A comprehensive phylogeny of extinct and extant Rhizomyinae (Rodentia): evidence for multiple intercontinental dispersals Raquel López-Antoñanzas, Lawrence Flynn, Fabien Knoll

To cite this version:

Raquel López-Antoñanzas, Lawrence Flynn, Fabien Knoll. A comprehensive phylogeny of extinct and extant Rhizomyinae (Rodentia): evidence for multiple intercontinental dispersals. Cladistics, Wiley, 2013, 29 (3), pp.247 - 273. ￿10.1111/j.1096-0031.2012.00426.x￿. ￿hal-01920777￿

HAL Id: hal-01920777 https://hal.archives-ouvertes.fr/hal-01920777 Submitted on 17 Dec 2020

HAL is a multi-disciplinary open access L’archive ouverte pluridisciplinaire HAL, est archive for the deposit and dissemination of sci- destinée au dépôt et à la diffusion de documents entific research documents, whether they are pub- scientifiques de niveau recherche, publiés ou non, lished or not. The documents may come from émanant des établissements d’enseignement et de teaching and research institutions in France or recherche français ou étrangers, des laboratoires abroad, or from public or private research centers. publics ou privés. Cladistics

A comprehensive phylog eny of extinct and extant Rhizomyinae (Rodentia): evidence for multiple intercontinental dispersals

Journal:For Cladistics Peer Review

Manuscript ID: CLA-12-05-0553.R2

Manuscript Type: Article

Date Submitted by the Author: n/a

Complete List of Authors: López-Antoñanzas, Raquel; Museo Nacional de Ciencias Naturales-CSIC, Paleobiology Flynn, Lawrence; Harvard University, Peabody Museum Knoll, Fabien; Museo Nacional de Ciencias Naturales-CSIC, Paleobiology

Paleontology, Phylogeny, Systematics, Taxonomy, Dispersal < Keywords: Biogeography < Applications

Cladistics Page 1 of 68 Cladistics

1 2 3 Abstract 4 5 6 The subfamily Rhizomyinae is known from the Late Oligocene up to the present. 7 8 Today, this group comprises six species, which live in southern Asia and eastern Africa. 9 10 Despite the current moderate diversity of the rhizomyines, they had a greater 11 diversification and wider distribution in the past from Asia, their land of origin, to 12 13 Africa where they entered during the Early Miocene. So far 33 fossil species can be 14 15 referred to this group. A cladistic analysis involving fossil and living species has been 16 carried out. Prokanisamys spp. turned out to be the most basal taxa of the ingroup. This 17 18 analysis calls theFor monophyly Peer of several genera Review into question and allows proposing a 19 20 phylogenetic definition of the tribes Tachyoryctini and Rhizomyini. It also provides 21 information about the origin of the African rhizomyines and allows inferring multiple 22 23 dispersal phenomena from Asia to Africa in Early and Late Miocene times. 24 25 26 Keywords: Mammalia; Systematics; Cladistics; Cenozoic; Eurasia; Africa 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Cladistics Cladistics Page 2 of 68

1 2 3 Table of contents 4 5 Introduction 6 7 Material and methods 8 9 Systematics 10 11 Genus Rhizomys Gray, 1831 12 13 Rhizomys sinensis Gray, 1831 14 15 16 Rhizomys sumatrensis (Raffles, 1821) . 17 18 Rhizomys pruinosusFor Blyth, 1851Peer. Review 19 20 Rhizomys (Brachyrhizomys) shansius (Teilhard de Chardin, 1942) . 21 22 23 Rhizomys (Brachyrhizomys) shajius Flynn, 1993. 24 25 Genus Cannomys Thomas, 1915a. 26 27 Cannomys badius (Hodgson, 1841) . 28 29 30 Genus Tachyoryctes (Rüppell, 1836) . 31 32 Tachyoryctes splendens Rüppell, 1836 . 33 34 Tachyoryctes macrocephalus (Rüppell, 1842) . 35 36 37 Tachyoryctes pliocaenicus Sabatier, 1978 . 38 39 Tachyoryctes konjiti Sabatier, 1982 . 40 41 Tachyoryctes makooka Wesselman, Black et Asnake, 2009 . 42 43 44 Genus Anepsirhizomys Flynn, 1982a. 45 46 Anepsirhizomys opdykei Flynn, 1982a . 47 48 Anepsirhizomys pinjoricus (Hinton, 1933) . 49 50 51 Genus Miorhizomys Flynn, 2009. 52 53 Miorhizomys nagrii (Hinton, 1933) . 54 55 Miorhizomys micrus (Flynn, 1982a) . 56 57 58 Miorhizomys blacki (Flynn, 1982a) . 59 60 Cladistics Page 3 of 68 Cladistics

1 2 3 Miorhizomys pilgrimi (Hinton, 1933) . 4 5 Miorhizomys harii (Prasad, 1968) . 6 7 Miorhizomys tetracharax (Flynn, 1982a) . 8 9 10 Miorhizomys choristos (Flynn, 1982a) . 11 12 Genus Eicooryctes Flynn, 1982 . 13 14 Eicooryctes kaulialensis Flynn, 1982 . 15 16 17 Genus Protachyoryctes Hinton, 1933. 18 For Peer Review 19 Protachyoryctes tatroti Hinton, 1933 . 20 21 22 Genus Rhizomyides Bohlin, 1946. 23 24 Rhizomyides sivalensis (Lydekker, 1884) . 25 26 Rhizomyides carbonnelli (Brandy, 1979) . 27 28 Rhizomyides platytomeus Flynn, Heintz, Sen et Brunet, 1983 . 29 30 31 Rhizomyides mirzadi (Lang et Lavocat, 1968) . 32 33 Rhizomyides punjabiensis (Colbert, 1933) . 34 35 36 Rhizomyides lydekkeri (Hinton, 1933) and Rhizomyides saketiensis (Gupta, Verma et 37 Tewari, 1978) . 38 39 40 Genus Kanisamys Wood, 1937. 41 42 Kanisamys indicus Wood, 1937 . 43 44 45 Kanisamys sivalensis Wood, 1937 . 46 47 Kanisamys nagrii Prasad, 1968 . 48 49 Kanisamys potwarensis Flynn, 1982a . 50 51 52 Genus Nakalimys Flynn et Sabatier, 1984. 53 54 Nakalimys lavocati Flynn et Sabatier, 1984 . 55 56 Genus Pronakalimys Tong et Jaeger, 1993. 57 58 59 60 Cladistics Cladistics Page 4 of 68

1 2 3 Pronakalimys andrewsi Tong et Jaeger, 1993. 4 5 Genus Prokanisamys de Bruijn, Hussain et Leinders, 1981. 6 7 Prokanisamys arifi de Bruijn, Hussain et Leinders, 1981 . 8 9 10 Prokanisamys kowalskii ( Lindsay, 1996 ). 11 12 Prokanisamys benjavuni ( Mein et Ginsburg, 1985 ). 13 14 Prokanisamys major Wessels et de Bruijn, 2001 . 15 16 17 Prokanisamys sp. Wessels, Fejfar, Peláez-Campomanes et de Bruijn, 2003 . 18 For Peer Review 19 Phylogenetic analysis 20 21 22 Previous work 23 24 New analysis 25 26 Discussion 27 28 29 Prokanisamys 30 31 The African Pronakalimys and Nakalimys 32 33 Kanisamys 34 35 36 Protachyoryctes 37 38 Rhizomyides 39 40 “Rhizomyides ” mirzadi and “ R”. punjabiensis 41 42 43 Miorhizomys 44 45 The crown group 46 47 Tribe Tachyoryctini. 48 49 Tribe Rhizomyini. 50 51 52 Conclusion 53 54 Acknowledgements 55 56 57 References 58 59 60 Cladistics Page 5 of 68 Cladistics

1 2 3 4 López-Antoñanzas, R. et al. 1 5 6 7 8 A comprehensive phylogeny of extinct and extant 9 10 Rhizomyinae (Rodentia): evidence for multiple 11 12 intercontinental dispersals 13 14 Formatted: English (U.S.) 15 1 2 1 16 Raquel López-Antoñanzas , Lawrence J. Flynn and Fabien Knoll 17 18 1 For Peer Review 19 Departamento de Paleobiología, Museo Nacional de Ciencias Naturales-CSIC, c/ 20 José Gutiérrez Abascal 2, Madrid 28006, Spain. 21 2Peabody Museum of Archaeology and Ethnology, Harvard University, Cambridge, 22 23 MA 02138. E-mails (RLA): [email protected] ; (LJF): [email protected], 24 (FK): [email protected] 25 26 27

28 29 30 Running title:Phylogeny of rhizomyineRhizomyinae 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Cladistics Cladistics Page 6 of 68

1 2 3 4 López-Antoñanzas, R. et al. 2 5 6 7 8 Introduction 9 10 11 Living Rhizomyinae are fossorial muroid rodents from southern Asia and eastern 12 Africa (Fig. 1) that belong to the family Spalacidae (Jansa and Weksler, 2004; 13 14 Gogolevskaya et al., 2010 ). Despite their seceretive way of life, they are important to 15 the people who live in the same zones because, like other rodents (e.g., the cane rat 16 17 Thryonomys; López-Antoñanzas et al., 2004), they are agricultural pests, but also can be 18 a source of protein. For Peer Review 19 20 The living Asian forms (bamboo rats ) include Rhizomys (a trispecific genus) and 21 Cannomys (a monospecific genus). There is a single living African genus Tachyoryctes, 22 23 the mole rat, which includes 2 extant species, although that diversity may be greatly 24 underestimated (Musser and Carleton, 2005). No fossil of Cannomys has been found to 25 26 date. However, the Asian Rhizomys (Brachyrhizomys ) and the African Tachyoryctes are 27 known since Late Miocene time. Although the rhizomyines are today moderately 28 29 diversified, this subfamily, which is known from Late Oligocene times, has had a 30 flourishing history as reflected by the finding discovery of over 30 fossil species 31 32 distributed in 11 genera ( Pronakalimys , Nakalimys , Prokanisamys , Kanisamys , 33 Rhizomyides , Protachyoryctes , Tachyoryctes , Eicooryctes , Miorhizomys, 34 Anepsirhizomys Rhizomys Brachyrhizomys 35 and ( )) in Africa, Thailand, Afghanistan, 36 Pakistan, India, and China. 37 The systematic position of the rhizomyines is controversial. Some authors have 38 39 placed them in a distinct Family Rhizomyidae (Miller and Gidley, 1918; Ellermann, 40 1940, 1941; Simpson, 1945; Ellerman and Morrison-Scott, 1951, 1966; Colbert and 41 42 Hooijer, 1953; Flynn 1982a, 1982b, 1990, 1993; Flynn and Guo-qinQi , 1982; Flynn and 43 Sabatier, 1984; Flynn et al., 1983, 1995; Alemseged and Geraads, 2000; Wessels and de 44 45 Bruijn, 2001), including the tachyoryctines with them. Other authors considered the 46 rhizomyines as members (Lydekker, 1885; Palmer, 1897; Colbert, 1935) or a subfamily 47 48 (Thomas, 1896; Tullberg, 1899; Musser and Carleton, 2005; Flynn, 2009) of the 49 Spalacidae. As Winge (1887) originally did when he created the Tribe Rhizomyini, 50 51 some other specialists considered the rhizomyines as a tribe within the Muridae. 52 McKenna and Bell (1997) considered the rhizomyines a subfamily within the Muridae 53 54 and subdivided it into two tribes: the Rhizomyini and the Tachyoryctini. The results of 55 56 57 58 59 60 Cladistics Page 7 of 68 Cladistics

1 2 3 4 López-Antoñanzas, R. et al. 3 5 6 7 recent analyses of mitochondrial and nuclear genes support a monophyletic Spalacinae 8 9 plus Rhizomyinae, with Myospalacinae as sister group (Jansa and Weksler, 2004; 10 Gogolevskaya et al., 2010) and leave little doubt about the inclusion of the rhizomyines 11 12 in the Family Spalacidae.The results of recent analyses of nucleotide sequences support 13 a monophyletic Spalacinae plus Rhizomyinae, with Myospalacinae as sister group 14 15 (Norris et al., 2004; Gogolevskaya et al., 2010; see also Jansa and Weksler, 2004) and 16 leave little doubt about the inclusion of the rhizomyines in the Family Spalacidae. 17 18 The similarity in many morphologicalFor Peertraits among the threeReview subfamilies probably 19 stems from their common origin ( Gogolevskaya et al., 2010 ) as well as convergence to 20 21 a fully subterranean way of life through geological time. 22 However, the relationships within the rhizomyines (Asian bamboo rats and African 23 24 mole rats) remain unclear. Some authors considered that the Asian rhizomyines are 25 distantly related to the African representatives of the group (Endo et al., 2001; Musser 26 27 and Carleton, 2005) and, therefore, they separate them into two different subfamilies: 28 the Rhizomyinae (Asian) and the Tachyoryctinae (African). Other rodentologists 29 30 considered that African fossil rhizomyines do not appear far from the Asian ones from a 31 morphological viewpoint and they include them in a single subfamily, the Rhizomyinae 32 33 (McKenna and Bell, 1997; Flynn, 2009; López-Antoñanzas, 2012; López-Antoñanzas 34 and Wesselmann, in press). Furthermore, the origin of Tachyoryctes is controversial. 35 Some authors advocated an independent African origin for this taxon (Ellerman, 1940, 36 37 1941; Lavocat, 1978; Endo et al., 2001), whereas others defended an Asian origin 38 (Black, 1972; Jacobs, 1978; Sabatier, 1978; Munthe, 1980; De Bruijn et al., 1981; 39 40 Flynn, 1982a, 1986, 1990, 2009; Flynn et al., 1990; Wessels and De Bruijn, 2001; 41 Haile-Selassie et al., 2004; Wesselman et al., 2009). All these yet unresolved issues 42 43 motivated us to provide a substantiated hypothesis of the mutual relationships within 44 this subfamily based on cladistic methods. 45 46 47 48 49 Material and methods 50

51 52 The study presented herein is based on the examination of original specimens, casts, and Formatted: Indent: First line: 0" 53 data from the literature. We examined the following: skulls of extant Rhizomys sinensis 54 55 56 57 58 59 60 Cladistics Cladistics Page 8 of 68

1 2 3 4 López-Antoñanzas, R. et al. 4 5 6 7 (ZMB 16768, ZMB 17413, ZMB 45362, ZMB 43372, ZMB 43373, ZMB 17411, ZMB 8 9 16770, ZMB 17417, ZMB 16766 in the MB and C.G.1934-1237, C.G.1912-851 in the 10 MNHN), R. pruinosus (DH. 280, C.G. 1929-259, C.G. 1929-261, C.G.1892-1360, C.G. 11 12 1929-263, C.G.1958-724, C.G.1962-2297, C.G.1962-2295, C.G. 1929-266, C.G. 1929- 13 272, C.G. 1929-260, C.G. 1929-267, C.G. 1993-48, C.G. 1929-264, C.G. 1937-1251, 14 15 C.G. 1929-265, C.G. 1929-270, C.G. 1929-340, C.G. 1993-02 in the MNHN), R. 16 sumatrensis (ZMB 1847, ZMB A1316, ZMB 38840, ZMB 48085, ZMB 21249 in the 17 18 MB C.G.1993-49, C.G. 1962-2294,For C.G. Peer2007-162, C.G. 1980-235,Review A 7.216, A7.215 in 19 the MNHN), Cannomys badius (ZMB 3424, ZMB 44768, ZMB 44769 in the MB and 20 21 CG-2000-761, CG-1860-382 in the MNHN), Tachyoryctes macrocephalus (ZMB 22 36656, ZMB 55165, ZMB 36649 in the MB and C.G. 1972-73, C.G. 1972-74, C.G. 23 24 1972-75, C.G. 1972-76, C.G. 1974-202 in the MNHN), T. splendens (ZMB 12, ZMB 25 78, ZMB 79, ZMB 80, ZMB 81, ZMB 83, ZMB 88, ZMB 114, ZMB 117, ZMB 3404, 26 27 ZMB 3403, ZMB 9015, ZMB 16048, ZMB 28069, ZMB 36601-13, ZMB 36618, ZMB 28 36620-36621, ZMB 36624-36634, ZMB 36636-36637, ZMB 36640-36646, ZMB 29 30 36648, ZMB 36650-36655, ZMB 46257, ZMB 48300, ZMB 72559-72593, ZMB 7721 31 in the MB and C.G. 1993-217, C.G. 1993-231, C.G. 1993-220, C.G. 1993-238, C.G. 32 33 1933-2752, C.G. 2000-685, C.G. 1961-33, C.G. 1972-10, C.G. 1972-27, C.G. 1972-39, 34 C.G. 1972-51, C.G. 1972-64, C.G. 1972-65, C.G. 1972-68, C.G. 1993-222, C.G. 1993- 35 225, C.G. 1993-241, C.G. 1993-246, C.G. 1993-250, C.G. 1993-253, C.G. 1993-254, 36 37 C.G. 1993-257, C.G. 1993-263, C.G. 1993-265, C.G. 1993-268, C.G. 1993-272, C.G. 38 1993-274, C.G. 1993-289, C.G. 1993-302, C.G. 1993-303, C.G. 1993-308, C.G. 1993- 39 40 323, C.G. 1933-2746, C.G. 1933-2756, C.G. 1905-385, C.G. 1933-2745, C.G. 1933- 41 2748, C.G. 1960-454, C.G. 1962-2292, C.G. 1970-210, C.G. 1972-16, C.G. 1972-19, 42 43 C.G. 1972-44, C.G. 1972-46, C.G. 1972-50, C.G. 1972-54, C.G. 1972-58, C.G. 1993- 44 221, C.G. 1993-223, C.G. 1993-232, C.G. 1993-238, C.G. 1993-245, C.G. 1993-281, 45 46 C.G. 1993-288, C.G. 1993-294, C.G. 1993-304, C.G. 1993-307, C.G. 1993-318, C.G. 47 1993-322, C.G. 1960-453, C.G. 1962-2290, C.G. 1972-14, C.G. 1972-25, C.G. 1972- 48 49 29, C.G. 1972-41, C.G. 1972-57, C.G. 1974-202, C.G. 1991-629, C.G. 1993-217, C.G. 50 1993-224, C.G. 1993-239, C.G. 1993-240, C.G. 1993-249, C.G. 1993-261, C.G. 1993- 51 52 270, C.G. 1993-292b, C.G. 1993-295, C.G. 1993-296, C.G. 1993-316, C.G. 1961-921, 53 C.G. 1972-11, C.G. 1972-15, C.G. 1972-17, C.G. 1972-24, C.G. 1972-33, C.G. 1972- 54 55 56 57 58 59 60 Cladistics Page 9 of 68 Cladistics

1 2 3 4 López-Antoñanzas, R. et al. 5 5 6 7 37, C.G. 1972-45, C.G. 1972-52, C.G. 1972-53, C.G. 1972-69, C.G. 1981-533, C.G. 8 9 1988-119, C.G. 1993-226, C.G. 1993-227, C.G. 1993-228, C.G. 1993-230, C.G. 1993- 10 235, C.G. 1993-237, C.G. 1993-242, C.G. 1993-247, C.G. 1993-259, C.G. 1993-260, 11 12 C.G. 1993-267, C.G. 1993-283, C.G. 1993-284, C.G. 1993-286, C.G. 1993-287, C.G. 13 1993-309, C.G. 1993-310, C.G. 1993-315, C.G. 1993-317, C.G. 1993-320, C.G. 1933- 14 15 2747, C.G. 1972-06, C.G. 1972-21, C.G. 1972-26, C.G. 1972-30, C.G. 1972-34, C.G. 16 1972-43, C.G. 1972-59, C.G. 1972-62, C.G. 1972-66, C.G. 1972-67, C.G. 1993-231, 17 18 C.G. 1993-234, C.G. 1993-236,For C.G. 1993-256, Peer C.G. 1993-275, Review C.G. 1993-278, C.G. 19 1993-285, C.G. 1993-290, C.G. 1993-291, C.G. 1993-300, C.G. 1993-305, C.G. 1993- 20 21 314, C.G. 1962-2291, C.G. 1972-09, C.G. 1972-12, C.G. 1972-31, C.G. 1972-35, C.G. 22 1972-40, C.G. 1972-47, C.G. 1972-61, C.G. 1988-242, C.G. 1993-219, C.G. 1993-220, 23 24 C.G. 1993-248, C.G. 1993-251, C.G. 1993-258, C.G. 1993-262, C.G. 1993-273, C.G. 25 1993-276, C.G. 1993-282, C.G. 1993-299, C.G. 1993-313, C.G. 1993-321, C.G. 1933- 26 27 2749, C.G. 1972-18, C.G. 1972-20, C.G. 1972-23, C.G. 1972-42, C.G. 1972-48, C.G. 28 1972-55, C.G. 1972-56, C.G. 1972-71, C.G. 1993-218, C.G. 1993-229, C.G. 1993-243, 29 30 C.G. 1993-264, C.G. 1993-269, C.G. 1993-271, C.G. 1993-277, C.G. 1993-279, C.G. 31 1993-292, C.G. 1993-297, C.G. 1993-298, C.G. 1993-306, C.G. 1993-311, C.G. 1993- 32 33 319, C.G. 1993-324, C.G. 1972-07, C.G. 1972-08, C.G. 1972-13, C.G. 1972-22, C.G. 34 1972-28, C.G. 1972-32, C.G. 1972-36, C.G. 1972-38, C.G. 1972-49, C.G. 1972-60, 35 C.G. 1972-63, C.G. 1972-70, C.G. 1993-233, C.G. 1993-244, C.G. 1993-252, C.G. 36 37 1993-255, C.G. 1993-266, C.G. 1993-280, C.G. 1993-293, C.G. 1993-301, C.G. 1993- 38 312, C.G. 1933-2752, in the MNHN), isolated teeth, maxillary fragments and mandible 39 40 fragments of the following extinct species: Prokanisamys arifi (casts of HGSP-116-22, 41 HGSP-116-371, HGSP-116-83, HGSP-116-101, HGSP-116-123, HGSP-116-125, 42 43 HGSP-116-14, HGSP-116-37, HGSP-116-42, HGSP-116-52, HGSP-116-61 in the 44 BSP), Kanisamys indicus (casts of GSI-D271 in the BSP), K. nagrii (casts of AMNH 45 46 30.000 and AMNH 39323 in the BSP), K. potwarensis (AMNH 39321), Rhizomyides 47 punjabiensis (AMNH 19762 and cast of GSI-D287 in the BSP), R. sivalensis (casts of 48 49 D97, D275 and BM (NH) 15926 in the BSP) Protachyoryctes tatroti (cast of GSI-D272 50 in the BSP), Anepsirhizomys pinjoricus (casts of GSI-D278 and GSI-D280 in the BSP), 51 52 Miorhizomys nagrii (casts of GSI-D273 and GSI-D285 in the BSP), M. tetracharax 53 (upper incisor of AMNH 39326, cast of AMNH 39325 in the BSP), M. pilgrimi (lower 54 55 56 57 58 59 60 Cladistics Cladistics Page 10 of 68

1 2 3 4 López-Antoñanzas, R. et al. 6 5 6 7 incisor of AMNH 39327 and cast of D-278 in the BSP). In addition, we reviewed 8 9 unpublished original material of the following taxa recovered under the Harvard 10 University-Geological Survey of Pakistan field program on the Potwar Plateau, Pakistan 11 12 from 1994-2000: Prokanisamys major, Kanisamys indicus, K. potwarensis, K. nagrii, K. 13 sivalensis , and Rhizomyides punjabiensis . 14 15 First, second, and third lower molars are designated as m1, m2, and m3, respectively, 16 and first, second, and third upper molars as M1, M2, and M3, respectively. The 17 18 terminology used in the toothFor descriptions Peer follows the rodent Review dental terminology of 19 Mein and Freudenthal (1971), with some adjustments, and Flynn (1982a) (see Fig. 2). 20 21 The cladistic analysis treated as ingroup all extant and fossil species of this subfamily 22 known to date. The taxonomic units chosen are the valid species of Rhizomyini and 23 24 Tachyoryctini, namely Rhizomys sumatrensis , R. sinensis , R. pruinosus , R. 25 (Brachyrhizomys) shansius , R. (Brachyrhizomys) shajius , Cannomys badius , 26 27 Miorhizomys nagrii , M. micrus , M. blacki , M. pilgrimi , M. harii , M. tetracharax , M. 28 choristos , Protachyoryctes tatroti , Tachyoryctes makooka , T. konjiti, T. pliocaenicus , T. 29 30 macrocephalus , T. splendens , Rhizomyides sivalensis , R. punjabiensis , R. carbonnelli , 31 R. platytomeus , R. mirzadi, Kanisamys indicus , K. nagrii , K. sivalensis , K. potwarensis , 32 33 Eicooryctes kaulialensis , Anepsirhizomys opdykei , A. pinjoricus , Pronakalimys 34 andrewsi , Nakalimys lavocati , Prokanisamys kowalskii , P. arifi , P. benjavuni and P. 35 major , P. sp. (Jebel Zelten), (see Supplementary File 1”Systematics” below ). 36 37 Debruijnia arpati , a basal Spalacinae according to the phylogenetic analysis of Flynn 38 (2009), is selected as outgroup. A total of 44 phylogenetically informative characters 39 40 (mainly of dental morphology) have been coded (Supplementary File 1). 24 characters 41 are binary, whereas 20 are multistate. Owing to the lack of a priori information, all Formatted: Font: Italic 42 43 characters were unordered and equally weighted (Fitch optimality criterion). As some 44 species are known so far from only a few specimens, the influence of intraspecific 45 46 variation in the scoring of the characters could not be assessed. 47 1. Hypsodonty: Crown height on at least one side of the tooth greater than length or 48 49 width of the tooth. This character has been scored as (0) absent; (1) moderate; (2) high. 50 2. Size: Length of the m2: (0) 1.5-2.5 mm; (1) 2.5-3.5 mm; (2) 3.5-4.5 mm; (3) 4.5- 51 52 5.5 mm; (4) >5,5 mm 53 3. Lophodonty: (0) weak; (1) moderate; (2) high 54 55 56 57 58 59 60 Cladistics Page 11 of 68 Cladistics

1 2 3 4 López-Antoñanzas, R. et al. 7 5 6 7 4. Masseteric crest: (0) with long anterior extension (1) with short anterior 8 9 extension (2) without anterior extension 10 5. Lower masseteric crest: (0) inflated under m2; (1) not inflated 11 12 6. Dentary depth: (0) Shallow; (1) Moderate; (2) Deep 13 7. Infraorbital foramen: (0) with ventral slit; (1) with abbreviated ventral slit; (2) 14 15 without ventral slit 16 8. Longitudinal ornamentation on the i: (0) double ridge; (1) single ridge; (2) 17 18 absent For Peer Review 19 9. Anteroconid on the m1: (0) discernible); (1) absent (fused with the anterolophid) 20 21 10. Mure on the m1: (0) present; (1) constricted; (2) absent. 22 11. Labial anterolophid on the m1: (0) present; (1) absent 23 24 12. Mesolophid on the m1: (0) long and complete and separated from the 25 hypolophid; (1) short and well-separated from the hypolophid; (2) short and migrated 26 27 towards the hypolophid; (3) absent; (4) a long continuation of the protoconid 28 13. Posterior protoconid-metaconid connexion on the m1: (0) present; (1) absent 29 30 14. Anterosinusid on the m1: (0) present; (1) absent 31 15. Mure on the m2: (0) present; (1) constricted; (2) absent 32 33 16. Mesolophid on the m2: (0) long and complete and separated from the 34 hypolophid; (1) short and well-separated from the hypolophid; (2) short and migrated 35 towards the hypolophid; (3) absent; (4) a long continuation of the protoconid 36 37 17. Labial anterolophid on the m2: (0) present; (1) absent 38 18. Hypolophid on the m2: (0) not isolated; (1) isolated 39 40 19. Protosinusid on the m2: (0) present; (1) absent 41 20. Anterosinusid on the m2: (0) present; (1) absent 42 43 21. m3 : (0) Reduced (shorter than the m2); (1) Enlarged (equal or larger than the 44 m2) 45 46 22. Labial anterolophid on the m3: (0) present; (1) absent 47 23. Mure on the m3: (0) present; (1) absent 48 49 24. Mesolophid on the m3 : (0) long and complete and separated from the 50 hypolophid; (1) short and well-separated from the hypolophid; (2) short and migrated 51 52 towards the hypolophid; (3) absent; (4) a long continuation of the protoconid 53 54 55 56 57 58 59 60 Cladistics Cladistics Page 12 of 68

1 2 3 4 López-Antoñanzas, R. et al. 8 5 6 7 25. Posterosinusid on the m3: (0) large; (1) small; (2) absent (posterolophid fused 8 9 with hypolophid) 10 26. Entoconid on the m3: (0) not isolated; (1) isolated 11 12 27. Protosinusid on the m3: (0) present; (1) absent 13 28. Anterosinusid on the m3: (0) present; (1) absent 14 15 29. Roots on the M1: (0) three; (1) four; (2) Hypselodont. The development of four 16 roots on the M1 is a modern Rhizomyinae trait 17 18 30. Protosinus on the M1:For (0) present; Peer (1) absent Review 19 31. Mesoloph on the M1: (0) absent; (1) short; (2) long and complete; (3) long and 20 21 complete but divided with a buccal cusp 22 32. Mesostyle on the M1: (0) may be present; (1) absent 23 24 33. Ectoloph on the M1: (0) present; (1) absent 25 34. Anterocone on the M1: (0) distinct; (1) fused with the anteroloph 26 27 35. Metaloph on the M1: (0) present and distinct from the posteroloph; (1) fused 28 with the posteroloph 29 30 36. Posterosinus on the M1: (0) present; (1) absent 31 37. Mesoloph on the M2: (0) absent; (1) short; (2) long and complete; (3) long and 32 33 complete but divided with a buccal cusp 34 38. Longitudinal crest on the M2: (0) present; (1) absent 35 39. Ectoloph on the M2: (0) present; (1) absent 36 37 40. Metaloph on the M2: (0) present and distinct from the posteroloph; (1) early 38 fused with the posteroloph; (2) absent 39 40 41. Posterosinus on the M2: (0) present; (1) absent 41 42. Anterosinus on the M3: (0) open; (1) enclosed by anteroloph and paracone; (2) 42 43 absent 44 43. Mesoloph on the M3: (0) absent; (1) short; (2) long and complete; (3) long and 45 46 complete but divided with a buccal cusp 47 44. Posterior lake on the M3: (0) large; (1) small; (2) absent 48 49 The data matrix (Table 1) was built under Mesquite 2.6 (Maddison and Maddison, 50 2009) and processed with TNT (Goloboff et al., 2008) with the "traditional search" 51 52 option (using TBR). Branch support was estimated through two complementary indices: 53 54 55 56 57 58 59 60 Cladistics Page 13 of 68 Cladistics

1 2 3 4 López-Antoñanzas, R. et al. 9 5 6 7 Bremer Support (Bremer, 1994) and Relative Bremer Support (Goloboff and Farris, 8 9 2001). 10

11 12 Abbreviations : 13

14 15 AMNH: American Museum of Natural History, New York; BAM: Bamian, 16 Afghanistan; BMNH: The Natural History Museum, London, England; BSP: 17 18 Bayerische StaatssammlungFor für Paläontologie Peer und Historische Review Geologie, Munich, 19 Germany; C.G: Catalogue général du Laboratoire des Mammifères et Oiseaux, MNHN, 20 21 Paris, France; DMR: Department of Mineral Ressources, Bangkok, Thailand; FSL: 22 Université Claude Bernard-Lyon 1, Villeurbanne, France; GSI, Geological Survey of 23 24 India, Calcutta, India; H-GSP: Howard University-Geological Survey of Pakistan; 25 IVAU: Department of Earth Sciences, Utrecht University, Utrecht, Netherlands. IVPP: 26

27 Institute of Vertebrate Palaeontology and Palaeoanthropology, Beijing, China; KNM: 28 Kenya National Museums, Nairobi, Kenya; MNHN: Muséum national d’Histoire 29 30 naturelle, Paris, France; NME: National Museum of Ethiopia, Addis Ababa, Ethiopia; 31 PMAE: Peabody Museum of Archaeology and Ethnology, Cambridge, USA; PMNH: 32 33 Pakistan Museum of Natural History, Islamabad, Pakistan; PEC: Pul-e Charkhi, Kabul, 34 Afgahnistan; SMF: Mammalogie-Sammlung, Forschungsinstitut und Naturmuseum 35 Senckenberg, Frankfurt am Main; UM2: Université des Sciences et Techniques du 36 37 Languedoc-Montpellier 2, Montpellier, France; Y-GSP: Yale-Geological Survey of 38 Pakistan, Quetta, Pakistan; YPM: Yale Peabody Museum of Natural History, New 39 40 Haven, USA; Z: Zinda Pir area, Pakistan; ZMB: Zoologische Sammlung, Museum für 41 Naturkunde der Humboldt-Universität, Berlin, Germany; 42 43 44 45 46 Systematics 47 48 49 Asian rhizomyines are widespread. They live in China, northern and northeastern 50 India, Nepal, Bhutan, southeastern Bangladesh, Myanmar, Laos, Thailand, Cambodia, 51 52 Vietnam, Malaysia and Sumatra (Fig. 1) where they are known from a wide variety of 53 habitats (from bamboo forest to cultivated land, and up to 4000 m elevation) (IUCN, 54 55 56 57 58 59 60 Cladistics Cladistics Page 14 of 68

1 2 3 4 López-Antoñanzas, R. et al. 10 5 6 7 2010). African rhizomyines are nowadays discontinuously distributed in the centre and, 8 9 especially, the eastern part of the continent (Democratic Republic of the Congo, 10 Uganda, Rwanda, Burundi, Tanzania, Kenya, Ethiopia, Somalia and possibly South 11 12 Sudan and Eritrea) (Fig. 1) and they are also recorded from a wide variety of habitats 13 (tropical moist forest and open woodland to savanna, grasslands and agricultural and 14 15 pasture areas, sometimes above 4000 m) (IUCN, 2010). 16 Musser and Carleton (1993, pp. 685, 2005, pp. 913–915) have provided a census of 17 18 the extant species of RhizomysFor and Cannomys Peer recognized asReview valid (see also Ellerman 19 and Morrison-Scott, 1951, pp. 550–553, 1966, pp. 521–522). They concluded that 3 20 21 species of Rhizomys and a single species of Cannomys are well defined at present: 22 Rhizomys sinensis, R. pruinosus, R. sumatrensis and Cannomys badius . The taxonomy 23 24 of Tachyoryctes is more controversial and the number of species included in this taxon 25 ranges from 14 (Allen, 1939, pp. 361–363; Ellerman, 1941) to 2 (Misonne, 1971, p. 7; 26 27 Rahm, 1980, pp. 23–27; Bucher, 1982, p. 478; Corbet and Hill, 1991). Musser and 28 Carleton (1993, pp. 685-687; 2005, pp. 922-926) considered that the drastic reduction 29 30 from 14 to 2 species deserves systematic attention , that has never been performed and 31 therefore they cautiously recognized 11 (Musser and Carleton, 1993) and or 13 (Musser 32 33 and Carleton, 2005) species. However, the recent morphometric approach given by 34 Beolchini and Corti (2004) supports the hypothesis of Misonne (1971) of recognizing 35 only 2 species in the genus Tachyoryctes : T. macrocephalus and T. splendens . Because 36 37 our analysis is morphological, we evaluate these two species only , although we admit 38 that T. splendens might actually be a cryptic species complex . 39 40 The morphology of the teeth of fossil rhizomyines has been described in detail 41 (Colbert, 1933, 1935; Teilhard de Chardin, 1942; Bohlin, 1946; Colbert and Hooijer, 42 43 1953; Black, 1972; Jacobs, 1978; Sabatier, 1978, 1979, 1982; Brandy, 1979a, 1979b; 44 Munthe, 1980; de Bruijn et al., 1981; Flynn, 1982a, 1982b, 1983, 1993, 2009; Flynn 45 46 and Guo-qinQi , 1982; Wessels et al., 1982, 2003, 2008; Flynn et al., 1983, 1990; Sen, 47 1983; Flynn and Sabatier, 1984; Mein and Ginsburg, 1985, 1997; Tong and Jaeger, 48 49 1993; Lindsay, 1996; Wessels and de Bruijn, 2001; Wesselman et al., 2009; López- 50 Antoñanzas and Wesselman, in press). Therefore, in this work, we only describe the 51 52 teeth of the three modern genera and give a short historical background for the fossil 53 ones as well as some pertinent additional details. 54 55 56 57 58 59 60 Cladistics Page 15 of 68 Cladistics

1 2 3 4 López-Antoñanzas, R. et al. 11 5 6 7

8 9 Genus Rhizomys Gray, 1831. Type species: Rhizomys sinensis Gray, 1831 10 Referred species: Rhizomys pruinosus , R. sumatrensis 11 12 13 On the basis of a skull and skin of a glirine quadruped brought from China by the 14 15 amateur naturalist J. Reeves, Gray (1831, p. 95) described the new genus Rhizomys as 16 follows: “ Dentes primores 2/2 maximi, elongati, triangulares, acutati: molares 3/3 3/3 17 18 radicati, subcylindrici, coronisFor transversim Peer subparallelim porcatis;Review superiores internè 19 lobati. Caput magnum. Oculi parvi aperti. Auriculæ nudæ conspicuæ. Corpus crassum 20 21 subcylindricum. Pedes breves validi, digitis 5-5. Cauda mediocris, crassa, nuda” . That 22 is: Incisors 2/2 very large, elongate, triangular, pointed: molars 3/3 three rooted, 23 24 subcylindrical, with crowns with transversal subparrallel furrows; upper teeth lobated 25 internally. Head large. Eyes small open. External ears naked visible. Body massive 26 27 subcylindrical. Feet short robust, with 5-5 digits. Tail average, thick, naked. 28 29 30 Rhizomys sinensis Gray, 1831 . Gray (1831) recognized the new genus and new 31 species of rodent Rhizomys sinensis . The brief Latin diagnosis he gave (Gray, 1831: 32 33 95):“Rhiz. pallidè cinerascens unicolor ” can be translated as follows: “Rhiz. pale ash- 34 grey single-colored”. The type specimens are BMNH 49b (skin) and BMNH 1161a 35 (skull) (P. Jenkins, pers. comm. 2012). 36 37 The molars of Rhizomys sinensis are short rooted and lack the mure. They show 38 unilateral hypsodonty and a strong wear gradient with first molars well worn by the 39 40 time the third molars erupt (Fig. 3A-B). The orange upper incisors are recurved and 41 ventrally directed and lack major ornamentation. M2 and m2 are the largest teeth of the 42 43 upper and lower tooth row, respectively. The dental pattern of the upper molars is 44 tetralophodont with four lophs (anteroloph, protoloph, mesoloph, and 45 46 metaloph+posteroloph). By wear, the labial reentrants are close d-off and three enamel 47 lakes are isolated. The central one may be confluent with the lingual reentrant, forming 48 49 a narrow and elongated central reentrant that may divide the tooth into two parts. The 50 lower molars are lower crowned than the upper ones. M1 undergoes heavy wear earlier 51 52 than posterior molars; the latter continue to erupt past its occlusal surface until, in old 53 age, the molars are worn to the same level. The lower incisors lack major ornamentation 54 55 56 57 58 59 60 Cladistics Cladistics Page 16 of 68

1 2 3 4 López-Antoñanzas, R. et al. 12 5 6 7 and their enamel is flat. The mandibular foramen is well posterior to the distal part of 8 9 m3 (located a little above the m3 at the level of the tip of the coronoid process). The 10 mental foramen is situated anteriorly to the anterior part of m1 at about the midline of 11 12 the dentary. The pattern of m1 is pentalophodont (anterolophid, metaconid, mesolophid, 13 hypolophid, and posterolophid) whereas m2 and m3 show four major lophs 14 15 (anterolophid-metalophid, mesolophid, short hypolophid, and long posterolophid). With 16 wear, the lingual and labial reentrants close off, isolating various enamel lakes, the 17 18 number of which dependingFor depends on thePeer state of wear ofReview the teeth. 19 20 21 Rhizomys sumatrensis (Raffles, 1821) . Raffles (1821, p. 258) named this species as 22 Mus sumatrensis on the basis of a drawing and a specimen forwarded from Malacca 23 24 (Malasya) by W. Farquhar. The reason he chose sumatrensis as a species name when, 25 by that time, it had not been found in Sumatra remains unknown. He gave the following 26 27 diagnosis (Raffles, 1821, p.258): “The body is about seventeen inches in length, ten 28 inches in circumference, and the height at the shoulder about five. The tail is six inches 29 30 long, tapering, and blunt at the points, naked and scaly. The body is covered with rough 31 greyish hair, brownish on the back. The head round and lighter coloured. Incisors large, 32 33 two in each jaw. Eyes small. Ears naked. Fore-feet four-toed; hind-feet with a short fifth 34 toe.” Later on, Gray (1831, p. 95) included Mus sumatrensis in the new genus Rhizomys 35 and he gave the following brief Latin diagnosis: “Pallidè fuscus, pilis raris albidis 36 37 interspersis; corporis lateribus pedibusque saturatioribus; genis pallidioribus, occipite 38 nigrescenti lineâ longitudinali albâ, pectore albido ”. This can be translated as: “Pale 39 40 brown, with some hairs scattered whitish; with body broad and feet darker; with cheeks 41 paler, back of the head blackish with a white longitudinal stripe, chest whitish”. The 42 43 type specimen was possibly lost in transit (P. Jenkins, pers. comm. 2012). 44 The teeth of Rhizomys sumatrensis and R. sinensis are quite similar and only a few 45 46 differences can be discerned.The molars of R. sumatrensis are high crowned, although 47 not as high as those of R. sinensis , but they have longer roots and much smaller M3 than 48 49 R. sinensis . Furthermore, M1 of R. sumatrensis is the largest tooth, whereas in R. 50 sinensis it is M2. 51 52 53 54 55 56 57 58 59 60 Cladistics Page 17 of 68 Cladistics

1 2 3 4 López-Antoñanzas, R. et al. 13 5 6 7 Rhizomys pruinosus Blyth, 1851 . This species was named by Blyth (1851, p. 519) 8 9 after the examination of specimens collected by R. Frith in the Khasi hills (Cherrapunji, 10 India). Blyth differentiated this species from R. sumatrensis in being “much less robust, 11 12 having a much shorter tail, and a dense coat of fine soft fur instead of a thin coat of 13 bristly fur”. However, he pointed out that their structural characters were essentially the 14 15 same. The syntypes are supposed to be in the collection of the Indian Museum, Kolkata 16 (P. Jenkins, pers. comm. 2012), but we were unable to obtain confirmation. 17 18 The dental pattern of theFor teeth of R. pruinosusPeer is similar Reviewto that of the remaining 19 species of the genus. There is only a minor difference concerning the smaller size of the 20 21 molars. 22 23 24 Rhizomys (Brachyrhizomys) shansius (Teilhard de Chardin, 1942). Teilhard de 25 Chardin (1942), on the basis of four lower mandibles from the Pliocene of Taoyang 26 27 (Yushe Basin, Shanxi, China) named the new genus Brachyrhizomys , which was seen as 28 a primitive type of Rhizomys , and the new species B. shansius . Later on, Flynn (2009) 29 30 after a complete and detailed study of this species concluded that, even though it is 31 distinct from all living species of Rhizomys , it is sufficiently closely related to them for 32 33 Brachyrhizomys to be treated as a subgenus within Rhizomys (see Fig. 4F-G). The 34 holotype of this species, IVPP 31.096, a right dentary with incisor, is housed in the 35 IVPP. Additional material of this species has been found at other localities of the Yushe 36 37 Basin (Flynn, 1993; Flynn, 2009). 38

39 40 Rhizomys (Brachyrhizomys) shajius Flynn, 1993. The holotype and only known 41 specimen of this taxon (V8920), a right dentary fragment with m2-m3 comes from the 42 43 Late Miocene YS 156 locality of the Tancun subbasin (south of Yushe) (Flynn, 1993). 44 Smaller than B. shansius , it is housed in the IVPP. 45 46 47 Genus Cannomys Thomas, 1915 a. Type species: Cannomys badius (Hodgson, 1841) 48 49 50 Cannomys badius (Hodgson, 1841) . This species was coined as Rhizomys badius by 51 52 Hodgson (1841, p. 60) on the basis of a male specimen obtained from “some miles 53 north of the great valley” of Nepal. He gave the following brief Latin 54 55 56 57 58 59 60 Cladistics Cladistics Page 18 of 68

1 2 3 4 López-Antoñanzas, R. et al. 14 5 6 7 diagnosis:” Unicolor: pilis intern ē nigri cantibus schistaceis: longitudo Corporis 8 8 9 uncia, caudæ 2 ⅝, capitis 2 ¼: Pedis antici (cum ungue) 1 1/16 : pedis postici 1 ⅜, 10 auriculæ ¼”. This can be translated as: “One-colored: with hairs internal blackish slate- 11 12 colored: length of body 8 inches, tail 2 ⅝, head 2 ¼: anterior foot (with claw) 1 1/16 : 13 posterior foot 1 ⅜, external ear ¼”. The type specimen is BMNH 1843.1.12.61 (skin and 14 15 skull) (P. Jenkins, pers. comm. 2012). 16 Later on, Thomas (1915a) created the new genus Cannomys for which he took 17 18 Rhizomys badius as type, butFor in which he Peer also included two Review species now seen as junior 19 synonyms. He gave the following diagnosis: “Size comparatively small. Palm and sole- 20 21 pads normal, well defined, not granulated. Mammæ 2-2=8. Anteorbital foramina and 22 posterior nares much as in Rhizomys . Incisors forming a segment of a large circle, their 23 24 points thrown strongly forwards. Molars decreasing in size backwards, the first 25 decidedly larger than the second, its grinding surface not worn lower than that of the 26 27 latter”. The species Cannomys pater coined by Thomas (1915b) is today also considered 28 a junior synonym of C. badius (Musser and Carleton, 2005). 29 30 The upper molars of Cannomys badius show unilateral hypsodonty. The upper 31 incisors are proodont and lack major ornamentation. The dental pattern of the M1 is 32 33 tetralophodont (protocone-anteroloph, protocone-protoloph, hypocone-mesoloph and 34 hypocone-metaloph-posteroloph), whereas the M2 and M3, much shorter than the M1, 35 only have three transverse lophs (protocone-anteroloph-protoloph, hypocone-mesoloph 36 37 and hypocone-metaloph-posteroloph) (Fig. 3D). Through wear, the lophs join at the 38 margin of the tooth isolating two transversely elongated enamel lakes. The lower molars 39 40 are lower crowned than the upper ones. The lower incisors lack major ornamentation 41 and their enamel is flat. The mandibular foramen is well posterior to the distal side of 42 43 the m3 (it is located a little higher than the m3 at the level of the extremity of the 44 coronoid process). The mental foramen is situated anteriorly to the anterior side of the 45 46 m1 at about the midline of the dentary. The dental pattern of the m1 is pentalophodont 47 with five transverse lophids (anterolophid, protoconid-metalophid, protoconid- 48 49 mesolophid, hypoconid-hypolophid and hypoconid-posterolophid), whereas that of the 50 m2 and m3 has three lophids (anterolophid-metalophid, mesolophid and hypolophid- 51 52 posterolophid) (Fig. 3C). By wear, all lingual reentrants are closed off forming lingual 53 enamel lakes. 54 55 56 57 58 59 60 Cladistics Page 19 of 68 Cladistics

1 2 3 4 López-Antoñanzas, R. et al. 15 5 6 7

8 9 Genus Tachyoryctes (Rüppell, 1836). Type species: Tachyoryctes splendens (Rüp pell, 10 1836) 11 12 Referred species: Tachyoryctes macrocephalus , Tachyoryctes konjiti, Tachyoryctes 13 pliocaenicus, Tachyoryctes makooka. 14 15 16 Tachyoryctes splendens Rüppell, 1836 . Rüp pell (1836, p.36) named a new species of 17 18 rodent that he discovered inFor Ethiopia (Fig. Peer 5). In his work heReview pointed out that this rodent 19 resembled very much Bathyergus from Cape of Good Hope in the skull, crown and size 20 21 of the teeth, toes and claws. However, he also noticed some singularities in the 22 Ethiopian animal like the small size of its pinna and its longer and more rounded tail. 23 24 Since very few specimens of the genus Bathyergus were known at that time, he 25 refrained from naming a new genus, even though unnamed eminent London er and 26 27 Parisian experts thought otherwise. Nevertheless, in case someone wanted to transfer his 28 species into a new genus, he added as a footnote the following thought: “Will man mit 29 30 aller Gewalt diese Thier zu einer eigenen Gattung erheben, so dürfte man solche 31 passender Weise Tachyoryctes benennen, von ταχύς, schnell, und όρύκτης , Gräber ”, 32 33 which means (“If one would by any means erect a new genus for this animal, he might 34 name it appropriately Tachyoryctes from ταχύς, fast, and όρύκτης , burrower” ). Rüppell 35 (1842) reallocated this species to the genus Rhizomys including in it the new taxon 36 37 Rhizomys macrocephalus (see below) as well. The genus Tachyoryctes was accepted in 38 Heck (1851a, p.464) in a section written by C. Girard (Heck, 1851b, p. v). Heuglin and 39 40 Fitzinger (1866, pp.32-33) and later authors also accepted Tachyoryctes as valid and 41 included both T. splendens and T. macrocephalus in it. The lectotype of T. splendens is 42 43 SMF 4317 (head and skin). 44 45 46 Tachyoryctes splendens has ever growing teeth showing unilateral hypsodonty. 47 Upper and lower incisors are orange and flattened, and the lower incisor has a 48 49 prominent ridge. The proodont upper incisors may also bear a fine longitudinal ridge. 50 The upper molars lack the mure and the the anterior part of the teeth is isolated from the 51 52 posterior one. When unworn, the M1 show a long anteroloph, a distinct and isolated 53 protoloph, a long mesoloph that joins the hypocone and a metacone that connects 54 55 56 57 58 59 60 Cladistics Cladistics Page 20 of 68

1 2 3 4 López-Antoñanzas, R. et al. 16 5 6 7 labially to a short posteroloph. The morphology of the M2 and M3 is similar to that of 8 9 the M1 except for the anterior part of the teeth in which there is a single anterior crest 10 (anteroloph-protoloph) (Fig. 3F). Rapidly during wear, the lophs join at the margin of 11 12 the tooth isolating three (on the M1) or two (on the M2 and M3) enamel lakes, which 13 will disappear with further attrition. When unworn, the m1 shows a cuspid-like isolated 14 15 anterolophid, an isolated metaconid, the entoconid connected to the protoconid and an 16 arcuate posterolophid, which is a long continuation of the hypoconid. The m2 and m3 17 18 have the anterolophid fusedFor with the metaconid Peer and the teeth Review show three oblique lophids 19 (anterolophid-metaconid, entoconid-protoconid and hypoconid-posterolophid). The 20 21 mure is lacking (Fig. 3G). 22 23 24 Tachyoryctes macrocephalus (Rüppell, 1842) . Rüppell (1842) named the new species 25 Rhizomys macrocephalus on the basis of a rodent from Shoah province Province 26 27 (southern Ethiopia) that was sent to him. The lectotype is SMF 4318 (skull and skin). 28 This species was soon reallocated to the genus Tachyoryctes (see above). The 29 30 morphology of the cheek teeth of T. macrocephalus is similar to that seen in T. 31 splendens . The difference between these taxa rests in the much greater size of the 32 33 former . 34 35 Tachyoryctes pliocaenicus Sabatier, 1978 . This species was named on the basis of 36 37 eight mandibles (some of them incomplete) and a fragmentary skull with right and left 38 M1-M2 from the Pliocene Hadar Formation, Ethiopia (Sabatier, 1978). The holotype, a 39 40 right mandible with m1-m3 (AL.288.29), is housed in the NME. 41

42 43 Tachyoryctes konjiti Sabatier, 1982 . This taxon was named by Sabatier (1982; see 44 also Sabatier, 1979) on the basis of a hundred isolated molars from the Pleistocene 45 46 locality of Garba VIII (Melka Kunturé, Ethiopia). Its holotype (Garba VIII B 16), an 47 isolated M1, is housed in the NME. Additional material of T. konjiti has been found in 48 49 other Pleistocene Ethiopian localities as Garba IV, Gombore I γ, Simbiro, Gombore II 50 B/C, Garba III E, Garba X and Garba VIIIB (Sabatier, 1979; Galloti et al ., 2010). 51 52 53 54 55 56 57 58 59 60 Cladistics Page 21 of 68 Cladistics

1 2 3 4 López-Antoñanzas, R. et al. 17 5 6 7 Tachyoryctes makooka Wesselman, Black et Asnake, 2009 . This taxon was coined on 8 9 the basis of various mandibles, maxillary fragments, isolated molars as well as some 10 postcranial elements (Wesselman et al., 2009). Its holotype comes from the Late 11 12 Miocene Digiba Dora Vertebrate Locality 1 (Middle Awash, Ethiopia) and is housed in 13 the NME. This species has been also recovered from various Late Miocene-Pliocene 14 15 localities of the Middle Awash (ALA-VP-2, AME-VP-1, AMW-VP-1, ASK-VP-3, 16 GAW-VP-1, GAW-VP-3, KUS-VP-1, STD-VP-2, WKH-VP-1) (Wesselman et al., 17 18 2009). The morphology of Forthe teeth of T. Peermakooka has been Review described in detail by 19 López-Antoñanzas and Wesselman (in press). 20 21 22 Genus Anepsirhizomys Flynn, 1982. Type species: Anepsirhizomys opdykei Flynn, 23 24 1982 25 Referred species: Anepsirhizomys pinjoricus 26 27 28 Anepsirhizomys opdykei Flynn, 1982 . This taxon was named on the basis of a left 29 30 dentary fragment with m1-m2 from the middle Pliocene locality DP70 (Bunha River 31 area, Potwar Plateau, Pakistan) (Flynn, 1982a). Additional material of this species has 32 33 not been found to date. 34 35 Anepsirhizomys pinjoricus (Hinton, 1933) . Hinton (1933), in the course of the 36 37 preparation of a monograph on Indian fossil rodents, which was never actually 38 published, studied some rodent remains belonging to the GSI. This work led him to 39 40 discover various new species of Rhizomys , one of which he named Rhizomys pinjoricus . 41 Black (1972) reallocated this species to the genus Rhizomyides created by Bohlin 42 43 (1946), which unfortunately he miswrote as “ Rhizomyoides ” (see below). It was not 44 until 1990 that this taxon was suspected to be an Anepsirhizomys (Flynn et al., 1990). 45 46 The holotype of this species (G.S.I. D278), a fragmentary mandible with m2-m3, is 47 from the Late Pliocene of the Simla area, India. It is housed in the BMNH (paleontology 48 49 collection). Additional material of this species has been recorded from the early 50 Pleistocene near the base of the “Pinjor” strata (Nadah village, Simla Hills, India) 51 52 (Flynn et al., 1990). 53 54 55 56 57 58 59 60 Cladistics Cladistics Page 22 of 68

1 2 3 4 López-Antoñanzas, R. et al. 18 5 6 7 Genus Miorhizomys Flynn, 2009. Type species: Miorhizomys nagrii (Hinton, 1933) 8 9 Referred species: Miorhizomys micrus , M. blacki , M. pilgrimi, M. harii , M. tetracharax , 10 M. choristos . 11 12 The detailed study of the oldest fossil skull and postcrania of Rhizomys 13 (Brachyrhizomys ) shansius from the Yushe Basin carried out by Flynn (2009) led him 14 15 to realize that even if Brachyrhizomys from the Siwaliks of the Indian subcontinent 16 share some synapomorphies with the former species (and extant Rhizomys ), there are 17 18 also noteworthy differencesFor between them. Peer Indeed, the Siwalik Review bamboo rats lack 19 important synapomorphies of living rhizomyines and of R. (Brachyrhizomys ) shansius 20 21 like such as a high infraorbital foramen without open ventral slit (Flynn, 2009). Thus, he 22 created the new genus Miorhizomys for large Siwalik bamboo rats and their Miocene 23 24 relatives from southern China. 25 26 27 Miorhizomys nagrii (Hinton, 1933) . Hinton (1933) on the basis of some remains he 28 found among the material belonging to the GSI named four new species, which he 29 30 considered as belonging to the genus Rhizomys . One of them was Rhizomys nagrii , 31 which was based on a left dentary fragment with broken i and m2-m3 (GSI D273) from 32 33 the Late Miocene lower alternations of Haritalyangar, Simla Hills, India. Black (1972) 34 described in detail the material briefly diagnosed by Hinton (1933). In his revision, 35 Black transferred this species to the genus Rhizomyides Bohlin, 1946, mispelling as 36 37 Rhizomyoides (see below). Afterwards R. nagrii was reallocated to Brachyrhizomys 38 (Flynn, 1982a). When Flynn (2009) named the new genus Miorhizomys he designated 39 40 this as its genotypic species. The holotype of this taxon is housed in the BMNH 41 (paleontology collection). 42 43 Additional material of this species has been recorded from the Late Miocene of 44 Kaulial, Pakistan, as well as from localities Y270, Y260, Y182, Y330, Y261, Y317, 45 46 Y464, Y262, Y892 of the lower Dhok Pathan Fm., Potwar Plateau, Pakistan (Flynn 47 1982a) and from the coeval vertebrate locality of Lufeng (about 9 km north of Lufeng 48 49 city) in southern China (Flynn and Qi , 1982). Furthermore, variousSeveral isolated 50 molars of Miorhizomys nagrii described by Jacobs (1978) under the name Rhizomyoides 51 52 cf. R. nagrii had beenwere collected and described by Jacobs (1978) by him atfrom the 53 Late Miocene Dhok Pathan hominoid locality YGSP 182A, about 9.2 Myr. 54 55 56 57 58 59 60 Cladistics Page 23 of 68 Cladistics

1 2 3 4 López-Antoñanzas, R. et al. 19 5 6 7

8 9 Miorhizomys micrus (Flynn, 1982) . The holotype and sole specimen recorded of this 10 species (YGSP 5494) is a left dentary fragment with broken i1, m1-m3 from the Late 11 12 Miocene (9.3 Myr) lower Dhok Pathan Formation in the Potwar Plateau, Pakistan 13 (Flynn 1982a). The taxon was originally ascribed to Brachyrhizomys , but was 14 15 reallocated to Miorhizomys by Flynn (2009). This material is currently housed at the 16 PMAE and it should eventually be returned to Pakistan and be deposited at the PMNH. 17 18 For Peer Review 19 Miorhizomys blacki (Flynn, 1982) . Brachyrhizomys blacki was named by Flynn 20 21 (1982a) on the basis of a single right dentary fragment with broken i1, m2-m3 recorded 22 from the Late Miocene locality Y464 of the Dhok Pathan Formation in the Potwar 23 24 Plateau, Pakistan. The taxon was originally ascribed to Brachyrhizomys , but was 25 reallocated to Miorhizomys by Flynn (2009). This specimen is currently housed at the 26 27 PMAE and it should eventually be deposited at the PMNH. Unfortunately no additional 28 material of this species has been found to date. 29 30 31 Miorhizomys pilgrimi (Hinton, 1933) . In 1933, Hinton named the new species 32 33 Rhizomys pilgrimi for a partial right mandible with m2-m3 from the late Miocene lower 34 alternations, Haritalyangar, India. This taxon has known the same taxonomic history as 35 Miorhizomys nagrii . The holotype is housed in the BMNH (paleontology collection). M. 36 37 cf. pilgrimi has been found at the Late Miocene (9.4-7.9 Myr) localities Y270, Y17, 38 Y19, Y118, Y954, Y992, and Y1000 of the Dhok Pathan Formation in the Potwar 39 40 Plateau, Pakistan (Flynn, 1982a), and at the coeval Chinese locality of Lufeng (Flynn 41 and Qi, 1982). 42 43 44 Miorhizomys harii (Prasad, 1968) . This taxon was named based on a single right 45 46 dentary fragment with m1-m2 recovered from Haritalyangar, India. Its age is unknown, 47 but Flynn (1982b) suggested an age close to 9 Myr (current time scale) on the basis of 48 49 its stage of evolution. This species was originally considered as belonging to the genus 50 Rhizomys . It was subsequently transferred to the genus Brachyrhizomys (Flynn, 1982b) 51 52 and later to Miorhizomys (Flynn, 2009). No additional material of M. harii has been 53 found to date. 54 55 56 57 58 59 60 Cladistics Cladistics Page 24 of 68

1 2 3 4 López-Antoñanzas, R. et al. 20 5 6 7 8 9 Miorhizomys tetracharax (Flynn, 1982) . This taxon was originally named as 10 Brachyrhizomys tetracharax (Flynn, 1982a), but Flynn (2009) included it in 11 12 Miorhizomys .The holotype of this species (YGSP 4810) consists of an associated palate 13 with right and left I, M1-M3, right dentary fragment with i, m1-m3 and right half 14 15 occiput with auditory region recorded from the Late Miocene (8.7 Myr) locality Y204 16 from the Dhok Pathan Formation in the Potwar Plateau, Pakistan (Flynn, 1982a). It is 17 18 currently housed at the PMAEFor and it should Peer be deposited eventuallyReview at the PMNH. 19 More material of Miorhizomys tetracharax was found in the same formation at localities 20 21 Y260 and Y388 (Flynn, 1982a), and more recently at Y631, Y952, and Y954. In India, 22 Late Miocene remains of this species (Fig. 4E) were recorded from the Hari Scarp, 23 24 Punjab (Flynn, 1982a), and from the “hominoid interval” of the Haritalyangar Scarp, 25 Himachal Pradesh (Flynn et al., 1990). Finally, M. tetracharax was also found at the 26 27 Late Miocene vertebrate locality of Lufeng in southern China (Flynn and Qi , 1982). 28 29 30 Miorhizomys choristos (Flynn, 1982) . The holotype of this species (YGSP 4053), an 31 associated snout with both I1, left and right dentary fragments with broken i1, m1-m3, a 32 33 distal left humerus, a proximal left ulna, a right femur portion and cranial fragments is 34 housed at the PMAE and it should be eventually deposited at the PMNH (Flynn, 35 1982a). It comes from the Late Miocene (8.3 Myr) locality Y172, Dhok Pathan 36 37 Formation, Potwar Plateau, Pakistan. At that time, the holotype was the sole material 38 found of this taxon. Later on, additional specimens were recovered from Y457, Y535, 39 40 and in northern India from the the “hominoid interval” of the Haritalyangar Scarp, 41 Himachal Pradesh (Flynn et al., 1990). 42 43 44 Genus Eicooryctes Flynn, 1982 . Type species: Eicooryctes kaulialensis Flynn, 1982 45 46 47 Eicooryctes kaulialensis Flynn, 1982 . Flynn (1982a) named this taxon on the basis of 48 49 some cranial and dentary fragments from the Late Miocene (8.0-7.2 Myr) localities 50 Y434, Y437 and Y438 of the upper Dhok Pathan Formation, Kaulial Kas, Potwar 51 52 Plateau, Pakistan. Its holotype, a palate and snout with broken incisors and left and right 53 M1-M2 (YGSP 15320), is currently housed at the PMAE and it should eventually be 54 55 56 57 58 59 60 Cladistics Page 25 of 68 Cladistics

1 2 3 4 López-Antoñanzas, R. et al. 21 5 6 7 returned to Pakistan where it would be deposited at the PMNH. Recently additional 8 9 material of this species was been found at locality Y898. 10

11 12 Genus Protachyoryctes Hinton, 1933. Type species: Protachyoryctes tatroti Hinton, 13 1933 14 15 16 Protachyoryctes tatroti Hinton, 1933 . This species was named by Hinton (1933) on 17 18 the basis of a partial right mandibleFor with m1-m3Peer (GSI D272), Review which is housed in the 19 BMNH (paleontology collection). Hinton (1933) indicated that this material comes from 20 21 a Pliocene horizon near Tatrot village, Salt Range, Pakistan. However, its provenance 22 has been thrown into doubt since, in particular, the discovery of new material belonging 23 24 to this taxon at the Late Miocene (7.8-7.1 Myr) localities Y434, Y909, and Y369 in the 25 Dhok Pathan Formation of the Potwar Plateau, Pakistan (Flynn, 1982a). Flynn (1982a) 26 27 pointed out the presence directly beneath Tatrot of much older rocks (more than 6 Myr), 28 which he suspected was the provenance of the holotype of P. tatroti . These doubts are 29 30 supported by the great resemblance between this species and the Late Miocene 31 Tachyoryctes makooka . 32 33 34 Genus Rhizomyides Bohlin, 1946. Type species: Rhizomyides sivalensis (Lydekker, 35 1884) 36 37 Synonymy: Rhizomyoides Black, 1972 38 Referred species: Rhizomyides punjabiensis , R. carbonnelli , R. platytomeus , R. 39 40 mirzadi 41 This genus was created by Bohlin (1946) to include the two Siwalik rhizomyines 42 43 known by him at that time (R. sivalensis (Lydekker, 1884) and R. punjabiensis (Colbert 44 1933)), which were considered, in those days as belonging to the genus Rhizomys . 45 46 Hinton (1933) had named four new species of “ Rhizomys ” from the Siwaliks ( R. nagrii , 47 R. pinjoricus , R. lydekkeri and R. pilgrimi ), but this was apparently unknown to Bohlin. 48 49 According to Bohlin (1946), R. sivalensis and R. punjabiensis were congeneric and 50 clearly distinct from Rhizomys . Thus, he proposed the genus Rhizomyides and chose 51 52 Rhizomyides sivalensis as the type species. Black (1972) in his review of the fossil 53 rodents from the Siwaliks mistook this genus name as Rhizomyoides . This spelling error 54 55 56 57 58 59 60 Cladistics Cladistics Page 26 of 68

1 2 3 4 López-Antoñanzas, R. et al. 22 5 6 7 was repeated by subsequent authors (see e.g. Gupta et al., 1978; Vasishat et al., 1978; 8 9 Brandy 1979a) until review by Flynn (1981). 10

11 12 Rhizomyides sivalensis (Lydekker, 1884) . Lydekker (1878) briefly described two 13 mandibular rami of a new rodent from Jabi, Punjab, Pakistan. Lydekker (1878) thought 14 15 this material could probably represent a new species, but he cautiously proposed the 16 name R. sivalensis and it turned out his thoughts were confirmed. As this new taxon was 17 18 named with uncertainty it cannotFor be taken Peer into account ( article Review 67.2.5 of the 19 International International Commission on Zoological Nomenclature, code of zoological 20 21 nomenclature, 20001999: article 67.2.5 ) and it was Lydekker (1884) who made the 22 name Rhizomys sivalensis available. Its holotype (GSI D97) is a left dentary fragment 23 24 with i1 , m2-m3 from Jabi. It is housed in the BMNH (paleontology collection). 25 Additional material of this species has been recorded from the Potwar Plateau, Pakistan 26 27 and from the Upper Siwaliks of the Siwalik Hills, India (Flynn, 1982a). 28 29 30 Rhizomyides carbonnelli (Brandy, 1979) . Brandy (1979 a) named this taxon 31 Rhizomyoides carbonnelli repeating the spelling error made by Black in 1972 in the 32 33 generic name Rhizomyides created by Bohlin in 1946 (see above). The holotype of this 34 species (PEC 101) is a right mandible with i, m1-m3 from the Late Miocene-Early 35 Pliocene of Pul-e Charkhi, Kabul, Afghanistan (Brandy, 1979a, 1979b, 1981). It is 36 37 currently housed in the collections of the UM2 (B. Marandat, pers. comm. 2011). 38

39 40 Rhizomyides platytomeus Flynn, Heintz, Sen et Brunet, 1983 . This species was 41 named on the basis of a left dentary with i, m1-m3 (AFG 059) recorded from the middle 42 43 Pliocene locality of Lataband, Sarobi Basin, Afghanistan (Flynn et al., 1983). It is 44 currently housed in the paleontological collections of the MNHN (S. Sen, pers. comm. 45 46 2011). Unfortunately no additional material of this taxon has been found to date. 47 48 49 Rhizomyides mirzadi (Lang et Lavocat, 1968) . Lang and Lavocat (1968) named the 50 new species Rhizomys mirzadi on the basis of a right mandibular fragment with m2 51 52 (BAM 1) and Kanisamys bamiani based on an isolated m1 (BAM 2) from the Late 53 Miocene Bamian Basin (Afghanistan). Brandy (1981) reallocated Rhizomys mirzadi to 54 55 56 57 58 59 60 Cladistics Page 27 of 68 Cladistics

1 2 3 4 López-Antoñanzas, R. et al. 23 5 6 7 the genus Protachyoryctes . The reexamination of these two specimens and an 8 9 unpublished third one (BAM 3), led Flynn (1983) to gather all them in a sole species, 10 which he transferred to the genus Rhizomyides . Thus, Kanisamys bamiani is a synonym 11 12 of Rhizomyides mirzadi. The holotype (BAM 1) is currently housed in the collections of 13 the MNHN (S. Sen, pers. comm. 2011). 14 15 16 Rhizomyides punjabiensis (Colbert, 1933) . Colbert (1933) described a small 17 18 rhizomyine that he named RhizomysFor punjabiensis Peer. Bohlin (1946)Review coined the new genus 19 Rhizomyides and allocated this species to it. R. punjabiensis was later reassigned to the 20 21 genus Brachyrhizomys (Flynn 1981, 1982a, 1982b, 1984, 1985). Flynn (1986, 1990) 22 later recognized the appropriate attribution Rhizomyides , making it the oldest known 23 24 species of this genus. The holotype (AMNH 19762), a right mandibular ramus with the 25 m2 and m3 and the posterior portion of the incisor, and only record of this species in 26 27 those days, comes from the Late Miocene (Flynn, 1982a) Lower Siwalik beds of 28 Bilaspur, Punjab, India (Colbert, 1933). Additional material of this species has been 29 30 recorded from the early Late Miocene of Pakistan, near Kaulial and at the localities 31 Y797, Y259 and Y450 from the Nagri Formation, Potwar Plateau (Flynn, 1982a) 32 33 34 35 Rhizomyides lydekkeri (Hinton, 1933) and Rhizomyides saketiensis (Gupta, Verma 36 et Tewari, 1978) . Rhizomyides lydekkeri was named by Hinton (1933) as Rhizomys 37 38 lydekkeri . Its holotype (BMNH 15925) is a partial mandible with m1-m3 that comes 39 from the Pliocene Upper Siwaliks of Northern India. Colbert (1935) considered this 40 41 specimen together with some others (BMNH 15926, 15927 and 15927a) as Rhizomys sp. 42 and pointed out that they probably did not belong to Rhizomys sivalensis because of 43 44 their larger size and more recent age. However, Black (1972) synonymized the two 45 species, and Flynn (1982b) considered them to represent the genus Rhizomyides . 46 47 48 Gupta et al., (1978) named the new species Rhizomyoides saketiensis on the basis of 49 two dentary fragments from the Pliocene/Pleistocene Tatrot beds at Saketi Park 50 51 (Himachal Pradesh). The holotype (GSI 19549) is a partial right mandibular ramus with 52 m2 and m3. Flynn (1983) reallocated this species to the genus Rhizomyides . It is 53 54 unknown if R. lydekkeri and R. saketiensis represent the same species (Flynn, 1982b; 55 56 57 58 59 60 Cladistics Cladistics Page 28 of 68

1 2 3 4 López-Antoñanzas, R. et al. 24 5 6 7 Flynn et al., 1983). Pending a re-examination of Hinton's (1933) original material, we 8 9 have preferred not to not include these two taxa in our analysis. 10 11

12 13 14 Genus Kanisamys Wood, 1937. Type species: Kanisamys indicus Wood, 1937 15 Referred species: Kanisamys nagrii , K. sivalensis , K. potwarensis 16 17 In 1932, G. E. Lewis of Yale University collected some fossil rodents from the 18 Siwalik deposits from the ChinjiFor and Nagri Peer zones of what wasReview then North India, which 19 20 he gave to A. E. Wood for study. Some of these specimens turned out to belong to a 21 new genus that Wood (1937) named Kanisamys . 22 23 24 Kanisamys indicus Wood, 1937 . The holotype of this species (YPM 13810), a right 25 26 mandible with m1-m3, comes from the Chinji Zone, South south of Chinji Village, now 27 Pakistan (Wood, 1937). Additional material of this taxon has been found in Pakistan 28 29 from Middle Miocene localities in the Chinji Formation (Flynn et al., 1995). 30 Furthermore, a fragmentary mandible (specimen D 271) from near Chinji (Attock 31 32 District) considered by Hinton (1933) as Theridomys was reallocated to K. indicus by 33 Black (1972). K. indicus has also been recorded from the locality 18 in the Chinji 34 35 Formation at Daud Khel (Munthe, 1980), from the locality H-GSP 107 in the Chinji 36 Formation, Banda Daud Shah (Wessels et al., 1982) and at various localities in the 37 Lower Manchar formation, Sind (Wessels and de Bruijn, 20032001 ). 38 39 40 Kanisamys sivalensis Wood, 1937 . The holotype of this species (YPM 13801), a left 41 42 mandible with m1-m3, comes from Lower Alternations, East of Haritalyangar, India 43 (Wood, 1937). Additional material of this taxon has been recorded from various Late 44 45 Miocene localities in the Upper Nagri and lower Dhok Pathan formations, Potwar 46 Plateau, Pakistan (Jacobs 1978; Flynn, 1982a; Flynn et al., 1995). 47 48 49 Kanisamys nagrii Prasad, 1968 . This species was named on the basis of a single left 50 51 dentary fragment (GSI 18086) with i1M2-3 (Prasad, 1968), which was refigured by 52 Flynn (1982b). It comes from the Late Miocene Lower Alternations, East of 53 54 Haritalyangar, India. Additional material of this taxon has been recorded in the same 55 56 57 58 59 60 Cladistics Page 29 of 68 Cladistics

1 2 3 4 López-Antoñanzas, R. et al. 25 5 6 7 area (Fig. 4A-C) as well as from the upper portion of the Chinji Formation, and the Late 8 9 Miocene Nagri Formation, Potwar Plateau, Pakistan (Flynn, 1982a). 10

11 12 Kanisamys potwarensis Flynn, 1982 . This species was named by Flynn (1982a) on 13 the basis of various isolated molars (Fig. 4D). The holotype of this taxon (YGSP 8379), 14 15 a left M3, comes from the Middle Miocene locality Y491 in the Chinji Formation, 16 Potwar Plateau, Pakistan. Additional material of this taxon from the same area is now 17 18 recognized at localities Y709,For Y733, Y430, Peer Y491, Y668, Y59,Review Y641, and Y651 and 19 Y41 (14.3 to 13.4 Myr). Kanisamys cf. potwarensis has been mentioned from the Lower 20 21 Siwalik Group of Ramnagar, Udhampur District, Jammu and Kashmir, India (Parmar 22 and Prasad, 2006; Sehgal and Patnaik in press). 23 24 25 Genus Nakalimys Flynn et Sabatier, 1984. Type species. Nakalimys lavocati Flynn 26 27 and Sabatier, 1984 28 29 30 Nakalimys lavocati Flynn et Sabatier, 1984 . This taxon was named on the basis of a 31 palate and mandible, two dentaries, an isolated lower incisor and some lower and upper 32 33 molars from the early Late Miocene site of Nakali, Kenya (Flynn and Sabatier, 1984). 34 The holotype (KNM-NA 7700) is a palate and mandible with left and right M1-M3 and 35 i1 m1-m3. Nakalimys cf. lavocati has been found in the coeval site of Chorora, Ethiopia 36 37 (Geraads, 1998). 38

39 40 Genus Pronakalimys Tong et Jaeger, 1993. Type species. Pronakalimys andrewsi 41 Tong and Jaeger, 1993 42 43 44 Pronakalimys andrewsi Tong et Jaeger, 1993. This species was named on the basis 45 46 of 13 partial lower jaws (7 with m1-m3, 5 with m1-m2 and 1 with m3) and a right 47 maxilla with M1-M3 from the Middle Miocene site of Fort Ternan, Kenya (Tong and 48 49 Jaeger 1993). The holotype (KNM-FT64-007) is a right maxilla. No additional material 50 of this taxon has been found so far. 51 52 53 54 55 56 57 58 59 60 Cladistics Cladistics Page 30 of 68

1 2 3 4 López-Antoñanzas, R. et al. 26 5 6 7 Genus Prokanisamys de Bruijn, Hussain et Leinders, 1981. Type species: 8 9 Prokanisamys arifi de Bruijn, Hussain et Leinders, 1981 10 Referred species: Prokanisamys kowalskii , P. benjavuni , P. major, P. sp. Libya 11 12 Except for a few remains of Prokanisamys sp. described by Wessels et al., (2003) 13 from the Lower Miocene of Libya, all the specimens of this genus found to date come 14 15 from the Asian continent (Pakistan and Thailand). 16 17 18 Prokanisamys arifi de Bruijn,For Hussain Peer et Leinders, 1981 .Review This species was defined on 19 the basis of some isolated molars (6 m1, 11 m2, 11 m3, 6 M1, 6 M2 and 16 M3) from 20 21 the Lower Miocene H-GSP 116 locality of the Murree Formation, Banda Daud Shah, 22 Kohät, Pakistan (de Bruijn et al., 1981). The holotype (specimen nº22) is a right m1 23 24 currently housed at the IVAU. Additional material of this species has been found at the 25 localities Z126, Z122, Z120 and Z 124 of the Chitarwata and Vihowa Formations, 26 27 Zinda Pir Dome, Pakistan, as well as at the localities H-GSP 81114a, H-GSP 8424, H- 28 GSP 81114, H-GSP 8426 (Sehwan Section) and H-GSP 8107a, H-GSP 8209, H-GSP 29 30 8311 and H-GSP 8106 (Gaj River Section) of the Lower Manchar Formation, Sind, 31 Pakistan (Wessels and de Bruijn 20032001 ; Lindsay et al., 2005; Wessels 2009). 32 33 34 Prokanisamys kowalskii (Lindsay, 1996 ). Lindsay (1996) coined a new species of 35 Eumyarion as E. kowalskii on the basis of various molars and some isolated incisors 36 37 recorded from the latest Oligocene (Lindsay et al., 2005; Flynn, 2009) locality Z113, 38 Zinda Pir Dome, Pakistan. Wessels and de Bruijn (2001) transferred E. kowalskii to the 39 40 genus Prokanisamys because of the greater closeness to the latter genus than to 41 Eumyarion from Europe and Western Asia. The holotype of this species (PMNH 42 43 Z113/679) is a right isolated M1. 44 45 46 Prokanisamys benjavuni (Mein et Ginsburg, 1985 ). The preliminary results of a field 47 campaign carried out in Li Mae Long, Lamphun district, Thailand by P. Mein and L. 48 49 Ginsburg in 1984 in collaboration with the DMR and the University of Chiang Mai 50 (Thailand) lead them to name a new species of rhizomyine: Kanisamys benjavuni (Mein 51 52 and Ginsburg, 1985). Jacobs et al. (1989) reallocated this species to the genus 53 Prokanisamys , a conclusion with which Mein and Ginsburg (1997) agreed after being 54 55 56 57 58 59 60 Cladistics Page 31 of 68 Cladistics

1 2 3 4 López-Antoñanzas, R. et al. 27 5 6 7 able to sample more specimens of this taxon. Prokanisamys benjavuni comes from Li, 8 9 Lamphun district, Thailand. This locality was first considered as Early Miocene until 10 the works of Chaimanee et al. (2007) proved it to be Middle Miocene in age. The 11 12 holotype of this taxon (T Li 203), a left isolated m3 (the single complete m3 kown at the 13 time), is housed at the collections of the FSL under the number FSL 67035 (Mein, pers. 14 15 comm. 2011). Additional material of this species has also been found in the MaeMoh 16 Basin, northern Thailand ( Chaimanee et al., 2007) as well as in some Pakistani late 17 18 Early Miocene levels (5, 6 Forand 6 bis) of DeraPeer Bugti (Welcomme Review and Ginsburg, 1997; 19 Welcomme et al., 2001) and at H-GSP 8114a, H-GSP 8114, H-GSP 8424 and H-GSP 20 21 8426 localities in the Lower Manchar Formation, Sehwan Section (Wessels and de 22 Bruijn, 2001). 23 24 25 Prokanisamys major Wessels et de Bruijn, 2001 . The holotype of this species 26 27 (specimen nº 4522) is an isolated left M1 from the Middle Miocene locality H-GSP 28 81.14 of the Lower Manchar Formation, Sind, Pakistan. It is currently housed at the 29 30 IVAU. Additional material of this species has been found in the same area at late Early 31 and Middle Miocene localities (H-GSP 8106, H-GSP 8227, H-GSP 8224, H-GSP 8424, 32 33 H-GSP 8214, H-GSP 8425, H-GSP 8426, H-GSP 8427) as well as at some localities (Y- 34 GSP 591, Y-GSP 592, Y-GSP 642, Y-GSP 501, Y-GSP 589, Y-GSP 491, Y-GSP 640, 35 Y-GSP 641, Y-GSP 496, Y-GSP 634) of similar age on the Potwar Plateau, Pakistan 36 37 (Flynn, 1986; Jacobs et al., 1990; Wessels and de Bruijn, 2001). The specimens from 38 this latter area were first identified as Prokanisamys benjavuni (Flynn, 1986; Jacobs et 39 40 al., 1990) until the discovery of the species Prokanisamys major (Wessels and de 41 Bruijn, 2001). 42 43 44 Prokanisamys sp. Wessels, Fejfar, Peláez-Campomanes et de Bruijn, 2003 . Two 45 46 paleontological field campaigns carried out in 1983 and 1997 in various Miocene levels 47 of the Marada Formation, Jebel Zelten, Libya, resulted in the discovery of some large 48 49 and small mammals. Among the latter, five isolated teeth (3M1, 1m2, 1m3) of a species 50 of Prokanisamys were found at the Early Miocene levels ATH4B, ATH5A1 and 51 52 ATH7A2 (Wessels et al., 2003). This finding is of relevance not only as the only record 53 54 55 56 57 58 59 60 Cladistics Cladistics Page 32 of 68

1 2 3 4 López-Antoñanzas, R. et al. 28 5 6 7 of Prokanisamys found to date in Africa, but also for being the oldest record of a 8 9 rhizomyine found in this continent. 10

11 12 Phylogenetic analysis 13 14 15 Previous work 16 17 18 Flynn (1990). The first attemptFor at deciphering Peer the relationships Review of the extinct and 19 20 extant Rhizomyinae (Rhizomyidae sensu Flynn, 1982a) based on the synapomorphies of 21 the different genera of rhizomyines (a group then at the family level) was carried out by 22 Flynn (1990). According to the hand-generated cladogram (Fig. 6), the Asian 23 24 Prokanisamys was the most basal genus of the ingroup, followed by the East African 25 Nakalimys . The analysis supported the monophyly of (Kanisamys , Protachyoryctes , 26 27 Eicooryctes , Tachyoryctes ), which Flynn (1990) considered as the subfamily 28 Tachyoryctinae. In addition, the clade (“ Brachyrhizomys ”, Rhizomys , Cannomys ) would 29 30 constitute the subfamily Rhizomyinae. Flynn (1990) inferred two plausible relationships 31 for the sister genus of Kanisamys , Rhizomyides : either it belongs to the Tachyoryctinae 32 33 or it is the sister taxon to the clade constituted by the Tachyoryctinae and Rhizomyinae. 34 35 36 New analysis 37 38 39 32 most parsimonious trees were generated with a length of 158 and a relatively high 40 degree of homoplasy (CI = 0.456 and RI = 0.819). The strict and semi-strict consensus 41 42 trees are identical and largely resolved (Fig. 7A) with five polytomies each. In the 43 majority-rule consensus tree one of these polytomies disappears and one, although 44 45 persisting, involves a reduced number of taxa (Fig. 7B). Bremer and relative Bremer 46 support indices (Goloboff and Farris, 2001; Goloboff et al., 2003) for each node are 47 48 indicated on the cladogram in the figure 7A. 49 The transformations supporting the topology of this tree (under the ACCTRAN and 50 DELTRAN optimizations) are listed in Table 12. Each internal node is discussed below, 51 52 beginning from the most basal (whenever both unambiguous and ambiguous 53 synapomorphies support a given node, only the former are mentionned). 54 55 56 57 58 59 60 Cladistics Page 33 of 68 Cladistics

1 2 3 4 López-Antoñanzas, R. et al. 29 5 6 7 Node 62. Prokanisamys spp. + more derived taxa. This node is based on a single 8 9 non-exclusive synapomorphy under ACCTRAN: mesolophid on the m2 short and well- 10 separated from the hypolophid. 11 12 Node 61. Prokanisamys spp. This clade is sustained by two exclusive and 13 unambiguous synapomorphies, i.e weak lophodonty and mesolophid on the m3 short 14 15 and well-separated from the hypolophid, and two non-exclusive and unambiguous 16 synapomorphies, i.e. mesolophid on the m1 short and well-separated from the 17 18 hypolophid (a convergenceFor with Nakalimys Peer lavocati , the node Review 53 and some individuals 19 of Miorhizomys nagrii and M. choristos ) and anterosinus on M3 enclosed by anteroloph 20 21 and paracone (a convergence with node 58). 22 Node 60. Pronakalimys andrewsi + more derived taxa. This node is supported by one 23 24 exclusive and unambiguous synapomorphy, i.e. ectoloph on M2 absent, and one non- 25 exclusive and unambiguous synapomorphy, i.e. ectoloph on M1 absent (a convergence 26 27 with Prokanisamys kowalskii ). 28 Node 59. Nakalimys lavocati + more derived taxa. One exclusive and unambiguous 29 30 synapomorphy supports this node, i.e. anteroconid on m1 absent, as well as one non- 31 exclusive and unambiguous synapomorphy, i.e. m3 enlarged (a convergence with 32 33 Prokanisamys major ). 34 Node 58. ( Kanisamys potwarensis , K. indicus ) + more derived taxa. This node is 35 supported by two exclusive and unambiguous synapomorphies, i.e. labial anterolophid 36 37 on m1 absent and mesoloph on M2 long and complete. It is also sustained by three non- 38 exclusive and unambiguous synapomorphies, i.e. moderate hypsodonty (a convergence 39 40 with Prokanisamys benjavuni , Tachyoryctes pliocaenicus and T. konjiti ), presence of 41 mesostyle on M1 (a convergence with Rhizomyides punjabiensis and Miorhizomys 42 43 nagrii ) and anterosinus enclosed by anteroloph and paracone on M3 (convergence with 44 node 61). 45 46 Node 57. Kanisamys potwarensis + K. indicus . One non-exclusive and unambiguous 47 synapomorphy supports this clade, i.e. mesolophid long and complete and separated 48 49 from the hypolophid on m2 (a reversal). 50 Node 56. Kanisamys nagrii + more derived taxa. This node is supported by five 51 52 exclusive and unambiguous synapomorphies: high lophodonty, labial anterolophid 53 absent on m2 and m3, protosinusid absent on the m2 and m3. Two additional non- 54 55 56 57 58 59 60 Cladistics Cladistics Page 34 of 68

1 2 3 4 López-Antoñanzas, R. et al. 30 5 6 7 exclusive and unambiguous synapomorphies are present, i.e. posterior protoconid- 8 9 metaconid connection absent on m1 (a convergence with some individuals of 10 Kanisamys indicus and node 44) and mesolophid short and migrated toward the 11 12 hypolophid on m3 (a convergence with Pronakalimys andrewsi ). 13 Node 55. Kanisamys sivalensis + more derived taxa. Two non-exclusive and 14 15 unambiguous synapomorphies support this node, i.e. metaloph fused with posteroloph 16 on M1 (a convergence with Kanisamys potwarensis ) and posterosinus absent on the M1 17 18 (a convergence with KanisamysFor potwarensis Peer). Review 19 Node 54. Miorhizomys micrus + ( Rhizomyides mirzadi , R. punjabensis ) + 20 21 (Tachyoryctes makooka , Protachyoryctes tatroti ) + ( R. carbonnelli (R. platytomeus , R. 22 sivalensis )) + more derived taxa. This node is supported by five exclusive and 23 24 unambiguous synapomorphies, i.e. dentary depth moderate, mesostyle on M1 absent, 25 anterocone fused with anteroloph on M1, metaloph early fused with posteroloph on M2 26 27 and posterosinus absent on M2. 28 Node 53. Rhizomyides mirzadi + R. punjabensis . Two non-exclusive and unambiguous 29 30 synapomorphies support this node, i.e. masseteric crest with short anterior extension 31 32 (Prokanisamys major and maybe other species of this genus, Miorhizomys micrus, M. 33 34 nagrii , M. harii , Eicooryctes kaulialensis and Cannomys badius ) and mesolophid short 35 36 and well-separated from the hypolophid on m1 (a convergence with node 61, Nakalimys 37 38 lavocati , Kanisamys potwarensis and some specimens of Miorhizomys nagrii and M. 39 40 choristos ). 41 42 Node 52. Rhizomyides carbonnelli + ( R. platytomeus , R. sivalensis ). This node is 43 supported by two non-exclusive and unambiguous synapomorphies, i.e. lower 44 45 masseteric crest inflated under m2 (a convergence with Protachyoryctes tatroti and 46 node 45) and presence of posterior protoconid-metaconid connection on m1 (a highly 47 48 homoplastic character). 49 Node 51. Rhizomyides platytomeus + R. sivalensis . Two non-exclusive and 50 51 unambiguous synapomorphies support this node, i.e. length of m2 between 4.5 and 5.5 52 mm (a convergence with some individuals of Miorhizomys choristos , M. tetracharax 53 54 55 56 57 58 59 60 Cladistics Page 35 of 68 Cladistics

1 2 3 4 López-Antoñanzas, R. et al. 31 5 6 7 and node 48) and mesolophid long and complete and separated from the hypolophid on 8 9 m2 (a reversal). 10 Node 50. Protachyoryctes tatroti + Tachyoryctes makooka . This node is supported 11 12 two non-exclusive and unambiguous synapomorphies, i.e. mesolophid short and 13 migrated toward the hypolophid (a convergence with some specimens of Kanisamys 14 15 nagrii, K. sivalensis, Eicooryctes kaulialensis and Tachyoryctes pliocaenicus ) and mure 16 on the m2 constricted (an homoplastic character that is also found in Rhizomyides 17 18 platytomeus , Miorhizomys Forharii , M. tetracharax Peer, Rhizomys Review(Brachyrhizomys ) shajius 19 and Eicooryctes kaulialensis ). 20 21 Node 49. Miorhizomys choristos + M. tetracharax + M: harii + M. blacki + M. 22 nagrii + Rhizomys (Brachyrhizomys) shajius + more derived taxa . This node is 23 24 supported by no less than four exclusive and unambiguous synapomorphies, i.e. 25 masseteric crest without anterior extension, M1 hypselodont and M2 and M3 absent, 26 27 and six non-exclusive and unambiguous synapomorphies, i.e. longitudinal 28 ornamentation on i absent (a convergence with Miorhizomys micrus ), mure on m1 29 30 constricted (a convergence with Tachyoryctes makooka ), mesolophid long and complete 31 and separated from hypolophid on m2 (a reversal), posterosinusid small on m3 (a 32 33 convergence with Miorhizomys micrus , some individuals of M. pilgrimi and 34 Tachyoryctes pliocaenicus ), anterosinusid absent on m3 (a convergence with some 35 specimens of Kanisamys nagrii and either with the node 52 or, as a parallelism, with 36 37 Rhizomyides platytomeus and R. carbonnelli ) and protosinus absent on M1 (a 38 convergence with Prokanisamys benjavuni and some individuals of Rhizomyides 39 40 punjabiensis ). 41 Node 48. Rhizomys sinensis + Tachyoryctes splendens + all the descendants of their 42 43 common ancestor (crown group). This clade is sustained by only one exclusive and 44 unambiguous synapomorphy, i.e. posterior lake absent on M3, but four non-exclusive 45 46 and unambiguous synapomorphies, i.e. m2 length between 4.5 mm and 5.5 mm (a 47 convergence with node 51 and some individuals of Miorhizomys choristos and M. 48 49 tetracharax ), mure on m3 absent (a convergence with Miorhizomys choristos and M. 50 tetracharax ), posterosinusid on m3 absent (a convergence with Miorhizomys choristos 51 52 and M. tetracharax ) and mesoloph long and complete on M3 (a reversal). 53 54 55 56 57 58 59 60 Cladistics Cladistics Page 36 of 68

1 2 3 4 López-Antoñanzas, R. et al. 32 5 6 7 Node 47. Anepsirhizomys opdykei + more derived taxa. Two non-exclusive and 8 9 unambiguous synapomorphies support this node, i.e. high hypsodonty (a convergence 10 with node 41) and posterior protoconid-metaconid connection on m1 absent (a highly 11 12 homoplastic character). 13 Node 46. Anepsirhizomys pinjoricus + more derived taxa. A single exclusive and 14 15 unambiguous synapomorphy supports this node, i.e. masseteric crest with long anterior 16 extension. 17 18 Node 45. Eicooryctes kaulialensisFor + morePeer derived taxa. ThisReview node is supported by 19 two exclusive and unambiguous synapomorphies, i.e. m2 length between 2.5 mm and 20 21 3.5 mm and mesolophid on m2 absent. 22 Node 44. Tachyoryctes pliocaenicus + T. konjiti + T. macrocephalus + T. splendens . 23 24 Three non-exclusive and unambiguous synapomorphies support this node, i.e. mure on 25 m1 absent (a convergence with node 42), posterior protoconid-metaconid connection on 26 27 m1 absent (a highly homoplastic character) and anterosinusid on m1 present (a 28 reversal). 29 30 Node 43. Miorhizomys pilgrimi + more derived taxa. This clade is sustained by only 31 one non-exclusive and unambiguous synapomorphy, i.e. mesolophid on m2 that is a 32 33 long continuation of the protoconid (a convergence with Miorhizomys choristos ). 34 Node 42. Rhizomys (Brachyrhizomys) shansius + more derived taxa. This node is 35 supported by three exclusive and unambiguous synapomorphies, i.e. deep dentary, 36 37 infraorbital foramen without ventral slit and hypolophid on m2 isolated, and one non- 38 exclusive and unambiguous synapomorphy, i.e. anterosinusid on m2 absent (a reversal). 39 40 Node 41. Cannomys badius + more derived taxa. This node is supported by one 41 exclusive and unambiguous synapomorphy, i.e. mesolophid on m1 that is a long 42 43 continuation of the protoconid, and two non-exclusive and unambiguous 44 synapomorphies, high hypsodonty (a convergence with node 47) and entoconid on m3 45 46 isolated (a convergence with Miorhizomys choristos and M. tetracharax ). 47 Node 40. Rhizomys pruinosus + R. sinensis + R. sumatrensis . This node is supported 48 49 by two exclusive and unambiguous synapomorphies, i.e. mesoloph on M1 and M3 long 50 and complete but divided, in early wear, with a buccal cusp, and one non-exclusive and 51 52 unambiguous synapomorphy, i.e. mesoloph on M2 long and complete but divided, in 53 early wear, with a buccal cusp (a convergence with Miorhizomys choristos ). 54 55 56 57 58 59 60 Cladistics Page 37 of 68 Cladistics

1 2 3 4 López-Antoñanzas, R. et al. 33 5 6 7

8 9 10 11 Discussion 12 13 14 Prokanisamys 15 Prokanisamys is the oldest known genus of the rhizomyines. The results of this work 16 17 support the monophyly of this taxon. They do not confirm the position ofthat the most 18 ancient species of the genus,For P. kowalskii Peer, ias the most basal Review one within this clade. Even 19 20 though this Late Oligocene species shows the first phase of the rhizomyine dentition, its 21 lophodonty is incipient and its hypsodonty is hardly greater than that of typical muroids. 22 We , therefore , hypothesize that further work will eventually resolve P. kowalskii as the 23 24 most basal member of the genus. 25

26 27 The African Pronakalimys and Nakalimys 28 Pronakalimys and Nakalimys branch sequentially between Prokanisamys and 29 30 Kanisamys . Pronakalimys and Nakalimys are indeed much closer morphologically to 31 Early and Middle Miocene Asian fossil rhizomyines than to the African Tachyoryctes . 32 33 The occurrence of Prokanisamys sp. from the Early Miocene of Jebel Zelten (Fejfar and 34 Horá ček 2006; Wessels et al., 2003, 2008) provides evidence of an Early Miocene 35 36 migration of rhizomyines from Asia to Africa. However, the topology of our tree 37 suggests that the latter taxon Prokanisamys sp. did not give rise directly to 38 39 Pronakalimys , which would be the fruit of a second immigration of rhizomyines into 40 Africa. As suggested by Flynn and Sabatier (1984), Nakalimys lavocati is more basal 41 42 than Kanisamys indicus and not directly ancestral to Tachyoryctes . 43 44 45 Kanisamys 46 Wessels and de Bruijn (2001) pointed out that ancestor-descendent relationships 47 Kanisamys 48 within this genus were not clear. In fact, the genus is not monophyletic 49 according to our results. However, K. indicus and K. potwarensis do form a clade and 50 are, therefore, true Kanisamys . In contrast, “K. ” nagrii and “K. ” sivalensis insert 51 52 sequentially on the stem leading to more derived rhizomyines. Flynn (1982a, 1984, 53 1986) was, therefore, correct in considering K. potwarensis most closely related to K. 54 55 56 57 58 59 60 Cladistics Cladistics Page 38 of 68

1 2 3 4 López-Antoñanzas, R. et al. 34 5 6 7 indicus than to other species then considered as belonging to Kanisamys . His opinion 8 9 that “K”. nagrii is more basal than “K”. sivalensis and that the former is closer to the 10 latter than to either K. indicus or K. potwarensis is confirmed. Flynn (1982a, 1984, 11 12 1986) also suggested that “K”. nagrii and “K”. sivalensis are more derived than K. 13 indicus and K. potwarensis , which is substantiated by our results. However, the 14 15 suggestion that K. indicus would be a direct ancestor of “K”. sivalensis (Black, 1972; 16 Flynn 1982a) is not supported. 17 18 For Peer Review 19 Protachyoryctes 20 21 The sister species Protachyoryctes tatroti and “Tachyoryctes ” makooka are both 22 hypsodont, have reduced but distinct mesolophids and have retained the protosinus on 23 24 the M1. This latter trait is distinctive of the most basal genera of the subfamily like 25 Prokanisamys , Kanisamys , Pronakalimys , Nakalimys and Rhizomyides (see Flynn 26 27 1982a). It becomes less marked in P. tatroti and “T. ” makooka (in which the protosinus 28 is shallow). The topology of our cladogram implies that “T.” makooka should be 29 30 reallocated to the genus Protachyoryctes (Fig. 7) as already hypothesized by López- 31 Antoñanzas and Wesselmann (in press). From these results a unidirectional dispersal 32 33 event from southern Asia to Africa at the origin of “T.” makooka can be inferred 34 between 6 and 8 Myr (Figs. 8 and 9). Contrary to the suggestion of Black (1972), and 35 what its name may suggest, Protachyoryctes is not closer to Tachyoryctes (sensu 36 37 stricto) than to Rhizomys. 38

39 40 Rhizomyides 41 As currently understood, Rhizomyides is a paraphyletic taxon. True Rhizomyides , 42 43 with R. sivalensis as type, constitute the sister group of Protachyoryctes according to the 44 majority-rule consensus tree (Fig. 7B). 45 46 As Flynn (1982a, 1986, 1990) pointed out, Rhizomyides and Kanisamys share many 47 traits (e.g., similar degree of hypsodonty, inclined masseteric crest, a strongly rounded 48 49 I1, a large posterior enamel lake on M3). This fact led Flynn (1983, 1986) to think that 50 the origin of Rhizomyides could have been earlier than reflected by its first occurence in 51 52 the fossil record and that it could have evolved from an early Kanisamys . This is not 53 fully supported by our results. The fact that Rhizomyides sivalensis retained a strong 54 55 56 57 58 59 60 Cladistics Page 39 of 68 Cladistics

1 2 3 4 López-Antoñanzas, R. et al. 35 5 6 7 mesolophid on the M2 (usually reduced in K. sivalensis ) led Flynn (1982a) to suggest 8 9 that R. sivalensis may have evolved from Kanisamys outside Pakistan and later 10 immigrated to the Potwar Plateau. This hypothesis is congruent with our results given 11 12 that the species most closely related to R. sivalensis (R. carbonnelli and R. platytomeus ) 13 were both found in Afghanistan (Brandy, 1979a, 1979b, 1981; Flynn et al., 1983) . In the 14 15 Late Miocene a form close to R. carbonnelli would have evolved there to give rise to 16 the lineages of R. platytomeus and R. sivalensis the latter taxon subsequently 17 18 immigrating to the Potwar PlateauFor (Fig. 8).Peer We agree with theReview suggestion of Flynn 19 (1983) and Flynn et al. (1983) according to which there is probably neither direct 20 21 relationship between R. carbonnelli and R. sivalensis nor between R. platytomeus and 22 Tachyoryctes . 23 24 25 “Rhizomyides ” mirzadi and “R”. punjabiensis 26 27 The generic attribution of “Rhizomyides ” mirzadi and “R”. punjabiensis has been 28 controversial. “R”. mirzadi has been considered as belonging to Kanisamys (Jacobs, 29 30 1978), Protachyoryctes (Brandy, 1981) and Rhizomyides (Flynn, 1983). “R”. 31 punjabiensis was considered Rhizomyides by Bohlin, (1946), reassigned to 32 33 Brachyrhizomys by Flynn (1982a) and returned to Rhizomyides by Flynn (1986, 1990). 34 Our results suggest that a new genus name is required for these two taxa, which are 35 sister species (Fig. 8). 36 37 38 Miorhizomys 39 40 Another interesting result of the present work is that the genus Miorhizomys does not 41 appear monophyletic. Flynn (1982a) suggested that M. micrus is a conservative 42 43 descendant of the “Rhizomyides ” punjabiensis lineage. This hypothesis is not greatly 44 different from the topology suggested by our tree, and M. micrus is the most basal 45 46 species of Miorhizomys as currently understood. Interestingly enough, Rhizomys 47 (Brachyrhizomys ) shajius nests within the various species of Miorhizomys , and does not 48 49 appear closely related to R. (B.) shansius . This is not very surprising because the former 50 species is primitive in molar narrowness, low hypsodonty and retention of a mure, and 51 52 short mesolophid on the m3), and is represented by only one jaw (Flynn 1993). Note 53 54 55 56 57 58 59 60 Cladistics Cladistics Page 40 of 68

1 2 3 4 López-Antoñanzas, R. et al. 36 5 6 7 that R. (B.) shajius , while older (5.9 Myr vs 4-3.3 Myr, see Flynn, 2009), it occurs in the 8 9 same basin as R. (B.) shansius . 10 The majority-rule consensus tree does not solve the interrelationships between 11 12 Miorhizomys blacki , M. nagrii and Rhizomys (Brachyrhizomys ) shajius (Fig. 7B). 13 Nevertheless, it shows that M. tetracharax, M. choristos and M. harii form a 14 15 monophyletic group of which M. choristos is the most basal species and in which M. 16 tetracharax and M. harii are sister taxa. However, in the strict consensus tree (Fig. 7A), 17 18 the interrelationships betweenFor M. choristos Peer, M. tetracharax ,Review M. harii , M. blacki , M. 19 nagrii and Rhizomys (Brachyrhizomys) shajius remain unsolved. The number of true 20 21 Miorhizomys species remains therefore unsettled. M. micrus , M. blacki , M. nagrii and R. 22 (B.) shajius show some burrower adaptations. An increase of specialization toward a 23 24 fossorial way of life is shown in the more derived M. tetracharax, M. choristos and M. 25 harii , a trend continued by the most derived species of the subfamily. 26 27 28 The crown group 29 30 Another interesting result is the evidence of two important lineages originating from 31 the basal node of the crown group (48). One corresponds to all taxa originating from 32 33 node 47 (Anepsirhizomys opdykei + more derived taxa). This clade is characterized by 34 the synapomorphies of having high hypsodonty and lacking posterior protoconid- 35 metaconid connection on the m1. We understand this clade as the tribe Tachyoryctini 36 37 and we provide it with the following phylogenetic al definition: The most inclusive clade 38 containing Tachyoryctes splendens (Rüppell 1836) but not Rhizomys sinensis Gray, 39 40 1831 . The other lineage originates from node 43 (Miorhizomys pilgrimi + more derived 41 taxa). It is characterized by the synapomorphy of having the mesolophid on the m2 as a 42 43 long continuation of the protoconid. This clade is the tribe Rhizomyini, whose 44 phylogenetic definition is: The most inclusive clade containing Rhizomys sinensis Gray, 45 46 1831 but not Tachyoryctes splendens (Rüppell 1836). 47 The divergence between the tribes Tachyoryctini and Rhizomyini can be estimated as 48 49 having occurred about 10 Myr (Fig. 8), because the oldest record of Miorhizomys 50 pilgrimi is near 9.5 Myr. 51 52 53 54 55 56 57 58 59 60 Cladistics Page 41 of 68 Cladistics

1 2 3 4 López-Antoñanzas, R. et al. 37 5 6 7 Tribe Tachyoryctini. This tribe is constituted by Anepsirhizomys opdykei and more 8 9 derived taxa. Our work sheds light on the controversial taxonomic position of 10 Anepsirhizomys . This genus of unclear affinity has been considered a rhizomyine as 11 12 well as a tachyoryctine (Flynn et al., 1990). According to the results of our analysis, 13 Anepsirhizomys is the most basal genus within the tribe Tachyoryctini. The reallocation 14 15 of “ Rhizomys ” pinjoricus to Anepsirhizomys as suggested by Flynn et al. (1990) and 16 Flynn (2009) is not supported by this work. Anepsirhizomys pinjoricus is actually more 17 18 derived and should be reallocatedFor into a genusPeer of its own. Review 19 The African Tachyoryctini (T. pliocaenicus and more derived species) would come 20 21 from an immigration of a species close to Eicooryctes kaulialensis in Africa. This is in 22 agreement with the fact that Eicooryctes displays many derived characters shared by 23 24 and even, for some of them, accentuated in the more modern species of Tachyoryctes 25 (e.g., hypsodonty, absence of mesolophid on the m2 and m3, lack of protosinus on the 26 27 M1). 28 29 30 Tribe Rhizomyini. This clade is constituted by Miorhizomys pilgrimi plus more 31 derived taxa. The origin of the Rhizomyini is to be found at the beginning of the Late 32 33 Miocene in a Miorhizomy -like form. According to our consensus trees, the most basal 34 taxon of this clade is Miorhizomys pilgrimi . 35 After studying Brachyrhizomys shansius , Flynn (2009) concluded that this taxon 36 37 should be treated as a subgenus of Rhizomys (see above). For this reallocation to be 38 accepted, the genus Cannomys would have to be synonymized with Rhizomys . 39 40 However, the former differs from the latter in its smaller size and in having proodont 41 incisors and smooth sole pads (Nowak, 1999). Therefore, we favor the restablishment of 42 43 the genus Brachyrhizomys. Cannomys is the sister-group of Rhizomys . 44 45 46 Conclusion 47 48 49 The most basal representatives of the subfamily Rhizomyinae belong s to 50 Prokanisamys . The oldest species of this genus is P. kowalskii from the Late Oligocene 51 52 of Pakistan. This record is followed in age by P. arifi and P. major from Pakistan. 53 These two species persisted into the Middle Miocene of Pakistan, where they were 54 55 56 57 58 59 60 Cladistics Cladistics Page 42 of 68

1 2 3 4 López-Antoñanzas, R. et al. 38 5 6 7 sympatric with Kanisamys indicus . In Thailand, P. benjavuni is recorded in the Middle 8 9 Miocene. During Late Miocene times, the rhizomyines experienced their greatest 10 diversification with 14 species distributed across five genera ( Kanisamys , Rhizomyides , 11 12 Protachyoryctes , Eicooryctes and Miorhizomys ) in Pakistan, India and China. Our 13 results call the monophyly of several of these genera into question. The Asian Pliocene 14 15 rhizomyine record suggests that the diversity of these animals dropped there at this 16 time. In fact, they are only represented by four species belonging to four genera 17 18 (Rhizomyides , BrachyrhizomysFor, Anepsirhizomys Peer and a still unamedReview genus for “ A.” 19 pinjoricus ), which occurred in Afghanistan, Pakistan, India and China. 20 21 African fossil Rhizomyinae are scarce; the oldest and sole Early Miocene 22 representative is Prokanisamys sp. from Jebel Zelten (Libya). This recent discovery 23 24 provides evidence of an Early Miocene migration of members of this family between 25 Asia and Africa through the “ Gomphotherium landbridge”. In Africa, a single species of 26 27 rhizomyine is known from the Middle Miocene (Pronakalimys andrewsi from Fort 28 Ternan, Kenya), two from the Late Miocene (Nakalimys lavocati from Nakali, Kenya, 29 30 and the recently discovered Protachyoryctes makooka from the Middle Awash, 31 Ethiopia). From the Pliocene Hadar Formation (Afar, Ethiopia) a single species of 32 33 Tachyoryctes is known (T. pliocenicus ). The most recent fossil species of Tachyoryctes 34 known to date is T. konjiti , which comes from the Pleistocene of Melka-Kunturé 35 (Ethiopia). 36 37 Together with other rodents such as the Ctenodactylidae (López-Antoñanzas and 38 Knoll, 2011), Diatomyidae (López-Antoñanzas, 2011), basal Mur oideaae (López- 39 40 Antoñanzas, 2009), Zapodidae (López-Antoñanzas and Sen, 2006) and Thryonomyidae 41 (López-Antoñanzas et al., 2004), the rhizomyines provide a fine example of long 42 43 distance dispersal of rodents consistent with geological evidence, indicating the 44 establishment of an Early Miocene corridor between Afro-Arabia and Eurasia. 45 46 Rhizomyines survive in Asia and Africa, and their phylogenetic relationships imply the 47 ancestry to be found in Asia. This subfamily underwent a wide geographic expansion 48 49 during Miocene times. From a morphological viewpoint African fossil rhizomyines 50 appear close to the Asian ones, and the extant African rhizomyines share more 51 52 synapomorphies with certain Pakistani Late Miocene species than with the older 53 African taxa. The hypothesis that this subfamily dispersed more than once from Asia 54 55 56 57 58 59 60 Cladistics Page 43 of 68 Cladistics

1 2 3 4 López-Antoñanzas, R. et al. 39 5 6 7 into Africa could explain those facts and is supported by our results. A firstInitial 8 9 geo dispersal of Prokanisamys from Pakistan to East Africa would have taken place at 10 the beginning of the Early Miocene. From this first event Prokanisamys sp. from Jebel 11 12 Zelten would have originated. The second and third dispersalist events may have taken 13 place not earlier than 18 Myr and would have given rise to the Middle Miocene 14 15 Pronakalimys and to the Late Miocene Nakalimys (Fig. 9). The hypothesis that African 16 Nakalimys could have originated from the dispersal into Africa of an Asian Kanisamys 17 18 by Middle Miocene times isFor not supported Peer by our results. Review 19 With respect to the Late Miocene Protachyoryctes makooka and the African 20 21 Tachyoryctini, two unidirectional dispersal events from southern Asia to Africa took 22 place in Late Miocene times (not earlier that 8.2 Myr). The first one would have been at 23 24 the origin of Protachyoryctes makooka , whereas the derivation of the African 25 Tachyoryctini (Tachyoryctes spp.) would have come from an independent entery of this 26 27 group into Africa (Fig. 9). 28 29 30 Acknowledgements 31 32 33 We sincerely thank C. Denys (Muséum National d’Histoire Naturelle, Paris), F. Mayer 34 and N. Lange (Museum für Naturkunde der Humboldt-Universität, Berlin), G. Rößner 35 36 (Staatssamlung für Paläontologie und Geologie, Munich), P. Holroyd (University of 37 California, Berkeley), J. Galkin, J. Meng, R. O’Leary and C. Mehling (American 38 39 Museum of Natural History, New York) for having made available the rhizomyine 40 material under their care. D. Brandy (Neuville sur Touques), H. de Bruijn (Utrecht 41 42 University, Utrecht), P. Jenkins (The Natural History Museum, London), B. Marandat 43 and L. Marivaux (Université Montpellier II, Montpellier), P. Mein (Université Claude 44 45 Bernard, Lyon I, Villeurbanne), S. Sen (Muséum national d’Histoire naturelle, Paris), P. 46 Shepherd (British Geological Survey, Nottinghamshire), J. Thomas and E. Fara 47 48 (Université de Bourgogne, Dijon) answered all our inquiries. W. Wessels (Utrecht 49 University, Utrecht) kindly allowed us to use the photos of Prokanisamys arifi in our 50 figure 1. M. R. Dawson (Carnegie Museum of Natural History, Pittsburgh) and an 51 52 anonymous reviewer enhanced this work through careful, critical reading. The sojourns 53 in Munich (RLA and FK) and Paris (RLA) were funded by the Alexander von 54 55 56 57 58 59 60 Cladistics Cladistics Page 44 of 68

1 2 3 4 López-Antoñanzas, R. et al. 40 5 6 7 Humboldt Foundation through sponsorships of renewed research stays in Germany and 8 9 by the EDIT Gender Action Plan, respectively. This research received support from the 10 SYNTHESYS Project (http://www.synthesys.info/), which is financed by European 11 12 Community Research Infrastructure Action under the FP7 "Capacities" Program. RLA 13 and FK are currently supported by the Ramón y Cajal Program and the research projects 14 15 CGL2011-24829 and CGL2009-12143, respectively, of which they are PI. 16 17 18 References For Peer Review 19 20 21 Alemseged, Z., Geraads, D. 2000. A new Middle Pleistocene fauna from the Busidima– 22 Telalak region of the Afar, Ethiopia. C.R. Acad. Sci., Earth and Planetary Sciences 23 24 331, 549–556. 25 Allen, G. M. 1939. A checklist of African mammals. Bull. Mus. Comp. Zool. Harvard 26 27 Univ. 83, 1–763. 28 Beolchini, F., Corti, M. 2004. The taxonomy of the genus Tachyoryctes : a geometric 29 30 morphometric approach. Ital. J. of Zool. 71, 35–43. 31 Black, C. C. 1972. Review of fossil rodents from the Neogene Siwalik Beds of India 32 33 and Pakistan. Palaeontology 15, 238–266. 34 Blyth, E. 1851. Notice of a collection of Mammalia, Birds, and Reptiles, procured at or 35 36 near the station of Chérra Punji in the Khásia hills, north of Sylhet. J. Asiat. Soc . 37 Bengal 20, 517–524. 38 39 Bohlin, B. 1946. The fossil mammals from the Tertiary deposits of Taben-Baluk, 40 Western Kansu, Part II: Simplicidentata, Carnivora, Artiodactyla, Perissodactyla, and 41 42 Primates. Palaeontol. Sinica C 8b, 1–259. 43 Brandy, L.D. 1979a. Rongeurs nouveaux du Néogène d’Afghanistan. C.R. Acad. Sci., 44 45 Série D 289, 81–83. 46 Brandy, L.D. 1979b. Etude de rongeurs muroides du néogène supérieur et du 47 48 quaternaire d'Europe, d'Afrique du nord et d'Afghanistan: évolution, biogéographie, 49 correlations. PhD thesis: Université des Sciences et Techniques du Languedoc. 50 Brandy, L.D. 1981. Rongeurs muroïdés de Néogène supérieur d'Afghanistan : évolution, 51 52 biogéographie, corrélations. Paleovertebrata 11, 133–179. 53 Bremer, K. 1994. Branch support and tree stability. Cladistics 10, 295–304. 54 55 56 57 58 59 60 Cladistics Page 45 of 68 Cladistics

1 2 3 4 López-Antoñanzas, R. et al. 41 5 6 7 Bucher, J. E. 1982. Family Rhizomyidae. In: Honacki, J., Kinmam, K. E., Koeppl, J. W. 8 9 (Eds.), Mammals Species of the World, First Edition. Lawrence: Association of 10 Systematics Collections, pp. 477-478. 11 12 Chaimanee, Y., Yamee, C., Marandat, B., Jaeger, J. J. 2007. First Middle Miocene 13 Rodents from the Mae Moh Basin (Thailand): Biochronological and 14 15 Paleoenvironmental Implications. Bull. Carnegie Mus. 39, 157-163. 16 Colbert, E. H. 1933. Two new rodents from the Lower Siwalik beds of India. Am. Mus. 17 18 Novit. 633, 1–6. For Peer Review 19 Colbert, E. H. 1935. Siwalik Mammals in the American Museum of Natural History. 20 21 Trans. Am. Phil. Soc. New Series 26, 1–401. 22 Colbert, E. H., Hooijer, D. A. 1953. Pleistocene mammals from the limestone fissures 23 24 of Szechwan, China. Bull. Am. Mus. Nat. Hist. 102, 1–134. 25 Corbet, G. B., Hill, J. E. 1991. A World List of Mammalian Species, Third edition. 26 27 Oxford University Press, Oxford. 28 De Bruijn, H., Hussain, S. T., Leinders, J. M. 1981. Fossil rodents from the Murree 29 30 formation near Banda Daud Shah, Kohat, Pakistan. Proc. Kon. nederl. Akad. 31 Wetensch. B 84, 71–99. 32 33 Ellerman, J.R. 1940. The families and genera of living rodents. Volume I. Rodents other 34 than Muridae. London: British Museum (Natural History). 35 Ellerman, J.R. 1941. The families and genera of living rodents. Volume II. Family 36 37 Muridae. London: British Museum (Natural History). 38 Ellerman, J. R., Morrison-Scott, T. C. S. 1951. Checklist of Palaearctic and Indian 39 40 Mammals, 1758 to 1946, First Edition. British Museum (Natural History), London. 41 Ellerman, J. R., Morrison-Scott, T. C. S. 1966. Checklist of Palaearctic and Indian 42 43 Mammals, 1758 to 1946, Second Edition. British Museum (Natural History), 44 London. 45 46 Endo, H., Satoh, K., Cuisin, J., Stafford, B., Kimura, J. 2001. Morphological adaptation 47 of the masticatory muscles and related apparatus in Asian and African Rhizomyinae 48 49 species. Mammal Study 26, 101–108. 50 Fejfar, O., Horá ček, I. 2006. The Early Miocene mammalian assemblages in Jebel 51 52 Zelten, Libya. Lynx 37, 95–105. 53 54 55 56 57 58 59 60 Cladistics Cladistics Page 46 of 68

1 2 3 4 López-Antoñanzas, R. et al. 42 5 6 7 Flynn, L. J. 1981. Biostratigraphy and systematics of Siwalik Rhizomyidae (Rodentia). 8 9 Unpublished PhD dissertation, University of Arizona, Tucson. 10 Flynn, L. J. 1982a. Systematic revision of Siwalik Rhizomyidae (Rodentia). Geobios 11 12 15, 327–389. 13 Flynn, L. J. 1982b. A revision of fossil rhizomyid rodents from northern India and their 14 15 correlation to a rhizomyid biochronology of Pakistan. Geobios 15, 583–588. 16 Flynn, L. J. 1983. Sur l'âge de la faune de Vertébrés du bassin de Bamian, Afghanistan. 17 18 C.R. Acad. Sci., Série IIFor 297, 687–690. Peer Review 19 Flynn, L. J. 1984. Mosaic evolution in a family of fossorial rodents. In: Buffetaut, E., 20 21 Mazin, J. M., Salmon, E. (Eds.), Actes du Symposium paléontologique Georges 22 Cuvier. Montbéliard: Ville de Montbéliard. pp. 185–195. 23 24 Flynn, L. J. 1985. Evolutionary patterns and rates in Siwalik Rhizomyidae (Rodentia). 25 Acta Zool. Fenn. 170, 141–144. 26 27 Flynn, L. J. 1986. Species longevity, stasis, and stairsteps in rhizomyid rodents. Univ. 28 Wyoming Contr. Geol, Spec. Pap. 3, 273–285. 29 30 Flynn, L. J. 1990. The natural history of rhizomyid rodents. In: Nevo, E., Reig, O. A. 31 (Eds.), Evolution of Subterranean Mammals at the Organismal and Molecular 32 33 Levels. A. R. Liss, New York. pp. 155–183. 34 Flynn, L. J. 1993. A new bamboo rat from the Late Miocene of Yushe Basin. Vert. 35 Palasiatica 31, 95–101. 36 37 Flynn, L. J. 2009. The Antiquity of Rhizomys and Independent Acquisition of Fossorial 38 Traits in Subterranean Muroids. Bull. Am. Mus. Nat. Hist. 331, 128–156. 39 40 Flynn, L. J., Barry J. C., Morgan M. E., Pilbeam D., Jacobs L. L., Lindsay E. H. 1995. 41 Neogene Siwalik mammalian lineages: Species longevities, rates of change, and 42 43 modes of speciation. Palaeogeogr. Palaeocl. Palaeoeco.115, 149–264. 44 Flynn, L. J., Heintz, E., Sen, S., Brunet, M. 1983. A new Pliocene tachyoryctine 45 46 (Rhizomyidae, Rodentia) from Lataband, Sarobi Basin, Afghanistan. Proc. Kon. 47 nederl. Akad. Wetensch. B 86, 61–68. 48 49 Flynn, L. J., Qi, G. 1982. Age of the Lufeng, China, hominoid locality. Nature 298, 50 746–747. 51 52 Flynn, L. J., Sabatier, M. 1984. A Muroid Rodent of Asian affinity from the Miocene of 53 Kenya. Journal of Vertebrate Paleontology 3, 160–165. 54 55 56 57 58 59 60 Cladistics Page 47 of 68 Cladistics

1 2 3 4 López-Antoñanzas, R. et al. 43 5 6 7 Flynn, L. J., Sahni, A., Jaeger, J. J., Singh, B., Bhatia, S. B. 1990. Additional fossil 8 9 rodents from the Siwalik Beds of India. Proc. Kon. nederl. Akad. Wetensch. 93, 7– 10 20. 11 12 Gallotti, R., Collina, C., Raynal, J. P., Kieffer, G., Geraads, D., Piperno, M. 2010. The 13 Early Middle Pleistocene Site of Gombore II (Melka Kunture, Upper Awash, 14 15 Ethiopia) and the Issue of Acheulean Bifacial Shaping Strategies. Afr. Archaeol. 16 Rev. 27, 291–322. 17 18 Geraads, D. 1998. RongeursFor du Miocène Peersupérieur de Chorora Review (Ethiopie): Cricetidae, 19 Rhizomyidae, Phiomyidae, Thryonomyidae, Sciuridae. Paleovertebrata 27, 203–216. 20 21 Geraads, D., Alemseged, Z., Reed, D., Wynn J., Roman, D. C. 2004. The Pleistocene 22 fauna (other than Primates) from Asbole, lowerAwash Valley, Ethiopia, and its 23 24 environmental and biochronological implications. Geobios 37, 697–718. 25 Gogolevskaya, I. K., Veniaminova, N. A., Kramerov, D. A. 2010. Nucleotide sequences 26 27 of B1 SINE and 4.5SI RNA support a close relationship of zokors to blind mole rats 28 (Spalacinae) and bamboo rats (Rhizomyinae). Gene 460, 30–38. 29 30 Goloboff, P., Farris, J. 2001. Methods for quick consensus estimation. Cladistics 17, 31 26–34. 32 33 Goloboff, P., Farris, J. S., Källersjö, M., Oxelman, B., Ramírez, M. J., Szumika C. A. 34 2003. Improvements to resampling measures of group support. Cladistics, 19, 324– 35 332. 36 37 Goloboff, P., Farris, J., Nixon K. C. 2008. TNT, a free program for phylogenetic 38 analysis. Cladistics 24, 774–786. 39 40 Gray, J.E. 1831. Characters of three new genera, including two new species of 41 Mammalia from China. Proc. Comm. Zool. Soc. London 1, 94–98. 42 43 Gupta, S. S., Verma, B. C., Tewari, A. P. 1978. A new fossil rodent Rhizomyoides Formatted: English (U.S.) 44 saketiensis sp. nov. from the Tatrot member (Astian) of Siwalik fossil park, Saketi, 45 46 Sirmur distt, Himachal Pradesh. J. Paleont. Soc. India 21-22, 112–115. 47 Haile-Selassie, Y., Wolde Gabriel, G., White, T. D., Bernor, R. L., DeGusta, D. L., 48 49 Renne, P., Hart, W. K., Vrba, E., Ambrose, S., Howell, F. C. 2004. Mio-Pliocene 50 mammals from the Middle Awash, Ethiopia. Geobios 37, 536–552. 51 52 Heck, J. G. 1851a. Iconographic Encyclopaedia of Science, Literature, and Art. Vol II: 53 Botany, Zoology, Anthropology, and Surgery. New York, R. Garrigue. 54 55 56 57 58 59 60 Cladistics Cladistics Page 48 of 68

1 2 3 4 López-Antoñanzas, R. et al. 44 5 6 7 Heck, J. G. 1851b. Iconographic Encyclopaedia of Science, Literature, and Art. Vol I: 8 9 Mathematics and Astronomy, Physics and Meteorology, Chemistry, Mineralogy, 10 Geognosy and Geology. New York, R. Garrigue. 11 12 Heuglin, T. von, Fitzinger, L. J. 1866. Systematische Übersicht der Säugethiere 13 Nordost- Afrika's mit Einschluß der arabischen Küste, des rothen Meeres, der 14 15 Somáli- und der Nilquellen- Länder, südwärts bis zum vierten Grade nördlicher 16 Breite. Sitzungsberichte Math.-Nat. Classe Kais. Akad. Wissensch., Abt. 1. 54, 537– 17 18 611. For Peer Review 19 Hinton, M. A. C. 1933. Diagnoses of new genera and species of rodents from Indian 20 21 Tertiary deposits. Ann. Mag. nat. Hist. Ser. 10. 12, 620–622. 22 Hodgson, B.H.1841. New species of Rhizomys discovered in Nepal. Calcutta J. Nat. 23 24 Hist. 2, 60–62. 25 IUCN. 2010. IUCN red list of threatened species. Version 2010.1. Available at 26 27 www.iucnredlist.org 28 International Commission on Zoological Nomenclature. 1999. International Code of 29 30 Zoological Nomenclature. Natural History Museum London, Fourth Edition. The 31 Natural History Museum, London. 32 33 Jacobs, L. L. 1978. Fossil Rodents (Rhizomyidae and Muridae) from Neogene Siwalik 34 deposits, Pakistan. Mus. North. Ariz. Bull. 52, 1–103. 35 36 Jacobs, L. L., Flynn, L. J. and Dows, W. R. 1989. Neogene rodents of Southern Asia. 37 Nat. Hist. Mus. Los Angeles County, Science Series 33, 157–177. 38 39 Jacobs L. L., Flynn L. J., Downs W. R. & Barry J. C. 1990. Quo vadis , Antemus ? The 40 Siwalik muroid record. In: Lindsay, E. H., Fahlbusch, V., Mein, P. (Eds.), European 41 42 Neogene mammal chronology. Plenum Press, New York. pp. 573–586. 43 Jansa, S. A., Weksler, M. 2004. Phylogeny of muroid rodents: relationships within and 44 45 among major lineages as determined by IRBP gene sequences. Mol. Phylogenet. 46 Evol. 31, 256–276. 47 48 Lang, J., Lavocat, R. 1968. Première découverte d’une faune de vertébrés dans le 49 tertiaire d’Afghanistan et datation de la série de Bamian. C.R. Acad. Sci., Série D 50 266, 79–82. 51 52 53 54 55 56 57 58 59 60 Cladistics Page 49 of 68 Cladistics

1 2 3 4 López-Antoñanzas, R. et al. 45 5 6 7 Lavocat, R. 1978. Rodentia and Lagomorpha. In: Maglio, V.J., Cooke, H.B.S. (Eds.), 8 9 Evolution of African mammals. Harvard University Press, Cambridge, MA, pp. 69– 10 89. 11 12 Lindsay, E. H. 1996. A new eEumyarionine cricetid from Pakistan. Acta Zool. Cracov. 13 39, 27–40. 14 15 Lindsay, E. H., Flynn, L. J., Cheema, I. U., Barry, J. C., Downing, K., Rajpar, A. R., 16 Raza, S. M. 2005. Will Downs and the Zinda Pir Dome. Palaeontol. Electron. 8, 19A, 17 18 18 pp. For Peer Review 19 López-Antoñanzas, R. 2009. First Potwarmus from the Miocene of Saudi Arabia and 20 21 the early phylogeny of murines (Rodentia: Muroidea). Zool. J. Linn. Soc. 156, 664– 22 679. 23 24 López-Antoñanzas, R. 2011. First Diatomyid out of Asia. Naturwissenschaften 98, 117– 25 123. 26 27 López-Antoñanzas, R. 2012. Dental morphology and wear pattern in Tachyoryctes 28 (Spalacidae, Rodentia). Mammalia. 72. 29 30 López-Antoñanzas, R., Sen, S. 2006. New jumping mouse (Rodentia, Zapodidae) from 31 the Lower Miocene of Saudi Arabia and a contribution to the knowledge of zapodid 32 33 phylogeny. J. Vertebr. Paleont. 26, 170–181. 34 López-Antoñanzas, R., Knoll, F. 2011. A comprehensive phylogeny of the gundis 35 (Ctenodactylinae, Ctenodactylidae). J. Syst. Paleontol. 9, 379–398. 36 37 López-Antoñanzas, R., Wesselman, H. B. In press. Tachyoryctes makooka 38 (Tachyoryctini, Spalacidae, Rodentia) and its bearing on the phylogeny of the 39 40 Tachyoryctini. Paleontology. 41 López-Antoñanzas, R., Sen, S., Mein, P. 2004. Systematics and phylogeny of the cane- 42 43 rats (Rodentia, Thryonomyidae). Zool. J. Linn. Soc. 142, 423-444. 44 Lydekker, R. 1878. Notice of Siwaliks mammals. Rec. Geol. Sur. India 6, 64–104. 45 46 Lydekker, R. 1884. Rodents and new ruminants from the Siwaliks, and synopsis of 47 Mammalia. Palaeontol. Indica 10, 105-134. 48 49 Lydekker, R. 1885. Catalogue fossil Mammalia British Museum (Natural History), Part 50 1. British Museum (Natural History), London. 51 52 McKenna, M. C., Bell, S. K. 1997. Classification of Mammals above the Species level. 53 Columbia University Press, New York. 54 55 56 57 58 59 60 Cladistics Cladistics Page 50 of 68

1 2 3 4 López-Antoñanzas, R. et al. 46 5 6 7 Maddison, W. P., Maddison, D. R. 2009. Mesquite: a modular system for evolutionary 8 9 analysis, Version 2.6. Mesquite Project, Vancouver. 10 Mein, P., Freudenthal, M. 1971. Les Cricetidae (Mammalia, Rodentia) du Néogène 11 12 moyen du Vieux–Collonges. Partie 1, Le genre Cricetodon Lartet, 1851. Scripta Geol. 13 5, 1–51. 14 15 16 Mein, P., Ginsburg, L. 1985. Les rongeurs miocènes de Li (Thaïlande). C.R. Acad. Sci., 17 18 Série II 301, 1369–1374.For Peer Review 19 Mein, P., Ginsburg, L. 1997. Les mammifères du gisement miocène inférieur de Li Mae 20 21 Long, Thaïlande: systematique, biostratigraphie et paléoenvironnement. 22 Geodiversitas 19, 783–844. 23 24 Miller, G.S., Gidley, J.W. 1918. Synopsis of the supergeneric groups of rodents. J. 25 Wash. Acad. Sci. 8, 431–448. 26 27 Misonne, X. 1971. Order Rodentia. In: Meester, J., Setzer, H.W. (Eds.), The Mammals 28 of Africa: an identification manual. Smithsonian Institution Press, Washington, 1–39. 29 30 Munthe, J. 1980. Rodents of the Miocene Daud Khel Local Fauna, Mianwali District, 31 Pakistan. Part 1. Sciuridae, Gliridae, Ctenodactylidae, and Rhyzomyidae. Milwaukee 32 33 Publ. Mus. Contr. Biol. Geol. 34, 1–36. 34 Musser, G. G., Carleton, M. D. 1993. Family Muridae. In: Wilson, D. E. Reeder, D. M. Formatted: English (U.S.) 35 36 (Eds.), Mammal Species of the World, Second Edition. The Johns Hopkins 37 University Press, Maryland, pp. 501–755. 38 39 Musser, G. G., Carleton, M. D. 2005. Superfamily Muroidea. In: Wilson D. E., Reeder, 40 D. M. (Eds.), Mammal Species of the World, Third Edition. The Johns Hopkins 41 42 University Press, Maryland, pp.894–1531. 43 Norris, R. W., K. Zhou, C. Zhou, G. Yang, C. W. Kilpatrick, R. L. Honeycutt. 2004. 44 45 The phylogenetic position of the zokors (Myospalacinae) and comments on 46 the families of muroids (Rodentia). Mol. Phy. Evol. 31, 972–978. 47 48 Nowak, R. M. 1999. Muridae: Rats, Mice, Hamsters, Voles, Lemmings, and Gerbils. In: 49 Nowak R. M. (Ed.), Walker’s mammals of the World, 6th Edition. The Johns 50 Hopkins University Press, Maryland, pp.1344–1616. 51 52 Palmer, T. S. 1897. A list of the generic and family names of rodents. Proc. Biol.Soc. 53 Wash. XI, 241–270. 54 55 56 57 58 59 60 Cladistics Page 51 of 68 Cladistics

1 2 3 4 López-Antoñanzas, R. et al. 47 5 6 7 Parmar, V., Prasad, G.V.R. 2006. Middle Miocene rhizomyid rodent (Mammalia) Formatted: English (U.K.) 8 9 from the Lower Siwalik Subgroup of Ramnagar, Udhampur District, Jammu and 10 Kashmir, India. N. Jb. Geol. Paläont., Abh. 6, 371–384. 11 12 Prasad, K. N. 1968. The vertebrate fauna from the Siwalik beds of Haritalyangar, 13 Himachal Pradesh, India. Palaeont. Indica 39, 1–79. 14 15 Raffles, T.S. 1821. Descriptive Catalogue of a Zoological Collection. Trans. Linn. Soc. 16 Lond. XIII 239–274. 17 18 Rahm, U.1980. Die AfrikanischeFor Wurzelratte Peer Tachyoryctes .Review Ziemsen, A., Lutherstadt 19 Wittenberg. 20 21 Rüppell, E. 1836. Neue Wirbelthiere zu der Fauna von Abyssinien gehörig. Part 7. 22 Schmerber, S., Frankfurt am Main, pp.17–36. 23 24 Rüppell, E. 1842. Säugethiere aus der Ordnung der Nager, beobachtet im nordöstlichen 25 Africa. Mus. Senckenb. 3, 99–116. 26 27 Sabatier, M. 1978. Un nouveau Tachyoryctes (Mammalia, Rodentia) du bassin Pliocene 28 de Hadar (Ethiopie). Geobios 11, 95–99. 29 30 Sabatier, M. 1979. Les rongeurs des sites aà hominides de Hadar et Melka-Kunturé 31 (Ethiopie). PhD thesis: Université des Sciences et Techniques du Languedoc. 32 33 Sabatier, M. 1982. Les rongeurs des sites pléistocènes de Melka-Kunturé, Ethiopie, 34 Abbay 11, 45–64. 35 Sehgal, R. K., Patnaik, R. In press. New muroid rodent and Sivapithecus dental remains 36 37 from the Lower Siwalik deposits of Ramnagar (JandK, India): Age implication, 38 Quatern. Int. 39 40 Sen, S. 1983. Rongeurs et lagomorphes du gisement pliocène de Pul-e Charkhi, bassin 41 de Kabul, Afghanistan. Bull. Mus. Hist. Nat. Paris Séries 5C, 33–74. 42 43 Simpson, G. G. 1945. The principles of classification and a classification of mammals. Formatted: French (France) 44 Bull. Am. Mus. Nat. Hist. 85, 1–350. 45 46 Teilhard de Chardin, P. 1942. New rodents of the Pliocene and lower Pleistocene of 47 north China. Pub. Inst. Géo-Biol. 9, 1–101. 48 49 Thomas, O. 1896. On the genera of rodents: an attempt to bring up to date the current 50 arrangement of the Order. Proc. Zool. Soc. Lond. 1012–1028. 51 52 Thomas, O. 1915a. Notes on the Asiatic Bamboo-Rats (Rhizomys, etc.). Ann. Mag. Nat. 53 Hist. XVI, 56–61. 54 55 56 57 58 59 60 Cladistics Cladistics Page 52 of 68

1 2 3 4 López-Antoñanzas, R. et al. 48 5 6 7 Thomas, O. 1915b. Further notes on Asiatic Bamboo-Rats. Ann. Mag. Nat. Hist. XVI, 8 9 313–317. 10 Tong, H., Jaeger J.J. 1993. Muroid rodents from the Middle Miocene Fort Ternan 11 12 locality (Kenya) and their contribution to the phylogeny of muroids. 13 Palaeontographica, Abteilung A 229, 51–73. 14 15 Tullberg, T. 1899. Über das System der Nagethiere: eine phylogenetische Studie. 16 Akademische Buchdruckerei. 17 18 Vasishat, R. N., Gaur R., ChopraFor S. R. K.Peer 1978. Community Review structure of Middle Sivalik 19 vertebrates from Haritalyangar (H.P.), India. Palaeogeogr. Palaeocl. Palaeoeco. 23, 20 21 131–140. 22 Welcomme J. L., Ginsburg, L. 1997. Mise en evidence de l’Oligocène sur le territoire 23 24 des Bugti (Balouchistan, Pakistan). C.R. Acad. Sci., Série IIa 325, 999–1004. 25 Welcomme, J. L., Benammi, M., Crochet, J. Y., Marivaux, L., Métais, G., Antoine, P. 26 27 O., Balouch, I. 2001. Himalayan Forelands: palaeontological evidence for Oligocene 28 detritic deposits in Bugti Hills (Balochistan, Pakistan). Geol. Mag. 138, 397–405. 29 30 Wesselman, H. B., Black, M. T., Asnake, M. 2009. Small Mammals. In: Haile-Selassie, 31 Y., WoldeGabriel, G. (Eds.), Ardipithecus kadabba : Late Miocene Evidence from 32 33 the Middle Awash, Ethiopia (The Middle Awash Series). University of California 34 Press, London, pp. 105–133. 35 Wessels, W. 2009. Miocene rodent evolution and migration. Muroidea from Pakistan, 36 37 Turkey and Northern Africa. Geol. Ultraiectina 307, 1–290. 38

39 40 Wessels, W., de Bruijn H. 2001. Rhizomyidae from the Lower Manchar Formation 41 (Miocene, Pakistan). Ann. Carnegie Mus. 70, 143–168. 42 43 Wessels,W., de Bruijn, H., Hussain, S. T., Leinders, J. 1982. Fossil rodents from the 44 Chinji Formation, Banda Daud Shah, Kohat, Pakistan. Proc. Kon. nederl. Akad. 45 46 Wetensch. B 85, 337–364. 47 Wessels, W., Fejfar, O., Peláez-Campomanes, P., De Bruijn, H. 2003. Miocene small 48 49 mammals from Jebel Zelten, Libya. Coloq. Paleont.Vol. Extr., 699–715. 50 Wessels, W., Fejfar, O., Peláez-Campomanes, P., Van Der Meulen, A., De Bruijn, H., 51 52 El-Arnauti, A. 2008. The age of the small mammal faunas from Jabal Zaltan, Libya. 53 Garyounis Sci. Bull., spec. issue 5, 129–138. 54 55 56 57 58 59 60 Cladistics Page 53 of 68 Cladistics

1 2 3 4 López-Antoñanzas, R. et al. 49 5 6 7 Winge, H. 1887. Jordfundne og nulevende Gnavere (Rodentia) fra Lagoa Santa, Minas 8 9 Geraes, Brasilien. E Museo Lundii, University of Copenhagen 1, 1–178. 10 Wood, A. E. 1937. Fossil rodents from the Siwalik beds of India. Am. J. Sci. 34, 64–76. 11 12 13 14 15 16 17 18 For Peer Review 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Cladistics Cladistics Page 54 of 68

1 2 3 Fig. 1. Distribution of the extant Rhizomyines (after IUCN data) and fossil localities. 1, 4 5 Taoyang, Yushe Basin, Shanxi, China; 2, Lufeng, Chuxiong Yi, Yunnan, China; 3, Mae 6 7 Long and Mae Moh, Lamphang Province, Thailand; 4, 5, Haritalyangar, India; 6, 8 9 10 Punjab, India; 7, Ramnagar, India; 8, Potwar Plateau (Dhok Pathan-Khaur, Chinji- 11 12 Nagri, Tatrot-Hasnot regions), Pakistan; 9, Banda Daud Shah (Pakistan); 10, Zinda Pir 13 14 Dome (Pakistan); 11, 12, Sind (Pakistan); 13, Lataband, Sarobi Basin, Afghanistan; 14, 15 16 Pul-e Charkhi Kabul, Afghanistan; 15, Bamian Basin, Afghanistan; 16, Jebel Zelten 17 18 (Libya); 17, KadaFor Hadar, Hadar Peer Formation, Ethiopia;Review 18, Digiba Dora Vertebrate 19 20 21 Locality 1 (Middle Awash, Ethiopia); 19, Melka Kunturé, Ethiopia; 20, Nakali, Kenya; 22 23 21, Fort Ternan, Kenya. 24 25 26 27 Fig. 2. Rhizomyidae dental terminology used in this paper. Prokanisamys arifi —A. Left 28 M1, H-GSP 8311/4359. —B. Left M2, H-GSP 8311/4369. —C. Left M3, H-GSP 29 30 8311/4381. —D. Left m1, H-GSP 8311/4294. —E. Left m2, H-GSP 8311/4316. —F., 31 32 Left m3, H-GSP 8311/4334. Scale bar equals 1 mm. 33

34 35 Fig. 3. Cheek teeth of Rhizomys sinensis , Cannomys badius and Tachyoryctes 36 37 splendens . —A. Lower cheek teeth of Rhizomys sinensis (specimen C.G. 1912-851) 38 with m1-m3. —B. Upper cheek teeth of the same specimen with M1-M3. —C. Lower 39 40 cheek teeth of Cannomys badius (specimen C.G. 1860-382) with m1-m3. —D. Upper 41 42 cheek teeth of the same specimen with M1-M3. —E. Lower cheek teeth of 43 Tachyoryctes splendens (specimen C.G. 1993-217) with m1-m3. —F. Upper cheek 44 45 teeth of the same specimen with M1-M3. Scale bar equals 2mm. 46 47 48 Fig. 4. Cheek teeth of Kanisamys nagrii , Kanisamys potwarensis , Miorhizomys 49 50 tetracharax , Rhizomys (Brachyrhizomys) sansius .—A. Maxilla of Kanisamys nagrii 51 52 (specimen AMNH 30000) with right M1-M3. —B. Maxilla of Kanisamys nagrii 53 (specimen AMNH 30000) with left M1-M2. —C. Mandible of Kanisamys nagrii 54 55 (specimen AMNH 39323) with m1-m3. —D. Maxilla of Kanisamys potwarensis 56 57 (specimen AMNH 39321) with M1-M3. —E. Mandible of Miorhizomys tetracharax 58 59 60 Cladistics Page 55 of 68 Cladistics

1 2 3 (specimen AMNH 39325) with m1-m3). —F. Mandible of Rhizomys (Brachyrhizomys) 4 5 shansius (specimen AMNH 117337) with m1-m3. —G. Maxilla of Rhizomys 6 (Brachyrhizomys) shansius (specimen AMNH 117337). 7 8 9 10 Fig. 5. Reconstruction of Bathyergus splendens provided by Rüppell (1836). 11 12 13 Fig. 6. Relationships of extant and extinct Rhizomyines provided by Flynn (1990). 14 15 16 Fig. 7. Consensus trees generated by the cladistic analysis of the Rhizomyidae 17 18 performed in thisFor paper (matrix Peer in Appendix Review2). —A. strict and semi-strict consensus 19 20 trees, nodes are designed in bold by numbers 40 to 62, Bremer and Relative Bremer 21 Indices are showed at the appropriate nodes. —B, majority-rule consensus tree. The 22 23 trees have a length of 158 steps, a consistency index (CI) of 0.456 and a retention index 24 25 (RI) of 0.819. 26 27 28 Fig. 8. Calibrated phylogeny based on the cladistic relationships presented in Fig. 7 and 29 30 the recorded temporal ranges (black). Grey bars represent missing ranges and missing 31 ancestral lineages. Error bars indicate age uncertainty. Chronostratigraphical data from 32 33 Sabatier (1979), de Bruijn et al. (1981), Flynn (1982a, 1983, 1986), Flynn et al. (1983, 34 35 1990, 1995, 2009), Sen (1983), Wessels et al. (2003), Chaimanee et al. (2007), 36 Wesselman et al (2009) and Wessels (2009). 37 38 39 40 Fig. 9 Geographic distribution of rhizomyine rodents plotted onto the strict consensus of 41 the most parsimonious trees of our analysis using Mesquite ver. 2.6 (Maddison and 42 43 Maddison, 2009) as a two-state character. 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Cladistics Cladistics Page 56 of 68

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 For Peer Review 19 20 21 22 23 24 25 Fig. 1. Distribution of the extant Rhizomyines (after IUCN data) and fossil localities. 1, Taoyang, Yushe 26 Basin, Shanxi, China; 2, Lufeng, Chuxiong Yi, Yunnan, China; 3, Mae Long and Mae Moh, Lamphang 27 Province, Thailand; 4, 5, Haritalyangar, India; 6, Punjab, India; 7, Ramnagar, India; 8, Potwar Plateau 28 (Dhok Pathan-Khaur, Chinji-Nagri, Tatrot-Hasnot regions), Pakistan; 9, Banda Daud Shah (Pakistan); 10, 29 Zinda Pir Dome (Pakistan); 11, 12, Sind (Pakistan); 13, Lataband, Sarobi Basin, Afghanistan; 14, Pul-e 30 Charkhi Kabul, Afghanistan; 15, Bamian Basin, Afghanistan; 16, Jebel Zelten (Libya); 17, Kada Hadar, 31 Hadar Formation, Ethiopia; 18, Digiba Dora Vertebrate Locality 1 (Middle Awash, Ethiopia); 19, Melka 32 Kunturé, Ethiopia; 20, Nakali, Kenya; 21, Fort Ternan, Kenya. 96x52mm (300 x 300 DPI) 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Cladistics Page 57 of 68 Cladistics

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 For Peer Review 19 20 21 22 23 24 25 26 27 28 29 Fig. 2. Rhizomyidae dental terminology used in this paper. Prokanisamys arifi —A. Left M1, H-GSP 30 8311/4359. —B. Left M2, H-GSP 8311/4369. —C. Left M3, H-GSP 8311/4381. —D. Left m1, H-GSP 8311/4294. —E. Left m2, H-GSP 8311/4316. —F., Left m3, H-GSP 8311/4334. Scale bar equals 1 mm. 31 134x86mm (300 x 300 DPI) 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Cladistics Cladistics Page 58 of 68

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 For Peer Review 19 20 21 22 23 24 25 26 Fig. 3. Fig. 3. Cheek teeth of Rhizomys sinensis , Cannomys badius and Tachyoryctes splendens . —A. Lower 27 cheek teeth of Rhizomys sinensis (specimen C.G. 1912-851) with m1-m3. —B. Upper cheek teeth of the 28 same specimen with M1-M3. —C. Lower cheek teeth of Cannomys badius (specimen C.G. 1860-382) with 29 m1-m3. —D. Upper cheek teeth of the same specimen with M1-M3. —E. Lower cheek teeth of Tachyoryctes 30 splendens (specimen C.G. 1993-217) with m1-m3. —F. Upper cheek teeth of the same specimen with M1- 31 M3. Scale bar equals 2mm. 32 102x58mm (300 x 300 DPI) 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Cladistics Page 59 of 68 Cladistics

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 For Peer Review 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 Fig. 4. Cheek teeth of Kanisamys nagrii , Kanisamys potwarensis , Miorhizomys tetracharax , Rhizomys (Brachyrhizomys) sansius .—A. Maxilla of Kanisamys nagrii (specimen AMNH 30000) with right M1-M3.—B. 48 Maxilla of Kanisamys nagrii (specimen AMNH 30000) with left M1-M2. —C. Mandible of Kanisamys nagrii 49 (specimen AMNH 39323) with m1-m3. —D. Maxilla of Kanisamys potwarensis (specimen AMNH 39321) with 50 M1-M3. —E. Mandible of Miorhizomys tetracharax (specimen AMNH 39325) with m1-m3). —F. Mandible of 51 Rhizomys (Brachyrhizomys) shansius (specimen AMNH 117337) with m1-m3. —G. Maxilla of Rhizomys 52 (Brachyrhizomys) shansius (specimen AMNH 117337). 53 250x350mm (300 x 300 DPI) 54 55 56 57 58 59 60 Cladistics Cladistics Page 60 of 68

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 For Peer Review 19 Fig.5 Reconstruction of Bathyergus splendens provided by Rüppell (1836) 20 35x14mm (300 x 300 DPI) 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Cladistics Page 61 of 68 Cladistics

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 For Peer Review 19 20 21 22 23 24 25 26 27 28 29 Fig. 6. Relationships of extant and extinct Rhizomyines provided by Flynn (1990). 30 55x36mm (300 x 300 DPI) 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Cladistics Cladistics Page 62 of 68

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 For Peer Review 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Cladistics Page 63 of 68 Cladistics

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 For Peer Review 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Cladistics Cladistics Page 64 of 68

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 For Peer Review 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Cladistics Page 65 of 68 Cladistics

1 2 3 Table 1 4 5 Character/taxon matrix used in the analysis of relationships of all extinct and extant 6 7 species of Rhizomyinae. Characters are listed in material and methods. Character 8 9 10 scoring: 0, 1 and 2, conditions of character; -, uncodable character; ?, character state 11 12 uncertain. 13 14 Table 2 15 16 Synapomorphies common to the 32MPTs plotted on the strict consensus tree. Exclusive 17 18 synapomorphiesFor are indicated Peer in bold. Italics Review indicate ambiguous synapomorphies. 19 20 21 Node numbers are shown in Figure 7. 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Cladistics Cladistics Page 66 of 68

1 2 3 4 5 6 7 8 9 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 31 32 33 34 35 36 37 38 39 40 41 42 43 44

10 Debruijnia arpati 0 0 1 ? ? ? ? ? 0 0 0 0 0&1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 0 0 0 0 2 0 0 0 0 0 2 0 11 2 4 2 2 1 2 2 2 1 2 1 4 1 1 2 4 1 1 1 0 1 1 1 4 2 1 1 1 1 1 3 1 1 1 1 1 3 1 1 1 1 2 3 2 12 Rhizomys sumatrensis 13 Rhizomys sinensis 2 4 2 2 1 2 2 2 1 2 1 For 4 1 1 2 Peer 4 1 1 1 0 0 Review 1 1 4 2 1 1 1 1 1 3 1 1 1 1 1 3 1 1 1 1 2 3 2 14 Rhizomys pruinosus 2 4 2 2 1 2 2 2 1 2 1 4 1 1 2 4 1 1 1 0 1 1 1 4 2 1 1 1 1 1 3 1 1 1 1 1 3 1 1 1 1 2 3 2 15 R. (Brachyrh.) shansius 1 3 2 2 1 2 2 2 1 2 1 0 1 1 2 4 1 1 1 0 1 1 1 4 2 0 1 1 1 1 2 1 1 1 1 1 2 1 1 1 1 1 2 2

16 R. (Brachyrh.) shajius 1 2 2 2 ? 1 ? ? ? ? ? ? ? ? 1 2 1 0 1 1 1 1 0 2 1 0 1 1 ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? 17 Cannomys badius 2 0&1 2 1 1 2 2 2 1 2 1 4 1 1 2 4 1 1 1 ? 1 1 1 4 2 1 1 1 1 1 2 1 1 1 1 1 2 1 1 1 1 2 2 2 18 Miorhizomys nagrii 1 1&2 2 1 1 1 1 2 1 0 1 0&1 1 1 0 0 1 0 1 1 0 1 0 0&2 1 0 1 1 0 1 2&3 0 1 1 1 1 2 0 1 1 1 1 1&2 1 19 20 Miorhizomys micrus 1 1 2 1 1 1 ? 2 1 0 1 0 1 0 0 2 1 0 1 0 1 1 0 2 1 0 1 0 ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? 21 Miorhizomys blacki 1 2 2 2 1 1 1 2 ? ? ? ? ? ? 0 0 1 0 1 1 1 1 0 2 1 0 1 1 ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? 22 Miorhizomys pilgrimi 1 3 2 2 1 1 1 2 ? ? ? ? ? ? 2 4 1 0 1 1 1 1 1 4 1&2 0 1 1 ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? 23 Miorhizomys harii 1 2 2 1 1 ? ? 2 1 1 1 0 1 0 1 1 0 0 1 1 ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?

24 Miorhizomys tetracharax 1 2&3 2 2 1 1 1 2 1 1 1 0 1 0 1 0 1 0 1 0 1 1 1 4 2 1 0 0 1 ? ? ? ? ? ? ? 2 1 1 ? 0 1 1 1 25 Miorhizomys choristos 1 2&3 2 2 1 1 1 2 1 1 1 0&1 1 1 2 4 1 0 1 1 1 1 1 4 2 1 1 1 1 1 0 1 1 1 1 1 3 1 1 1 1 0 1 1 26 1 2 2 0 0 1 ? 1 1 0 1 2 1 0 1 2 1 0 1 0 1 1 0 2 0 0 1 0 0 0 2 1 1 0 1 1 2 0 1 1 1 ? ? ? 27 Protachyoryctes tatroti 28 Tachyoryctes makooka 1 1&2 2 ? ? ? 1 ? 1 1 1 2 1 0 1 2 1 0 1 0 1 1 0 2 0 0 1 0 0 0 2 1 1 1 1 1 2 0 1 1 1 1 2 2 29 T. pliocaenicus 1 0&1 2 0 0 1 1 1 1 2 1 2 1 0 2 2&3 1 0 1 0 1 1 1 3 1 0 1 0 2 1 2 1 1 1 1 1 2 1 1 1 1 2 2 2 30 T.macrocephalus 2 2 2 0 0 1 1 1 1 2 1 3 1 0 2 3 1 0 1 1 1 1 1 3 2 0 1 1 2 1 2 1 1 1 1 1 2 1 1 1 1 2 2 2 31 Tachyoryctes splendens 2 0&1 2 0 0 1 1 1 1 2 1 3 1 0 2 3 1 0 1 1 1 1 1 3 2 0 1 1 2 1 2 1 1 1 1 1 2 1 1 1 1 2 2 2

32 Rhizomyides sivalensis 1 3 2 0 0 1 ? 1 1 0 1 0 0 0 0 0 1 0 1 1 1 1 0 2 0 0 1 0 0 0 2 1 1 1 1 1 2 0 1 1 1 1 ? 0 33 Miorh. punjabiensis 1 1 1 1 1 1 ? 1 1 0 0 1 1 0 0 1&2 1 0 1 1 1 0 0 2 0 0 1 0 0 0&1 1&2 0 1 1 1 1 2 0 1 1 1 1 2 0 34 1 2 2 0 0 1 ? 1 1 0 1 0 0 ? 0 2 1 0 1 1 1 1 0 2 0 0 1 1 ? ? ? ? ? ? ? ? 2 0 1 0 ? ? 0 ? 35 Rhizomyides carbonnelli 36 Rhizomyides playtomeus 1 3 2 0 0 1 ? 1 1 0 1 0 0 ? 1 0 ? 0 1 1 1 ? ? 2 0 ? 1 1 ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? 37 Tachyoryctes konjiti 1 1 2 ? ? ? ? 1 1 2 1 3 1 0 2 3 1 0 1 1 1 1 1 3 2 0 1 1 2 1 2 1 1 1 1 1 2 1 1 1 1 2 2 2 38 Kanisamys indicus 1 0 1 0 1 0 ? 0 1 0 0&1 0 0&1 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 1 0 1 0 0 0 1&2 0 1 0 0 1 1 0 39 40 41 42 43 44 45 46 Cladistics 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Page 67 of 68 Cladistics

1 2 3 4 5 6 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 31 32 33 34 35 36 37 38 39 40 41 42 43 44 7 1 0&1 2 0 1 0 1 1 1 0 0&1 0&2 1 0 0 2 1 0 1 0 1 0&1 0 2 0 0 1 0&1 0 0 1 0 1 0 0 0 2 0 1 0 0 1 1 0 8 Kanisamys nagrii 9 Kanisamys sivalensis 1 1 2 0 1 0 1 1 1 0 0&1 0&2 1 0 0 2 1 0 1 0 1 1 0 2 0 0 1 0&1 0 0 1&2 0 1 0 1 1 2 0 1 0 0 1 1 0 10 Kanisamys potwarensis 1 0&1 1 0 1 0 ? 1 1 0 1 1 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 1 0 0 0 1 1 2 0 1 1 0 1 1 0&1 11 Eicooryctes kaulialensis 2 1 2 1 0 1 ? 1 1 1 1 2 0 1 1 2&3 1 0 1 1 1 1 0 3 0 0 1 1 2 1 2 1 1 1 1 1 2 1 1 1 1 ? ? ? 12 Anepsirhizomys opdykei 2 4 2 2 1 1 ? ? 1 1 1 0 0 ? 2 0 1 0 1 0 ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? 13 Anepsirhizom. pinjoricus 2 3 2 0 ? 1 ? 1 1 ? 1 For 0 0 1 2 Peer 0 1 0 1 1 1 Review 1 1 0 2 0 1 ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? 14 Pronakalimys andrewsi 0 0 1 0 0 1 ? ? 0 0 0 0 0 0 0 2 0 0 0 0 0 0 0 2 0 0 0 0 0 0 1 1 1 0 0 0 1 0 1 0 0 0 1 2 15 0 1&2 1 0 0 1 ? 1 1 0 0 1 0 0 0 2 0 0 0 0 1 0 0 0&1 0 0 0 0 0 0 1 1 1 0 0 0 0&1 0 1 0 0 0 1 0 16 Nakalimys lavocati 17 Prokanisamys kowalskii 0 0 0 ? ? ? ? 0 0 0 0 1 0 0 0 1 0 0 0 0 0 0 0 1 0 0 0 0 0 0 1 1 1 0 0 0 1 0 0 0 0 0 1 0 18 Prokanisamys arifi 0 0 0 ? ? ? ? 0 0 0 0 0&1 0 0 0 1 0 0 0 0 0 0 0 1 0 0 0 0 0 0 1 1 0 0 0 0 0&1 0 0 0 0 1 1 0 19 Prokanisamys benjavuni 1 0 1 ? ? ? ? ? 0 0 0 1 0 0 0 1 0 0 0 0 0 0&1 0 1&3 0 0 0 0 0 1 0&1 1 0 0 0 0 1 0 0 0 0 1 0&1 0 20 Prokanisamys major 0 0 0 1 1 0 ? ? 0 0 0 1 0 0 0 1 0 0 0 0 1 0&1 0 0&1 0 0 0 0 0 0 1 1 0 0 0 0 1 0 0 0 0 1 1 0

21 Prokanisamys JebelZelten 0 0 0 ? ? ? ? ? ? ? ? ? ? ? 0 1 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0&1 1 0 0 0 0 ? ? ? ? ? ? ? ? 22 Rhizomyides mirzadi 1 1&2 2 1 1 ? ? ? 1 0 1 1 0 0 0 2 1 0 1 1 ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? 2 0 1 1 ? ? ? ? 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 Cladistics 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Cladistics Page 68 of 68

1 2 3 Acctran Deltran 4 Rhizomys sumatrensis 5 Rhizomys sinensis 21(1
0) 21(1
0) 6 Rhizomys pruinosus 7 Rhizomys (Brachyrh.) shansius 42(2


1) Rhizomys (Brachyrh.) shajius 16(0
2) 15(0
1) ; 16(0
2) 8 Cannomys badius 4(2
1) 2(3


0&1) ; 4(2
1) 9 Miorhizomys nagrii 4(2
1); 10(1


0) ; 14(0
1) ; 21(1
0); 29(1


0) ; 32(1
0); 4(2
1); 10(1


0) ; 15(1
0) ; 21(1
0); 29(1


0) ; 32(1
0); 38(1


0) 10 38(1


0) 11 Miorhizomys micrus 2(2
1); 4(0
1); 8(1
2); 20(1
0) ; 25(0
1) 2(2
1); 4(0
1); 8(1
2); 25(0
1) Miorhizomys blacki 15(1
0) 12 Miorhizomys pilgrimi 13 Miorhizomys harii 4(2
1); 14(1
0) ; 16(0
1); 17(1


0) 4(2
1); 15(0
1) ; 16(0
1); 17(1


0) 14 Miorhizomys tetracharax 14(1
0) ; 20(1
0); 23(0
1); 24(2
4); 25(1
2); 26(0
1); 15(0
1) ; 20(1
0); 23(0
1); 24(2
4); 25(1
2); 26(0
1); 15 27(1


0) ; 28(1
0); 41(1


0) 27(1


0) ; 28(1
0); 41(1


0) Miorhizomys choristos 15(1
2) ; 16(0
4); 23(0
1); 24(2
4); 25(1
2); 14(0
1); 15(0
2) ; 16(0
4); 23(0
1); 24(2
4); 25(1
2); 16 26(0
1); 31(2


0) ; 37(2
3); 42(1
0) 26(0
1); 31(2


0) ; 37(2
3); 42(1
0) 17 Protachyoryctes tatroti 34(1


0) 5(1
0); 34(1


0) 18 Tachyoryctes makooka 10(0
1) For Peer Review10(0
1) ; 43(1
2) ; 44(0
2) 19 T. pliocaenicus 1(2
1); 12(3


2) ; 20(1
0); 25 (2


1); 28(1
0) 1(2
1); 12(3


2) ; 20(1
0); 25(2


1) ; 28(1
0) T.macrocephalus 2(1
2) 2(1
2) 20 Tachyoryctes splendens 21 Rhizomyides sivalensis 28(1
0) 22 Rhizomyides punjabiensis 3(2


1) ; 11(1
0) 2(2
1) ; 3(2


1) ; 11(1
0); 22(1
0) ; 32(1
0) ; 43(1
2) 23 Rhizomyides carbonnelli 40(1


0) 28(0
1) ; 40(1


0) ; 43(1


0) 24 Rhizomyides playtomeus 15(0
1) 15(0
1); 28(0
1) Tachyoryctes konjiti 1(2
1) 1(2
1) 25 Kanisamys indicus 8(1


0) 8(1


0) 26 Kanisamys nagrii 27 Kanisamys sivalensis 2(0
1) 28 Kanisamys potwarensis 12(0
1); 33(1


0) ; 35(0
1); 36(0
1); 40(0
1) 12(0
1); 33(1


0) ; 35(0
1); 36(0
1); 40(0
1) Eicooryctes kaulialensis 4(0
1); 15(2


1) ; 23(1


0) ; 25(2


0) 4(0
1); 12(0
2) ; 15(2


1) ; 23(1


0) ; 25(2
0) 29 Anepsirhizomys opdykei 2(3
4); 20(1
0) 2(3
4); 20(1
0) 30 Anepsirhizom. pinjoricus 24(2
0) 31 Pronakalimys andrewsi 24(0
2); 44(0
2) 24(0
2); 44(0
2) 32 Nakalimys lavocati 12(0
1) 2(0


1&2) ; 12(0
1) 33 Prokanisamys kowalskii 33(0
1); 42(1
0) 33(0
1); 42(1
0) Prokanisamys arifi 34 Prokanisamys benjavuni 1(0
1); 3(0


1) ; 30(0
1) 1(0
1); 3(0


1) ; 30(0
1) 35 Prokanisamys major 21(0
1) 4(0
1) ; 21(0
1) 36 Prokanisamys sp. (Libya) 37 Rhizomyides mirzadi 13(1
0) 13(1
0) 38 40 2(0


4) ; 31(2


3) ; 37(2
3); 43(2


3) 2(3
4); 31(2


3) ; 37(2
3); 43(2


3) 41 1(1
2); 2(3


0) ; 12(0


4) ; 26(0
1) 1(1
2); 12(0


4) ; 26(0
1); 42(1
2) 39 42 6(1


2) ; 7(1


2) ; 18(0


1) ; 20(1
0) 6(1


2) ; 7(1


2) ; 10(1
2); 18(0


1) ; 20(1
0) 40 43 10(1
2) ; 16(0
4); 24 (0


4) 16(0
4); 24(2
4) 41 44 10(1
2); 12(2


3) ; 13(0
1); 14(1
0) 10(1
2); 12(0


3) ; 13(0
1); 14(1
0); 42(1
2) 42 45 2(3


1) ; 12(0
2) ; 16(0


3) ; 24(0


3) 2(3


1) ; 5(1
0); 16(0


3) ; 24(2


3) ; 29(1
2) 46 4(2


0) ; 5(1
0) 4(2


0) ; 8(2
1) 43 47 1(1
2); 8(2
1) ; 13(1
0); 29(1
2) 1(1
2); 13(1
0) 44 2(2
3); 15(1
2); 23(0
1); 24(2
0) ; 25(1
2); 42(1
2) ; 2(2
3); 14(0
1) ; 15(0
2) ;23(0
1);25(1
2);43(1
2); 48 45 43(1
2); 44(1


2) 44(1


2) 4(0


2) ; 8(1
2); 10(0
1); 14(0
1) ; 15(0
1) ; 16(2
0); 25(0
1); 4(0


2) ; 8(1
2); 10(0
1);16(2
0); 20(0
1); 25(0
1); 46 49 28(0
1); 29(0


1) ; 30(0
1); 38(0


1) ; 43(2


1) ; 44(0


1) 28(0
1); 29(0


1) ; 30(0
1); 38(0


1) ; 44(0


1) 47 50 5(1
0) ; 12(0
2); 15(0
1); 20(1
0) ; 44(0
2) 12(0
2); 15(0
1) 48 51 2(2
3);16(2
0) 2(2
3);16(2
0) 49 52 5(1
0); 13(1
0); 28(0
1) ; 43(2


0) 5(1
0); 13(1
0); 20(0
1) 50 53 2(2
1) ; 4(0
1); 12(0
1); 22(1
0) ; 32(1
0) 4(0
1); 12(0
1); 20(0
1) 2(1
2) ; 6(0


1) ; 20(0
1) ; 32(0


1) ; 34(0


1) ; 40(0


1) ; 41(0


1) ; 2(0


2) ;6(0


1) ; 31(1
2) ; 32(0


1) ; 34(0


1) ; 40(0


1) ; 54 51 43(1
2) 41(0


1) 52 55 31(1
2) ; 35(0
1); 36(0
1) 35(0
1); 36(0
1) 3(1


2) ; 13(0
1); 17(0


1) ; 19(0


1) ; 22(0


1) ; 24(0
2); 53 56 3(1


2) ; 13(0
1); 17(0


1); 19(0


1) ; 22(0


1) ; 24(0
2); 27(0


1) 54 27(0


1) 57 2(1


0) ; 16(2
0) 16(2
0) 55 58 1(0
1); 11(0


1) ; 32(1
0); 37(1


2) ; 42(0
1) 1(0
1); 11(0


1) ; 32(1
0); 37(1


2); 42(0
1) 56 59 2(0
1) ; 9(0


1) ; 21(0
1) 8(0
1); 9(0


1) ; 21(0
1) 57 60 8(0
1) ; 16(1


2) ; 33(0
1); 39(0


1) 16(0
2) ; 33(0
1); 39(0


1) 58 61 3(1


0) ; 4(0
1) ; 12(0
1); 24(0


1) ; 42(0
1) 3(1


0) ; 12(0
1); 16(0
1) ; 24(0


1) ; 42(0
1) 59 62 16(0
1) 60 Cladistics