Published October 24, 2019

RESEARCH Exploring the Genetic Diversity of Wild Populations in the Upper Midwestern United States

Lorraine Rodríguez-Bonilla,* Fabian Rodríguez Bonilla, Daniel Matusinec, Eric Wiesman, Sean D. Schoville, Amaya Atucha, and Juan Zalapa*

L. Rodríguez-Bonilla, A. Atucha, and J. Zalapa, Dep. of Horticulture, ABSTRACT Univ. of Wisconsin, Madison, WI 53706; F.R. Bonilla, Dep. of Biology, breeding continuously evolves to satisfy Univ. of Puerto Rico Mayaguez, Mayaguez, Puerto Rico 00680; D. the needs of the growing population, but in Matusinec, Dep. of Atmospheric and Oceanic Sciences and Political many crops, it has resulted in a decline in the Science, Univ. of Wisconsin, Madison, WI 53706; E. Wiesman and genetic diversity available. Therefore, increasing J. Zalapa, USDA-ARS, Vegetable Crops Research Unit, Univ. of knowledge of the range, genetic relationships, Wisconsin, Madison, WI 53706; S.D. Schoville, Dep. of Entomology, and diversity among crop wild relatives (CWR) Univ. of Wisconsin, Madison, WI 53706. Received 5 June 2019. Accepted is essential for the efficient use of available 30 Sept. 2019. *Corresponding authors ([email protected]; germplasm in breeding programs. In cran- [email protected]). Assigned to Associate Editor Joseph Robins. berries ( macrocarpon Ait.), most Abbreviations: CTAB, cetyltrimethylammonium bromide; CWR, cultivars share the same genetic background crop wild relative(s); EDTA, ethylenediaminetetraacetic acid; PCA, based on only a few wild selections. This limits principal component analysis; PCoA, principal coordinate analysis; the breeding pool for selection to support the PC, principal component; PCR, polymerase chain reaction; RAPD, cranberry industry. Therefore, we studied 36 random amplified polymorphic DNA; SSR, simple sequence repeat. wild populations of V. macrocarpon and V. oxycoccos L. across Wisconsin and Minne- lant breeding is crucial for food production and food sota using 32 simple sequence repeat (SSR) security; however, multiple cycles of selection can result in markers. We found high levels of heterozygosity P in both species, despite previous molecular the loss of genetic diversity (Cheema, 2018). This genetic diver- markers studies revealing low genetic varia- sity loss leads to crop homogeneity, which makes food systems tion. In V. macrocarpon, a total of 294 alleles vulnerable to biotic and abiotic stresses, as has been the case of and moderate to high levels of heterozygosity many crops, such as bananas (Musa acuminata Colla, M. balbisiana

(observed [HO] = 0.51, total [HT] = 0.66) were Colla) and potatoes (Solanum tuberosum L.) (Ullstrup, 1972; Rowe, found. As expected for outcrossed polyploid 1984; National Research Council, 1993; Dempewolf et al., 2017).

species (4x), higher levels of heterozygosity (HO A way to reduce homogeneity is to find, characterize, and use = 0.81, HT = 0.83) were found in V. oxycoccos wild plant genetic resources as a source of traits for crop improve- than in V. macrocarpon. A comparison between ment (Brozynska et al., 2016). Crop wild relatives (CWR) have wild V. macrocarpon with a group of cultivated provided wild germplasm with many valuable traits for breeding and experimental hybrids found a distinct sepa- cultivated crops, such as resistance to pests, as is the case of resis- ration between both groups, suggesting there is tance to rice blast (Pyricularia grisea) (Liu et al., 2002) and resistance diversity in the wild that needs to be explored and incorporated into breeding programs. The to wheat blast (Magnaporthe oryzae Triticum pathotype) (Anh et results of our studies are the first to explore in al., 2018). Crop wild relatives also possess traits for tolerance to depth the genetic diversity of wild cranberry abiotic stress, such as salt tolerance from soybean [Glycine max (L.) populations in the upper midwestern United Merr.] to tobacco (Nicotiana tabacum L.) (Ji et al., 2010) and cold States and provide novel information to support in situ conservation efforts to protect CWR of Published in Crop Sci. 59:2413–2428 (2019). one of the few North American native crops. doi: 10.2135/cropsci2019.06.0367

© 2019 The Author(s). This is an open access article distributed under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). crop science, vol. 59, november–december 2019 www.crops.org 2413 sweetening resistance in potato (Hamernik et al., 2009), MATERIALS AND METHODS and traits influencing yield and quality such as in tomato Plant Materials (Solanum lycopersicum L.) (Tanksley et al., 1996; Brozynska Plant material of 572 samples of wild (putatively et al., 2016; Zhang et al., 2017). Therefore, exploring and collected as V. macrocarpon or V. oxycoccos) was obtained from 36 understanding the genetic diversity of CWR of locations across Wisconsin (22) and Minnesota (14) (Table 1). native to North America can help us prepare to face future The locations visited were public lands, Department of agricultural challenges and help boost local production. Natural Resources State Natural Areas, State Scientific Areas, In the United States, very few species used for food or Forest Service lands. Permits were obtained to sample and production are native to North America, and even fewer collect in locations in both states. Using satellite imagery in Google Earth Pro and data provided by the Department of have CWR in nature (Khoury et al., 2013). The American Natural Resources of both states, we located potential collec- cranberry (Vaccinium macrocarpon Ait.) is a diploid, woody tion sites. Prior to visiting the populations, satellite imagery perennial species native to North America with vast amounts was surveyed to define the area of collection to design the of unexplored wild populations and wild relatives sharing best strategy to sample the overall genetic diversity at each site overlapping ranges and environments (Darrow and Camp, while avoiding clones. In the sites, coordinates were logged per 1945; Vander Kloet, 1988). Cranberry domestication and sample, cuttings of a single vine ?20 cm long were obtained, cultivation began in the 1800s with several documented and samples were collected at least 5 m apart to ensure that wild selections (Chandler and Demoranville, 1958; Dana, plant tissue collected for each sample came from a single plant 1983; Eck, 1990). Most cultivated cranberries share a similar and/or genotype. Plant tissue was stored in a bag with wet genetic background; in fact, all hybrid cultivars released in towels, labeled, and kept in a cold bag. A total of 20 plants were the history of cranberry breeding have been derived from collected by location when possible based on the geography ‘Early Black’, ‘Howes’, ‘McFarlin’, and ‘Searles’ (the “Big and environmental conditions. Leaf material was then stored at −4°C for DNA extractions. Plant material from cultivated Four”) and a few other native selections (Eck 1990). V. macrocarpon was obtained from the USDA-ARS, Vegetable Vaccinium oxycoccos L. is mostly a tetraploid species, Crop Research Unit, Cranberry Genetics and Genomics which grows as a small evergreen perennial vine that Laboratory at the University of Wisconsin–Madison. The produces small overwintering berries that have a similar germplasm collected in this study will be kept ex situ and flavor profile to cranberries and exhibit a superior anti- housed in the USDA-ARS, Vegetable Crop Research Unit oxidant profile (Vander Kloet, 1988; Jacquemart, 1997; in Madison, WI, and selected accessions will be added to the Česonienė et al., 2011; Brown et al., 2012). As one of the National Clonal Germplasm Repository collection. V. macrocarpon wild relatives in the continental United States, V. oxycoccos provides great potential for developing DNA Extraction interspecific hybrids to transfer desirable traits into cran- Genomic DNA extractions were performed using leaf tissue per berry, such as tolerance or resistance to abiotic stressors. a cetyltrimethylammonium bromide (CTAB) method (Doyle Because V. oxycoccos has a circumboreal distribution and is and Doyle 1987) modified in our laboratory to isolate high- quality, clean DNA. A total of 10–20 mg of frozen tissue was adapted to high latitudes and harsh winters (Areškevičiūtė et al., 2006), the species could be an excellent source of ground by hand using liquid N and transferred to a 2.0-ml tube, and 700 ml of 2% CTAB extraction buffer (20 mM ethylenedi- cold hardiness genes, as is the case of other wild relatives aminetetraacetic acid [EDTA], 0.1 M Tris-HCl pH 8.0, 1.4 M to crops such as rice (Oryza sativa L.) (Shakiba et al., 2017) NaCl, 2% CTAB) was added and mixed by inversion. The and potato (Hamernik et al., 2009). solution was then incubated at 65°C for 45 min. After incu- Therefore, exploring, studying, and using wild cran- bation, 400 ml of a chloroform-isoamyl alcohol (24:1) solution berry diversity in breeding is crucial to the future of was added to the tubes and gently mixed by inversion. Samples cranberry breeding and production. To meet emerging were then centrifuged for 5 min at 14,000 rpm, 500 ml of the challenges, there is a need to identify and transfer needed supernatant was transferred to a fresh 1.5-ml tube containing traits into cultivated cranberries, as well as develop hybrids 50 ml of 10% CTAB buffer, and the sample was mixed by inver- with novel traits (Vorsa et al., 2008). In the United States, sion. A total of 750 ml of cold isopropanol (100%) was added, little has been done to analyze and understand the wild and then samples were incubated from 2–48 h at a tempera- genetic diversity of native crops such as V. macrocarpon and ture of −20°C. Samples were gently mixed by inversion and V. oxycoccos. This study represents the most comprehensive centrifuged at 14,000 rpm for 20 min. After centrifugation, the supernatant was discarded, and the resulting pellet was air dried analysis of wild populations of cranberries in Wisconsin for 5 min. The pellet was then washed with 700 ml of cold 70% and the first insight into the diversity present in Minne- ethanol to clean the DNA. The solution was then vortexed and sota. Knowledge about the distribution of genetic diversity centrifuged at 14,000 rpm for 4 min, the ethanol was discarded, in the vastly unexplored wild populations in Wisconsin and the pellet was air dried for 24 h. The DNA was then resus- and Minnesota will allow us to identify valuable diversity pended in 100 ml TE 10:1 buffer (10 mM Tris-HCl pH 8.0, pockets to collect, preserve, and use this germplasm in 1 mM EDTA pH 8.0) plus 5 ml of ribonuclease (RNase 10 mg cranberry breeding programs. ml−1) in each tube and was incubated at 37°C for 2 h prior

2414 www.crops.org crop science, vol. 59, november–december 2019 to storage at −20°C. The DNA was then quantified using a in multiple studies, and were the best markers among 697 SSRs Nanodrop ND-1000 spectrophotometer. Samples were then tested across different cranberry samples and other Vaccinium diluted to 10 ng ml−1 with deionized distilled water for simple species (Schlautman et al., 2015a, 2015b, 2017, 2018; Rodri- sequence repeat (SSR) amplifications. guez-Bonilla et al., 2019). Since all the markers used in this study were previously used in cranberry linkage mapping, the SSR Markers and Multiplex PCR Conditions markers have been proven to segregate in a Mendelian fashion A total of 32 markers (supplemental materials) previously (Schlautman et al., 2015a, 2015b, 2017, 2018). The 32 markers designed and assessed for transferability between V. macrocarpon are evenly distributed across the cranberry genome, and at and V. oxycoccos were used in this study (Schlautman et al., least two markers represent each of the 12 cultivated cranberry 2015a, 2015b; Rodriguez-Bonilla et al., 2019). Since we are linkage groups. For the purpose of simplicity, SSR fragments amplifying the markers now in two closely related cranberry will hereafter be referred to as alleles in V. oxycoccos. The multi- species, one of the disadvantages with SSRs is that if a polymor- plex panels were designed by developing SSR primer pairs. phism is obtained it may not necessarily be due to corresponding These pairs were clustered into combinations of one to three homologous chromosomes, particularly when working with markers, considering nonoverlapping fragment (i.e., alleles) polyploids, but due to nonspecific amplification of nonhomolo- ranges and annealing temperatures based on previous studies gous chromosomes. However, all the 32 markers used in this (Georgi et al., 2012; Zhu et al., 2012; Schlautman et al., 2015a, study have proven consistent, with clear amplification patterns 2015b, 2017, 2018; Rodriguez-Bonilla et al., 2019). To boost

Table 1. Cranberry (Vaccinium macrocarpon and V. oxycoccos) populations and individuals collected in Wisconsin and Minnesota. Population Population name code N Species Latitude Longitude State ° N ° W Botany Bog BB 20 V. oxycoccos 47.36802402 93.60217603 MN Big Lake Minnesota BLM 13 V. oxycoccos 47.510 0270 4 94.59074297 MN Beltrami Road BRR 20 V. oxycoccos 48.26355502 94.10271504 MN Beltrami Road M BRRM 20 V. oxycoccos 48.30934301 94.24009804 MN Chippewa National Forest MN Chip 20 V. macrocarpon 47.6 58 51497 94.47189102 MN Hole in the Bog HB 19 V. oxycoccos 47.31079901 94.21872402 MN Iron Springs Bog ISB 3 V. oxycoccos 47.253 678 01 95.24143497 MN Lova Lake LL 15 V. macrocarpon/V. oxycoccos 46.57545104 94.55575397 MN Long Lake B LOLB 20 V. oxycoccos 46.650917 93.46198099 MN Long Lake M LOLM 20 V. macrocarpon 46.647796 93.46040502 MN Little Rice Lake LRL 20 V. macrocarpon/V. oxycoccos 47.70 82710 3 92.439135 MN Pennington Bog PB 20 V. oxycoccos 47.49 93 6 501 94.47868899 MN River Road I PRF 20 V. oxycoccos 48.42440198 94.90923098 MN River Road II TR 21 V. oxycoccos 48.42439402 94.89678302 MN Bear Lake BL 18 V. macrocarpon/V. oxycoccos 45.609398 91.81301 WI Chippewa River† CHPR 1 V. oxycoccos 46.00559 90.8275 WI Flambeau Forest FBF 7 V. oxycoccos 45.69066002 90.77997297 WI Frog Lake FRLK 16 V. macrocarpon/V. oxycoccos 46.12128 90.0073 WI Goblers Lake GL 18 V. oxycoccos 45.6328 89.8744 WI Gypsy Lake R GYPL 33 V. macrocarpon/V. oxycoccos 45.80585 89.49159499 WI Hayward† HW 1 V. macrocarpon 46.05793 91.7306 WI Jaquet Lake JL 12 V. oxycoccos 45.793367 88.625817 WI Kelley Lynn Bog KLB 21 V. oxycoccos 45.45832 89.2752 WI Kemp Road KMP 12 V. oxycoccos 45.845886 89.662259 WI Lincoln/Oneida Locp 30 V. oxycoccos 45.46933 89.3226 WI Land O’ Lakes LOL 3 V. oxycoccos 46.1414 89.4917 WI Manitowish Waters MWB 16 V. oxycoccos 46.13512 89.7373 WI N-Bradley† NB 1 V. oxycoccos 45.5397 89.814333 WI Powell Marsh State Natural Area PMS 20 V. oxycoccos 46.09658 89.909 WI Powell Marsh Wildlife Refuge PMW 23 V. macrocarpon/V. oxycoccos 46.08051 89.8815 WI Spread Eagle Barrens SEB 17 V. macrocarpon/V. oxycoccos 45.86595 88.18705 WI Shelp Lake SHPL 11 V. macrocarpon/V. oxycoccos 45.776667 89.020617 WI Swamp Road SR 17 V. oxycoccos 45.406667 89.796233 WI Wabaso Lake WABL 39 V. macrocarpon/V. oxycoccos 45.9726 90.0009 WI Winchester† WIN 2 V. oxycoccos 46.19062 89.9051 WI † Populations of only one to two were not analyzed but are shown for location and collection purposes. crop science, vol. 59, november–december 2019 www.crops.org 2415 the potential of our combinations and maximize resources Jost differentiation (D_est), and inbreeding coefficient (Gis) for our fragment analysis, each reaction was combined with were obtained from the software GenoDive (Meirmans and Van four polymerase chain reaction (PCR) multiplex reactions Tienderen, 2004). For the dendrograms, the R stats package and a fluorescent dye (M13-FAM-, HEX-, NED-, or PET- (R Core Team, 2018) was used to obtain Euclidean genetic labeled primers). The forward primers were designed with the distance and clustering. The dendrograms were then visualized M13 sequence (5-CACGTTGTAAAACGAC-3) for fluores- using the packages dendextend (Galili, 2015) and circlize (Gu cent labeling of PCR products (Schuelke 2000), and the PIG et al., 2014). sequence (5-GTTTCTT-3) was attached to the reverse primers to promote full adenylation of SSR fragments during PCR (Brownstein et al., 1996). The development of primers with an RESULTS attached M13 sequence and a fluorescently labeled M13 tail is Genetic Relationships of Wild Populations more cost effective than using direct fluorescent primer labeling of Cranberries (Guichoux et al., 2011). By not using this direct fluorescence, Vaccinium macrocarpon and V. oxycoccos are sympatric and we created any marker combinations and indirectly label them can occur at the same time in the same space (Vander with any fluorescence needed. The PCR contained 3.5 ml of Kloet 1983). Therefore, we found nine populations in 1´ Jumpstart RedTaq Ready Mix (Sigma), 1 ml of molecular- −1 which both species were present throughout our sampling. grade H2O, 2.0 ml of 10 ng ml DNA, 0.5 ml of 5 mM forward primer, 0.5 ml of 50 mM reverse primer, and 0.5 ml of 0.5 mM Although all samples were putatively classified into species M13-FAM, M13-HEX, M13-NED, and M13-PET primer. in the field based on leaf morphology, we determined The 1.5 ml of forward and reverse primer was divided by the species membership, as V. macrocarpon and as V. oxycoccos, number of multiplexed SSRs (i.e., 1 ml or 0.75 ml of forward using genetic relationships based on 32 SSR markers. A and reverse primer mix from each SSR were added when three dendrogram based on Euclidean genetic distance was or four markers were multiplexed together, respectively). Ther- constructed (Figure 1A), and we observed a clear differ- mocycling conditions consisted of 94°C for 3 min; 33 cycles entiation between samples identified asV. macrocarpon of 94°C 15 s, 55°C for 1 min, and 72°C for 2 min; 72°C for and samples identified as V. oxycoccos. To further confirm 30 min; and 4°C thereafter. The PCR products were visual- our results, we performed a PCA and again observed a ized on a 1% agarose gel, and then 1–3 ml of multiplexed PCR clear differentiation between the two species. Using these product from each of the four M13 dyes were pooled using the results, we divided our data into two main groups (V. DY632 ladder from BioVentures. A total of 25 ml in 1500 ml of macrocarpon and V. oxycoccos individuals) to further study formamide (Hi-Di Formamide from Life Technologies) were added per plate of 96 wells. The poolplexed mixture was sent to and obtain diversity statistics by species. the University of Wisconsin–Madison Biotechnology Center DNA sequencing facility for fragment analysis using an ABI Overall Genetic Diversity of Wild Cranberries 3730 fluorescent sequencer (Applied Biosystems). Allele geno- When comparing the two species (Table 2), we observed typing was performed using GeneMarker 2.63 (SoftGenetics). that V. oxycoccos had the highest number of mean alleles per locus (28.18), private alleles (583), and heterozygosity

Data Analysis indexes (HO, heterozygosity within populations [HS], The allelic information obtained from the genotyping was and total heterozygosity [HT]) with values close to 0.81. formatted as a GenAlEx (Peakall and Smouse, 2012) input file. Vaccinium macrocarpon had a mean of 8.87 alleles per locus of, This file was then converted to a geneclone object to run in the 25 private alleles, and heterozygosity values (HO, HS, and R statistical software (R Core Team, 2018) package Popula- H ) ranging from 0.50–0.64. Both groups had low levels tion Genetics in R (poppr) (Kamvar et al., 2014). Using poppr, T of inbreeding demonstrated by the low Gis values of 0.00 we estimated the observed number of private alleles. For the for V. oxycoccos and −0.03 for V. macrocarpon. In terms of the principal component analysis (PCA), the packages ade4 (Dray level of fixation, which measures population differentiation and Dufour, 2007) and adegenet (Jombart, 2008) were used based on the missing allele replacement method, which replaces caused by genetic structure, both groups had low values, missing data with the mean allele frequencies. The package with V. macrocarpon having slightly higher values for Gst factoextra (Kassambara and Mundt, 2016) was used to visu- (0.22), G¢st(Nei) (0.24), G¢st(Hed) (0.48), and D_est (0.32). alize the PCA. The packages magrittr (Bache and Wickham, 2014), pegas (Paradis, 2010), and ape (Paradis et al., 2004) were Genetic Diversity in Wild also used to visualize the data. For the genetic distance-based Vaccinium macrocarpon principal coordinate analysis (PCoA), the package polysat was In V. macrocarpon, the total number of alleles was 284. All used to calculate a distance matrix based on Bruvo’s genetic the markers were polymorphic, and the number of alleles distance (Bruvo et al., 2004; Clark and Jasieniuk, 2011), and per locus ranged from 2–18 with a mean value of 8.87 package ggfortify (Tang et al., 2016) was used to visualize the (Fig. 2A). The observed heterozygosity per locus ranged first two coordinates of the PCoA. Calculations of genetic from 0.08–0.97 with a mean value of 0.51 (Fig. 2B). The diversity statistics such as number of alleles, observed heterozy- total heterozygosity per locus values ranged from 0.08– gosity (H ), fixation index G( ), Nei corrected fixation index O st 0.89 with a mean value of 0.64 (Fig. 2C). The inbreeding [G¢st(Nei)], Hedrick standardized fixation index G[ ¢st(Hed)],

2416 www.crops.org crop science, vol. 59, november–december 2019 Fig. 1. Euclidean distance dendrogram and principal component analysis (PCA) of cranberry (Vaccinium macrocarpon and V. oxycoccos) wild accessions. PC, principal component. per locus coefficient ranged from −0.42–0.60 with a mean spread of the individuals as suggested by the levels of value of −0.01 (Fig. 2D). The values of heterozygosity heterozygosity in Table 3. within populations per locus ranged from 0.05–0.77 and a mean value of 0.50 (data not shown). The populations Genetic Diversity in Wild Vaccinium oxycoccos with the highest levels of genetic diversity were FRLK, In V. oxycoccos, the total number of alleles was 902. All the GYPL, and PMW with values of 0.58 (Table 3). markers were polymorphic, and the number of alleles per locus A PCoA based on Bruvo genetic distance (Fig. 3) ranged from 3 to 56 with a mean value of 28.18 (Fig. 4A). showed a clear separation between V. macrocarpon indi- The observed heterozygosity (HO) per locus ranged from viduals from Minnesota (blue) and Wisconsin (red), with 0.41–0.98 with a mean value of 0.81 (Fig. 4B). The total Principal Component 1 (PC1) explaining 3.22% and heterozygosity (HT) values ranged from 0.45–0.95 with a

Principal Component 2 (PC2) explaining 2.34% of the mean value of 0.83 (Fig. 4C). The inbreeding coefficientG ( is) variation. In both populations, we saw a wide genetic ranged from −0.34–0.40 with a mean value of 0.08 (Fig. 4D).

Table 2. Overall genetic diversity statistics for two cranberry (Vaccinium macrocarpon and V. oxycoccos) species. Statistic V. oxycoccos V. macrocarpon Definition N 433 139 Total individuals per species NA 902 284 Total number of alleles Num 28.18 8.87 Mean number of alleles PA 583 25 Private alleles

HO 0.81 0.51 Observed heterozygosity

HS 0.81 0.49 Heterozygosity within populations

HT 0.83 0.64 Total heterozygosity

Gis 0.00 −0.03 Inbreeding coefficient

Gst 0.02 0.22 Fixation index 0.02 0.24 Nei, corrected fixation index G¢st(Nei) 0.12 0.48 Hedrick, standardized fixation index G¢st(Hed)

D_est 0.10 0.32 Jost, differentiation crop science, vol. 59, november–december 2019 www.crops.org 2417 Fig. 2. Genetic diversity statistics per locus for Vaccinium macrocarpon.

Table 3. Genetic diversity statistics per population of Vaccinium macrocarpon. Populations with only one to two samples after individuals were sorted into species were not analyzed.

Population N† Num‡ HO§ HS¶ HT# Gis†† BL 13 2.28 0.30 0.29 0.29 −0.01 Chip 20 2.50 0.51 0.34 0.34 −0.49 GYPL 13 3.25 0.58 0.52 0.52 −0.10 LL 7 3.41 0.55 0.59 0.59 0.07 LOLM 20 2.38 0.43 0.39 0.39 −0.09 LRL 17 3.13 0.48 0.52 0.52 0.07 PMW 6 2.25 0.58 0.49 0.49 −0.18 WABL 37 3.31 0.55 0.53 0.53 −0.05 † Number of samples.

‡ Mean number of alleles.

§ Observed heterozygosity.

¶ Heterozygosity within populations.

# Total heterozygosity.

†† Inbreeding coefficient.

2418 www.crops.org crop science, vol. 59, november–december 2019 Fig. 3. Principal coordinate analysis of Vaccinium macrocarpon based on Bruvo genetic distance.

The values of heterozygosity within populations ranged from the United States, we found high levels of heterozygosity 0.48–0.94 and had a mean value of 0.81 (data not shown). and mean allele numbers (Table 5). We found higher levels When looking at the genetic diversity by population in V. of observed heterozygosity and lower inbreeding index oxycoccos, most populations exhibited very high levels of genetic for wild populations (HO = 0.99, Gis = −0.31) than for diversity, averaging 0.80, when compared with V. macrocarpon cultivated cranberries (HO = 0.69, Gis = 0.15). Although (Table 4). In the PCoA of V. oxycoccos accessions (Fig. 5), we cultivated accessions had a higher number of alleles than observed some spatial separation by state, although V. oxycoccos the wild accessions overall, the majority of these alleles showed less differentiation than V. macrocarpon. Principal were fixed for the cultivated accessions, whereas the wild Component 1 explained 2.76% of the variation, whereas PC2 accessions had higher heterozygosity and lower inbreeding explained 2.24%. coefficient. We performed a PCA using eight SSR markers to Comparing the Relatedness between Wild compare existing cranberry cultivars and breeding selec- vs. Cultivated Vaccinium macrocarpon tions diversity with wild V. macrocarpon diversity identified When examining the diversity present in the wild V. in the current study, with PC 1 explaining 13.7% and PC macrocarpon population samples in Wisconsin and Minne- 2 explaining 7.6% of the variation (Fig. 6). We observed sota vs. a collection of cultivars and wild selections from a clear differentiation between our Wisconsin and

crop science, vol. 59, november–december 2019 www.crops.org 2419 Fig. 4. Genetic diversity statistics per locus for Vaccinium oxycoccos.

Minnesota wild accessions and the cultivated accessions, selections more than half a century ago, there is great need which includes all the major cultivars, wild accessions to incorporate new traits to enhance current materials (mostly eastern wild germplasm), and breeding selec- (Eck 1990). Because of this, we studied several unexplored tions (Fajardo et al., 2013; Zalapa et al., 2015; Schlautman wild cranberry populations in Wisconsin and Minnesota et al., 2018) (Fig. 6). In the PCA, we observed higher to understand the genetic diversity to increase the avail- heterozygosity and more differentiation and spread in ability of germplasm for cranberry breeding programs. the Wisconsin and Minnesota wild accessions than in the cultivated samples. Overview of the Genetic Relationship and Diversity of Wild Vaccinium macrocarpon DISCUSSION and Vaccinium oxycoccos Populations Although cranberry is a perennial fruit crop of great in Wisconsin and Minnesota importance with economic and human health benefits, Vaccinium macrocarpon and V. oxycoccos share many morpho- breeding has been hampered by the narrow phenotypic logical traits, which makes identification in the field and genetic diversity of cultivars and breeding materials challenging. Therefore, we designated field cuttings to each (Diaz-Garcia et al., 2019). Since the majority of culti- of the species, V. macrocarpon and V. oxycoccos, based on leaf vars used in breeding were developed from a few native size and shape (Smith et al., 2015) and then corroborated

2420 www.crops.org crop science, vol. 59, november–december 2019 these classifications using genetic information based on 32 hybridization of the two know ancestral species, V. macro- SSR markers. We constructed a Euclidean distance-based carpon and V. microcarpum (Turcz. ex Rupr.) Schmalh, or dendrogram and a PCA of all the accessions collected from the polyploidization with a third species that could (Fig. 1A–1B). In both, we clearly saw a differentiation now be extinct (Darrow and Camp 1945; Diaz-Garcia et between the two species. This clear differentiation between al., 2019). The involvement of V. macrocarpon in the specia- V. macrocarpon and V. oxycoccos has been previously reported tion of V. oxycoccos is further suggested by the number of using amplified fragment length polymorphisms (AFLPs) private alleles present when compared with V. oxycoccos. (Smith et al., 2015) and SSRs (Zalapa et al., 2015). The Similar results were observed in a study using allozymes, striking differentiation among the two species could be due in which they found that V. oxycoccos shared predominant to their different, but interconnected, evolutionary histo- alleles with V. macrocarpon (Mahy et al., 2000) (Table 2). ries, where V. macrocarpon could be an ancestral diploid In terms of overall genetic diversity for both species, we species, while V. oxycoccos is a more recently derived, mostly observed that V. oxycoccos, as expected due to its polyploid allotetraploid species (Smith et al., 2015; Diaz-Garcia et al., (4´) and outcrossing nature, was highly diverse with high

2019). We looked at the number of private alleles present in heterozygosity (HO, HT, and HS) levels. Such high levels of both species and found that V. oxycoccos had more private heterozygosity have been observed in other crops such as alleles (598) than V. macrocarpon with only 25. The high sweet potato and cassava (Manihot esculenta Crantz) (Fregene number of private alleles in V. oxycoccos could be explained et al., 2003; Rodriguez-Bonilla et al., 2014). Vaccinium macro- by its putative allopolyploid origin resulting from the carpon populations were quite heterozygous for a diploid

Table 4. Genetic diversity statistics per population of Vaccinium oxycoccos. Populations with only one to two samples after individuals were sorted into species were not analyzed.

Population N† Num‡ HO§ HS¶ HT# Gis†† BL 5 6.06 0.86 0.82 0.82 −0.04 FBF 7 6.18 0.81 0.75 0.75 −0.08 GL 18 11.3 0.82 0.82 0.82 0 GYPL 20 10.56 0.86 0.81 0.81 −0.05 JL 12 11.25 0.77 0.85 0.85 0.09 KLB 21 10.78 0.83 0.81 0.81 −0.02 KMP 12 6.53 0.83 0.78 0.78 −0.06 Locp 30 10.4 0.8 0.8 0.8 0 MWB 16 10 0.81 0.82 0.82 0.01 PMS 20 11.87 0.84 0.83 0.83 0 PMW 17 9.78 0.8 0.81 0.81 0 SEB 16 7.6 5 0.8 0.76 0.76 −0.05 SHPL 10 9 0.77 0.81 0.81 0.04 SR 17 9.75 0.77 0.79 0.79 0.02 LOL 3 4.78 0.77 0.81 0.81 0.04 LU 3 3.96 0.8 0.8 0.8 0 BB 20 12 0.83 0.83 0.83 −0.01 BLM 13 8.93 0.85 0.8 0.8 −0.05 BRR 20 9.5 0.82 0.79 0.79 −0.03 BRRM 20 9.06 0.82 0.77 0.77 −0.06 HB 19 11.56 0.8 0.81 0.81 0.01 LL 8 8.56 0.79 0.83 0.83 0.05 LOLB 20 11.59 0.81 0.81 0.81 −0.01 LRL 3 5.43 0.85 0.85 0.85 0 PB 20 10.25 0.8 0.8 0.8 0 PRF 20 11.75 0.82 0.83 0.83 0.01 TR 21 12.09 0.79 0.82 0.82 0.03 FRLK 13 7.15 0.85 0.77 0.77 −0.11 † Number of samples.

‡ Mean number of alleles.

§ Observed heterozygosity.

¶ Heterozygosity within populations.

# Total heterozygosity.

†† Inbreeding coefficient. crop science, vol. 59, november–december 2019 www.crops.org 2421 Fig. 5. Principal coordinate analysis of Vaccinium oxycoccos based on Bruvo genetic distance. plant that can self-pollinate and reproduce (Bruederle et al., nectar to pollinators such as bumblebees, which are able to 1996). Despite being highly heterozygous, V. macrocarpon transfer the pollen between cranberry flowers. This pollen

(HO = 0.5) exhibited lower levels of heterozygosity than V. movement maintains gene flow between populations and oxycoccos (HO = 0.8). The high diversity observed in wild can introduce new genetic variation (Liu et al., 2015). populations of cranberries may be the result of different Aside from pollinators potentially increasing diversity mechanisms that allow cranberries to maintain and enrich in undisrupted natural areas, open spaces are inviting for their genetic diversity. Cranberries provide pollen and bird species such as cranes as well as other animals. These

Table 5. Genetic diversity comparison among wild V. macrocarpon Wisconsin and Minnesota and cranberry cultivars.

Population N† Num‡ HO§ HS¶ HT# Gis†† Wild 139 10.87 0.99 0.76 0.76 −0.31 Cultivated 197 19.25 0.69 0.82 0.82 0.15 † Number of samples.

‡ Mean number of alleles.

§ Observed heterozygosity.

¶ Heterozygosity within populations.

# Total heterozygosity.

†† Inbreeding coefficient.

2422 www.crops.org crop science, vol. 59, november–december 2019 Fig. 6. Principal component analysis (PCA) comparing cultivated and wild individuals of Vaccinium macrocarpon. PC, principal component. animals that consume wild cranberry fruits and then move of alleles per locus ranging from 2–18 (Fig. 2A), whereas or migrate to other locations can disperse the cranberry other researchers found one to two alleles per locus seeds (Barzen et al., 2018). Cranberries are also often found using random amplified polymorphic DNA (RAPD) near large and interconnected bodies of water, and when markers (Novy and Vorsa, 1995; Stewart and Excoffier, fruit become detached from a cranberry plant, they float 1996) and allozymes (Mahy et al., 2000). These results due to their locular cavities and may travel long distances, could be explained by the large number of hypervariable which can result in chance seedlings increasing migration and codominant SSR markers used, as well as the large and diversity among populations. number of samples collected in this study. When looking at the data by population, we also found that most popu- Vaccinium macrocarpon Genetic Diversity lations were moderately to highly diverse, and diversity Based on our results, we found that V. macrocarpon is more levels and alleles per loci were comparable with those of diverse than previously thought by Bruederle et al. (1996) the six Wisconsin wild populations described by Zalapa et and Stewart and Excoffier (1996). We found high numbers al. (2015). Previous research in wild cranberries suggested

crop science, vol. 59, november–december 2019 www.crops.org 2423 that the low levels of heterozygosity and diversity could genotypic traits useful for the development of new be explained by self-pollination and inbreeding, especially breeding lines for cranberry production. As demon- after several inbreeding cycles (Bruederle et al., 1996; strated in Table 2 by the high number of alleles per locus Zalapa et al., 2015). However, our results indicate that (28), high levels of heterozygosity (0.8), and low levels of wild populations of V. macrocarpon likely outcross, leading inbreeding (0.00), the studied wild V. oxycoccos Minne- to moderate to high levels of heterozygosity (Table 2). sota and Wisconsin populations were highly diverse. This can be further confirmed by the low levels of As mentioned before, the polyploidization of multiple inbreeding observed suggested by the Gis values (−0.49– species and the outcrossing nature of V. oxycoccos has 0.07) (Table 3). been crucial to maintain and enrich the genetic diver- The low levels of heterozygosity observed in some sity of the species (Darrow and Camp 1945; Diaz-Garcia populations could be due to sampling bias, long-term et al., 2019). Mahy et al. (2000) analyzed the genetic evolutionary forces, and recent environmental events. For diversity of V. macrocarpon and V. oxycoccos populations example, Stewart and Excoffier (1996) and Bruederle et al. using allozymes and observed that, for several loci, V. (1996) believed a genetic bottleneck during the Pleistocene oxycoccos exhibited higher levels of heterozygosity (i.e., glaciation accounted for the low levels of diversity found in 0.5–0.6) than V. macrocarpon. Our study is the first to use the wild populations of V. macrocarpon they studied. Addi- codominant SSR markers to investigate diversity of the tionally, low levels of diversity could be the result of recent species using multiple populations. The previous study events, including environmental changes due to agricul- only used one population consisting of 31 individuals to ture, urbanization, and other anthropogenic activities. For test cross-amplification of eight SSR markers and differ- example, the declining number and quality of pollinators in entiation of the two species, but diversity statistics were an area due to human activity, such as a nearby construction not given besides similar number of alleles per locus and or a major road disrupting an area, has had an effect in other total number of alleles (Zalapa et al., 2015). Diversity plant species, such as lower fruit set in Halimium halimifo- statistics were comparable with those in other polyploids lium (L.) Willk. (Suárez-Esteban et al., 2014). Previous species such as sweetpotato [Ipomoea batatas (L.) Lam.] studies that found low levels of genetic variation in wild (Rodriguez-Bonilla et al., 2014) and cassava (Fregene et populations of cranberry were surprised by their low diver- al., 2003). Česonienė et al. (2013) analyzed the genetic sity results (Stewart and Excoffier, 1996; Bruederle et al., diversity of V. oxycoccos clones in Lithuanian reserves 1996). Although V. macrocarpon can self-pollinate and can using RAPD markers on two populations and found resist some levels of inbreeding, inbreeding in cranberry considerable levels of genetic diversity, suggesting that often results in low fertility rates and inbreeding depression, V. oxycoccos could be a promising crop for cultivation in signaling that cranberry is mostly an outcrossing species different locations across central Europe. This study also (Bruederle et al., 1996). Therefore, as demonstrated by our demonstrates that there could be invaluable resources results, natural cranberry genetic variation is much higher for breeding in wild locations such as the ones surveyed than observed in previous studies (Bruederle et al., 1996). in this study. In the Bruvo genetic distance PCoA for We suggest that such diversity was not detected in previous V. oxycoccos, we observed a closer relationship among V. studies mostly due to the use of molecular markers with oxycoccos individuals than in V. macrocarpon (Fig. 5). This low power and clonal sampling complications (Stewart and could be explained by the colonization history of cran- Excoffier, 1996; Mahy et al., 2000). berry. It has long been believed that most wild cranberry In the Bruvo genetic distance PCoA (Fig. 3), we populations resulted from a bottleneck that occurred observed that most individuals grouped together based in the Pleistocene and that they eventually populated on geographical location (Minnesota vs. Wisconsin). We receding glacial areas (Stewart et al., 1995). More exten- observed some minor clustering among some individuals sive sampling of new locations—for example, in the in some populations that could be explained by the low Driftless Area of Wisconsin—should be performed to fixation indices (0.2–0.3) in those populations, indicating investigate the possibility of a bottleneck and possible inbreeding. We also observed low variance explaining the colonization and cranberry spread (Stewart et al., 1995). PCoA clustering; these results are consistent with multiple population genetic studies that found similarly low levels Comparing the Relatedness between Wild of explained variance, as genetic matrices are highly and Cultivated Vaccinium macrocarpon dimensional (Patterson et al., 2006) Most cranberry cultivars today come from eastern wild selections made decades ago. Therefore, most cultivars Vaccinium oxycoccos Genetic Diversity share the same genetic background (Eck, 1990). A survey Vaccinium oxycoccos is one of the wild relatives of cran- published by Gallardo et al. (2018) identifies the needs and berry, which is one of the few crops native to North goals for cranberry breeding. Breeding needs vary consid- America. Therefore, V. oxycoccos may be a source of erably by growing location, but some of the most salient

2424 www.crops.org crop science, vol. 59, november–december 2019 needs across locations include disease resistance and fruit of great use to propagate and conserve germplasm (Lloyd quality, including anthocyanin content, fruit size, and flavor and McCown, 1980; Engelmann, 1991). We successfully due to acid and sugar content. Additionally, in contrast introduced to tissue culture (McCown and Zeldin, 2005) with other cranberry growing regions, Wisconsin, which a total of 359 genetically unique genotypes from our is the most important cranberry producer in the world with collection sites in Wisconsin and Minnesota. At least two 8093.71 ha (20,000 acres) from a total of 20,234.282 ha to three copies per genotype have been kept as backup for (50,000 acres) worldwide, faces pests such as the cranberry a total of 1305 plants. The unique genotypes from vastly fruit worm (Acrobasis vaccinii). Also, temperature challenges unexplored cranberry regions of Wisconsin and Minne- in terms of extreme summer heat and winter cold, as well as sota will be available for breeding in the USDA-ARS, fluctuating winter temperatures, are leading to permanent Vegetable Crop Research Unit in Madison, WI, Cran- frost risk (Dana, 1983; Roper and Vorsa, 1997). berry Genetics and Genomics Laboratory. Additionally, Because of these needs, we assessed how much of the selected accessions will be added to the National Clonal diversity present in our wild populations has been harnessed Germplasm Repository collection. for breeding and cultivar development. Zalapa et al. (2015) used 12 SSR markers and observed that there were consid- CONCLUSIONS erable levels of genetic diversity present in Wisconsin wild In this study, we used a set of markers capable of distin- populations compared with cultivars based on the number guishing between V. macrocarpon and V. oxycoccos. We of private and unique alleles, as well as levels of observed demonstrated the presence of untapped and previously heterozygosity. In our study, we compared 139 wild acces- undiscovered cranberry diversity in natural areas in the sions collected in Wisconsin and Minnesota, with 195 states of Wisconsin and Minnesota. The plant material cultivated cranberries including wild and breeding selec- used in this study was also composed of cranberry CWR tions originating mostly from eastern US states such as (V. oxycoccos), which may possess genotypic traits useful Massachusetts and New Jersey (Fajardo et al., 2013; Zalapa for the development of new breeding lines for cranberry et al., 2015; Schlautman et al., 2018). We found higher levels production. Particularly, since all of these populations of observed heterozygosity in wild populations than in were sampled in Minnesota and Wisconsin, they might cultivars and breeding selections. As expected, we observed possess unique adaptations to the largest cranberry produc- lower levels of inbreeding in wild populations than in culti- tion area in the world, central Wisconsin. Due to the lack vars. In the wild, cranberries are mostly outcrossing, which of exploration, Wisconsin growers have been forced to is not the case of plants used in breeding programs (Fajardo use cranberry cultivars developed for other growing areas. et al., 2013; Zalapa et al., 2015). As selection cycles occur, Therefore, the challenges faced by growers in our state more genetic diversity is lost, as has been the case in many are unique and include intense frost risk and cranberry crops, such as maize (Zea mays L.) (Whitt et al., 2002). fruit worm, among others. This study demonstrates the To further compare between cultivated breeding vast diversity and range of cranberry populations in two materials and wild V. macrocarpon accessions, we performed mostly unexplored and underutilized midwestern US a PCA in which we observed a clear separation between states. We believe that there could be many other states species on PC1 (Fig. 6). Most cranberry wild selections that possess wild areas with unique and unexplored popu- were from eastern germplasm, and a few have been bred to lations that could be excellent targets not only to establish create a narrow genetic pool in the current hybrid cultivars. a germplasm collection, but also for characterization. The reduction of diversity is an inevitable consequence of developing cultivars adapted to homogenous production Supplemental Material environment (Zalapa et al., 2015). The wild populations, Supplemental material is available online for this article. as mentioned above, are naturally outcrossing in an open environment that allows for chance seedlings, and not Conflict of Interest in a homogeneous environment such as a cranberry bed. The authors declare that there is no conflict of interest. These results, as well as the levels of heterozygosity, high- light how little of the present diversity is currently being Acknowledgments used in cranberry breeding. This project was supported by the USDA-ARS (Project no. 5090-21220-004-00-D provided to J. Zalapa); Ocean Spray Development of an In Vitro Cranberries, the Wisconsin Cranberry Growers Association, Cranberry Collection and the Cranberry Institute. The authors want to thank the Wisconsin Departments of Natural Resources and the US For- Conservation of plants in the field or greenhouse condi- est Service for providing permits and coordinates to find cran- tion presents formidable challenges, many times resulting berry growing areas. Special thanks to Minnesota Department in the loss of plant genetic resources conserved (Engel- of Natural Resources personnel Erica Rowe, Mark Cleveland, mann, 2011). Therefore, using tissue culture techniques is and Charles Tucker for all their help with permits, providing crop science, vol. 59, november–december 2019 www.crops.org 2425 the researchers a place to stay in Norris Camp, and coordi- Chandler F., and I. Demoranville. 1958. Cranberry varieties of nates to cranberry locations. We want to thank the Long Lake North America. Bull. 513. Massachusetts Agric. Exp. Stn., Conservation Center for allowing the researchers to collect on Amherst. their lands. Lorraine Rodríguez-Bonilla wants to thank the Cheema, A.K. 2018. Plant breeding its applications and future American Association of Hispanics in Higher Education and prospects. Int. J. Eng. Technol. Sci. Res. 5:88–94. the CULTIVAR program for providing funding for travel- Clark, L.V., and M. Jasieniuk. 2011. POLYSAT: An R package for ing, and well as the University of Wisconsin–Madison SciMed polyploid microsatellite analysis. Mol. Ecol. Resour. 11:562– Graduate Research Scholars (GRS) funding. J. Zalapa would 566. doi:10.1111/j.1755-0998.2011.02985.x like to express his gratitude through Psalm 136:1. The authors Dana, M.N. 1983. Cranberry cultivar list. Fruit Var. J. 37:88–95. also thank the anonymous reviewers who helped enhance the Darrow, G.M., and W.H. Camp. 1945. Vaccinium hybrids and the development of new horticultural material. Bull. Torrey Bot. quality of this paper. Club 72:1–21. doi:10.2307/2481260 Dempewolf, H., G. Baute, J. Anderson, B. Kilian, C. Smith, and References L. Guarino. 2017. Past and future use of wild relatives in Anh, V.L., Y. Inoue, S. Asuke, T.T.P. Vy, N.T. Anh, S. Wang, et al. Crop breeding. Crop Sci. 57:1070–1082. doi:10.2135/crop- 2018. Rmg8 and Rmg7, wheat genes for resistance to the wheat sci2016.10.0885 blast fungus, recognize the same avirulence gene AVR-Rmg8. Diaz-Garcia, L., L. Rodriguez-Bonilla, J. Rohde, T. Smith, and Mol. Plant Pathol. 19:1252–1256. doi:10.1111/mpp.12609 J. Zalapa. 2019. Pacbio sequencing reveals identical organ- Areškevičiūtė, J., A. Paulauskas, L. Česonienė, and R. Daubaras. elle genomes between American cranberry (Vaccinium macro- 2006. Genetic characterisation of wild cranberry (Vaccinium carpon Ait.) and a wild relative. Genes 10:291. doi:10.3390/ oxycoccos) from Čepkeliai reserve by the RAPD method. genes10040291 Biologija 2006:5–7. Doyle, J.J., and J.L. Doyle. 1987. CTAB DNA extraction in plants. Bache, S.M., and H. Wickham. 2014. magrittr: A forward-pipe Phytochem. Bull. 19:11–15. operator for R. R package version 1. Comprehensive R Arch. Dray, S., and A.-B. Dufour. 2007. The ade4 package: Implement- Network, Vienna. ing the duality diagram for ecologists. J. Stat. Softw. 22(4). Barzen, J., T. Thousand, J. Welch, M. Fitzpatrick, and T. Triet. Eck, P. 1990. The American cranberry. Rutgers Univ. Press, New 2018. Determining the diet of whooping cranes (Grus ameri- Brunswick, NJ. cana) through field measurements. Waterbirds 41:22–34. Engelmann, F. 1991. In vitro conservation of tropical plant doi:10.1675/063.041.0104 germplasm: A review. Euphytica 57:227–243. doi:10.1007/ Brown, P.N., C.E. Turi, P.R. Shipley, and S.J. Murch. BF00039669 2012. Comparisons of large (Vaccinium macrocarpon Ait.) Engelmann, F. 2011. Use of biotechnologies for the conservation and small (Vaccinium oxycoccos L., Vaccinium vitis-idaea of plant biodiversity. In Vitro Cell. Dev. Biol. Plant 47:5–16. L.) cranberry in British Columbia by phytochemical doi:10.1007/s11627-010-9327-2 determination, antioxidant potential, and metabolomic pro- Fajardo, D., J. Morales, H. Zhu, S. Steffan, R. Harbut, N. Bassil, filing with chemometric analysis. Planta Med. 78:630–640. et al. 2013. Discrimination of American cranberry cultivars doi:10.1055/s-0031-1298239 and assessment of clonal heterogeneity using microsatellite Brownstein, M.J., J.D. Carpten, and J.R. Smith. 1996. Modu- markers. Plant Mol. Biol. 31:264–271. doi:10.1007/s11105- lation of non-templated nucleotide addition by Taq DNA 012-0497-4 polymerase: Primer modifications that facilitate genotyping. Fregene, M.A., M. Suarez, J. Mkumbira, H. Kulembeka, E. Biotechniques 20:1004–1010. doi:10.2144/96206st01 Ndedya, A. Kulaya, et al. 2003. Simple sequence repeat Brozynska, M., A. Furtado, and R.J. Henry. 2016. Genom- marker diversity in cassava landraces: Genetic diversity and ics of crop wild relatives: Expanding the gene pool for crop differentiation in an asexually propagated crop. Theor. Appl. improvement. Plant Biotechnol. J. 14:1070–1085. doi:10.1111/ Genet. 107:1083–1093. doi:10.1007/s00122-003-1348-3 pbi.12454 Galili, T. 2015. dendextend: An R package for visualizing, adjust- Bruederle, L.P., M.S. Hugan, J.M. Dignan, and N. Vorsa. 1996. ing and comparing trees of hierarchical clustering. Bioinfor- Genetic variation in natural populations of the large cran- matics 31:3718–3720. doi:10.1093/bioinformatics/btv428 berry, Vaccinium macrocarpon Ait.(). Bull. Torrey Gallardo, R.K., P. Klingthong, Q. Zhang, J. Polashock, A. Atu- Bot. Club 123:41–47. doi:10.2307/2996305 cha, J. Zalapa, et al. 2018. Breeding trait priorities of the cran- Bruvo, R., N.K. Michiels, T.G. D’Souza, and H. Schulenburg. berry industry in the United States and Canada. HortScience 2004. A simple method for the calculation of microsatellite 53:1467–1474. doi:10.21273/HORTSCI13219-18 genotype distances irrespective of ploidy level. Mol. Ecol. Georgi, L., R.H. Herai, R. Vidal, M.F. Carazzolle, G.G. Pereira, 13:2101–2106. doi:10.1111/j.1365-294X.2004.02209.x J. Polashock, and N. Vorsa. 2012. Cranberry microsatel- Česonienė, L., R. Daubaras, I. Jasutienė, J. Venclovienė, and I. lite marker development from assembled next-generation Miliauskienė. 2011. Evaluation of the biochemical compo- genomic sequence. Mol. Breed. 30:227–237. doi:10.1007/ nents and chromatic properties of the juice of Vaccinium mac- s11032-011-9613-7 rocarpon Aiton and Vaccinium oxycoccos L. Plant Foods Hum. Gu, Z., L. Gu, R. Eils, M. Schlesner, and B. Brors. 2014. circlize Nutr. 66:238–244. doi:10.1007/s11130-011-0241-5 implements and enhances circular visualization in R. Bioin- Česonienė, L., R. Daubaras, A. Paulauskas, J. Žukauskienė, and M. formatics 30:2811–2812. doi:10.1093/bioinformatics/btu393 Zych. 2013. Morphological and genetic diversity of European Guichoux, E., L. Lagache, S. Wagner, P. Chaumeil, P. Léger, O. cranberry (Vaccinium oxycoccos L., Ericaceae) clones in Lithu- Lepais, et al. 2011. Current trends in microsatellite geno- anian reserves. Acta Soc. Bot. Pol. 82:211–217. doi:10.5586/ typing. Mol. Ecol. Resour. 11:591–611. doi:10.1111/j.1755- asbp.2013.026 0998.2011.03014.x

2426 www.crops.org crop science, vol. 59, november–december 2019 Hamernik, A.J., R.E. Hanneman, Jr., and S.H. Jansky. 2009. Patterson, N., A.L. Price, and D. Reich. 2006. Population struc- Introgression of wild species germplasm with extreme resis- ture and eigenanalysis. PLoS Genet. 2:e190. doi:10.1371/jour- tance to cold sweetening into the cultivated potato. Crop Sci. nal.pgen.0020190 49:529–542. doi:10.2135/cropsci2008.04.0209 Peakall, R., and P.E. Smouse. 2012. GenAlEx 6.5: Genetic Jacquemart, A.-L. 1997. Vaccinium oxycoccos L. (Oxycoccus palustris analysis in Excel. Population genetic software for teach- Pers.) and Vaccinium microcarpum (Turcz. ex Rupr.) Schmalh. ing and research: An update. Bioinformatics 28:2537–2539. (Oxycoccus microcarpus Turcz. ex Rupr.). J. Ecol. 85:381–396. doi:10.1093/bioinformatics/bts460 doi:10.2307/2960511 R Core Team. 2018. R: A language and environment for statistical Ji, W., Y. Zhu, Y. Li, L. Yang, X. Zhao, H. Cai, and X. Bai. 2010. computing. R Found. Stat. Comput., Vienna. Over-expression of a glutathione S-transferase gene, GsGST, Rodriguez-Bonilla, L., H.E. Cuevas, M. Montero-Rojas, from wild soybean (Glycine soja) enhances drought and salt tol- F. Bird-Pico, D. Luciano-Rosario, and D. Siritunga. erance in transgenic tobacco. Biotechnol. Lett. 32:1173–1179. 2014. Assessment of genetic diversity of sweet potato in doi:10.1007/s10529-010-0269-x Puerto Rico. PLoS One 9:e116184. doi:10.1371/journal. Jombart, T. 2008. adegenet: A R package for the multivariate pone.0116184 analysis of genetic markers. Bioinformatics 24:1403–1405. Rodriguez-Bonilla, L., J. Rohde, D. Matusinec, and J. Zalapa. doi:10.1093/bioinformatics/btn129 2019. Cross-transferability analysis of SSR markers devel- Kamvar, Z.N., J.F. Tabima, and N.J. Grünwald. 2014. Poppr: An oped from the American cranberry to other Vaccinium species R package for genetic analysis of populations with clonal, of agricultural importance. Genet. Resour. Crop Evol. (in partially clonal, and/or sexual reproduction. PeerJ 2:e281. press). doi:10.1007/s10722-019-00826-1 doi:10.7717/peerj.281 Roper, T.R., and N. Vorsa. 1997. Cranberry: Botany and hor- Kassambara, A., and F. Mundt. 2016. Package factoextra: Extract ticulture. In: J. Janick, editor, Horticultural reviews. and visualize the results of multivariate data analyses. Com- Vol. 21. John Wiley & Sons, Oxford, UK. p. 215–249. prehensive R Arch. Network, Vienna. doi:10.1002/9780470650660.ch7 Khoury, C.K., S. Greene, J. Wiersema, and N. Maxted. 2013. An Rowe, P. 1984. Breeding bananas and plantains. Plant Breed. Rev. inventory of crop wild relatives of the United States. Crop 2:135–155. doi:10.1002/9781118060995.ch4 Sci. 53:1496–1508. doi:10.2135/cropsci2012.10.0585 Schlautman, B., G. Covarrubias-Pazaran, L. Diaz-Garcia, M. Ior- Liu, G., G. Lu, L. Zeng, and G.L. Wang. 2002. Two broad‐ izzo, J. Polashock, E. Grygleski, et al. 2017. Construction of a spectrum blast resistance genes, Pi9(t) and Pi2(t), are physi- high-density American cranberry (Vaccinium macrocarpon Ait.) cally linked on rice chromosome 6. Mol. Genet. Genomics composite map using genotyping-by-sequencing for multi- 267:472. doi:10.1007/s00438-002-0677-2 pedigree linkage mapping. G3: Genes, Genomes, Genet. Liu, M., S.G. Compton, F.-E. Peng, J. Zhang, and X.-Y. Chen. 2015. 7:1177–1189. doi:10.1534/g3.116.037556 Movements of genes between populations: Are pollinators Schlautman, B., G. Covarrubias-Pazaran, L.A. Diaz-Garcia, J. more effective at transferring their own or plant genetic mark- Johnson-Cicalese, M. Iorrizo, L. Rodriguez-Bonilla, et al. ers? Proc. Biol. Sci. 282:20150290. doi:10.1098/rspb.2015.0290 2015a. Development of a high-density cranberry SSR linkage Lloyd, G., and B. McCown. 1980. Commercially-feasible micro- map for comparative genetic analysis and trait detection. Mol. propagation of mountain laurel, Kalmia latifolia, by use of Breed. 35:177. doi:10.1007/s11032-015-0367-5 shoot-tip culture. Combined Proc., Int. Plant Propagators’ Schlautman, B., G. Covarrubias-Pazaran, L. Rodriguez-Bonilla, Soc. 30:421–427. K. Hummer, N. Bassil, T. Smith, and J. Zalapa. 2018. Genetic Mahy, G., L.P. Bruederle, B. Connors, M. Van Hofwegen, and diversity and cultivar variants in the NCGR cranberry (Vac- N. Vorsa. 2000. Allozyme evidence for genetic autopoly- cinium macrocarpon Aiton) collection. J. Genet. 97:1339–1351. ploidy and high genetic diversity in tetraploid cranberry, doi:10.1007/s12041-018-1036-3 Vaccinium oxycoccos (Ericaceae). Am. J. Bot. 87:1882–1889. Schlautman, B., D. Fajardo, T. Bougie, E. Wiesman, J. Polashock, doi:10.2307/2656840 N. Vorsa, et al. 2015b. Development and validation of 697 McCown, B.H., and E.L. Zeldin. 2005. Vaccinium spp. cranberry. In: novel polymorphic genomic and EST-SSR markers in the R.E. Litz, editor, Biotechnology of fruit and nut crops. CABI, American cranberry (Vaccinium macrocarpon Ait.). Molecules Wallingford, UK. p. 247–261. doi:10.1079/9780851996622.0247 20:2001–2013. doi:10.3390/molecules20022001 Meirmans, P.G., and P.H. Van Tienderen. 2004. genotype and Schuelke, M. 2000. An economic method for the fluorescent genodive: Two programs for the analysis of genetic diversity of labeling of PCR fragments. Nat. Biotechnol. 18:233–234. asexual organisms. Mol. Ecol. Notes 4:792–764. doi:10.1111/ doi:10.1038/72708 j.1471-8286.2004.00770.x Shakiba, E., J.D. Edwards, F. Jodari, S.E. Duke, A.M. Baldo, P. National Research Council. 1993. Managing global genetic Komiliev, et al. 2017. Genetic architecture of cold tolerance resources: Agricultural crop isues and policies. Natl. Acad. in rice (Oryza sativa) determined through high resolution Press, Washington, DC. genome-wide analysis. PLoS One 12:e0172133. doi:10.1371/ Novy, R.G., and N. Vorsa. 1995. Identification of intracultivar journal.pone.0172133 genetic heterogeneity in cranberry using silver-stained RAPDs. Smith, T.W., C. Walinga, S. Wang, P. Kron, J. Suda, and J. Zalapa. HortScience 30:600–604. doi:10.21273/HORTSCI.30.3.600 2015. Evaluating the relationship between diploid and tetra- Paradis, E. 2010. pegas: An R package for population genetics with ploid Vaccinium oxycoccos (Ericaceae) in eastern Canada. Bot- an integrated–modular approach. Bioinformatics 26:419–420. any 93:623–636. doi:10.1139/cjb-2014-0223 doi:10.1093/bioinformatics/btp696 Stewart, C.N., and L. Excoffier. 1996. Assessing population Paradis, E., J. Claude, and K. Strimmer. 2004. APE: Analyses of genetic structure and variability with RAPD data: Applica- phylogenetics and evolution in R language. Bioinformatics tion to Vaccinium macrocarpon (American cranberry). J. Evol. 20:289–290. doi:10.1093/bioinformatics/btg412 Biol. 9:153–171. doi:10.1046/j.1420-9101.1996.9020153.x crop science, vol. 59, november–december 2019 www.crops.org 2427 Stewart, C.N., C. Neal Stewart, and E.T. Nilsen. 1995. Pheno- Vorsa, N., J. Johnson-Cicalese, and J. Polashock. 2008. A typic plasticity and genetic variation of Vaccinium macrocarpon, blueberry by cranberry hybrid derived from a Vaccinium the American cranberry. I. Reaction norms of clones from darrowii ´ (V. macrocarpon ´ V. oxycoccos) intersectional central and marginal populations in a common garden. Int. J. cross. Acta Hortic. 810:187–190. doi:10.17660/ActaHor- Plant Sci. 156:687–697. doi:10.1086/297291 tic.2009.810.24 Suárez-Esteban, A., M. Delibes, and J.M. Fedriani. 2014. Unpaved Whitt, S.R., L.M. Wilson, M.I. Tenaillon, B.S. Gaut, and E.S. roads disrupt the effect of herbivores and pollinators on the Buckler, IV. 2002. Genetic diversity and selection in the reproduction of a dominant shrub. Basic Appl. Ecol. 15:524– maize starch pathway. Proc. Natl. Acad. Sci. USA 99:12959– 533. doi:10.1016/j.baae.2014.08.001 12962. doi:10.1073/pnas.202476999 Tang, Y., M. Horikoshi, and W. Li. 2016. ggfortify: Unified inter- Zalapa, J.E., T.C. Bougie, T.A. Bougie, B.J. Schlautman, E. Wies- face to visualize statistical results of popular R packages. R J. man, A. Guzman, et al. 2015. Clonal diversity and genetic 8:474–489. doi:10.32614/RJ-2016-060 differentiation revealed by SSR markers in wild Vaccinium Tanksley, S.D., S. Grandillo, T.M. Fulton, D. Zamir, Y. Eshed, macrocarpon and Vaccinium oxycoccos. Ann. Appl. Biol. 166:196– V. Petiard, et al. 1996. Advanced backcross QTL analysis in a 207. doi:10.1111/aab.12173 cross between an elite processing line of tomato and its wild Zhang, H., N. Mittal, L.J. Leamy, O. Barazani, and B.-H. Song. relative L. pimpinellifolium. Theor. Appl. Genet. 92:213–224. 2017. Back into the wild: Apply untapped genetic diversity doi:10.1007/BF00223378 of wild relatives for crop improvement. Evol. Appl. 10:5–24. Ullstrup, A.J. 1972. The impacts of the southern corn leaf blight doi:10.1111/eva.12434 epidemics of 1970–1971. Annu. Rev. Phytopathol. 10:37–50. Zhu, H., D. Senalik, B.H. McCown, E.L. Zeldin, J. Speers, J. doi:10.1146/annurev.py.10.090172.000345 Hyman, N. Bassil, K. Hummer, P.W. Simon, and J. Zalapa. Vander Kloet, S.P. 1983. The taxonomy of Vaccinium § Oxycoccus. 2012. Mining and validation of pyrosequenced simple Rhodora 85:1–43. sequence repeats (SSRs) from American cranberry (Vaccinium Vander Kloet, S.P. 1988. The genus Vaccinium in North America. macrocarpon Ait.). Theor. Appl. Genet. 124:87–96. doi:10.1007/ Agric. Canada, Ottawa. s00122-011-1689-2

2428 www.crops.org crop science, vol. 59, november–december 2019