108 
Fishery Bulletin 119(2-3) 
Quaternary period (from 2.58 MYA to present). In the 
Early to Middle Pleistocene (0.78—2.58 MYA), the main 
islands of Japan were connected to the continent of Asia, 
and the Ryukyu Islands were connected to the continent 
of Asia through Taiwan, by large land bridges, repeatedly 
(Kizaki and Oshiro, 1977, 1980; Ujiie, 1986; Kimura, 1996, 
2000). Clade A likely became differentiated in the semi- 
closed environment of the East China Sea and dispersed 
northward when the southern straits of the Sea of Japan 
opened. Clade B might have extended its distribution 
northward along the Pacific coast after divergence from 
clade A and divided into clades B1 and B2 approximately 
0.50—0.80 MYA in the Middle Pleistocene. The divergence 
of clades B1 and B2 may have been caused by changes 
in the dynamic of the ancient Oyashio Current, which 
strengthened during 0.48-0.34 MYA (Matsuzaki et al., 
2014); afterward, the Kuroshio Current dominated 0.78— 
0.85 MYA (Kang et al., 2010; Gallagher et al., 2015). Clade 
B extended its distribution northward along the coast of 
Honshu with the Kuroshio Current but was prevented 
from further extension to the north by the strength- 
ened Oyashio Current approximately 0.5 MYA, and the 
Kuroshio (clade B1) and the Oyashio (clade B2) lineages 
diverged. 
Although the results of the mismatch distribution anal- 
ysis, the neutrality tests (Table 4), and the combination of 
high h and low x values (Table 2) (Grant and Bowen, 1998) 
indicate a population expansion of clade A (Table 4, Fig. 
5A), the possibility of a constant population could not be 
rejected in the BSP (Fig. 5D). However, the BSP for clade A 
indicates that N, notably increased after LGM, and clade 
A may have expanded its distribution from the East China 
Sea to the Sea of Japan after LGM. The gradual decrease 
of genetic diversity (h in mtDNA analysis and Hy in SSR 
analysis) in populations of polkadot skate from the Danjo 
Islands in the East China Sea northward to Niigata Pre- 
fecture in the Sea of Japan (Table 2, Fig. 4, Suppl. Table 2 
[online only]) may have been due to a founder event or to 
differences in adaptive abilities, such as tolerance to cold 
temperatures as frequencies of restricted haplotypes (hap- 
lotypes al and a3) increase with increasing latitude (Fig. 4). 
Similar declines in genetic diversity toward the north in 
the Sea of Japan have been reported for the sailfin sand- 
fish (Arctoscopus japonicus) (see Shirai et al., 2006) and 
a rocky intertidal goby, Chaenogobius gulosus (see Hirase 
and Ikeda, 2014). 
Although sum of squared deviations and the Harpend- 
ing’s raggedness index are not significant and Fu’s F's does 
not indicate certain departures from neutrality in clade 
B1 (Table 4), the mismatch distributions (Fig. 5B) roughly 
have an L shape, indicating that clade B1 had experienced 
a population bottleneck (Rogers and Harpending, 1992). 
The combination of low h and low zm values observed in 
clade B1 (Table 2) also indicates a recent population bot- 
tleneck or founder event for a single or a few mtDNA 
lineages (Grant and Bowen, 1998). However, no signal 
of population bottleneck is apparent in the BSP (Fig. 
5E). In any case, N, in clade B1 did not notably increase 
after LGM, and JN, in clade B1 is considerably smaller at 
present than JN, for the other clades, according to values 
for 0, (Table 4) and the BSP. The low genetic diversity in 
both mtDNA and SSR analyses for clade B1 was related 
to the small N, of the clade. In clade B2, the population 
expansion could not be inferred by using neutrality tests 
(Table 4); however, results of the mismatch distribution 
analysis (Table 4, Fig. 5C) and the combination of high h 
and low x values (Table 2) indicate that the population 
expanded (Rogers and Harpending, 1992; Grant and 
Bowen, 1998). Although the possibility of a constant popu- 
lation could not be rejected in the BSP (Fig. 5F), the BSP 
for clade B2, as well as that for clade A, indicates that N, 
increased after LGM (Fig. 5D). 
Implications for management and conservation 
The application of genetic and evolutionary measures in 
the conservation and management of natural resources 
has led to the identification of management units (MUs), 
which represent populations or groups of populations that 
are connected by low levels of gene flow and are function- 
ally independent (Moritz, 1994, 2002). Significant popula- 
tion structure in the polkadot skate was observed around 
Japan in our study, and we suggest considering the exis- 
tence of a minimum of 3 MUs for this species: the group 
of populations in the Sea of Japan and the East China 
Sea, the population along the southern Pacific coast, and 
the population along the northern Pacific coast. Each MU 
possesses unique genetic traits and, therefore, should be 
treated independently. Particularly, the MU for the pop- 
ulation along the southern Pacific coast is characterized 
by genetic diversity that is lower than that of the other 2 
MUs and should be monitored and managed more care- 
fully. Ecological studies, such as those that focus on age, 
growth, reproduction, and movement patterns, for each 
MU are also necessary to establish appropriate conserva- 
tion and management strategies. 
Conclusions 
We examined the genetic population structure of polkadot 
skate around Japan. Results of the mtDNA analysis of 
polkadot skate indicate that this species has 3 lineages, 
clades A, B1, and B2. Clade A consists of populations in 
the Sea of Japan and the East China Sea. Clades B1 and 
B2 comprise populations in the Pacific Ocean along the 
southern and northern coasts of Japan, respectively. This 
genetic differentiation is also supported by results from 
SSR analysis. Results of this study indicate that the diver- 
gence between clades A and B occurred because of isola- 
tion of the East China Sea in the Early Pleistocene and 
that the divergence between clades B1 and B2 occurred 
likely because of changes in the dynamics of ancient water 
currents in the Middle Pleistocene. Each of the 3 geneti- 
cally discrete management units of polkadot skate around 
Japan possess unique genetic traits and evolutionary his- 
tory, and each should be treated independently in manage- 
ment of this species in the future. 
