Karyotype (2n) and allozymc divrrsity at 37 gene loci \vrrc determined in ti9 subterrmean tilde rats in Turkry helongirig to the two supcrspccics: the ancestor Spa1n.x lur~rudorr (11 = 55; 20 populations) and the descendant S. t~/rierrhqr (11 = 14; liwr populations). \Ve idciitificd remarkable variation rif diploid clirnmos~itii~ niniil)rrs in the S. Ieucodirrr suprrspccics: 2n = :18. 10. 50, 3, 60 and 62: and in thr S. ulrrenhtrg? supersprcirs: 2rr = 52. 56 and 58. Genrtic diversity indices were low 011 average in Imth S. Itirrridon and ,S. ehrtnbrrp supcrspccirs: Allrle diversity, . 4 = 1.081 and 1.074; polymorphism. F5Y; = 0.077 and 0.068; hctcroqgosity. II = 0.038 ;nid 0.027; and gcnc diversity, H, = 0.038 and 0.034, rrspeclivcly. H ranged from 0 in nirsir or srmimcsic rc+ns to 0.088 in arid Anatolia.W c considrr the populations with dilterent diploid chr(imr)some nuinhers, 2n. as good thloqical spccies. Kar);ot)pic diversity may tnark rxtcnsivc ecological speciation. Nei's grnctic distances. I ) (average 0.174, range 0.002 0.422) and ecngeographical criteria suqyst that a l~n~s t e;ich popitlation may rrprcsriit a diferrnt biological species, but c-ritiral ruturc trsting is necessary to support this claim. Kalyot)prs and allozymes are ti on ran do nil^ distlihutcd across 'l'urkey. displaying rcinar-kahle correlations \s\ith climatic and biotic Factors. Both 2rr and H are significantly correlated with aridity stress (2n/rainlill, r, = -0.74; P < 0.001), and in our regon also with climatic unpredictahility.Thrsc results support the niche-width genetic \ariatinu l~ypotl~csis in space and time. Climatic selection in 'I'urkey appears to he a inajor architect of kai7otype and genetic ~allozymc) diversity and divcrgcncc in mcik rat cv(iIiitioii, i n both speciation and adaptation.