Genetic mosaicism and population connectivity of edge-of-range Halodule wrightii populations

Journal Pre-proof Genetic mosaicism and population connectivity of edge-of-range Halodule wrightii populations Gina Digiantonio (Conceptualization) (Validation) (Formal analysis) (Investigation)Data Curation)Writing - Original Draft)Project Administration), Linda Blum (Conceptualization) (Methodology)Writing Review and Editing) (Supervision), Karen McGlathery (Conceptualization) (Resources)Writing Review and Editing) (Supervision), Kor-jent van DijkData Curation)Formal Analysis)Writing - Review and Editing), Michelle Waycott (Resources)Writing - Review and Editing)

PII:

S0304-3770(19)30224-4

DOI:

https://doi.org/10.1016/j.aquabot.2019.103161

Reference:

AQBOT 103161

To appear in:

Aquatic Botany

Received Date:

29 January 2019

Revised Date:

25 April 2019

Accepted Date:

23 September 2019

Please cite this article as: Digiantonio G, Blum L, McGlathery K, van Dijk K-jent, Waycott M, Genetic mosaicism and population connectivity of edge-of-range Halodule wrightii populations, Aquatic Botany (2019), doi: https://doi.org/10.1016/j.aquabot.2019.103161

This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. © 2019 Published by Elsevier.

Genetic mosaicism and population connectivity of edge-of-range Halodule wrightii populations

Gina Digiantonio1*, Linda Blum1, Karen McGlathery1, Kor-jent van Dijk2, Michelle Waycott2, 3 1

Department of Environmental Sciences, University of Virginia, 291 McCormick Rd, Charlottesville, VA 22903, U.S.A. 2 School of Biological Sciences, The University of Adelaide, North Terrace, Adelaide, SA 5005, Australia. 3 State Herbarium of South Australia, Department for Environment and Water, GPO Box 1047, Adelaide, SA 5001, Australia. *

Corresponding author. [email protected]

Edge-of-range populations of Halodule wrightii were highly clonal Remote populations (North Carolina, Bermuda) likely originated from founder effects Genetic differentiation followed an isolation-by-distance relationship Putative aneuploidy was detected in 27 percent of genets Somatic mutations may enhance genetic diversity in H. wrightii clonal lineages

-p

    

ro of

Highlights

Abstract

ur

na

lP

re

The genetic population structure of Halodule wrightii at locations in Florida, North Carolina, and Bermuda was investigated using 11 polymorphic microsatellite loci on samples representing 15 sites. We measured allelic diversity and genotypic richness and determined population differentiation and gene flow using principal components and k-means population clustering. Halodule wrightii sites were highly clonal with a mean genotypic richness of 0.09. Genetic differentiation followed a statistically significant isolation-by-distance relationship. Population clustering identified two groups 1) Bermuda and Florida Bay, and 2) the Indian River Lagoon, Gulf of Mexico, and North Carolina. Results from this study indicate that vegetative growth is important for H. wrightii at multiple spatial scales and that isolated populations of H. wrightii likely originated from founder effects. In addition, many of the identified clones included samples that displayed variable copy number of some loci, suggestive that there may be an abnormal chromosome complement and/or may be indicative of aneuploidy. Given the low genotypic diversity observed overall, genetic diversity accumulated without sexual reproduction through genetic mosaicism and other derived somatic mutations will affect fitness and clonal persistence.

Jo

Keywords: seagrass; genetic diversity; microsatellites

1. Introduction

Climate change can alter the persistence and natural range of species (Pinsky et al., 2013) and is contributing to the well documented global decline in seagrass populations (Halpern et al., 2008; Waycott et al., 2009). Enhanced seagrass population resilience and associated ecosystem services have been documented for seagrasses as a result of increased genetic diversity (Reusch and Williams, 1998; Hughes and Stachowicz, 2004; Reusch et al., 2005; Hughes and Stachowicz, 2011; Reynolds et al., 2012). There has been an observed decrease in genetic diversity at edge-of-range seagrass populations (Olsen et al., 2004; Diekmann and Serrão, 2012; Sinclair et al., 2016), although not in all systems studied (Alberto et al., 2008; McMahon et al., 2017). Attributable causes of genetic diversity declines in edge-of-range

populations may result from multiple and cumulative factors including dispersal limits or barriers to connectivity (Kendrick et al., 2012; McMahon et al., 2014), extreme environmental conditions (Thomson et al., 2014), and niche specialization (Sexton et al., 2009).

-p

ro of

Halodule wrightii is a tropical/subtropical seagrass species that occupies an extensive geographic range and spans from southern Brazil to North Carolina U.S.A. in the western Atlantic Ocean (Short et al., 2010). It has been proposed that H. wrightii has the potential for range expansion northward along the U.S. Atlantic Coast due to increasing temperature and an associated dieback of the seagrass Zostera marina that dominates these ecosystems (Micheli et al., 2008). The genetic structure of H. wrightii populations has been investigated in a limited number of regions, in particular the Texas Gulf Coast, using a variety of molecular markers. One study using Random Amplified Polymorphic DNA (RAPD) markers showed each ramet (sampled at 3-m intervals in Texas and Florida) as genetically distinct (Angel, 2002). However, other studies utilizing Amplified Fragment Length Polymorphisms (AFLPs), RAPDs, and microsatellite markers detected low to moderate levels of genetic diversity along the Texas coast (Travis and Sheridan, 2006; Larkin et al., 2008, 2017). While H. wrighii along the western Gulf of Mexico coast generally show high clonality and weak population structure, these trends in genetic diversity may not be indicative of the broader H. wrightii range due to local environmental dynamics (Larkin et al., 2017). For example, lower genetic diversity was detected in samples of the seagrass Thalassia testudinum from the Low Laguna Madre, Texas compared to less peripheral regions of the species’ biogeographical range (van Dijk et al., 2018).

ur

na

lP

re

An additional factor for consideration in studies of H. wrightii genetic diversity is the variation in chromosome numbers detected for the species. In contrast to the original chromosome report of 44 (den Hartog et al., 1979), da Silva et al. (2017) detected total H. wrightii chromosome counts of 24, 38, and even 39, with 38 as the most common count. The presence of chromosome numbers not aligning to a single series suggests that this series is unstable and/or another chromosome inheritance mechanism is likely to be occurring, including the generation of abnormal chromosome number cells, i.e. aneuploid cells (e.g. da Silva et al., 2017). Indeed, laboratory based observations suggest segments of chromosomal DNA may be exchanged between neighboring cells in H. wrightii meristematic tissue during prophase, a process called cytomixis (da Silva et al., 2017). As a result of the chromosomal exchange, cells derived from cytomixis can be aneuploid, polyploid, or contain no nuclei (Mursalimov et al., 2013). Thus, an individual may contain varied genotypes (genetic mosaicism, e.g. Buss, 1985).

Jo

Genetic mosaicism is considerably more common than has been assumed in the past (PinedaKrch and Lehtilä, 2004). Mosaics have been detected in several seagrass species, often in populations where a greater reliance on clonality have been detected (Reusch and Boström, 2011; Becheler et al., 2014; Sinclair et al., 2016). Genetic mosaicism has been positively correlated with low population scale genotypic diversity (clonal richness < 0.20) in edge-ofrange populations of Z. marina (Reusch and Boström, 2011). The formation of genetic mosaics within individual organisms can now be more readily detected due to the availability of high resolution genotypic DNA based markers that can be screened from small amounts of DNA (Gajecka, 2016). The purpose of this study was to use microsatellite markers to assess the genetic diversity and regional differentiation of H. wrightii from 15 sites in Florida, North Carolina, and Bermuda. These sites expand the range of H. wrightii genetic diversity studied as previous studies of the genetic population structure did not include temperate populations in North Carolina. H.

wrightii populations in North Carolina and Bermuda are at the far northern latitude of the species’ range and therefore are considered edge-of-range for this study. The sampled populations may display previously undetected genetic diversity patterns given the correlations between low clonal richness, edge-of-range locations, and genetic mosaicism. While microsatellites are unable to provide definitive detection of chromosomal changes, they are useful for studies of the genetic structure of populations and may still detect unusual allele trends.

ro of

We hypothesized that due to the location of study sites along the northern edge of the distribution range, populations rely predominantly on asexual reproduction rather than sexual reproduction, following the edge-of-range effect observed in some other seagrasses. Genetic mosaicism was not expected due to the lack of mosaicism observed in previous genetic population studies. We also predicted that because H. wrightii occurs in isolated locations in Bermuda and North Carolina and is dioecious with negatively buoyant seeds, it will exhibit limited genetic connectivity and follow a strong isolation-by-distance relationship across our sampled range. 2. Methodology 2.1 Sample collection

na

lP

re

-p

We collected samples from 15 sites along the U.S. Atlantic coast in the Gulf of Mexico (1 site), Florida Bay (5), Indian River Lagoon (3), North Carolina (4), and Bermuda (2) (Figure 1). Sites in the Indian River Lagoon, Florida Bay, and North Carolina were established at varied spatial scales from tens to thousands of meters apart to capture spatial structure and connectivity of populations (Table A.1). At each site, seagrass was sampled from 2 or 3 (Research Dock site only) quadrats at a target distance of 50 m apart. Actual distances between quadrats at each site ranged from 35 – 200 m due to how continuous Halodule growth in the population was. Each quadrat was sectioned with dive flags into a 5-m by 5-m grid. At each quadrat, a shoot was collected every meter of the grid for a total of 72 shoots per site (Figure A.1). Exceptions to standardized sampling were made in Wabasso Causeway (Indian River Lagoon) where quadrats were sampled randomly due to seagrass patchiness and Bermuda locations that were collected in a non-quadrat format by previous researchers. Epiphytes were manually scrapped-off of shoots. The shoots were placed in desiccant and stored, dried, until the DNA was extracted following the approach of Chase and Hills (1991). 2.2 Determination of genotype

Jo

ur

To prepare samples for DNA extraction, a piece of the youngest dried tissue was placed in a 96 deep-well plate (USA Scientific, Ocala, Florida, USA), sealed with strip caps (Qiagen,Valencia, California, USA), and shipped to the University of Wisconsin-Madison Biotechnology Center DNA Sequencing Facility (Madison, Wisconsin, USA) where DNA extraction was completed. The Cetyltrimethyl Ammonium Bromide (CTAB) method, as described in Saghai-Maroof et al. (1984), was performed with minimal modification to extract DNA from 40-50 mg of tissue. Following elution, a 1.5:1 (v:v) mixture of Axygen Clean-Seq beads (Corning Life Sciences, Corning, New York, USA) to extracted DNA sample was used to clean the DNA of any inhibitory compounds. The experimenters used Quant-IT PicoGreen fluorescent dye (Thermo Fisher, Waltham, Massachusetts, USA) to quantify the DNA (Thermo Fisher). We standardized the DNA to 5 ng µL-1, except in a final rerun of samples where 9:1 and 3:1 dilutions of RNase-free water:DNA were made. Results of representative samples were verified with comparisons to

in-house extractions following the manufacturer's protocol of Qiagen DNeasy Plant Mini extraction kits. We amplified microsatellite markers with direct PCR in a 3-panel multiplex setup using 11 microsatellite primers developed by Larkin et al. (2012; personal communication) (Table A.2). Primer combinations were organized into multiplex panels using Multiplex Manager v1.2 (Holleley and Geerts, 2009). PIG-tails (5′-GTTTCT-3′) were attached to reverse primers to reduce stutters, and fluorescent tags were added to forward primers to allow for a nested Polymerase Chain Reaction (PCR) protocol (Brownstein et al., 1996; Schuelke, 2000). Each reaction well contained 1 µL of DNA template and 9 µL of master mix consisting of 0.1 µL forward and reverse primer, 5 µL TypeIt Microsatellite master mix (Qiagen), and 3.2 µL RNase-free water. PCR protocol was optimized using a touch-down approach from 60◦C to 54◦C to better target primer annealing temperatures. Amplification was confirmed with gel electrophoresis using a 1.5% agarose gel.

-p

ro of

Fragment length analysis of PCR products was conducted by the Georgia Genomics Facility (Athens, Georgia, USA) using a capillary-based 3730xl DNA analyzer (Applied Biosystems, Foster City, California, USA) with an internal GGF_LIZ500 size (custom made for GGF) standard. We processed and scored the electropherograms in Geneious R9.0.5. Some electropherograms displayed abnormal fragment length variations that were not consistent with standard diploid genotypes and these were rerun to reduce the risk of scoring PCR artifacts. 2.3 Statistical analyses

na

lP

re

The presence of abnormal fragment length variation was challenging for statistical analyses due to uncertainty in the allele assignment (allelic dosing) and uncertain ploidy levels. For example, an electropherogram displaying alleles ABC could be AABC, ABBC, ABCC, or ABCX (where X is a null allele) (Dufresne et al., 2014). Allelic dosage assignments affect allele frequency statistics, which is the basis for other genetic diversity measures. Thus, discrepancies in allelic dosage propagate throughout statistical analyses. To overcome this deficiency while preserving the maximum amount of information about the allelic composition of the populations, dosage uncertainties were incorporated into the interferences of population genetic parameters.

Jo

ur

GenoDive 2.0 vb15 (Meirmans and Van Tienderen, 2004) allows for correction of unsure allele dosage for certain allele frequency tests, population fixation indexes, AMOVA testing, and Principal Components Analysis (Meirmans, 2013). A maximum likelihood method for every locus and population, modified from De Silva et al. (2005) by Meirmans and Van Tienderen (2004), was used to resample multiple copies (> 2 per locus) of alleles into diploid identities for STRUCTURE analysis that required data to be of a single ploidy level. Resampling overestimates the number of heterozygotes and therefore was not performed for other analyses. Missing data may bias the results of Principal Components Analysis (PCA) and STRUCTURE tests by artificially grouping samples with missing data at the same loci. For this reason, the missing values (< 8%) were filled in using randomly drawn alleles from relative allelic frequencies for each population in those analyses, as recommended by Meirmans (2013). Preliminary clonal identity of all samples readily identified as diploid was calculated in GenClone 2.0 (Arnaud-Haond and Belkhir, 2006) and the probabilities of randomly drawing the genotype, ρgen, and of a clone resulting from a sexual event, ρsex, were calculated (Arnaud-Haond and Belkhir, 2006). Unknown ploidy genotypes and samples with missing values were manually assigned to clones using sorting functions in Microsoft Excel v. 2013

to identify individuals with shared multilocus genotypes. Samples that exhibited null alleles and so were not confidently assigned to a clone were genotyped again. If adequate amplification did not occur, the sample was removed from the dataset. The resulting list of potentially unique genotypes was reanalyzed to increase the dataset robustness. Final clonal identity for each population was determined in RClone (Bailleul et al., 2016). Replicate genotypes were removed from further analyses to avoid biasing allelic frequencies. Allelic frequencies and common genetic diversity parameters, including the number of alleles (N), effective number of alleles (Neff), expected subpopulation heterozygosity (Hs), and expected total heterozygosity (Ht) were calculated with correction for unsure allele dosage in GenoDive. Observed heterozygosity, inbreeding coefficients, and population divergence from Hardy-Weinberg equilibrium could not be accurately calculated with the presence of uncertain ploidy levels or putative aneuploidy.

re

-p

ro of

Partitioning of genetic diversity between and among populations was performed in an Analysis of Molecular Variance (AMOVA) using a ploidy-independent Infinite Allele Model without permutations. Matrices of population differentiation were calculated based on Nei’s Corrected Fixation Index, G’ST (Nei, 1987), the Corrected Standardized Fixation Index, G’’ST (Meirmans and Hedrick, 2011), and Jost’s D (Jost, 2008). Jost’s D population differentiation matrix was converted in a manner similar to Sundqvist et al. (2016) to estimate relative gene flow and depict population connectivity over evolutionary time periods. Bermuda sites were excluded from the analysis due to small sample size. Gene flow was plotted using the qgraph function in the qgraph package in R (Epskamp et al., 2016). A standard Mantel test was performed using Jost’s D differentiation matrix and pairwise geographic distances, excluding Bermuda sites, to test for isolation-by-distance. Geographic distances were determined from Google Earth using the shortest in-water path between sites. Principal Component Analysis (PCA) covariance matrices were generated in GenoDive and outputs were plotted in DataGraph.

ur

na

lP

An admixture model of assumed population size K was run in STRUCTURE for population sizes K=1 to K=8 (Pritchard et al., 2000). Individuals were assigned probabilistically to populations based on allelic frequencies using a Markov Chain Monte Carlo (MCMC) method with default initial values, burn-in length of 50,000 steps, run length of 250,000 steps, and 15 iterations per K. Variations of dataset and parameter set inputs were tested to compare between potentially biased runs due to resampling and missing data fills. Two datasets were run with no missing data, resampled genotypes, and with and without location set as an a-priori assumption. An additional dataset was run under the same settings but contained missing data. Resulting files were further analyzed with CLUMPAK software to visualize population structure and estimate Best K based on the ΔK method from Evanno et al. (2005) (Kopelman et al., 2015).

Jo

3. Results

The highest probability of randomly detecting the same genotype given the allelic diversity was 1:8260 in Terrapin Bay, and the highest probability of a clone resulting from a sexual event was 1:30 in St. Joseph’s Bay. Across all sites, 112 unique genotypes were identified among the 955 samples screened (Table 1, 11%). Uncertainty in ploidy level or putative aneuploidy as variable locus allele copy number (cf. da Silva et al., 2017) was detected in 30 of the 112 genets (27%). Aneuploidy or variable locus allele copy number was detected in all sites except Wabasso Causeway (Indian River Lagoon), St. Joseph’s Bay (Gulf of Mexico), and both Bermuda sites. Terrapin Bay (Florida Bay) and the Middle Marsh sites (North Carolina) had the largest proportion of samples with an abnormal number of alleles. Samples

that displayed variation in ploidy levels were generally estimated to be triploid, with a small number (< 7%) of tetraploids based on allelic frequency calculation in GenoDive. The consistency of detection of uncertain ploidy levels across the data collected among runs, repeated analysis of samples, and as many as 7 loci in an individual make it unlikely that the unusual number of alleles were due to scoring error (Figure A.2). Regardless of which quadrat the sample came from, all samples at a given site were pooled together during clone assignment because initial screening of samples showed the presence of identical clones across quadrats. Both Bermuda locations were excluded from some analyses because only a single clone was found at each site. Genotypic richness was constrained between 0 for Bermuda sites to 0.20 for Pelican Key (Table 1). While a higher proportion of sites in Bermuda and North Carolina showed extensive clonality, other regions showed greater, but variable, genotypic richness.

ro of

The number of alleles per locus ranged from 2 to 25 as compared with 2 to 8 for Syringodium filiforme (Cymodoceaceae) (Bijak et al., 2014) and 3 to 17 for Thalassia testudinum (Hydrocharitaceae) (van Dijk et al., 2007). The locus with 25 alleles commonly revealed alleles reflecting polyploidy. Per population, the average number of alleles per locus ranged from 1.2 for the Bermuda sites to 4.82 for Pelican Key (Florida Bay). In general, the effective number of alleles followed a similar pattern where Neff was highest in Florida Bay sites and lowest in Bermuda. Average total heterozygosity over all populations was 0.50.

na

lP

re

-p

Measures of population differentiation, G’ST, G’’ST, and Jost’s D (Table A.3), varied in absolute value but showed similar trends. Nei’s G’ST consistently had the lowest values and G’’ST the highest, as expected given that G’’ST is standardized to the maximum theoretical GST to account for the underestimation of G’ST in populations with more than 2 alleles. Both GST variants excluded Bermuda due to low sample size and showed St. Joseph’s Bay (Gulf of Mexico) clones as being the most differentiated from clones at other sites. The highest differentiation based on Jost’s D occurred between St. Joseph’s Bay and Bermuda, with clones at Bermuda sites as the most differentiated from clones at other sites. The connectivity analysis showed that weakest gene flow occurred between St. Joseph’s Bay and Florida Bay (Figure 1). The highest relative gene flow occurred within the Indian River Lagoon and between the Indian River Lagoon and North Carolina. Linear regression suggested that genetic distance increased by a factor of 0.178 per unit increase in geographic distance, indicating isolation-by-distance occurred (p = 0.005). Geographic distance accounted for 22% of the variance in genetic distance (R2 = 0.220).

Jo

ur

The RhoST of 0.297 (AMOVA) indicated population structuring has occurred. The majority of variance (70.3%) was attributed to the within-population variation. Principal component axes 1 and 2 represented 27.7% and 9.0% of the observed variance (Figure 2). Visual inspection suggests clustering within individual sites. The closest between-site grouping occurred between Bermuda and Florida Bay. Interpretation of the PCA results is supported by STRUCTURE analysis. STRUCTURE output for missing, non-missing, location prior and no a-prior settings indicated a best population structure at K= 2, followed by K = 3 (e.g. ΔK of 270.7 for K= 2, ΔK of 77.9 for K= 3, and ΔK of 7.9 for K= 4), where K is the number of populations (Figure A.3). There was a consistent order of magnitude difference between the ΔK values of K = 2 and K= 3 across runs. At K= 2, Florida Bay and Bermuda sites were grouped as a population and all other sites were grouped as the second population. North Carolina separated from Florida Bay at K= 3, and the Indian River Lagoon sites separated from St. Joseph’s Bay at K= 4. 4. Discussion

Our data support the hypothesis that edge-of-range populations rely predominantly on asexual reproduction, leading to local clonality, as 3 of the 4 North Carolina sites and both of the Bermuda sites demonstrated genotypic richness below the median value of all sampled locations. Nevertheless, all study sites displayed low genotypic richness (R = 0.00 – 0.20). These results suggest that widespread clonality is a common survival strategy across the different sites sampled. Larkin et al. (2017) also detected relatively low genotypic richness at a Bermuda site (0.03), but found more variable clonal richness (0.02– 0.81) in populations along the Texas coast at latitudes comparable to the Indian River Lagoon. Differences between this study and that of Larkin et al. (2017) could be due to the spatial scales at which sampling occurred, from 6-m by 30-m transects at each site by Larkin et al. (2017) compared to the smaller 5-m by 5-m quadrats of this study.

re

-p

ro of

Alternatively, differences between genotypic richness may be due to varied local conditions. For example, the Indian River Lagoon experienced algal blooms in 2015 and 2016 (including during the sample collection period in 2016) that impacted seagrass cover (Morris et al., 2018). Similarly, St. Joseph’s Bay (Gulf of Mexico) experienced an intense red tide in the fall of 2015 (sampled in 2016; R = 0.06) (Heck et al., 2018), and Florida Bay experienced an extreme salinity event and seagrass die-off in 2015 that particularly impacted the area near the Terrapin Bay site (sampled in 2015; R = 0.01) (Hall et al., 2016). While these populations are not at the edge of the species’ latitude, such stressors could cause seagrass populations to be at their edge of environmental tolerance. Repeated measurements would be needed to determine the impact of these stressful environmental events on genetic diversity, however, environmental stressors could decrease genotypic richness through the loss of seagrass genotypes that are not resilient to the stressor or through increased bed reliance on asexual reproduction.

na

lP

Low clonal richness at local scales could be due to bed recruitment that is dominated by rhizome expansion rather than seed-based recruitment. Rhizome expansion facilitates persistence under suboptimal seed recruitment conditions (Kendrick et al., 2012). Although dioecious, H. wrightii functions as a self-incompatible plant, and when pollination success is low, vegetative expansion allows plants to ‘wait’ for more optimal conditions for seed recruitment (Kudoh et al., 1999; Honnay and Jacquemyn, 2008). This could be categorized as a ‘bet-hedging’ strategy (Philippi and Seger, 1989), or waiting for windows of opportunity (sensu Eriksson and Fröborg, 1996). Such strategies are likely to be common in seagrasses that are recruitment limited such that clonality provides for long-term resilience in the face of environmental variability (Kendrick et al., 2012; Unsworth et al., 2015).

Jo

ur

The importance of vegetative growth for H. wrightii at multiple spatial scales is evident from the low genotypic richness also observed at the quadrat scale across sites in this study. We detected a common multi-locus genotype among 3 Florida Bay sites (Research Dock, Pelican Key, and Porjoe Key). This suggests the formation of a single genotype, inferred to be a clone, that has spread across hundreds to thousands of meters. Detection of such clones in seagrass systems is becoming more commonplace (Arnaud-Haond et al., 2012). Extensive clonal expansion has been shown for T. testudinum in Mexico where a clone extended nearly 250 m (van Dijk and van Tussenbroek, 2009) and over an even greater scale of 47 km in the Indian River Lagoon in Florida (Bricker et al., 2018). For distances that exceed the maximal rhizome growth extent for each species, clonal dispersal can be facilitated by vegetative fragmentation. Vegetative fragments of H. wrightii can remain viable for up to 4 weeks, providing a recruitment window for fragments to move long distances, settle, and re-root (Hall et al., 2006).

The detection of genotypes with variable allele copy number (> 2 per locus) can be explained by variable chromosome numbers and/or ploidy levels. This phenomenon suggests an unstable genome and/or another chromosome inheritance mechanism. One such mechanism potentially occurring is the generation of particular cells with abnormal chromosome numbers through cytomixis (e.g. da Silva et al., 2017). Cytomixis leads to the formation of genetic mosaics, including some aneuploid karyotypes. Typically, aneuploidy would be mistaken as PCR stutters or poor-quality traces that can occur with the genotyping methodology used in this study. However, such samples in our study lacked consistency in apparent ploidy level across many loci and electropherograms, supporting evidence for genetic mosaicism as a real biological feature of these plants and not a result of methodological problems. This interpretation of the data is consistent with laboratory observations of somatic mutations (likely resulting from cytomixis) in other H. wrightii samples (da Silva et al., 2017).

re

-p

ro of

While edge-of-range populations of H. wrightii displayed low clonal diversity as expected, this study differed by detecting and accounting for intra-individual diversity in statistical analyses. At many of the populations studied, somatic mutations replaced sexual reproduction as the prevailing force behind genetic diversity. However, the implications of genetic mosaicism on the long-term fitness of individuals and clones are unclear (Pineda-Krch and Lehtiliä, 2004). Genetic mosaicism could contribute to plant fitness and population resilience by increasing heterozygosity and genetic variation, could be neutral, or could have a detrimental influence by altering the dosages of genes involved with plant processes and pathways (de Storme and Mason, 2014). Genetic variability is an important factor for the adaptability of edge-of-range populations, though further incorporation of genetics with models, empirical studies, and other evolutionary drivers are needed to clearly demonstrate the importance of population genetic diversity to range expansion (Sexton et al., 2009).

Jo

ur

na

lP

The processes underlying the propagation of somatic mutations at the population scale require further study. Under some rhizome branching patterns, H. wrightii aneuploid-type genotypes and genetic mosaicism have the potential to revert during vegetative expansion and increase differentiation (Reusch and Boström, 2011). Genetic homogeneity can be restored within a new shoot if it is the result of a single, dividing cell line. However, this process would increase differentiation because genetic homogeneity would not extend to the original clone (Reusch and Boström, 2011). The rhizomes of H. wrightii become highly branched during vegetative expansion (Bell et al., 2014), which would facilitate such differentiation. Further study of the formation, propagation, and role somatic mutations play in clonal survival of H. wrightii edge-of-range populations will inform discussions on the evolutionary trajectory of the species. Increasingly, meiotic and mitotic processes that lead to genome duplication and chromosome change are found to be mechanisms that underlie speciation and hybridization events (de Storme and Mason, 2014). Given the lack of clonal richness within the study sites, investigation of genetic differentiation and gene flow across sites is important for understanding long-term diversity trends. A statistically significant isolation-by-distance relationship was found, indicating that geographic distance contributed to genetic differentiation. Unexpectedly, Mosquito Lagoon and Parrish Park (both northern Indian River Lagoon) populations displayed lower differentiation from a North Carolina site than suggested by their distance from the nearest inlet and the Indian River Lagoon’s flushing period of 230 days (Smith, 1993). Mosquito Lagoon and Parrish Park are likely experiencing genetic drift due to their isolation, dispersal limitation, and potential inbreeding (e.g. Kendrick et al., 2012; McMahon et al., 2014; Bricker et al., 2018). Across sites, moderate genetic structure (RhoST = 0.297) was detected

despite higher variability within populations than between populations. Intuitively this is expected as H. wrightii has negatively buoyant seeds that are released below the sediment (McMillan, 1985), thereby limiting seed dispersal. Low fixation values typically occurred within regions, which is consistent with low differentiation at local scales (< 100 km) observed across other seagrasses (Kendrick et al., 2012).

re

-p

ro of

Cluster analysis showed that clones from the Bermuda sites were most like those in Florida Bay despite the large geographic distance between them. This supports a long-distance dispersal scenario of Florida Bay material (seed or vegetative fragment) to Bermuda. This connectivity, while detectable, is rare as evidenced by the low genetic diversity among all Bermuda samples, i.e. they are clonal. In combination with the cluster results, the extreme clonality of the Bermuda populations suggests a genetic bottleneck resulting from a founder effect. Alternatively, a greater genotypic diversity may have existed historically in the Bermuda region but was lost to the population over time because some genets were unable to persist in the conditions of the edge-of-range habitat. The latter scenario would seem less probable as the lack of aneuploidy in the highly clonal Bermuda sites suggests that the establishment of H. wrightii was likely too recent for major accumulation of somatic mutations within the populations. In fact, H. wrightii is excluded in papers on the ecology of Bermuda from 1918, 1935, and 1936 (Bernatowicz, 1952). The first report of H. wrightii in Bermuda was by Ostenfeld in 1927, though it was not substantiated with specimens (Bernatowicz, 1952). During the late 1940s, H. wrightii was noted in several Bermuda bays, mentioned in a paper by Moore (n.d. ca 1947), and collected for specimen comparison (Bernatowicz, 1952). Thus, it is possible this species was not present in Bermuda until the last century, and only became established coincident with greater human mediated boat movements.

ur

na

lP

Connectivity within each of the regions is high, as would be expected. Between regions, our genetic based connectivity data supports evidence for likely dispersal pathways at the regional scale of H. wrightii in the north-eastern sector of its range. Bermuda is largely isolated from the remaining sampled regions, although is likely to have originated from the Florida Bay region. North Carolina appears to be most similar to Indian River Lagoon, which is also to be expected as it is in the direction of current movements. Florida Bay has high connectivity to the adjacent Indian River Lagoon locations. The Gulf of Mexico sample is difficult to place as few genotypes were observed, and thus further sampling would be required to better resolve the nature of connectivity between this region and the more eastern locations.

Jo

This study expanded the area investigated for H. wrightii genetic population structure along edge-of-range locations to include temperate populations and found low clonal richness with the detection of single genotypes spread hundreds to thousands of meters. Halodule wrightii populations displayed a low clonal richness, but somatic mutations contribute genetic variation within clonal lineages. Previous population studies of H. wrightii genetic diversity did not account for uncertainty in allelic dosage, likely dismissing putative aneuploidy as PCR stutters. Future studies should take the possibility of aneuploidy into consideration when performing genetic diversity analyses and investigate the potential impacts of these somatic mutations to the resilience of clonal lineages and evolutionary processes of H. wrightii. The geographically isolated populations of Bermuda and North Carolina showed few unique genotypes and may have resulted from founder effects from different source populations. Population differentiation at geographically distant locations have the potential to facilitate

evolutionary adaptations and survivorship in variable, sometimes stressful environments, supporting the concept that H. wrightii is a highly resilient species.

6. Funding Sources Funding for this research was provided by the Jones Conservation Fund and the University of Virginia. Funding sources had no involvement with the study design, collection, analysis and interpretation of data, writing of the report, or decision to submit this article for publication.

The authors claim no conflicts of interest. Author contributions

-p

ro of

Gina Digiantonio: Conceptualization, Validation, Formal analysis, Investigation, Data Curation, Writing - Original Draft, Project Administration. Linda Blum: Conceptualization, Methodology, Writing - Review and Editing, Supervision. Karen McGlathery: Conceptualization, Resources, Writing - Review and Editing, Supervision. Kor-jent van Dijk: Data Curation, Formal Analysis, Writing - Review and Editing. Michelle Waycott: Resources, Writing - Review and Editing

5. Acknowledgements

Jo

ur

na

lP

re

The authors would like to acknowledge Margot Miller for her assistance in the laboratory, Ainsley Calladine for technical support, and Alexandra Bijak and Kimberly Holzer for providing samples from Bermuda (Bermuda Dept. of Conservation Services License no. 1504-16-22 and Bermuda Department of Environment and Natural Resources permit number 130607, respectively).

7. Literature Cited Alberto, F., Massa, S., Manent, P., Diaz-Almela, E., Arnaud-Haond, S., Duarte, C.M., Serrao, E.A., 2008. Genetic differentiation and secondary contact zone in the seagrass Cymodocea nodosa across the Mediterranean–Atlantic transition region. J. Biogeogr. 35, 1279-1294. Angel, R., 2002. Genetic diversity of Halodule wrightii using random amplified polymorphic DNA. Aquat. Bot. 74, 165–174. Arnaud-Haond, S., Belkhir, K., 2006. GenClone: a computer program to analyse genotypic data, test for clonality and describe spatial clonal organization: program note. Mol. Ecol. Notes 7, 15–17. doi:10.1111/j.14718286.2006.01522.x

ro of

Arnaud-Haond, S., Duarte, C.M., Diaz-Almela, E., Marba, N., Sintes, T., Serrao, E.A., 2012. Implications of extreme life span in clonal organisms: millenary clones in meadows of the threatened seagrass Posidonia oceanica. Plos One 7, e30454 30410 pp. Bailleul, D., Stoeckel, S., Arnaud-Haond, S., 2016. RClone: a package to identify multilocus clonal lineages and handle clonal data sets in R . Methods Ecol. Evol. 7, 966–970. doi:10.1111/2041-210X.12550

-p

Becheler, R., Benkara, E., Moalic, Y., Hily, C., Arnaud-Haond, S., 2014. Scaling of processes shaping the clonal dynamics and genetic mosaic of seagrasses through temporal genetic monitoring. Heredity 112, 114-121.

re

Bell, S.S., Middlebrooks, M.L., Hall, M.O., 2014. The value of long‐term assessment of restoration: support from a seagrass investigation. Restoration Ecol. 22, 304310.

lP

Bernatowicz, A.J., 1952. Marine monocotyledonous plants of Bermuda. Bull. Mar. Sci. 2, 338-345.

na

Bijak, A.L., Dijk, K. van, Waycott, M., 2014. Development of microsatellite markers for a tropical seagrass, Syringodium filiforme (Cymodoceaceae). Appl. Plant Sci. 2, 1400082. doi:10.3732/apps.1400082

ur

Bricker, E., Calladine, A., Virnstein, R., Waycott, M., 2018. Mega clonality in an aquatic plant—a potential survival strategy in a changing environment. Front. Plant Sci. 9, 435.

Jo

Brownstein, M., Carpten, J., Smith, J., 1996. Modulation of non-templated nucleotide addition by Taq DNA polymerase: primer modifications that facilitate genotyping. Biotechniques 11, 1004-1110. Buss, L.W.,1985. The uniqueness of the individual revisited. In: Jackson, J.B.C., Buss, L.W., Cook, R.E. (Eds.), Population biology and evolution of clonal organisms. Yale University Press, New Haven, pp. 467-506. Chase, M. W., Hills, H. H., 1991. Silica gel: an ideal material for field preservation of leaf samples for DNA studies. Taxon 40, 215-220. da Silva, S.L., Magalhães, K.M., de Carvalho, R., 2017. Karyotype variations in seagrass (Halodule wrightii Ascherson—Cymodoceaceae). Aquat. Bot. 136, 5255.

De Silva, H.N., Hall, A.J., Rikkerink, E., McNeilage, M.A., Fraser, L.G., 2005. Estimation of allele frequencies in polyploids under certain patterns of inheritance. Heredity 95, 327-334. de Storme, N., Mason, A., 2014. Plant speciation through chromosome instability and ploidy change: cellular mechanisms, molecular factors and evolutionary relevance. Curr. Plant Biol. 1, 10–33. doi:10.1016/j.cpb.2014.09.002 den Hartog, C., van Loenhoud, P.J., Roelofs, J.G.M., van de Sande, J.C.P.M., 1979. Chromosome numbers of three seagrasses from the Netherlands Antilles. Aquat. Bot. 7, 267-271. Diekmann, O.E., Serrão, E.A., 2012. Range-edge genetic diversity: Locally poor extant southern patches maintain a regionally diverse hotspot in the seagrass Zostera marina. Mol. Ecol. 21, 1647-1657.

ro of

Dufresne, F., Stift, M., Vergilino, R., Mable, B.K., 2014. Recent progress and challenges in population genetics of polyploid organisms: an overview of current state-of-the-art molecular and statistical tools. Mol. Ecol. 23, 40–69. doi:10.1111/mec.12581

-p

Epskamp, S., Cramer, A.O., Waldorp, L.J., Schmittmann, V.D., Borsboom, D., 2012. qgraph: network visualizations of relationships in psychometric data. J. Stat. Softw. 48, 1-18.

re

Eriksson, O., Fröborg, H., 1996. “Windows of opportunity” for recruitment in longlived clonal plants: experimental studies of seedling establishment in Vaccinium shrubs. Canadian J. Bot. 74: 1369-1374. doi: 10.1139/b96-166.x

lP

Evanno, G., Regnaut, S., Goudet, J., 2005. Detecting the number of clusters of individuals using the software structure: a simulation study. Mol. Ecol.14, 2611–2620. doi:10.1111/j.1365-294X.2005.02553.x Gajecka, M. 2016. Unrevealed mosaicism in the next-generation sequencing era. Mol. Genet. Genomics 291, 513-530.

na

Hall, L.M., Hanisak, M.D., Virnstein, R.W., 2006. Fragments of the seagrasses Halodule wrightii and Halophila johnsonii as potential recruits in Indian River Lagoon, Florida. Mar. Ecol. Prog. Ser. 310, 109–117.

ur

Hall, M.O., Furman, B.T., Merello, M., Durako, M.J., 2016. Recurrence of Thalassia testudinum seagrass die-off in Florida Bay, USA: initial observations. Mar. Ecol. Prog. Ser. 560, 243-249.

Jo

Halpern, B.S., Walbridge, S., Selkoe, K.A., Kappel, C.V., Micheli, F., D’Agrosa, C., Bruno, J.F., Casey, K.S., Ebert, C., Fox, H.E., Fujita, R., Heinemann, D., Lenihan, H.S., Padin, E.M.P., Perry, M.T., Selig, E.R., Spalding, M., Steneck, R., Watson, R.A., 2008. A global map of human impact on marine ecosystems. Science 319, 948-952. Heck, K., Byron, D., Brucker, J., Kebart, K., Yarbro, L.A., Johnsey, E., Carlson, P., 2018. Summary report for St. Joseph Bay. In: Yarboro L.A., Carlson P.A., (eds). Seagrass integrated mapping and monitoring report 3: 3–10. doi: 10.13140/RG.2.2.12366.05445.

Holleley, C., Geerts, P., 2009. Multiplex Manager 1.0: a cross-platform computer program that plans and optimizes multiplex PCR. Biotechniques 46, 511–517. doi:10.2144/000113156 Honnay, O., Jacquemyn, H., 2008. A meta-analysis of the relation between mating system, growth form and genotypic diversity in clonal plant species. Evol. Ecol. 22, 299-312. Hughes, A.H., Stachowicz, J.J., 2004. Genetic diversity enhances the resistance of a seagrass ecosystem to disturbance. PNAS 101, 8998-9002. Hughes, A.R., Stachowicz, J.J., 2011. Seagrass genotypic diversity increases disturbance response via complementarity and dominance. J. Ecol. 99, 445-453. Jost, L., 2008. GST and its relatives do not measure differentiation. Mol. Ecol. 17, 4015-4026.

ro of

Kendrick, G.A., Waycott, M., Carruthers, T.J.B., Cambridge, M.L., Hovey, R., Krauss, S.L., Lavery, P.S., Les, D.H., Lowe, R.J., Vidal, O.M. i, Ooi, J.L.S., Orth, R.J., Rivers, D.O., Ruiz-Montoya, L., Sinclair, E.A., Statton, J., van Dijk, J.K., Verduin, J.J., 2012. The central role of dispersal in the maintenance and persistence of seagrass populations. BioScience 62, 56–65. doi:10.1525/bio.2012.62.1.10

re

-p

Kopelman, N.M., Mayzel, J., Jakobsson, M., Rosenberg, N.A., Mayrose, I., 2015. Clumpak : a program for identifying clustering modes and packaging population structure inferences across K. Mol. Ecol. Resour. 15, 1179–1191. doi:10.1111/1755-0998.12387

lP

Kudoh, H., Shibaike, H., Takasu, H., Whigham, D. F., Kawano, S., 1999. Genet structure and determinants of clonal structure in a temperate deciduous woodland herb, Uvularia perfoliata. J. Ecol. 87, 244-257. Larkin, P., Heideman, K.L., Parker, J.E., Hardegree, B., 2008. Genetic structure of Halodule wrightii populations from the Laguna Madre region in the Western Gulf of Mexico. Gulf Mexico Sci. 124-129

na

Larkin, P., Schonacher, T., Barrett, M., Paturzzio, M., 2012. Development and characterization of microsatellite markers for the seagrass Halodule wrightii. Conserv. Genet. Resour. 4, 511–513. doi:10.1007/s12686-011-9587-0

Jo

ur

Larkin, P.D., Maloney, T.J., Rubiano-Rincon, S., Barrett, M.M., 2017. A map-based approach to assessing genetic diversity, structure, and connectivity in the seagrass Halodule wrightii. Mar. Ecol. Prog. Ser. 567, 95-107. doi:10.3354/meps12037 McMahon, K., van Dijk, K.-j., Ruiz-Montoya, L., Kendrick, G.A., Krauss, S.L., Waycott, M., Verduin, J., Lowe, R., Statton, J., Brown, E., Duarte, C.M., 2014. The movement ecology of seagrasses. Proc. R. Soc. Lond. Ser. B. Biol. Sci. 281, 20140878. doi: 10.1098/rspb.2014.0878 McMahon, K.M., Evans, R.D., van Dijk, K.J., Hernawan, U., Kendrick, G.A., Lavery, P.S., Lowe, R., Puotinen, M., Waycott, M., 2017. Disturbance is an important driver of clonal richness in tropical seagrasses. Front. Plant Sci. 8, Article number 2026.

McMillan, C., 1985. The seed reserve for Halodule wrightii, Syringodium filiforme and Ruppia maritima in Laguna Madre, Texas. Contrib. Marine Sci. 28, 141-149. Meirmans, P.G., 2013. GenoDive (version 2.0 b23) manual: software for analysis of population genetic data. Universiteit van Amsterdam. Accessed from http://www.bentleydrummer.nl/software/GenoDiveManual.pdf. Meirmans, P.G., Hedrick, P.W., 2011. Assessing population structure: FST and related measures: Invited technical review. Mol. Ecol. Resour. 11, 5–18. doi:10.1111/j.1755-0998.2010.02927.x Meirmans, P.G., Van Tienderen, P.H., 2004. GENOTYPE and GENODIVE: two programs for the analysis of genetic diversity of asexual organisms. Mol. Ecol. Notes 4, 792–794. doi:10.1111/j.1471-8286.2004.00770.x

ro of

Micheli, F., Bishop, M.J., Peterson, C.H., Rivera, J., 2008. Alteration of seagrass species composition and function over two decades. Ecol. Monogr. 78, 225– 244.

Moore, H.B., n.d. Ecological guide to Bermuda inshore water. Bermuda Biol. Sta. Res.

-p

Morris L.J., Hall L.M., Chamberlain R.C., Jacoby C.A., 2018. Summary report for the northern Indian River Lagoon. In: Yarboro L.A., Carlson P.A., (eds). Seagrass integrated mapping and monitoring report 3: 3–10. doi: 10.13140/RG.2.2.12366.05445.

re

Mursalimov, S.R., Sidorchuk, Y.V., Deineko, E.V., 2013. New insights into cytomixis: specific cellular features and prevalence in higher plants. Planta 238, 415–423. doi: 10.1007/s00425-013-1914-0.x

lP

Nei, M., 1987. Molecular Evolutionary Genetics. Columbia University Press, New York.

na

Olsen, J.L., Stam, W.T., Coyer, J.A., Reusch, T.B.H., Billingham, M., Bostrom, C., Calvert, E., Christie, H., Granger, S., La Lumiere, R., Milchakova, N., OudotLe Secq, M.P., Procaccini, G., Sanjabi, B., Serrao, E., Veldsink, J., Widdicombe, S., Wyllie-Echeverria, S., 2004. North Atlantic phylogeography and large-scale population differentiation of the seagrass Zostera marina L. Mol. Ecol. 13, 1923-1941.

ur

Philippi, T., Seger, J. 1989. Hedging one’s evolutionary bets, revisited. Trends in Ecol. Evol. 4, 41-44. Pineda‐Krch, M., Lehtilä, K., 2004. Costs and benefits of genetic heterogeneity within organisms. J. Evol. Biol. 17, 1167-1177.

Jo

Pinsky, M.L., Worms, B., Fogarty, M.J., Sarmiento, J.L., Levin, S.A., 2013. Marine taxa track local climate velocities. Science 341, 1239-1242. Pritchard, J.K., Stephens, M., Donnelly, P., 2000. Inference of population structure using multilocus genotype data. Genetics 155, 945–959. Reusch, T.B.H., Boström, C., 2011. Widespread genetic mosaicism in the marine angiosperm Zostera marina is correlated with clonal reproduction. Evol. Ecol. 25, 899–913. doi:10.1007/s10682-010-9436-8 Reusch, T.B.H., Ehlers, A., Hammerli, A., Worm, B., 2005. Ecosystem recovery after climatic extremes enhanced by genotypic diversity. Proc. Natl. Acad. Sci. U.S.A. 102, 2826-2831.

Reusch, T.B.H., Williams, S.L., 1998. Variable responses of native eelgrass Zostera marina to a non-indigenous bivalve Musculista senhousia. Oecologia 113, 428441. Reynolds, L.K., McGlathery, K.J., Waycott, M., 2012. Genetic diversity enhances restoration success by augmenting ecosystem services. PLoS ONE 7, e38397. doi:10.1371/journal.pone.0038397 Saghai-Maroof, M.A., Soliman, K.M., Jorgensen, R.A., Allard, R.W., 1984. Ribosomal DNA spacer-length polymorphisms in barley: Mendelian inheritance, chromosomal location, and population dynamics. Proc. Nat. Acad. Sci. 81, 8014-8018. Schuelke, M., 2000. An economic method for the fluorescent labeling of PCR fragments. Nature Biotechnol. 18, 233–234.

ro of

Sexton, J.P., McIntyre, P.J., Angert, A.L., Rice, K.J., 2009. Evolution and ecology of species range limits. Annu. Rev. Ecol. Evol. Syst. 40, 415–436. doi:10.1146/annurev.ecolsys.110308.120317

-p

Short, F.T., Carruthers, T.J.R., van Tussenbroek, B., Zieman, J. 2010. Halodule wrightii. The IUCN Red List of Threatened Species 2010: e.T173372A7001725. doi:10.2305/IUCN.UK.20103.RLTS.T173372A7001725.en. Downloaded on 14 April 2018.

re

Sinclair, E.A., Statton, J., Hovey, R., Anthony, J.M., Dixon, K.W., Kendrick, G.A., 2016. Reproduction at the extremes: Pseudovivipary, hybridization and genetic mosaicism in Posidonia australis (Posidoniaceae). Annals Bot. 117, 237-247. Smith, N.P., 1993. Tidal and nontidal flushing of Florida’s Indian River Lagoon. Estuaries 16, 739-746.

lP

Sundqvist, L., Keenan, K., Zackrisson, M., Prodöhl, P., Kleinhans, D., 2016. Directional genetic differentiation and relative migration. Ecol. Evol. 6, 34613475.

na

Thomson, J.A., Burkholder, D.A., Heithaus, M.R., Fourqurean, J.W., Fraser, M.W., Statton, J., Kendrick, G.A., 2015. Extreme temperatures, foundation species, and abrupt ecosystem change: an example from an iconic seagrass ecosystem. Global Change Biol. 21, 1463-1474.

ur

Travis, S.E., Sheridan, P., 2006. Genetic structure of natural and restored shoalgrass Halodule wrightii populations in the NW Gulf of Mexico. Mar. Ecol. Prog. Ser. 322, 117–127.

Jo

Unsworth, R.K., Collier, C.J., Waycott, M., Mckenzie, L.J., Cullen-Unsworth, L.C., 2015. A framework for the resilience of seagrass ecosystems. Mar.Pollut. Bull. 100, 34-46. van Dijk, K.J., Bricker, E., van Tussenbroek, B.I., Waycott, M., 2018. Range-wide population genetic structure of the Caribbean marine angiosperm Thalassia testudinum. Ecol. Evol. 8, 9478-9490. van Dijk, J. K., Waycott, M., van Tussenbroek, B. I., Ouborg, J., 2007. Polymorphic microsatellite markers for the Caribbean seagrass Thalassia testudinum Banks ex König. Mol. Ecol. Notes 1, 89-91.

van Dijk, J.K., van Tussenbroek, B.I., 2010. Clonal diversity and structure related to habitat of the marine angiosperm Thalassia testudinum along the Atlantic coast of Mexico. Aquat. Bot. 92, 63–69. doi:10.1016/j.aquabot.2009.10.005

Jo

ur

na

lP

re

-p

ro of

Waycott, M., Duarte, C.M., Carruther, T.J.B., Orth, R.J., Dennison, W.C., Olyarnik, S., Calladine, A., Fourqurean, J.W., Heck, Jr., K.L., Hughes, A.R., Kendrick, G.A., Kenworthy, W.J., Short, F.T., Williams, S.L., 2009. Accelerating loss of seagrasses across the globe threatens coastal ecosystems. PNAS 106, 1237712381. doi: 10.1073/pnas.0905620106

Table 1. Genetic diversity estimates for H. wrightii based on 11 microsatellite loci where N = number of samples, G = number of genets, Gp = number of uncertain ploidy genets, R = genotypic richness, Na = average number of alleles per locus Neff = effective number of alleles, Hs = heterozygosity within populations, and Ht = total heterozygosity. Location abbreviations represent Florida Bay (FB), Indian River Lagoon (IRL), Gulf of Mexico (GOM), North Carolina (NC) and Bermuda (BRM).

Location N

G

Gp

R

Na

Neff

Hs

Ht

Research Dock

FB

102

13

1

0.12

4.46

2.81

0.63

0.63

Pelican Key

FB

70

15

4

0.20

4.82

2.93

0.59

0.59

Porjoe Key

FB

70

10

2

0.13

4.36

2.64

0.62

0.62

Duck Key

FB

69

10

1

0.13

3.55

2.12

0.48

0.48

Terrapin Bay

FB

68

2

2

0.01

2.55

1.86

0.54

0.54

Wabasso Causeway IRL

71

5

0

0.06

Parrish Park

IRL

66

10

3

0.14

Mosquito Lagoon IRL

66

13

4

0.18

St. Joseph's Bay

GOM

70

5

0

0.06

Middle Marsh 1

NC

59

6

3

Middle Marsh 2

NC

67

3

Middle Marsh 3

NC

58

6

North Carolina

NC

70

12

Bailey's Bay

BRM

19

Well Bay

BRM

30

2.38

0.60

0.60

3.64

2.50

0.52

0.52

4.55

3.42

0.64

0.64

2.55

1.91

0.38

0.38

0.09

2.55

2.17

0.57

0.57

3

0.03

2.36

2.00

0.56

0.56

5

0.09

2.91

2.56

0.57

0.57

2

0.16

3.09

1.86

0.45

0.45

1

0

0.00

1.20

1.27

0.20

0.20

1

0

0.00

1.20

1.27

0.20

0.20

30

Ave. 0.09

3.12

2.25

0.50

0.50

Jo

re

lP 112

-p

3.09

na 955

ur

Overall:

ro of

Site

f oo pr ePr na l Jo ur

Figure 1. H. wrightii sample collection locations and relative gene flow between populations. Connectivity results are based on Jost’s D. Thicker arrows depict greater gene flow scaled proportionally. Bermuda sites were not included in the gene flow analysis due to low sample size and are presented in the call-out map. Site abbreviations are as follows: St. Joseph’s Bay (SJ), Research Dock (RD), Pelican Key (PK), Porjoe Key (PJ), Duck Key (DK), Terrapin Bay (TRP), Parrish Park (PP), Wabasso Causeway (WC), Mosquito Lagoon (ML), North Carolina (NC), Middle Marsh 1 (MM1), Middle Marsh 2 (MM2), Middle Marsh 3 (MM3), Bailey’s Bay (BRM), Well Bay (WB).

f oo pr ePr na l Jo ur

Figure 2. Principal components analysis of H. wrightii populations. Different colors represent sites within locations (different shapes): Florida Bay (circle), Indian River Lagoon (square), Gulf of Mexico (star), North Carolina (triangle) and Bermuda (pentagon).

na l

Jo ur

oo

pr

e-

Pr

f

8. Appendix A

Location

Site

Abbrev.

Latitude (N)

Longitude (W)

Year Collected

Appx. distance (m) 1 to 2:

Appx. distance (m) from: RD1 N/A 130 80 550 550 5,750 5,750 10,825 10,825 30,000 30,000

Florida Bay 25°5'15.58" 25°5'17.48" 25°5'14.14" 25°5'31.31" 25°5'31.83" 25°8'9.80" 25°8'11.27" 25°10'47.11" 25°10'46.23" 25°9'52.50" 25°9'56.00"

80°27'11.12" 80°27'6.94" 80°27'12.63" 80°27'20.57" 80°27'18.72" 80°28'24.43" 80°28'25.20" 80°29'24.49" 80°29'23.51" 80°44'3.10" 80°44'3.20"

2015 2015 2015 2015 2015 2015 2015 2015 2015 2015 2015

N/A 130 80 N/A 55 N/A 50 N/A 35 N/A 100

SJ1 SJ2

29°45'59.75" 29°46'1.94"

85°24'14.47" 85°24'15.03"

2016 2016

N/A 70

-p

N/A 75 N/A 200 N/A 40

MM1_1 MM1_2 MM2_1 MM2_2 MM3_1 MM3_2 NC1 NC2

34°41'24.74" 34°41'25.55" 34°41'25.68" 34°41'26.94" 34°41'51.45" 34°41'51.82" 34°40'32.30" 34°40'31.49"

76°37'10.27" 76°37'12.24" 76°37'9.97" 76°37'8.62" 76°35'48.16" 76°35'46.26" 76°34'33.81" 76°34'32.71"

2015 2015 2015 2015 2015 2015 2016 2016

N/A 55 N/A 50 N/A 50 N/A 40

WB BRM

32°21'9.00" 32°20'16.27"

64°39'44.51" 64°43'26.05"

2013 2015

N/A N/A

80°25'40.99" 80°25'38.47" 80°47'33.24" 80°47'27.26" 80°50'12.91" 80°50'12.53"

2016 2016 2016 2016 2016 2016

re

27°45'15'56" 27°45'16.69" 28°37'30.98" 28°37'34.93" 28°52'9.17" 28°52'7.92"

lP

WC1 WC2 PP1 PP2 ML1 ML2

ur

Jo

ro of

RD1 RD2 RD3 PK1 PK2 PJ1 PJ2 DK1 DK2 TRP1 TRP2

na

Research Dock 1 Research Dock 2 Research Dock 3 Pelican Key 1 Pelican Key 2 Porjoe Key 1 Porjoe Key 2 Duck Key 1 Duck Key 2 Terrapin Bay 1 Terrapin Bay 2 Gulf of Mexico St. Joseph's Bay 1 St. Joseph's Bay 2 Indian River Lagoon Wabasso Causeway 1 Wabasso Causeway 2 Parrish Park 1 Parrish Park 2 Mosquito Lagoon 1 Mosquito Lagoon 2 North Carolina Middle Marsh 1 Q1 Middle Marsh 1 Q2 Middle Marsh 2 Q1 Middle Marsh 2 Q2 Middle Marsh 3 Q1 Middle Marsh 3 Q2 North Carolina 1 North Carolina 2 Bermuda Well Bay Bailey's Bay

Table A.1. H. wrightii sampling locations coordinates, collection dates, and distances. Refer to Figure 1 for a detailed site map. Distance gradients between sites were established for Florida Bay, the Indian River Lagoon, and North Carolina (far right column).

WC1 N/A 45 100,000 100,000 135,000 135,000 MM1_1 N/A 55 30 80 2,250 2,300 4,300 4,340

JN615001

HW190b*

KT002048

HW196

JN615002

HW200

JN615003

HW208

KT278676

HW212

JN615005

HW214

JN615006

HW222* HW228* HW232*

KT002049 KT002050 KT002051

f

HW190

239-257

Multiplex 2

Dye

1

PET

oo

JN615000

ATG

pr

HW188

Range 151-229

VIC

CA

113-161

3

FAM

GA

N/A

1

FAM

GA

262-278

1

FAM

CAT

183-192

1

NED

TAGA

232-268

2

NED

CT

191-205

3

NED

CAT

284-305

3

VIC

CA

212-244

2

PET

TAGA

N/A

1

VIC

CAT

262-304

2

FAM

CT

265-267

3

PET

e-

JN614999

Repeat CA

Pr

HW180

Primer sequence (5'-3') F: AGCACTCGCTTACTCCAACAC R: TCCCATTCTTTAGGTTCAACG F: GTGGAGGCCGAACTGTATCT R: CGACCTTCATCCTAATCATCG F: ACCTTCATAAATGGCAACTTG R:CAACTTGGTTCTGGTAGTCATC F: ATGACGAATCCCGAGGTAT R: CTCACCCACGTTAAAGCACAAT F: ATGACGAATCCCGAGGTAT R: CACCCACGTTAAAGCACAAT F: ACAACCTAGATCATCCTCACAC R: AGCAGGAAGTCAAGAGATAGG F: ACAACCTAGATCATCCTCACAC R: AGCAGGAAGTCAAGAGATAGG F: TGCCTTTCCCAACTTTTC R: CTAGGGGTGCTTATGTAGGGT F: ATGGATGTTCATTGAGTTTGAC R: CAAGGCTAAGGTAGTGGACC F: TCCTCTATCAATGGGATTTAGA R: GGGTGGCTATGTATCGA F: CCAGCAACAAGACAAATGTAT R: CTATAAGGATTAGGACAAGCACAC F: AAGACGGCATTGGAAAATAAG R: CTGGTATCATCGGAAGCACTGT F: AGCACCCTTCATTCCAAC R: CTCTGCCAATCTTCTTCTTCTACA

na l

Genbank JN609256

Jo ur

Locus HW166

Table A.2. Microsatellite primers for H. wrightii population study. Primer development is described in Larkin et al., 2012. Primer sequences with asterisks were obtained through a personal communication with Larkin.

St. Joseph's Bay (SJ)

NC

Middle Marsh 1 (MM1) Middle Marsh 2 (MM2) Middle Marsh 3 (MM3) North Carolina (NC)

0.00 0.01 0.13 0.24

0.30

0.33

0.40 0.21 0.64 0.27 0.29 0.34 0.56 0.36 0.36

0.43 0.26 0.66 0.31 0.30 0.39 0.59 0.25 0.25

Bailey's Bay (BRM) Well Bay (WB)

RD

PK

pr

0.00 -0.01 0.05 0.10 0.26

Jo ur

GOM

BRM

oo

f IRL

Research Dock (RD) Pelican Key (PK) Porjoe Key (PJ) Duck Key (DK) Terrapin Bay (TRP) Wabasso Causeway (WC) Parrish Park (PP) Mosquito Lagoon (ML)

e-

FB

0.00 0.20 0.15

0.00 0.45

0.00

0.30

0.50

0.22

0.00

0.34 0.18 0.58 0.27 0.26 0.35 0.51 0.26 0.26

0.63 0.39 0.85 0.48 0.44 0.54 0.70 0.15 0.15

0.33 0.24 0.42 0.32 0.33 0.40 0.50 0.49 0.49

0.05 0.10 0.40 0.17 0.07 0.15 0.21 0.74 0.74

0.00 0.12 0.25 0.14 0.14 0.15 0.28 0.82 0.82

0.00 0.34 0.13 0.14 0.17 0.32 0.66 0.66

0.00 0.38 0.39 0.35 0.46 1.07 1.07

0.00 -0.11 -0.09 0.18 0.69 0.69

0.00 -0.05 0.11 0.66 0.66

0.00 0.12 0.76 0.76

0.00 0.85 0.85

0.00 0.00

0.00

PJ

DK

TRP

WC

PP

ML

SJ

MM1

MM2

MM3

NC

BRM

WB

Pr

0 < x < 0.05 0.05 - 0.09 0.1 - 0.25 0.25 < x

Very little differentiation Little differentiation Moderate differentiation Strong differentiation Very great differentiation

na l

x≤0

Table A.3. Pairwise genetic differentiation between populations using Jost’s D. Abbreviations along x-axis refer to populations listed in the first column.

ro of -p re

Jo

ur

na

lP

Figure A.1. Schematic of sampling protocol. The 5 shaded circles in this schematic represent the sampling sites in Florida Bay, which were located along a distance gradient. At each site two 5-m by 5-m quadrats, represented by squares, were sampled at known distances. A shoot was collected at each meter of the quadrats, shown by the hexagons.

ro of

A

Jo

ur

na

lP

re

-p

B

Figure A.2. Scoring examples of H. wrightii diploid and aneuploid samples. Both panels show amplification of 5 loci (colored above in blue, yellow, green, and red) form Multiplex 1. Panel A shows homozygous (1 peak per locus) and heterozygous (2 peaks per locus) diploid samples from Parrish Park. Panel B shows samples from Middle Marsh with diploid, triploid, and tetraploid genotypes.

ro of -p re

Jo

ur

na

lP

Figure A.3. STRUCTURE assignment of populations for K =2 to K =4.