Chromosomal Evolution of the Genus Nannospalax (Palmer 1903) (Rodentia, Muridae) from Western Turkey

Home , Spalax

Turkish Journal of Zoology Turk J Zool (2013) 37: 470-487 http://journals.tubitak.gov.tr/zoology/ © TÜBİTAK Research Article doi:10.3906/zoo-1208-25

Chromosomal evolution of the genus Nannospalax (Palmer 1903) (Rodentia, Muridae) from western Turkey

Ferhat MATUR*, Faruk ÇOLAK, Tuğçe CEYLAN, Murat SEVİNDİK, Mustafa SÖZEN Department of Biology, Faculty of Arts and Sciences, Bülent Ecevit University, Zonguldak, Turkey

Received: 29.08.2012 Accepted: 17.02.2013 Published Online: 24.06.2013 Printed: 24.07.2013

Abstract: We used 33 blind mole rats belonging to 10 different chromosomal races from 10 localities in western Turkey. We applied G- and C-banding techniques to compare chromosomal races as well as clarifying relationships between them. We discussed cytogenetic similarities and differences between chromosomal races. We concluded that 2n = 60C is the ancestor of the other chromosomal races. However, as a result of ongoing evolution processes 2n = 38 and 2n = 60K have become ancestors to chromosomal races on their peripherals. We discovered which rearrangements contribute to the evolution of such a complex chromosomal race system in a genus. With this study we provide a comprehensive comparison of the 10 chromosomal races and perform a cladistic analysis using chromosomal rearrangement character states. According to our tree, chromosomal races with a low diploid number formed a monophyletic group.

Key words: Blind mole rat, comparative cytogenetic, G- and C-banding, chromosome differentiation, phylogeny, Anatolia

1. Introduction assumed that ancestral karyotype diverged into the 2n The genus Nannospalax includes blind rodents that have = 60W and R chromosomal races, and independent adapted to living underground. Currently more than 30 translocations of short arms of some chromosomes caused chromosomal races have been determined in Turkish blind this differentiation. Arslan et al. (2011) studied variation mole rats but there is still doubt about the taxonomy of this of C- and AgNOR-banding of 3 chromosomal races (2n = taxon (Nevo et al., 1994; Sözen et al., 1998a, 1998b, 1999, 40, 58, and 60) of N. xanthodon from southern Anatolia. 2000a, 2000b, 2006a, 2006b, 2011; Sözen, 2004; Matur and They found differentiation among the chromosomal Sözen, 2005; Kankiliç et al., 2007; Ivanitskaya et al., 2008; races. Matur et al. (2011) banded 4 chromosomal races Arslan et al., 2011). Kandemir et al. (2012) discussed the with 2n = 50 from different localities in Anatolia. These taxonomic name problem of blind mole rats and so here 4 chromosomal races had the same diploid numbers but we will follow Kandemir et al. (2012). their G-banding patterns were different. The complements Ivanitskaya and Nevo (1998) analyzed Jordanian blind of all these chromosomal races included 2 identical mole rats using C-, G-, and AgNOR-banding techniques metacentric autosomes and the sex chromosomes were and compared these data with previous results obtained also always the same. The studied chromosomal races in Turkish and Israeli blind mole rats. They found should have their own evolutionary pathway but they have that NF values were useful for differentiation due to a common ancestor. pericentric inversions and centromeric shifts. So far, only Dobigny et al. (2004) indicated that chromosomal a few banding studies of Turkish N. nehringi have been data have been underutilized in phylogenetic research, performed. These were conducted in populations from and chromosomal changes could be used as a character. Malatya (Ivanitskaya et al., 1997) and Kastamonu and We set out to identify chromosomal characters that Çankırı provinces (Ivanitskaya et al., 2008). Additionally, could be used to reconstruct the evolutionary history of a banding study was performed by Ivanitskaya et al. (1997) these blind mole rats. The aim of the present study was with southeastern Anatolian blind mole rats (N. ehrenbergi) to compare 9 chromosomal races of N. xanthodon and using G-, C-, and AgNOR-banding techniques. Ivanitskaya a chromosomal race from N. leucodon by determining et al. (2008) assigned the 2n = 60 populations in Turkey to which Robertsonian translocations are prevailing (fissions 2 chromosomal races as 2n = 60W and 2n = 60R, based vs. fusions). By finding the main chromosomal changing on G-bands, C-bands, AgNOR staining, fluorochrome mechanism, we may explain chromosomal evolution of staining, and FISH of telomeric and rDNA probes. They the genus Nannospalax in western Turkey. * Correspondence: [email protected] 470 MATUR et al. / Turk J Zool

2. Materials and methods and C- (Sumner, 1972) banding techniques were applied In this study, 33 animals were studied from 10 localities to each specimen. Pictures of metaphases were taken using (Table 1; Figure 1). According to their geographical a Canon DP 21 digital camera. location in Turkey, these chromosomal races were The G-banding patterns allowed us to assess all the designated as N for north (52N, 54N, 58N), W for west chromosomal homologies among chromosomal races and (50W), E for east (50E), Tr for Thrace (56Tr), C for central to identify the structural differences among karyotypes. (60C), and K for Kastamonu (60K). We recognized the 2n The 2n = 60J fromN. ehrenbergi from Jordan was used = 36, 2n = 38, 2n = 40, 2n = 50W, 2n = 52N, 2n = 54N, as an outgroup (Ivanitskaya and Nevo, 1998). In order 2n = 56W, 2n = 58N, and 2n = 60C chromosomal races to determine whether fusion or fission is the main from N. xanthodon and 2n = 56Tr from N. leucodon. rearrangement we identified the specific arm combination Karyotypes were prepared from bone marrow according of a particular metacentric (Figure 2). If an arm was to Ford and Hamerton (1956). Then G- (Seabright, 1971) included in different metacentrics, fusion was accepted

Table 1. Localities, sample size (M: males, F: females), diploid chromosome numbers (2n), and chromosomal arm numbers (NF) of animals examined.

2n NF Localities Province N 36 70 Kemer Cemetery AYDIN 3 38 74 Kırkağaç, Gelenbe MANİSA 3 40 72 Beyşehir 12 km SW KONYA 4 50W 74 Alaşehir MANİSA 5 52N 72 Yalova YALOVA 3 54N 70 Eflani KASTAMONU 2 56W 72 Kula 7 km S MANİSA 5 56Tr 78 Hayrabolu KIRKLARELİ 3 58N 72 Taşköprü KASTAMONU 3 60C 78 Kızılcasöğüt 1 km S UŞAK 2

30 35

56 Tr 38 40 36 56W 50W 52N 60K 54N 60 58N 100 0 100 200 300 km

Figure 1. Map of the study area in Turkey and the geographical distribution of the chromosomal races studied.

471 MATUR et al. / Turk J Zool

Populat on A Populat on B Populat on C

a c e a c a F ss on

b d f b d e f b c d e f

Metacentr c chromosomes

Populat on B Populat on C

a c d c e Populat on A

Fus on b e a b c d e f f d b a f

Acrocentr c chromosomes

Figure 2. Schematic diagram showing how to specify the rearrangement that plays a main role. If in a hypothetical Population A we always found “a” and “b” arms together like Population B and C then we can claim that fission is responsible for differentiation, or if the reverse situation is observed—“a” and “b” arms are combined with different arms in different condition such as Population B or Population C—it can be said the fusion is the main rearrangement responsible. as the responsible mechanism (references within Sumner 3. Results (2003)). We coded all the characters as indicated by Dobigny 3.1. Karyotype results of C-banding patterns and et al. (2004). First, the chromosomes or chromosomal G-banding comparisons segments were treated as characters, and their presence/ After C- and G-banding, heterochromatin variation in absence or the changes they had undergone represented the heterochromatin distribution (Figures 3a–j) and the character states. Secondly, the chromosomal changes rearrangements among chromosomal races (Figures themselves were considered to represent the characters. 4a–i) were identified. In karyotypes of the 2n = 60, 9 of Using both approaches, we treated 30 chromosome states 30 chromosomes were bi-armed. The X chromosome and 60 chromosomal rearrangements as absent/present. was submetacentric, while the Y chromosome was The matrix of chromosomal characters (Table 2) was subtelocentric. According to the C-band pattern of the 2n analyzed by maximum parsimony using the heuristic = 60, 16 pairs of chromosomes (pairs 1, 2, 3, 5, 6, 7, 10, search option in PAUP 4.0b.10 (Swofford, 2001) with 11, 14, 17, 18, 19, 20, 22, 23, and 26) had heterochromatin bisection–reconnection (TBR) and 10,000 random taxon areas (Figure 3a). The X chromosomes had a centromeric addition replicates. Bootstrap resampling (Felsenstein, heterochromatin area. 1985) was applied to assess the support for individual The karyotype of the 2n = 36 showed 17 pairs of bi- nodes using 10,000 bootstrap replicates with 100 random armed chromosomes. The X chromosome was a middle- additions. TBR was also conducted by both a NJ search sized submetacentric and the Y chromosome was a with 10,000 heuristic bootstrap analysis and NJ bootstrap small-sized acrocentric. C-band results showed interstitial analysis in PAUP 4.0b.10 (Swofford, 2001) dibranch blocks in 7 pairs (pairs 3, 6, 10, 11, 13, 14, and 17) and swapping. The karyotype preparations and animals centromeric heterochromatin in 2 pairs (pairs 7 and 12) examined were deposited in the Department of Biology, (Figure 3b). Two chromosomal rearrangements relative to Faculty of Arts and Sciences, Bülent Ecevit University. 2n = 38 were recognized from G-banding patterns. These

472 MATUR et al. / Turk J Zool

Table 2. The matrix of first 30 out of 90 chromosomal characters identified in Nannospalax and used for the phylogenetic reconstruction. The karyotype of 2n = 60 Jordanian (Ivanitskaya and Nevo, 1998) has been used as outgroup. These 30 characters are present/absent in chromosomes; others are rearrangement and every identified rearrangement is coded as 1.

2n/Chr.no. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30

60 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 40 1 0 1 0 0 1 1 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 1 1 1 50W 1 0 0 0 0 0 0 0 1 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 38 1 0 0 0 0 0 1 1 1 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 36 1 0 0 0 0 0 1 1 1 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 52N 1 1 0 0 0 1 1 1 1 0 1 1 1 1 0 1 1 1 0 1 0 1 0 1 1 1 1 1 1 1 56W 1 0 1 1 0 1 0 1 1 1 1 0 1 1 1 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 54N 0 1 0 0 1 0 1 1 1 0 0 0 0 0 0 1 1 1 1 1 1 0 0 0 1 1 0 1 1 1 58N 0 1 0 0 0 0 1 1 1 0 0 0 0 0 0 1 1 0 1 1 1 0 0 0 1 1 0 1 1 1 56Tr 1 1 1 0 0 0 0 1 0 0 0 1 1 0 0 1 1 1 1 1 1 0 0 0 1 0 0 1 1 1 60K 0 1 0 0 1 0 1 1 1 1 0 1 0 0 0 1 1 0 1 1 1 0 0 0 1 1 0 1 1 1 Out 1 0 1 0 0 0 0 1 0 1 0 1 0 0 0 1 0 1 1 0 0 0 0 0 0 0 1 0 1 0 were the inversion in pair 8 and a Robertsonian fusion: autosomes, resulting in NF =74. The X chromosome was Rb(8.16) (Figure 4a). a middle-sized submetacentric and the Y chromosome The NF value of the 2n = 38 was determined to be 74. The was a small-sized acrocentric. C-band results showed X chromosome was a middle-sized submetacentric, while heterochromatin blocks in 9 pairs (pairs 2, 3, 5, 7, 10, 11, the Y chromosome was a small-sized acrocentric. C-band 13, 18, and 22) (Figure 3e). G-banding results showed that patterns showed heterochromatin blocks in 11 pairs (pairs the chromosomal race also derives from the 2n = 38, as 1, 2, 3, 4, 5, 7, 8, 9, 10, 11, and 14), while 2 pairs (pairs 13 was the case for 2n = 36. Six Robertsonian fissions (Rbf) and 15) had pericentromeric heterochromatin blocks. The were responsible for the evolution of this chromosomal X chromosome had centromeric heterochromatin (Figure race. These were as follows: Rbf in 3, Rbf in 5, Rbf in 6, Rbf 3c). Comparison of the G-banding pattern led to the in 11, Rbf in 13, and Rbf in 14 (Figure 4d). recognition of rearrangements relative to the karyotype of The NF value of the 2n = 52 was 70. This chromosomal 2n = 60, i.e. pericentric inversions in pairs 2 and 3. Arms race had 19 pairs of bi-armed and 16 pairs of acrocentric forming the large metacentrics of 2n = 38 correspond to chromosomes. The X chromosome was a middle-sized 24 pairs of chromosomes of the 2n = 60 karyotype after submetacentric, while the Y chromosome was a small- Robertsonian fusion: Rb(2, 11), Rb(10, 16), Rb(15, 28), sized acrocentric. According to C-band results, 10 pairs Rb(3, 29), Rb(21, 27), Rb(4, 14), Rb(13, 20), Rb(5, 22), (pairs 1, 3, 4, 5, 6, 7, 8, 10, 19, and 21) had centromeric Rb(6, 19), Rb(17, 23), Rb(25, 30), and Rb(18, 24) (Figure heterochromatin blocks (Figure 3f). Comparison 4b). of the G-banding patterns led to the recognition of The 2n = 40 had 15 bi-armed chromosomes and 4 rearrangements relative to the karyotype of 2n = 60 acrocentric chromosomes, resulting in NF = 72. The X of pericentric inversions on pairs 1, 2, and 4. Four chromosome was a middle-sized submetacentric and Robertsonian fusions take place as follows: Rb(5, 21), the Y chromosome was a small-sized acrocentric. This Rb(3, 10), Rb(15,23), and Rb(4, 19) (Figure 4e). chromosomal race had C positive heterochromatin blocks The 2n = 54N populations were determined to be 2n = in 7 pairs (pairs 5, 7, 8, 11, 12, 13, and 14) (Figure 3d). The 54, NF = 74, NFa = 70. The X chromosome was a medium- individual chromosomal arms were correctly identified by sized submetacentric and the Y chromosome was a small- G-banding. The 10 pairs of metacentrics resulted from Rb sized acrocentric (Sözen et al., 2006). The autosomal set fusion of chromosomes identified in the 2n = 60. These consists of 3 groups: 6 pairs of meta-submetacentric, were as follows: Rb(12, 14), Rb(10, 11), Rb(4, 24), Rb(5, 3 pairs of subtelocentric, and 17 pairs of acrocentric 15), Rb(2, 21), Rb(16, 20), Rb(26, 27), Rb(13, 22), Rb(17, chromosomes. The C-banding pattern of the 2n = 54 19), and Rb(18, 23) (Figure 4c). shows centromeric heterochromatin on 14 pairs (pairs 4, In the karyotype of the 2n = 50, there were 11 pairs 6, 7, 8, 10, 11, 12, 13, 14, 15, 16, 17, 20, and 23) (Figure 3g). of bi-armed chromosomes and 13 pairs of acrocentric According to G-banding results, this chromosomal race

473 MATUR et al. / Turk J Zool

(a)

2 3 4 5 6

12 9 10 11 7 8

13 14 15 16 17 18

19 20 21 22 23 24

26 27 28 29 25 X Y

(b)

1 2 3 4 5 6

7 8 9 10 11 12

14 15 13

X X

16 17

474 MATUR et al. / Turk J Zool

(c)

3 4 5 6 1 2

7 8 9 10 11 12

15 16 13 14

X Y

17 18

(d)

3 4 5 1 2

8 9 6 7 10

13 14 15 11 12

16 17 18 19

X X

475 MATUR et al. / Turk J Zool

(e)

3 4 5 6 1 2

9 10 11 12 7 8

18 13 14 15 16 17

23 24 19 20 21 22

Y X

(f)

1 2 3 4 5 6

9 7 8 X Y

15 13 14 10 11 12

21 17 18 19 20 16

25 22 23 24

476 MATUR et al. / Turk J Zool

(g)

5 6 3 4 2 1

11 8 9 10 7

15 16 12 13 14 17

18 19 20

X X

(h)

3 4 2 1

X Y

5 6 7

10 11 12 13 9 8

15 16 17 18 19 14

20 21 22 23 24 25

27 26

477 MATUR et al. / Turk J Zool

(i)

6 5 2 3 4 1

7 8 X Y

13 14 9 10 11 12

17 18 19 20 15 16

21 22 23 24 25 26

(j)

5 6 4 3 2 1

10 11 12 7 8 9

16 17 18 13 14 15

24 22 23 19 20 21

25 26 27 28 X Y

Figure 3. C-banding results of the following chromosomal races: a) 2n = 60, b) 2n = 36, c) 2n = 38, d) 2n = 40, e) 2n = 50W, f) 2n = 52N, g) 2n = 54N, h) 2n = 56W, i) 2n = 56Tr, and j) 2n = 58N).

478 MATUR et al. / Turk J Zool

(a) 38-36

1 2 2 3 3 4 4 5 5 8 6 10

16 9 18 1 14 10 11 11 6 12 7

13 12 14 14 7 15 15 16 17 17

X X 19 18

(b) 38-60 1 2 3 4 5 7 6 11 16 28 29 27

1 2 20 3 21

9 12 10 11 19 8 12 14 20 22 19 1 13 5 4 6

14 15 16 13 8 23 30 18 7 9

17 25 24

17 18 26 X X

479 MATUR et al. / Turk J Zool

12 10

7 6 24 9 7 10 9 8 21 15

4 2 5

11 3 16 12 13 15 20 26 17 23

16 27 19 18

14 22 18 25 17 28 19 29 20 30 13

(d) 50W-38 10 12 22

5 1 10 17 5 2 19 11 8 9 1 9

15 16

8 7 3 3 4 2 6 6 13 20 14 11

4 14 19 12 13 15 21 16 24 18 23 17

X Y X Y

480 MATUR et al. / Turk J Zool

(e) 19 21 52N-60

6 6 4 9 5 2 3 3 4 5 1 1

11 12 8 7 8 9 10 7 13 12 14 23

15 18 16 16 17 17 18 11 13 25 14 28

19 20 20 29 21 27 22 22 23 30 24 24

25 26

X X

(f) 54-60K 5 4

1 2 3 3 4 6 5 7 6 8 7 9 8 10 1 2

13 11 14 13 10 12 1 16 14 17 15 18 16 19 9 11

21 25 22 24 28 17 21 18 22 19 23 20 2 26 23 27

25 26 29 20 27 30

481 MATUR et al. / Turk J Zool

(g) 56W-60 7

5 2 2 3 3 4 6 8 1 4 5 11

X X

10 11 10 7 8 12 9 6 11 9

13 19 16 24 12 27 14 13 1 15

1 22 18 21 20 25 21 18 22 23 17 1

26 28 27 29 28 30 23 17 24 20 25 26

(h) 60K -58N

1 1 2 2 3 3 4 16 5 6 7 4 9 8 5 6

9 10 10 12 12 8 13 14 14 7 15 1 16 2 17 17

22 21 21 1 23 2 2 11 2 24 18 13 19 1 2 20

X X

27 28 2 27 29 2 26 25 30 29

482 MATUR et al. / Turk J Zool

(i) 60 -60K

2 1 1 2 2 3 4 4 6 5 5 6 7 14

8 8 1 9 1 10 12 11 1 12 14 13 1 14

19 15 21 16 26 17 24 18 28 19 17 20 16 21

22 22 20 23 23 24 18 25 29 26 27 27 25 28

X X

30 29

9 30 Figure 4. Chromosome banding comparisons between the following chromosomal races: a) 2n = 38 and 2n = 36, b) 2n = 38 and 2n = 60, c) 2n = 40 and 2n = 60, d) 2n = 50W and 2n = 38, e) 2n = 52N and 2n = 60, f) 2n = 54N and 2n = 60K, g) 2n = 56W and 2n = 60, h) 2n = 58N and 2n = 60K, i) 2n = 60 and 60K. Rearrangements in figures were coded as 1 in the analyses. evolved from the 2n = 60K with 3 Robertsonian fusions, 3i). Both the X and Y chromosomes were negative for namely Rb(5, 1), Rb(4, 2), and Rb(12, 11) (Figure 4f). heterochromatin. Results from the karyotype of the 2n = 56W populations In the 2n = 58N, the NF = 74 and the NFa = 70. showed that this chromosomal race had 7 pairs of biarmed The X chromosome was a submetacentric and the Y chromosomes and 21 pairs of acrocentric chromosomes. chromosome was a small subtelocentric. The autosomal set The NF value was 72. The X chromosome was a middle- contains a pair of metacentric, 6 pairs of submetacentric, sized submetacentric and the Y chromosome was a small- and 20 pairs of acrocentric chromosomes. Eight pairs sized acrocentric. Blocks of C heterochromatin occur on 8 (pairs 6, 7, 18, 19, 20, 21, 22, and 23) had centromeric pairs (pairs 2, 4, 6, 7, 9, 14, 16, and 19) (Figure 3h). Arms C-positive heterochromatin blocks (Figure 3j). While forming the metacentric of the 2n = 56 correspond to 4 the Y chromosome was negative, the X chromosome was pairs of chromosomes of the karyotype of the 2n = 60C C-positive. According to the G-banding results, the 2n = after Robertsonian fusion, namely Rb(5, 7) and Rb(12, 16), 58N evolved from the 2n = 60K by a Robertsonian fusion and an inversion in 7 (Figure 4g). of 2 acrocentrics at Rb(6, 11) (Figure 4h). In the 2n = 56Tr, the NF = 78 and NFa = 74. The X According to G-banding results, we were able to identify chromosome was a submetacentric and the Y chromosome 10 deletions (pairs 2, 3, 4, 5, 12, 13, 22, 24, 25, and 27) and was a small-sized acrocentric. The autosomal set 4 centromeric shifts (pairs 10, 14, 17, 18) as rearrangement contains 10 pairs of bi-armed and 17 pairs of acrocentric between 2n = 60–60K comparison (Figure 4i). chromosomes. The C-banding patterns of the 2n = 3.2. Chromosomal phylogeny 56Tr chromosomal race showed positive staining in the We used chromosomal rearrangements as characters to centromeric region in 22 pairs (pairs 1, 2, 3, 4, 5, 6, 7, 8, 9, hypothesize on the phylogenetic relationships among 10, 11, 12, 13, 14, 15, 17, 18, 20, 21, 22, 25, and 26) (Figure chromosomal races. In this analysis, 57 informative

483 MATUR et al. / Turk J Zool characters were found. Figure 5 shows the most Abant, Mudurnu, Nallıhan, Seben, Yeniçağa, Kartalkaya, parsimonious tree found (tree length = 118, consistency and Mengen (Bolu), and by Matur and Sözen (2005) index = 0.75, CI excluding uninformative characters from Gölpazarı, Taraklı, Geyve, and Yenipazar (Bilecik). = 0.66). Five clades occurred. Clade 1, labeled as west, The new localities given here clarify the geographic range included the 2n = 36, 38, 40, and 50. In this clade, the 2n of this chromosomal race over a larger area. According = 38 and 36 were clustered together and the 2n = 50 was to previous records, the locality Yalova recorded here connected to them but the 2n = 40 was separated clearly is the most westerly point for the 2n = 52 chromosomal from them. Clade 2, labeled as Middle, included the 2n race. Clarifying the reasons for such a distribution and = 52, 56, and 60. In this clade, the 2n = 56W and 60 were separation of chromosomal races may help in identifying clustered together. Clade 3, labeled as north, included the the reasons for the parapatric and allopatric distribution 2n = 54, 58, and 60K (60R). Clade 4, labeled as Thrace, patterns found in blind mole rats. The River Sakarya, included the 2n = 56Tr, and the last clade, 5, included the flowing between the Yalova (NF = 72) and the Bolu and outgroup (2n = 60 Jordan). Bilecik populations (NF = 70), acts as a barrier between the 2 chromosomal races. These differentiations may have 4. Discussion occurred after the formation of the river. Two new localities for the 2n = 36 chromosomal race Populations with the 2n = 56 karyotype were found were recorded and its distributional area was extended from Uşak and Manisa. Previously, Kankılıç et al. (2010) southward to the town of Yatağan. Moreover, based on one recorded 2n = 56; in the present study, we expanded the male specimen from Yatağan, the Y chromosome of this distribution of this chromosomal race to Manisa. chromosomal race is determined as a small acrocentric for The 2n = 60C karyotype has mostly been recorded the first time (Figure 3). The 2n = 36 was first reported from central Anatolia (Sözen et al., 2006a; Kankılıç et al., by Sözen et al. (1999) from Bayındır in western Anatolia 2007). The 2n = 60C had several different NF values in its based on a single female specimen. Tez et al. (2002) gave distributional area (Sözen et al., 2006a). new distributional data for the 2n = 38 chromosomal race A comparative analysis of G- and C-banded from western Anatolia and did not find the 2n = 36 in chromosomes of N. xanthodon and N. leucodon from 11 western Anatolia. populations revealed a considerable degree of homology The 2n = 52, NF = 70 karyotype from northwestern between long arms of autosomes and a general similarity Anatolia was reported by Sözen (2004) from Karamürsel, in the amount of heterochromatin material. The patterns

94/97 60 94/94 56W

52N

-/70 54N

100/10 0 60K

58N

56tr

40 -/- 50W 100/10 0 88/81 38

36 out

0.03 Figure 5. Neighbor-joining tree based on chromosomal rearrangement in Nannospalax. Numbers on branches show neighbor joining results and bootstraps results. Node support from NJ and MP bootstrapping is shown for main groups. ‘–’ indicates nonsignificant support, <70% for bootstrap analyses.

484 MATUR et al. / Turk J Zool of rearrangements and, consequently, C-heterochromatin studied further. block locations were found to be important for Ivanitskaya et al. (1997) analyzed the chromosomal discriminating between chromosomal races. races of N. xanthodon and N. ehrenbergi from southern In Turkey, geographic features may act as an isolation Anatolia. They found that heterochromatin addition was mechanism and thus separate chromosomal races (Matur the main process along with pericentric inversions and and Sözen, 2005). Most chromosomal races are, however, Robertsonian rearrangements. In Israel, the 2n = 52 is the differentiated without a barrier. Although chromosomal ancestor of other chromosomal races of N. ehrenbergi due races are close together geographically in Turkey, to Robertsonian fissions, which turn chromosomal races hybridization does not occur (Nevo et al., 1994; Coşkun, into new ones (Nevo et al., 2001). Assuming that the 2n = 2003; Sözen, 2004; Matur and Sözen, 2005; Sözen et al., 60C is the ancestor of others, our results revealed that the 2006a, 2006b; Kankılıç et al., 2007; Ivanitskaya et al., Robertsonian fusion is the main mechanism responsible 2008). We neither recorded any hybrid specimens in the for chromosomal evolution in blind mole rats in the study area nor found polymorphism in G- and C-banding western part of Turkey. Furthermore, the Robertsonian patterns. Each individual from each of the respective fission, deletion, and inversion have a minor effect on chromosomal races was identical. This means that all the chromosomal evolution in blind mole rats. chromosomal races have been isolated from each other 4.1. Chromosomal evolution pathways and fixation or selection pressure is a strong evolutionary The 2n = 60C had the largest area of distribution (Figure force on the evolution of new chromosomal races. 6). Because of its distribution, this chromosomal race is Previous banding studies (Ivanitskaya et al., 1997, 2008) considered to be the ancestral chromosomal race to the suggested a high level of chromosomal divergence, which others. It gave rise to new races, and then chromosomal means that it is possible to consider these chromosomal differentiation might have acted as a postmating isolation races as having well-differentiated chromosomal lineage mechanism. New chromosomal races may reach fixation within the xanthodon group. Our phylogenetic tree and then differentiate as separate species. This supports indicates that each group (west, north, middle, and the process of peripatric speciation, which is considered to Thrace) is well differentiated (Figure 5). These groups be the common means of speciation for blind mole rats in should include one or more undescribed species when Turkey (Nevo, 1991; Sözen, 2004). With this assumption,

60 K 54 N 56 N 58 N 56 N 56Tr 50 N

58 52 N 54 C

38 60

56W

36 60

50W 56 40 58 S 56 S 52 S 56Ehr

Figure 6. Geographic distributions of chromosomal races of Nannospalax in western Turkey. In the figure the numbers indicate chromosome number, and the letters indicate the position of the chromosomal races (S: south; N: north; E: east, W: west; Tr: Thrace; C: central).

485 MATUR et al. / Turk J Zool we evaluated G- and C-banding results to develop and leucodon was closer to xanthodon rather than to scenarios for chromosomal evolution in blind mole rats. ehrenbergi. Hadid et al. (2012) also used molecular tools to We discuss possible scenarios for the general evolution of draw a maximum likelihood tree of blind mole rats from blind mole rats in west and north populations in Turkey in 3 different species. In their results, Aydın (2n = 36) and detail below. Beyşehir (2n = 40) grouped together. The 2n = 60K and 38 may have originated independently from 2n = 60C as peripheral populations 5. Conclusion at different times. While the main mechanism of the 2n = This study presents the first detailed banding results on 38 race differentiation was Robertsonian fusion, for the 2n blind mole rats in Turkey. Useful data for chromosomal = 60K it was deletion. However, deletions hardly play an evolution of blind mole rats in Turkey were collected. important role in chromosomal evolution; we think that Banding results indicated that the 2n = 60C is the additional forces (i.e. isolation from main population and ancestral chromosomal race and the 2n = 38 and 2n = selection pressure on small population) also caused the 60K are the secondary ancestral chromosomal races. differentiation. Then those 2 chromosomal races reached Moreover, we agreed with Ivanitskaya et al. (1997) that fixation. Later peripheral populations of the 2n = 60K Robertsonian fusions are the main mechanism responsible and 38 started to adaptively be distributed through new for chromosomal evolution of blind mole rats in Turkey. empty areas when they reached the place where other Additionally, we showed that Robertsonian fissions and chromosomal races were derived. The 2n = 36 and 50W pericentric inversions, and deletion are minor forces in may have been derived from the 2n = 38. While the 2n the chromosomal evolution of blind mole rats in Turkey. = 36 was derived by Robertsonian fusions (Figure 4a), Which factor causes these rearrangements is still not the 2n = 50W resulted from 6 fissions (Figure 4d). The conclusively known. On the other hand, recently published 2n = 40 was derived from the 2n = 60C but this race molecular studies (Krystufek et al., 2011; Hadid et al., includes the west group (Figure 5). However, the reason 2012; Kandemir et al., 2012) showed that differentiation why they occurred in the west group as a paraphyletic of chromosomal races of blind mole rats in Anatolia was group is still a mystery. In the north group, the 2n = already done and each chromosomal race might be a good 60K became the ancestral chromosomal race of the 2n candidate to be a separate species. However, more detailed = 54N and 58N after Robertsonian fusions. The north molecular studies using techniques such as genome group evolved as a monophyletic group. Sözen (2004) analysis and mtDNA are still required. indicated that the 2n = 50N, 54N, 56N, and 58N from northern Turkey had identical chromosomal morphology. Acknowledgments According to our parsimony tree results, N. xanthodon This study was supported financially by TÜBİTAK (TBAG grouped polyphyletically, because the 2n = 56Tr (N. HD–164–101T225) and Bülent Ecevit University (Nr: leucodon) grouped together with an Anatolian sample 2008–13–06–01) grants. We also thank the 3 anonymous (N. xanthodon). Kandemir et al. (2012) investigated the reviewers for their contribution during the evaluating phylogeny of Turkish blind mole rats using molecular process. This study is a part of the karyological and tools. They found similar tree topology. Chromosomal cytogenetical results presented in the PhD thesis of Ferhat races with small diploid numbers were grouped together, Matur (2009).

References

Arslan, A., Akan, Ş. and Zima, J. 2011. Variation in C-heterochromatin Hadid, Y., Németh, A., Snir, S., Pavlíček, T., Csorba, G., Kázmér, and NOR distribution among chromosomal races of mole rats M., Major, Á., Mezhzherin, S., Rusin, M., Coşkun, Y., Nevo, (Spalacidae) from Central Anatolia, Turkey. Mamm. Biol. 76: E. 2012. Is evolution of blind mole rats determined by climate 28–35. oscillations? PLoS One 7(1): 30043. Coşkun, Y. 2003. A study on the morphology and karyology of Ivanitskaya, E., Coskun, Y. and Nevo, E. 1997. Banded karyotypes Nannospalax nehringi (Satunin, 1898) (Rodentia: Spalacidae) of mole rats (Spalax, Spalacidae, Rodentia) from Turkey: a from North-eastern Anatolia, Turkey. Turk. J. Zool. 27: 171–176. comparative analysis. J. Zool. Sys. Evol. Res. 35: 171–177. Ivanitskaya, E. and Nevo, E. 1998. Cytogenetics of mole rats of the Dobigny, G., Ducroz, J.F., Robinson, T.J. and Volobouev, V. 2004. Spalax ehrenbergi superspecies from Jordan (Spalacidae, Cytogenetics and cladistics. Syst. Biol. 53: 470–484. Rodentia). Z. Saeuget. 63: 336–346. Felsenstein, J. 1985. Confidence limits on phylogenies: an approach Ivanitskaya, E., Sozen, M., Rashkovetsky, L., Matur, F. and Nevo E using the bootstrap. Evolution 39: 783–791. 2008. Discrimination of 2n = 60 Spalax leucodon cytotypes Ford, C. and Hamerton, J. 1956. A colchicine hypotonic citrate, squash (Spalacidae, Rodentia) in Turkey by means of classical and sequence for mammalian chromosomes. Stain Tech. 31: 247– molecular cytogenetic techniques. Cytogenet. Genome Res. 251. 122: 139–149.

486 MATUR et al. / Turk J Zool

Kankilic, T., Kankilic, T., Colak, R., Colak, E. and Karatas, A. 2007. Sozen, M. and Kivanc, E. 1998a. Two new karyotypic forms of Spalax Karyological comparison of populations of the Spalax leucodon leucodon (Nordmann, 1840) (Mammalia: Rodentia) from Nordmann, 1840 superspecies (Rodentia: Spalacidae) in Turkey. Z Saugetierkd. -Mamm. Biol. 63: 307–310. Turkey. Zool. Middle East 42: 15–24. Sozen, M. and Kivanc, E. 1998b. A new karyotype of Spalax leucodon Kankılıç, T., Kankılıç, T., Seker, P.S., Çolak, R., Selvi, E. and Çolak E. cilicicus Mehely, 1909 (Mammalia: Rodentia) from the type 2010. Contributions to the karyology and distribution areas of locality in Turkey. Israel J. Zool. 44: 53–55. cytotypes of Nannospalax leucodon (Rodentia: Spalacidae) in Sozen, M., Colak, E., Yigit, N., Ozkurt, S. and Verimli, R. 1999. Western Anatolia. Acta Zool. Bulg. 62: 161–167. Contributions to the karyology and taxonomy of the genus Kandemir, İ., Sözen, M., Matur, F., Martinkova, N., Kankılıç, T., Spalax Guldenstaedt, 1770 (Mammalia: Rodentia) in Turkey. Çolak, F., Özkurt, Ş.Ö., and Çolak, E. 2012. Phylogeny of Z. Sauget. 64: 210–219. species and cytotypes of mole rats (Spalacidae) in Turkey Sozen, M., Colak, E. and Yigit, N. 2000a. Contributions to the inferred from mitochondrial cytochrome b gene sequences. karyology and taxonomy of Spalax leucodon nehringi Folia Zool. 61(1): 25–33. Satunin, 1898 and Spalax leucodon armeniacus Mehely, 1909 Kryštufek, B., Ivanitskaya, E., Arslan, A., Arslan, E. and Bužan, E.V. (Mammalia: Rodentia) in Turkey. Z. Sauget. 65: 309–312. 2011. Evolutionary history of mole rats (genus Nannospalax) Sözen, M., Yiğit, N. and Çolak, E. 2000b. A study on karyotypic inferred from mitochondrial cytochrome b sequence. Biol. J. evolution of the genus Spalax Guldenstaedt, 1770 (Mammalia: Linnean Soc. 105(2): 446–455. Rodentia) in Turkey. Israel J. Zool. 46: 239–242. Matur, F. and Sozen, M. 2005. A karyological study on subterrranean Sozen, M., Sevindik, M. and Matur, F. 2006a. Karyological and some mole rats of the Spalax leucodon Nordmann, 1840 (Mammalia: morphological characteristics of Spalax leucodon Nordmann, Rodentia) superspecies in northwestern Turkey. Zool. Middle 1840 (Mammalia: Rodentia) superspecies around Kastamonu East 36: 5–10. province, Turkey. Turk. J. Zool. 30: 205–219. Matur, F., Colak, F., Sevindik, M. and Sozen, M. 2011. Chromosome Sozen, M., Matur, F., Colak, E., Ozkurt, S. and Karatas, A. 2006b. differentiation of four 2n = 50 chromosomal forms of Turkish Some karyological records and a new chromosomal form for mole rat, Nannospalax nehringi. Zool. Sci. 28: 61–67. Spalax (Mammalia: Rodentia) in Turkey. Folia Zool. 55: 247– Nevo, E. 1991. Evolutionary theory and processes of active speciation 256. and adaptive radiation in subterranean mole rats, Spalax Sözen, M., Çataklı, K., Eroğlu, F., Matur, F. and Sevindik, M. 2011. ehrenbergi superspecies, in Israel. Evol. Biol. 25: 1–125. Distribution of chromosomal forms of Nannospalax nehringi Nevo, E., Filippucci, M.G., Redi, C., Korol, A. and Beiles, A. 1994. (Satunin, 1898) (Rodentia: Spalacidae) in Çankırı and Çorum Chromosomal speciation and adaptive radiation of mole-rats provinces, Turkey. Turk. J. Zool. 35(3): 367–374. in Asia Minor correlated with increased ecological stress. Proc. Sumner, A. 1972. A simple technique for demonstrating centromeric Natl. Acad. Sci. USA 91: 8160–8164. heterochromatin. Exp. Cell. Res. 75: 304–306. Nevo, E., Ivanitskaya, E. and Beiles, A. 2001. Adaptive radiation of Sumner, A. 2003. Chromosomes: Organization and Function. blind subterranean mole rats: naming and revisiting the four Blackwell Publishing, Oxford, UK. sibling species of the Spalax ehrenbergi superspecies in Israel: Spalax galili (2n = 52), S. golani (2n = 54), S. carmeli (2n = Swofford, D. 1999. PAUP*. Phylogenetic Analysis Using Parsimony (* 58) and S. judaei (2n = 60). Bachkhuys Publishers, Leiden, The and other Methods). In. Sunderland, MA: Sinauer Associates. Netherlands. Tez, C., Gunduz, I. and Kefelioglu, H. 2002. Note: New data on the Seabright, M. 1971. A rapid banding technique for human- distribution of 2n = 38 Spalax leucodon (Nordmann, 1840) chromosomes. Lancet 2: 971–972. cytotype in Turkey. Isr. J. Zool. 48: 155–159. Sozen, M. 2004. A karyological study on subterranean mole rats of the Spalax leucodon Nordmann, 1840 superspecies in Turkey. Mamm. Biol. 69: 420–429.

487