AMERICAN MUSEUM Novitates PUBLISHED by the AMERICAN MUSEUM of NATURAL HISTORY CENTRAL PARK WEST at 79TH STREET, NEW YORK, N.Y

AMERICAN MUSEUM Novitates PUBLISHED BY THE AMERICAN MUSEUM OF NATURAL HISTORY CENTRAL PARK WEST AT 79TH STREET, NEW YORK, N.Y. 10024 Number 2849, pp. 1-17, figs. 1-5, tables 1-4 June 27, 1986

Two New Catostomid Fishes (Cypriniformes) from the Northern Sierra Madre Occidental of Mexico

DARRELL J. SIEBERT1 AND W. L. MINCKLEY2

ABSTRACT Two new species of Catostomus (sensu stricto) new taxa exhibit low pharyngeal tooth counts, pos- are described from northwest Mexico. One inhab- sibly indicating a close relationship between them. its northeastern headwater streams in the Rio Ya- They are hypothesized to be members of an old, qui drainage and the other is widely distributed in regional ichthyofauna that stretches from central the southern and southeastern parts of the Rio Mexico to northwest United States. Yaqui and in the upper Rio Mayo basin. These

INTRODUCTION Catostomid fishes ofMexico remain poor- species of Catostomus not obviously related ly known, especially those distributed in the to other species in the region are described mountain vastness of the northwestern por- here. Their presence in streams ofthe north- tion ofthat country. Six described species are ern Sierra Madre Occidental has been known currently recognized from the region. Most for some time (Miller, 1959, 1976), but only are wide ranging, and some, on further study, recently have they been taken in numbers or may prove to constitute more than a single of specimen quality adequate to define their taxon (see, e.g., Ferris et al., 1982). Two new characteristics, ecology, and distribution.

' Graduate Student, Department of Ichthyology, American Museum of Natural History; Department of Biology, City College, 138th Street and Convent Avenue, New York, New York 10031. 2 Professor ofZoology, U.S. Fish and Wildlife Service, Dexter National Fish Hatchery, Dexter, New Mexico 88230; Department of Zoology, Arizona State University, Tempe, Arizona 85287.

MATERIALS AND METHODS was instrumental in their success. We thank Specimens were collected during ichthyo- Dean A. Hendrickson, Robert Rush Miller, logical surveys of the Rio Yaqui basin (Hen- and C. L. Smith, for reading and commenting drickson et al. [1981]) and are deposited at on the manuscript. the American Museum of Natural History (AMNH), Arizona State University (ASU), SYSTEMATIC DESCRIPTIONS and the University of Michigan Museum of Catostomus leopoldi, new species Zoology (UMMZ). Comparative materials are Bavispe sucker also housed at those institutions. Work in Figures IA, B, 2A, 3A, C, E; tables 1-3 Mexico was under auspices of the Endan- gered Species Office, U.S. Fish and Wildlife Catostomus sp., Miller, 1959, p. 215 ("headwaters Service (USFWS), Albuquerque, New Mex- of both the Yaqui and Casas Grandes"); Miller ico. Collections were authorized by Permit in Koehn, 1969, p. 28 ["undescribed species ... No. 3618, issued by Departamento de Pesca, subgenus Catostomus (of Catostomus)," possible hybridization with Pantosteus plebeius (Baird Mexico, D.F., Mexico. and Girard) in Rio Casas Grandes basin]. Hen- Counts and body measurements follow drickson et al. [1981], pp. 76-77 (distribution, Hubbs and Lagler (1970), with the following habitat notes). Minckley and Brown, 1982, p. exceptions and additions: (1) interpectoral 223 (listed, Guzman basin and Rio Yaqui drain- width, distance between pectoral fin inser- age). Minckley et al., 1986 (listed, "Catostomus tions; (2) interpelvic width, distance between sp. [Yaqui-Guzman]; from the uppermost rios pelvic fin insertions; (3) pectoral fin to pelvic Bavispe-Casas Grandes"). fin length, distance between fin insertions; (4) lip length, distance from anteriormost point DIAGNOSIS: A species ofCatostomus (sensu ofupper to posteriormost extent oflower lip; stricto) with moderately small scales, usually (5) lip width, widest measurement across lips; 78 to 88 in the lateral line, and 9 or 10 dorsal (6) upper lip length, distance at midpoint from fin rays. Lips large and elliptical. Lower lip outer to inner edge; and (7) lower lip length, completely cleft, with lobes widely separated distance from anterior to posterior edge. at mandibular symphysis. Lips abundantly Counts of pharyngeal teeth include enumer- papillose, five or more rows ofpapillae cross- ation of gaps where teeth were obviously ing midline of upper and seven or more of missing. papillae rows crossing longitudinal line across Osteological line drawings are from camera longest part of lower lobes. Pharyngeal teeth lucida tracings of individual bones. Mea- more than 32, but fewer than 40 per arch. surements for calculation of of TYPES: Holotype: UMMZ 213391, a gravid proportions female, 151.0 mm standard length (SL), from tripus parts are also from such drawings; graph Rio Gavilan (locally Arroyo Moctezuma), paper was overlaid so that a line designated Chihuahua, Mexico (latitude 2951'N, lon- as the y-axis passed through the anterior- and gitude 108°24'W, 2180 m), June 21, 1978, posteriormost points and this distance was D. A. Hendrickson, J. E. Johnson, R. R. and measured to the nearest millimeter. Width F. H. Miller, W. L. and P. H. Minckley, and was obtained by measurement of a perpen- D. J. Siebert. Paratopotypes: UMMZ 213392, dicular to length that passed through lines 12 females and 8 males, 80.0-137.5 mm SL, tangent to the widest parts of the bone that ASU 10559, 42 specimens, 83.0-29.0 mm were parallel to the y-axis. Tripus arm lengths SL, and AMNH 50708SD, dry skeleton. Ad- were measured as a distance between ante- ditional paratypes: ASU 10560, mixed lot of rior- and posteriormost points ofan arm and 10 specimens, 22.0-137.0 mm SL, and a line perpendicular to a y-axis that passed AMNH 51302SD, dry skeleton, from Rio through the apex of the articulating process. Gavilan (locally Arroyo Moctezuma) between Rancho Puente de Agua and El Co- ACKNOWLEDGMENTS lorado, Chihuahua, Mexico (latitude 29°53'N, James E. Johnson, USFWS, arranged for longitude 108°24'W, 2130 m), May 23,1978; research funding. Dean A. Hendrickson, AMNH 50709, mixed lot of 44 specimens, ASU, was in charge of field operations, and 42.0-139.0 mm SL, from Rio Negro, ca. 1.0 1986 SIEBERT AND MINCKLEY: CATOSTOMID FISHES 3

Fig. 1. Photographs of female holotype (A) and male paratype (B) of Catostomus leopoldi, new species. Note oblique orientation of mouth. km upstream from bridge on road from Chu- overall shortness, and typically has a fat or huichupa to Molino de Aserrar Huaynopa, puffy aspect when compared to other Catos- Chihuahua, Mexico (latitude 29°33'N, lon- tomus, especially in the suborbital (cheek) re- gitude 108°25'W, 1990 m), July 23, 1978; gion. ASU 10561, mixed lot of47 specimens, 43.0- The mouth is inferior and slightly oblique. 164.3 mm SL, and AMNH 51301SD, dry The highly papillose lips are large (fig. 2A), skeleton, from Arroyo de la Nortefna, ca. 3 their respective length and width 38 and 48 km NW El Norte, Chihuahua, Mexico (lati- percent of head length. The lower lobes are tude 29°39'N, longitude 108°24'W, 2100 m), divided by a median cleft, commonly slitlike July 22-23, 1978. in Catostomus, but broader in C. leopoldi, DESCRIPTION: Proportional measurements widely separating the lobes from one another of Catostomus leopoldi are shown in table 1 at their points of attachment on either side and its body form and color pattern in alcohol of the mandibular symphysis. are depicted in figure IA, B. The body is well The dorsal fin origin is about midway along rounded, giving the species a plump appear- SL, rarely slightly before the midpoint. Sex- ance, especially about the head. The head is ual dimorphism is pronounced (fig. IA, B), deep and broad, giving an impression of with tuberculate males having longer and eoN-- - t%Ifin N r sf 0 V r-0s - t- %0 - - 0

4 o4 r6 14 (6 r; cl co e6 . l C; 14 i c; cl wi l e£on5-io

- - ¢~~~~~~Cpo-v W-Noe-4 tC-vNC oo- C4, 06Cv o oe o6 o o W; o; t vo tfX

_ o _ t} £ UN t-- tt to0 oot0 %o ON0 - - N O e- - I O N ON % es - ei - -4 - - --7-- - - -_ _N - C _ _`N A ohS A o, C A | T 0 2 s 1 $ 4 4 A 42 @ %0 0 4 W - 0 N O' ON t (~~~O ~ ~~~~~eo O t- -, - N N toNO tn ON ON 4 - tl t- A010 -i - r- M - ~~~~4 10 V 0 V W z5 o~~~N o~o o x oW N N o _ o 0000 0 0 0 0 0 0 _ O q N 0 W - o e_ o 0 o 0 0 O o e v - %0o X 0% *ION as WI 0-*et t-ooen en N e-l'-O en t 0o t - o as - 1 II 1 1, , sr-,__^_N______N _0~ _-I WIWI>ON NO . l el - - - V 0 enO Cq o VON - N 0 t- Ch

4 IA (4- C - - - -4 -- -4-4 -u- C - olci6cl0"e;0je;Ci-_in el&~ ; en&Z6t: 6C6o- Co-oel; W oo 4) ca v c>l tt O Oa eq 00 - - 00 00 oo en-t} -V ON N eq. WI 0%O tn en 0 s

> ~.H~~~~~~~o-~~~~e von _ en _ C 0 C)_ 0i - en_enC_ en t-_ en r WI ON _ < a £ £oo e- c'q 00 C'l c1 e >-m ^-00ot

- 9 e0% e-eO0Oe\0N % IAN00 el 0 0 NON c enO - 00No

*.m eq -4 4I-$- 4 -

~d ,O0 O 0- >00 0ne % e 00re tn 4 n T h e 0 % 00 S en

-_o- . e.r. -4 -e 04 en in 04 -44 -4 en -0C -4 C - - .4 4 -4 4 -4 - 2

o-2 ~ in-- - -£ ay o - -o o-4 u~~ ~ - 11 £t- t- W - o" Nf M ro~~~ 1_ N~00 )IA C} ' ~00 ~ef N- 00 . ~~l O }e

4 tn e( . Z n t- 0 n WI 00 0 _- q el CNen N en£_O 10 Os N 0 Os tn 0 Q0 o ooQ e- oO0 - '0o'z o0o 0% 0% C 004 00 _-^ 000 0S01114MAI2 S

0%~~~~~~~~~~W @ ~ IeO£O0 % 0 r0O 00000-0%0 oN 0 0 %O 00 e 4)I~~~~~~~%4)4)_mu~ ~ (U(__n___0o_ N~~~~~~~~~) -: *e _,~ ;~ I00JJ~-~~o0 ~ - S; Q~~~0Q

4 em £0 'o 0% et) 0%Ne o %0 %o £t df CUT'o %o r}N C}vo 0ON C .IN N 0% - c

Is C- cn C7. 0 ONN 00en 1. t' ON C) 00 14 00 M r-m t - In o

0-t 00 - - oN -e4 --en C4 - - C--a --I e- - -4__ _~ _~ '~nt --r t 0e 0-- e 000 e.s tr0o c e 0000 No e _ ____

- -eN --oo e -co oc%I--- t 0%- >Q - o % ooe~I% 4 A r A 0 A oo O N - 0 0% 0000A e 00 CO lt tn ON un (Dt0N n % (- o C) (C C W) r- t%N 'o4n en 0-oN eeoot- 10 _11--__ - _ en o-- N---e~N~~e--N--- _ ---%1 --oo en 00 N N - % N n el e ot WI N en 0% ON 0% 'IO a0 08t C 00

C: 00%0 t08 00en oqeo unt - \0 eo -i "T N N-m as r-tbi sFoo e _ <, _~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~tN _ o 00 0t0N 0 ' 0 e 0 o N On en 00 _r0_ -00 11tn 7t0enWI en Nt en el" t- 00 -rq- C)O-4nt 0 t- i t- CD 0 -t- nen _ s14 c00000C* nt04004-00 ^ 00 0 .0M -_ _ -CM0 _ N0% 00_ _ t0 0- t ON ra, 'fICM V0 ~~~-CM--e~~~~~~~~~~~~~~~~~~CM ~t Mt- CM-CM--N - en eON rf0NCM m Nouc00 -0 C v %00N N 000aN'1 c00

-CM--r~~~~~~~C M--C M---~~~~~t C --CMi - X~~~~~~~~~~W 0e t- el000 tn r-t, 0 e O 08n ON V F %o 0 00 as

1 O4.~~~~~~~~~~~~0A- CMCS_ 1t C14N--Ct-ON tI00 r-A CN4 010-- O_NF0--Oicro r-N- ---00 _ - I - - mo oOO-oO_ O0-CM00 v V 0- ONo 0 Oe0 N C cNCM- 0O N 0n00-0O.O 0- d Ut- N N rq-0n 00

t~~~~~~~~~~~CVooen V a en vo tn en o.ONin 0 .t . o . . o. o.o. tn 0 eq , %Oe t~Ot ~~~~~~~- eq:1. C4 en WI--CM£ tl-O 0 0 -t <00 cr %O tn 081VCM%O 0 O -% °O- V oN tn ON ene0%00 N 0 0t%e V4 rL- CZ% %D C ^ ~

E 00t u N~ _ __ N__ __ O00%0%^^teM O O NC-o^d -%^^ 0 ;-- 0^-o 00 O\- rUoo~~~~~~~~s_o-tFN£ e e£ F o o_ o _ o£ e t £o xoW o0 S! 0% NB-.^N£o X e t X £ £ o > X N F b X b o m £ >~~~~~~~~~10N -oo .t eg - t

5-N :^ ol2 o' s ;'' z 0 _0 <> .ot__et£o ooo0

5 6 AMERICAN MUSEUM NOVITATES NO. 2849

Fig. 2. Photographs of lips of male paratype of Catostomus leopoldi, new species (A) and a female paratype of C. cahita, new species (B). Note difference in lip size (relative to head length). more expansive dorsal, pelvic, and anal fins 1977, for terminology). Tooth size decreases (table 1). and compression increases rapidly dorsad. The skull is slightly compressed dorsoven- Struts bracing the anteriolateral expan- trally. The frontoparietal fontanelle is var- sions of the basioccipital process against the iously developed, ranging from moderate ventral side of the braincase are short and width to a narrow slit. The dorsal surface of thick. the pterotic is broad, as in other Catostomus. The neural complex ofthe specialized We- The mandible (fig. 3A) is characteristic of berian apparatus is rather low and pleural Catostomus, with a moderately long gnathic ribs of the fourth vertebral centrum unite to ramus that is not sharply deflected ventrally form an acute angle, as in other Catostomus (as contrasted with Pantosteus; Smith, 1966). (in part, Nelson, 1948). The tripus is sub- The body of the maxilla is slender, with a triangular when the anterior and posterior long, posteriorly directed dorsal process that arms are excluded, its width about 45 percent articulates with the prepalatine cartilage (fig. of total length of bone. The posterior arm is 3C). The median dorsal process of the max- relatively long, about 60 percent oftotal length illary body is rather low. The ridge to which (fig. 3E). Width of the ascending process on the maxillodentary ligament attaches on the a line connecting articular tips ofthe anterior mesial side ofthe ventral arm is robust as in and posterior arms is about equal to length other Catostomus, and in the genus Pantos- ofthe thickened, distal portion ofthe anterior teus. arm. The pharyngeal bones, modified fifth cer- Scale and fin-ray counts of the holotype atobranchials characteristic ofcyprinoid fish- and paratypes are given in tables 2 and 3. No es, are heavy and deeply arched, with 34 to significant differences appear in these features 39 teeth (X = 36.3, s = 2.5, n = 3). The first among specimens from different localities. few ventral teeth are large and show little ETYMOLOGY: The name leopoldi is patron- dorsoventral compression (see Eastman, ymic, proposed to honor Aldo Leopold, whose 1986 SIEBERT AND MINCKLEY: CATOSTOMID FISHES 7

1mm

- 1 ...... -

F lmm Fig. 3. Line drawings of left mandible (mesial view), maxilla, and tripus of paratype of Catostomus leopoldi, new species (AMNH 51301SD) (A, C, E, respectively) and paratype of C. cahita, new species (AMNH 51303SD) (B, D, F, respectively). contributions to natural history of North Grandes, Chihuahua, immediately east ofand America and to the science of ecology are adjacent to tributaries ofthe Rio de Bavispe. immeasurable. Surprisingly, our cursory ex- We have not examined those specimens, nor amination of the literature found no species have we been able to find catostomids other named for Leopold. It seems especially ap- than Pantosteus plebeius Baird and Girard propriate to name this Mexican sucker for (Rio Grande Sucker) in that region. him since his "Song of the Gavilan" (Leo- HABITAT AND ASSOCIATED SPECIES: Habi- pold, 1940) was based in part on experiences tats range from pools in high-gradient head- along the stream (locally Arroyo Moctezuma) water streams flowing through Madrean that serves as type locality for the new species. Montane Forest to those in lower gradient The common name, Bavispe sucker, is for creeks of Montane Meadow Grassland the subbasin of the Rio Yaqui watershed in (Brown, 1982), mostly higher than 2000 m which the new species lives. elevation. Streams ranged from about 1 m to RANGE: Catostomus leopoldi is restricted 3 m in width. Depths ofpools where the suck- to headwater streams of the Rio de Bavispe er was concentrated ranged to 2 m, but most drainage, the major northeastern tributary of were less than a meter. Bottoms were stony, the vast Rio Yaqui watershed (fig. 4). Miller with cobble and gravel on riffles and sub- (1959) reported suckers, presumably of this stantial deposits offine gravels, sand, silt, and species, in headwaters of the Rio Casas sometimes organic materials in pools. The 00 N- r 'r en tn ~o ON ol c; c; C; n it r- ah el ell 0 *n en 00 Cu 04 - -_ tg - %N - 00 00 00 00 1 0 - 'r - co 0 (O 00 00 00 0 _-_

In00O Ch .1 0oN P o-o C) 1) N ON

enlt 0% 0 N I_ -_ _4 _q tn t g ld N N666- 0% N 00 V1) -4 _- _4 N I .- rn 000% . 10 00 N 1- 00 00 0%

0- llqt n e -4 00 000o 0C 00 1-4 I 0% e- N 4) ,-q 00 -4 00 C14 N C1 N 0 C) c0 00 C; C; -- o C) 000 C14 C14 wa .. C4 _. _4 qt 00 00 1-4 1.4 N N - U, z * 00

lqt tn O 0 00 oo 00 ,soN 0% - 00 r- N -4 en N N eq r 00 - -r 1-4 1-4 o 0% q *o Ntn 0% 0% 00 N 00 00 -4 en N tl-No It 000 eli r4) 'uoo oo If) N -4 en N cqb N C 81 00 '0 N - 0 00 N ON N tn 00 C- N eN r N 0 ~o NC,CN

lt en 0 o en 0 en 0 6 6 e en e- en) en en au Cu _ ,o. -_ - %o - C ,4 ont Un F- 0 0o GO 0t o Ddw: F-4 ~o F- 0

8 Cq O 00O0 0 0 0

- 0 en en 0*t 0 *t %0 O q ti -- L) N 1- en 1-4 ft4 -oo £ -S 1.0 Io 00ON 1% t I - - 0 m m 4n No .- 0 rF -4 - CM "- v ~1 4 1-

0 - oo oN 0% %0 CO)o *M *4 CM 6 6 6 U

_3 _ o o _4 in 0%S m Cq Ol *- oo*6 N * NI-t:06 92 eq e t3 qA CM4 CM CM4 z cm O~~~~.4 :t 0 0 1 0 - 0 U, 0 en 0 4-b engi0 PC 0. I3 k N CM C ,0 s 0 ao Ocoo ~0 %0 0 M eq 0 C. 0 8 o CM4 r- N N IC 00- in 't eq 0 CMtM

aA O O ON 0 0 - -1 t- _ _4 _4 t 8 C 1 0 2 re '0 t S 6 6o 0.t0 0 C 61 00 0%a -_ oo o . N *% % '0 o-_C;0o0 2 U4C

0 cre o%o 0 0%C as F 0 0%oN O% -0-_

NCM en 0 en 0 en °I 6 -, 6 0 0 4 tn Un 8 0 io 0° a I 2 ; F- C-

9 10 AMERICAN MUSEUM NOVITATES NO. 2849

Catostomus cahita, new species Cahita sucker Figures 2B, 3B, D, F, 5A, B; tables 1-3 Pantosteus plebeius (Baird and Girard), Meek, 1902, p. 75 (misidentification, fide Miller, 1976); Meek, 1904, p. 30 ("Minaca," misidentification, fide Miller, 1976). Catostomus sp., Miller, 1976, p. 5 ("an undescribed species of Catostomus"). Hendrickson et al. [1981], pp. 76-77 ("southern and southeastern headwaters of the Rio Yaqui system"). Minckley and Brown, 1982, p. 233 (listed, rios Yaqui and Mayo systems). Minckley et al., in press (listed "Catostomus sp. [Yaqui-Mayo]; from the rios Papigochic-Mayo headwaters"). DIAGNOSIS: A species ofCatostomus (sensu stricto) with moderately fine scales, usually 78 to 92 in the lateral line, and 9 or 10 dorsal fin rays (rarely 11). Lips small, the lower lobes completely cleft, but closely appressed at the midline. Lips with four or fewer rows of papillae crossing the midline on upper lips and eight or fewer rows crossing a longitudinal line on longest part of lower lobes. Pharyn- geal teeth more than 29, but fewer than 35 per arch. TYPES: Holotype. UMMZ 213393, a female, 148.1 mm SL, from Arroyo Ahumado, a tributary of Rio Tomochic, Chihuahua, Fig. 4. Drainage map of rios Yaqui and Mayo Mexico, 21.8 km from Tomochic along road of northwestem Mexico showing distribution of to Basaseachic (latitude 2821'N, longitude Catostomus leopoldi, new species (stars) and C. 107058'W, ca. 2000 m), June 16, 1978, D. A. cahita, new species (circles); open symbols indicate type localities. Hendrickson, J. E. Johnson, R. R. and F. H. Miller, W. L., P. H., and R. L. Minckley, and D. J. Siebert. Paratopotypes: UMMZ 213394, 4 females and 5 males, 101.1 to 140.5 mm fish was associated with cover in the form of SL, ASU 10561, 14 specimens, 46.0-105.0 boulders or undercut banks, typically over mm SL, and AMNH 51303SD, dry skeleton. clean, gravel or sand bottoms (Hendrickson Additional paratypes: AMNH 51305, mixed et al. [1981]). lot of 47 specimens, 26.0 to 144.0 mm SL, The Bavispe sucker is syntopic with Pan- and AMNH 51304SD, dry skeleton, from a tosteus plebeius in Arroyo Moctezuma and tributary ofRio Concheino, Chihuahua, Mex- occurred with Catostomus bernardini Girard ico, 16.5 km E Yepachic on road to Basa- (Yaqui sucker) in Rio Negro, but was the only seachic (latitude 28°22'N, longitude catostomid taken in arroyos de la Norteiia 108017'W, ca. 1800 m), June 15, 1978; ASU and San Antonio. Native cyprinids occurred 10563, mixed lot of 47 specimens, 37.0 to widely with the new species, as did the un- 101.3 mm SL, and AMNH 51306SD, dry described Yaqui trout, Salmo sp., at two lo- skeleton, from tributary of Rio Candameiia calities (table 4). Introduced species, Salmo above Salto de Basaseachic, Chihuahua, gairdneri Richardson and Lepomis macro- Mexico, at highway bridge on road to La Jun- chirus (Rafinesque), each co-occurred with C. ta, 8.5 km from the entrance of Basaseachic leopoldi as a few individuals at separate lo- Parque Nacional. calities. DESCRIPTION: Color pattern in alcohol and 1986 SIEBERT AND MINCKLEY: CATOSTOMID FISHES I1I body form of Catostomus cahita are shown TABLE 4 in figure 5A, B and proportional measure- Species Associated with Catostomus leopoldi, New ments are given in table 1. The general ap- Species and C. cahita, New Species pearance of the Cahita sucker is much the Compiled from Hendrickson et al. (1981). Intro- same as other western American Catostomus duced species are indicated by an asterisk (*). species. The body is terete and head length c. enters SL about four times. leo- C. The mouth is horizontal and inferior, with Associate poldi cahita papillose lips that resemble other Catostomus Salmonidae (fig. 2B). Lips are relatively small, their re- Salmo gairdneri Richardson* X X spective length and width 27 and 36 percent Salmo sp. X of head length. The dorsal fin, with 9 or 10, rarely 1 1, fin Catostomidae Catostomus bernardini Girard X X rays, is located about midway along SL. Tu- Pantosteus plebeius (Baird and Gi- berculate males possess a longer dorsal fin rard) X than females (fig. 5A, B), with a larger base; males have longer and more expansive anal, Cyprinidae pectoral, pelvic, and caudal fins as well (table Agosia chrysogaster Girard X X 1). Campostoma ornatum Girard X X open, al- Codoma ornata Girard - X The frontoparietal fontanelle is Gila sp.a - X though the posterior extension ofthe parietal G. robusta Baird and Girard X X portion is closed in some specimens, parietals Notropisformosus (Girard) X X meeting at the dorsal midline. The pterotics Pimephales promelas Rafinesque* - X are rooflike as in other Catostomus. Ante- riorly directed spines of the lateral ethmoids Ictaluridae extend as far forward as the anterior margins Ictalurus pricei (Rutter) - X ofthe lateral expansions ofthe supraethmoid. Crypinodontidae The ascending process ofthe dentary bone Cyprinodon sp. - X rises sharply from the mandibular body (fig. Centrarchidae 3B). The gnathic ramus to mandibular body Micropterus salmoides (Lac6pede)* - X angle is obtuse, as in other Catostomus, and Lepomis macrochirus Rafinesque* X X orientation of the mandibular symphysis is parallel to the plane of the gnathic ramus. a Reported as Gila pulchra (Girard) by Hendrickson The maxilla is robust, with a thick neck and et al. (1981). heavy head (fig. 3D). Its ventral keel is well developed and the mesial bifurcation of the articular tips of the anterior and posterior ventral arm where the maxillodentary liga- arms is about twice the length of the slender, ment attaches is pronounced. distal portion of the anterior arm. Its poste- The pharyngeal arch is robust and strongly rior arm is short, less than 60 percent oftotal curved. Tooth number is low for the genus length (fig. 3F). Catostomus, ranging from 29 to 32 (x = 30.3, Scale and fin-ray counts for the holotype s = 1.5, n = 3), and the ventralmost teeth and paratypes from three localities are given are large and little compressed. Tooth size in tables 2 and 3. Type material from the Rio decreases and dorsoventral compression in- Yaqui basin shows a consistent trend for creases rapidly along the row to result in rel- higher counts than in the other two popula- atively tiny teeth dorsally. tions, both of which are from the Rio Mayo The pleural ribs of the fourth centrum of drainage. Further investigations may dem- the Weberian apparatus unite to form an an- onstrate uniqueness ofsuckers from these two gle of less than 900 and are slightly retrorse, independent Pacific drainages. However, with broad blunt ends. The neural complex specimens are similar and consistent in os- is low. The tripus is rather inflated, its width teological characters and in their differences about 50 percent of its total length. Width of from both Catostomus leopoldi and the wide- the ascending process on a line connecting spread C. bernardini. 12 AMERICAN MUSEUM NOVITATES NO. 2849

Fig. 5. Photographs of female holotype (A) and male paratype (B) of Catostomus cahita, new species.

ETYMOLOGY: The epithet, cahita, a noun (Brown, 1982). It was recorded in greatest in apposition, is after the name for the abundance, however, in small streams at high subfamily of Uto-Aztec language spoken by and intermediate elevations. It avoids tur- Amerinds inhabiting the region in which the bulence and characteristically lives in pools, new species lives (Crumrine, 1980). The sug- but sometimes was taken from deeper runs. gested common name, Cahita sucker, follows Associated fishes ranged from Yaqui trout.at from the trivial epithet. high elevations, through most cyprinids RANGE: Catostomus cahita is widely dis- known from the region, to an undescribed tributed in a diversity ofstreams in the south- pupfish (Cyprinodon sp.) in the Rio Papi- ern and southeastern parts ofthe Rio Yaqui, gochic (table 4). principally in the Rio Papig6chic drainage, and in the upper Rio Mayo basin, Chihua- COMPARISONS hua, Mexico (fig. 4). HABITAT AND ASSOCIATED SPECIES: The Catostomus leopoldi and C. cahita are Cahita sucker inhabits streams of variable readily distinguishable from each other and size and configuration. It often is syntopic from other congeners. The large lips (fig. 2A) with Catostomus bernardini, especially in and slightly oblique mouth orientation ofthe larger systems, and ranges from high-eleva- Bavispe sucker, as opposed to smaller lips tion habitats in Madrean Montane Forest to (fig. 2B) and a horizontal mouth in C. cahita, lower elevations in Sinaloan Thornscrub always serve to separate the two new species. 1986 SIEBERT AND MINCKLEY: CATOSTOMID FISHES 13

Osteologically, the ascending process of the occur naturally in Pacific drainages ofnorth- mandible rises more abruptly in the Cahita western Mexico. Carpiodes carpio (Rafines- sucker than in the Bavispe sucker, the ante- que) has been introduced and is established rior portion ofits maxilla is more robust, and in the Rio Yaqui system (Hendrickson et al. its tripus is inflated rather than subtriangular [1981]). None ofthese is likely to be confused (excluding anterior and posterior arms) (fig. with either of the new taxa. Catostomus in- 3). signis and P. clarki inhabit the Gila River One of the striking features of both new basin, lower Colorado River system (Minck- taxa is their unusually low number of pha- ley, 1980b, 1980c), and M. austrinum oc- ryngeal teeth. Other Catostomus, with the cupies drainages far south and east of the possible exception of Catostomus wigginsi range of either of the two new species (Jen- Herre and Brock (Opata sucker), generally kins, 1980). have more than 45 teeth on each arch (East- Catostomus leopoldi is syntopic with C. man, 1977; Siebert, orig. data). As few spec- bernardini and Pantosteus plebeius in differ- imens were skeletonized, numbers of indi- ent parts of its range (Hendrickson et al. viduals examined for this character are [1981]). So far as is known, it is not syntopic admittedly small (two specimens from Ar- with both at any one locality. Catostomus royo Moctezuma and one from Rio de la bernardini has more dorsal (11-13) and pel- Nortefna for C. leopoldi; one specimen each vic (18-20, both fins) fin rays than C. leopoldi from Arroyo Ahumado, Rio Concheino, and and has larger sales as well (lateral line scales Rio Candameina for C. cahita). However, they 60-80, z = 72.8, s = 7.99, n = 39, scales came from widely separated localities in each around caudal peduncle fewer than 25, n = species' range, which lends credence to ob- 39; Siebert, orig. data). Lips of C. bernardini served counts not being seriously affected by are smaller, its caudal peduncle more shallow local population influences or small sample (x = 8.8% SL, s = 3.81, n = 39), and its sizes. Low tooth counts might also be attrib- dorsal fin base longer (Fx = 19.4% SL, s = uted to subadult specimens since tooth num- 10.67, n = 10) than that of C. leopoldi. bers increase with age in suckers. However, Catostomus wigginsiwas reported from the there is no evidence that this character Rio Yaqui system based on a single specimen changes after sexual maturity (Eastman, 1977) captured remote from localities inhabited by and all specimens examined for numbers of C. leopoldi (Hendrickson et al. [1981]). The teeth were adults. record was considered invalid by Van De- Possession offew teeth on each pharyngeal vender et al. (1985) who suspected it resulted arch may indicate that these two taxa are from transposition ofspecimens or labels. The more closely related to one another than either Opata sucker is nonetheless readily distin- is to other Catostomus. Tooth size relation- guished from the Bavispe sucker by its fewer ships along the arch may be another indica- lateral line scales (F = 72.3, s = 2.94, n = tion of close relationship. Although most 10; Siebert, orig. data), smaller lips, larger catostomids demonstrate considerable dif- head (x = 29.2% SL, s = 14.52, n = 10), and ferences between ventral and dorsal tooth size, shallow median cleft of the lower lip (with the disparity is exaggerated in these two taxa. up to four rows ofpapillae crossing the mid- So few teeth also raises the possibility that line). C. eopoldi and C. cahita are most closely Pantosteus plebeius, unlike Catostomus related to the genus Pantosteus, which also leopoldi, possesses typical Pantosteus lip tends to have fewer than 40 teeth per pha- structure (shallow median cleft between low- ryngeal arch (Siebert, orig. data). er lobes and well-developed lateral notches Six other catostomids, Catostomus ber- between upper and lower lips) and a closed nardini, Catostomus insignis Baird and Gi- frontoparietal fontanelle. rard (Sonoran sucker), C. wigginsi, Pantos- One skeletonized specimen from Arroyo teus plebeius, P. clarki (Baird and Girard) Moctezuma (AMNH 56171SD) almost cer- (desert mountainsucker), and Moxostoma tainly represents a hybrid between Catosto- austrinum (Bean) [west Mexican redhorse; mus leopoldi and Pantosteus plebeius. The including M. mascotae Regan (Buth, 1978)] frontoparietal fontanelle is nearly closed, with 14 AMERICAN MUSEUM NOVITATES NO. 2849 only a small anterior slit, a condition inter- allowed formation of extensive, linear dis- mediate between the two species at this lo- tributions of ancestral Catostomus. Such ex- cality. A probability of hybridization be- panded ranges have precedence in modern tween P. plebeius and a Catostomus suckers (viz., Pantosteus plebeius and the C. (presumably C. leopoldi) in the Rio Casas bernardini complex, see below). This hy- Grande drainage, Chihuahua, Mexico, was pothesized distribution pattern would have detected in hemoglobins by Koehn (1969). been established by early or middle Miocene. Smith (1966) had earlier reported P. plebeius Rates and magnitudes of displacements as the only Pantosteus not known to hybrid- along what is now the San Andreas Trans- ize, although introgression between P. ple- form since Eocene are well documented to beius and P. discobolus was hypothesized and fall within time constraints and relatively great recently documented (Smith et al., 1983). Our distances required to transport modern fishes specimen thus corroborates occurrence of from basin to basin along the continental natural hybridization between P. plebeius and margin. What is now the Sierra Nevada of a Catostomus (sensu stricto), adding infor- California-Nevada existed as an eroding mation to biology of the sucker genus Pan- highland in Miocene and Pliocene and was tosteus. Zuckerman et al. (1984) recently possibly interconnected with the northern recorded P. plebeius hybridizing with non- Sierra Madre Occidental due to the interven- native C. commersoni (Lacepede) in the up- ing presence of what are now the Transverse per Rio Grande of Colorado. Ranges ofCalifornia. The latter represent mi- Catostomus cahita is broadly syntopic with croplate(s) that later migrated west-north- C. bernardini. Meristic characters and dorsal west and were rotated and uplifted in appar- fin length that distinguish the Yaqui sucker ent response to Miocene extensional opening from C. leopoldi also separate it from the of basins to the west. Uplift, westerly tilting, Cahita sucker. Pantosteus plebeius and C. and clockwise rotation of the Sierra Nevada wigginsi are not known to be syntopic with proceeded from south to north beginning 18 C. cahita, but features that distinguish them mya, and has been mostly accomplished since from C. leopoldi also differentiate them from 5 mya. the Cahita sucker. We cursorily examined possibilities for relationships among the two new Mexican Ca- ZOOGEOGRAPHIC RELATIONS tostomus and species to the north and north- The Sierra Madre Occidental was progres- west. Striking mosaics of resemblances in sively formed by volcanic activity from 30 scalation, body shape, pigmentation, and lip to 22 million years ago (mya) in the north structure are evident among Catostomus leo- and 10 to 8 mya in the south (McDowell and poldi, C. cahita, C. fumeiventris Miller Clabaugh, 1979; Cameron et al., 1980). The (Owens Valley sucker, occurring on the mountains are surrounded on the east, north, southeast end of the Sierra Nevada, Califor- and west by basin and range topography that nia [Miller, 1973]), and further to the C. mi- developed from east to west between 27 and crops group ofsuckers (Moyle and Marciochi, 7 mya (Chapin, 1979; Dokka and Merriam, 1975) of the northern Sacramento River and 1982). Montane habitats in northwestern adjacent drainages (C. microps Rutter [Pitt Mexico have thus been available to fishes for River, California], C. warnerensis Snyder a long period of time. [Warner Valley, Oregon], C. tahoensis Gill Minckley et al. (1986) reviewed tecto- and Jordan [Lahontan Basin, Nevada, Cali- nism along coastal provinces ofwestern North fornia, Oregon], and Catostomus sp. [Sur- America and documented continuity of a prise Valley, Nevada]). Disruption of distri- southwestern North American ichthyofauna butions attained either by vicariance, stretching north from Mexico into California. dispersal, or a combination ofthe two, would They considered it possible that transport have occurred through westward displace- northward on tectonically displaced coastal ment and rotation of mountain masses, de- formed and form- velopment ofextension zones reflected in ba- terranes, dispersal through sin and range structure between the northern ing basin and range areas, or both, could have Sierra Madre Occidental and the southerm 1986 SIEBERT AND MINCKLEY: CATOSTOMID FISHES 15

Colorado Plateau, and extension and hori- ico (Miller, 1976; Hendrickson, 1984). Other zontal displacement along the broad San An- southwestern freshwater fishes that may be dreas Transform (Minckley et al., 1986). of comparable age include those mentioned Distributional relations paralleling those in the preceding paragraph, and perhaps some just described were suggested for species of salmonids (Salmo spp.). the subgenus Temeculina of Gila (G. pulchra [Girard], G. pupurea [Girard], Gila sp., and LITERATURE CITED G. ditaenia Miller of Mexico, and G. orcutti Barbour, Clyde D., and Robert Rush Miller [Eigenmann and Eigenmann] and G. crassi- 1978. A revision ofthe Mexican Cyprinid fish cauda [Baird and Girard] of California) ad- genusAlgansea. Misc. Publ. Mus. Zool., jacent to the Sierra Madre Occidental and Univ. Michigan, no. 155, pp. 1-72. west of the Sierra Nevada. Pantosteus ple- Brown, David E. (ed.) beius of Mexico and P. santaanae (Snyder) 1982. Biotic communities of the American of California may also show a comparable Southwest-United States and Mexico. distribution (Minckley et al., 1986). Phylo- Desert Plants, Special Issue, vol. 4, no. genetic relationships of Temeculina to the 1-4, pp. 1-342. south are with the central Mexican genus Al- Buth, Donald G. gansea (Barbour and Miller, 1978). Another 1978. Biochemical systematics of the Mox- ostomatini (Cypriniformes, Catostomi- probable relative of Algansea (D. A. Hen- dae). Dissertation Abstr. Internat., vol. drickson, personal commun.), Agosia chrys- 39, no. 5. ogaster, ranges from Sinaloa, Mexico, north- Cameron, Maryellen, W. C. Bagby, and Kenneth ward to the Bill Williams River, Arizona, east L. Cameron ofthe Sierra Nevada and the Mohave Desert 1980. Petrogenesis of voluminous mid-Ter- (Minckley, 1980a). This pattern also appears tiary ignimbrites of the Sierra Madre both east and west of the Sierra Nevada Occidental, Chihuahua, Mexico. Con- among cyprinodontoids. Livebearers of the trib. Minerol. Petrol., vol. 74, no. 3, pp. family Goodeidae occur in central Mexico, 271-284. and then across a substantial hiatus marked Chapin, Charles E. by isolates 1979. Evolution of the Rio Grande Rift: a and fossil occurrences to reappear summary. In Robert E. Reiker (ed.), Rio in California and Nevada as primitive, egg-lay- Grande rift: tectonics and magmatism, ing representatives (Parenti, 1981) (northern pp. 1-5. Washington, D.C., Amer. Geo- populations are referred to Empetrichthyidae physical Union, 438 pp. by Miller and Smith, 1986). Pupfishes (genus Crumrine, N. Ross Cyprinodon) arc from northern Mexico 1980. Cahitians. In Dictionary ofIndian tribes through Arizona to eastern California (Mil- of the Americas, pp. 375-387. Ameri- ler, 1981), and Minckley et al. (1986) further can Indian Publishers, vol. 1, 580 pp. suggested examination of possible relation- Dokka, Roy K., and Richard H. Merriam of western North American Fundulus 1982. Late cenozoic extension ofnortheastern ships Baja California, Mexico. Geol. Soc. with those of southern and central Mexico Amer. Bull., vol. 93, no. 5, pp. 371-378. (Miller, 1955). Eastman, Joseph T. We therefore propose that Catostomus leo- 1977. The pharyngeal bones and teeth of ca- poldi and C. cahita (and likely C. wigginsi of tostomid fishes. Amer. Midl. Natural- the Rio Sonora basin) represent relicts or de- ist, vol. 97, no. 1, pp. 68-88. rivatives ofan old regional fauna. Pantosteus Ferris, Stephen D., Donald G. Buth, and Gregory plebeius, on the basis of its vast distribution S. Whitt from the uppermost Rio Grande basin, 1982. Substantial genetic differentiation U.S.A., to the Rio Mezquital of Mexico among populations of Catostomus ple- (Smith, 1966; Minckley, 1980c), may also be beius. Copeia, no. 2, pp. 444-449. an Hendrickson, Dean A. old faunal component, and the complex 1984. Distribution records of native and ex- of catostomids represented in the Rio Yaqui otic fishes in Pacific drainages ofnorth- by C. bernardini has a similarly wide distri- ern Mexico. Jour. Arizona-Nevada bution from the Bill Williams and Gila ba- Acad. Sci., vol. 18, no. 2, pp. 33-38. sins, U.S.A., to the Rio San Lorenzo ofMex- Hendrickson, Dean A., W. L. Minckley, Robert 16 AMERICAN MUSEUM NOVITATES NO. 2849

Rush Miller, Darrell J. Siebert, and Patricia H. (genus Cyprinodon) in the American Minckley Southwest. In Robert J. Naiman and "1980" [1981]. Fishes of the Rio Yaqui basin, David L. Soltz (eds.), Fishes in North Mexico and United States. Jour. Ari- American deserts, pp. 39-94. Wiley, zona-Nevada Acad. Sci., vol. 15, no. 3, New York, i-xi + 552 pp. pp. 65-106. Miller, Robert Rush, and Michael L. Smith Hubbs, Carl L., and Karl F. Lagler 1986. The origin and geography of the fishes 1970. Fishes of the Great Lakes region. Ann ofcentral Mexico. In Charles H. Hocutt Arbor, Univ. Michigan Press, 213 pp. and Edward 0. Wiley (eds.), Zoogeog- Jenkins, Robert E. raphy of freshwater fishes of North 1980. Moxostoma austrinum (Bean), west America, pp. 487-517. Wiley-Intersci- Mexican redhorse. In David S. Lee et ence, New York. al. (eds.), Atlas of North American Minckley, W. L. freshwater fishes, p.413. Raleigh, North 1980a. Agosia chrysogasterGirard, longfin dace. Carolina State Museum, i-x + 854 pp. In David S. Lee et al. (eds.), Atlas of Koehn, Richard K. North American freshwater fishes, p. 1969. Hemoglobins of fishes ofthe genus Ca- 141. Raleigh, North Carolina State Mu- tostomus in western North America. seum, i-x + 854 pp. Copeia, no. 1, pp. 21-30. 1980b. Catostomus clarki (Baird and Girard), Leopold, Aldo desert sucker. In David S. Lee et al. 1940. Song of the Gavilan. Jour. Wildlife (eds.), op. cit., p. 373. Raleigh, North Management, vol. 4, no. 3, pp. 329- Carolina State Museum, i-x + 854 pp. 332. 1980c. Catostomus insignis Baird and Girard, McDowell, Fred W., and Stephen E. Clabaugh Sonoran sucker. In David S. Lee et al. 1979. Ignimbrites of the Sierra Madre Occi- (eds.), op. cit., p. 380. Raleigh, North dental and their relation to the tectonic Carolina State Museum, i-x + 854 pp. history of western Mexico. In Charles 1980d. Catostomus plebeius Baird and Girard, E. Chapin and Wolfgang E. Elston (eds.), Rio Grande sucker. In David S. Lee et Ash-flow tuffs, pp. 113-124. Geol. Soc. al. (eds.), op. cit., p. 387. Raleigh, North Amer., Spec. Pap. no. 180, 211 pp. Carolina State Museum, i-x + 854 pp. Meek, Seth E. Minckley, W. L., and David E. Brown 1902. A contribution to the ichthyology of 1982. Part 6. Wetlands. In David E. Brown Mexico. Field Columbia Mus., Zool. (ed.), Biotic communities of the Amer- Serv., vol. 3, no. 6, pp. 63-128. ican Southwest-United States and 1904. The fresh-water fishes of Mexico north Mexico, pp. 222-287, 333-341, + Lit- of the isthmus of Tehuantepec. Ibid., erature Cited. Desert Plants, Special Is- Zool. Ser., vol. 5, pp. i-lxii, 1-252. sue, vol. 4, no. 1-4, pp. 1-342. Miller, Robert Rush Minckley, W. L. , Dean A. Hendrickson, and Carl 1955. An annotated list of the cyprinodontid E. Bond fishes of the genus Fundulus, with the 1986. Geography of western North American description ofFunduluspersimilis from freshwater fishes: description and rela- Yucatan. Occas. Papers Mus. Zool., tion to intercontinental tectonics. In Univ. Michigan, no. 568, pp. 1-25. Charles H. Hocutt and Edward 0. Wi- 1959. Origin and affinities of the freshwater ley (eds.), Zoogeography of freshwater fish fauna ofwestern North America. In fishes of North America, pp. 519-613. C. L. Hubbs (ed.), Zoogeography, pp. Wiley-Interscience, New York. 187-222. Amer. Assoc. Adv. Sci. Symp., Moyle, Peter B., and Alan Marciochi vol. 51, 1958. 1975. Biology of the Modoc sucker, Catosto- 1973. Two new fishes, Gila bicolorsnyderi and mus microps, in northern California. Catostomus fumeiventris, from the Copeia, no. 3, pp. 556-560. Owens River basin, California. Occas. Nelson, Edward M. Papers Mus. Zool., Univ. Michigan, no. 1948. The comparative morphology of the 667, pp. 1-19. Weberian apparatus of the Catostomi- 1976. An evaluation of Seth E. Meek's con- dae and its significance in systematics. tribution to Mexican ichthyology. Jour. Morphol., vol. 83, no. 1, pp. 225- Fieldiana, Zool., vol. 69, no. 1, pp. 1- 251. 31. Parenti, Lynne R. 1981. Coevolution of deserts and pupfishes 1981. A phylogenetic and biogeographic anal- 1986 SIEBERT AND MINCKLEY: CATOSTOMID FISHES 17

ysis ofcyprinodontiform fishes (Teleos- Van Devender Thomas R., Amadeo M. Rea, and tei, Atherinomorpha). Bull. Amer. Mus. Michael L. Smith Nat. Hist., vol. 168, pp. 341-557. 1985. The Sangamon interglacial vertebrate Smith, Gerald R. fauna from Rancho la Brisca, Sonora, 1966. Distribution and evolution ofthe North Mexico. Trans. San Diego Soc. Nat. American catostomid fishes of the sub- Hist., vol. 21, no. 2, pp. 23-55. genus Pantosteus, genus Catostomus. Zuckerman, Lawrence D., Ted Smith, and Robert Misc. Publ. Mus. Zool., Univ. Michi- J. Behnke gan, no. 129, pp. 1-132. 1984. Hybridization between Catostomusple- Smith, Gerald R., John G. Hall, Richard K. Koehn, beius and C. commersoni in the Rio and David J. Innes Grande basin, Colorado. Program and 1983. Taxonomic relationships of the Zuni Abstr. of64th Ann. Meeting, Amer. Soc. Mountain sucker, Catostomus discobo- Ichthyologists and Herpetologists, Nor- lus yarrowi. Copeia, no. 1, pp. 32-48. man, Oklahoma, p. 219.

IRecent issues of the Novitates may be purchased from the Museum. Lists ofback issues of the Novitates, Bulletin, and Anthropological Papers published during the last five years are available free of charge. Address orders to: American Museum ofNatural History Library, Department D, Central Park West at 79th St., New York, New York 10024. i