A Genetic Method for Distinguishing Cryptic Pocillopora Species in French Polynesia Without Sequencing Francesca M Cohn, Erika C Johnston, Scott C Burgess, Jordan A Sims, Krithika Layagala, Pierrick Harnay, Hollie M Putnam, Adrienne M S Correa Abstract Pocillopora is a widespread, dominant reef‐building coral genus in the Indo‐Pacific; many species within this group exhibit high morphological similarity and plasticity. Given this, genetic tools are needed to robustly identify Pocillopora individuals to the species level. Quick and accurate identification approaches for Pocillopora species are critical to estimating biodiversity patterns under current and future environmental challenges. In recent years, the mitochondrial open reading frame (mtORF) and a histone region (PocHistone) have been validated using genome‐wide data to become the most widely used species‐level markers for Pocillopora. However, Sanger sequencing of a large number of samples can be prohibitively expensive and sequencing facilities are not always readily available. Therefore, we present restriction fragment length polymorphism (RFLP) digests here that identify the six species of Pocillopora ( P. acuta , P. cf. effusa, P. grandis , P. meandrina , P. tuahiniensis, and P. verrucosa ) found in French Polynesia, without sequencing. In uninformed validation tests (in silico and in vitro), our protocol identified each Pocillopora species with 100% accuracy. Given their cost‐effective, rapid nature, the tailoring of additional RFLP digest protocols to identify cryptic coral species in reef regions around the world will support foundational reef science, conservation and restoration initiatives. This study presents restriction fragment length polymorphism (RFLP) digests that identify the six species of Pocillopora ( P. acuta , P. cf. effusa, P. grandis , P. meandrina , P. tuahiniensis, and P. verrucosa ) found in French Polynesia, without sequencing. In uninformed validation tests (in silico and in vitro), our protocol identified each Pocillopora species with 100% accuracy. Introduction Pocillopora (Lamarck, 1816) is a genus of scleractinian reef‐building corals that is widespread across reefs and reef zones in the greater Indo‐Pacific. Pocillopora is notoriously plastic in its morphology (Todd ; Pinzón et al. ; Marti‐Puig et al. ; Paz‐García et al. ; Gélin et al. ; Johnston et al. ; Johnston et al. ), with environmental factors such as light intensity and water motion (Mass and Genin ; Todd ; Enríquez et al. ; Edmunds and Burgess ) contributing to overall skeletal morphology. However, despite gross morphological similarity (Figure ), cryptic species differ in various ways, such as the corallivore grazing pressure they experience (Burgess et al. ), their reproductive timing (Harnay et al. ), Symbiodinaceae communities (Johnston, Cunning, and Burgess ) and responses to temperature (Burgess et al. ) and light (Edmunds and Burgess ). These differences contribute to a growing body of evidence that morphologically cryptic Pocillopora species do not necessarily share equivalent ecological niches in the reef system (Edmunds and Burgess ; Johnston, Cunning, and Burgess ; Johnston, Wyatt, et al. ). Furthermore, interspecific differences in bleaching support the value of Pocillopora as a model for cryptic species resistance and resilience (natural and with human intervention) to climate change (Burgess et al. , ; Hughes et al. ). The ability to quickly and accurately identify cryptic Pocillopora species within a given morphotype (Oury et al. , ; Voolstra et al. ) is therefore critical to estimating reef biodiversity and understanding how communities of morphologically indistinguishable species may contribute to the robustness of reefs against future environmental challenges (e.g., through portfolio effects, Burgess et al. , Edmunds and Burgess ). To date, approaches using genetic markers (Flot et al. ; Pinzón and LaJeunesse ; Pinzón et al. ; Marti‐Puig et al. ; Schmidt‐Roach et al. ; Gélin et al. ) have delineated species within Pocillopora, resolving phylogenetic relationships between historically confused types (Table ; Johnston, Cunning, and Burgess ; Oury et al. ). A standardized barcoding method using the mitochondrial open reading frame (mtORF; Flot et al. ; Pinzón et al. ; Johnston et al. ) and a histone region (PocHistone; Johnston et al. ) to delineate Pocillopora meandrina and Pocillopora grandis, which share mtORF haplotype 1, have in recent years become the most widely used species‐level markers for Pocillopora (Table ). Importantly, multiple independent genomic datasets and phylogenetic analyses have organized each mtORF haplotype and PocHistone variant into validated species, allowing researchers to identify species based on these markers, rather than whole genome sequencing (Johnston, Cunning, and Burgess ; Oury et al. ; Voolstra et al. ). However, even Sanger sequencing the mtORF region may be prohibitively expensive for studies involving large sample sizes (Johnston et al. ), and sequencing is currently infeasible at many reef locations. A restriction fragment length polymorphism (RFLP) digest and gel protocol was therefore developed to ameliorate these issues in Hawai‘i (Johnston et al. ); this protocol supported reliable identification of Pocillopora spp. by leveraging interspecific nucleotide differences of the mtORF (Flot et al. ) and PocHistone regions (Johnston et al. ) without sequencing. Johnston et al. () were the first to uniquely identify P. meandrina and P. grandis using a histone region. The RFLP protocol developed by Johnston et al. () was designed to differentiate Pocillopora species in Hawai‘i (Pocillopora acuta, Pocillopora damicornis , P. grandis , Pocillopora ligulata, P. meandrina , and Pocillopora verrucosa ), but the complete protocol has not been validated for Pocillopora spp. in other geographic regions, which often harbor additional Pocillopora haplotypes and species (Oury et al. ; Connelly et al. ). Several enzymes that were formerly identified as sufficient to differentiate Pocillopora in Hawaiʻi are confounded by de novo mtORF haplotypes present at other locations due to the presence of different species. For example, the restriction enzyme SacI‐HF does not cut haplotype 8a, a haplotype not present in Hawai‘i, but which is a variant of P. meandrina in French Polynesia (Johnston et al. ; Johnston, Cunning, and Burgess ). Additionally, in Pocillopora samples from Hawai‘i (Johnston et al. ), the use of the restriction enzyme AlwNI differentiates P. ligulata (haplotype 6) from all other Pocillopora species, but this enzyme also cuts mtORF haplotype 11 (Pocillopora cf. effusa), which is not present in Hawai‘i (Forsman et al. ; Gélin et al. ). However, in French Polynesia, AlwNI cannot be used to differentiate P. cf. effusa because P. cf. effusa consists of two mtORF haplotypes, haplotypes 2 and 11 (Johnston, Cunning, and Burgess ), and AlwNI will not cut mtORF haplotype 2. RFLP protocols that accurately discern cryptic Pocillopora species in other regions of the Pacific (besides Hawai‘i; Johnston et al. ) are not yet available; extensive knowledge of existing Pocillopora species in a given region is a prerequisite for RFLP protocol development. Here, we describe an RFLP protocol that uses previously established mtORF and PocHistone regions to identify species of Pocillopora in French Polynesia, broadening the geographic applicability of RFLP methods for cryptic species identification. The diversity of Pocillopora species on Mo'orean reefs has been extensively documented (Forsman et al. ; Johnston, Cunning, and Burgess ; Johnston, Wyatt, et al. ; Burgess et al. ), and therefore RFLP methods could be developed for regionally specific mtORF haplotypes and species (e.g., P. cf. effusa: haplotypes 2, 11; P. meandrina haplotype 8a; P. tuahiniensis: haplotype 10). Additionally, Pocillopora are a dominant reef building genus in this region (Burgess et al. , , ; Johnston, Wyatt, et al. ; Edmunds and Burgess ) contributing to ecological recovery (Tsounis and Edmunds ; Burgess et al. ) following disturbance events including bleaching events, cyclones, and crown of thorns sea star outbreaks (Pratchett et al. , ; Edmunds et al. ). Much research to date, however, has documented trends at the genus level (e.g., Pratchett et al. ; Speare et al. ) because species identification of Pocillopora using morphology is unreliable. We anticipate that the relatively inexpensive, rapid and sequencing‐independent protocol presented here for the identification of cryptic Pocillopora species will facilitate species level monitoring and research efforts on French Polynesian reefs. Methods Mo'orean reefs harbor broadly distributed, common Pocillopora species, as well as several endemic mtORF haplotypes (following the delineation and nomenclature established by Forsman et al. ; Pinzón et al. ). There are six species, containing 17 associated mtORF haplotypes, that are common on reefs around Mo'orea (Edmunds et al. ; Burgess et al. , , ; Johnston, Cunning, and Burgess ; Johnston and Burgess ): P. acuta : 5a (Rouzé et al. ; Burgess et al. ); P. cf. effusa: 2, 11 (Burgess et al. ); P. grandis : 1a (Johnston, Cunning, and Burgess ; Johnston, Wyatt, et al. ; Burgess et al. ); P. meandrina : 1a (same mtORF as P. grandis , but different PocHis amplicon), 1c, 1d, 1e, 8a, 9 (Burgess et al. ); P. tuahiniensis: 10 (Johnston, Cunning, and Burgess ; Johnston and Burgess ; Burgess et al. ); and P. verrucosa : 3a, 3b, 3c, 3d, 3e, 3f, 3h (Table ; Burgess et al. ). Pocillopora damicornis and Pocillopora woodjonesi have also been reported at low abundances based on gross morphology in the Society Islands (Edmunds et al. ), and Pocillopora damicornis was reported at low levels in the Southern and Cook Islands based on the mitochondrial ATP synthase subunit 6 (atp6) and putative control region (pmapcr, Mayfield et al. ). However, given the lack of genetic detection of P. damicornis and P. woodjonesi using widely applied genetic approaches (mtORF and PocHistone) despite extensive studies in Mo'orea (Burgess et al. , , ; Johnston et al. ; Johnston, Wyatt, et al. ; Harnay et al. ), these putative species were not included in the RFLP protocol presented here. To develop a RFLP‐based species identification protocol for Pocillopora species in Mo'orea and the greater French Polynesian region, we used an iterative process of: (1) identifying species‐defining single nucleotide polymorphisms (SNPs) in the mtORF and PocHistone regions (Johnston, Cunning, and Burgess ; Johnston, Wyatt, et al. ) and in silico identification of restriction enzymes that target these sites, and (2) testing the utility of candidate restriction enzymes via in vitro digestion of DNA from previously sequenced Pocillopora samples. The digestion protocol was optimized based on the ability of its component restriction enzymes to distinguish Pocillopora species and haplotypes based on the fragment band sizes they produced. Lastly, in silico and in vitro uninformed validation tests (via temporary redaction of species IDs) were run on sample sets to ensure accuracy and usability of the finalized enzyme protocol. In Silico SNP and Enzyme Identification Methods Species‐specific Pocillopora single nucleotide polymorphisms (SNPs) were identified from a Geneious Prime (v.2024.0.02) alignment of the 17 unique mtORF haplotypes (1a, 1c, 1d, 1e, 2, 3a, 3b, 3c, 3d, 3e, 3f, 3h, 5a, 8a, 9, 10, 11; Data ) documented from Mo'orea based on ~4500 samples drawn from Burgess et al. (), Johnston, Wyatt, et al. (), Burgess et al. (), S.C. Burgess et al. (unpublished), and P. Harnay et al. (unpublished). These mtORF haplotypes are anchored to type material based on morphometric comparisons and location (Schmidt‐Roach et al. ; Johnston and Burgess ), except for Pocillopora effusus (Glynn et al. ; Pinzón et al. ), which we therefore refer to as P. cf. effusa in this study. Candidate restriction enzymes were tested in silico using the Restriction Analysis feature of Geneious Prime. Restriction enzymes were selected for testing and validation based on their potential to differentiate two or more of the focal Pocillopora species. A similar selection process was attempted using an alignment of the PocHistone amplicon for haplotypes associated with each focal Pocillopora species based on a total of 41 samples (Johnston et al. ; Data ). However, high intra‐haplotype variability in the PocHistone amplicon limited this approach as exemplified by the attempted HgaI digest of PocHistone described below (Figure ). Therefore, the PocHistone amplicons were only used as previously described by Johnston et al. () in a XhoI digest to differentiate P. grandis and P. meandrina . To determine the location of each enzyme's cut sites and all expected band sizes of the cut mtORF amplicon, the Flot et al. () primers were mapped onto the P. grandis mitochondrial genome (GenBank accession NC_009798.1). This reference genome was added to the mtORF alignment to determine where predicted cut sites fell within the mtORF amplicon. In Vitro Enzyme Testing Methods A total of eight candidate enzymes were identified for further testing in silico: AciI (New England Biolabs (NEB), MA, USA, #R0551S), BanI (NEB, #R0118S), BceAI (NEB, #R0623S), BseYI (NEB, #R0635S), EcoRV‐HF (NEB, #R3195S), HgaI (NEB, #R0154S), NlaIV (NEB, #R0126S), SacI‐HF (NEB, #R3156S), and XhoI (NEB, #R0146S). Each enzyme was used in in vitro restriction digests applied to mtORF amplicons generated from 42 samples of Pocillopora species representing 12 haplotypes (1a, 2, 3a, 3b, 3e, 3f, 3h, 5a, 8a, 10, 9, 11), collected from 5, 10, and 20 m depth at four sites across the fringing, back, and forereef zones of Mo'orea (Burgess et al. ; Johnston, Wyatt, et al. ; Burgess et al. , S.C. Burgess et al. unpublished). Additionally, the associated PocHistone amplicon of each of the above samples was tested with XhoI to test the digest's ability to differentiate P. grandis and P. meandrina species among haplotypes from French Polynesia. The utility of restriction enzyme digests in accurately parsing Pocillopora species was assessed based on: (1) the extent to which the RFLP ‘fingerprints’ produced by an enzyme matched the band size expected based on in silico analysis of sequence data; and (2) the ability of human users to accurately visually distinguish distinct banding patterns in gels. In preparation for restriction enzyme digest tests, 50 μL aliquots of mtORF amplicons were generated from DNA extractions of the 42 Pocillopora samples using primers FatP6.1 (5′‐TTTGGGSATTCGTTTAGCAG‐3′) and RORF (5′‐SCCAATATGTTAAACASCATGTCA‐3′; Flot et al. ). Polymerase chain reactions (PCRs) were run using Taq PCR Master Mix Kits (Qiagen, Hilden, Germany) and contained 25 μL of master mix, 22 μL of H2O, 1 μL of each primer (forward and reverse, 10 μM), and 1 μL of sample DNA (~20–30 ng/μL). The thermocycler protocol included initial denaturation at 94°C for 60 s; 40 cycles of 94°C for 30 s, 53°C for 30 s, and 72°C for 75 s; and a final extension of 72°C for 5 min (Flot et al. ; Johnston et al. ). PocHistone amplicons were also generated from DNA from each Pocillopora sample using PocHistoneF (5′‐ATTCAGTCTCACTCACTCACTCAC‐3′) and PocHistoneR primers (5′‐TATCTTCGAACAGACCCACCAAAT‐3′; Johnston et al. ) using the same reagents and reaction volumes detailed above for mtORF reactions. However, the following thermocycler program was used in PocHistone PCR reactions: initial denaturation at 94°C for 60 s; 40 cycles of 94°C for 30 s, 53°C for 30 s, and 72°C for 1 min; and a final extension of 72°C for 5 min. Amplicons were stored at −20°C prior to undergoing restriction enzyme digestion. Restriction Enzyme Digestion Conditions for In Vitro Testing To perform enzyme digestions, 5 μL of PCR amplicons generated from each Pocillopora species were digested with 0.5 μL of digestion buffer and 0.5 μL of one restriction enzyme or 1 μL of two combined restriction enzymes for 1 h at 37°C, followed by a 20‐min enzyme denaturation step. Reaction reagents and denaturation temperatures varied for each restriction enzyme assay and are detailed in Table . Following digestion, 5 μL of digested PCR amplicons were run on a 2% agarose gel at 100 V for 45 min. Length of DNA fragments produced was determined by comparison to a 50 bp DNA ladder (NEB, #N3236). Gels were stained with Invitrogen SYBR Safe DNA gel stain (ThermoFisher Scientific, MA, USA) and visualized on a GelDoc Go imaging system (Bio‐Rad, CA, USA). The restriction enzyme NlaIV was previously validated for grouping P. damicornis and P. acuta in a RFLP protocol for Hawai‘i (Johnston et al. ) and was not included in the in vitro enzyme identification process; this enzyme was included, however, in the in vitro uninformed validation. In Silico and In Vitro Uninformed Validation Methods In silico uninformed validations of BseYI/AciI, EcoRV‐HF, and NlaIV were next performed on an mtORF haplotype dataset generated from 673 Pocillopora samples collected in French Polynesia (Johnston, Wyatt, et al. ), a subset of the ~4500 samples used to construct the alignment used for the In Silico SNP and Enzyme Identification Methods. Using the Restriction Analysis feature of Geneious Prime 2024.0.02, mtORF haplotypes were digested with BseYI/AciI, and each haplotype was categorized as either P. verrucosa , P. tuahiniensis, or one of the other four Pocillopora species. A Restriction Analysis digest was then conducted with EcoRV‐HF, and each haplotype was categorized as either P. cf. effusa or one of the other five Pocillopora species. A final Restriction Analysis digest was then run with NlaIV, and each haplotype was categorized as (1) P. tuahiniensis or P. verrucosa , (2) P. grandis , P. meandrina , or P. cf. effusa, or (3) P. acuta . The predicted species for each sample was then compared against the sample metadata to confirm that the correct species had been identified by the in silico analyses. XhoI was not included in the uninformed test because PocHistone sequences were not available from Johnston, Wyatt, et al. (). However, XhoI cutting of PocHistone amplicons was extensively shown to distinguish P. grandis and P. meandrina by Johnston et al. () and it was established that XhoI digested the P. grandis haplotype, while leaving all P. meandrina haplotypes uncut (including the endemic 8a haplotype) during in silico enzyme testing. An in vitro uninformed validation was then conducted on 36 samples of Pocillopora, representing eight haplotypes from P. acuta (haplotype 5a; n = 5 from Mo'orea and n = 1 from University of Rhode Island (URI) aquarium), P. meandrina (haplotypes 1a, 8a; n = 5 each), P. cf. effusa (haplotype 11; n = 5), P. grandis (haplotype 1a; n = 5), P. tuahiniensis (haplotype 10; n = 5), and P. verrucosa (haplotypes 3a, n = 1; 3b, n = 4) sampled at 2 m depth from the lagoonal reefs of the north shore of Mo'orea (Figure , Table ). Initial species identifications were performed at URI by Sanger sequencing the mtORF region using FatP6.1 and RORF primers (Data ). Samples with haplotype 1 were then amplified using PocHistone primers, digested with XhoI, and visualized as described above for species‐level identification. DNA aliquots of each sample (with species IDs redacted) were sent to the University of California, Berkeley, where the uninformed validation of this protocol was performed. PocHistone and mtORF amplicons were generated and each of the four digests plus an additional alternative BceAI/BanI digest (Materials and Methods : BanI/BceAI Digest) were run according to the methods described in In Vitro Enzyme Identification. Results In Silico SNP and Enzyme Identification Here, we detail restriction enzyme cut sites used, based on the presence of fixed SNPs that differentiate Pocillopora species. Note that because some of the enzymes in this protocol have staggered cut sites (e.g., BseYI), the positions described here are calculated from the 5′ forward primer (FatP6.1 or forward PocHistone) to the 5′‐most cut site on the top strand. The top strand is then used to identify the position of all subsequent cut sites on each haplotype. P. acuta (haplotype 5a) was differentiated from all other Pocillopora species in Mo'orea via mtORF using a previously identified fixed Guanine SNP at position 341 recognized by NlaIV and validated in Johnston et al. (). This polymorphism additionally distinguished P. meandrina , P. grandis , and P. cf. effusa from all other species for further differentiation by a fixed guanine SNP at position 312 in P. cf. effusa, P. acuta , P. meandrina , and P. grandis . All species share a cut site for this restriction enzyme at positions 514 and 515. To differentiate P. cf. effusa (haplotype 2, 11) from all other species, a fixed cytosine SNP at position 342 was identified as a cut site of the restriction enzyme EcoRV‐HF in the multispecies mtORF alignment. This fixed cytosine SNP at 342 bp in P. cf. effusa is not shared with any other species. Pocillopora verrucosa (haplotypes 3a, 3b, 3c, 3d, 3e, 3f, 3h) and P. tuahiniensis (haplotype 10) were differentiated from each other and all other species using a co‐digest of AciI and BseYI targeting SNPs in the mtORF alignment. To differentiate P. verrucosa from most species, a fixed cytosine SNP at position 210 was identified in P. verrucosa and P. tuahiniensis and is recognized by AciI. The same restriction cut site is shared by all species at position 546. To simultaneously differentiate P. tuahiniensis from all other species, a fixed cytosine SNP is used at position 398 and recognized by BseYI. Pocillopora grandis and P. meandrina are easily confused at small colony sizes and share the 1a mtORF haplotype. The Johnston et al. () protocol uses a XhoI digest of the PocHistone amplicon to differentiate these species, as XhoI cuts P. grandis , but not P. meandrina, due to a fixed thymine polymorphism at position 279. In the process of designing this protocol, it was identifi ed. that the XhoI digest of P. grandis amplicons produces either two or three bands when run out on a gel. Further sequencing (restriction site associated DNA sequencing, RADseq) of Pocillopora samples from Mo'orea has revealed that some P. grandis individuals are heterozygotic for the XhoI restriction cut site (5′C^TCGAG3′), suggesting that some individuals may be monozygotic for absence of the cut site. Of 194 samples collected from Mo'orea that were identified as P. meandrina by sequencing the mtORF region followed by the PocHistone RFLP protocol (i.e., these samples were monozygotic for absence of the XhoI cut site), only one sample out of 194 (0.5% of colonies analyzed), was erroneously identified as P. grandis following the analysis of RADseq data (Materials and Methods : RADseq Histone Analysis). In Vitro Enzyme Testing EcoRV‐HF (mtORF), BseYI/AciI (mtORF), NlaIV (mtORF), and XhoI (PocHistone) were selected for this protocol because they produced RFLP ‘fingerprints’ that matched the band sizes expected based on in silico analysis of sequence data and were easily distinguishable visually on gels (Table , Figure ). Other enzymes, such as SacI‐HF and HgaI, were not selected for the protocol because haplotypes of a given Pocillopora species exhibited different banding patterns when digested (Figures , ). While other enzymes were possible candidates for the final protocol, including BceAI/BanI to identify P. cf. effusa and group together P. grandis (Haplotype 1), P. meandrina (Haplotype 1a, 1c, 1d, 1e, 8a, 9), and P. acuta (Haplotype 5a), these were ultimately not selected for the protocol because they increased the number of enzymes required to identify all Pocillopora species sequentially (Materials and Methods : BanI/BceAI Digest, Figure , Table ). The EcoRV‐HF (mtORF), BseYI/AciI (mtORF), NlaIV (mtORF), and XhoI (PocHistone) RFLP digests are combinatorially able to unambiguously identify the six Pocillopora species present in Mo'orea. For the majority of Pocillopora samples, only one digest was needed to identify the corresponding species after the sample had been amplified at the mtORF or PocHistone regions. The EcoRV‐HF digest identified P. cf. effusa (mtORF); the NlaIV digest identified P. acuta (mtORF); the BseYI/AciI digest identified each of P. tuahiniensis and P. verrucosa , respectively (mtORF); the XhoI digest identified P. grandis (PocHistone). Three digests and two markers are required to identify P. meandrina : NlaIV (mtORF), EcoRV‐HF (mtORF), and XhoI (PocHistone) (Table ). Using these enzymes, an unknown Pocillopora sample from French Polynesia can be identified to species using a sequential protocol of up to six steps (Finalized Sequential Protocol). In Silico and In Vitro Uninformed Validation of the Protocol The in silico uninformed validation of BseYI/AciI, EcoRV‐HF, and NlaIV digestions identified 100% of the samples (n = 673) accurately based on each digest. Additionally, the in vitro uninformed validation identified 100% of samples (n = 36) to species accurately by RFLP, as confirmed by prior (uninformed) sequencing of the mtORF region of each sample (Figure , Data ). Finalized Sequential Protocol For a collection of unidentified Pocillopora samples, the sequential protocol is (Figure ): - PCR amplify the mtORF region of each sample using the FatP6.1 and RORF primers. - Digest the mtORF amplicons with the restriction enzyme NlaIV to (1) distinguish P. acuta from all other species, (2) group P. tuahiniensis/ P. verrucosa , and (3) group P. cf. effusa/ P. meandrina / P. grandis . - For samples that are in the P. tuahiniensis/ P. verrucosa group, digest the mtORF amplicons with the restriction enzymes BseYI and AciI to distinguish P. tuahiniensis and P. verrucosa . - For samples that are in the P. cf. effusa/ P. meandrina / P. grandis group, digest the mtORF amplicons with the restriction enzyme EcoRV‐HF to distinguish P. cf. effusa from P. meandrina / P. grandis . - For samples that are in the P. meandrina / P. grandis group, PCR amplify the PocHistone amplicon using the PocHistoneF and PocHistoneR primers. - Digest the PocHistone amplicons with the restriction enzyme XhoI to distinguish P. meandrina and P. grandis . Discussion This study presents a restriction enzyme digest method to rapidly and unambiguously differentiate the six species of cryptic Pocillopora present in Mo'orea, French Polynesia, based on their mtORF and PocHistone haplotypes. Taken together with the methods described in Johnston et al. (), this study indicates that the development of additional region‐specific, rapid, and sequencing‐free RFLP protocols holds merit for cryptic coral species identification in geographic reef regions (e.g., the wider Pacific and Red Sea for Pocillopora). In Mo'orea, the morphological similarity and plasticity of Pocillopora species have historically posed a major roadblock to species‐specific research (Figure ; Johnston et al. ; Johnston, Cunning, and Burgess ; Burgess et al. , ). For example, P. tuahiniensis, P. verrucosa, and P. meandrina share similar gross skeletal morphologies (Figure ; Johnston, Wyatt, et al. ; Johnston and Burgess ; Burgess et al. ) and an overlapping depth distribution (2–20 m; Burgess et al. ). The method presented here provides a single digest of the mtORF amplicon with BseYi/AciI to identify P. verrucosa and P. tuahiniensis, and distinguish between them (Table ). P. meandrina and P. grandis are also often difficult to distinguish at smaller sizes, and both can be confused with P. verrucosa (Burgess et al. ). XhoI digest of the PocHistone amplicon identifies P. grandis samples; additional digests of mtORF amplicons with NlaIV, EcoRV‐HF, and BseYI/AciI, respectively, are required to distinguish P. meandrina from P. verrucosa (Table ). Finally, adult P. grandis and P. cf. effusa can attain large colony sizes and be morphologically indistinguishable (figure 1; Johnston, Wyatt, et al. ). These species can be separated with a single digest of the mtORF amplicon with EcoRV‐HF to identify P. cf. effusa samples, followed by an additional XhoI digest of the PocHistone amplicon to identify P. grandis (Table ). As suggested by these examples, the methods presented in this paper can be used individually to identify specific target species, or combined to identify all species, making the protocol highly customizable to the identification needs of any project involving French Polynesian Pocillopora species. However, in most cases (especially for coral recruits), the finalized sequential protocol (Figure ) must be run to accurately identify all Pocillopora species by process of elimination. The biggest advantage of the method presented here is that it can be conducted at many field sites, while eliminating the need for access to sequencing facilities for species‐level identification of Pocillopora. Remote reefs, including some reefs of French Polynesia, can present logistical challenges for molecular work and highlight the need for field‐adaptable genetic identification protocols such as the one we present here. With the advent of small, modular, and open‐source PCR technology (e.g., prebuilt kits, DIY arduino or raspberry pi‐based thermocyclers), the methods described in this paper allow for fast and efficient identification of Pocillopora species at sites without sequencing facilities (see example protocols at Cohn et al. ). This RFLP protocol provides a cost‐effective alternative to the amplicon Sanger sequencing that was previously required to resolve morphologically similar Pocillopora species (Johnston et al. ). At the time of publication, the enzymes and associated buffers used in these methods cost approximately $4.20 USD per sample to run all four digests without any enzyme dilutions—roughly two‐thirds the cost of Sanger sequencing ($7.00 USD total to sequence in both directions). Additionally, depending upon the concentration of amplicons, 1:1 enzyme dilutions in water may be used to cut RFLP protocol costs in half (to $2.10 USD per sample); this is roughly one‐third of Sanger sequencing costs. Therefore, the RFLP protocol provides a flexible range of pricing that depends on the enzymes required, as opposed to a fixed sequencing rate that may be unnecessarily costly (i.e., when trying to distinguish between only two species, such as P. verrucosa and P. tuahiniensis). We detected variation in the PocHistone amplicon used to differentiate P. grandis from P. meandrina . Unlike the mtORF amplicon, a mitochondrial marker, the PocHistone amplicons are generated from nuclear DNA and may therefore be heterozygous. We found that P. grandis amplicons in French Polynesia produce either two or three bands following digestion, suggesting that some of the population is heterozygous for the XhoI cut site (5′ C/TCGAG 3′; Figure ). While this does not interfere with identification of P. grandis using our RFLP protocol, this finding indicates that there may be P. grandis colonies that are homozygous for the absence of the XhoI cut site. However, variation in this P. grandis cut site that leads to the enzyme not recognizing the site (i.e., C/TCGCG, as is observed in P. meandrina ) appears to be very rare—of the 194 colonies identified as either P. meandrina or P. grandis , only one (0.5%) of the P. grandis colonies was homozygous for the XhoI cut site absence (C/TCGCG) (Materials and Methods : RADseq Histone Analysis). We also found a PocHistone amplicon anomaly among P. acuta samples. This included a 345 bp indel starting at bp position 538 (identified by Sanger sequencing; GenBank accession PZ381813), resulting in a substantially longer PocHistone amplicon (~1000 bp) compared to the 669 bp observed in other species (Figure ). Additionally, another P. acuta anomaly exists where another band ~500 bp is present, sometimes with the 1014 bp indel band and expected 669 bp band. However, this does not impact the ability to identify P. grandis with the XhoI digest, as only P. grandis presents the diagnostic 287 and 382 bp bands when digested (Figure ). Despite these caveats, the uninformed validation used in this study was able to identify Pocillopora samples (n = 36) to species with 100% accuracy (Figure ). In terms of mtORF variation, it is noteworthy that in French Polynesia, P. verrucosa is represented by a large number of haplotypes (3a, 3b, 3c, 3d, 3e, 3f, 3h), and many of these haplotypes have few SNP differences between them (e.g., 3b and 3d, 3h and 3e: one SNP difference; 3a and 3b: two SNP differences). In Mo'orea, these haplotypes all correspond to P. verrucosa (Johnston, Cunning, and Burgess ). However, in some study locations outside of French Polynesia, some of these haplotypes represent distinct species. For example, mtORF haplotype 3a and 3d are identified as Pocillopora lacera in the Eastern Tropical Pacific (Oury et al. , ; Connelly et al. ), whereas haplotypes 3c, 3e, 3g, 3h, and 7a are identified as Pocillopora favosa in the Arabian Peninsula (Oury et al. ). For Pocillopora species beyond Mo'orea, mtORF haplotypes have been characterized at neighboring islands, including Tetiaroa, Tahiti, and Maiao (Edmunds et al. ), with no new haplotypes identified at these locations. The Tara Pacific expedition (2016–2018) sampled colonies resembling a typical P. meandrina morphology spanning 18,000 km of the Pacific Ocean, including Mo'orea, Tenoko, Tekava, and Kamaka from the Gambiers, and Aitutaki from the Cook Islands. Samples of Pocillopora species from the expedition did not identify any haplotypes unaccounted for in our methods, further supporting the applicability of our method to the larger region of French Polynesia. However, we do caveat that study sites located further from Mo'orea have the potential to contain species or haplotypes not represented in our results, which may reduce the accuracy of the RFLP protocol presented here at those locations. For example, P. damicornis , identified using variable sites with the mitochondrial ATP synthase subunit 6 (atp6) and putative control region (pmapcr), has been found in low abundance in the Southern and Cook Islands of French Polynesia and is not included in these methods (Mayfield et al. ). However, if applying the RFLP protocol here to the southern islands, it is likely that an additional digest of Tsp45I can be added after the NlaIV digest to distinguish P. acuta from P. damicornis as indicated in Johnston et al. (), although the mtORF haplotypes of P. damicornis in French Polynesia are unconfirmed. Conclusion This paper presents and validates a highly accurate, rapid, and low‐cost sequence‐free method for differentiating the cryptic Pocillopora species present in Mo'orea and the broader geography of French Polynesia using mtORF and PocHistone amplicons. We demonstrate that RFLP protocols can be developed to identify cryptic Pocillopora species in regions other than the Hawaiian Islands; additional protocols could be tailored for other reef regions in the Pacific and the Red Sea (Oury et al. ). As cryptic species are increasingly identified across coral genera (Grupstra et al. ; Rassmussen et al. ), approaches for RFLP protocol development, such as those presented here, are likely to become increasingly necessary globally. Disclosure Benefit‐Sharing Statement: The results of this research, and all associated data and protocols, are open access and are being shared with the provider communities and the broader scientific community. This research can support conservation and restoration activities related to the organisms being studied—a priority concern in the provider community. Contributor Information Francesca M. Cohn, Email: fmc28@stanford.edu. Adrienne M. S. Correa, Email: amsc@berkeley.edu.