Verrill’s type confidently; therefore* although A. striata is likely to be a valid species* its exact identity must remain unresolved. Our A. aff. striata is likely to be an undescribed species* but given the morphological variability within the species* further sampling is required to confirm this. In our series* PN59* from ~ 25 m depth in the lagoon at Ant Atoll* Federated ºtates of Micronesia * has a boưlebrush morphology and sparser radial corallites than PN63 or PL59* which were collected from shallower depths and have a bushy* upright morphology reminiscent of the interpretation of A. striata by ºhirai (1980) and Wallace (1999). Given that both PN59 and PN63 were collected from the same site but different depths* the boưlebrush version could represent a deep-water morphology of this species.

Figures

Figure 2. Maximum likelihood (ML) phylogeny based on the 50% complete edge-trimmed alignment matrix (2032 loci). A* inset phylogeny depicts the relationships among all six major Acropora clades (sensu Cowman et al. 2020) and the three subclades within clade I. Clade I-C is indicated by the dashed box. Ŋe position of Acropora selago (Clade VI) and Acropora donei (Clade III) are indicated. B* ML phylogeny of clade I-C. Ultrafast bootstrap support and gene/site concordance factors are indicated at the nodes (see key). Primary species hypotheses (PºHs) are indicated by clade labels. Field codes for each specimen (tip labels) also reflect the collection location: Great Barrier Reef (73-* 74-* 76-* GBR); Coral ºea (75-); Lord Howe Island* south-east Australia (81-)* Fiji (101-* FJ)* Tonga (TG); Rarotonga* Cook Islands (CI); Kimbe Bay* Bismarck ºea* Papua New Guinea (69-* KM); Palau (PL); Pohnpei (PN); Ningaloo Reef* Western Australia (88-); Ryukyu Islands* Japan (54-); Chagos (BFC); and Mayoưe (MY).

Figure 4. Nominal species of Acropora identified in this study. A* B* Acropora kenti (Brook 1892) topotype GBR130/G80247* Myrmidon Reef* GBR. C* A. kenti lectotype NHM 1892.6.8.202* Ŋursday Island* Australia. D* E* Acropora echinata Dana* 1846 topotype FJ139/G78322* Viti Levu* Fiji. F* A. echinata lectotype USNM 275* Fiji. G* H* Acropora tenuis Dana* 1846 topotype FJ127/G78323* Viti Levu* Fiji. I* A. tenuis holotype USNM 259* Fiji.J* K* Acropora yongei Veron & Wallace* 1984 topotype GBR85/G78644* Lodestone Reef* GBR. L* A. yongei holotype G55079* Britomart Reef* GBR.A field image (less column)* skeletal image (middle column) and type specimen (right column) are shown for each species. M–V* specimens examined in this study that could not be assigned to nominal species: PN63 Acropora aff.striata* Ant Atoll* Micronesia (M* N); KM27/G71345 Acropora aff.echinata* Kimbe Bay* Papua New Guinea (O* P); 54-5921/RUMF-ZG-04792 Acropora akajimensis Veron* 1990 topotype specimen placed within Acropora sp(p). clade* Ryukyu Islands* Japan (Q* R); 54-6036/RUMF-ZG-04796 Acropora sp.* Ryukyu Islands* Japan (º* T); and BFC219 Acropora aff.tenuis ‘WIO’* Chagos* Indian Ocean (U* V).

Figure 5. Acropora tenuissima. Holotype MTQ G78705 in the field at Myrmidon Reef* Great Barrier Reef (A* B); paratype MTQ G78343 from Namaqumaqua* Viti Levu* Fiji (C); holotype skeletal details (D* E); and scanning electron microscope images of axial corallite (F)* radial corallites (G) and coenosteum (H) of the holotype.

Figure 6. Acropora rongoi. Holotype MTQ G78411 in the field the lagoon at Rarotonga* Cook Islands (A* B) and in the laboratory (C* D); paratype MTQ G78418 in the field on the outer reef slope at Papua* Rarotonga* Cook Islands (E* F); and scanning electron microscope images showing axial (G) and radial corallite and coenosteum details (H) of the holotype.