Species in the genus Nuphar Sm. (Nymphaeaceae) are freshwater aquatic plants distributed in the temperate zones of the Northern Hemisphere. The genus is one of the most problematic aquatic macrophytes to identify; seven to 20 species have been recognized worldwide (Cook, 1996; Padgett, 2007), and six species with one or two varieties or forms are distributed in Japan (Shiga et al., 2006; Shiga, 2007; Shiga and Kadono, 2015). In the Saijo Basin (Hiroshima Prefecture, western Japan), three taxa of Nuphar are distributed sympatrically (Nuphar japonica DC., N. oguraensis Miki var. akiensis Shimoda, and N. ×saijoensis (Shimoda) Padgett & Shimoda; Padgett et al., 2002), according to the classification of species based on genetic and morphological variations (Shiga, 2007). Nuphar japonica is widely distributed throughout Japan. By contrast, N. oguraensis var. akiensis grows in a few remote regions of western Japan (this variety also appears in Korea and Taiwan), and N. ×saijoensis, which originated from natural hybridization between N. japonica and N. oguraensis (Padgett et al., 2002), grows in several regions of western Japan (Shiga, 2007). Both N. oguraensis and N. oguraensis var. akiensis are listed as N. pumila subsp. oguraensis (Miki) Padgett in Padgett (2007).

In recent years, aquatic plants growing in irrigation ponds, such as Nuphar species, are disappearing owing to land reclamation, repair work, water quality deterioration, and invasion of exotic species. Consequently, N. oguraensis is currently included in the Japanese Red Data Book (Ministry of the Environment Japan, 2015). Thus, understanding the genetic diversity of Nuphar species will play a key role in its future management. Although microsatellite markers have been developed in two relatives of N. japonica (N. lutea [Ouborg et al., 2000] and N. submersa [Yokogawa et al., 2012]), their utility for N. japonica and N. oguraensis has been shown to be limited (Yokogawa et al., 2012). Therefore, we have developed polymorphic genomic microsatellite markers for use in genetic investigations of the three Nuphar taxa of the Saijo Basin.

METHODS AND RESULTS

We collected plant samples from three populations of N. japonica (Sawahara, Kouno, and Doinouchisako-shita ponds), one population of N. oguraensis var. akiensis (Rakan Pond), and one population of N. ×saijoensis (Imori-shita Pond) in the Saijo Basin, Hiroshima Prefecture, Japan (Appendix 1); each population was from a separate pond. The geographic distance between individual ponds ranged from 2.3 to 8.4 km, with an average of 5.8 km. We selected Saijo Basin as the study site because it is unique in having three sympatrically distributed Nuphar taxa. Sample size was eight or 12 plants per population (48 plants in total). We allowed at least 10 m between sampled individuals to avoid duplicated sampling from the same genet. Pond size and the clonal nature of the species led to small sample sizes for each population. Total genomic DNA was isolated from 30–50 mg of leaf tissue from each plant by using the DNA Suisui–VS extraction buffer (RIZO, Tsukuba, Ibaraki, Japan).

DNA extracted from one N. japonica plant collected in Kouno Pond was used for library preparation with a TruSeq Nano DNA Library Prep Kit (Illumina, San Diego, California, USA). Sequencing was performed on a MiSeq Benchtop Sequencer (Illumina) in the 2 × 300-bp read mode, yielding 9,135,890 reads. Prior to assembly, reads were trimmed of adapters and poor quality bases using Trimmomatic version 0.32 (Bolger et al., 2014). Then, reads were assembled into 2,380,573 contigs with fastq-join (Aronesty, 2011). MSATCOMMANDER version 1.0.8 software (Faircloth, 2008) identified 26,667 contigs with dinucleotide motifs with a minimum of 15 repeats. Primers were then designed using Primer3 version 2.2.3 software (Rozen and Skaletsky, 1999) with default settings.

We tested 72 primer pairs. PCR mixtures (10 µL) contained 5 ng of DNA, 0.2 µL of KOD FX Neo polymerase (Toyobo, Osaka, Japan), 5 µL of 2× PCR buffer, 0.4 mM dNTPs, and 0.2 µM of each primer (forward primer with a barcoded split tag [BStag], reverse primer, and fluorescent BStag primer). Each BStag comprised a basal region common among six BStags, a three-nucleotide “barcode” sequence, and a mismatched nucleotide in the middle position. A BStag was added at the 5′ end of the forward primer for each locus to enable postlabeling (Shimizu and Yano, 2011). We labeled the BStag primers with fluorescent dyes to create F9GAC-FAM (5′-CTAGTATCAGGACGAC-3′), F9GTC-VIC (5′-CTAGTATGAGGACGTC-3′), F9TAC-NED (5′-CTAGTATCAGGACTAC-3′), F9GCC-PET (5′-CTAGTATTAGGACGCC-3′), F9CCG-FAM (5′-CTAGTATTAGGACCCG-3′), and F9AGG-VIC (5′-CTAGTATTAGGACAGG-3′) (Shimizu and Yano, 2011). Fragments were amplified in a Veriti Thermal Cycler (Life Technologies, Carlsbad, California, USA) as follows: initial denaturation at 94°C for 2 min; 30 cycles of denaturation at 98°C for 10 s, annealing at 60°C for 30 s, and extension at 68°C for 30 s; 12 cycles of 98°C for 10 s, 49°C for 30 s, and 68°C for 10 s; and a final extension at 68°C for 7 min. The size of the PCR products was measured using an ABI PRISM 3130xl Genetic Analyzer and GeneMapper software (both from Life Technologies). We conducted a PCR amplification trial twice for each of four randomly selected individuals from each of the three Nuphar taxa, and 15 of 72 primer pairs showed a clear, strong, single band for each allele in N. japonica. (The sequences of these 15 primer pairs are listed in Table 1.) Fifteen loci were amplified in N. japonica and in N. ×saijoensis, and 11 loci were amplified in N. oguraensis var. akiensis (Table 2).

Table 1.

Characteristics of the 15 polymorphic microsatellite markers developed for Nuphar japonica.^a

Evaluation of genetic polymorphism within all 32 N. japonica plants showed that the 15 loci had moderate levels of polymorphism. The total number of alleles per locus in these populations ranged from two to nine (mean ± SE: 3.47 ± 0.48). Expected heterozygosity per locus was generally high, ranging from 0.50 to 0.78 (0.56 ± 0.02). At the population level, the number of alleles ranged from two to six (2.47 ± 0.13), observed heterozygosity from 0.50 to 1.00 (0.84 ± 0.02), and expected heterozygosity from 0.45 to 0.74 (0.54 ± 0.01). The levels of polymorphism were moderate in N. japonica and N. ×saijoensis and relatively low in N. oguraensis var. akiensis (Table 2). Deviations from Hardy–Weinberg equilibrium in each population and linkage disequilibrium were tested using GENEPOP software, Web version 4.2 (Raymond and Rousset, 1995). Among all loci in all five populations, 11 locus–population combinations deviated significantly from Hardy–Weinberg equilibrium (P < 0.01; Table 2). Significant linkage disequilibrium (P < 0.05) was observed among six locus pairs in N. japonica populations (NJ03362 and NJ14763, NJ04340 and NJ14763, NJ04340 and NJ15714, NJ04340 and NJ18322, NJ14763 and NJ15714, and NJ14763 and NJ18010). Such lack of equilibrium and significant linkage disequilibrium could be explained by the small number of samples in each population and clonal reproduction by rhizome growth in Nuphar taxa.

CONCLUSIONS

In this study, a total of 15 polymorphic microsatellite markers for N. japonica were developed. Eleven of them (five polymorphic) also amplified in N. oguraensis var. akiensis and 15 (14 polymorphic) amplified in N. ×saijoensis. These markers will be useful both for investigating gene flow among the three taxa of Nuphar in the Saijo Basin and for determining the effects of habitat networks on levels of genetic diversity within the populations of these taxa. The results may have important implications for aquatic plant conservation and restoration.

Table 2.

Genetic variation of the 15 polymorphic microsatellite loci in three populations of Nuphar japonica, one population of N. oguraensis var. akiensis, and one population of N. ×saijoensis.^a