Submitted 18 November 2019Accepted 12 January 2020Published 18 February 2020

Corresponding authorsKaho H Tisthammer ktistsfsueduKtisthammergmailcomRobert J Toonenrjtoonengmailcomtoonenhawaiiedu

Academic editorAnastazia Banaszak

Additional Information andDeclarations can be found onpage 17

DOI 107717peerj8550

Copyright2020 Tisthammer et al

Distributed underCreative Commons CC-BY 40

OPEN ACCESS

Genetic structure is stronger acrosshuman-impacted habitats than amongislands in the coral Porites lobataKaho H Tisthammer12 Zac H Forsman3 Robert J Toonen3 andRobert H Richmond1

1Kewalo Marine Laboratory University of Hawaii at Manoa Honolulu HI United States of America2Department of Biology San Francisco State University San Francisco CA United States of America3Hawaii Institute of Marine Biology University of Hawaii at Manoa Kaneohe HI United States of America

ABSTRACTWe examined genetic structure in the lobe coral Porites lobata among pairs of highlyvariable and high-stress nearshore sites and adjacent less variable and less impactedoffshore sites on the islands of Olsquoahu and Maui Hawailsquoi Using an analysis ofmolecular variance framework we tested whether populations were more structuredby geographic distance or environmental extremes The genetic patterns we observedfollowed isolation by environment where nearshore and adjacent offshore populationsshowed significant genetic structure at both locations (AMOVA FST = 004sim019P lt 0001) but no significant isolation by distance between islands Strikingly coralsfrom the two nearshore sites with higher levels of environmental stressors on differentislands over 100 km apart with similar environmentally stressful conditions weregenetically closer (FST = 00 P = 073) than those within a single location less than2 km apart (FST = 004sim008 P lt 001) In contrast a third site with a less impactednearshore site (ie less pronounced environmental gradient) showed no significantstructure from the offshore comparison Our results show much stronger supportfor environment than distance separating these populations Our finding suggeststhat ecological boundaries from human impacts may play a role in forming geneticstructure in the coastal environment and that genetic divergence in the absence ofgeographical barriers to gene flow might be explained by selective pressure acrosscontrasting habitats

Subjects Conservation Biology Ecology Genetics Marine BiologyKeywords Coral reefs Anthropogenic impacts Population genetics Hawaii Lobe coral Isolationby distance Isolation by environment Local adaptation Gene flow

INTRODUCTIONCoral reefs are centers of marine biodiversity and productivity that provide a variety ofecosystem services of substantial cultural and economic value to humankind yet coralreefs worldwide are under serious threat as a result of human activities (Hughes et al 2010Graham 2014) Average global coral cover has declined dramatically in the past 100 yearsdue to a range of impacts such as sedimentation pollution overfishing disease outbreaksand climate change (Hughes et al 2010 Richmond ampWolanski 2011 Graham 2014)

How to cite this article Tisthammer KH Forsman ZH Toonen RJ Richmond RH 2020 Genetic structure is stronger across human-impacted habitats than among islands in the coral Porites lobata PeerJ 8e8550 httpdoiorg107717peerj8550

Such effects are particularly pronounced in nearshore marine habitats which areincreasingly exposed to reduced water quality due to human activities (Wenger et al 2015)Recent rapid coastal development along with coastal industrial and recreational activitieshave resulted in introducing sediments nutrients and a variety of chemical pollutantsto the nearshore environments (Smith et al 2008b Van Dam et al 2011 Wenger et al2015) These local stressors often create a steep environmental gradient of water qualityfrom nearshore toward offshore areas and lsquosigns of coral health impairmentrsquo are usuallydetected alongwith the gradient (eg Smith et al 2008bThompson et al 2014 Ennis et al2016) Additionally nearshore marine habitats are naturally exposed to higher fluctuationsin temperature pH and other environmental variables creating contrasting environmentalconditions relative to more stable offshore environments (Gorospe amp Karl 2011 Guadayolet al 2014) Some corals however continue to thrive in such nearshore lsquosuboptimalrsquohabitats (Morgan et al 2016) indicating that these individuals can withstand suchstressors

What impact does an ecological landscape with such a strong gradient have on thegenetics of the organisms lsquoIsolation by distancersquo (IBD Wright 1943) predicts thatthe degree of genetic differentiation increases with geographic distance due primarilyto dispersal limits (Slatkin 1993 Selkoe et al 2016) Isolation by distance is a neutralprocess in which dispersal limits gene flow and the scale over which genetic structureaccumulates lsquoIsolation by environmentrsquo (IBEWang amp Bradburd 2014) describes a patternin which genetic differentiation increases with environmental differences independentof geographic distance Isolation by environment is a process that emphasizes the role ofenvironmental heterogeneity and ecology in forming genetic structure likely because ofnatural selection (Orsini et al 2013 Wang amp Bradburd 2014) Isolation by environmentcan be generated by different processes including natural selection sexual selectionreduced hybrid fitness and biased dispersal examples of the terms describing a specificcase of IBE include lsquoisolation by adaptationrsquo (IBA Nosil Funk amp Ortiz-Barrientos 2009)lsquoisolation by colonizationrsquo (IBC De Meester et al 2002) and lsquoisolation by resistancersquo(IBR McRae amp Beier 2007) IBA and IBC emphasize the role of selection in forminggenetic structure and IBR describes correlation of genetic distance and resistance distance(ie friction to dispersal) (McRae amp Beier 2007) IBE along with related terms resultin a pattern where genetic distance increases as ecological distance increases but notwith geographic distance for most loci (Orsini et al 2013 Wang amp Bradburd 2014)Theoretically IBA IBC and other processes will result in different distributions of geneticvariation across landscapes (Orsini et al 2013) though in reality multiple processesalmost always contribute to structuring genetic variation and pinpointing the possibleunderlying processes may be difficult (Selkoe et al 2016) For coastal marine ecosystemsoften the distances between the impacted nearshore and un-impacted offshore sites arerelatively small with no apparent dispersal barrier between adjacent sites for broadcastspawning species with pelagic larval development providing an excellent opportunity tostudy IBE

Several studies have shown coral genetic divergence occurring along ecological gradientsFor example Carlon and Budd (Carlon amp Budd 2002) described a pair of incipient species

Tisthammer et al (2020) PeerJ DOI 107717peerj8550 224

in the coral Favia fragum associated with strong ecological gradients The two types arelargely restricted to alternate seagrass and adjacent coral reef habitats but retain phenotypicdistinction in a narrow zone of ecological overlap (Carlon amp Budd 2002) Subsequent workshowed that the morphologies were heritable and selection appeared to limit gene flowbetween the ecomorphs (Carlon et al 2011) Carlon et al (2011) postulated that divergentselection for lsquolsquoTallrsquorsquo and lsquolsquoShortrsquorsquo ecomorphs of these inbred and brooding corals was drivingthe diversification of this coral via an ecological model of speciation (sensu Rundle amp Nosil2005) Genetic divergence across nearshore and costal headland habitats has also beenobserved for a broadcast spawning coral Porites lobata in American Samoa (Barshis et al2010) and for brooding corals Seriatopora hystrix in Australia (Bongaerts et al 2010) andPorites astreoides in the Florida Keys (Kenkel et al 2013) Additionally Gorospe amp Karl(2015) found a significant genetic cline in Pocillopora damicornis along a depth gradientacross a 40 m diameter patch reef in Hawailsquoi

Porites lobata (Dana 1846) the study species occurs over a wide geographic range inthe tropical Pacific Ocean (Veron 2000) and several studies have documented a pattern ofisolation by distance across archipelagic (Polato et al 2010) or broader scales (Baums et al2012) with little evidence of restricted gene flow among geographically proximate reefs atinter-island distances (but see Barshis et al 2010) This massive coral is also known for itsrobustness for example P lobata shows a high tolerance for sedimentation (Stafford-Smith1993) and bleaching (Levas et al 2013) and a colony can recover from partial mortalitydue to tissues residing deep within the perforate skeleton a phenomenon referred to as thelsquoPhoenix effectrsquo (Roff et al 2014) Porites lobata is one of the most dominant scleractiniancoral species in Hawailsquoi (Franklin Jokiel amp Donahue 2013) Additionally P lobata showshigh fidelity to a specific endo-symbiont Symbiodinium Clade C15 which allows us tofocus on responses of the host coral to environmental differences (LaJeunesse et al 2004Smith et al 2008a Barshis et al 2010 Fabina et al 2012)

Skeletal morphological differences for P lobata appear to tell a similar storyas for F fragum (Forsman et al 2015 Tisthammer amp Richmond 2018) thoughenvironmental gradients have not yet been tested in population genetic studiesof this species in Hawailsquoi A growing number of studies indicate that IBE maybe more ubiquitous in the sea than previously assumed Therefore we pose thequestion of whether there is reason to believe non-IBD patterns might be observedin contrasting habitats in Hawailsquoi In order to test our hypothesis we implementa non-traditional population genetics sampling design that allows us to considerwhether geography vs environmental gradients explain overall patterns of geneticdifferentiation

Maunalua Bay Hawailsquoi Olsquoahu was selected as a study site due to the existenceof a strong environmental gradient large-scale urbanization in adjacent watershedshas caused severe deterioration in the health and extent of its nearshore coral reefsover the last century (Wolanski Martinez amp Richmond 2009) Corals that survivein these affected nearshore areas are under chronic stress and a previous surveyshowed significantly different cellular stress responses of individual colonies along theenvironmental gradient of pollutants and sedimentation from the inner bay toward

Tisthammer et al (2020) PeerJ DOI 107717peerj8550 324

offshore (Wolanski Martinez amp Richmond 2009 Richmond 2009 Storlazzi et al 2010) Atthe nearshore site of Maunalua Bay the suspended sediment concentration periodicallyexceeds several hundred mgL and the run-off water introduces toxicants such asbenzo[a]pyrene benzo[k]fluoranthene phenanthrene and alpha-chlordane (WolanskiMartinez amp Richmond 2009 Richmond 2009 Storlazzi et al 2010) The temperaturesalinity and turbidity likewise all show higher ranges of values and fluctuations atthe nearshore site than the offshore site (Storlazzi et al 2010) The distance betweenthe studied offshore and nearshore sites was less than 2 km across the extent of thisgradient and water movement in the areas suggests no dispersal barrier between the sites(Presto et al 2012)

Wahikuli West Maui was selected as another study site due to the existence of asharp environmental gradient that also runs from nearshore to offshore Although lesswell characterized than Olsquoahu the coral reefs off West Maui have a similar gradientof human impacts and experienced a dramatic decline in their coral cover fromland-based sources of anthropogenic stressors over the last several decades (Rodgerset al 2015) Substantial deterioration in the health of West Mauirsquos coral reefs haslead Wahikuli and Honokowai watersheds of West Maui to be designated as prioritysites for conservation and management by the United States Coral Reef Task Force(USCRTF) and the State of Hawailsquoi (Williams et al 2014) The Wahikuli study siteis directly exposed to terrestrial run-off due to its topography and current patterns(Storlazzi et al 2006) causing high turbidity especially after heavy rains Despite theirproximity the nearshore area at Wahikuli has markedly different water quality thanoffshore reefs roughly 300 m away (Fig S2) In contrast the nearshore area at theHonokowai site is less affected by runoff because it does not receive any direct streamdischarge resulting in consistently lower turbidity than the Wahikuli nearshore site(Vargas-Angel 2017)

Porites lobata populations from Maunalua Bay Olsquoahu (hereafter Olsquoahu) and WahikuliMaui (hereafter Maui1) represent two locations with a strong environmental gradientwhereas Honokowai Maui (hereafter Maui2) represents a similar paired nearshore-offshore comparison site but without a strong environmental gradient (Fig 1) Nearshoreand offshore sites at Olsquoahu and Maui1 sites are both characterized by highly contrastingenvironments in proximity with dramatic differences in water quality and sedimentationloads from anthropogenic impacts We undertook a genetic analysis of P lobata acrossthese sites to explore the possibility of isolation by environment By comparing coralscollected from heavily impacted nearshore environments to nearby congeners from moreoceanic conditions we sought to distinguish the roles played by ecology and anthropogenicimpacts to the environment on the genetics of coral populations in contrast to geographicaldistance limiting dispersal among similar habitats on adjacent islands We predictedthat the genetic structure of coral populations from areas with a strong environmentalgradient would follow IBE rather than IBD At each location we assessed the degreesof genetic differentiation and genetic diversity of P lobata between adjacent stronglyanthropogenically impacted lsquohigher-stressrsquo nearshore and lsquolower-stressrsquo offshore sites andcompared them within and between locations to understand the effects of habitat types

Tisthammer et al (2020) PeerJ DOI 107717peerj8550 424

Figure 1Maps of sampling locations

The Main Hawaiian Islands from which samples of Porites lobata were sampled Paired sitesat each location are designated as Nearshore (N) or Offshore (O) and geographic distancesbetween the sampling sites are shown in gray

PeerJ reviewing PDF | (2019114272311NEW 3 Jan 2020)

Manuscript to be reviewed

Figure 1 Maps of sampling locations The Main Hawaiian Islands from which samples of Porites lo-bata were sampled Paired sites at each location are designated as Nearshore (N) or Offshore (O) and geo-graphic distances between the sampling sites are shown in gray

Full-size DOI 107717peerj8550fig-1

anthropogenic impacts and geographical distance on the genetic structure of reef buildingcorals

MATERIALS amp METHODSSpecies identificationDue to its high morphological variation the genus Porites is notorious for its difficultiesin distinguishing between its species (eg Veron 1995 Veron 2000 Forsman et al 2009Forsman et al 2015) Genetic delineation of some Porites including P lobata has beenchallenging due to cryptic species and polymorphic or hybrid species complexes (egVeron amp Pichon 1982 Forsman et al 2009) Although Porites corallites are small irregularand can be highly variable micro-skeletal (corallite) structures have been proposed to beimportant for species identification therefore we examined the corallites of all collectedsamples to confirm our taxonomic identifications (Veron amp Pichon 1982 Veron 2000)(Fig S3) In Hawailsquoi the only Porites species with a similar colony morphology to P lobatais Porites evermanni (there are no records of Porites lutea inHawailsquoi although Fenner (2005)synonymized P evermanni and P lutea they represent two distinct genetic clades (Forsmanet al 2009) P evermanni is genetically distinct from P lobata (Forsman et al 2009)Clade V) and P lutea has a distinct corallite skeletal morphology compared to P lobata(Veron 2000)

Tisthammer et al (2020) PeerJ DOI 107717peerj8550 524

Coral samplingSmall fragments (1 cm2) of P lobata tissue samples were collected from live coloniesbetween February 2013 to May 2017 at the following sampling sites in Hawailsquoi (a)lsquoOlsquoahursquomdashnearshore (n= 22) and offshore (n= 21) sites at Maunalua Bay Olsquoahu(21261sim21278N 157711W) (b) lsquoMaui1rsquomdashnearshore (n= 21) and offshore (n= 23)sites off the Hanakaorsquoo Beach Park West Maui (Wahikuli 2095N 15668W) and(c) Maui2mdashnearshore (n= 23) and offshore (n= 20) sites off the Honokowai BeachPark West Maui (Honokowai 2090N 15669W) (Fig 1 Fig S1) Samples weretaken from coral colonies at least two meters apart at each site and after samplingeach coral colony was photographed and tagged to avoid resampling of the same colonyThe collected tissue samples were either flash frozen in liquid nitrogen on shore andsubsequently stored at -80 preserved in DMSO buffer Gaither et al 2011) or storedin 100 ethanol Genomic DNA was extracted from each coral tissue sample using theQiagenreg DNeasy Blood amp Tissue Kit Coral samples were collected under the State ofHawailsquoi Division of Aquatic Resources Special Activity Permit 2013-26 2014-64 2015-06and 2017-16

PCRFor the samples from Olsquoahu the following three regions of coral host DNA were PCR-amplified (1) sim400 bp coral mitochondrial putative control region (CR) with primersCRf and CO3r (Vollmer amp Palumbi 2002) (2) sim700 bp coral nuclear ITS1-58S-ITS2region (ITS) with primers ITSZ1 and ITSZ2 (Forsman et al 2009) and (3) sim1500 bpcoral nuclear histone region spanning H2A to H4 (H2) with novel primers zH2AH4f(5prime-GTGTACTTGGCTGCYGTRCT-3prime) and zH4Fr (5prime-GACAACCGAGAATGTCCGGT-3prime) H2 was developed to create a genetic marker that allow direct sequencing of postPCR products to efficiently assess small-scale population genetic structure because (1)the mitochondrial genome of P lobata exhibits very little sequence variability (lt002polymorphic sites (Tisthammer et al 2016)) due to its extremely slow evolutionaryrate (Shearer et al 2002) and (2) even though high polymorphism in ITS is a desirabletrait sequencing of ITS requires time-consuming cloning and analyzing the multi-copy gene poses analytical challenges as it deviates from a standard diploid modelH2 consists of partial coding region of H2A approximately 1000 bp introns andpart coding region of H4 After successful development of the H2 marker Mauipopulations were analyzed with only H2 for the above reasons (see File S1 for moredetails)

H2 was amplified under the following conditions 96 C for 2 min (one cycle) followedby 34 cycles consisting of 96 C for 20 s 585 C for 20 s and 72 C for 90 s and a finalextension at 72 C for 5min H2 amplifications (25microl) consisted of 05microl of DNA template02 microl of GoTaqreg DNA Polymerase (Promega Madison WI) 5 microl of GoTaqreg ReactionBuffer 16 microl of 50 mM MgCl2 2 microl of 10 mM dNTPmix 16 microl of each 10 mM primerand nuclease-free water to volume For samples with multiple bands approximately1500-bp PCR products were extracted from agarose gels after electrophoresis and purifiedusing the UltraCleanreg 15 DNA Purification Kit (MO BIO Laboratories Carlsbad CA)

Tisthammer et al (2020) PeerJ DOI 107717peerj8550 624

according to the manufacturerrsquos instruction The rest of the PCR products were purifiedwith UltraCleanreg PCR Clean-Up Kit (MO BIO Laboratories) and sequenced directly inboth directions on the ABI 3730xl DNA Analyzer Clone libraries were created for eachamplified ITS region using the pGEMreg-Easy Vector System (Promega) Positive insertswere verified by PCR using SP6 and T7 primers and plasmids (2ndash5 per library) were treatedwith UltraCleanreg 6MinuteMini Plasmid Prep Kit (MOBIO Laboratories) and sequencedon an ABI-3130XL Genetic Analyzer sequencer For Maui samples H2 was amplified andsequenced using the same method as described above

Sequence analysesResulting DNA sequences were aligned using Geneiousreg 618 (Biomatters Ltd AucklandNewZealand) Polymorphic sites withinH2 regionswere identified usingGeneiousreg (FindHeterozygotes option) and confirmed by eye Middle sections as well as both ends of H2were then trimmed to 1352 bp (for Olsquoahu sequences) or 1221 bp (for combined Olsquoahuand Maui analysis) to minimize missing nucleotides among samples H2 was phased usingthe program PHASE 21 (Stephens Smith amp Donnelly 2001) and SeqPHASE (Flot 2010)The analysis of molecular variance and other population genetic statistics were estimated inArlequin 35 (Excoffier amp Lischer 2010) and TCS 121 (Clement Posada amp Crandall 2000)The global AMOVA with a weighted average over loci with permutation tests was used asimplemented in Arlequin 35 using pairwise difference as a distance computation methodFor H2 both phased and non-phased sequences were run with AMOVA which producedthe same statistical results and therefore only the results from the phased sequences arepresented here Up to five coral ITS sequences were successfully cloned and sequenced percolony and the entire data set was used for calculation of population statistics treatingeach cloned sequence as a haplotype Attempts have been made to conduct genetic analysisusing ITS by (a) treating each sequence as a haplotype (inclusivity) (b) making a consensussequence per individual (consensus by plurality) or (c) using a hierarchal PERMANOVA(Barshis et al 2010) In this study we ran AMOVA using ITS by both (a) and (b) methodswhich produced the same statistical outcome and hence the results from inclusivity (a)are presented in this paper To address the unequal sample sizes (28 vs 44) between thesites in Maunalua Bay the analysis was repeated after resampling to the equal sample size(28) for 10 times All DNA sequences were inspected for the possibility of multi-sampledindividuals and all sampled colonies were considered as separate individuals (genets) sinceno two individuals from a single site shared the same haplotypes (H2) Mantelrsquos test forisolation by distancewas runon the samples inR (R Core Team 2014) using pairwise geneticdistance with 5000 bootstrap permutations Rarefaction analysis was conducted in AnalyticRarefaction 211 (Holland 2009) The network analysis was conducted in SplitsTreev4142 using UncorrectedP as a distance method NeighborNet as a network computingmethod

Tisthammer et al (2020) PeerJ DOI 107717peerj8550 724

Table 1 Analysis of Molecular Variance (AMOVA) for Porites lobata collected fromOlsquoahu (Maunalua Bay)

Source of Variation Variancecomponents

Variance FST

ITS Between populations 227 1918(n= 70) Within populations 956 8082

019

H2 Between populations 029 715(n= 43) Within populations 130 3188

Within individuals 249 60960072

CR Between populations 0034 849(n= 20) Within populations 0370 915

0086(P = 015)

NotesPopulations here refer simply to individuals sampled within the same sampling location with sample size in parentheses below the marker Variance refers to the proportionof genetic variation explained by each comparison and bold numbers with are significant at P lt 0001

RESULTSNearshore vs offshore comparison of genetic structure and diversityof P lobata populationsOlsquoahu (Maunalua Bay)For Olsquoahu P lobata populations the degree of genetic differentiation was estimated usinganalysis ofmolecular variance (AMOVA (Excoffier amp Lischer 2010)) between the nearshoreand offshore sites using three genetic markers CR ITS and novel H2 developed for thisstudy The AMOVA results for both nuclear makers revealed clear genetic differentiationbetween the two sites (ITS FST = 019 P lt 0001 H2 FST = 0072 P lt 0001) (Table 1)The mitochondrial marker (CR) did not detect significant differentiation (FST = 0086P = 015) which was not surprising due to its extremely low variability in corals andcnidarians in general (Shearer et al 2002) The numbers of shared haplotypes (alleles)between the nearshore and offshore Olsquoahu populations were also low out of 37 ITShaplotypes identified from the 70 total sequences only three (81) were shared betweenthe sites For H2 there were 54 unique haplotypes out of 86 total phased sequences andonly 5 sequences (93) were shared between the sites (Table 2) The pattern of geneticstructure was visualized using network analysis which revealed sequences clustering intothree major groups in both ITS and H2 markers which consisted of one cluster dominatedby the nearshore individuals the second one dominated by the offshore individuals andthe last group with approximately mixed origins (Fig 2) For CR three haplotypes wereidentified from 27 sequences all of which were present at both sites Interestingly the mostcommon haplotype was the most dominant one at the nearshore site while the secondcommon haplotype was the dominant haplotype at the offshore site though the AMOVAas well as the exact test results were not significant (P = 015 and 008 respectively) (Fig 3)

The pattern of genetic diversity also differed between the nearshore and offshorepopulations The degree of genetic diversity was higher at the offshore site percent privatealleles (pA) percent polymorphic sites (poly) and nucleotide diversity level (π) werealmost twice as high in the offshore population as in the nearshore one based on ITS(Table 2) Standardizing sample size by random resampling confirmed that this was not

Tisthammer et al (2020) PeerJ DOI 107717peerj8550 824

Table 2 Population genetic statistics of Porites lobata fromOlsquoahu (Maunalua Bay)

Sites ITS (707 bp)

n A pA poly DA DP i π θπ θs

OlsquoahuNearshore

28 13(46)

10(36)

45(64)

10plusmn 00095 0259plusmn 0182 31 00167plusmn 0009 1164plusmn 544 360plusmn 145

OlsquoahuOffshore

42 27(64)

24(57)

70(10)

10plusmn 00052 0343plusmn 0192 50 00340plusmn 0017 2403plusmn 1078 604plusmn 206

OlsquoahuOffshore

(28) 217(78)

193(69)

652(92)

485 00337plusmn 0017 237plusmn 1196 537plusmn 200

H2 (1352 bp)

n A pA poly Ho He(DA)

DP hom π θπ θs

OlsquoahuNearshore

22(44)

28(64)

23(52)

27(20)

0773 0965 0120plusmn 0151 5(23)

000553plusmn 0003 7483plusmn 396 6207plusmn 209

OlsquoahuOffshore

21(42)

31(74)

26(62)

27(20)

0810 0977 0162plusmn 0179 4(19)

000558plusmn 0003 7554plusmn 400 6275plusmn 212

CR (366 bp)

n A pA poly DA DP π θπ θs

OlsquoahuNearshore

13 3(23)

0(0)

2(05)

10plusmn 00302 0282plusmn 0000 000154plusmn 00015 05641plusmn 0551 06445plusmn 0485

OlsquoahuOffshore

14 3(21)

0(0)

2(05)

045plusmn 00270 0451plusmn 0124 00056plusmn 0003 09011plusmn 0747 06289plusmn 0474

NotesSample size (n for H2 the number in parentheses represents the number of phased sequences) number of haplotypes (A) number of private haplotypes (pA) number of poly-morphic sites (poly) mean overall gene diversity (DA plusmn SD) mean gene diversity for polymorphic sites only (DP plusmn SD) observed heterozygosity (HO) expected heterozygosity(He) number of indels (i) number of homozygous individuals (hom) nucleotide diversity (π plusmn SD) theta estimator 1 (θπ expected heterozygosity at a nucleotide position es-timated from the mean π) theta estimator 2 (Watterson estimator θs)Standardized values to the minimum sample size of 28

an artifact of a larger sample size of the offshore population (Table 2) Rarefaction analysisof ITS sequences also confirmed that allelic richness of the offshore population (Richness= 216 plusmn 21 at n= 28 95 CI [188ndash245]) was clearly higher than that of the nearshorepopulation (Richness = 13) (Fig 4) The level of genetic diversity in H2 also appearedslightly higher in the offshore populations the number of haplotypes (A) the number ofprivate allele (pA) the heterozygosity level (HO) the number of heterozygous individualsand mean gene diversity (DA DP) all had marginally higher values in the offshore samplesthough the differences were small (Table 2) In both nuclear markers θπ (the expectedheterozygosity estimated from the average nucleotide diversity) was higher than θs (thetheta estimated from the number of segregating sites)

MauiAt the two study locations on the island of Maui patterns of genetic structure of P lobatapopulations between the nearshore and offshores sites were analyzed using the novel H2marker (see Material and Methods) At Maui1 where a strong environmental gradientexists significant genetic differentiation was detected (FST = 00415 P = 00308) butthe level of genetic diversity was comparable between the two sites (Table 3) At Maui2

Tisthammer et al (2020) PeerJ DOI 107717peerj8550 924

84

16

a ITS b H2

NearshoreOffshore

80

(12)80

(3)20

(13)36

(23)64

8416

(16)

(3)

57(8)69

(18)

31

(20)

23

77(6)

57(16)

(12)43

Figure 2 Allele networks for Porites lobata sampled fromOlsquoahu (Maunalua Bay) NeighborNet phy-logenetic networks generated by SplitsTree based on (A) ITS and (B) H2 color-coded by Nearshore (yel-low) and Offshore (blue) sampling locations as shown in Fig 1 Pie charts represent the proportion of se-quences within each cluster

Full-size DOI 107717peerj8550fig-2

69(9)

(2)(2)

(7)

(4)(3)

31 22

7860

40

NearshoreOffshore

Figure 3 Mitochondrial haplotype network for Porites lobata sampled fromOlsquoahu (Maunalua Bay)Haplotype network based on mitochondrial putative control region (CR) color-coded by Nearshore andOffshore sampling locations as shown in Fig 1

Full-size DOI 107717peerj8550fig-3

which had much less contrasting environmental conditions between the nearshore andoffshore sites the AMOVA results found no significant genetic differentiation between thenearshore and offshore sites (FST = 00019 P = 0991) The level of genetic diversity atMaui2 appeared slightly higher in the nearshore population which had higher numbersof haplotypes (A) polymorphic sites (poly) and heterozygous individuals (Table 3) Thetheta estimators of Maui2 also showed a different pattern from the Olsquoahu and Maui1populations with higher values of θs than those of θπ at both nearshore and offshore sites

Tisthammer et al (2020) PeerJ DOI 107717peerj8550 1024

0 10 20 30 40

05

1015

2025

30

Number of ITS sequences sampled

Allel

ic ric

hnes

s

OffshoreNearshore

Figure 4 Rarefaction curve of allelic richness of ITS sequences from Porites lobata sampled fromnearshore and offshore populations on Orsquoahu (Maunalua Bay) Allelic richness from individualscollected at Nearshore and Offshore sampling sites of Olsquoahu as shown in Fig 1 The gray bar indicates the95 confidence interval of the allelic richness estimation for the offshore population at equivalent samplesize of n= 28

Full-size DOI 107717peerj8550fig-4

Table 3 Population genetic statistics of Porites lobata fromMaui1 andMaui2 sites

H2 (1221 bp)

n A pA poly hom Ho He π θπ θs

Maui1Nearshore

18(36)

31(86)

15(83)

27(22)

3(17)

0833 0987 000596plusmn 00032 7271plusmn 387 6511plusmn 225

Maui1Offshore

22(44)

35(80)

18(82)

31(25)

4(18)

0818 0986 000590plusmn 00031 7209plusmn 382 7126plusmn 234

Maui2Nearshore

23(46)

38(86)

43(93)

65(53)

3(13)

0870 0988 000614plusmn 00035 7494plusmn 396 14790plusmn 448

Maui2Offshore

20(40)

30(75)

37(93)

33(27)

4(20)

0800 0967 000481plusmn 00026 5878plusmn 319 7758plusmn 257

NotesBased on 1319 bp of H2 sequence data Sample size (n represents the number of phased sequences) number of haplotypes (A) number of private haplotypes (pA) number ofpolymorphic sites (poly) mean overall gene diversity (DA plusmn SD) mean gene diversity for polymorphic sites only (DP plusmn SD) observed heterozygosity (HO) expected heterozy-gosity (He) number of indels (i) number of homozygous individuals (hom) nucleotide diversity (π plusmn SD) theta estimator 1 (θπ expected heterozygosity at a nucleotide posi-tion estimated from the mean π) theta estimator 2 (Watterson estimator θs) Standardized values to the minimum sample size of 28

Olsquoahu vs MauiInter-island genetic structure as well as comparison of nearshore and offshore populationswere conducted using H2 marker The hierarchical AMOVA did not detect significantstructure between the Olsquoahu and Maui populations (FCT = 0007 P = 027) but thetwo Maui locations showed significant differentiation (FSC = 0063 P = 0000) based onH2 (Table S2A) The patterns of genetic diversity suggested an overall lower variabilityin the Olsquoahu population the numbers of haplotypes (A) polymorphic sites (poly) and

Tisthammer et al (2020) PeerJ DOI 107717peerj8550 1124

Table 4 Population genetic statistics of Porites lobata from the islands of Oahu andMaui

H2 (1221 bp)

n A pA poly hom Ho He π θπ θs

Maui1 (pooled) 40(80)

63(79)

76(95)

38(31)

7(159)

0825 0986 00060plusmn 00032 7385plusmn 387 7672plusmn 226

Maui2 (pooled) 43(86)

68(79)

82(95)

82(67)

7(162)

0837 0983 000571plusmn 00030 6972plusmn 367 16316plusmn 438

Maui (pooled) 83(166)

126(76)

159(815)

96(79)

14(161)

0831 0987 000604plusmn 00031 7380plusmn 384 16355plusmn 396

Olsquoahu (pooled) 43(86)

54(627)

47(87)

35(29)

9(209)

0791 0974 000643plusmn 00033 7844plusmn 408 6964plusmn 206

NotesBased on 1221 bp of H2 sequence data Sample size (n represents the number of phased sequences) number of haplotypes (A) number of private haplotypes (pA) number ofpolymorphic sites (poly) mean overall gene diversity (DA plusmn SD) mean gene diversity for polymorphic sites only (DP plusmn SD) observed heterozygosity (HO) expected heterozy-gosity (He) number of indels (i) number of homozygous individuals (hom) nucleotide diversity (π plusmn SD) theta estimator 1 (θπ expected heterozygosity at a nucleotide posi-tion estimated from the mean π) theta estimator 2 (Watterson estimator θs) Standardized values to the minimum sample size of 28

heterozygous individuals were all smaller on Olsquoahu although nucleotide diversity (π)levels were relatively similar between Olsquoahu and Maui (Table 4)

Pairwise FST comparisons between all combinations revealed that the nearshorepopulations from Olsquoahu and Maui1 with a high level of environmental stress weregenetically closer to each other than to their respective nearby offshore populationsand similarly the offshore populations from Olsquoahu and Maui1 were genetically closer toeach other than to their respective nearshore populations (Table 5 Olsquoahu and Maui1populations) Assessing by habitat types the nearshore and offshore populations of Olsquoahuand Maui1 also exhibited significant genetic differentiation (FST = 0065 P = 0000Table S2B) The results also revealed that Maui2 corals which showed no significantstructure between the nearshore and offshore sites turned out to be rather geneticallyunique compared to the rest of the populations However the FST values indicated that theMaui2 nearshore population was genetically closer to other offshore populations (FST =0027ndash0035) than to other nearshore populations of Olsquoahu and Maui1 (FST = 014ndash018)suggesting collectively that Maui2 corals at both sites were genetically closer to the offshorepopulations (Table 5)

Genetic structure of P lobata across islands was also visualized using network analysiswhich revealed three major clusters of H2 sequences similar to the results from the Olsquoahupopulations (Fig 5) Grouping by habitat-based genetic groups Cluster 1 was dominatedby the offshore type (including Maui2-nearshore) (88) Cluster 2 was dominatedby nearshore individuals (74) and Cluster 3 had approximately same proportion ofnearshore and offshore types depicting separation of offshore and nearshore individualsespecially for Olsquoahu and Maui1 populations No clear pattern was observed based ongeographic locations (Fig S4)

DISCUSSIONPrevious work on P lobata at the Archipelagic scale (Polato et al 2010) found IBDalthough habitat-level variation was not examined At a much smaller spatial scale

Tisthammer et al (2020) PeerJ DOI 107717peerj8550 1224

Table 5 Pairwise FST values for Porites lobata sampled fromOahu andMaui

Olsquoahu N Olsquoahu O Maui1 N Maui1 O Maui2 N Maui2 O

Olsquoahu N ndash 0000 0324 0000 0000 0000Olsquoahu O 0079 ndash 0008 0666 0012 0046Maui1 N 0001 0055 ndash 0004 0000 0000Maui1 O 0077 minus0007 0052 ndash 0015 0014Maui2 N 0177 0035 0139 0027 ndash 0072Maui2 O 0205 0026 0170 0030 0013 ndash

NotesPairwise FST values were estimated using AMOVA in Arlequin with 5000 permutations Below diagonal= FST values Abovediagonal= P valuesBold values are significant and asterisks represent significance values (P lt 005 P lt 001 P lt 0001)N Nearshore O Offshore

Hm5

B9

HN4

Wd13

HO1WN8

C8

HN10

HN17-2

C5WN3

WO3

C16

M11

M10M1

WN1

WO8C10

B11

HN17-3

Hs9

C14 HN17-5HN5

HO5C6

HO12

Hm9WN12

C1

HO4 HN6WN5

Hs5

HN3

WO1C3

HN9Hs3

WO5

HO11C7Hm4

Hm6 Wd1

HN8 Wd9

N13

WO9

N16 N2 B10 B5 Wd5 Wd2 B2

Wd4HO3

HM17-2

Wd8 Hs6 Wd12

HO6 Wd10

WO7Ws5

N11 B3 M9

M3Ws10 M2

N17

Ws2

C15

C12 Wd11

WN6

Hs4

Ws9

WO6

Ws3

HM17-3 N6

WN9

M4 WO4Hs1Hm7

M8

WN7

Ws11 Ws4 Ws6 M7 WO10 WO2 N1 N10

N12

HO2N7

Wd3

HM17-1

Ws8

Hm8

Wd7

Hs7

Wd6

C9HO10

N4

C2

Ws1 C4WN11

HN2

HN1

WN4

HO9

Hs10

HO8Hs8

HN17-4WN2

N3

M12

B7

HN7

0001

B11

M10

Nearshore

Offshore(Maui2-Nearshore)

A

C

B

(28)

(4)

74

26

(7)

5248

)

12

88

M4 MM4Ws6s6WssssWsWWsWWs6WWWW 6Ws4 s44W 4 WWWWWs111s11WssWsWW

M8WN9WN9WN9 M2M2Ws100sW 0

Ws2WsWWss

W 2

N17171N1M9M9WWN11WN

N11

M1

M1212

WN1N1

WN2WWs8WW 8

WWM11

N3

N4N4

M7 M7 M7 MMMM7MM6 6 M6 M6 MM666 N10N N1 N N

WN6W

N12NNN12NNWNW

WN33 N6 WN7WW

Ws33

6 WddWddWs99sss

s1Wss1 11W 1WWWWNN16166N11C

N222 22N2N2 N2226

N13NNN13NN77N7N77NN7NNNN HN8 HN8HNHN8N888

WN8WNWd1Wdd1WWdWW3

WN5

WN12

WN4W

000C12 212CCCC 2222CCC9

Wd10WWd 00d 0W 0WWdW

WO77W

HO6 WC222

W 33C22

C15555C15Ws3

WO6OOOOO6C5HM17-

HM17-1M 77

C161C161N44

WO3

WO22 WO10 WOWO1WO1WO1WWWWO4 W

B3 3 M M B

C14

C10C10WO8

M3M3MM3MMMM3

B2 Wd222 2 Wd5 W B5 WC44C4 C44C4

Wd1313Ws55WssWsWO7Ws55sW

B100B B BB2 B2 B 0 0 Hm666 66

HO3HO3HO3

HO22O2HWO99O9WOWWWWO9OO

HOOH C9

C8

HO8HO

C7

HO10WO5O5

Wd9WWWd9 Hm5Hm4H 4B9B9

Hs8HsHOHs8s

HO1HO1

Hm7

B7B7BB7B5

Hm8

HO11

C3 WO1

12dWdWWd1Wd11112Wd1112d122d112WWWW

Wd3Wd7

HO4HOHm9H

C1

Wd666d66666HM17-2

Wd4Wd4dd4d4dWdd4WdWdWdWd4Wd4

HHN17-5HH 5 Wd8

HO9HHHO9HHH

C6HO12HO

4

HN5HH5

H

HN7HHHHHHHHHNHNNNHNNHNHNNN

Hs177

HN6Hs556HH56

HN1NNHNNNNNNNNNNNN WHNN W666666 666 HsHsHsHss6 W WW

d66d68

HN2N22 HH

HN4

HN100

HN17-2HN17-2

HN9

Hs1010H

Hs3

HN3H

Hs7

HO5

11Wd1W

HN17-41N1N12N1

HN17-1N1HN17-4N12N1

HN17-3HN17-33-3HH

Hs9

Hs4

Figure 5 Diagrams of neighbor-net tree networks for Olsquoahu andMaui P lobata populations based onunphased H2 sequences Genetic structure is stronger across human-impacted habitats than among is-lands in the coral Porites lobata NeighborNet phylogenetic networks generated by SplitsTree color-codedby habitat-based genetic clusters Blue represents the offshore group (including Maui2-nearshore popula-tion) and yellow represents the two genetically-close nearshore populations of Olsquoahu and Maui1 The piecharts show the proportion of sequences present in each group The gray numbers in () represent the pro-portion of Maui2 nearshore population

Full-size DOI 107717peerj8550fig-5

considered here with explicit sampling of contrasting habitats from across this strongenvironmental gradient the pattern of genetic structure we observed for P lobata inHawailsquoi did not follow IBD (Mantel Test r =minus0091 P = 054) Instead a pattern

Tisthammer et al (2020) PeerJ DOI 107717peerj8550 1324

that resembles IBE was revealed with a correlation between habitat types irrespective ofgeographic distance the pairwise FST values revealed that offshore individuals from twoseparate islands (gt100 km) were genetically closer to each other than to their geographicallyclosest nearshore individuals (03ndash2 km) and nearshore individuals from two islands werealso typically genetically closer to each other than to corals from adjacent offshore sites(Table 5 Fig 5) Other studies that examined genetic differentiation onmultiple scales showthat patterns of IBD can be scale dependent For example Gorospe and Karl found distance(2013) and depth (2015) dependent patterns of coral genetic relatedness on a single patchreef but this pattern was no longer detectable for analyses on an inter-reef scale (Gorospe ampKarl 2013) Furthermore because of intra-reef patterns of genetic diversity and clonalityit is possible that sampling design may bias population-level patterns of genetic diversity(Gorospe Donahue amp Karl 2015) However we do not think this potential bias presents aproblem in our system since P lobata is a broadcasting spawning species and shows littleclonality (Boulay et al 2012 Boulay et al 2013 Schweinsberg Tollrian amp Lampert 2016)particularly in Hawailsquoi (Polato et al 2010)

Similar patterns of small-scale genetic structure have been observed in several reefbuilding corals (Barshis et al 2010 Bongaerts et al 2010Carlon amp Lippeacute 2011) In the caseof Favia fragum the lsquolsquoTallrsquorsquo ecomorph is a seagrass specialist withmorphological adaptationsto minimize sediment impacts whereas the lsquolsquoShortrsquorsquo ecomorph shows morphologicalspecializations for coral reef habitats that decrease its fitness in seagrass beds (Carlonamp Budd 2002) These traits are highly heritable and divergent selection appears to bedriving reproductive isolation among these morphs and ongoing diversification in theseincipient species (Carlon amp Lippeacute 2011) Here we find a similar pattern of repeatedgenetic differentiation across strong ecological gradients driven by anthropogenic impacts(between adjacent sites 03ndash2 km apart) but little evidence of differentiation among similarhabitats more than 100 km distant Moreover corals from a control site that lacks sucha strong environmental gradient (Maui2) did not show the significant genetic structuredocumented at the other sites which further suggests the potential role of environment informing the observed genetic patterns (IBE)

Additionally for the Olsquoahu populations differences in microskeletal morphology havebeen reported between the nearshore and offshore sites (Tisthammer amp Richmond 2018)similar to the correlation seen between genetic distance and microskeletal morphologyacross broader geographic regions in P lobata (Forsman et al 2015) The most noticeabledifference was the height of pali (inner vertical skeletal structure that usually exists in a setof eight in P lobata) within a corallite (the structure associated with individual polyps)nearshore corals had taller andmore pronounced pali than the offshore ones Because exactfunctions and heritability of the traits are unknown whether the observed morphologicaldifferences are due to divergent selection cannot be answered at this point However thestudy suggested that the differences might be due to potential beneficial roles played bylarger pali in shedding sediments in turbid water (Tisthammer amp Richmond 2018) similarto the case of the Caribbean coral F fragum (Carlon amp Budd 2002 Carlon amp Lippeacute 2011Carlon et al 2011) Correlation between the morphological and genetic distances reported

Tisthammer et al (2020) PeerJ DOI 107717peerj8550 1424

in the study is consistent with the idea of the likely role of environment in observeddivergence

Although we cannot rule out the possibility of prevailing oceanographic currents orbarriers playing a role in producing our observed pattern of genetic patchiness detailedstudies on water movement suggests little dispersal barrier between the nearshore andoffshore sites In the bay at the Olsquoahu site surface currents primarily flow west due tothe prevailing trade-winds (offshore to nearshore) The below surface current movementseems to be more complex and is generally towards the east (nearshore to offshore)with the presence of small eddies at least during the summer (Presto et al 2012) Eddieswould increase the larval retention time in the bay during the summer spawning seasonespecially for Porites species that produce neutrally buoyant gametes (Hunter 1988) ForMaui sites the detailed water movement around the study areas revealed the presence ofshear zones between the shallower (lt15 m) nearshore and the deeper portions of the reefdue to opposing flow directions during the coral reproductive season (summer) suggestinglocal retention of larvae (Storlazzi et al 2006 Storlazzi amp Field 2008) While these shearzones could contribute as a gene flow barrier between the shallower and deeper areas ouroffshore sites were located within the nearshore southward-flow dominated area based ontheir depths and distances from the shore thereby eliminating the probability of existenceof strong oceanographic barriers between the nearshore and offshore study sites

It is particularly interesting to find not only this pattern of clear genetic partitioningwithin a bay is repeated on a neighbor island but also the genetic similarity exists betweensites with similar histories of anthropogenic stress from the two separate islands whichpoints to the possibility of local processes acting to homogenize these geographically-separated sites The nearshore sites at Olsquoahu and Maui1 have experienced deterioratingwater and substrate quality due to terrestrial runoff from urbanization of adjacentwatersheds over the past century This repeated pattern implies a possibility of similarunderlying processes at both locations We suspect that the environmental decline haslikely limited new recruitment to the affected nearshore sites (Puritz amp Toonen 2011)and may have placed the populations under local selection similar to what was seen inthe Caribbean coral F fragum (Carlon amp Budd 2002 Carlon amp Lippeacute 2011) The geneticmarkers used here are unlikely to be the direct targets of selection and therefore theobserved genetic-environment association may result from linkage or be attributed tocoupling of endogenous (intrinsic) and exogenous barriers as theorized by Bierne etal (2011) According to Bierne et al (2011) intrinsic barriers that are independent ofenvironment often play an important role in forming genetic-environmental associationsthrough coupling with exogenous factors rather than ecologically based divergent selectionalone Although our understanding on how selection affects population differentiationat neutral markers is limited at this point (Bierne et al 2011) the coupling hypothesismay explain parallel genetic differentiation across independent putatively neutral markersobserved in our study Furthermore there appears to be reduced genetic diversity in thenearshore habitats (Table 2 Table S1) which may represent a subset of the standinggenetic variation of the larger population thereby representing a limited number ofindividuals capable of surviving in the nearshore habitats Regardless of the precise

Tisthammer et al (2020) PeerJ DOI 107717peerj8550 1524

underlying mechanism the role of environment in shaping the observed genetic divergenceappears more important than geographic distance in this case

Additional work is also needed to determine the specific environmental drivers likelyto result in selection across these environmental gradients that generate the observedpattern As discussed earlier there are many factors both natural and anthropogenicthat contrast between nearshore and offshore environments (eg salinity irradiance UVexposure temperature pH wave exposure nutrients and biological community) Alsofine-scale water movements that the oceanographic-scale studies cannot capture (isolationby resistanceThomas et al 2015) may be contributing to physical barriers to reproductionAny of these factors could contribute to create genetic partitioning between nearshore andoffshore sites For example a comparable pattern of genetic structure has been observedacross a particularly strong temperature gradient between the back-reef and forereef Plobata populations in the areas with negligible terrestrial runoff or pollution in AmericanSamoa (Barshis et al 2010) However in Hawailsquoi we did not see such differentiationbetween nearshore and offshore sites at our less-impacted reference (Maui2) which raisesa question of a potential role of other local abiotic or biotic factors affecting the observedpattern It will be interesting to continue to observe whether this differentiation is transientand of little evolutionary importance or whether it progresses towards incipient speciationas it appears to have done in the Caribbean coral F fragum (Carlon amp Budd 2002 Carlonamp Lippeacute 2011) Indeed our finding may shed light on the common pattern of lsquochaoticgenetic patchinessrsquo (Johnson amp Black 1982) so commonly reported among populationgenetic studies of marine organisms in which geographically proximate populations showgreater genetic differentiation than those from distant sites (eg Johnson amp Black 1984Hogan Thiessen amp Heath 2010 Toonen amp Grosberg 2011 Eldon et al 2016)

The ecological diversification of reef building corals over a small spatial scale despiteongoing gene flow also provides a rare example of genetic divergence in the absenceof spatial barriers to gene flow indicating that divergent natural selection can act asan evolutionary driver of reproductive isolation (Carlon amp Budd 2002 Bongaerts et al2010 Bongaerts et al 2011 Carlon et al 2011) Our results may extend another exampleof P lobata in Hawailsquoi to these findings which showed potential occurrence of similardiversification process across steep environmental gradients driven This may representthe initial stages of adaptive diversification as seen in other marine species from theHawaiian Archipelago (eg limpets Bird et al 2011 Bird 2011) There is clearly somegenetic connectivity among adjacent islands congruent to previous studies (Polato et al2010) and hence the observed differentiation across these steep ecological gradients inspite of high dispersal potential (Storlazzi amp Field 2008 Presto et al 2012) may be theearly phases of speciation with gene flow (Feder Egan amp Nosil 2012) Whether this initialstage of divergence among habitats is transient or has the potential to progress to laterstages remains to be seen but our results and others (Bird et al 2011 Bird 2011) indicatethat this initial stage can be realized even in a broadcasting species with high dispersalpotential Together these results add to the growing evidence that the initial phase ofspeciation is possible without geographic isolation and lend support to the hypothesis

Tisthammer et al (2020) PeerJ DOI 107717peerj8550 1624

that ecological speciation (sensu Bowen et al 2013) may be more common in the sea thanbelieved previously

Finally the potential reduction in intraspecific genotypic diversity within the populationsof corals under higher levels of environmental stressors can be a warning for the future ofreefs Focusing on the usual metrics of coral reef health and resilience specifically coralcover and species diversity misses the very concerning loss of genotypic diversity essentialto population resilience in a changing world The importance of molecular data is apparentin tracking the invisible changes that can lead to local extinctions and hence such toolsneed to be further refined and more broadly applied to the helping ensure the persistenceof coral reefs as a legacy for the future

CONCLUSIONSOur results show that the broadly distributed broadcast spawning coral P lobata exhibitsvery fine-scale (03sim2 km) genetic structure and environmental drivers across habitat typesappear to have a stronger effect in forming such genetic structure (IBE) than geographicdistances (IBD) in coastal areas that are heavily affected by anthropogenic stressorsGenetic similarity found between Olsquoahu andMaui1 nearshore populations suggest that theobserved genetic structure may be driven by similar underlying evolutionary mechanismsincluding adaptation to environmentally driven factors Our results also highlight theimportance of thorough sampling among habitats at small scales we could easily overlooksuch important local genetic differences and may mistakenly conclude that populationsare uniform across the landscape without thorough sampling Although our samples arefrom limited locations these results demonstrate that understanding small-scale geneticvariation and diversity can provide important information on the ecological basis of geneticdiversity and differentiation which must be understood to effectively implement futurecoral reef conservation efforts

ACKNOWLEDGEMENTSWe thank the following people for assistance in collecting corals M Stiber V Sindorf FSeneca J Murphy J Martinez A Lyman K Richmond V Houmllzer N Spies and A IrvineSpecial thanks to T Oliver for his support and advice ZHF would like to thank the SeaverInstitute for support

ADDITIONAL INFORMATION AND DECLARATIONS

FundingFinancial support came from the National Fish and Wildlife Foundation (Grant Number34413) the Hawaii Department of Health (MOA 13-502) and the National Oceanic andAtmospheric Administration Coral Reef Ecosystem Studies Program (NA17NMF4520144)to Robert H Richmond ZacH Forsmanwas supported by awards from the Seaver Instituteand NSF-OA1416889 to Robert JToonen The funders had no role in study design datacollection and analysis decision to publish or preparation of the manuscript

Tisthammer et al (2020) PeerJ DOI 107717peerj8550 1724

Grant DisclosuresThe following grant information was disclosed by the authorsNational Fish and Wildlife Foundation Grant Number 34413Hawaii Department of Health MOA 13-502NationalOceanic andAtmospheric Administration Coral Reef EcosystemStudies ProgramNA17NMF4520144Seaver Institute and NSF-OA1416889

Competing InterestsRobert J Toonen is an Academic Editor and Section Editor for Aquatic Biology for PeerJ

Author Contributionsbull Kaho H Tisthammer conceived and designed the experiments performed theexperiments analyzed the data prepared figures andor tables authored or revieweddrafts of the paper and approved the final draft

bull Zac H Forsman conceived and designed the experiments performed the experimentsanalyzed the data authored or reviewed drafts of the paper and approved the final draft

bull Robert J Toonen analyzed the data authored or reviewed drafts of the paper andapproved the final draft

bull Robert H Richmond conceived and designed the experiments analyzed the dataauthored or reviewed drafts of the paper and approved the final draft

Field Study PermissionsThe following information was supplied relating to field study approvals (ie approvingbody and any reference numbers)

Coral samples were collected under the State of Hawaii Division of Aquatic ResourcesSpecial Activity Permits 2013-26 2014-64 2015-06 and 2017-16

Data AvailabilityThe following information was supplied regarding data availability

All raw sequence data are available at GenBankITS KY493091ndashKY493160 H2 KY502280ndashKY502376 and MF629151ndashMF629152 CR

KY502373 KY502374 and KY502375The phased H2 sequences used in the analysis are available on Figshare Toonen Rob

(2019) Phased_H2_sequence_datadocx figshare Dataset httpsdoiorg106084m9figshare10315430v1

Supplemental InformationSupplemental information for this article can be found online at httpdxdoiorg107717peerj8550supplemental-information

REFERENCESBarshis DJ Stillman JH Gates RD Toonen RJ Smith LW Birkeland C 2010 Protein

expression and genetic structure of the coral Porites lobata in an environmentally

Tisthammer et al (2020) PeerJ DOI 107717peerj8550 1824

extreme Samoan back reef does host genotype limit phenotypic plasticityMolecularEcology 19140297ndash140297 DOI 101111j1365-294X201004574x

Baums IB Boulay JN Polato NR Hellberg ME 2012 No gene flow across the East-ern Pacific Barrier in the reef-building coral Porites lobataMolecular Ecology215418ndash5433 DOI 101111j1365-294X201205733x

Bierne NWelch J Loire E Bonhomme F David P 2011 The coupling hypothesiswhy genome scans may fail to map local adaptation genesMolecular Ecology202044ndash2072 DOI 101111j1365-294X201105080x

Bird CE 2011Morphological and behavioral evidence for adaptive diversificationof sympatric Hawaiian Limpets (Cellana spp) American Zoologist 51466ndash473DOI 101093icbicr050

Bird CE Holland BS bowen BW Toonen RJ 2011 Diversification of sympatricbroadcast-spawning limpets (Cellana spp) within the Hawaiian archipelagoMolecular Ecology 202128ndash2141 DOI 101111j1365-294X201105081x

Bongaerts P Riginos C Hay KB Van OppenMJ Hoegh-Guldberg O Dove S 2011Adaptive divergence in a scleractinian coral physiological adaptation of Seriatoporahystrix to shallow and deep reef habitats BMC Evolutionary Biology 11303DOI 101111j1365-294X201105081x

Bongaerts P Riginos C Ridgway T Sampayo EM Van OppenMJH Englebert NVermeulen F Hoegh-Guldberg O 2010 Genetic divergence across habitats in thewidespread Coral Seriatopora hystrix and its associated Symbiodinium PLOS ONE5e10871 DOI 101371journalpone0010871t001

Boulay JN Cortes J Nivia-Ruiz J Baums IB 2012High genotypic diversity of the reef-building coral Porites lobata (Scleractinia Poritidae) in Isla del Coco National ParkCosta Rica Revista de Biologiacutea Tropical 60279ndash292

Boulay JN Hellberg ME Cortes J Baums IB 2013 Unrecognized coral species diversitymasks differences in functional ecology Proceedings of the Royal Society B BiologicalSciences 28120131580ndash20131580 DOI 101111j1365-294X200502553x

Bowen BW Rocha LA Toonen RJ Karl SA Laboratory TT 2013 The originsof tropical marine biodiversity Trends in Ecology amp Evolution 28359ndash366DOI 101016jtree201301018

Carlon DB Budd AF 2002 Incipient speciation across a depth gradient in a scleractiniancoral Evolution 562227ndash2242 DOI 101111j0014-38202002tb00147x

Carlon DB Budd AF Lippeacute C Andrew RL 2011 The quantitative genetics ofincipient speciation heritability and genetic correlations of skeletal traits inpopulations of diverging favia fragum ecomorphs Evolution 653428ndash3447DOI 101111j1558-5646201101389x

Carlon DB Lippeacute C 2011 Estimation of mating systems in Short and Tall ecomorphs ofthe coral Favia fragumMolecular Ecology 20812ndash828DOI 101111j1365-294X201004983x

Clement M Posada D Crandall KA 2000 TCS a computer program to estimate genegenealogiesMolecular Ecology 91657ndash1659 DOI 101046j1365-294x200001020x

Tisthammer et al (2020) PeerJ DOI 107717peerj8550 1924

DeMeester L Goacutemez A Okamura B Schwenk K 2002 The Monopolization Hypoth-esis and the dispersalndashgene flow paradox in aquatic organisms Acta Oecologica23121ndash135 DOI 101016S1146-609X(02)01145-1

Eldon B Riquet F Yearsley J Jollivet D Broquet T 2016 Current hypotheses to explaingenetic chaos under the sea Current zoology 62551ndash566 DOI 101093czzow094

Ennis RS Brandt ME Grimes KRW Smith TB 2016 Coral reef health response tochronic and acute changes in water quality in St Thomas United States Virgin Is-landsMarine Pollution Bulletin 111418ndash427 DOI 101016jmarpolbul201607033

Excoffier L Lischer HEL 2010 Arlequin suite ver 35 a new series of programs toperform population genetics analyses under Linux and WindowsMolecular EcologyResources 10564ndash567 DOI 101111j1755-0998201002847x

Fabina NS PutnamHM Franklin EC Stat M Gates RD 2012 Transmission modepredicts specificity and interaction patterns in coral-Symbiodinium networks PLOSONE 7e44970 DOI 101371journalpone0044970s006

Feder JL Egan SP Nosil P 2012 The genomics of speciation-with-gene-flow Trends inGenetics 28342ndash350 DOI 101016jtig201203009

Fenner D 2005 Corals of Hawaii Honolulu Mutual Publishing LLCFlot JF 2010 SeqPHASE a web tool for interconverting PHASE inputoutput files

and FASTA sequence alignmentsMolecular Ecology Resources 10162ndash166DOI 101111j1755-0998200902732x

Forsman ZH Barshis DJ Hunter CL Toonen RJ 2009 Shape-shifting corals molecularmarkers show morphology is evolutionarily plastic in Porites BMC EvolutionaryBiology 945 DOI 1011861471-2148-9-45

Forsman ZWellington GM Fox GE Toonen RJ 2015 Clues to unraveling the coralspecies problem distinguishing species from geographic variation in Porite-sacross the Pacific with molecular markers and microskeletal traits PeerJ 3e751DOI 107717peerj751supp-1

Franklin EC Jokiel PL DonahueMJ 2013 Predictive modeling of coral distributionand abundance in the Hawaiian IslandsMarine Ecology Progress Series 481121ndash132DOI 103354meps10252

Gaither MR Szaboacute Z CrepeauMW Bird CE Toonen RJ 2011 Preservation of corals insalt-saturated DMSO buffer is superior to ethanol for PCR experiments Coral Reefs302(2)329ndash333

Gorospe KD DonahueMJ Karl SA 2015 The importance of sampling design spatialpatterns and clonality in estimating the genetic diversity of coral reefsMarineBiology 1625917ndash928 DOI 101007s00227-015-2634-8

Gorospe KD Karl SA 2011 Small-scale spatial analysis of in situ sea temperaturethroughout a single coral patch reef Juornal of Marine Biology 20111ndash12DOI 1011552011719580

Gorospe KD Karl SA 2013 Genetic relatedness does not retain spatial pattern acrossmultiple spatial scales dispersal and colonization in the coral Pocillopora damicor-nisMolecular Ecology 223721ndash3736 DOI 101111mec12335

Tisthammer et al (2020) PeerJ DOI 107717peerj8550 2024

Gorospe KD Karl SA 2015 Depth as an organizing force in Pocillopora damicornisintra-reef genetic architecture PLOS ONE 10e0122127ndash17DOI 101371journalpone0122127

GrahamNAJ 2014Habitat complexity coral structural loss leads to fisheries declinesCurrent Biology 24R359ndashR361 DOI 101016jcub201403069

Guadayol Ograve Silbiger NJ DonahueMJ Thomas FIM 2014 Patterns in temporalvariability of temperature oxygen and pH along an environmental gradient in acoral reef PLOS ONE 9e85213 DOI 101371journalpone0085213

Hogan JD Thiessen RJ Heath DD 2010 Variability in connectivity indicated by chaoticgenetic patchiness within and among populations of a marine fishMarine EcologyProgress Series 417263ndash275 DOI 103354meps08793

Holland S 2009 Analytic rarefaction Version 2 software Athens Hunt MountainSoftware

Hughes TP GrahamNAJ Jackson JBC Mumby PJ Steneck RS 2010 Rising tothe challenge of sustaining coral reef resilience Trends in Ecology amp Evolution25633ndash642 DOI 101016jtree201007011

Hunter CL 1988 Genotypic diversity and population structure of the Hawaiian reefcoral Porites compressa PhD Dissertation University of Hawaii at Manoa

JohnsonMS Black R 1982 Chaotic genetic patchiness in an intertidal limpetSiphonaria spMarine Biology 70157ndash164 DOI 101007BF00397680

JohnsonMS Black R 1984 Pattern beneath the chaos the effect of recruitment ongenetic patchiness in an intertidal limpet Evolution 381371ndash1383DOI 101111j1558-56461984tb05658x

Kenkel CD Goodbody-Gringley G Caillaud D Davies SW Bartels E Matz MV 2013Evidence for a host role in thermotolerance divergence between populations of themustard hill coral (Porites astreoides) from different reef environmentsMolecularEcology 224335ndash4348 DOI 101111mec12391

LaJeunesse T Thornhill D Cox E Stanton F Fitt W Schmidt G 2004High diversityand host specificity observed among symbiotic dinoflagellates in reef coral commu-nities from Hawaii Coral Reefs 23596ndash603 DOI 101007s00338-004-0428-4

Levas SJ Grottoli AG Hughes A Osburn CL Matsui Y 2013 Physiological andBiogeochemical traits of bleaching and recovery in the mounding species of coralPorites lobata implications for resilience in mounding corals PLOS ONE 8e63267DOI 101371journalpone0063267t006

McRae BH Beier P 2007 Circuit theory predicts gene flow in plant and animal popula-tions Proceedings of the National Academy of Sciences of the United States of America10419885ndash19890 DOI 101073pnas0706568104

Morgan KM Perry CT Smithers SG Johnson JA Daniell JJ 2016 Evidence of extensivereef development and high coral cover in nearshore environments implicationsfor understanding coral adaptation in turbid settings Scientific Reports 629616DOI 101038srep29616

Tisthammer et al (2020) PeerJ DOI 107717peerj8550 2124

Nosil P Funk DJ Ortiz-Barrientos D 2009 Divergent selection and heterogeneousgenomic divergenceMolecular Ecology 18375ndash402DOI 101111j1365-294X200803946x

Orsini L Vanoverbeke J Swillen I Mergeay J DeMeester L 2013 Drivers of popula-tion genetic differentiation in the wild isolation by dispersal limitation isolationby adaptation and isolation by colonizationMolecular Ecology 225983ndash5999DOI 101111mec12561

Polato NR Concepcion GT Toonen RJ Baums IB 2010 Isolation by distance acrossthe Hawaiian Archipelago in the reef-building coral Porites lobataMolecular Ecology194661ndash4677 DOI 101111j1365-294X201004836

Presto KM Storlazzi CD Logan JB Reiss TE Rosenberger KJ 2012 Coastal circulationand potential coral-larval dispersal in Maunalua Bay Oahu Hawaii-measurementsof waves currents temperature and salinity June-2010 US Geological SurveyOpen-File Report 2012-1040

Puritz JB Toonen RJ 2011 Coastal pollution limits pelagic larval dispersal NatureCommunications 211ndash8 DOI 101038ncomms1238

R Core Team 2014 R a language and environment for statistical computing Vienna RFoundation for Statistical Computing

Richmond RH 2009HCRI Project Progress Report FY2008 Watersheds impacts oncoral reefs in Maunalua Bay Oahu Hawaii

Richmond RHWolanski E 2011 Coral research past efforts and future horizons InDubinsky Z Stambler N eds Corals reefs an ecosystem in transition DordrechtSpringer 3ndash12

Rodgers KS Jokiel PL Brown EK Hau S Sparks R 2015 Over a decade of change inspatial and temporal dynamics of Hawaiian coral reef communities 1 Pacific Science691ndash13 DOI 1029846911

Roff G Bejarano S Bozec Y-M Nugues M Steneck RS Mumby PJ 2014 Poritesand the Phoenix effect unprecedented recovery after a mass coral bleachingevent at Rangiroa Atoll French PolynesiaMarine Biology 1611385ndash1393DOI 101007s00227-014-2426-6

Rundle HD Nosil P 2005 Ecological speciation Evolution 8336ndash352DOI 101111j1461-0248200400715x

Schweinsberg M Tollrian R Lampert KP 2016 Genetic variation in the massive coralPorites lobataMarine Biology 1631ndash12 DOI 101007s00227-016-3022-8

Selkoe KA DrsquoAloia CC Crandall ED Iacchei M Liggins L Puritz JB Heyden vonder S Toonen RJ 2016 A decade of seascape genetics contributions to ba-sic and applied marine connectivityMarine Ecology Progress Series 5541ndash19DOI 103354meps11792

Shearer TL Van OppenMJH Romano SLWoumlrheide G 2002 Slow mitochondrial DNAsequence evolution in the Anthozoa (Cnidaria)Molecular Ecology 112475ndash2487DOI 101046j1365-294X200201652x

SlatkinM 1993 Isolation by distance in equilibrium and non-equilibrium populationsEvolution 47(1)264ndash279 DOI 101111j1558-56461993tb01215x

Tisthammer et al (2020) PeerJ DOI 107717peerj8550 2224

Smith TB Nemeth RS Blondeau J Calnan JM Kadison E Herzlieb S 2008b Assessingcoral reef health across onshore to offshore stress gradients in the US Virgin IslandsMarine Pollution Bulletin 561983ndash1991 DOI 101016jmarpolbul200808015

Smith LWWirshing HH Baker AC Birkeland C 2008a Environmental versus geneticinfluences on growth rates of the corals Pocillopora eydouxi and Porites lobata(Anthozoa Scleractinia) 1 Pacific Science 6257ndash69DOI 1029841534-6188(2008)62[57EVGIOG]20CO2

Stafford-SmithMG 1993 Sediment-rejection efficiency of 22 species of Australianscleractinian coralsMarine Biology 115229ndash243 DOI 101007BF00346340

Stephens M Smith NJ Donnelly P 2001 A new statistical method for haplotype recon-struction from population data American Journal of Human Genetics 68978ndash989DOI 101086319501

Storlazzi CD Field ME 2008 Winds waves tides and the resulting flow patterns andfuxes of water sediment and coral larvae off West Maui Hawaii USGS Open-FileReport 2008-1215

Storlazzi CD McManusMA Logan JB McLaughlin BE 2006 Cross-shore ve-locity shear eddies and heterogeneity in water column properties over fring-ing coral reefs West Maui Hawaii Continental Shelf Research 26401ndash421DOI 101016jcsr200512006

Storlazzi CD Presto KM Logan JB Field ME 2010 Coastal circulation and sedimentdynamics in Maunalua Bay Oahu Hawaii USGS Open-File Report 2010-1217

Thomas L KenningtonWJ Stat MWilkinson SP Kool JT Kendrick GA 2015Isolation by resistance across a complex coral reef seascape Proceedings of the RoyalSociety B Biological Sciences 28220151217 DOI 101098rspb20151217

Thompson A Schroeder T Brando VE Schaffelke B 2014 Coral community responsesto declining water quality Whitsunday Islands Great Barrier Reef Australia CoralReefs 33923ndash928 DOI 101007s00338-014-1201-y

Tisthammer KH Forsman ZH Sindorf VL Massey TL Bielecki CR Toonen RJ 2016The complete mitochondrial genome of the lobe coral Porites lobata (AnthozoaScleractinia) sequenced using ezRADMitochondrial DNA Part B 1247ndash249DOI 1010802380235920161157770

Tisthammer KH Richmond RH 2018 Corallite skeletal morphological variation inHawaiian Porites lobata Coral Reefs 65560ndash514 DOI 101007s00338-018-1670-5

Toonen RJ Grosberg RK 2011 Causes of chaos spatial and temporal genetic hetero-geneity in the intertidal anomuran crab Petrolisthes cinctipes Phylogeography andPopulation Genetics in Crustacea 75ndash107

Van Dam JW Negri AP Uthicke S Mueller JF 2011 Chemical pollution on coral reefsexposure and ecological effects In Saacutenchez-Bayo F van den Brink PJ Mann RMeds Ecological impact of toxic chemicals Bussum The Netherlands Bentham SciencePublishers 187ndash211

Vargas-Angel B 2017Maui_Baseline_report_2017 1ndash49DOI 107289V5SP-PIFSC-17-001

Tisthammer et al (2020) PeerJ DOI 107717peerj8550 2324

Veron JEN 1995 Corals in space and time the biogeography and evolution of the Sclerac-tinia Ithaca Cornell University Press

Veron JEN 2000 Corals of the world Townsville Australian Institute of Marine ScienceVeron JEN PichonM 1982 Scleractinia of Eastern Australia part IV family Poritidae

Canberra Australian National University PressVollmer SV Palumbi SR 2002Hybridization and the evolution of reef coral diversity

Science 2962023ndash2025 DOI 101126science1069524Wang IJ Bradburd GS 2014 Isolation by environmentMolecular Ecology 235649ndash5662

DOI 101111mec12938Wenger ASWilliamson DH Da Silva ET Ceccarelli DM Browne NK Petus C Devlin

MJ 2015 Effects of reduced water quality on coral reefs in and out of no-takemarine reserves Conservation Biology 30142ndash153 DOI 101111cobi12576

Williams I Vargas BWhite D Callender T 2014West Maui Wahikuli amp HonokowaiPriority Watershed Area Reef Condition Report Sceintific Report by Hawaii Divi-sion of Aquatic Resources Ridge 2 Reef Initiative NOAA Coral Reef ConservationProgram

Wolanski E Martinez JA Richmond RH 2009 Quantifying the impact of watershedurbanization on a coral reef Maunalua Bay Hawaii Estuarine Coastal and ShelfScience 84259ndash268 DOI 101016jecss200906029

Wright S 1943 Isolation by distance Genetics 28114ndash138

Tisthammer et al (2020) PeerJ DOI 107717peerj8550 2424