Full text of DOI:10.1007/BF02875332

Vol. 44 No. 7
SCIENCE IN CHINA (Series D)
July 2001
Suckers (Fish, Catostomidae) from the Eocene of China
account for the family’s current disjunct distributions
CHANG Mee-mann (张弥曼)1, MIAO Desui (苗德岁)2, CHEN Yiyu (陈宜瑜)3,
ZHOU Jiajian (周家健)1 & CHEN Pingfu (陈平富)1
1.Institute of Vertebrate Paleontology and Paleoanthropology, Chinese Academy of Sciences, Beijing 100044, China;
2. Natural History Museum, University of Kansas, Lawrence, KS 66045-2454, USA;
3. Institute of Hydrobiology, Chinese Academy of Sciences, Wuhan 430027, China
Correspondence should be addressed to Chang Mee-mann (email: Zhangmm@public. bta.net.cn)
Received December 18, 2000
Abstract Unequivocal Eocene suckers from China are for the first time reported here. This dis-
covery demonstrates that catostomids of the Eocene Epoch (some 55—35 Ma ago) are scattered
widely on mainland Asia as well as western North America. The present day disjunct distribution
pattern of catostomids, with 68 extant species widespread in North America and the northern part
of Middle America and only two in the restricted areas of Asia, is the result of their post-Eocene
decline in Asia due to the competitive pressure from cyprinids, their Late Cenozoic radiation in
North America, and the vicariant and dispersal events triggered by the changed biogeographic
landscape. All of these prove to be a historical product of the geological, biological, and climatic
changes throughout the Cenozoic.
Keywords: China, catostomids, Eocene, biogeography.
Suckers (Catostomidae) are now one type of the most widespread freshwater fishes in North
America, with a southern extension into Mexico and Guatemala. Among the family’s 69 extant
species[1,2], only two occur elsewhere: one in northeastern corner of Siberia and the other in the
Yangtse River and Minjiang River (Fujian Province), China[3,4]. This disjunct distribution (fig. 1(a))
has long puzzled ichthyologists and biogeographers alike. Although Darlington’s[5] scenario of
—
eastern Asian origin of catostomids and their subsequent dispersal routes gained acceptance[6 8], it
is currently considered problematic because it lacked a robust phylogenetic analysis and an ade-
quate fossil record. Surprisingly, with an almost blank record of fossil suckers in east Asia but
—
quite a few from the then so-called “Oligocene” and “Miocene” in North America[9 11], Darling-
ton[5] postulated that “catostomids originated in eastern Asia, moved primarily from Asia to North
America, and (except for Myxocyprinus) have been replaced in Asia by cyprinids; that catostomids
have radiated secondarily in North America; and that one Catostomus has returned to the near
corner of Asia.” At the time of his writing, the Tertiary fossil record of catostomids in North
America comprised a genus †Amyzon with five species from what was then thought to be the Oli-
gocene or Miocene, but now known as the Eocene, of Nevada, Colorado, and British Columbia.
Since then, however, catostomids have been known from the Eocene and even Paleocene localities

578
SCIENCE IN CHINA (Series D)
Vol. 44
throughout western North America[12,13] (fig. 1(b)). They were all referred to the genus †Amyzon[15
—
18] (fig. 2(b)), though the monophyly of the genus[17] and the validity of some included species[19]
were questioned. Suckers belonging to modern genera did not appear until the middle or late
Miocene (approximately 15 Ma ago); once they did, they occurred on both sides of the Continen-
tal Divide of North America[13,20,21] (fig. 1(b)).
Fig. 1. Distribution of modern suckers ((a), hatched areas), modified from Nelson[14] and Tertiary suckers (b), showing
the Eocene catostomids spreading across mainland Asia and western North America, and Late Tertiary catostomids re-
stricted to North America. Black spots indicate early Tertiary localities, triangles indicate late Tertiary localities.
Prior to this report, fossil suckers were conspicuously absent from China except a few disar-
ticulated operculars and vertebrae assigned as Catostomus sp. from the middle Eocene of Inner
Mongolia, China[1,22]. In fact, the fossil record of suckers from Asia as a whole was sketchy.
Sytchevskaya[7] described a late Eocene or early Oligocene sucker, †Vasnetzovia, from the Far
East coastal area of Russia but did not mention its pharyngeal teeth. Other fossil suckers described
by the same author from the early-middle Eocene and early Oligocene of Zaysan Basin, East Ka-
zakhstan, were based on detached pharyngeal teeth alone. Among her findings, all but an ictiobine,

No. 7
SUCKERS FROM THE EOCENE OF CHINA
579
†Vasnetzovia, were dismissed by Smith[1]. Besides, their occurrences were restricted to northern
Asia[7].
1
Systematics
Here we describe a newly recognized catostomid from Xiawanpu, Xiangxiang, Hunan Prov-
ince, southern China based on abundant materials previously referred to Osteochilus, a Recent
cyprinid[23,24]
.
Order Cypriniformes
Family Catostomidae
Genus †Amyzon Cope, 1872
Species †Amyzon hunanensis (Cheng), 1962 (fig. 2(a))
Emended diagnosis. Eocene †Amyzon with a combination of the following characters: ter-
minal mouth; frontoparietal fontanelle present; dorsal fin elongate and emarginate with 19—20
principal rays; caudal fin shallowly emarginated with round-tip lobes and 16 branched rays; anal
fin with 10 principal rays; total number of vertebrae including the Weberian apparatus 34—35, 14
—15 being caudal; anterior dorsal profile comparatively rounded, body depth to standard length
ratio 0.35—0.43; head length to standard length ratio 0.27—0.29; ratio of caudal peduncle depth
to standard length 0.13—0.18.
Description and comparison. †Amyzon hunanensis is a comparatively small fish with
maximum body length of 200 mm, whereas the body length of †A. gosiutensis (regarded by
Bruner[19] as junior synonym of †A. aggregatum) from Green River, Wyoming[18] may exceed 300
mm. The anterior dorsal profile seems more rounded than in †A. aggregatum. It has an edentulous,
terminal mouth, as in all other †Amyzon (M. V. H. Wilson, pers. comm.). The upper jaw margin is
bordered by both the premaxilla and maxilla, as in other catostomids. The smooth and somewhat
concave medial margin of the frontal indicates presence of an open frontoparietal fontanelle. The
dorsal fin is elongate and slightly emarginate behind the first, longest fin ray, with 19—20 prin-
cipal rays, less than that in †A. aggregatum (21—27) from Horsefly, British Columbia[16] and †A.
gosiutensis (22—24) from Green River, Wyoming[18]. The caudal fin is shallowly emarginate. Its
upper and lower lobes have rounded tips. There are 16 branched caudal rays. The caudal skeleton
is of the cyprinid and catostomid type. The compound ural centrum (possibly a fusion of the first
preural and the two ural centra) has an urostyle and a short neural spine, with 3—5 upper hypu-
rals and one epural. The anal fin has 10 principal rays and thus, in most cases, has more rays than
that in †A. aggregatum (8—10) and †A. gosiutensis (7 —8). The total number of vertebrae in-
cluding the Weberian apparatus is 34—36, of which 14—15 are caudal. The number is compara-
ble to or less than that in †A. aggregatum (36—41) but comparable with that in †A. gosiutensis
(34—35). The maximum depth of the body is at the origin of the dorsal fin or just in front of it.

580
SCIENCE IN CHINA (Series D)
Vol. 44
The depth to standard length ratio is 0.35—0.43, close to that in †A. aggregatum (around 0.41)
and †A. gosiutensis (0.36—0.44). The head length to standard length ratio is 0.27—0.29 and is
slightly smaller than that in both †A. aggregatum (0.30) and †A. gosiutensis (0.29—0.33). The
depth of peduncle to standard length ratio is 0.13—0.18, and thus falls within the range of †A.
aggregatum (0.094—0.189) and similar to that of †A. gosiutensis (0.14—0.17). Among the
aforementioned characters the component of the upper jaw margin and the number of the principal
branched caudal fin rays are two more easily recognized differences between catostomids and cy-
prinids in addition to the pharyngeal teeth. In cyprinids, the upper jaw margin is bordered by the
premaxilla alone and there are 17 principal branched caudal fin rays.
Fig. 2. Two species of †Amyzon: †A. aggregatum Wilson (a) from Horsefly Mine L88, BC, Canada (Royal Ontario Mu-
seum 11051[16]), a complete fish; and †A. hunanensis (Cheng) (b) from Xiawanpu, Xiangxiang, Hunan, China (V12571.1), a
nearly complete fish with tip of snout lacking.
Remarks. We recently discovered material from Huadian, Jilin Province, northeastern
China with an unusually well preserved pharyngeal bone carrying one row of numerous and
closely set teeth (fig. 3). This diagnostic feature unmistakably reveals its identity as a sucker and
thus shows, unambiguously for the first time, the Asiatic occurrence of Eocene suckers. We also
found a few tiny, compressed pharyngeal teeth on one specimen among hundreds from Songzi,
Hubei Province, after nearly 20 years of searching and exquisite preparation. The fish, at the same

No. 7
SUCKERS FROM THE EOCENE OF CHINA
581
time, has many other characters similar to †Amyzon. Having compared these with the previously
—
described Chinese fossils that were lumped under a living cyprinid genus Osteochilus[22 25], we
came to realize that all of them in fact belong to Catostomidae[26,27]. Particularly, †Osteochilus
hunanensis Cheng, 1962, from the middle Eocene of Hunan is strikingly similar to †Amyzon ag-
gregatum (with †A. gosiutensis as its junior synonym) (compare fig. 2(a) with fig. 2(b)). The few
differences between them include a more rounded anterior dorsal profile and a few minor meristic
ones as discussed above. However, the specimen from Huadian differs from other Chinese ca-
tostomids in its unusually long anal fin. Therefore, while we tentatively refer †Osteochilus hu-
nanensis to †Amyzon here, we prefer to describe the Huadian specimen in a more detailed study.
Fig. 3. Catostomidae indet. from Huanan, northeast China, pharyngeal bone with teeth (V12572.2).
2
Discussion
In retrospect, at least two factors could be allotted to the previous misidentifications of the
Chinese fossil suckers as cyprinids. First, not only are catostomids morphologically similar to
cyprinids but also the former share a part of the latter’s ecological niches. In fact, the first extant
sucker species was initially also described as a cyprinid, Cyprinus catostomus[28]. Cyprinids are
characterized by fewer but larger pharyngeal teeth arranged in one to three rows on the pharyngeal
bones[29], in contrast with catostomids’ one row of numerous, laterally compressed, and compara-
tively small pharyngeal teeth[30]. On imperfect fossil specimens and in the absence of pharyngeal
bones and teeth, it is not always easy to distinguish them. Second, cyprinids are one of the most
dominant groups of extant freshwater fishes in Asia and have a reasonably good Asiatic fossil re-
cord from the middle Eocene onward. Catostomids, on the other hand, were then known to be a
minor element in the Asian ichthyofaunas, both extant and extinct. Again in the absence of pha-

582
SCIENCE IN CHINA (Series D)
Vol. 44
ryngeal bone and teeth and the specimens with well-preserved skeleton, it was only natural for the
earlier authors[23,24] to refer their specimens to cyprinids instead of catostomids.
With recent discoveries and amended identifications of the previously described taxa, the
Eocene and early Oligocene catostomids are now known to be widely spread across mainland
Asia, from Mongolia in the north to Guangdong Province of China in the south and from
eastern Kazakhstan in the west to the Far East region of Russia and Wutu, Shandong Province,
China in the east (fig. 1(b)). Suckers are often found the most abundant fishes at the localities
where they occur, on both sides of the Pacific. Therefore, it is clear that during the Eocene
and early Oligocene, catostomids had a wide, transpacific distribution over Asia and western
North America.
Except two surviving species, Myxocyprinus asiaticus and Catostomus catostomus, suckers
have disappeared from Asia since early Oligocene. After the withdrawal of the Mid-Continental
Seaway in the Paleocene, North America integrated into an extensive land but did not immediately
allow suckers to exploit the various habitats on the eastern side of the Continental Divide, owing
to the structural disturbances created by the Cretaceous-Eocene Laramide Orogeny and its ripple
effects during the mid-Tertiary[31,32]. And due to the high eustatic sea levels much of the Atlantic
and Gulf Coastal Plains were inundated during most of the Eocene[33]. Thus, it was not until the
Oligocene did suckers spread to eastern North America. From the Miocene on, suckers of living
species of Cycleptus, Ictiobus, Carpiodes, and many genera from the subfamily Catostominae
started to conquer freshwater systems of the entire territory of North America. Catostominae, with
a total of 59 species, comprises more than 80% of the species number of the whole family. Ca-
tostominae is restricted to North America and the northern part of Middle America except for one
species, Catostomus catostomus, which also occurs in a very restricted part of east Siberia, pre-
sumably representing a postglacial dispersal of the species[1,13,15,31]
.
A biotic reorganization took place approximately at the Eocene/Oligocene boundary due to
the global climatic change from warm and humid to cool and arid[34]. Conceivably, with this gen-
eral reorganization ended the predominance of catostomids in the freshwater fish fauna of Asia, as
did many other groups in the biota. Then why do suckers still continue to exist, expand, and even
in some way, flourish in North America while most of their kin have vanished from Asia? The
main factor contributing to the catostomids’ decline in Asia is most probably the competitive
pressure from cyprinids[5]. Both of them are of the few primary freshwater fish groups and are
occupying similar ecological niches. When cyprinids first appeared in Asia in the middle Eo-
cene[7,25,35], catostomids still dominated the scene. The Oligocene record of fossil fishes and other
terrestrial animals are comparatively rare in China as well as in other parts of the world. The fish
remains of this epoch from the known localities mainly consist of detached pharyngeal teeth of
cyprinids, usually found in large quantity and along with the remains of small mammals. The fos-
sil-bearing deposits are relatively coarse. No remains of suckers have ever been found from any
Oligocene or later deposits of Asia. By the early Miocene cyprinids began to constitute the major-

No. 7
SUCKERS FROM THE EOCENE OF CHINA
583
ity of the freshwater fossil fish record from this area. Many Miocene and Pliocene fossil fishes are
—
from lacustrine deposits and thus well preserved[35 39]. Cyprinids are dominant in the Recent
freshwater ichthyofauna of Asia. Most of them are known to have a temperature preference for
spawning, possibly adapted for the Monsoon climate. The uplift of the Himalayas and the Qing-
hai-Tibet Plateau beginning at the Eocene/Oligocene boundary created such a climate in south-
eastern Asia and caused the large eastward flowing rivers to develop; this climate and topographi-
cal changes may have favored cyprinids, probably at the expense of catostomids.
In North America, cyprinids appeared in the middle to late Oligocene, later than they did in
Asia. They seem to have reached North America from Asia across Beringia. Only by the late
Miocene were they widely distributed in western North America[40]. Of the 2010 living cyprinid
species (belonging to 210 genera) of the world, some 1270 are known from Eurasia and only 270
from North America[41]. Nearly seventy percent of North American living cyprinid species dwell
in the eastern region, a result of their recent explosive radiation[6]. It appears that the cyprinids’
later arrival and blossoming in North America may account for catostomids’ continued success
there.
Suckers are only one of several fish groups which had a transpacific distribution during the
Early Tertiary[7,35,42]. The transpacific distribution pattern started in the Late Cretaceous when a
series of continental collisions in Beringia provided a broad passage for the dispersal of terrestrial
and freshwater biota[32]. After the faunal reorganization at the Eocene/Oligocene boundary, many
groups became extinct in Asia, e.g. Amiidae, Hiodontidae, and most of Catostomidae, but contin-
ued to exist in North America. Some forms belonging to these groups survive there even up to the
present day.
In addition, this distribution pattern of freshwater fishes also mimics that of other vertebrate
groups, especially mammals. Beard[43] even went so far as to call it “East of Eden” model and
deem North America functioned primarily as a biogeographical cul-de-sac of Asia since at least
Late Cretaceous. All in all, fossil record and historical geology now appear to favor Darlington’s[5]
dispersal hypothesis on historical biogeography of catostomids. However, in terms of the dispersal
model, one could equally argue that catostomids might have originated in North America and their
two crossings of Beringia were in the same direction: both from North America to Asia. Just as the
available fossil evidence seems to corroborate Darlington’s scenario, so the same evidence seems
to be congruous with this alternative one.
Alternatively, this contradiction could be reconciled by adopting a vicariance model in light
of Smith’s[1] cladistic analysis of catostomids. Smith’s[1] cladistic analysis largely resolved the
intrarelationships of the family Catostomidae. In his phylogenetic scheme (fig. 4), the Early Terti-
ary †Amyzon is grouped with Recent ictiobines, i.e., Ictiobus + Carpiodes, to form the most primi-
tive of the three subfamilies of the Catostomidae——Ictiobinae. The Chinese rouge fish Myxo-
cyprinus is the sister group of the North American Cycleptus, and they together form the second
subfamily Cycleptinae. The rest and the majority of species of suckers fall into the most advanced

584
SCIENCE IN CHINA (Series D)
Vol. 44
subfamily Catostominae. As we stated earlier, the
Eocene primitive ictiobine suckers of †Amyzon
type had a wide distribution across Asia and west-
ern North America through the Bering Land
Bridge. This wide longitudinal distribution may
be due to global warming and a declining
pole-to-equator temperature gradient during the
Eocene[34]. In the subfamily Cycleptinae, Cyclep-
tus may be represented by the Oligocene fossils
from Montana, but the purported Asian Cycleptus
based on isolated teeth was considered misidenti-
Fig. 4. Phylogeny of the Catostomidae, simplified
from Smith[1]
.
fied[1]. No fossil Myxocyprinus has ever been reported anywhere. The only two living species of
the subfamily have a disjunct distribution, with Myxocyprinus asiaticus in the Yangtse and Minji-
ang rivers and Cycleptus elongatus mainly in the Mississippi valley. Consequently, the split be-
tween them might have been caused by the inundation of the Bering Land Bridge, thus the separa-
tion of Asia from North America. This, therefore, would represent a vicariant event. However,
Catostomus’s occurrence in Siberia after the desiccation of the Bering Strait during Pleistocene
and catostomids’ spread to the eastern North America after its reunion with the western counter-
part appear to be dispersal events. With this combination of vicariance and dispersal as influenced
by the geological events and climatic changes it is possible to explain the current disjunct distribu-
tions of catostomids without invoking any scenario based on their presumed “center of origin.”
Acknowledgments The authors are very grateful to M. V. H. Wilson for careful reading the manuscript, sharing with the
authors his knowledge on North American †Amyzon, giving the authors useful suggestions and permission to use the photo of
†Amyzon aggregatum. Thanks are also due to G. R. Smith, E. O. Wiley, W. Dimmick, and C. R. Robins for helping the authors
track down some pertinent references, F. Cross for providing Recent catostomid skeletons representing all the three subfamilies
from North America, G. Arratia, C. G. Zhang and X. B. Yu for reviewing the manuscript, and J. Zhang for technical assistance.
This work is supported by the National Natural Science Foundation of China (Grant No. 48932010) and the Special Funds for
Major State Basic Research Projects of China (Grant No. G2000077708).
References
1. Smith, G. R., Phylogeny and biogeography of the Catostomidae, freshwater fishes of North America and Asia, in Sys-
tematics, Historical Ecology, and North American Freshwater Fishes(ed. Mayden, R. L.), Stanford: Stanford Univ. Press,
1992, 778—826.
2. Burr, B. M., Mayden, R. L., A new species of Cycleptus (Cypriniformes: Catostomidae) from Gulf slope Drainages of
Alabama, Mississippi, and Louisiana, with a review of the distribution, biology, and conservation status of the genus, Bull.
Alabama Mus. Nat. Hist., 1999, 20: 19—57.
3. Laboratory of Hydrobiology, Hubei, Fishes of Changjiang(in Chinese), Beijing: Science Press, 1976, 1—14, 150—152.
4. Lian, Z. S., Cypriniformes: Catostomidae, in Fishes of Fujian Province (in Chinese), Part I. Fuzhou: Fujian Science and
Technology Press, 1984, 226—228.
5. Darlington, P. J. Jr., Zoogeography: the Geographical Distribution of Animals, New York: Wiley, 1957, 33.
6. Gilbert, C. R., Composition and derivation of the North American freshwater fauna, Florida Science, 1976, 39: 104—111.
7. Sytchevskaya, E. K., Paleogene freshwater fish fauna of the USSR and Mongolia, Joint Soviet-Mongolian Paleontol. Exp.

No. 7
SUCKERS FROM THE EOCENE OF CHINA
585
Trans (in Russian with English summary), 1986, 29: 1—157.
8. Briggs, J. C., Biogeography and Plate Tectonics, New York: Elsevier Science Publishers, 1987.
9. Cope, E. D., On the Tertiary coal and fossils of Osino, Nevada, Proc. Amer. Philos. Soc., 1872, 12: 478—481.
10. Cope, E. D., The vertebrata of the Tertiary formations of the West, Book 1, in Rept. U. S. Geol. Suev. Terr. (ed. Hayden, F.
V.), 1884, 1009.
11. Cope, E. D., Fossil fishes from British Columbia, Proc. Acad. Nat. Sci. Phila., 1893, 45: 401—402.
12. Wilson, M. V. H., Oldest known Esox (Pisces: Esocidae), part of a new Paleocene teleost fauna from western Canada,
Can. J. Earth Sci., 1980, 17(3): 307—312.
13. Cavender, T. M., Review of the fossil history of North American freshwater fishes, in The Zoogeography of North Ameri-
can Freshwater Fishes (eds., Hocutt, C. H., Wiley, E. O.), New York: Wiley, 1986, 699—724.
14. Nelson, J. S., Fishes of the World, 2nd ed, New York: Wiley, 1984, 127—130, 409.
15. Cavender, T. M., Freshwater fish remains from the Clarno Formation, Ochoco Mountains of North-central Oregon, The
Ore. Bin., 1968, 30: 125—141.
16. Wilson, M. V. H., Middle Eocene freshwater fishes from British Columbia, Life Sci. Contr. Roy. Ontario. Mus., 1977, 113:
1—61.
17. Wilson, M. V. H., Year classes and sexual dimorphism in the Eocene catostomid fish †Amyzon aggregatum, J. Vertebr. Pa-
leontol., 1984, 3(3): 137—142.
18. Grande, L., Eastman, J. T., Cavender, T. M., †Amyzon gosiutensis, a new catostomid fish from the Green River Formation,
Copeia, 1982, 1982: 523—532.
19. Bruner, J. C., Comments on the genus †Amyzon (Family Catostomidae), J. Paleont., 1991, 65: 678—686.
20. Cross, F. B., Mayden, R. L., Stewart, J. D., Fishes in the western Mississippi drainage, in The Zoogeography of North
American Freshwater Fishes (eds., Hocutt, C. H., Wiley, E. O,). New York: Wiley, 1986, 363—412.
21. Miller, R. R., Smith, G. R., Origin and geography of the fishes of Central Mexico, in The Zoogeography of North Ameri-
can Freshwater Fishes, (eds., Hocutt, C. H., Wiley, E. O.,) New York: Wiley, 1986, 487—518.
22. Hussakof, L., The fossil collected by the Central Asiatic expedition, Amer. Mus. Novit., 1932, 553: 1—19.
23. Tang, X., A new fossil fish from Linli, Hunan, Vert. PalAsiat. (in Chinese), 1959, 1: 211—213.
24. Cheng, C. C., Fossil fishes from the Early Tertiary of Hsiang-Hsiang, Hunan, with discussion of age of the Hsiawanpu
Formation, Vert. PalAsiat (in Chinese with English summary), 1962, 6: 333—343.
25. Wang, J. K., Li, G. F., Wang, J. S., The Early Tertiary fossil fishes from Sanshui and its adjacent basin, Guadong, Palaeon-
tologia Sinica (in Chinese with English summary), whole number 160, New Ser. C, 1981, 22: 1—90.
26. Chang, M. M., Zhou, J. J., A brief survey of the Chinese Eocene ichthyofauna, Kaupia, Darmstadter Beitrage zur Natur-
geschichte, 1993, Heft 2: 157—162.
27. Chang, M. M., Chen, Y. Y., Late Mesozoic and Tertiary ichthyofaunas from China and some puzzling patterns of distribu-
tion, Vert. PalAsiat, 2000, 38: 161—175.
28. Forster, J. R., An account of some curious fishes sent from Hudson Bay, Philosophical Transactions of the Royal Society,
London, 1773, 63: 149—160.
29. Chu, Y. T., Comparative studies on the scales and on the pharyngeals and their teeth in Chinese cyprinids, with particular
reference to taxonomy and evolution, Biol. Bull. St. John’s Univ., 1935, 2: 1—290.
30. Eastman, J. T., The pharyngeal bones and teeth of catostomid fishes, Amer. Midl. Nat., 1977, 97: 68—88.
31. Smith, G. R., Late Cenozoic freshwater fishes of North America, Ann. Rev. Ecol. Syst., 1981, 12: 163—193.
32. Briggs, J. C., Introduction to the zoogeography of North American fishes, in The Zoogeography of North American
Freshwater Fishes (eds., Hocutt, C. H., Wiley, E. O), New York: Wiley, 1986, 1—16.
33. Prothero, D. R., The chronological, climatic, and paleogeographic background to North America mammalian evolution, in
Evolution of Tertiary Mammals of North America (eds., Janis, C., Scott, K. M., Jacobs, L. L), Vol. I: Terrestrial Carnivores,
Ungulates, and Ungulatelike Mammals, Cambridge: Cambridge University Press, 1998, 9—36.
34. Janis, C. M., Tertiary mammal evolution in the context of changing climates, vegetation, and tectonic events, Ann. Rev.

586
SCIENCE IN CHINA (Series D)
Vol. 44
Ecol. Syst., 1993, 24: 467—500.
35. Liu, H. T., Su, D. C., Pliocene fishes fishes from Yüshe Basin, Shansi, Vert. PalAsiat. (in Chinese with English summary),
1962, 6: 1—25.
36. Zhang, M. M., Zhou, J. J., Qin, D. R., Tertiary fish fauna from coastal region of Bohai Sea, Mem. Inst. Vert. Paleont. Pa-
leoanth. Acad. Sin. (in Chinese with English summary), 1985, 17: 1—60.
37. Zhou, J. J., The Cyprinidae fossils from Middle Miocene of Shanwang basin, Vert. PalAsiat (in Chinese with English
summary), 1990, 28: 95—127.
38. Sytchevskaya, E. K., Neogene freshwater fish fauna of Mongolia, Joint Soviet-Mongolian, Paleontol. Exp. Trans. (in Rus-
sian with English summary), 1980, 39: 1—140.
39. Chang, M. M., Chen, Y. Y., Tong, H. W., A new Miocene xenocyprinine (Cyprinidae) from Heilongjiang Province, north-
east China and succession of late Cenozoic fish faunas of east Asia, Vert. PalAsiat (in Chinese with English sum-
mary) ,1996, 34: 165—183.
40. Cavender, T. M., The fossil record of the Cyprinidae, in Cyprinid Fishes Systematics, Biology and Exploitation (eds.,
Winfield, I. J., Nelson, J. S,), Great Britain: St. Edmundsbury Press, 1991, 34—54.
41. Nelson, J. S., Fishes of the World, 3rd ed, New York: Wiley, 1994.
42. Grande, L., Repeating patterns in nature, predictability, and “impact” in science, in Interpreting the Hierarchy of Nature:
from Systematic Patterns to Evolutionary Process Theories (eds., Grande, L., Rieppel, O.), New York: Academic Press,
1994, 61—84.
43. Beard, K. C., East of Eden: Asia as an important center of taxonomic origination in mammalian evolution, Dawn of the
Age of Mammals in Asia, Bull. Carnegie Mus. Nat. Hist. (eds., Beard, K. C., Dawson, M. R.), 1998, 34: 5—39.