SPECIAL PUBLICATIONS THE MUSEUM TEXA&jT^R9utioiiary Relationships of Spiny Pocket 7mce, Genus Liomys Hugh H. Ge noways No. 5 Hi in. December 1973 TEXAS TECH UNIVERSITY Grover E. Murray, President Glenn E. Barnett, Executive Vice President Regents. — Bill E. Collins (Chairman), J. Fred Bucy, R. Trent Campbell, Clint Formby, John J. Hinchey, Frank Junell, A. J. Kemp, Charles G. Scruggs, and Judson F. Williams. Academic Publications Policy Committee. — J . Knox Jones, Jr. (Chairman), Dilford C. Carter (Managing Editor), C. Leonard Ainsworth, Craig C. Black, Frank B. Conselman, Samuel E. Curl, Ray C. Janeway, W. R. Johnson, S. M. Kennedy, Thomas A. Langford, George F. Meenaghan, Harley D. Oberhelman, Robert L. Packard, and Charles W. Sargent. The Museum Special Publication No. 5 368 pp., 66 figs. 7 December 1973 $7.00 Special Publications of The Museum are numbered separately and published on an irregular basis under the auspices of the Dean of the Graduate School and Director of Academic Pub¬ lications, and in cooperation with the International Center for Arid and Semi-Arid Land Studies. Copies may be obtained on an exchange basis from, or purchased through, the Ex¬ change Librarian, Texas Tech University, Lubbock, Texas 79409. Texas Tech Press, Lubbock, Texas 1973 SPECIAL PUBLICATIONS THE MUSEUM TEXAS TECH UNIVERSITY Systematics and Evolutionary Relationships of Spiny Pocket Mice, Genus Liomys Hugh H. Genoways No. 5 December 1973 TEXAS TECH UNIVERSITY Grover E. Murray, President Glenn E. Barnett, Executive Vice President Regents. — Bill E. Collins (Chairman), J. Fred Bucy, R. Trent Campbell, Clint Formby, John J. Hinchey, Frank Junell, A. J. Kemp, Charles G. Scruggs, and Judson F. Williams. Academic Publications Policy Committee. — J. Knox Jones, Jr. (Chairman), Dilford C. Carter (Managing Editor), C. Leonard Ainsworth, Craig C. Black, Frank B. Conselman, Samuel E. Curl, Ray C. Janeway, W. R. Johnson, S. M. Kennedy, Thomas A. Langford, George F. Meenaghan, Harley D. Oberhelman, Robert L. Packard, and Charles W. Sargent. The Museum Special Publication No. 5 368 pp., 66 figs. 7 December 1973 $7.00 Special Publications of The Museum are numbered separately and published on an irregular basis under the auspices of the Dean of the Graduate School and Director of Academic Pub¬ lications, and in cooperation with the International Center for Arid and Semi-Arid Land Studies. Copies may be obtained on an exchange basis from, or purchased through, the Ex¬ change Librarian, Texas Tech University, Lubbock, Texas 79409. Texas Tech Press, Lubbock, Texas 1973 CONTENTS Introduction . 5 Methods, Materials, and Acknowledgments . 8 Nongeographic Variation . 16 Variation with Age . 16 Secondary Sexual Variation . 31 Individual Variation . 39 Molt . 40 Systematic Accounts . 44 Genus Liomys . 44 Key to the Species of Liomys . 44 Liomys irroratus . 47 Liomys irroratus alleni . 99 Liomys irroratus bulleri . 106 Liomys irroratus guerrerensis . 108 Liomys irroratus irroratus . 110 Liomys irroratus jaliscensis . 113 Liomys irroratus texensis . 116 Liomys irroratus torridus . 119 Liomys pictus . 124 Liomys pictus annectens . 175 Liomys pictus hispidus . 179 Liomys pictus pictus . 185 Liomys pictus plantinarensis . 192 Liomys spectabilis . 195 Liomys salvini . 202 Liomys salvini crispus . 235 Liomys salvini salvini . 237 Liomys salvini vulcani . 243 Liomys adspersus . 246 Status of Liomys centralis . 256 Specific Relationships . 257 External and Cranial Morphology . 257 Tooth Structure . 272 Morphology of Gians Penis . 283 Bacular Morphology . 289 Morphology of Spermatozoa . 293 Morphology of Hind Feet . 296 Comparative Karyology . 297 Ectoparasites . 301 Endoparasites . 304 Reproduction . 305 Pterygoid Structure . 310 Pelage . 311 Distribution . 314 Other Studies . 315 Evolutionary and Zoogeographic Relationships . 317 Literature Cited . 330 Appendix I — Gazetteer . 341 Appendix II — Ectoparasites of Liomys and Heteromys . 356 Appendix III — Matrix of characters used in the analyses of evolutionary relationships . 366 Appendix IV — Methods for scoring evolutionary characters . 368 Systematics and Evolutionary Relationships of Spiny Pocket Mice, Genus Liomys Hugh H. Genoways INTRODUCTION Spiny pocket mice of the genus Liomys are members of the rodent family Heteromyidae and together with the genus Heteromys form the subfamily Heteromyinae. Their geographic range extends from northern Sonora, in western Mexico, and southern Texas southward to the vicinity of the Panama Canal Zone. Within this area, members of the genus occur mainly in dry to arid situations being replaced in areas of rain forest and cloud forest by members of the genus Heteromys. The vernacular name for Liomys is based on the fact that many of their hairs have been modified in the form of stiff, aristiform spines. However, my field companions have other names for members of the genus such as “green mice” and others that would not be appropriate to publish, because of the ten¬ dency of these rodents to rot quickly after being trapped and because their thin skin tears easily during preparation. Although spiny pocket mice are relatively common inhabitants of much of Mexico and Central America, it was not until 1868 that representatives of the genus were described and not until 1902 that the generic name was proposed. Gray (1868) described the first two species that are presently included in the genus Liomys, although he placed them in the genus Heteromys Desmarest, 1817, where they remained until the genus Liomys was described. The two species ( Heteromys irroratus and Heteromys albolimbatus ) were based on only three specimens from Oaxaca — the holotype of irroratus from an unspecified locality in the state and the two specimens of albolimbatus from La Parada. A third species ( Heteromys adspersus ), now assignable to Liomys, was described by Peters (1874) from Panama. These three species together with other species of Heteromys known at the time were reviewed by Alston (1879-82) who recog¬ nized only two species, Heteromys desmarestianus and Heteromys longicaudatus , in Mexico and Central America. The species irroratus, albolimbatus, and adsper¬ sus were placed in the synonymy of H. longicaudatus. A fourth species ( Hete¬ romys alleni ) was described by Coues (in J. A. Allen, 1881:187-89) based on a specimen from Rio Verde, San Luis Potosi, although Allen in the same paper expressed some doubt as to the validity of the species in view of Alston’s findings. Later, however, Allen (1 891 :268-272) concluded on the basis of additional speci¬ mens from Brownsville, Texas, and Moroleon, Guanajuato, that Heteromys alleni was a valid species. The next descriptions of members of the genus appeared in 1893 when Thomas (1893^:329-32, 1 8936:233-34) named three species — Hete¬ romys bulleri, H. salvini, H. pictus — and a subspecies of H. salvini. In the first of these papers Thomas recognized all species named to that time and suggested 5 6 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY that species of the genus could be defined by a combination of the characteristics based on number of plantar tubercles and hairy or naked soles of hind feet. Four years later, J. A. Allen (1897) described another species, Heteromys hispidus, from Compostela, Nayarit, and listed known members of the genus, dividing them into three groups based on Thomas’ characteristics. It is of interest to note that Allen inadvertently omitted from his listing Heteromys pictus Thomas, 1 893, with which hispidus is now considered synonymous. The genus Liomys was formally described in 1902 by C. H. Merriam with Heteromys alleni Coues as the type species. Liomys was distinguished from Hete¬ romys by Merriam mainly on the basis of the absence of secondary lobes or per¬ manent enamel islands on the molars of Liomys that are present in Heteromys. In the same paper, Merriam described 1 1 new species and four new subspecies of Liomys and also Heteromys annectens, which is now considered to be a member of the genus Liomys. The following year Elliot ( 1 903a: 146-47, 19036:233) de¬ scribed two new species, one from Morelos ( Heteromys exiguus ) and the other from Veracruz ( Heteromys paralius). In these two papers as in his other major works of the next several years, Elliot (1904:368-82; 1905:316-23; 1907:344-47) considered Liomys to be a subgenus of Heteromys. Goldman (1904:82) char¬ acterized a species, Liomys parviceps, from Michoacan and in so doing accorded Liomys generic rank. In the next four years, J. A. Allen (1906:211, 251, 1908: 652) described two new species and a new subspecies ( Heteromys jaliscensis, Heteromys vulcani, Heteromys pictus escuinapae), and placed them in the genus Heteromys without commenting on the status of Liomys. Therefore, by 1911 when Goldman’s comprehensive work, “Revision of the spiny pocket mice (genera Heteromys and Liomys)” appeared, 25 species and five subspecies were recognized in the genus. Goldman recognized nine species and 17 subspecies, besides the nominal races, based upon his study of 1003 speci¬ mens. In addition, he described one new species, Liomys guerrerensis, from Omil- teme, Guerrero, and one new subspecies, Liomys irroratus pretiosus, from Met- laltoyuca, Puebla. Although Goldman did not accord them subgeneric rank, he did divide the genus into three species groups — irroratus group, pictus group, and crispus group. This was the last major work to appear dealing with Recent species of the genus until 1948, although in the intervening years descriptions of one species, Liomys anthonyi (Goodwin, 1932a:2), and three subspecies — Liomys salvini aterrimus (Goodwin, 1938:4), Liomys irroratus pullus (Hooper, 1947: 47), Liomys irroratus acutus (Hall and Villa, 1948:253) — appeared, and Liomys vulcani was reduced to subspecific status under Liomys salvini (Goodwin, 1946: 374). In 1948, Hooper and Handley (1948) presented a synopsis of the known races of Liomys irroratus and analyzed the trends in geographic variation within the species. Since 1 948, the only papers dealing with the taxonomy of the genus were a description of a new species, Liomys pinetorum (Goodwin, 19566:2-3), from Chiapas and a new subspecies, Liomys irroratus yautepecus (Goodwin, 1956a:7-8), from Oaxaca; the latter was subsequently arranged as a synonym of L. i. irroratus (Goodwin, 1969:148). The only nominal fossil member of the genus is Liomys centralis described by Hibbard (194 la: 349) from the Rexroad GENOWAYS— SYSTEMATICS OF LIOMYS 7 Fauna of western Kansas, although other fossil and subfossil material has been reported from San Josecito Cave, Nuevo Leon, by Cushing (1945:185) and Jakway (1958:320), and from caves in southern Tamaulipas by Koopman and Martin (1959:2, 6) and Dalquest and Roth (1970:226). Therefore, at the begin¬ ning of my study 1 1 Recent species, 22 Recent subspecies, and one fossil species were recognized in the genus. The purpose of this study was first to examine the taxonomic relationships of the species of the genus Liomys and, after forming a new systematic arrangement of the taxa, to examine the evolutionary relationship of the species within the genus and to evaluate their relationships to members of the genus Heteromys. The present work is easily divisible into three sections — nongeographic variation, geographic variation, and an assessment of evolutionary relationships — although no one section is independent from the other two. The section on nongeographic variation examines variation with age, secondary sexual variation, individual variation, and variation resulting from molt. Study of geographic variation and specific relationships using both univariate and multivariate statistics has shown that only four of the 1 1 species recognized prior to this study are actually distinct; these species are Liomys irroratus (seven subspecies), Liomys pictus (four sub¬ species), Liomys salvini (three subspecies), and Liomys adspersus (monotypic). In addition to these four species, one other monotypic species, Liomys spectabilis, recently described by me (Genoways, 1971) is recognized, resulting in a total of five species within the genus. Because a fossil record of the genus is nearly non¬ existent, I have attempted to analyze the relationships of these five species to each other and to members of the genus Heteromys by study of such characteris¬ tics as external and cranial morphology, structure of unworn premolars and molars, bacular morphology, morphology of glans penes, comparative karyology, sperm morphology, ectoparasite faunas, structure of pterygoid bone, pelage char¬ acteristics, morphology of hind feet, and reproductive patterns. The section on specific relationships also should be used as an appendix to the section on geo¬ graphic variation because all of these same characteristics were used in deter¬ mining the species to be recognized. I decided to place all of the specific char¬ acteristics in a single section for ease of comparison rather than scattering them throughout the sections on geographic variation within the accounts of individual species. 8 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY METHODS, MATERIALS, AND ACKNOWLEDGMENTS Discussed here are the methods and materials used in the analysis of nongeo¬ graphic and geographic variation or related to more than one part of the section on evolutionary relationships. Specific methods and materials used in individual parts of the section on evolutionary relationships are detailed in introductory re¬ marks to each of the parts. In the course of this study, 7186 specimens representing the five species of the genus Liomys were examined. The vast majority of these specimens were stand¬ ard museum skins and skulls accompanied by the appropriate field data. Addi¬ tional material consisted of complete skeletons, skins without skulls, skulls un¬ accompanied by skins, and specimens preserved in alcohol. All of the holotypes of nominal taxa of Liomys were examined except those of Liomys adspersus, Liomys irroratus, Liomys albolimbatus, Liomys salvini, Liomys salvini nigres- cens, and Liomys pictus, although the latter five were kindly examined in the British Museum (Natural History) for me by Dr. J. Knox Jones, Jr. The holotype of Liomys adspersus , which supposedly is deposited in the Berlin Museum, was beautifully figured by Peters (1 874) in his original description. From most adult specimens and from selected individuals of other age cate¬ gories four external measurements and weight were recorded from the specimen labels and 10 cranial measurements were taken by means of dial calipers. In or¬ der to clarify exactly in what manner measurements were taken, I have defined each below; letters in parenthesis are used to identify the measurement in some figures. All measurements in text are given in millemeters unless otherwise noted. Total length (TL). — Distance from tip of nose to tip of fleshy part of tail; recorded by preparator. Length of tail (TV). — Distance from joint between proximal tail vertebrae and sacrum to fleshy tip of tail; recorded by preparator. Length of hind foot (HF). — Distance from tip of longest claw to heel; recorded by prepa¬ rator. Length of ear. — Distance from bottom of notch to distalmost edge of fleshy part of ear; recorded by preparator; not used in analysis of nongeographic or geographic variation. Weight. — Total weight recorded in grams by preparator; not used in analysis of nongeo¬ graphic or geographic variation. Greatest length of skull (GLS). — From A to B in Fig. 1; greatest distance from the an- teriormost projection of the nasal bones to the posteriormost portion of the occipital bone. Zygomatic breadth (ZB). — From C to D on Fig. 1; greatest width across zygomatic arch¬ es at right angle to longitudinal axis of cranium. Interorbital constriction (IOC). — From E to F on Fig. 1; least width across the interor¬ bital constriction at right angle to longitudinal axis of cranium. Mastoid breadth (MB). — From G to H on Fig. 1; greatest width across mastoid processes at right angle to the longitudinal axis of cranium. Length of nasals (LN). — From A to I on Fig. 1; greatest distance from anteriormost pro¬ jection of nasal bones to the posteriormost projection of the nasals along their medial suture. Length of rostrum (LR). — From A to J on Fig. 1; greatest distance from notch lateral to lacrimal bone to anteriormost projection of nasal bone on the same side of the cranium. Length of maxillary toothrow (MTR). — From K to L on Fig. 1; distance from anterior lip of alveolus of the premolar to the posterior lip of the alveolus of M3. Depth of braincase (DBC). — From M to N on Fig. 1; least distance from basioccipital- basisphenoid complex to dorsal most portion of cranium. GENOWAYS— SYSTEM ATICS OF LIOMYS 9 A J E C Q G O B K L G Fig. 1. — Skull of Liomys pictus illustrating points between which measurements described in text were taken. Interparietal width (IW). — From O to P on Fig. 1; greatest transverse width measured from the lateralmost projections of the interparietal bone at right angle to longitudinal axis of cranium. Interparietal length (IL). — From Q to R on Fig. 1; greatest distance from anteriormost projection of interparietal bone to posteriormost border of interparietal bone; this measure¬ ment was always taken along medial line of cranium even when there was a notch in pos¬ terior border. Variation in color was studied by use of a Photovolt Photoelectric Reflection Meter, Model 610, on which reflectance values are recorded as a percentage of pure white (see Lawlor, 1965; Dunnigan, 1967). Readings of color reflectance for red, green, and blue were taken in the middorsal region of specimens with un¬ worn pelage. Study of variation in color was extremely difficult for several rea¬ sons. Although I attempted to use specimens collected only in June, July, Aug¬ ust, and September, those from many critical areas were collected at other times of the year, thus forcing me to use material obtained in several other months. An¬ other problem encountered was that many critical specimens were collected by Nelson and Goldman in the late 1890’s and early 1900’s whereas many others have been collected since 1950; the possible changes in color resulting from long¬ term museum storage, even under the best of circumstances, is unknown. The conditions under which specimens are housed at various museums introduces still another variable. Results of my analysis of variation in color should be viewed, therefore, with these variables in mind. 10 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY Four qualitative cranial characters were recorded and scored for each adult and subadult specimen examined for inclusion in the analysis of geographic vari¬ ation. Some authors (Berry and Searle, 1963; Hedges, 1969) have termed these “epigenetic polymorphism”; however, I have avoided use of this term because I felt epigenetic polymorphism carried connotations that could not be proven on the basis of my data. Three of these characters (shape of posterior termination of nasal bone, length of nasals in comparison with length of premaxillae, and con¬ dition of the posterior margin of the interparietal bone) were used by Goldman (1911) in his analysis of Liomys\ the fourth was based upon my observation that some individuals had the interparietal divided, whereas others did not. Another character (shape of interparietal bone) was used by Goldman, but I found this character to be so variable that it was unsatisfactory for analysis. The states for each of these characters, together with the score used in the analysis (see Fig. 2), are as follows: shape of posterior margin of nasals — emarginate (1), rounded (2), truncate (3); length of premaxillary bones — longer than nasals (1), equal to nasals (2); condition of interparietal bone — undivided (1), divided (2); posterior margin of interparietal bone — notched (1), slightly notched (2), unnotched (3). Statistical procedures requiring computer analyses were performed on the GE 635 computer at The University of Kansas; simpler procedures were carried out on the Olivetti Underwood Programma 101 in the Museum of Natural His¬ tory of Kansas. Univariate analyses of geographic variation were performed us¬ ing a program (UNIVAR) written and extensively used by Power (1970). This program yields standard statistics (mean, range, standard deviation, standard er¬ ror of the mean, variance, and coefficient of variation) and, when two or more groups are being compared, employs a single-classification analysis of variance or anova (F-test, significance level .05) to test for significant differences between or among the means (Sokal and Rohlf, 1969). When means were found to be sig¬ nificantly different, the Sums of Squares Simultaneous Test Procedure (SS-STP) developed by Gabriel (1964) was used to determine maximally nonsignificant subsets. Multivariate analyses were performed using the NT-SYS programs developed at The University of Kansas by F. J. Rohlf, R. Bartcher, and J. Kishpaugh. In all multivariate analyses, the OTUs were grouped localities discussed below and the values for each character were means for the measurement or mean score for qualitative cranial characters. Matrices of Pearson’s product-moment correlation were computed, and phenetic distance coefficients were derived from standard¬ ized character values. Cluster analyses were conducted using UPGMA (unweight¬ ed pair-group method using arithmetic averages) on the correlation and distance matrices and a phenogram was generated for each. Phenograms were compared with their respective matrices, and a coefficient of cophenetic correlation was computed for each comparison; I have shown only the phenetic distance pheno¬ grams because their coefficients of cophenetic correlation were larger than for the correlation phenogram and, generally, the results of the distance phenogram agreed more closely with results of other analyses. A matrice of correlation among characters then was computed, and the first three principal components extracted. GENOWAYS— SYSTEMATICS OF LIOMYS 1 1 A B 1 2 c 3 1 2 D Fig. 2. — Semidiagrammatic illustration of classes used in scoring qualitative cranial characters. A, shape of posterior margin of nasals (1, emarginate; 2, rounded; 3, truncate); B, length of premaxillary bones (1, longer than nasals; 2, equal to nasals); C, posterior mar¬ gin of interparietal bone (1, notched; 2, slightly notched; 3, unnotched); D, condition of interparietal bone (1, undivided; 2, divided). Both two-dimensional and three-dimensional projections of the OTUs onto the first three principal components were made; three-dimensional projections were drawn using a Benson-Lehner incremental plotter (Rohlf, 1968). Discussions of the theory underlying these tests were given by Sokal and Sneath (1963), Schnell (1970:42-44), and Atchley (1970:206-212). Choate (1970), Rising (1970), and Genoways and Jones (1971) have used these techniques in studies similar to mine. Discriminant function analyses were performed using the MULDIS subroutine of the NT-SYS system. This program used variance-covariance mathematics to 12 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY differentially weight characters relative to their within-group and between-groups variation. Only two reference samples were used for all discriminant analyses in this paper; these two reference samples were either samples of two species in which I was studying character differences or samples geographically removed from zones suspected of hybridization or intergradation with which a sample from the intermediate geographical area was compared. A discriminant multiplier was calculated using the reference samples for each character, and this was multi¬ plied by the value of its respective character; all such values were summed for each individual to yield its discriminant score. Where hybridization or intergra¬ dation was suspected, the discriminant scores were plotted on a frequency histo¬ gram to compare individuals of the two reference samples and to compare the test sample from the intermediate geographical area. A good discussion of dis¬ criminant functions was given by Jolicoeur (1959). Lawrence and Bossert (1969), Birney (1970), and Genoways and Choate (1972) have used this test in study¬ ing other mammalian groups. Ratio diagrams were prepared using a program written for Olivetti Under¬ wood Programma 101 by Sydney Anderson. Ratio diagrams are graphic methods of analysis first used by Simpson (1941) and discussed later by Simpson et al. (1960). Musser (1969a, 19696, 1970) has used this technique in several recent papers dealing with systematic problems of mammals. Other less frequently used tests such as Student’s /-test and Wilcoxon two-sample test are based upon pro¬ cedures outlined by Sokal and Rohlf (1969); all statistical terminology is also based upon that used by Sokal and Rohlf. In my analysis of evolutionary relationships, a Prim Network was computed. The Prim Network (Prim, 1957) is an expression of the phenetic relationships of the species based upon all characters considered (characters not weighted). The patristic difference (the length of the line connecting two species and the corre¬ sponding numerical value) represents the proportional overall difference that separates the species. Angles between cladistic events are arbitrary. This network also provides a framework for an inference of primitive character states and an¬ cestral forms. I had planned to also produce a Wagner Diagram, but this was not done because a satisfactory method could not be found for weighting the many binary and coded characters used in my analyses that do not vary within species. Theories underlying these techniques and their application are discussed in de¬ tail by Ferris (1966, 1967, 1970), Kluge (1969), and Kluge and Ferris (1969); Lawlor (1971) has recently used these analyses in studying members of the genus Peromyscus from the islands in the Gulf of California. My analysis was carried out using a computer program supplied by J. S. Ferris to S. R. Edwards. I am deeply indebted to the following institutions and curators who made material in their care available to me for study. Abbreviations preceding the names of institutions are used in the accounts beyond to identify the source of specimens. AMNH — American Museum of Natural History, New York City (Richard G. Van Gelder and Sydney Anderson) ANSP — Academy of Natural Sciences of Philadelphia, Philadelphia (Robert R. Grant, Jr.) GENOWAYS— SYSTEMATICS OF LIOMYS 13 BMNH — British Museum (Natural History), London (J. E. Hill) CAS — California Academy of Sciences, San Francisco (Robert T. Orr) ENCB — Escuela Nacional de Ciencias Biologicas, Mexico, D. F. (Ticul Alvarez) FMNH — Field Museum of Natural History, Chicago (Joseph C. Moore) KU — Museum of Natural History, The University of Kansas, Lawrence (J. Knox Jones, Jr.) LACM — Los Angeles County Museum, Los Angeles (Donald R. Patten) LSU — Museum of Natural Science, Louisiana State University, Baton Rouge (George H. Lowery, Jr.) MCZ — Museum of Comparative Zoology, Harvard University, Cambridge (Barbara Law¬ rence) MSU — The Museum, Michigan State University, East Lansing (Rollin H. Baker) MVZ — Museum of Vertebrate Zoology, The University of California, Berkeley (William Z. Lidicker, Jr., and Seth B. Benson) MWU — Department of Biology, Midwestern University, Wichita Falls, Texas (Walter W. Dalquest) ROM — Royal Ontario Museum, Toronto (Randolph L. Peterson and James R. Tamsitt) SDNHM — San Diego Natural History Museum, San Diego (Joseph R. Jehl, Jr.) TCWC — Texas Cooperative Wildlife Collection, Texas A&M University, College Station (Dilford C. Carter and William B. Davis) TNHC — Texas Natural History Collection, The University of Texas, Austin (W. W. New¬ comb) TTU — Texas Tech University, Lubbock (Robert L. Packard and Robert J. Baker) UA — University of Arizona, Tucson (E. Lendell Cockrum) UCLA — University of California, Los Angeles, including the Donald R. Dickey Collection (Thomas R. Howell) UMMZ — Museum of Zoology, The University of Michigan, Ann Arbor (Emmet T. Hooper and William H. Burt) UNAM — Instituto de Biologia, Universidad Nacional Autonoma de Mexico, Mexico, D. F. (Bernardo Villa-R.) UNM — Museum of Southwestern Biology, University of New Mexico, Albuquerque (James S. Findley) USNM — United States National Museum, including the Biological Surveys Collection, Washington, D. C. (Charles O. Handley, Jr., Henry W. Setzer, and John L. Paradiso) WBC — Wayland Baptist College, Plainview, Texas (J. Hoyt Bowers) YPM — The Peabody Museum of Natural History, Yale University, New Haven (Charles G. Sibley) I would like to thank the following agencies for grants-in-aid of travel to the major mammalogical collections in North America: the Committee on Systematic and Evolutionary Biology (National Science Foundation Program GB-4446X administered by Dr. George W. Byers) at The University of Kansas; the Com¬ mittee on Grants-in-aid of Research, The Society of the Sigma Xi; the Watkins Museum of Natural History Fund, Museum of Natural History, The University of Kansas; and The Graduate School, The University of Kansas. Much of the material of Liomys and Heteromys deposited in the Museum of Natural History at Kansas was collected under the aegis of a contract (DA-49-1 93-MD-22 15) to Dr. J. Knox Jones, Jr., from the U.S. Army Medical Research and Development Command. Parts of my graduate training were supported by an 1 1 -month train¬ eeship from the Committee on Systematic and Evolutionary Biology (National Science Foundation Program GB-4446X1 administered by Dr. J. Knox Jones, Jr.), at The University of Kansas and a research assistantship supported by a contract (DA-49-193-MD-2215) to Dr. J. Knox Jones, Jr., from the U.S. Army Medical 14 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY Research and Development Command. Photographic and other supplies were purchased with funds supplied by Research Awards Committee of the Kansas Academy of Science and the Division of Biological Sciences, The University of Kansas. Computer time allotted to me by the Division of Biological Sciences enabled use of the GE 635 computer at The University of Kansas. I am especially grateful to Dr. J. Knox Jones, Jr., who supervised my research and critically reviewed this manuscript. Dr. George W. Byers and Dr. Sydney Anderson also reviewed the manuscript and made many helpful suggestions. T. H. Swearingen and B. Siler prepared the cranial and dental drawings. Some drawings and measurements of special structures were made with the aid of a Wild Heerbrugg Stereomicroscope fitted with a drawing tube, kindly made avail¬ able to me by Dr. F. B. Cross. Dr. Richard F. Johnston placed at my disposal a Bausch and Lomb Spectronic 505 recording spectrophotometer. The section on karyology could not have been completed without the help of Dr. Robert J. Baker, who gave freely his supplies, equipment, and time, and Dr. James L. Pat¬ ton and Dr. Alfred L. Gardner. Mr. Eustorgio Mendez of the Gorgas Memorial Laboratory supplied live specimens of Liomys adspersus and Heteromys desma- restianus from Panama. I would like to extend my gratitude to the many people who collected material used in this report with particular thanks to Percy L. Clif¬ ton and J. R. Alcorn, who made large and significant collections for The Uni¬ versity of Kansas. Dr. Craig C. Black made available the holotype of Liomys centralis and tendered his opinion on the status of that fossil species. My early graduate training was done under the guidance of Dr. E. R. Hall. Several of my colleagues at The University of Kansas, especially Drs. Elmer C. Birney, Jerry R. Choate, Carleton J. Phillips, and James D. Smith were helpful in giving technical assistance and willingly engaged in stimulating discussions pertinent to my research. Special thanks are extended to the following specialists who identified ecto¬ parasites of Liomys and Heteromys and checked the lists of ectoparasites in Ap¬ pendix II: Dr. K. C. Emerson (Anoplura), Dr. J. Kethley (Cheyletidae), Dr. C. E. Yunker (Dermanyssidae), Drs. G. M. Kohls and E. K. Jones (Ixodides), Dr. R. W. Strandtmann (Laelapidae), Dr. B. McDaniel (Listrophoridae), Dr. R. Traub (Siphonaptera), and Dr. R. B. Loomis (Trombiculidae). Finally, I especially acknowledge my wife, Joyce, who prepared many of the illustrations, typed the drafts of this manuscript, and provided the moral support necessary to see this project through to its completion. The systematic accounts are introduced by remarks on the distribution, diag¬ nosis, comparisons with other species, and general ecology of the species without reference to infraspecific variation. These sections are followed by the analysis of geographic variation, both univariate and multivariate. For the statistical analyses, it was necessary to group specimens from several localities in order to have large enough samples for testing. In grouping localities, I attempted to keep the included geographic area as small as possible, not to cross any major phys¬ iographic boundaries, and not to cross previously recognized taxonomic bound¬ aries. The localities grouped for each sample are listed in the appropriate species GENOWAYS— SYSTEM ATICS OF LIOMYS 15 account in an abbreviated form; precise localities can be found in lists of speci¬ mens examined. Following the analyses of geographic variation are the taxonomic conclusions based on these analyses. The synonymies in the subspecific accounts are arranged with the original description first, followed by the first use of the presently employed name combi¬ nation and, finally, any junior synonyms. Although the International Code of Zoological Nomenclature (Article 51b) prohibits the use of a comma between a scientific name and the name of a user subsequent to the original author, I have done so in my synonymies because it has been, and still is, a common practice in the mammalogical literature. Following the synonymies are a brief description of the holotype of the subspecies, distribution of the subspecies, and comparisons with other subspecies within the species. The remarks section contains comments on relationship and intergradation with other subspecies of the species, on rela¬ tionship with sympatric species, and on infrasubspecies variation. In the lists of specimens examined the countries and states are listed alphabetically, and within each political unit the towns or physiographic features are listed alphabetically. If more than one locality is listed with reference to a particular place-name, the localities are arranged with the most northwesterly first and the most south¬ easterly last. A gazetteer is provided in Appendix I. Specimens examined are followed by a list of additional records from the literature and a list of marginal records. The marginal records are plotted on the distribution maps for the appro¬ priate species (excepting those in italics, which have not been plotted because undue crowding of symbols would have resulted). The first marginal record listed is the northernmost for the subspecies, and subsequent localities are listed in a clockwise manner from this starting point in the fashion employed by Hall and Kelson (1959). At this point it seems appropriate to define my ideas about two concepts — those of the species and the subspecies — used extensively in this paper. The “bio¬ logical” species concept has been followed insofar as it was possible to do so. This concept may be simply stated that species are “groups of actually or poten¬ tially interbreeding natural populations, which are reproductively isolated from other such groups” (Mayr, 1966). In the genus Liomys, I have found no diffi¬ culty in applying this concept because no individual that could be identified as a hybrid between species was found. My concept of a mammalian subspecies was most succinctly stated by Lidicker (1962:169, see also 1960:161-63) as follows: “A subspecies is a relatively homogeneous and genetically distinct portion of a species which represents a separately evolving, or recently evolved, lineage with its own evolutionary tendencies, inhabits a definite geographical area, is usually at least partially isolated, and may intergrade gradually, although over a fairly narrow zone, with adjacent subspecies.” This is an appealing concept because it views a subspecies as a population that has made initial steps toward speciation (although most do not complete the process) and applies an evolutionary philoso¬ phy to infraspecific variation rather than viewing it solely as geographic variation. 16 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY NONGEOGRAPHIC VARIATION Four types of nongeographic variation — variation with age, secondary sexual variation, individual variation, and seasonal and maturational molts — are dis¬ cussed in the following section. Appreciation of these types of variation is neces¬ sary in studies of the kind presented herein, because they must be taken into account and their effects thus minimized or eliminated when geographic variation is considered. Variation with Age When variation with age is present in a taxon, it is necessary in studies of geo¬ graphic variation of the taxon to identify those age categories that represent “adults” in that the individuals have stopped or nearly stopped growing. Speci¬ mens from selected populations of three species of Liomys ( irroratus , pictus, and salvini) were assigned to one of five age categories and external and cranial meas¬ urements were recorded for all individuals in order to determine what mensural variation with age occurred in these populations. The age categories are defined below, listed in sequence of increasing age. Numbers given after dental characters refer to toothrows of Liomys salvini shown in Fig. 3, which illustrate the char¬ acters, and toothrows illustrated by Reeder (1953:62) for Liomys pictus. I — M3 not fully erupted; deciduous premolars present (I); braincase domed; rostrum proportionally short; cranial bones only loosely articulated; juvenile pelage. II — M3 fully erupted; deciduous premolars present (II); braincase becoming flattened; cranial bones firmly articulated but all sutures plainly evident; juvenile or adult pelage (many individuals molting from juvenile pelage). III — Permanent premolar present but little worn; Ml revealing little evi¬ dence of wear, the median valley still completely separating the anterior and posterior lophs (III and 1-3 of Reeder); most sutures still evident; presphenoid- basioccipital suture usually open, but closed in a few individuals; temporal ridges only weakly defined especially in the parietal region; adult pelage. IV — Ml heavily worn so that only a dentine lake surrounded by a rim of enamel remaining (some individuals of L. salvini were judged to be adults when Ml evinced enough wear so that the anterior and posterior lophs joined at least at one end, if the other criteria were met) (IV and 4-5 of Reeder); presphenoid- basioccipital suture closed; sutures between nasals and frontals becoming ob¬ scured as are those between interparietal and parietals; temporal ridges well defined, even in the parietal region. V — All molars so heavily worn that no enamel pattern remains on their oc¬ clusal surface, excepting possibly a small enamel island on M3 (V and 7-10 of Reeder). Each of the 1 3 external and cranial characters were tested separately for males and females with single classification anova to determine if any of the means of the age categories were significantly different at P<. 05. If the means were found to be significantly different, the simultaneous sums of squares testing procedure GENOWAYS— SYSTEMATICS OF LIOMYS 17 i n m i z i Fig. 3. — Left maxillary toothrows of Liomys salvini illustrating wear patterns for the five age categories. See text for description of age categories. was used to find maximally nonsignificant subsets. The results of these tests are discussed below and shown in Table 1. Liomys irroratus. — Length of maxillary toothrow, depth of braincase, inter¬ parietal width, and interparietal length for both males and females were the only four characters tested that were found to have no significant difference between the means of all age categories. Nonoverlapping subsets of I, II-III, and IV-V were exhibited by four characters for both sexes (total length, length of tail, length of nasals, and length of rostrum for males and total length, greatest length of skull, length of nasals, and length of rostrum for females) of the remaining nine. Age category I formed a subset that differed significantly from the other four categories in length of hind foot of females. In greatest length of skull for males, nonoverlapping subsets of I-II, III, and IV-V were formed and subsets I, II-III-IV, and V were formed by age categories for interorbital constriction of females. The remaining characters (length of hind foot, zygomatic breadth, inter¬ orbital constriction, and mastoid breadth for males and length of tail, zygomatic breadth, and mastoid breadth for females) displayed more complex patterns of overlapping subsets, but in all for the females, and one for the males (mastoid breadth) category I formed a subset distinct from the other four categories. In only one character — interorbital breadth of females — of all those tested was a significant difference found between categories IV and V. In only four instances (length of hind foot of females and depth of braincase, interparietal width, and interparietal length of males) were the means of groups IV and V not found to be the largest. Group I was found to have the smallest mean in all characters except interparietal width of males and interparietal length of females, and in both the . analysis of variance was nonsignificant. Liomys p ictus. — Males were found to have no significant difference between the means of the age categories in two characters (interparietal width and inter- 18 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY parietal length) and females in only one character (length of maxillary tocthrow). Nonoverlapping subsets of I, II-III, and IV-V were found in five characters for males (total length, length of tail, greatest length of skull, length of nasals, and length of rostrum) and two characters for females (greatest length of skull and length of rostrum). Two other patterns of nonoverlapping subsets were illustrated in this species, I-II-III and IV-V for zygomatic breadth of males and interorbital constriction of females and I-II, III, and IV-V for zygomatic breadth of females. The remaining characters (length of hind foot, interorbital constriction, mastoid breadth, length of maxillary toothrow, and depth of braincase for males and total length, length of tail, length of hind foot, zygomatic breadth, length of nasals, depth of braincase, interparietal width, and interparietal length for females) showed complex patterns of overlapping subsets, but in two of the characters for females (length of nasals and interparietal width) category I was separate from the other four. In no case was category IV found to be significantly different from V. Either category IV or V had the largest mean for all characters studied and for all measurements category I had the smallest mean. Liomys salvini. — In one measurement for males (length of maxillary toothrow) and three for females (length of maxillary toothrow, interparietal width, and interparietal length) no significant difference was found between the various age categories. Nonoverlapping subsets of I, II, and III-IV-V were shown in total length for females. Greatest length of skull and zygomatic breadth for males have subsets of I, II-III, and IV-V and age categories for depth of braincase for males were divided into two subsets (I-II-III and IV-V). Two measurements for males (length of nasals and length of rostrum) and three for females (greatest length of skull, length of nasals, and length of rostrum) have nonoverlapping subsets of I, II, III, and IV-V. However, the majority of measurements for Liomys salvini show a complex pattern of overlapping subsets. Seven measurements for males (total length, length of tail, length of hind foot, interorbital constriction, mastoid breadth, interparietal width, and interparietal length) and six for females (length of tail, length of hind foot, zygomatic breadth, interorbital constriction, mastoid breadth, and depth of braincase) exhibit overlapping subsets. In three characters with overlapping subsets for both males (total length, length of tail, and mastoid breadth) and females (length of tail, zygomatic breadth, and mastoid breadth), category I formed a subset that was distinct from the remaining categories. In no measurement was age category IV found to be significantly different from V. In all cases either IV or V had the largest mean and category I had the smallest mean. Conclusions. — In the three species studied, age categories IV and V represent the largest individuals and individuals of the two groups are inseparable based on size. Specimens falling into these two categories have been used as “adults” throughout the remainder of my study and form the basis for the analysis in inter¬ specific and intraspecific variation. For Liomys irroratus and L. pictus, the three remaining categories appear to fall into two groups — I and II-III. For Liomys salvini the situation is not as clear, but it would appear that the three remaining categories can best be considered as distinct from each other. GENOWAYS— SYSTEMATICS OF LIOMYS 19 Table 1. — Variation with age in external and cranial measurements of three species of Liomys. The three species (in. the order that they appear in the table) are Liomys irroratus from central Jalisco, Liomys pictus from western Jalisco, and Liomys salvini from the de¬ partments of Carazo and Managua, Nicaragua. Statistics given are number, mean, two standard errors of mean, range, coefficient of variation, F, and Fs. Age classes for the males are listed first for each measurement followed by those for females; the age classes are listed in decreasing order with the largest mean first. Groups of means that were found to be significantly different at P <.05 were tested with the sums of squares-simultaneous testing procedure to find the nonsignificant subsets. Groups of means that were found to be not significantly different at PC. 05 are marked ns. Measurements, sex, and age Fs Results classes N Mean ±2 SE Range CV F SS-STP Liomys irroratus Total length Male V 4 251.5 ± 9.00 (240.0-262.0) 3.6 41.22 IV 18 235.1 ± 5.07 (216.0-253.0) 4.6 2.60 III 13 219.2 ± 5.01 (202.0-231.0) 4.1 II 7 205.1 ± 4.20 (195.0-210.0) 2.7 I 4 184.0 ± 8.16 (174.0-194.0) 4.4 Female V 8 228.6 ± 5.21 (215.0-237.0) 3.2 27.24 IV 26 225.3 ± 3.78 (207.0-251.0) 4.3 2.53 III 19 215.1 ± 3.75 (201.0-233.0) 3.8 II 6 207.8 ± 4.39 (198.0-212.0) 2.6 I 3 175.3 ± 16.71 (166.0-192.0) 8.3 Length of tail Male V 4 127.3 ± 8.22 (118.0-138.0) 6.5 30.89 IV 18 1 18.6 ± 3.26 (106.0-130.0) 5.8 2.60 III 13 1 11.2 ± 2.74 (100.0-118.0) 4.5 II 7 102.6 ± 3.51 (95.0-107.0) 4.5 I 4 88.5 ± 4.80 (84.0-94.0) 5.4 Female IV 26 1 12.6 ± 2.48 (102.0-131.0) 5.6 19.74 V 8 1 12.3 ± 2.58 (105.0-116.0) 3.3 2.53 III 19 107.1 ± 2.80 (97.0-1 18.0) 5.7 II 6 104.3 ± 3.21 (97.0-107.0) 3.8 I 3 82.3 ± 9.68 (77.0-92.0) 10.2 20 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY Table 1. — Continued. Length of hind foot Male V 4 29.9 1.03 (28.5-31.0) 3.5 4.13 III 14 29.4 0.35 (28.0-30.0) 2.2 2.60 IV 17 29.0 0.54 (26.0-30.5) 3.8 II 7 28.9 0.52 (28.0-30.0) 2.4 I 4 27.6 ± 0.75 (26.5-28.0) 2.7 Female II 6 28.8 ± 0.33 (28.0-29.0) 1.4 7.91 III 19 28.6 0.28 (27.0-29.5) 2.2 2.52 IV 28 28.2 ± 0.32 (27.0-30.0) 3.0 V 8 28.2 ± 0.53 (27.0-29.0) 2.7 I 3 Greatest length of skull 26.0 ± 2.08 (24.5-28.0) 6.9 Male V 3 33.2 ± 0.92 (32.6-34.1) 2.4 31.34 IV 18 31.9 0.51 (30.4-34.1) 3.4 2.60 III 14 30.7 ± 0.23 (29.9-31.3) 1.4 II 7 29.3 ± 0.74 (27.8-30.3) 3.4 I 4 27.8 ± 0.51 (27.1-28.3) 1.8 Female V 8 31.6 ± 0.51 (30.8-33.0) 2.3 54.08 IV 25 31.4 0.23 (30.3-32.6) 1.9 2.53 III 19 30.4 ± 0.29 (29.0-31.3) 2.1 II 6 29.7 ± 0.28 (29.2-30.1) 1.2 I 3 26.2 ± 1.30 (25.5-27.5) 4.3 Zygomatic breadth Male V 3 16.3 ± 0.24 (16.1-16.5) 1.3 20.39 IV 16 15.4 ± 0.29 (14.8-16.6) 3.8 2.62 III 14 14.8 ± 0.14 (14.3-15.1) 1.8 II 7 14.5 ± 0.26 (14.0-15.0) 2.4 I 3 13.7 ± 0.29 (13.5-14.0) 1.8 Female V 7 15.5 ± 0.21 (15.0-15.9) 1.8 17.79 IV 20 15.2 ± 0.21 (14.6-16.3) 3.1 2.56 II 6 14.8 ±_ 0.12 (14.6-15.0) 1.0 III 17 14.6 ± 0.17 (13.9-15.2) 2.4 I 3 13.6 ± 0.37 (13.4-14.0) 2.4 GENOWAYS— SYSTEMATICS OF LIOMYS Table 1. — Continued. Interorbital constriction Male V 4 8.3 ± 0.10 (8. 2-8. 4) 1.2 10.1 1 IV 18 7.9 ± 0.18 (7. 2-8. 7) 4.8 2.59 III 14 7.6 ± 0.86 (7. 3-7. 9) 2.1 II 7 7.6 ± 0.27 (7. 2-8. 2) 4.8 I 4 7.1 ± 0.20 (6.8-7. 2) 2.8 Female V 8 7.9 0.16 (7. 7-8. 3) 2.8 13.15 IV 27 7.6 ± 0.98 (7.2-8. 1) 3.4 2.52 III 19 7.5 ± 0.86 (7. 2-7.8) 2.5 II 6 7.5 ± 0.13 (7. 3-7.7) 2.1 I 3 6.9 ± 0.18 (6. 7-7.0) 2.2 Mastoid breadth Male V 3 14.7 ± 0.29 (14.5-15.0) 1.7 9.66 IV 18 14.5 ± 0.20 (13.8-15.4) 3.0 2.62 II 6 14.2 ± 0.22 (13.8-14.5) 1.9 III 14 14.1 ± 0.95 (13.8-14.4) 1.3 I 3 13.4 ± 0.46 (13.0-13.8) 3.0 Female V 8 14.4 ± 0.19 (13.8-14.6) 1.9 22.67 IV 25 14.3 ± 0.93 (13.8-14.7) 1.6 2.53 II 6 14.1 ± 0.99 (13.9-14.2) 0.9 III 19 14.0 ± 0.11 (13.7-14.5) 1.6 I 3 13.1 ± 0.41 (12.7-13.4) 2.7 Length of nasals Male V 4 12.9 ± 0.35 (12.5-13.3) 2.7 31.26 IV 18 12.4 ± 0.32 (11.4-13.4) 5.5 2.59 III 14 1 1.4 ± 0.23 (10.9-12.3) 3.8 II 7 10.8 0.34 (10.3-11.6) 4.1 I 4 9.6 ± 0.57 (8.8-10.0) 5.9 Female V 8 12.6 ± 0.39 (1 1.6-13.2) 4.4 33.00 IV 28 12.1 ± 0.22 (1 1.0-13.5) 4.9 2.52 III 19 11.5 ± 0.18 (10.5-12.4) 3.4 II 6 11.1 ± 0.33 (10.8-1 1.9) 3.7 I 3 9.1 ± 0.55 (8. 7-9. 6) 5.2 22 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY Table 1. — Continued. Length of rostrum Male V 4 14.7 ± 0.34 (14.5-15.2) 2.3 38.73 IV 18 14.1 0.30 (13.2-15.4) 4.5 2.59 III 14 13.1 0.16 (12.6-13.5) 2.3 II 7 12.5 ± 0.33 (11.8-13.0) 3.5 I 4 11.5 ± 0.38 (11.0-11.9) 3.3 Female V 8 14.0 ± 0.23 (13.7-14.7) 2.3 48.11 IV 28 13.7 ± 0.15 (13.0-14.9) 2.9 2.52 III 19 13.1 ± 0.19 (12.3-13.7) 3.1 II 6 12.7 ± 0.15 (12.4-12.9) 1.5 I 3 10.7 Length of maxillary toothrow ± 1.11 (10.0-11.8) 9.0 Male IV 18 5.1 0.08 (4.8-5. 4) 3.1 2.41 III 14 5.1 ± 0.07 (4.8-5. 3) 2.7 V 4 5.0 ± 0.14 (4.9-5. 2) 2.8 3.29 Female IV 28 5.1 ± 0.06 (4.8-5. 4) 3.3 0.58 V 8 5.0 ± 0.22 (4.4-5. 2) 6.0 3.18 III 19 5.0 Hh 0.08 (4. 8-5.4) 3.4 Depth of braincase Male III V 14 2 8.8 8.7 ± 0.16 (8. 4-9.4) (8.7) 3.4 1.89 2.63 IV 18 8.7 ± 0.13 (8. 2-9.2) 3.2 II 5 8.6 ± 0.20 (8. 4-9.0) 2.5 I 3 8.3 ± 0.31 (8. 1-8.6) 3.2 Female V 8 8.7 ± 0.09 (8. 5-8. 8) 1.4 1.90 IV 25 8.6 ± 0.08 (8. 2-9.0) 2.3 2.54 III 19 8.6 ± 0.06 (8. 3-8. 8) 1.6 II 6 8.6 ± 0.14 (8. 5-9.0) 2.0 I 2 8.4 ± 0.10 (8. 3-8.4) 0.8 Interparietal width Male III 14 8.8 ± 0.24 (7. 5-9.3) 5.1 0.51 IV 18 8.7 ± 0.21 (7. 9-9.4) 5.1 2.61 II 6 8.6 ± 0.35 (7. 9-9.0) 5.1 I 4 8.6 ± 0.26 (8. 2-8.8) 3.1 V 3 8.5 ± 0.44 (8. 2-8. 9) 4.5 GENOWAYS— SYSTEMATICS OF LIOMYS 23 Table 1. — Continued. Female V 8 8.7 ± 0.20 (8. 2-9.0) 3.3 0.77 111 19 8.5 ± 0.20 (7. 9-9.3) 5.0 2.54 11 6 8.5 0.35 (7. 9-9.0) 5.0 IV 26 8.4 ± 0.16 (7. 8-9.5) 4.8 I 3 8.3 ± 0.20 (8.2-8. 5) 2.1 Interparietal length Male III 14 3.7 ± 0.12 (3. 4-4.0) 5.8 1.38 V 3 3.6 ± 0.44 (3. 2-3. 9) 10.4 2.61 IV 18 3.6 ± 0.12 (3.0-4. 1) 6.8 II 6 3.5 ± 0.15 (3. 2-3. 7) 5.4 I 4 3.5 ± 0.29 (3. 2-3. 9) 8.1 Female V 8 3.7 ± 0.16 (3. 3-4.0) 6.0 0.53 III 19 3.7 ± 0.15 (3.0-4.3) 9.0 2.54 IV 26 3.6 ± 0.12 (2.9-4. 1) 8.2 I 3 3.6 ± 0.50 (3. 3-4.1) 12.1 II 6 3.5 ± 0.33 (3. 0-4.1) Liomys p ictus 11.6 Total length Male V 2 253.5 ± 9.00 (249.0-258.0) 2.5 32.61 IV 33 240.3 ± 4.21 (218.0-264.0) 5.0 2.56 III 16 224.5 ± 4.22 (208.0-239.0) 3.8 II 10 214.2 ± 5.56 (199.0-226.0) 4.1 I 4 185.0 ± 15.43 (165.0-201.0) 8.3 Female V 2 232.0 ± 18.00 (223.0-241.0) 5.5 15.29 IV 25 229.5 ± 3.87 (212.0-248.0) 4.2 2.56 III 19 216.6 ± 4.61 (192.0-230.0) 4.6 II 5 207.0 ± 8.85 (196.0-220.0) 4.8 I 2 183.3 ± 23.50 (171.5-195.0) 9.1 Length of tail Male V 2 131.5 ± 1.00 (131.0-132.0) 0.5 22.08 IV 33 123.0 ± 2.88 (105.0-138.0) 6.7 2.56 III 16 115.0 ± 2.85 (109.0-127.0) 5.0 II 10 108.7 ± 4.41 (98.0-117.0) 6.4 I 4 92.1 ± 7.58 (82.5-101.0) 8.2 24 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY Table 1. — Continued. Female V 2 121.5 ± 9.00 (117.0-126.0) 5.2 9.20 IV 25 118.2 ± 2.30 (108.0-128.0) 4.9 2.56 III 19 111.6 ± 4.06 (91.0-125.0) 7.9 II 5 104.4 ± 7.50 (94.0-116.0) 8.0 I 2 93.5 ± 9.00 (89.0- 98.0) 6.8 Length of hind foot Male V 4 29.1 ± 1.31 (28.0-31.0) 4.5 3.84 IV 36 28.8 ± 0.37 (26.0-31.0) 3.9 2.56 II 11 28.4 ± 0.47 (27.0-29.5) 2.7 III 17 27.9 ± 0.42 (26.0-29.0) 3.1 I 4 27.5 ± 1.35 (25.5-28.5) 4.9 Female IV 27 28.7 ± 0.45 (27.0-31.0) 4.1 8.02 V 7 27.9 ± 0.81 (26.0-29.0) 3.8 2.54 III 20 27.7 ± 4.00 (26.0-29.5) 3.2 II 5 26.7 ± 1.41 (25.0-28.0) 5.9 I 2 25.0 ± 2.00 (24.0-26.0) 5.7 Greatest length of skull Male V 4 32.7 ± 0.67 (32.1-33.4) 2.0 54.04 IV 35 32.2 ± 0.26 (30.8-34.1) 2.4 2.51 III 19 30.4 ± 0.37 (28.5-31.8) 2.6 II 10 29.5 ± 0.56 (28.3-30.8) 3.0 I 4 27.7 ± 0.97 (26.6-28.7) 3.5 Female V 6 31.6 ± 0.74 (30.2-32.8) 2.9 27.76 IV 28 31.6 ± 0.28 (30.1-33.0) 2.3 2.54 III 19 30.1 ± 0.41 (27.7-31.3) 2.9 II 5 29.6 ± 0.73 (28.5-30.6) 2.8 I 2 26.5 ± 2.00 (25.5-27.5) 5.3 Zygomatic breadth Male V 3 15.1 ± 0.53 (14.6-15.4) 3.1 15.61 IV 27 15.0 ± 0.15 (14.1-15.9) 2.6 2.54 III 19 14.4 ± 0.21 (13.7-15.3) 3.2 il 10 14.0 ± 0.23 (13.4-14.7) 2.6 I 2 13.7 ± 0.20 (13.6-13.8) 1.0 GENOWAYS— SYSTEMATICS OF LIOMYS Table I. — Continued. Female V 5 14.7 ± 0.36 (14.1-15.0) 2.7 19.08 IV 26 14.6 ± 0.13 (14.1-15.2) 2.3 2.56 III 17 14.2 ± 0.14 (13.5-14.8) 2.1 II 5 13.6 ± 0.33 (13.3-14.2) 2.7 I 2 Interorbital constriction 13.2 ± 0.30 (13.0-13.3) 1.6 Male IV 36 7.8 ± 0.11 (7. 1-8.7) 4.2 9.27 V 4 7.7 ± 0.54 (7 1-8.4) 7.0 2.50 III 20 7.5 0.11 (7. 1-8.0) 3.4 II 12 7.2 ± 0.24 (6.2-7. 8) 5.7 I 4 7.1 ± 0.08 (7. 0-7. 2) 1.1 Female V 7 7.9 ± 0.19 (7. 5-8.3) 3.2 1 1.95 IV 28 7.6 ± 0.12 (7. 1-8.6) 4.1 2.54 III 20 7.3 ± 0.09 (6. 9-7. 6) 2.7 II 5 7.3 ± 0.08 (7. 2-7. 4) 1.2 I 2 6.8 ± 0.10 (6.7-6. 8) 1.0 Mastoid breadth Male IV 35 14.3 ± 0.13 (13.6-15.0) 2.6 10.60 V 4 14.2 ± 0.37 (13.7-14.6) 2.6 2.50 III 20 14.1 ± 0.14 (13.3-14.7) 2.2 II 12 13.7 ± 0.21 (13.2-14.4) 2.7 I 4 13.4 ± 0.31 (13.0-13.7) 2.3 Female IV 28 14.2 ± 0.13 (13.6-14.8) 2.3 7.74 V 7 14.0 ± 0.17 (13.5-14.2) 1.6 2.54 III 20 13.9 ± 0.17 (13.0-14.5) 2.7 II 5 13.5 ± 0.58 (12.8-14.5) 4.8 I 2 13.0 ± 0.10 (12.9-13.0) 0.5 Length of nasals Male V 4 13.3 ± 0.48 (12.6-13.7) 3.6 52.62 IV 36 13.0 ± 0.19 (1 1.7-14.3) 4.5 2.51 III 20 1 1.9 ± 0.21 (10.6-12.6) 3.9 II 12 1 1.5 ± 0.17 (1 1.0-12.1) 2.6 I 4 10.2 ± 0.33 (9.7-10.4) 3.3 I 26 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY Table 1. — Continued. Female IV 28 12.7 ± 0.23 (11.1-13.9) 4.7 21.19 V 6 12.7 ± 0.53 (11.9-13.3) 5.1 2.54 III 19 11.8 ± 0.22 (11.0-12.8) 4.0 II 5 11.3 ± 0.64 (10.3-12.3) 6.3 I 2 9.5 ± 1.50 (8.7-10.2) 11.2 Length of rostrum Male V 4 14.3 ± 0.05 (14.3-14.4) 0.3 50.31 IV 34 14.2 ± 0.18 (13.1-15.5) 3.6 2.51 III 19 13.1 ± 0.25 (11.8-14.1) 4.1 II 11 12.5 ± 0.31 (11.8-13.5) 4.1 I 4 11.3 ± 0.51 (10.7-11.9) 4.5 Female V 6 13.9 ± 0.47 (13.0-14.6) 4.1 23.73 IV 26 13.8 ± 0.20 (13.0-14.8) 3.6 2.55 III 18 13.0 ± 0.22 (12.2-13.6) 3.7 II 4 12.7 ± 0.41 (12.1-13.1) 3.3 I 2 10.9 Length of maxillary toothrow ± 0.80 (10.5-11.3) 5.2 Male V 4 5.1 ± 0.18 (4.9-5. 3) 3.6 3.55 IV 32 4.9 ± 0.06 (4.6-5. 4) 3.5 3.18 III 19 4.9 ± 0.08 (4.5-5. 1) 3.7 Female V 7 4.9 ± 0.19 (4.4-5. 1) 5.2 2.16 IV 28 4.9 ± 0.07 (4.5-5. 2) 3.9 3.18 III 19 4.8 ± 0.08 (4.4-5. 2) 3.6 Depth of braincase Male V 4 8.5 ± 0.41 (8. 1-8.9) 4.8 6.09 IV 35 8.3 ± 0.07 (7. 8-8.7) 2.4 2.50 III 19 8.2 ± 0.10 (7. 8-8.7) 2.7 II 12 8.0 ± 0.11 (7. 7-8.4) 2.3 I 4 8.0 ± 0.32 (7. 5-8.2) 4.0 Female V 6 8.3 ± 0.26 (8. 0-8.8) 3.8 3.85 II 5 8.2 ± 0.16 (8. 0-8. 5) 2.2 2.55 IV 27 8.2 ± 0.09 (7. 7-8.6) 2.8 III 19 8.1 ± 0.86 (7. 9-8. 5) 2.3 I 2 7.6 ± 0.80 (7. 2-8.0) 7.4 ns GENOWAYS— SYSTEMATICS OF LIOMYS 27 Table 1. — Continued. Interparietal width Male V 4 9.0 ± 0.22 (8. 7-9.2) 2.5 1.64 IV 35 8.9 ± 0.14 (7. 9-9.7) 4.6 2.50 III 20 8.7 ± 0.16 (8. 2-9.4) 4.0 II 12 8.6 ± 0.20 (8.0-9. 1) 3.9 I 4 8.6 ± 0.44 (8. 1-9.1) 5.2 Female V 7 8.9 ± 0.39 (8.0-9.5) 5.8 3.61 IV 28 8.8 ± 0.17 (7. 9-9.5) 5.2 2.54 II 5 8.5 ± 0.53 (7. 8-9.2) 7.0 III 20 8.4 ± 0.22 (7. 6-9. 5) 5.9 I 2 8.1 ± 0.10 (8.0-8. 1) 0.9 Interparietal length Male V 4 4.7 ± 0.22 (4. 5-5.0) 4.7 1.97 IV 34 4.5 ± 0.10 (3. 8-5.2) 6.7 2.51 II 12 4.5 ± 0.19 (4. 0-5.0) 7.2 III 20 4.4 ± 0.11 (4. 0-5.0) 5.6 I 4 4.2 ± 0.50 (3. 5-4.7) 12.0 Female V 7 4.7 ± 0.26 (4. 3-5.1) 7.3 8.12 IV 28 4.4 ± 0.96 (4.0-4.9) 5.7 2.54 II 5 4.3 ± 0.29 (3. 8-4.7) 7.6 III 20 4.3 ± 0.15 (3. 7-5.0) 8.0 I 2 3.5 ± 0.10 (3.4-3. 5) Liomys salvini 2.0 Total length Male IV 9 225.8 ± 6.70 (213.0-240.0) 4.5 23.06 V 4 224.3 ± 10.44 (213.0-235.0) 4.7 2.65 III 14 204.6 ± 8.12 (178.0-230.0) 7.4 II 8 194.5 ± 9.69 (172.0-217.0) 7.0 I 4 153.8 ± 15.46 (132.0-167.0) 10.1 Female IV 1 1 210.0 ± 5.89 (196.0-227.0) 4.7 21.18 V 5 207.2 ± 0.40 (207.0-208.0) 0.2 2.56 III 26 203.1 ± 5.71 (182.0-241.0) 7.2 II 12 188.5 4.19 (178.0-200.0) 3.8 I 3 145.7 ± 18.52 (129.0-161.0) 11.0 28 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY Table 1. — Continued. Length of tail Male V 4 115.5 ± 3.32 (113.0-120.0) 2.9 13.99 IV 9 111.6 ± 5.65 (94.0-121.0) 7.6 2.65 III 14 104.0 ± 5.05 (84.0-123.0) 9.1 II 8 97.6 ± 6.41 (81.0-112.0) 9.3 I 4 75.3 ± 12.28 (58.0-87.0) 16.3 Female IV 11 108.5 ± 3.86 (100.0-121.0) 5.9 25.70 III 26 106.5 ± 3.14 (96.0-128.0) 7.5 2.56 V 5 104.8 ± 3.37 (101.0-111.0) 3.6 II 12 93.8 ± 2.73 (86.0-103.0) 5.0 I 3 71.0 ± 6.93 (65.0-77.0) 8.5 Length of hind foot Male IV 9 27.2 ± 1.04 (25.0-30.0) 5.7 4.19 V 5 26.6 ± 0.80 (25.0-27.0) 3.4 2.63 II 10 26.0 ± 0.73 (24.0-27.0) 4.4 III 14 25.9 ± 0.38 (25.0-28.0) 3.8 I 5 24.8 ± 0.40 (24.0-25.0) 1.8 Female IV 13 26.2 0.56 (25.0-29.0) 3.9 2.83 II 14 25.9 ± 0.75 (22.0-27.0) 5.4 2.52 V 8 25.6 ± 0.75 (24.0-27.0) 4.1 III 30 25.5 ± 0.53 (23.0-28.0) 5.7 I 4 23.8 ± 1.26 (22.0-25.0) 5.3 Greatest length of skull Male V 5 31.2 ± 0.54 (30.7-32.2) 1.9 58.52 IV 12 30.9 ± 0.51 (29.3-32.2) 2.9 2.61 III 14 29.2 ± 0.45 (27.5-30.1) 2.9 II 9 28.1 ± 0.54 (27.0-29.8) 2.9 I 5 24.8 ± 0.85 (23.5-26.0) 3.8 Female V 8 30.2 ± 0.52 (28.9-31.1) 2.4 39.74 IV 1 1 30.1 ± 0.53 (28.7-31.3) 2.9 2.52 III 31 28.9 ± 0.29 (27.6-30.7) 2.8 1 II 13 27.7 ± 0.53 (25.9-29.6) 3.4 I 2 23.4 ± 0.30 (23.2-23.5) 0.9 GENOWAYS— SYSTEMATICS OF LIOMYS 29 Table 1. — Continued. Zygomatic breadth Male IV 9 14.3 ± 0.44 (13.5-15.7) 4.7 20.22 V 4 14.4 ± 0.47 (14.0-15.0) 3.3 2.73 III 8 13.5 ± 0.25 (13.2-14.2) 2.6 II 7 13.2 ± 0.33 (12.5-13.8) 3.4 I 4 11.9 ± 0.34 (1 1.5-12.2) 2.9 Female V 5 14.1 ± 0.46 (13.6-14.8) 3.6 19.21 IV 10 13.8 ± 0.27 (13.0-14.5) 3.1 2.63 III 16 13.4 ± 0.14 (12.9-13.9) 2.0 II 6 13.1 ± 0.32 (12.6-13.7) 3.0 I 4 Interorbital constriction 12.1 ± 0.43 (11.5-12.5) 3.6 Male IV 13 6.7 ± 0.18 (6. 2-7. 4) 4.8 13.71 V 5 6.7 0.29 (6.2-7. 1) 4.9 2.59 III 14 6.5 ± 0.14 (6.0-6.9) 3.9 II 10 6.2 ± 0.11 (5. 8-6. 4) 2.8 I 5 5.9 ± 0.12 (5. 7-6.0) 2.2 Female V 8 6.5 ± 0.17 (6. 1-6.8) 3.7 5.56 IV 15 6.5 ± 0.71 (6. 2-6. 8) 2.1 2.51 III 31 6.4 0.08 (5. 9-7.0) 3.6 II 15 6.3 ± 0.11 (5. 9-6.7) 3.4 I 4 6.1 ± 0.06 (6.0-6. 1) 1.0 Mastoid breadth Male IV 12 13.5 ± 0.25 (12.9-13.9) 2.5 23.76 V 5 13.4 ± 0.41 (13.1-14.2) 3.4 2.60 III 14 13.0 ± 0.14 (12.5-13.3) 2.0 II 10 12.9 ± 0.19 (12.5-13.5) 2.4 I 5 11.9 ± 0.08 (1 1.8-12.0) 0.7 Female V 8 13.4 ± 0.28 (12.8-13.9) 2.9 12.37 IV 15 13.3 ± 0.23 (12.7-14.2) 3.3 2.51 III 31 13.1 ± 0.12 (12.5-13.9) 2.5 II 13 12.8 ± 0.21 (12.0-13.5) 2.9 I 4 12.0 ± 0.33 (11.6-12.4) 2.7 I 30 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY Table 1. — Continued. Length of nasals Male V 5 12.1 ± 0.42 (11.5-12.8) 3.9 50.50 IV 13 11.7 ± 0.24 (11.0-12.4) 3.7 2.60 III 14 10.9 ± 0.32 (9.7-11.8) 5.5 1 11 9 9.9 0.31 (9.3-10.8) 4.7 1 1 5 8.3 ± 0.62 (7. 7-9.2) 8.2 1 Female V 8 11.6 0.37 (1 1.0-12.2) 4.5 41.06 IV 13 11.4 0.29 (10.4-12.2) 4.6 2.52 III 31 10.7 ± 0.16 (9.8-11.7) 4.1 1 II 15 10.1 ± 0.21 (9.5-11.1) 4.1 1 I 2 8.0 ± 0.10 (7. 9-8.0) 0.9 1 Length of rostrum Male V 5 13.6 ± 0.46 (13.1-14.2) 3.8 59.97 IV 13 13.2 ± 0.27 (12.4-14.3) 3.7 2.63 III 12 12.2 ± 0.31 (11.2-13.0) 4.4 1 II 8 11.4 0.30 (10.8-12.1) 3.7 1 I 5 9.7 ± 0.50 (9.1-10.5) 5.8 1 Female V 8 12.9 ± 0.25 (12.4-13.3) 2.7 44.05 IV 11 12.7 ± 0.30 (11.9-13.5) 4.0 2.53 III 29 12.0 ± 0.14 (11.3-12.9) 3.2 1 II 13 11.4 0.25 (10.6-12.2) 3.9 1 I 2 9.3 ± 0.50 (9.0-9. 5) 3.8 1 Length of maxillary toothrow Male V 5 4.5 0.17 (4. 2-4. 7) 4.2 0.61 ns IV 12 4.5 ± 0.10 (4. 3-4. 8) 3.8 3.37 III 12 4.5 0.10 (4. 2-4. 8) 3.9 Female V 8 4.6 ± 0.11 (4. 4-4. 9) 3.5 0.25 ns IV 16 4.6 ± 0.11 (4. 2-5.0) 4.9 3.21 III 24 4.6 ± 0.06 (4. 2-4. 9) 3.5 Depth of braincase Male V 4 8.3 ± 0.27 (8. 1-8.7) 3.3 6.88 IV 12 8.2 ± 0.24 (7. 6-9.0) 4.9 2.62 III 13 7.8 ± 0.13 (7. 5-8.2) 3.1 II 10 7.8 ± 0.10 (7.5-8. 1) 2.1 I 5 7.7 ± 0.14 (7. 5-7. 9) 2.1 GENOWAYS— SYSTEMATICS OF LIOMYS 31 Table 1. — Continued. Female V 8 8.2 ± 0.20 (7. 8-8.5) 3.4 5.19 IV 13 8.1 ± 0.24 (7. 5-9.2) 5.4 2.52 III 30 8.0 ± 0.11 (7. 4-8. 6) 3.6 II 12 7.8 ± 0.17 (7. 3-8. 5) 3.8 I 4 7.4 ± 0.17 (7. 2-7. 6) 2.3 Interparietal width Male IV 12 8.6 ± 0.27 (7. 9-9.5) 5.4 3.06 V 5 8.5 ± 0.55 (7. 7-9.1) 7.2 2.60 III 14 8.3 ± 0.18 (7. 9-9.0) 4.1 II 10 8.3 ± 0.27 (7. 7-9.2) 5.1 I 5 7.9 ± 0.20 (7. 6-8.2) 2.9 Female V 8 8.8 ± 0.21 (8. 3-9.1) 3.4 1.26 IV 14 8.7 ± 0.30 (7. 6-9.8) 6.4 2.52 II 13 8.6 0.31 (7. 6-9.2) 6.5 III 30 8.4 ± 0.18 (7. 2-9.4) 5.8 I 4 8.4 ± 0.22 (8. 1-8.6) 2.6 Interparietal length Male IV 12 4.0 ± 0.15 (3. 6-4.4) 6.7 4.95 V 5 3.9 0.24 (3. 6-4.2) 6.9 2.60 III 14 3.6 ± 0.20 (3.0-4.5) 10.3 II 10 3.6 ± 0.22 (3. 0-4.0) 9.8 I 5 3.3 ± 0.28 (3. 0-3. 7) 9.4 Female V 8 3.9 ± 0.17 (3. 5-4.3) 6.2 1.51 II 13 3.9 ± 0.19 (3. 3-4.4) 8.8 2.52 IV 14 3.8 ± 0.21 (3. 4-4. 7) 10.1 III 30 3.7 ± 0.12 (3.0-4.4) 8.4 I 4 3.6 ± 0.17 (3.4-3. 8) 4.8 ns ns Secondary Sexual Variation Adult males (age categories IV and V) of Liomys irroratus, L. p ictus, L. sal- vini, and L. adspersus were tested against adult females of the corresponding species using single classification anova to learn if the sexes were significantly different in size. The results of these tests are discussed below and shown in Table 2. Liomys irroratus. — Males were found to be significantly larger than females in seven (total length, length of tail, length of hind foot, greatest length of skull, interorbital constriction, mastoid breadth, and length of rostrum) of the 1 3 meas- 32 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY urements tested. In four (zygomatic breadth, length of nasals, depth of braincase, and interparietal width) of the remaining measurements, the means for males were larger than those for females; means for length of maxillary toothrow and interparietal length were the same for the two sexes. Liomys pictus. — Males were found to be significantly larger than females in six (total length, length of tail, greatest length of skull, zygomatic breadth, length of nasals, and length of rostrum) of the 1 3 measurements tested. Males had larger means in four other measurements (length of hind foot, interorbital constriction, mastoid breadth, and depth of braincase), with nonsignificant results, and the means for males and females were identical in three measurements (length of maxillary toothrow, interparietal width, and interparietal length). Liomys salvini. — Males were found to be significantly larger than females in six (total length, length of hind foot, greatest length of skull, zygomatic breadth, interorbital breadth, and length of rostrum) of the 13 measurements tested. Males had the larger mean in five measurements (length of tail, mastoid breadth, length of nasals, depth of braincase, and interparietal length) in which the means were not significantly different; in the remaining two measurements (length of maxil¬ lary toothrow and interparietal length), females had the larger mean. Liomys adspersus. — Males were found to be significantly larger than females in seven (total length, length of tail, greatest length of skull, zygomatic breadth, length of nasals, length of rostrum, and depth of braincase) of the 1 3 measure¬ ments tested. For three of the measurements (length of hind foot, length of maxil¬ lary toothrow, and interparietal width) with nonsignificant results, males had the largest mean, and for two (mastoid breadth and interparietal length) males and females had the same means. Females have, on the average, a broader interorbital constriction. Conclusions. — Males were found to be significantly larger than females in approximately half of the measurements tested and in most of the others where the means were not significantly different, males averaged larger than females. Based upon this information all data dealing with size has been treated separately for the sexes in all of the following analyses. The finding of significant secondary sexual variation in size in Liomys is of interest because no such significant variation has been found (or at least it was not thought to be present) in most measurements of other heteromyid genera studied — Dipodomys (Hall and Dale, 1939:49; Hall, 1946:403-433; Setzer, 1949:478; Genoways and Jones, 1971:268), Perognathus (Hall, 1946:357-377; Glass, 1947:175; Jones, 1953:518; Baker, 1954:342), and Microdipodops (Hall, 1941:241). Lidicker (1960:142-155), however, did find considerable secondary sexual variation in specimens of Dipodomys merriami as did Schmidly (1971: 110-111) in specimens of Dipodomys ordii from north Texas. In many instances, the one characteristic in which males and females were found to differ most was weight; Genoways and Jones (1971:268) found that males and females of Dipo¬ domys phillipsii were significantly different in total length and length of tail, but not in other measurements. The extent and a possible explanation of the signifi- GENOWAYS— SYSTEMATICS OF LIOMYS 33 Table 2. — Secondary sexual variation in external and cranial measurements in four species o/Liomys. The four species (in the order that they appear in the table) are Liomys irroratus from central Jalisco, Liomys pictus from western Jalisco, Liomys salvini from the depart¬ ments of Carazo and Managua, Nicaragua, and Liomys adspersus from Guanico, Los San¬ tos, Panama. Statistics given are number, mean, two standard errors of the mean, range, coefficient of variation, F, and Fs. Means for males and females that are significantly different at P<. 05 are marked with an asterisk', those that are not significantly different are marked ns. Measurements and sex N Mean ±2 SE Range CV Fs F Liomys irroratus Total length Male 22 238.0 5.18 (216.0-262.0) 5.1 17.52 * Female 34 226.1 ± 3.14 (207.0-251.0) 4.0 4.03 Length of tail Male 22 120.1 ± 3.31 (106.0-138.0) 6.5 17.69 * Female 34 112.5 ± 1.98 (102.0-131.0) 5.1 4.03 Length of hind foot Male 21 29.2 ± 0.49 (26.0-31.0) 3.8 14.27 * Female 36 28.2 ± 0.27 (27.0-30.0) 2.9 4.02 Greatest length of skull Male 21 32.1 ± 0.49 (30.4-34.1) 3.5 8.21 * Female 33 31.4 ± 0.21 (30.3-33.0) 2.0 4.03 Zygomatic breadth Male 19 15.6 ± 0.29 (14.8-16.6) 4.0 3.10 ns Female 27 15.3 ± 0.17 (14.6-16.3) 2.9 4.06 Interorbital constriction Male 22 8.0 ± 0.16 (7. 2-8. 7) 4.7 10.10 * Female 35 7.7 ± 0.09 (7.2-8. 3) 3.6 4.02 Mastoid breadth Male 21 14.5 ± 0.18 (13.8-15.4) 2.8 4.69 * Female 33 14.3 ± 0.08 (13.8-14.7) 1.7 4.03 Length of nasals Male 22 12.5 ± 0.28 (11.4-13.4) 5.3 2.10 ns Female 36 12.2 ± 0.20 (11.0-13.5) 5.0 4.02 34 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY Table 2. — Continued. Length of rostrum Male 22 14.2 ± 0.27 (13.2-15.4) 4.5 7.36 * Female 36 13.8 Length of maxillary toothrow ± 0.13 (13.0-14.9) 2.9 4.02 Male 22 5.1 ± 0.07 (4.8-5. 4) 3.2 1.25 ns Female 36 5.1 ± 0.07 (4. 4-5. 4) 4.0 4.02 Depth of braincase Male 20 8.7 ± 0.12 (8. 2-9. 2) 3.0 0.16 ns Female 33 8.6 ± 0.06 (8. 2-9.0) 2.1 4.03 Interparietal width Male 21 8.7 ± 0.19 (7. 9-9.4) 5.0 2.58 ns Female 34 8.5 ± 0.13 (7. 8-9. 5) 4.6 4.03 Interparietal length Male 21 3.6 ± 0.11 (3. 0-4.1) 7.1 0.02 ns Female 34 3.6 ± 0.10 (2. 9-4.1) Liomys p ictus 7.7 4.03 Total length Male 35 241.0 ± 4.12 (218.0-264.0) 5.1 15.80 * Female 27 229.7 ± 3.72 (212.0-248.0) 4.2 4.00 Length of tail Male 35 123.5 ± 2.80 (105.0-138.0) 6.7 7.27 * Female 27 118.5 ± 2.21 (108.0-128.0) 4.8 4.00 Length of hind foot Male 40 28.9 ± 0.35 (26.0-31.0) 3.9 1.95 ns Female 34 Greatest length of skull 28.5 ± 0.41 (26.0-31.0) 4.2 3.98 Male 39 32.3 ± 0.24 (30.8-34.1) 2.3 15.76 * Female 34 31.6 ± 0.26 (30.1-33.0) 2.4 3.98 Zygomatic breadth Male 30 15.0 ± 0.14 (14.1-15.9) 2.6 14.67 * Female 31 Interorbital constriction 14.6 ± 0.12 (14.1-15.2) 2.4 4.00 Male 40 7.8 ± 0.11 (7. 1-8. 7) 4.4 2.42 ns Female 35 7.6 ± 0.11 (7.1-8. 6) 4.2 3.98 GENOWAYS— SYSTEMATICS OF LIOMYS 35 Table 2. — Continued. Mastoid breadth Male 39 14.3 ± 0.12 (13.6-15.0) 2.6 2.83 ns Female 35 14.1 ± 0.1 1 (13.5-14.8) 2.3 3.98 Length of nasals Male 40 13.1 ± 0.18 (11.7-14.3) 4.4 7.86 * Female 34 12.7 ± 0.21 (11.1-13.9) 4.7 3.98 Length of rostrum Male 38 14.2 ± 0.16 (13.1-15.5) 3.4 11.81 * Female 32 13.8 Length of maxillary toothrow ± 0.18 (13.0-14.8) 3.7 3.99 Male 36 4.9 ± 0.06 (4.6-5. 4) 3.6 2.63 ns Female 35 4.9 ± 0.07 (4.4-5. 2) 4.1 3.99 Depth of braincase Male 39 8.3 ± 0.07 (7. 8-8.9) 2.7 2.66 ns Female 33 8.2 ± 0.08 (7. 7-8. 8) 3.0 3.98 Interparietal width Male 39 8.9 0.13 (7. 9-9.7) 4.5 0.01 ns Female 35 8.9 ± 0.16 (7. 9-9.5) 5.2 3.98 Interparietal length Male 38 4.5 ± 0.10 (3. 8-5.2) 6.7 0.02 ns Female 35 4.5 ± 0.10 (4.0-5. 1) Liomys salvini 6.4 3.98 Total length Male 12 224.8 ± 5.79 (213.0-240.0) 4.5 21.08 * Female 16 209.1 ± 4.05 (196.0-227.0) 3.9 4.21 Length of tail Male 12 1 12.2 4.24 (94.0-121.0) 6.5 3.80 ns Female 16 107.3 ± 2.92 (100.0-121.0) 5.4 4.22 Length of hind foot Male 13 27.0 ± 0.78 (25.0-30.0) 5.2 5.59 * Female 21 Greatest length of skull 26.0 ± 0.46 (24.0-29.0) 4.0 4.14 Male 16 30.9 ± 0.38 (29.3-32.2) 2.5 8.20 * Female 19 30.2 ± 0.37 (28.7-31.3) 2.7 4.13 36 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY Table 2. — Continued. Zygomatic breadth Male 12 14.4 ± 0.35 (13.5-15.7) 4.2 5.82 * Female 15 13.9 0.24 (13.0-14.8) 3.4 4.22 Interorbital breadth Male 17 6.7 0.16 (6. 2-7. 4) 4.9 5.71 * Female 23 6.5 ± 0.07 (6.1-6.8) 2.7 4.09 Mastoid breadth Male 16 13.5 ± 0.19 (12.9-14.2) 2.8 0.85 ns Female 23 13.3 ± 0.17 (12.7-14.2) 3.1 4.11 Length of nasals Male 17 11.8 ± 0.20 (1 1.0-12.4) 3.6 2.84 ns Female 21 11.5 ± 0.23 (10.4-12.2) 4.5 4.11 Length of rostrum Male 17 13.3 ± 0.23 (12.4-14.3) 3.6 11.15 * Female 19 12.8 Length of maxillary toothrow ± 0.21 (11.9-13.5) 3.5 4.12 Male 16 4.5 ± 0.09 (4.2-4. 8) 3.8 1.40 ns Female 24 4.6 ± 0.08 (4.2-5. 0) 4.4 4.10 Depth of braincase Male 15 8.3 ± 0.20 (7. 6-9.0) 4.6 1.44 ns Female 21 8.1 ± 0.17 (7. 5-9.2) 4.7 4.13 Interparietal width Male 16 8.6 ± 0.25 (7. 7-9.5) 5.9 0.86 ns Female 22 8.7 ± 0.20 (7.6-9. 8) 5.4 4.11 Interparietal length Male 16 3.9 ± 0.13 (3. 6-4.4) 6.7 0.82 ns Female 22 3.8 ± 0.14 (3. 4-4. 7) Liomys adspersus 8.8 4.11 Total length Male 18 265.9 ± 5.29 (248.0-285.0) 4.2 10.89 * Female 6 249.7 ± 5.99 (244.0-264.0) 2.9 4.28 Length of tail Male 18 136.6 ± 3.37 (123.0-148.0) 5.2 5.87 * Female 6 128.8 ± 4.54 (124.0-138.0) 4.3 4.28 GENOWAYS— SYSTEMATICS OF LIOMYS Table 2. — Continued. Length of hind foot Male 21 31.4 ± 0.74 Female 8 30.6 ± 0.92 (26.0-34.0) (29.0-32.0) 5.4 1.30 ns 4.3 4.21 Greatest length of skull Male 20 35.8 ± 0.64 Female 7 34.7 ± 0.24 (33.4-38.9) (34.3-35.1) 4.0 4.49 0.9 4.21 Zygomatic breadth Male 5 Female 3 17.0 ± 0.33 16.0 ± 0.23 (16.6-17.4) 2.1 19.88 (15.8-16.2) 1.3 5.59 Interorbital constriction Male 21 Female 8 7.4 ± 0.12 7.5 ± 0.12 (7. 0-8.0) 3.6 0.28 ns (7. 3-7. 8) 2.2 4.21 Mastoid breadth Male 20 Female 8 14.5 ± 0.14 14.5 ± 0.28 (14.0-15.3) (13.8-15.2) 2.1 0.02 ns 2.7 4.22 Length of nasals Male 20 Female 7 15.1 ± 0.31 14.1 ± 0.32 (13.9-17.3) (13.5-14.8) 4.6 11.49 3.0 4.22 Length of rostrum Male 14 Female 5 16.1 ± 0.44 15.2 ± 0.29 (15.0-17.9) (14.7-15.5) 5.2 5.55 2.1 4.41 Length of maxillary toothrow Male 20 5.4 ± 0.11 Female 8 5.3 ± 0.15 (4.9-5. 9) 4.6 0.53 ns (5. 1-5. 6) 3.9 4.22 Depth of braincase Male Female 20 9.5 ± 0.14 8 9.2 ± 0.16 (8.9-10.1) 3.3 6.57 (8.9-9. 5) 2.4 4.21 Interparietal width Male Female 15 8.3 ± 0.22 8 8.1 ± 0.30 (7. 2-8.9) (7.5-8. 5) 5.0 1.19 5.2 4.32 Interparietal length Male 19 4.4 ± 0.19 Female 8 4.4 ± 0.27 (3.4-4. 8) 9.3 0.01 (4.0-5.2) 8.7 4.24 38 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY abcdefghijklm Fig. 4. — Coefficients of variation (ordinate) of external and cranial measurements used in this study. These measurements (abscissa) are as follows: A, total length; B, length of tail; C, length of hind foot; D, greatest length of skull; E, zygomatic breadth; F, interorbital con¬ striction; G, mastoid breadth; H, length of nasals; I, length of rostrum; J, length of maxillary toothrow; K, depth of braincase; L, interparietal width; M, interparietal length. The plotted GENOWAYS— SYSTEMATICS OF LIOMYS 39 cance of secondary sexual variation in this group of rodents and others must await the time that more species have been thoroughly studied. Individual Variation External and cranial measurements. — The majority of the measurements used in this study revealed relatively low individual variation in the four species ana¬ lyzed. Most coefficients of variation were less than 6.0 (Table 2 and Fig. 4); these values are well within limits of those found for other small rodents (Long, 1968, 1969, 1970). Length of the interparietal bone was found to be the most variable parameter tested with coefficients of variation ranging from 6.4-9. 3, but width of the interparietal also was found to have relatively high individual variation. The high variability of these two measurements probably resulted from the fact that size and shape of the interparietal bone tends to be rather variable and, most im¬ portantly, because measurements were difficult to take, resulting in a relatively low degree of precision. Variation in external measurements averaged greater than in cranial measurements. This undoubtedly resulted from variation in taking of the external measurements by various collectors as pointed out by Sumner (1927:177-181). Cranial variation. — Individual cranial variations or abnormalities that I found in Liomys can be grouped as follows: 1) supernumerary or bregmatic bones, 2) abnormalities resulting from injuries, and 3) dental abnormalities from genetic or developmental sources and disease. Variation in the latter two groups was recorded for all specimens examined, but supernumerary cranial bones were recorded only for specimens stored in the collections of The University of Kan¬ sas. These bregmatic bones were found only along the suture between the parie- tals, at the junction of the frontal and parietal bones, and along the sutures be¬ tween parietals and the interparietal (Fig. 5). The occurrence of abnormalities is discussed below for each species in which they were found. Supernumerary bones were noted in only one of the more than 800 specimens examined of Liomys irroratus. This specimen was a subadult female from Ejido Santa Isabel, Tamaulipas (KU 57616). Of more than 2700 specimens examined, two were found to have cranial abnormalities resulting from injury (KU 39898, 59532), four individuals exhibited genetic or developmental abnormalities of teeth (UMMZ 96939, UNAM 3053, KU 96021, USNM 32732/44655), six showed dental erosion resulting from caries (KU 109233, 109236, 55908, 99780, 28055, 24011), and two specimens had dental attrition from causes other than caries (M VZ 91313, USNM 81411). Most interesting of these abnormalities was found in an adult female from Tequesquitengo, Morelos, in which the anterior and posterior lophs of the upper left premolar were never fused; however, the enamel covering of each loph was complete as far as could be determined. numbers refer to taxa as follows: 1, Liomys irroratus (males); 2, Liomys irroratus (females); 3, Liomys pictus (males); 4, Liomys pictus (females); Liomys salvini (males); 6, Liomys salvini (females); 7, Liomys adspersus (males); 8, Liomys adspersus (females). See also Table 2. 40 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY Fig. 5. — Interparietal bone and suture between parietal bones of three specimens of Liomys pictus to illustrate size and position of bregmatic bones. The specimens from left to right are KU 85813, Platanares, 10 mi. E Ruiz, Nayarit; KU 96055, 2.2 mi. NE Contla, Jalisco; KU 89497, 3 mi. NE El Fuerte, Sinaloa. In more than 950 specimens of Liomys pictus examined, supernumerary bones were noted in the crania of 16 individuals (Fig. 5). In more than 2850 specimens examined, one was found to have suffered a cranial injury (CAS 14351), three specimens exhibited genetic or developmental abnormalities of teeth (AMNH 172410, CAS 14337, KU 28451), one individual had dental caries (UA 10812), two specimens showed displaced teeth as the result of foreign material being lodged between them (KU 39824, 39829), and one had a high degree of attrition of peridental tissue (KU 62563). One noteworthy specimen was a subadult male from 5 miles south of Chiapa, Chiapas, in which the second upper molar on the left side consisted of a single loph (apparently the anterior). In nearly 500 specimens of Liomys salvini examined, supernumerary bones were found in the cranium of only one specimen, which was from km. 24 on high¬ way S Guatemala City, Guatemala, Guatemala (KU 83964). In a total of more than 1375 specimens of Liomys salvini examined, two were found to have cranial abnormalities resulting from injuries (KU 71 159, 71 181) and two individuals ex¬ hibited dental abnormalities resulting from developmental errors (UMMZ 96535, 96539). One (USNM 323670) specimen of Liomys adspersus of 194 examined was found to have foreign material lodged between its teeth. No other abnormal¬ ities were noted for this species. External variation. — Little individual or abnormal variation was observed in the external morphology of spiny pocket mice. The only external variation of note was the occurrence of white tips on tails of specimens of Liomys pictus from southern Sonora (five of 1 1 KU specimens from Alamos) and of Liomys irroratus from Michoacan (nine of 19 MVZ specimens from 4 mi. S Cuitzeo and two specimens from Patzcuaro at the same museum). Molt The sequence of molt described below was based on study of specimens of Liomys pictus housed in the Museum of Natural History at Kansas. Other species GENOWAYS— SYSTEMATICS OF LIOMYS 41 Fig. 6. — Generalized diagram of progression of molt in Liomys (from upper left to lower right). This sequence is discussed in text. of Liomys apparently follow a similar pattern of molt. I have seen only two pelages — Juvenile and adult — in museum specimens of Liomys-, however, accord¬ ing to Eisenberg (1963n f" o o 00 in O NO o n OO (N a — p o o o o o o o M o U 1 1 1 NO m 4-4 — M 4^ 00 NO rsi rr 4—4 r~~ rr NO rsi o o O o r- NO r- o d o o o o o o d c 43 M r- NO in ■c* r- r- C m n m oo ■*fr — m o ON r~ ON ON ON ON ON a1- c o o o o o o o u O U 1 1 1 1 1 1 1 rr o * c ON — cu T3 03 4) L_ X3 •a 'o m C/5 03 03 C/5 OS C Cm o X3 E 3 e/5 O O x: £ o o o J3 X. 03 E 00 00 00 3 C C u u u J J J 4» jc ^ 3 5 3 2 “ c £ « 2 75 - ■° <3 Cm ■- ° £ x: a M U a u 43 — Q - 4—* In 03 a u. Q 03 C/3 03 C G- o c ’5) u 03 E 43 4—) C/3 o a 90 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY I m I Fig. 12. — Two-dimensional projections of the first three principal components illustrating the phenetic position of the 36 samples of male Liomys irroratus. Top, component I plotted against compoeent II; bottom, component I plotted against component III. See Fig. 8 and text for key to samples. GENOWAYS— SYSTEMATICS OF LIOMYS 91 I Fig. 13. _ Two-dimensional projections of the first three principal components illustrating the phenetic position of the 39 samples of female Liomys irroratus. Top, component 1 plotted against component II; bottom, component I plotted against component III. See Fig. 8 and text for key to samples. 92 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY The third major cluster of samples of males, at the bottom of the distance phe- nogram, contains the largest individuals of the species. The samples are from two geographical areas; one is Omilteme, Guerrero (39), and the other is northern Jalisco, western Zacatecas, Durango, and Chihuahua (35, 36). The distance phenogram based on values for female Liomys irroratus also clusters the samples into three major groups, but the composition of these groups is somewhat different than those of males. The upper cluster in this phenogram contains samples from southern Texas, northeastern Mexico, Veracruz, Puebla, Morelos, Guerrero, northern Oaxaca, and northern Jalisco (1-5, 7-14, 17-20, 32) as it did for males, but includes also the samples from central and southern Jalisco (26-28) and southern Zacatecas (31) that were grouped with the Mexican Plateau samples for males. It should be noted also that sample 11, which was grouped with the Mexican Plateau samples based on values for males, is placed with the other Pueblan samples based on female values. The second major cluster of sam¬ ples includes those from the Mexican Plateau north of Mexico City into Nuevo Leon and Chihuahua (21-25, 29, 30, 33-36), from central and southern Oaxaca (15, 16), Omilteme, Guerrero (39), west-central Jalisco (38). The sample from the vicinity of Lago de Chapala (29) was placed with those from central Jalisco based on values for males and those from Omilteme (39) and northern Jalisco, western Zacatecas, Durango, and Chihuahua (35, 36) were placed in a separate group based on the data for males. The small sample from the Sierra del Tigre in Jalisco (25), which is represented only by female specimens, was placed in this group. The third major group is based upon a single specimen from Soyatlan del Oro, Jalisco (37), that has an unusually small interparietal bone, which probably accounts for its rather “distant” separation from the second cluster of samples. The amount of phenetic variation represented in the first three principal com¬ ponents for males and females, respectively, of Liomys irroratus was 63.79 and 60.73 for component I, 11.11 and 9.80 for component II, and 9.04 and 9.29 for component III. Results of a factor analysis showing characters influencing the first three components is given in Table 7. From the factor analysis, it can be seen that the first, and by far the most important, component is heavily influenced by general size; components II and III are influenced by size of the interparietal bone and the qualitative cranial characters. Examination of the two-dimensional plots (Figs. 12, 13) and the three-dimensional plots (Figs. 14, 15) of the samples reveal a pattern similar to that of the distance phenogram, but in some ways dif¬ fering from it. The samples on the right-hand side of the three-dimensional pro¬ jection for males represents the same group that was found in the upper part of the distance phenogram (1-5, 7-14, 17-20, 32) but grouped with these are the samples from central Jalisco (27, 28) and the sample from the vicinity of Tehua- can, Puebla (11), which were included in the middle cluster of the distance pheno¬ gram. The position of sample 1 1 is nearly intermediate between samples 1 3 and 14, to the south and west and sample 21 to the northwest. This pattern was evi¬ dent in several of the measurements in the univariate analysis. Samples 23, 29, and 31 lie between the two clusters of samples, although they were included with the central cluster in the distance phenogram. The positions of samples 29 and GENOWAYS— SYSTEMATICS OF LIOMYS 93 Fig. 14. — Three-dimensional projection of 36 samples of male Liomys irroratus onto the first three principal components based upon a matrix of correlation among the 13 external and cranial measurements and three qualitative cranial characters. Components I and II are indicated in the figure and component III is represented by height. See Fig. 8 and text for key to samples. 31 are intermediate between populations that are contiguous with them (27 and 30 for 29, and 27 and 33 for 31). The central group of samples are from two geo¬ graphic areas — Mexican Plateau north of Mexico City into Nuevo Leon in the northeast and central Zacatecas in the northwest (6, 21-24, 26, 30, 33, 34), and central and southern Oaxaca (15, 16). Still included with this group is the sample from southern Jalisco (26). The two samples (35, 36) from extreme northern Jalisco, western Zacatecas, Durango, and Chihuahua form a group by themselves as does the sample from Omilteme, Guerrero (39). The three-dimensional projection for females has on the right-hand side those samples that cluster in the upper part of the distance phenogram. These samples are from three geographic areas as follows: 1) southern Texas and northeastern Mexico as far south as central Veracruz (1-5, 7-10); 2) Puebla, Morelos, northern Oaxaca, and Guerrero (11-14, 17-20); and 3) southern, central, and northern Jalisco and southern Zacatecas (26-28, 31, 32). Sample 1 1, although it does fall within this group, is situated phenetically about halfway between samples from western Puebla and northern Oaxaca (13, 14) and the sample from Valle de Mexico. The samples in the center of the plot also are from three geographic areas as follows: 1) Mexican Plateau north to Nuevo Leon in the northeast and Chihuahua in the northwest (6, 21-25, 29, 30, 33-36); 2) central and southern Oaxaca (15, 16); and 3) La Laguna, Jalisco (38). The sample from extreme northern Jalisco and western Zacatecas (35) is somewhat removed from the re¬ mainder of the group. Samples 25 and 29 are located at the far right of this cen¬ tral group, which is of interest because they are contiguous with populations in central and southern Jalisco that are included in the group of samples at the right. The single specimen from Soyatlan del Oro, Jalisco (37), is separated from the other localities being located at the backside of the projection probably as the result of its small interparietal bone, which is a character that is heavily weighted 94 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY Fig. 15. — Three-dimensional projection of 39 samples of female Liomys irroratus onto the first three principal components based upon a matrix of correlation among the 13 external and cranial measurements and three qualitative cranial characters. Components I and II are indicated in the figure and component III is represented by height. See Fig. 8 and text for key to samples. in this component. The sample from Omilteme, Guerrero (39), located at the far left of the plot, is phenetically quite distant from other samples of Liomys irroratus. Variation in Color The range of color variation in Liomys irroratus (Fig. 16 and Table 8), as with other species of Liomys , is not extensive; the means for red reflectance ranged from 10.6 to 16.8, those for green reflectance from 5.8 to 9.3, and those for blue reflectance from 5.5 to 9.0 in the 46 samples studied for color (samples differ from those used in the univariate analysis because color values were not available for all univariate samples and color exhibits much more microgeographic varia¬ tion than do external and cranial measurements). As will be seen in other species, mice from low, arid situations are palest in color (highest reflectance), whereas those from high, moist areas are darkest. The palest samples of irroratus ex¬ amined were from southern Puebla (20) and coastal Tamaulipas (4, 9), whereas the darkest examined were from Omilteme, Guerrero (24), Cerro San Felipe, Oaxaca (16), Miquihauna, Tamaulipas (8), and Metlaltoyuca, Puebla (13). The samples from the Gulf coastal lowlands of northeastern Mexico and southern Texas generally have high readings, but samples from Veracruz, which grouped with them on mensural data, have relatively low color reflectance readings. Also, the sample from the vicinity of Monterrey and Linnares in Nuevo Leon has a much lower average color reading than do samples from coastal areas of Tamauli¬ pas. The range of values for red reflectance for samples from the vicinity of Mexi¬ co City northward on the Mexican Plateau is from 1 1.9 (in Queretaro) to 15.2 (in Zacatecas), indicating relatively little color variation in this area. Hooper (1947: 47) in describing L. i. pullus from the Valley of Mexico stated that it was much darker than mice from adjacent samples of L. i. alleni and more closely resembled GENOWAYS— SYSTEMATICS OF LIOMYS 95 a •2 a 2 3 <8. , % 1 x o •— CO '*-> 3 X C *, Q Sb Cb a 2 g 3 2 to .> VO ■*s c -2 s, 3 _o o o 3 t<5 b. o 33 C o o, X r- — oo r- X o X — • (N oo c o "t ON ■ct cri C — d _ ^ ^ ^ c 00 Os C 00 OS 00 00 sd Os 00 00 A 1 d o, co q q »A o o d d o 04 CO c sd C x C d sd oo C sd sd w WWW w w w w w UJ 00 GO r- X m oo cn i— < CO r- w OO (N +1 m ^r (N fN o rr o os so o CO coi co d sd ^ ^ _ _ v 4> CO CO o q >o q to cn CO >o O CO, 0X3 3 00 Os Os Os Os Os ON c Os OO OS 1 1 i o3 o CO q CO O CO tA o CO, O co co, Dd C 00 C C C C r-‘ sd oo C sd sd w WWW w' s— x CO 00 O 00 - 00 o CO O Os o CO o — < — • o X o >o OS OS OO d St sd C (N _ . _ . o CO q >o o o q o o CO, o o c A o d co o o cci ri Tf ^ o ro w o — — -* i—i s— < — -* — 1 W W w w ^ ^ • • • UJ Tf NO Tt 0C NO fN 00 w ro r- ON ON 00 lO 00 00 r- r- >o >o (N CO >o O to z — ^ 1-M 1— . w w •2 3 > •2 -2 Cb 3 b. Oc O 4j O 00 LU _l CO < H 03 C V | s O Cl u .hr « 3 * E 1) 03 f- H (N C/5 o "O 1) C _J c o i— l ^ a; a) U- 3 z m 03 O w CO j't' ^ a t a 3 > a. r E 03 c/5 o3 3 .9* •E 3 b y E «3 C/5 03 a 3 03 E 03 H c •o U X o > 4/ 3 z 03 < O . . c „ w c/5 c/5 ^ O S g it « o a. n a, ex o O •— rS *”* •— .£ "3 "E 3 9" 3 c 03 a X 3 H w N N c/5 3 3 _ » — > > — i > o o a. JH 3 c c rti 3 c3 c3 X Nct^cth|5::>uiur'3 CO >o x r— h 00 > > Cs © — (N Metlaltoyuca) 5 11.5 ± 0.95 (10.5-13.0) 9.2 6.1 ± 0.58 (5.5-7.0) 10.7 5.5 ± 0.55 (5.0-6.5) Table 8. — Continued. 96 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY — > rx ~ o ix O 00 00 I I ' - V >x >x >x >x >x ix 00 NO rx rx o o +1 +1 00 NO NO in rx ON (N 00 NO IX o o On 00 NO «x »n »n NO in »n ' — ' s — ' NO o cr in o o o +1 +1 +1 rx ON 00 x~ NO in ON X • r** 00 w oo x rX x — o o +1 +1 +1 NO o q o NO oo 00 ON NO Os in NO *— * in o XN m NO On 1 i © »n o ON cr NX w rx in NO On ^t >X o +1 +1 +1 NO o oo U ri NO 1 o NO — ; 00 ^ d o NX, IX o d 00 oo >x >x >x, >A NO NO NO un w ON r- x q o d ON ON d >x q ix o NO d NO NO w W ON 00 00 Tt XJ rf Tt x’ ON o on x*s V o o o d NO (N 1 »A »A in (N d Os i-H — r-' m IX r~ r- 00 r- o o d o +1 +1 +1 +1 in q 00 NO (N m ri o 1 O o 00 00 o d d OS in t"’ x— s. »n ix mj in IX 00 d NO NO 00 m o in »n IX NO r~’ in in d w' »n o O (N On in «x m -it o o o o o +1 +1 +1 +1 +1 (X Os o Os r-’ in NO C 00 r- m (N m 00 ri »n r-H IX IX S >x ON d x~ oc ix q IX ■ •x ON o os NO On v v /«s o in in o in in ^t rr m* in »— ■ i •n m o o in ri o ri »— *— S«X 00 O i— < cx r- 00 o 00 X- — o ri o o +1 +1 +1 +1 +1 »n Os q r- cn m’ — <" m* 00 n 00 in in c E .2 S t 3 rt u you ctf ctf o o o rt cn re x x x cu rt cd PU o o o ^ IO « o> G w x c cfl G T3 E n m o m o o o O o o q d 1 oo d 1 | on oo o\ ON ON OO r~‘ On ON 00 ON ON On o >n o d o q d. m 1 o 1 1 o 1 o 1 o i >^i iri d d d d r-i r-‘ t~-’ d d r- oo d w w v ' N — * w ' — ' o O in oo r-- CN ^5 m NO 1^5 NO © — : © o O d o o o o o o o o d o +1 +1 +1 +1 +1 + +1 +1 +1 +1 +i +1 +1 +1 +1 +1 m o — . CN t"; 00 — CN m m o o q o d oo d d oo 00 sd oo NO oo ON CN oo m CN >o _ CN m r- o , >o d CN Tt CN 00 o oo 00 cb OO CN «0 A »o o q o o IT) o o q o o o i On 1 r-’ 1 ON i 00 i On i o o ON r-" ON ON Os o o o o q tn i i o q o >A o o l o l 1 o NO r-* NO NO r-’ • 00 r-^ d d d 00 t-- 'w' 'w/' t — r- w w w w w 00 o ON CN NO 00 r- NO _ ON NO O (N 00 /-) o o CQ m (N NO A NO A A A A A ON d A A H q q — 1 — 1 >— i — — ■ ’—i — iA A o to to q q i o i to i »o 1 o 1 to 1 to i to i »o (N — * — (N r4 ro -a o Dl O O c N O O O O cd o o o o o o o o d o o u 3 C/5 CO a> c 00 o 15 .— 15 o 15 15 UJ 15 u ^“5 *— 5 'w' w' CN CO to NO r- OO m (O m CO CO CO CO o si 'z & ,c o o3 ca o 3 0Q .2 '5* O 3 c a kJ 03 .<2 3 J C 03 O ^ u o o ■3 a. 3 '3 ”<3 o -J 3 o C/5 o _ a u1 c 03 03 — u On cn O — Tf * § C/5 CN ’'fr «C < C/5 03 O O £ « 03 3 03 U O cd a N N l< C .3 C* 1/5 i/5 cd o -a c cd u o o iu -a o q/ ~ o .(u £ t T3 £ , x: - ^ l O o S o m 3 g> § * 3 i- 3 2 3 £ 3 Q x: Q U co N- to vo 97 98 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY Fig. 16. _ Geographic variation in red reflectance of the middorsal coloration of Liomys irroratus. The palest sample is represented by an open symbol, the darkest sample by a completely closed symbol, and the remaining samples by symbols that are expressed as a percentage of the difference between these extremes. See Table 8 for areas represented by symbols. L. guerrerensis from Omilteme, Guerrero. However, the sample from the geo¬ graphic range of pullus (25) had a mean red reflectance of 1 3. 1 (range, 1 2.0-1 5.0), which matches closely values for alleni in Queretaro and San Luis Potosi and, in fact, is paler than some of them. Compared with the specimens from Omilteme that had a mean of 10.6 for red reflectance with a range of 9.5 to 12.0, they are much paler. Samples from Jalisco had mean values for red reflectance between 12.7 and 14.7. Generally, there is relatively little variation in color in Liomys irroratus ; varia¬ tion that is present seems more related to local variation associated with dif¬ ferences in habitat than to genetic similarities. Essentially this same pattern is repeated for all species of Liomys. GENOWAYS— SYSTEMATICS OF LIOMYS 99 Taxonomic Conclusions Based upon my assessment of geographic variation in Liomys irroratus, I have identified seven separate units, which exhibit characteristic evolutionary ten¬ dencies and which intergrade (or have intergraded in the recent past) with con¬ tiguous units in relatively narrow zones; these seven units are the subspecies of L. irroratus here recognized. Four of these units are characterized by relatively large size. The largest individuals of the species and probably the phenetically most distinct occur in the vicinity of Omilteme, Guerrero, to which the trinomial Liomys irroratus guerrerensis Goldman, 1911, applies. Another highly distinctive subspecies, Liomys irroratus bulleri (Thomas, 1 893), occurs in west-central Jalis¬ co and is characterized by large size and a small interparietal bone. The nominate subspecies, Liomys irroratus irroratus (Gray, 1 868), occupies an area in central Oaxaca southward into the Sierra Madre del Sur of Oaxaca. The last of the sub¬ species characterized by large size occupies an extensive geographical area north¬ ward from Valle de Mexico onto the Mexican Plateau to eastern Jalisco in the west, southern Chihuahua in the northwest, and the Sierra Madre Oriental of Nuevo Leon in the northeast; the oldest available name for this taxon is Liomys irroratus alleni (Coues, 1881). A case could be made for recognition of the pop¬ ulations in western Zacatecas, extreme northern Jalisco, Durango, and southern Chihuahua under the name Liomys irroratus canus\ however, although these ani¬ mals are quite large, the change in size from that found in populations in San Luis Potosi forms a gradual cline with no distinct break that can be detected in the material at hand. Therefore, I feel the present arrangement best represents the relationships of these populations. The three remaining subspecies are characterized by medium to small size. Liomys irroratus texensis Merriam, 1902, occurs in the Gulf coastal lowlands of southern Texas, Nuevo Leon, Tamaulipas, eastern San Luis Potosi, and northern Veracruz. A second subspecies, Liomys irroratus torridus Merriam, 1902, occurs south of the Transverse Volcanic Belt in Puebla, Morelos, Guerrero, and northern Oaxaca. In southern and central Jalisco and southern Zacatecas is the medium¬ sized Liomys irroratus jaliscensis (J. A. Allen, 1906). Liomys irroratus alleni (Coues, 1881) 1881. Heteromys alleni Coues, in Allen, Bull. Mus. Comp. Zool., 8: 1 87, March. 191 1. Liomys irroratus alleni, Goldman, N. Amer. Fauna, 34:56, 7 September. 1902. Liomys canus Merriam, Proc. Biol. Soc. Washington, 15:44, 5 March; holotype from near Parral, Chihuahua. 1947. Liomys irroratus pullus Hooper, J. Mamm., 28:47, 17 February; holotype from Tlalpan, 2250 m., Distrito Federal. 1948. Liomys irroratus acutus Hall and Villa-R., Univ. Kansas Publ., Mus. Nat. Hist., 1:253, 26 July; holotype from 2 mi. W Patzcuaro, 7700 ft., M ichoacan. Holotype. — Probably an adult, sex unknown, skull in skin, MCZ 5889, from Hacienda Angostura, Rio Verde [may be near small village of this name, 33 km. NNE Rio Verde], San Luis Potosi; obtained on 26 February 1878 by Edward Palmer. Skin in relatively good condition, but condition of skull unknown. Measurements of holotype. — No measurements are available. 100 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY Distribution. — This subspecies has an extensive geographic range on the Mexi¬ can Plateau north of the Transverse Volcanic Belt (see Fig. 17). The southern limit of the range of alleni is in Tlaxcala, Mexico, Distrito Federal, and northern Michoacan from where it occurs northward into Hidalgo, Queretaro, Guanajuato, northeastern Jalisco, Aguascalientes, Zacatecas, Durango, southern Chihuahua, and San Luis Potosi; specimens referable to alleni also are known from the Sierra Madre Oriental of Tamaulipas and Nuevo Leon. Comparisons. — Liomys irroratus alleni is distinguished from Liomys irroratus jaliscensis by its larger size, which is especially noticeable in cranial measure¬ ments (compare values for samples 6, 21-25, 29, 30, 33-36 with those of 27, 28 and females from sample 26 in Table 3). None of the qualitative cranial char¬ acteristics appear to be sufficiently distinctive to distinguish alleni from jaliscensis, although samples of the latter do have a somewhat lower percentage of individuals with the posterior termination of the nasals truncate in shape than in most sam¬ ples of alleni (Table 6). From Liomys irroratus irroratus, the subspecies alleni can be distinguished by its smaller external size (compare values for samples 6, 21-25, 29, 30, 33-36 with samples 15, 16 in Table 3) and by the terminal shape of the nasals (Table 6). In 62 specimens examined of irroratus, only one had the posterior end of the nasals truncate, whereas samples of alleni have from 44.4 to 80.0 per cent of the in¬ dividuals with this characteristic (excepting in sample 24 where only three of 35 individuals have the nasals truncate). Cranial measurements of these two taxa do not appear to be distinctive. For comparison of alleni with bulleri, guerrerensis, texensis, and torridus, see accounts of those subspecies. Remarks. — As understood herein, the name Liomys irroratus alleni applies to those populations on the Mexican Plateau north of the Transverse Volcanic Belt that are characterized by large size and a relatively high percentage of individuals with the interparietal bone divided into halves and the posterior termination of the nasals truncate. This subspecies now includes the previously recognized races acutus, canus, and pullus. Hooper and Handley (1948: 18-19) pointed out that the population to which Goldman applied the name alleni was characterized by no unique feature, but each of the characters attained maximum or minimum devel¬ opment in adjoining areas. Those characters cited (Hooper and Handley, 1948: 18-19) as distinguishing acutus, canus, and pullus were found in my analysis to be non-existent, clinal in nature, or restricted to local situations. For example, all three of these nominal subspecies were supposedly larger than alleni. In the uni¬ variate analyses, the samples of acutus (sample 24) and pullus (sample 21) were found to be no larger than alleni, but the sample of canus (sample 36) was larger. The two intermediate samples (samples 34, 35) form a clinal bridge in many measurements between typical samples of alleni (33) and canus (36). In the multi¬ variate analysis for males, the two samples 35 and 36 are set off somewhat from the other samples of alleni, but in the analysis of females sample 36 fell near 34, 6, and 33. I have interpreted these results to mean that specimens formerly rec¬ ognized as L. i. canus are larger than typical L. /. alleni from San Luis Potosi, GENOWAYS— SYSTEMATICS OF LIOMYS 101 but that the differences are clinal in nature with no sharp steps in the cline. I here- fore, canus is best placed as a junior synonym of alleni. Two qualitative cranial characters appear to distinguish the sample of acutus (24) from other samples of alleni. A much smaller percentage of individuals of acutus (5.9 per cent) have a divided interparietal when compared with other samples of alleni (15.2 to 50 per cent). Also a much smaller percentage of the specimens of acutus (8.6 per cent) have truncate nasals posteriorly than do speci¬ mens in other samples of alleni (44.4 to 80.0 per cent). However, when all char¬ acters are considered at once in the multivariate analysis, the relationship of specimens formerly known as acutus is clearly with alleni. In coloration, specimens of acutus and pullus supposedly are darker than specimens of alleni (Hooper and Handley, 1948:18), whereas those of canus are somewhat paler (Hooper and Handley, 1948:22). The two samples of alleni have quite different mean values for red reflectance, 12.6 and 14.4, and a wide range of individual variation, 10.5 to 19.0. Material from near the type locality of pullus has a mean for red reflectance of 13.1, which is darker than in specimens from near the type locality of alleni and paler than material from near the city of San Luis Potosi. My sample of material of acutus may not be particularly repre¬ sentative of the race, but it averages slightly darker (12.5) in red reflectance than other samples of alleni as understood by Hooper and Handley. The two samples of canus have mean red reflectance readings (1 3.5 and 1 3.7) that are intermediate between the mean values for the two samples of alleni from San Luis Potosi. Based on these observations of the dorsal coloration, I see no merit in recognizing more than one race in this complex. Intergradation between L. i. alleni and L. i. jaliscensis is evident in specimens from near Lago de Chapala in Jalisco and Michoacan northward through Jalisco into extreme southern Zacatecas. Females in sample 25 from the vicinity of Maza- mitla, Jalisco, are somewhat smaller than those from adjacent Michoacan, but clearly they are larger than females from southern Jalisco. In the three- dimensional projection of the multivariate analysis, sample 25 falls at the lower edge of the group that includes other samples of alleni, but it is best placed with this group rather than with jaliscensis. Specimens in sample 29, especially the males, exhibit intermediate characters between alleni and jaliscensis in the uni¬ variate analysis. In the multivariate analysis, the males fall between alleni and jaliscensis and could be referred to either; however, the females, although inter¬ mediate, fall nearer alleni. Examination of the individual specimens in this sample reveals that those from Ocotlan and 10 mi. NE Yahualica in Jalisco are large and are clearly referable to alleni, whereas those from 6 mi. N and 4 mi. E Tepat- itlan, Jalisco, and 2 mi. SE La Palma, Michoacan, are intermediate in size. Speci¬ mens from the vicinity of Tepatitlan are assigned to alleni because they have qualitative cranial characters that suggest this relationship (three have the inter¬ parietal bone notched posteriorly, and two of three have the interparietal bone divided). The two males from near Zapotlanejo are of interest because they are from the westernmost locality in this area assigned to alleni. One specimen is relatively large (greatest length of skull 33.4), but the other is relatively small 102 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY (31.8). However, both have cranial characters (interparietal divided and notched, and nasals truncate) that suggest assignment to alleni is best until additional ma¬ terial indicates otherwise. The large series of specimens from 4 mi. NE Ocotlan contains only one adult and its cranial characters are inconclusive in indicating relationship with either subspecies. The same is true of specimens from La Palma, Michoacan. I have assigned these to alleni because their geographic origin seems to indicate that this was the best course. Specimens from extreme southern Zacatecas in the vicinity of Santa Rosa evince intergradation between jaliscensis and alleni. Males from this locality have mean values for seven measurements that are intermediate between values for alleni and jaliscensis and females are intermediate in five (total length, tail length, greatest length of skull, length of nasals, and depth of braincase for both sexes; hind foot and length of maxillary toothrow for males). Females, although aver¬ aging larger than females of jaliscensis , had mean values that fall nearer those of jaliscensis than alleni. In the multivariate analysis, males were intermediate be¬ tween alleni and jaliscensis, whereas the females were closer to samples of jali¬ scensis. The cranial characters of this sample also indicate its intermediate nature (Tables 5, 6). I have assigned these specimens to jaliscensis because females ap¬ pear to fit best here and the other evidence would allow assignment to either taxa. Remains of Liomys have been reported previously from San Josecito Cave, near Aramberri, Nuevo Leon, by Cushing (1945:185) and Jakway (1958:320). Jakway assigned a fragmentary cranium, four right dentaries, and one left dentary available to him to Liomys irroratus. The material that I was able to examine in the Vertebrate Paleontology Collection of the Los Angeles County Museum con¬ sisted of more than 1 50 of both right and left dentaries and 66 partial crania. Measurements of the San Josecito Cave material are given in Table 9 along with those of 20 specimens (10 males and 10 females) of Liomys irroratus alleni from the general vicinity of the cave. Note that the San Josecito material has a signifi¬ cantly longer maxillary toothrow and right mandibular toothrow than does the Recent material, although it is matched by measurements of other Recent popula¬ tions of alleni from Zacatecas, Durango, and Chihuahua (Table 3). The four other measurements that I was able to take on the cave material, including length of the left mandibular toothrow, showed no significant differences between the two groups. The San Josecito specimens are definitely representatives of Liomys irroratus and probably not much different from the Recent representatives of species occurring in the area today. Similar results were obtained by Hooper (1952:58-59, 195), who studied representatives of Reithrodontomys megalotis from the cave. Jakway (1958) assigned material of 10 other rodents to Recent species; only one fossil species of rodent, Orthogeomys onerosus (Russell, 1960) is recognized from San Josecito Cave material. Five adult males and eight adult females (nonpregnant) from Zacatecas, San Luis Potosi, and northeastern Jalisco averaged, respectively, 16.8 (16.0-18.0) and 16.0 (15.0-17.0) in length of ear, and 73.7 (70.0-78.8) and 63.7 (48.4-82.0) in weight. GENOWAYS— SYSTEMATICS OF LIOMYS a "a a S £ ~3 c a c 'O tu G o > 3 5 o ■> -3 G Ci §4 *5 * 5 Cj J2 2 1 § 2 «<5 r_\ .o s 3 >5 O C/D 3 c5 u O u u C/D >, E o o to 3 3 to 3 ON UJ _l 03 < H - u tu E a> > aJ a o D to O |— > c cd 00 c jo "c3 to 3 . — i 03 •— O to >. £ o so o o o ON d) r~’ d (N >oi d v-H ' — ^ 00 00 */"D NO NO o O o o o o o o o d o +1 +1 +1 +1 +1 +1 r*N m r- •'t m m 00 toi m ioi >oi o O o O O o n c -a 03 u u JO "O '5 to 03 >> u- o3 x 03 E »*— i o sz DO 00 c c ■o c 03 E o JZ 00 c V J 3 £ •a c o3 E 103 104 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY Specimens examined (707). — Aguascalientes: 3 mi. N Aguascalientes, 2000 m., 1 (MVZ); 18 mi. W, 2 mi. S Aguascalientes, 6000 ft., 1 (KU); 4V2 mi. NW Calvillo, 6000 ft., 11 (MVZ); Va mi. E state boundary, 8 mi. SW Calvillo, 4 (MVZ); 1 mi. N Chicalote, 1900 m., 2 (MVZ); Chicalote, 1 (USNM); 1 km. S La Labor (9 mi. by road N Calvillo), 1 (MVZ). Chihuahua: La Cienegilla Springs, 46.9 mi. S Jimenez, 2 (TTU); near Parral (31 miles northwest of Parral according to Goldman, 1951:124), 8 (USNM); 5 mi. E Parral, 5700 ft., 1 (KU); Santa Rosalia, 1 (USNM). Distrito Federal: Cerro Xaltepec, IV2 mi. NNW Zapotitlan, 2380 m., 2 (1 KU, 1 UNAM); Ciudad Universitaria, 1 (UNAM); Contreras, 2600 m., 3 (UMMZ); Pedregal de San Angel, 3 (ENCB); San Geronimo, 2400 m., 27 (UMMZ); San Gregorio Altapulco, 2270 m., 1 (KU); Va mi. NNE San Mateo Xalpa, 2390 m., 1 (KU); 200 m. N San Mateo Xalpa, 1 (UNAM); Tlalpan, 16 (2 FMNH, 11 UMMZ, 3 USNM); Vt km. N Xochitepec, 2250 m., 1 (UNAM). Durango: 5Vi mi. NNW Canatlan, 6400 ft., 1 (MSU); 1 mi. N Chorro, 6450 ft., 1 (KU); 8 mi. NW Durango, 6200 ft., 1 (KU); Durango, 1 (USNM); Inde, 5 (USNM); Navarro, 1 (UA); Rio Nazas, 10 mi. NNW Rodeo, 2 (KU); Rio Sestin, 4 (AMNH); Rosario, 6 (AMNH); 3 mi. SE Tepehuanes, 5840 ft., 6 (MSU). Guanajuato: Acambaro, 1 (USNM); Moroleon, 1 (USNM); Salvatierra, 5775 ft., 1 (KU); 5 km. N San Jose, 3 (WBC); Silao, 8 (USNM). Hidalgo: 10 mi. NW Actopan, 6000 ft., 2 (MSU); 2 km. E Actopan, 2 (ENCB); 12 km. S, 2 km. E Actopan, 8200 ft., 1 (KU); 1 km. N Epazoyucan, 2550 m., 3 (ENCB); Ixmi- quilpan, 5550 ft., 1 (KU); 6 mi. S Ixmiquilpan, 1 (ENCB); Maguey Verde, 8V2 mi. NE Zimapan, 7100 ft., 1 (MVZ); Marques, 2 (USNM); San Agustin, 1100 m., 3 (UMMZ); Tula, 2050 m„ 3 (1 FMNH, 2 USNM); Zacualtipan, 1800 m., 4 (UMMZ); Zimapan, 6 (3 UMMZ, 3 USNM). Jalisco: 2 mi. S Ayo el Chico, 5400 ft., 3 (KU); Belen de Refugio, 5700 ft., 5 (KU); 4V2 mi. NE Comanja de Corona, 8000 ft., 10 (KU); 9 mi. N Encarnacion, 1900 m., 12 (11 MVZ, 1 UNAM); SV2 mi. N, 2 mi. W Guadalupe de Victoria, 1 (MSU); Huascato, 3 (ENCB); 3 mi. S Huejucar, 5900 ft., 1 (KU); 5 mi. NE Huejuquilla, 6200 ft., 3 (KU); 1 mi. S Jalostotitlan, 5700 ft., 1 (KU); 7 mi. N Lago de Moreno, 1 (UNAM); 2 mi. WNW Lagos de Moreno, 6370 ft., 1 (KU); IV2 mi. N Mazamitla, 5 (UMMZ); V2 mi. NW Mazamitla, 2 (UMMZ); 4 mi. W Mazamitla, 6600 ft., 3 (KU); 3 mi. WSW Mazamitla, 1 (KU); 4 mi. NE Ocotlan, 5050 ft., 26 (24 KU, 2 MSU); 2 mi. WNW Ocotlan, 5000 ft., 4 (KU); Ocotlan, 15 (USNM); 1 mi. S Ocotlan, 5000 ft., 1 (KU); 6 mi. N, 4 mi. E Tepatitlan, 6400 ft., 2 (KU); 2 mi. S, V2 mi W Tepatitlan, 1 (KU); 3 mi. E Union de San Antonio, 6100 ft., 4 (KU); 10 mi. NE Yahualica, 5 (KU); 2 mi. E Zapotlanejo, 2 (KU). Mexico: Atlacomulco, 2500 m., 2 (UMMZ); 1.1 km. N Barrientos, 2300 m., 1 (ENCB); Cerro La Caldera, 1 1 mi. SE Mexico City, 2350 m., 7 (3 KU, 4 UNAM); 3 km. E, 2 km. S Chilpa, 2270 m., 4 (ENCB); 3.5 km. W Ecatepec, 2500 m., 3 (ENCB); Hacienda Cordoba, 2600 m., 1 (UMMZ); 23 km. E Mexico, 7500 ft., 20 (2 KU, 18 TCWC); 4 km. SSE San Rafael, 2460 m., 1 (KU); Temascaltepec, 4 (UMMZ); Tenancingo, 2050 m., 2 (FMNH); 5 mi. S, 1 mi. W Texcoco, 7350 ft., 1 (KU); 1 km. E Teotihuacan, 2420 m., 3 (ENCB); 2 km. NW Tlalr.epantla, 2300 m., 2 (ENCB); 4 km. ENE Tlalmanalco, 2290 m., 4 (2 KU, 2 UNAM); 2 km. NE Tlalpitzahuac, 1 (KU); 3 km. N Valle de Bravo, 4 (UNAM); U/2 mi. S Valle de Bravo, 6050 ft., 3 (KU); 3 mi. ENE Zoquiapan, 9200 ft., 1 (MSU). Michoacan: 3 mi. S Carapan, 6900 ft., 2 (MSU); 4 mi. S Cuitzeo, 1800 m., 19 (UMMZ); 9 km. W Jacona, 4 (ENCB); 4 km. SW Jacona, 1 (ENCB); 13 km. SW Jacona, 2 (ENCB); 2 mi. SE La Palma, SE side Lago de Chapala, 3 (KU); 2 mi. E La Palma, SE side Lago de Chapala, 1 (MSU); 12 km. W Morelia, 1 (UNAM); 15 mi. E Morelia, jet. of highway 4 and road to Tzitzio, 2150 m., 1 (UMMZ); 8 mi. NE Patzcuaro, 2 (MVZ); 3 mi. NW Patzcuaro, 6700 ft., 1 (MVZ); 2 mi. W Patzcuaro, 6700-7700 ft., 7 (MVZ); Patzcuaro, 6 (2 FMNH, 1 MCZ, 2 UMMZ, 1 USNM); 5 mi. S Patzcuaro, 7800 ft., 11 (4 MSU, 7 MVZ); Querendaro, 5 (USNM); 11 mi. W Quiroga, 1 (UMMZ); Tanc'itaro, 6000 ft., 1 (FMNH); 4Vi mi. NE GENOWAYS— SYSTEMATICS OF LIOMYS 105 Tarecuato, 6600 ft., 1 (KU); 1 mi. N Tinqiiindin, 6300 ft., 1 (KU); Zamora, 3 (USNM); 3 mi. E Zamora, 5480 ft., 1 (MSU); 10 km. E Zamora, 3 (ENCB). Nuevo Leon: Aramberri, 3600 ft., 1 (KU); 1 km. SW Aramberri, 3600 ft., 1 (KU); 5 mi. W Ascencion, 6500 ft., 1 (KU); Ibarilla, 35 mi. S Linares, 3000 ft., 7 (KU); Iturbide, Sierra Madre Oriental, 5000 ft., 5 (KU); Ojo de Agua, 1 (FMNH); Pablillo, 38 km. W, 28 km. S Linares, 7400 ft., 3 (KU); 1 Vi mi. N Zaragoza, 4500 ft., 3 (KU); 1 mi. S Zaragoza, 4600 ft., 1 (KU). Queretaro: Cadereyta, 2100 m., 4 (UMMZ); Jalpan, 11 (USNM); 5 mi. NE Pinal de Amoles, 6500 ft., 1 (MSU); Pinal de Amoles, 9 (3 UMMZ, 6 USNM); 15 km. N Queretaro, 2 (TCWC); 6 mi. E Queretaro, 7400 ft., 5 (KU); 6 mi. E Queretaro, 6450 ft., 1 (MSU); Tequisquiapan, 1 (USNM); Toliman, 1700 m., 15 (UMMZ). San Luis Potosi: Ahualulco, 6 (USNM); Alvarez, 8 (4 AMNH, 4 MCZ); 2 km. NE Arriaga, 2 (LSU); Bledos, 20 (LSU); Cerro Campanario, 1 (LSU); Cerro Penon Blanco, 7750 ft., 1 (LSU); Hacienda Angostura, 3 (1 MCZ, 2 USNM); Hacienda Capufin, 3 (LSU); Hacienda La Parada, 3 (USNM); Jesus Maria, 2 (USNM); Paso de San Antonio, 1 (LSU); 8 mi. SW Ramos, 6700 ft., 8 (KU); Rio Verde, 13 (1 LSU, 12 USNM); IV2 mi. E R'10 Verde, 3 (LSU); San Carlos, 1 (AMNH); 10 mi. NE San Luis Potosi, 3 (2 KU, 1 MSU); 11 km. S, 3(/2 km. E Santa Maria del Rio, 1800 m„ 1 (ENCB); Villar, 18(3 LSU, 15 USNM). Tamaulipas: Jaumave, 49 (3 FMNH, 23 KU, 22 UMMZ, 1 USNM); Miquihauna, 3 (USNM); Nicolas, 56 km. NW Tula, 5500 ft., 6 (KU); 16 mi. N, 6 mi. W Palmillas, 5500 ft., 1 (KU); 14 mi. N, 6 mi. W Palmillas, 5500 ft., 2 (KU). Tlaxcala: 5 mi. W Ciudad Tlaxcala, 3 (YPM). Veracruz: 10 km. SWJacales, 6500 ft., 2 (KU). Zacatecas: Berriozabal, 4 (USNM); 4 mi. NNW Chalchihuites, 7000 ft., 3 (CAS); Haci¬ enda San Juan Capistrano, 2 (USNM); 13 mi. N Jalpa, 5000 ft., 1 (KU); 3 mi. SW Jalpa, 4600 ft., 3 (KU); 1 mi. NE Noria de Angeles, 1 (CAS); 10 mi. N Rancho Grande, 6200 ft., 11 (MSU); 5 mi. SE Rio Grande, 6360 ft., 1 (MSU); Rio Nieves, 1 mi. N Rancho Grande, 1 (KU); 2 mi. W Sain Alto, 6900 ft., 2 (KU); 2 mi. ESE Trancoso, 1 (KU); 7 mi. SE Tran- coso, 1 (KU); Valparaiso, 20 (USNM); 9 mi. W Zacatecas, 7800 ft., 1 (CAS); 5 mi. NW Zacatecas, 7600 ft., 3 (KU); 8 mi. SE Zacatecas, 7225 ft., 4 (KU). Additional records. — Durango: 9 mi. N Durango (Baker and Greer, 1962:104). Jalisco: Encarnacion de Diaz (Grant, 1947:8; Johnson, 1962:417). Nuevo Leon: San Francisco (Koestner, 1941:12). Marginal records (localities in italics are not plotted on Fig. 17 because undue crowding of symbols would have resulted). — Chihuahua: Santa Rosalia; La Cienegilla Springs, 46.9 mi. S Jimenez. Durango: Inde; Rio Nazas, 10 mi. NNW Rodeo. Zacatecas: 5 mi. SE Rio Grande, 6360 ft.; 10 mi. N Rancho Grande, 6200 ft.; Rio Nieves, 1 mi. N Rancho Grande; 5 mi. NW Zacatecas, 7600 ft. San Luis Potosi: 8 mi. SW Ramos, 6700 ft.; Cerro Penon Blanco; Ahualulco; Villar; San Carlos. Tamaulipas: Nicolas, 56 km. NW Tula, 5500 ft. Nuevo Leon: 5 mi. W Ascencion, 6500 ft.; Pablillo, 38 km. W, 28 km. S Linares, 7400 ft. San Francisco; Ojo de Agua; Iturbide, Sierra Madre Oriental, 5000 ft.; Ibarilla, 35 mi. S Linares, 3000 ft.; Aramberri, 3600 ft.; I 1/2 mi. N Zaragoza, 4500 ft.; 1 mi. S Zara¬ goza, 4600 ft. Tamaulipas: Jaumave. San Luis Potosi: 1 Vi mi. E Rio Verde; Hacienda Capufin. Queretaro: Jalpan. Hidalgo: Maguey Verde, 8‘/2 mi. NE Zimapan, 7100 ft.; Zacualtipan. Veracruz: 10 km. SW Jacales, 6500 ft. Hidalgo: 1 km. N Epazoyucan, 2550 m. Tlaxcala: 5 mi. W Ciudad Tlaxcala. Mexico: 4 km. SSE San Rafael, 2460 m. Distrito Federal: San Gregorio Altapulco; 200 m. N San Mateo Xalpa. Mexico: Tenancingo, 2050 m.; Temascaltepec. Michoacan: 5 mi. S Patzcuaro, 7800 ft.; Tancitaro, 6000 ft.; 1 mi. N Tinqiiindin, 6300 ft. Jalisco: 3 mi. WSW Mazamitla; 4 mi. W Mazamitla, 6600 ft. Michoacan: 2 mi. SE La Palma, SE side Lago de Chapala. Jalisco: 1 mi. S Ocotlan, 5000 ft.; 2 mi. WNW Ocotlan, 5000 ft.; 2 mi. E Zapotlanejo; 2 mi. S, Vi mi. W Tepatitlan; 10 mi. NE Yahualica. Zacatecas: 3 mi. SW Jalpa, 4600 ft. Jalisco: 3 mi. S Huejucar, 5900 ft.; 5 mi. NE Huejuquilla, 6200 ft. Zacatecas: Hacienda San Juan Capistrano; 4 mi. NNW 106 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY Chalchihuites, 7000 ft. Durango: Durango; 8 mi. NW Durango; 3 mi. SE Tepehuanes, 5840 ft.; Navarro. Chihuahua: near Parral. Liomys irroratus bulleri (Thomas, 1893) 1893. Heteromys bulleri Thomas, Ann. Mag. Nat. Hist., ser. 6, 1 1:330, April. 191 1. Liomys bulleri, Goldman, N. Amer. Fauna, 34:61, 7 September. Holotype. — Adult female, skin in alcohol with skull removed, BMNH 93.3.6.39, from La Laguna, Sierra de Juanacatlan, Jalisco; obtained in December 1892 by A. C. Buller. Skin in alcohol in good condition; skull in good condition excepting a lower incisor broken. Measurements of holotype. — Greatest length of skull, 34.1 ; zygomatic breadth, 16.7; interorbital constriction, 8.6; mastoid breadth, 15.5; length of nasals, 13.4; length of rostrum, 15.3; length of maxillary toothrow, 5.9; depth of braincase, 10.0; interparietal width, 7.4; interparietal length, 3.9. Distribution. — Known only from the vicinity of La Laguna and Soyatlan del Oro in west-central Jalisco (see Fig. 17). Comparisons. — Liomys irroratus bulleri can be distinguished from the sub¬ species jaliscensis, texensis , and torridus by its larger external and cranial size (see Table 3). From the remaining subspecies of irroratus ( alleni , guerrerensis, and irroratus ) as well as those mentioned above, bulleri is distinguished by its small triangular or subtriangular interparietal bone. The small size of the inter¬ parietal is matched in only one sample of alleni (specimens formerly allocated to acutus ), but in these specimens the interparietal is semicircular in shape (Hall and Villa-R., 1948:254). Remarks. — Liomys irroratus bulleri is a large-sized subspecies known only from a restricted area in west-central Jalisco. It is potentially in contact only with the medium-sized L. i. jaliscensis to the east in central Jalisco. Until the present study, bulleri had been considered a monotypic species closely related to irrora¬ tus. The characters described by Goldman as distinctive of bulleri were its small and subtriangular interparietal bone and rounded termination of the nasals. The latter character is definitely not unique to bulleri because at least some individ¬ uals in all samples of Liomys irroratus exhibit this condition. The small size of the interparietal is rather unique for the species, but it is matched in a population in Michoacan (sample 24). Furthermore, because the size and shape of the inter¬ parietal bone is highly variable in Liomys irroratus, as it is in other species in the genus, I do not believe it is a character upon which specific distinction can be based. I have no evidence in my limited material to indicate intergradation of bulleri with jaliscensis, but the two taxa have a relationship that is similar to that between L. i. torridus and L. i. irroratus. On the basis of my findings, the relation¬ ships of bulleri are best expressed by considering it as a distinct subspecies of Liomys irroratus rather than retaining it as a monotypic species. In the multivariate analysis, samples 37 and 38 are relatively far removed from other samples of irroratus in the three-dimensional plot, especially in the second principal component. In the second component, the two most heavily weighted characters by far are length and width of the interparietal. GENOWAYS— SYSTEMATICS OF LIOMYS 107 Fig. 17. — Geographic distribution of subspecies of Liomys irroratus: 1, L. i. alleni; 2, L. i. bulleri; 3, L. i. guerrerensis; 4, L. i. irroratus; 5, L. i. jaliscensis; 6, L. i. texensis; 7, L. i. torrid us. Until the present study, Liomys irroratus bulleri was known by only two speci¬ mens from the type locality, La Laguna, Jalisco. However, five specimens in the University of Arizona collection from the vicinity of Soyatlan del Oro, Jalisco (sample 37), also appear to be referable to this taxon. The one adult female from this area is large as are the holotype and a topotype and all of the specimens have small interparietal bones (besides the adult female in Table 3, two subadult males and a subadult female had the following measurements: interparietal width, 7.0, 7.0, 6.4 and interparietal length 3.6, 3.3, 3.5). Specimens from the vicinity of Soyatlan del Oro extend the geographical range of bulleri approximately 60 kilometers to the southeast. The adult female from Soyatlan del Oro weighed 5 1 grams. Specimens examined (7). — Jalisco: La Laguna, Sierra de Juanacatlan, 2 (1 BMNH, 1 USNM); Rancho de Colomo, 3 km. E Soyatlan del Oro, 1 (UA); 3 km. N Soyatlan del Oro, 2 (UA); 5 km. W Soyatlan del Oro, 1 (UA); 5 km. S Soyatlan del Oro, 1 (UA). 108 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY Marginal records. — Jalisco: La Laguna, Sierra de Juanacatlan; 3 km. N Soyatlan del Oro\ Rancho de Colomo, 3 km. E Soyatlan del Oro; 5 km. S Soyatlan del Oro; 5 km. W Soyatlan del Oro. Liomys irroratus guerrerensis Goldman, 1911 1911. Liomys guerrerensis Goldman, N. Amer. Fauna, 34:62, 7 September. Holotype. — Subadult female, skin and skull, USNM 127523, from Omilteme, Guerrero; obtained 17 May 1903 by E. W. Nelson and E. A. Goldman. Skin in good condition; skull with broken zygoma. Measurements of holotype. — Total length, 255.0; length of tail, 127.0; length of hind foot, 34.0; greatest length of skull, 33.3; interorbital constriction, 8.6; mastoid breadth, 15.8; length of nasals, 12.5; length of rostrum, 14.9; length of maxillary toothrow, 6.0; depth of braincase, 9.2; interparietal width, 8.5; inter¬ parietal length, 3.6. Distribution. — Pacific slope of the Sierra Madre del Sur of Guerrero in the vicinity of the type locality (see Fig. 17). Comparisons. — Liomys irroratus guerrerensis can be distinguished from all other races of Liomys irroratus by its larger size, both externally and cranially (compare values of sample 39 with others in Table 3). In addition, the sample of guerrerensis is darker dorsally than any other sample of the species. Compared with adjacent samples of L. i. torridus (17, 18), guerrerensis has a much larger percentage of individuals with the posterior margin of the interparietal notched and the posterior termination of nasals rounded (Tables 5, 6). Remarks. — Until the present study, guerrerensis has been considered a mono- typic species, although it was long recognized as being closely related to irroratus. However, my study of variation within the species irroratus and the relationship of guerrerensis with irroratus has convinced me that this relationship is best ex¬ pressed by placing guerrerensis as a race of irroratus. The range of guerrerensis is confined to the wet montane forests of the Pacific slope of Sierra Madre del Sur in the vicinity of Omilteme, Guerrero. The nearest known localities of occurrence of other populations of irroratus (subspecies torridus ) are approximately 20 miles to the east in the vicinity of Chilpancingo, Guerrero. The relationship between L. guerrerensis and L. irroratus was investigated using discriminant function analysis. Because only seven specimens of guerrerensis were available for this analysis, the characters were divided into two sets and discriminant multipliers were generated for both sets. Two discriminant scores were produced for each specimen and these were plotted on a scatter diagram (Fig. 1 8). The reference sample of irroratus was taken from sample 1 8, which is composed of specimens from north of the Rio Balsas in Guerrero. Discriminant multipliers produced using the reference samples of irroratus and guerrerensis (Table 10) were used to produce discriminant scores for specimens from the vicinity of Chilpancingo, Guerrero. The discriminant scores for the specimens from Chilpancingo also are plotted on Fig. 1 8. For the first set of characters (total length, length of tail, length of hind foot, greatest length of skull, length of ros¬ trum and length of maxillary toothrow) the reference sample of guerrerensis had discriminant scores that ranged from 44.13 to 46.68, whereas the reference sam- GENOWAYS— SYSTEMATICS OF LIOMYS 109 (TL-TV-HF-GLS-LR-MTR) Fig. 18. — Scatter diagram based upon discriminant scores generated by a discriminant function analysis comparing Liomys irroratus torridus and Liomys irroratus guerrerensis. Two analyses had to be performed because only seven specimens of guerrerensis were available (characters used in each are indicated at the left and bottom of the diagram). The reference sample of torridus (closed circles) was based upon specimens from north of the Rio Balsas in Guerrero and the reference sample of guerrerensis (closed triangles) was based upon specimens from the vicinity of Omilteme, Guerrero. The specimens in the test sample (open circles) were from south of the Rio Balsas in Guerrero in the vicinity of Chil- pancingo. See text for discussion. pie of irroratus had scores ranging from 35.85 to 40.06; for the second set of characters (interorbital constriction, mastoid breadth, depth of braincase, inter¬ parietal width, interparietal length), the reference sample of guerrerensis had discriminant scores of 53.37 to 56.23, whereas irroratus had values of 46.23 to 50.64. When specimens from Chilpancingo were plotted on the scatter diagram, most of the specimens fell within the range of variation of the Liomys irroratus refer¬ ence sample; however, at least two fell outside the range and one of these is nearly perfectly intermediate between the guerrerensis and irroratus samples. The inter¬ mediate specimen (MVZ 106515) from Chilpancingo had discriminant scores of 110 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY Table 10. — Discriminant multipliers resulting from a discriminant function analysis com¬ paring Liomys irroratus torridus and Liomys irroratus guerrerensis. Characters were divided into two sets because of the small sample size of guerrerensis. First set of characters Second set of characters Measurements Discriminant multipliers Measurements Discriminant multipliers Total length 0.060 Interorbital Length of tail -0.037 constriction 0.516 Length of hind foot 0.181 Mastoid breadth 1.913 Greatest length Depth of braincase 2.054 of skull 0.362 Interparietal width -0.368 Length of rostrum -0.321 Interparietal length 0.559 Length of maxillary toothrow 3.219 42.49 and 5 1 .60 and the other specimen (FMNH 47559), also from Chilpancingo, had scores of 41 .32 and 47.92. The latter has a relatively low score for the second set of characters, owing to a relatively narrow interorbital constriction and a shallow depth of braincase. A third specimen (UMMZ 104999) from Chil¬ pancingo is of interest, but was not plotted on Fig. 18 because it lacks measure¬ ments for total length and length of tail. However, if the means for sample 17 are used for these missing measurements, values of 40.17 and 53.18 are obtained; the score for the first set of characters probably is low because the means used underestimate the original measurements of this specimen. The possibility of some intergradation between torridus and guerrerensis as well as biosystematic information that no differences were found between irrora¬ tus and guerrerensis in bacular morphology, karyotypic morphology, or pterygoid structure clearly indicate that the two taxa are closely related. On the basis of this evidence, I feel justified in recognizing guerrerensis as a subspecies of irroratus. Three adult males and three nonpregnant, adult females from the vicinity of Omilteme, Guerrero, had the following measurements, respectively: length of ear, 18, 19, 17, 19, 19, 18.5; weight, 81.7, 89.6, 71.7, 76.6, 65.2, 58.8. Specimens examined (18). — Guerrero: 15 km. SW Chilpancingo, 9000 ft., 2 (TCWC); 1 mi. NW Omilteme, 7260 ft., 2 (USNM); Omilteme, 7300 ft., 8 (2 ENCB, 3 KU, 1 MVZ, 2 USNM); 2 mi. E Omilteme, 6600 ft., 5 (KU); 1 mi. SW Omilteme, 1 (USNM). Marginal records. — Guerrero: 1 mi. NW Omilteme; 2 mi. E Omilteme ; 15 km. SW Chilpancingo; 1 mi. SW Omilteme. Liomys irroratus irroratus (Gray, 1 868) 1868. Heteromys irroratus Gray, Proc. Zool. Soc. London, p. 205, May. 1911. Liomys irroratus , Goldman, N. Amer. Fauna, 34:53, 7 September. 1868. Heteromys albolimbatus Gray, Proc. Zool. Soc. London, p. 205, May; lectotype from La Parada, Oaxaca (Thomas, 1927:552). 1956. Liomys irroratus yautepecus Goodwin, Amer. Mus. Novit., 1757:7, 8 March; holo- type from Rancho Sauce, San Pedro Jilotepec, 5000 ft., Oaxaca. GENOWAYS — SYSTEMATICS OF LIOMYS 1 1 1 Holotype. — Subadult of unknown sex, skin and skull, BMNH 59.7.10.2, from Oaxaca; obtained on an unknown date by A. Salie. Goldman (191 1:53-54) con¬ sidered material from the city of Oaxaca as typical of irroratus and the type local¬ ity is herewith restricted to Oaxaca, Oaxaca. Skin in rather poor condition with the tail broken off and attached to the skin by a string. The left zygomatic arch is broken and the cranium lacks the posterior portion including the interparietal, right bulla, and occipital and basicranial elements. Measurements of holotype. — Interorbital constriction, 9.1; length of nasals, 12.5; length of rostrum, 15.2; length of maxillary toothrow, 6.3; mastoid breadth, 15.2 ±. Distribution. — This subspecies is confined to central and south-central Oaxaca (Fig. 17). It occurs in mountainous areas and in the Valley of Oaxaca, but is not found on the coastal plain of southern Oaxaca where it is replaced by Liomys p ictus. Comparisons. — The nominate subspecies can be distinguished from the races jaliscensis, texensis, and torridus by larger external and cranial size — compare values for sample 15 and 16 ( irroratus ) with samples 26 to 28 (Jaliscensis ), to and 7 to 10 (texensis), and 11 to 14 and 17 to 20 ( torridus ) in Table 3. In addition, samples of irroratus have a large percentage of individuals with the posterior margin of the nasals emarginate (see Table 6) and a small percentage with a trun¬ cate margin, whereas the reverse is true in samples of jaliscensis. Most samples of jaliscensis have a somewhat greater percentage of individuals with the inter¬ parietal divided than do samples of irroratus (see Table 5), but it would be im¬ possible to separate the two subspecies on the basis of this character. Cranial characters of irroratus that tend to distinguish it from texensis are a greater incidence of notching of the posterior margin of the interparietal, a greater incidence of divided interparietal, and a lesser incidence of truncated posterior margin of the nasals (Tables 5, 6). The qualitative cranial differences that distin¬ guish irroratus and torridus are essentially the same as those that distinguish irroratus from texensis, although the percentage of individuals that exhibit notching of the posterior margin of the interparietal is not so different when populations of irroratus and torridus (especially sample 14) are compared as when populations of irroratus and texensis are compared. For comparisons of irroratus with alleni, bulleri, and guerrerensis see the accounts of those subspecies. Remarks. — This large-sized race was the first member of the genus, as it now is understood, to be described. In 1 868, J. E. Gray described Heteromys irroratus based on a single specimen from the state of Oaxaca, and in the same paper he described Heteromys albolimbatus on the basis of two specimens from La Parada, Oaxaca. Since Goldman’s revision of the genus, albolimbatus has been considered a junior synonym of irroratus, a conclusion with which I agree. It is of interest to note that Gray’s description of albolimbatus includes the following statement: “This species is known from the others by the greater softness of the fur, the greater slenderness of the hair, and the abundance of the elongated under-fur. Examination of the lectotype and lectoparatype in the British Museum revealed SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY 1 12 that these specimens were juveniles still in juvenile pelage, which accounts for the unusual nature of the pelage described by Gray. Goodwin (1956a: 7) described a subspecies, L. i. yautepecus, on the basis ot a single specimen from San Pedro Jilotepec in the coastal mountains of southern Oaxaca. Later, he (Goodwin, 1969:148) placed yautepecus as a junior synonym of irroratus. In the preceding univariate and multivariate analyses, sample 1 5 includes specimens believed to be typical of irroratus and sample 1 6 includes specimens from the range of yautepecus. The two samples constantly fell near each other, indicating little difference between them, and I, therefore, agree with Goodwin’s conclusion that yautepecus should be considered a junior synonym of irroratus. The only subspecies of the species with which Liomys irroratus irroratus is potentially in contact is torridus in northern Oaxaca and central Guerrero, but I have seen no specimens from northern Oaxaca that I would judge to be inter¬ grades between the two. Specimens from the vicinity of Chilpancingo, Guerrero, have been assigned to irroratus by some authors (see Hooper and Handley, 1948: 11); although these specimens average somewhat larger than other samples of torridus, their affinities are clearly with them. The exact relationship of these Guerreran specimens and with L. i. irroratus will not be clear until material be- ^ comes available from the mountainous areas of southeastern Guerrero and west- central Oaxaca. The average weights of seven adult males and five adult females (nonpregnant) of the subspecies were, respectively, 65.6 (54.8-76.6) and 54.6 (48.0-62.3); the mean length of ear for these same individuals was 17.4 (16.0-19.0) and 18.8 (17.0-21.0). Specimens examined (136). — Oaxaca: Benito Juarez Natl Park, S side Cerro San Felipe, 5500 ft., 1 (MVZ); Cerro San Felipe, 4 (1 UMMZ, 1 UNAM, 2 USNM); Chivaguela, 1 (AMNH); Ejutla, 1400 m., 3 (UMMZ); 8 mi. SSW Juchatengo, 6300 ft., 8 (MSU); La Cima, Oaxaca-Puerto Escondido Road, 5800 ft., 6 (CAS); La Parada, 2 (BMNH); 2 mi. SE Matatlan, 5950 ft., 4 (KU); Miahuatlan, 1100 m., 8 (UMMZ); 3 mi. W Mitla, 3 (KU); Monte Alban, 4 mi. SW Oaxaca, 6000 ft., 15 (10 MSU, 5 MVZ); Mt. Zempoaltepec, 5 (USNM); 7 mi. NNE Oaxaca, 1 (UMMZ); 6 mi. NNE Oaxaca, 6000 ft., 3 (UMMZ); Oaxaca, 36 (14 FMNH, 1 UMMZ, 21 USNM); 3 mi. ESE Oaxaca, 5 (KU); 4 mi. ESE Oaxaca, 5050 ft., 1 (KU); 10 mi. SE Oaxaca, 5000 ft., 1 (CAS); Km. 177.2 Oaxaca-Puerto Escondido Road, 5800 ft., 1 (CAS); Km. 183 Oaxaca-Puerto Escondido Road, 6000 ft., 1 (CAS); Rio Molino, Oaxaca-Puerto Angel Road, 7100 ft., 4 (CAS); Rio Jalatengo, Km. 178 Oaxaca-Puerto Angel Road, 4275 ft., 3 (CAS); San Lucas Ixcotepec, 1 (AMNH); San Pedro Jilotepec, 5000 ft., 1 (AMNH); Santa Catalina Quieri, 1 (AMNH); Santa Maria Candelaria, 5 (AMNH); Santo Tomas Quieri, 1 (AMNH); Santo Tomas Teipan, 1 (AMNH); Sierra Juarez, Ixtlan, 1 (UMMZ); Sola de Vega, l(UMMZ); 1 mi. E Tlacolula, 5500 ft., 5 (UMMZ); unspecified locality, 1 (BMNH); Yalalag, 1 (USNM); Zapotitlan, 1 (AMNH). Additional records. — Oaxaca (Goodwin, 1969:148): San Andres Chicahuaxtla; San Miguel Suchitepec; Santo Tomas Ocotepec; Yaitepec. Marginal records. — Oaxaca: Sierra Juarez; Yalalag; Mt. Zempoaltepec; San Pedro Jilo¬ tepec; Zapotitlan; Santa Maria Candelaria; Rio Jalatengo, Km. 178 Oaxaca-Puerto Angel Road, 4275 ft.; 8 mi. SSW Juchatengo, 6300 ft.; San Andres Chicahuaxtla; 7 mi. NNE Oaxaca. GENOWAYS— SYSTEMATICS OF LIOMYS 113 Liomys irroratus jaliscensis (J. A. Allen, 1906) 1906. Heteromys jaliscensis J. A. Allen, Bull. Amer. Mus. Nat. Hist., 22:251, 25 July. 1911. Liomys irroratus jalicensis [sic], Goldman, N. Amer. Fauna, 34:60, 7 September. Holotype. — Adult male, skin and skull, AMNH 26325, from Las Canoas, 7000 ft., Jalisco; obtained on 6 August 1905 by J. H. Batty. Both zygomatic arches are broken and the pterygoid processes are missing, but otherwise the skull is in good condition; the skin is in fairly good condition. Measurements of holotype. — The external measurements were recorded in inches. Greatest length of skull, 31.6; interorbital constriction, 8.0; mastoid breadth, 14. 1 ; length of nasals, 1 1.6; length of rostrum, 13.4; length of maxillary toothrow, 5.2; depth of braincase, 9.2; interparietal width, 8.7; interparietal length, 3.7. Distribution. — The subspecies jaliscensis is known from southern, central, and northern Jalisco and extreme southern Zacatecas and Nayarit (see Fig. 17). Comparisons. — From Liomys irroratus texensis, the subspecies jaliscensis is distinguished by its larger, longer cranium (measurements such as greatest length of skull, length of rostrum, and length of maxillary toothrow) and by its proportionately shorter nasals, which average less than those of texensis (com¬ pare samples 26-28 with 1-5, 7-10 in Table 3). Samples of texensis tend to have a greater percentage of individuals with the posterior margin of the interparietal unnotched than do samples of jaliscensis, whereas samples of jaliscensis have a greater percentage of individuals with the interparietal bone divided than do sam¬ ples of texensis (see Table 5). Liomys irroratus jaliscensis can be distinguished from Liomys irroratus tor- ridus by its somewhat larger overall size (compare samples 26-28 with 12-14, 17-20 in Table 3), but the difference is not so great as that between jaliscensis and alleni or torridus and alleni. Samples of jaliscensis have a much greater per¬ centage of individuals with the interparietal divided than do samples of torridus (see Table 5). It should be pointed out that these three subspecies — jaliscensis, texensis, and torridus — are geographically isolated from each other; they repre¬ sent populations of small-to-medium-sized individuals that are separated by populations of larger-sized individuals. For comparisons of jaliscensis with alleni, bulleri, guerrerensis, and irroratus, see the accounts of those subspecies. Remarks. — The specimens from sample 26 present a rather difficult situation taxonomically in that the males fall with samples of alleni in the multivariate analysis, whereas females fall with samples of jaliscensis (Figs. 12-15). Because the holotype and paratypes of jaliscensis are included in sample 26, the appli¬ cation of the name jaliscensis itself needs to be scrutinized. In the univariate analysis (Fig. 9 and Table 3), it can be seen that males of sample 26 definitely fall with samples of alleni in at least seven measurements (total length, length of tail, length of hind foot, greatest length of skull, zygomatic breadth, interorbital constriction, and mastoid breadth), but in at least four other measurements (length of nasals, length of rostrum, length of maxillary toothrow, and depth of braincase) the mean values for males are smaller than values for alleni and are 1 14 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY near those for males of other samples (27, 28) of jaliscensis. With the possible exceptions of zygomatic breadth and interorbital constriction, females from locality 26 have mean values that are similar to those of females from samples 27 and 28 and are much smaller than those for females of alleni. I have applied the name jaliscensis to specimens from southern, central, and northern Jalisco and extreme southern Zacatecas. I decided upon this course of action for several reasons, not the least of which is that it conveys a certain stabil¬ ity to the taxonomy of mice in this area. One of the alternatives to the recogni¬ tion of jaliscensis would be to place it in the synonymy of alleni and describe a new race from central Jalisco where more “typical” specimens of this small-sized subspecies are to be found; I have chosen not to follow this course of action. One good reason for recognition of jaliscensis is that females from southern Jalisco resemble females in other samples that are included in this taxon. Another reason is that the male holotype and paratypic males are relatively small, being of a size similar to males from central Jalisco (greatest length of skull of two paratypes, 32.4 and 32.3). One thing that weakens this argument is the fact that many of the localities included in this sample lie between the type locality and the central Jaliscan localities. The only possible explanation I can offer for this situation (large males re¬ sembling alleni and proportionately smaller females resembling samples from central Jalisco) is that some sort of character displacement is occurring in this population, because Liomys pictus in known from this area and, in fact, was taken along with irroratus at one of the included localities. A similar situation is seen in the zone of overlap between Liomys pictus and Liomys salvini in southeastern Oaxaca and northwestern Chiapas; in these species only the males evinced the effects of the character displacement (see section on specific relationships). This leaves unanswered the question of why the effects of character displacement are not seen in central Jalisco where irroratus and pictus also occur together. Specimens from the eastern margin of the geographic range of jaliscensis are of interest because of possible intergradation with alleni. Four adult specimens from 3 mi. NW Chapala, however, are small-sized and in all characters appear to be more or less typical of jaliscensis. One specimen from 7 km. SE Tonala is a juve¬ nile and the one from 8 mi. E Tizapan is a rather small subadult. These specimens have been assigned to jaliscensis mainly because their geographic origin suggested this would be the best arrangement at present. Additional material may show these suppositions to be erroneous. For additional discussion of intergradation between alleni and jaliscensis in eastern Jalisco and southern Zacatecas, see the account of alleni. There appears to be some confusion about the exact location of the type locality of jaliscensis, Las Canoas, Jalisco, in Allen’s (1906) paper in which the race was described. On page 238, he stated that Las Canoas was “on the table¬ land, about 40 miles west of Tuxpan, at an altitude of about 7000 feet” and at the bottom of the page and the top of the next he stated that collections were made “in southern Jalisco, a little west of Tuxpan.” Elsewhere in the paper, such as in the account of Platogeomys gymnurus, the location is given as “Las GENOWAYS— SYSTEMATICS OF LIOMYS 115 Canoas, near Zapotlan.” Finally in the description of Heteromys jaliscensis the placement of Las Canoas is given as “about 20 miles west of Zapotlan, Jalisco,” and it is this location that is cited by most subsequent authors. The location of Las Canoas at either 40 miles west of Tuxpan or 20 miles west of Zapotlan [ = Ciudad Guzman] seems to me to be highly unlikely especially in the case of the former. Both of these localities are west of other localities where jalis¬ censis is known to occur. It seems more likely to me that the type locality is a small town by the name, shown on some recent maps, that is 10 km. NW Tuxpan and 10 km. SSW Ciudad Guzman. This locality is near several others from which Liomys irroratus is known and is on the tableland east of the highland of Sierra Nevada de Colima on which Tuxpan and Ciudad Guzman are situated. The ear of 10 adult males and females from central Jalisco averaged, respec¬ tively, 15.4 (14.5-17.0) and 15.0 (14.5-16.0) in length. The weights of these same individuals averaged 54.2 (46.0-69.6) and 44.0 (30.8-52.2). Specimens examined (358). — Jalisco: 7 km. S Acatlan, 9 (ENCB); 3 mi. N Amatitan, 4050 ft., 1 (KU); 2 mi. N Amatitan, 4050 ft., 9 (KU); 2 mi. NNW Amatitan, 4000 ft., 2 (KU); T/2 mi. WNW Amatitan, 4100 ft., 6 (KU); Amatitan, 4050 ft., 1 1 (KU); 3 mi. NNW Ameca, 4300 ft., 6 (KU); 7 mi. W Ameca, 4000 ft., 23 (UMMZ); 6 mi. W Ameca, 4300 ft., 1 (UMMZ); Ameca, 10 (USNM); 1 mi. SW Ameca, 4000 ft., 36 (KU); 2 mi. SW Ameca, 4000 ft., 6 (KU); 1 mi. SSE Ameca, 4000 ft., 31 (KU); Arroyo de Gavalan, 4 (AMNH); Atemajac, 9 (USNM); 5 mi. W Atenquique, 6500 ft., 1 (MSU); Barranca de Tule, Sierra Nayarit, 2 (USNM); 6 mi. ENE Bolanos, 5350 ft., 6 (KU); 4 mi. ENE Bolanos, 4400 ft., 1 (KU); Bolanos, 2800 ft., 3 (KU); 1 mi. E Bolanos, 3350 ft., 1 (KU); 3 mi. NW Chapala, 5100 ft., 8 (UMMZ); 10 mi. W Ciudad Guzman, 6500 ft., 4 (1 UA, 3 UMMZ); 2 mi. N Ciudad Guzman, 14 (KU); 9/10 mi. S Ciudad Guzman, 5050 ft., 2 (KU); 3 Vi mi. W Etzatlan, 4400 ft., 1 (KU); Etzatlan, 1 (USNM); 2 6/10 mi. E Etzatlan, 4300 ft., 13 (KU); 3 mi. N Guadala¬ jara, 5100 ft., 6 (KU); 2 mi. N, Vi mi. W Guadalajara, 2 (KU); 5 mi. S Guadalajara, 1 (UA); 8 mi. S Guadalajara, 5050 ft., 2 (UMMZ); 20 km. S Guadalajara, 1 (ENCB); 13 mi. S, 15 mi. W Guadalajara, 17 (KU); 19 mi. SW Guadalajara, 1 (MSU); 27 mi. S, 12 mi. W Guadala¬ jara, 3 (KU); 2 Vi mi. ENE Jazrnrn, 6800 ft., 1 (KU); 9 mi. NW Jocotepec, 5240 ft., 2 (MSU); La Primavera, 12 mi. W Guadalajara, 5600 ft., 1 (UMMZ); Las Canoas, 10 (AMNH); 3 mi. W La Venta, 3 (KU); 2 mi. NW Magdalena, 4500 ft., 1 (KU); 1 mi. NW Mezquitic, 5000 ft., 2 (KU); N slope Nevado de Colima, 8000 ft., 1 (LACM); 6 mi. W San Marcos, 5400 ft., 6 (KU); 3 mi. ENE Santa Cruz de las Flores, 9 (KU); Sierra Nevada, 1 (USNM); 1 mi. NE Tala, 4400 ft., 4 (KU); 3 mi. W Tala, 4300 ft., 5 (KU); 4>/2 mi. W Teuchitlan, 4300 ft., 1 1 (KU); 8 mi. E Tizapan, 1 (MSU); 7 km. SE Tonala, 1 (ENCB); 3 Vi mi. NW Villa Guerrero, 5500 ft., 1 (KU); 3 mi. N Villa Guerrero, 5600 ft., 2 (KU); 4‘/2 mi. W Villa Guerrero, 5200 ft., 1 (KU); 3!/2 mi. WNW Zapoltitic, 5100 ft., 17 (KU); 2 7/10 mi. WNW Zapoltitic, 5000 ft., 3 (KU); Zapotlan, 2 (USNM). Zacatecas: 8 mi. S Moyahua, 5600 ft., 9 (CAS); 1 mi. N Santa Rosa, 3600 ft., 10 (MSU). Additional records. — Jalisco: Arroyo de Plantanar (J. A. Allen, 1906:251); 11 mi. SE Guadalajara (Levine et al., 1958:295); 21 mi. SW Guadalajara (Twente and Baker, 1951: 120). Nayarit: Ojo de Agua (J. A. Allen, 1906:251). Marginal records. — Jalisco: 1 mi. NW Mezquitic, 5000 ft.; 3 mi. N Villa Guerrero, 5600 ft. Zacatecas: 1 mi. N Santa Rosa; 8 mi. S Moyahua. Jalisco: Atemajac; 7 km. SE Tonala; 3 mi. NW Chapala; 8 mi. E Tizapan; 2 mi. N Ciudad Guzman; Zapotlan [= Ciudad Guz¬ man]; 9/10 mi. S Ciudad Guzman ; 2 7/10 mi. WNW Zapotiltic; Las Canoas; 5 mi. W Aten¬ quique; 2 1/2 mi. ENE Jazmin, 6800 ft .; 10 mi. W Ciudad Guzman, 6500 ft.; 27 mi. S, 12 mi. W Guadalajara; 7 km. S Acatldn\ 2 mi. SW Ameca, 4000 ft .; 7 mi. W Ameca, 4000 ft. Nayarit: Ojo de Agua. Jalisco: 6 mi. W San Marcos, 5400 ft.; 2 mi. NW Magdalena, 4500 ft.; Bolanos, 2800 ft. 116 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY Liomys irroratus texensis Merriam, 1902 1902. Liomys texensis Merriam, Proc. Biol. Soc. Washington, 15:44, 5 March. 1911. Liomys irroratus texensis , Goldman, N. Amer. Fauna, 34:59, 7 September. 1911. Liomys irroratus pretiosus Goldman, N. Amer. Fauna, 7 September; holotype from Metlaltoyuca, Puebla. Holotype. — Adult female, skin and skull, USNM 58670, from Brownsville, Cameron Co., Texas; obtained on 19 February 1894 by J. A. Loring. Skin and skull in good condition, except lacrimal bones missing. Measurements of holotype. — Total length, 231; length of tail, 1 14; length of hind foot, 30; greatest length of skull, 31.0; zygomatic breadth, 14.5; interorbital constriction, 8.0; mastoid breadth, 14.1; length of nasals, 12.8; length of maxil¬ lary toothrow, 5.2; depth of braincase, 8.6; interparietal width, 7.0; interparietal length, 4.0. Distribution. — Gulf coastal lowlands of Texas, Tamaulipas, Nuevo Leon, San Luis Potosi, Veracruz, and Puebla (see Fig. 17). The localities in Texas mark the northeastern limit of distribution for the species. Comparisons. — From Liomys irroratus alleni, the subspecies texensis can be distinguished by its much smaller size, both externally and cranially (compare values for samples 1-5, 7-10 with 6, 22, 33, 34 in Table 3). The only qualitative cranial character that tends to distinguish these two subspecies is the division of the interparietal bone, which occurs in low frequency in populations of texensis, excepting those in Veracruz (7, 10), but in more than 30 per cent of the individ¬ uals in adjacent populations of alleni. For comparisons of texensis with bulleri, guerrerensis, irroratus, jaliscensis, and torridus, see the accounts of those subspecies. Remarks. — This small- to medium-sized subspecies is confined to the Gulf coastal lowlands of northeastern Mexico and southern Texas. Intergradation is evident between texensis and alleni for a considerable distance along the eastern foothills of the Sierra Madre Oriental and the eastern edge of the Mexican Pla¬ teau. Specimens from the vicinity of Monterrey (previously assigned to alleni by Goldman, 1911:57) and from west of Allende and Linares in Nuevo Leon are somewhat larger than specimens from the lowlands to the east in Tamaulipas, but they are considerably smaller than specimens of alleni from contiguous popu¬ lations (sample 6). In Tamaulipas, specimens of texensis revealing intergradation with alleni were examined from Rancho Santa Rosa, Hidalgo, Villa Mainero (previously assigned to alleni by Alvarez, 1963:433), several localities to the west of Ciudad Victoria, El Carrizo, and El Encino. In San Luis Potosi, the zone of intergradation is not well delimited because specimens are lacking from critical areas. Specimens from the lowlands and eastern edge of the Plateau in the vicinity of El Salto and Ciudad del Maiz fit best with other populations of texensis, where¬ as specimens from near Rio Verde are clearly alleni. Also in Veracruz, Hidalgo, and Queretaro the width and exact location of the zone of intergradation are not clear, although specimens from Jalpan, Queretaro, are much smaller than others of alleni and could be as easily assigned to one subspecies as the other on the basis of size. However, because a high percentage (45.4) of these individuals have an interparietal bone that is divided, I have assigned them to alleni. GENOWAYS— SYSTEMATICS OF LIOMYS 117 The specimens from Metlaltoyuca, Puebla, the type locality of Liomys irro- ratus pretiosus described by Goldman (1911:58), are, 1 believe, intergrades be¬ tween texensis and alleni. Their relatively large size is explained by that fact. In the univariate and multivariate analyses, those specimens previously assigned to pretiosus clearly fall with typical texensis ; therefore, I regard pretiosus as a junior synonym of texensis. It is true that specimens from the vicinity of the type locality of pretiosus have the anterior portion of the nasals somewhat more in¬ flated than is typical of texensis , but I do not believe this is sufficient to warrant recognition of the race in light of the numerous similarities with texensis. Some geographic variation was noted within the subspecies texensis. Specimens from northern samples in Texas and northern Tamaulipas (1,2) averaged some¬ what larger than specimens from other samples, although those from central Vera¬ cruz (10) were relatively large. The smallest specimens of texensis, on the aver¬ age, were from the vicinity of Ebano, San Luis Potosi (8). Samples from along the Rio Grande in Texas and Tamaulipas (1,2) also differed from others in having a much higher frequency of individuals in which the posterior margin of the inter¬ parietal bone was unnotched. On the other hand, the two samples from Veracruz (7, 10) had a higher frequency (37.5 and 26.1, respectively) of individuals with the interparietal bone divided in half. A specimen (skull only) in the University of Michigan, Museum of Zoology (105005), which was acquired from the Cleveland Museum of Natural History, was supposedly collected at Ingram, Kerr County, Texas in 1914. This locality has not been plotted on Fig. 17 because the place is approximately 260 miles to the north-northwest of the nearest known locality and in a general habitat quite different from that along the Gulf coast of Texas. Also, relatively extensive trap¬ ping has been done in Kerr County by several mammalogists over the years and no other specimens have been obtained. It seems best to consider that this speci¬ men was mislabeled, unless additional material from the area becomes available. Aside from this specimen, the northernmost record for the subspecies is 10 mi. NW Raymondville in Willacy County (see Blair, 1949:201-202; 1952:241). Koopman and Martin (1959:2, 6) reported a subfossil maxillary fragment of Liomys irroratus from a small cave 1 km. S Aserradero del Paraiso, Tamaulipas. From late Pleistocene deposits in the nearby Cueva del Abra, Tamaulipas, Dal- quest and Roth (1970:226) recorded 49 jaws assignable to irroratus. Although I have not examined this material, it is probably referable to L. i. texensis , which commonly occurs in these areas today. Ten adult males and 10 nonpregnant, adult females from coastal regions of central and northern Tamaulipas had ears that averaged, respectively, 14.9 (1 3.5- 17.0) and 15.2 (13.0-18.0) in length and weights that averaged 48.1 (37.9-58.0) and 42.2 (38.0-50.0). Specimens examined (800). — Nuevo Leon: Cerro de la Silla, 1 (USNM); China. 1 (USNM); 20 km. NW General Teran, 900 ft., 5 (UMMZ); 8 mi. W Linares, 1 (MWU); 2l/z mi. W Linares, 1650 ft., 5 (KU); 7 mi. S, 16 mi. W Linares, 2200 ft., 1 (KU); 10 km. SSE Linares, 1 (KU); 20 km. NW Montemorelos, 1 (TCWC); Montemorelos, 4 (USNM); Mon¬ terrey, 9 (USNM); 3 mi. SW Monterrey, 1965 ft., 5 (KU); Rio Ramas, 4 mi. W Allende, 1700 ft., 3 (KU); San Pedro Santiago, 1 (MCZ). 118 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY Puebla: Metlaltoyuca, 7 (USNM); Pahuatlan, 1100 m„ 2 (UMMZ); 1 km. NW Zihua- teutla, 1 (KU). San Luis Potosi: 4 mi. SSW Ajinche, 1 (LSU); 1.5 mi. S Ajinche, 2 (LSU); 20 mi. W Antiguo Morelos (Tamaulipas), 2 (MVZ); Apetsco, 2 (LSU); 10 mi. E, 2 mi. N Ciudad del Maiz, 1 (KU); 4 km. NE Ciudad Valles, 3 (KU); Ebano, 5 (LSU); 14 km. SW Ebano, 10 (LSU); 10 mi. WSW Ebano, 7 (LSU); 19 km. SW Ebano, 8 (LSU); 22 km. SW Ebano, 8 (7 LSU, 1 ROM); 1 mi. E El Abra, 1 (KU); El Banito, 1 (FMNH); 3 mi. W El Naranjo, 1300 ft., 1 (MSU); 0.5 mi. E El Sabinito, 1200 ft., 1 (LSU); El Salto, 22 (18 AMNH, 3 LSU, 1 UNAM); Hacienda Limon, 10 mi. W Ebano, 1 (LSU); Huichihuayan, 1 (LSU); Mira- dor del Moctezuma, 6 km. N Tamazunchale, 350 m., 1 (UNAM); 10 km. E Platanito, 16 (LSU); Puerto del Lobos, 1 (LSU); Rio Axtla, 3 km. W Axtla, 1 (KU); 1 mi. N Tamazun¬ chale, 700 ft., 5 (MSU); Tamazunchale, 1 (AMNH); 3 km. S Tamazunchale, 7 (MWU); 21/a mi. N Tamuin, 1 (LSU); 3 mi. E Tamuin, 2 (LSU); Tanimul, 10 mi. ESE Valles, 2 (AMNH); 4 km. N Valles, 13 (LSU); 3 km. N Valles, 19 (LSU); 30 km. W Valles, 1 (MWU); Valles, 3 (USNM). Tamaulipas: Acuna, Sierra de Tamaulipas, 800 m., 7 (UMMZ); Altamira, 9 (1 KU, 8 USNM); 1 mi. S Altamira, 3 (KU); 2 Vi mi. S Altamira, 2 (TTU); 2Vi mi. SE Altamira, 1 (TTU); 10 km. N Antiguo Morelos, 22 (MWU); 8 km. NE Antiguo Morelos, 500 ft., 2 (KU); Antiguo Morelos, 9 (UMMZ); 3 Vi mi. S Antiguo Morelos, 900 ft., 2 (MSU); near Bagdad, 3 (USNM); Canada del Abra, 15 km. SSW Ciudad Mante, 2 (UNAM); Ciudad Mante, 1 (MWU); Km. 577 carretera Ciudad Mante y Ciudad Victoria, 1 (ENCB); 29 mi. N Ciudad Victoria, 3 (MVZ); 36 km. N, 10 km. W Ciudad Victoria [1 km. E El Barretal on Rio Puri- ficacion], 1 (KU); 15 mi. N Ciudad Victoria, 2 (KU); 15 km. N Ciudad Victoria, 1 (MWU); 12 km. N Ciudad Victoria, 1 (WBC); 70 km. S Ciudad Victoria and 2 km. W El Carrizo, 6 (5 KU, 1 UNAM); Cueva del Abra, 12 km. S Ciudad Mante, 6 (UNAM); Cueva de Quin¬ tero, 12 km. S Quintero, 250 m., 24 (UNAM); Ejido Santa Isabel, 2 km. W Pan American Highway, 3 (KU); 3 mi. W El Carrizo, 1500 ft., 1 (UMMZ); 53 km. N El Limon, and 12 km. S Rio Guayalejo, 4 (KU); 5 mi. NE Gomez Farias, 2 (UMMZ); Gomez Farias, 10 (AMNH); Hacienda Acuna, 2650 ft., 7 (MSU); Hacienda San Jose, near Soto la Marina, 2 (LSU); Hacienda Santa Engracia, 3 (FMNH); Hidalgo, 2 (USNM); 4 mi. N La Pesca, 5 (KU); 8 km. NW La Pesca, 2 (TNHC); 7 mi. W La Pesca, 25 ft., 10 (MSU); 6 mi. W La Pesca, 3 (MSU); 7 km. SW La Purisima, 1 (KU); Marmolejo, San Carlos Mts., 5 (UMMZ); Matamoros, 19 (USNM); 5.7 mi. S Matamoros, 3 (MVZ); Mulato, 1 (UMMZ); Pano Ayuctle, 5 mi. NE Gomez Farias, 2 (AMNH); Rancho Milagro, 11 mi. SW Cruillas, 1 (UMMZ); Rancho Pano Ayuctle, 25 mi. N El Mante, thence 3 km. W highway, 300 ft., 1 (KU); Rancho Pano Ayuctle, 6 mi. N Gomez Farias, 300 ft., 8 (KU); Rancho Pano Ayuctle, 3 mi. NE Gomez Farias, 2 (CAS); Rancho Santa Rosa, 25 km. N, 13 km. W Ciudad Vic¬ toria, 260 m., 2 (KU); 24 mi. ESE Reynosa, 5 (MVZ); Rio Corona, 19 mi. E Ciudad Vic¬ toria, 7 (MCZ); near headwaters of Rio Sabinas, 8 km. W, 10 km. N El Encino, 1 (KU); Rio Soto la Marina, Rancho Santa Ana, 8 mi. SW Padilla, 4 (MCZ); 7 km. S, 2 km. W San Fernando, 8 (7 KU, 1 UNAM); San Jose, San Carlos Mts., 1 (UMMZ); Sierra de Tamaulipas, 10 mi. W, 2 mi. S Piedra, 1200 ft., 37 (KU); Sierra Madre Oriental, 5 mi. S, 3 mi. W Ciudad Victoria, 1900 ft., 18 (KU); Soto la Marina, 35 (25 KU, 6 LSU, 4 USNM); 10 mi. NW Tampico, 1 (KU); 7 km. N Tampico, 2 (KU); Victoria, 6 (USNM); Villa Mainero, 1700 ft., 3 (KU). Veracruz: 1 km. ENE Coyutla, 5 (KU); 3 km. W Gutierrez Zamora, 300 ft., 7 (KU); Ixcatepec, 70 km. NW Tuxpan, 3 (KU); La Mar, 20 ft., 1 (KU); Miahuapa, 8 km. N Coyutla, 80 m„ 6 (KU); Miahuapa (La Tulapilla), 15 km. NNE Coyutla, 80 m., 3 (KU); Nautla, 10 (UMMZ); Ozulama, 500 ft., 3 (KU); Papantla, 4 (USNM); 4 km. E Papantla, 400 ft., 1 (KU); 9 km. E Papantla, 300 ft., 6 (KU); Piedras Clavadas, 75 km. NW Tuxpan, 2 (KU); Platon Sanchez, 800 ft., 4 (KU); Potrero Llano, 350 ft., 4 (KU); 3 km. SW San Marcos, 2 (KU); 5 mi. S Tampico (Tamaulipas), 4 (KU); 12V4 mi. N Tihuatlan, 300 ft., 8 (KU); 4 km. W Tlapacoyan, 1700 ft., 7 (KU); 35 km. NW Tuxpan, 2 (KU); 25 km. NW Tuxpan, 2 (KU); GENOWAYS— SYSTEMATICS OF LIOMYS 119 17 km. NW Tuxpan, 1 (KU); 15 km. NW Tuxpan, 1 (KU); 14 km. NW Tuxpan, 1 (KU); 6 km. N Tuxpan, 1 (KU); 5 km. NE Tuxpan, 1 (KU); 4 km. NE Tuxpan, 3 (KU). Texas: Alamo, Hidalgo Co., 3 (LSU); 8 mi. S Alamo, Hidalgo Co., 1 (USNM); 4 mi. N Brownsville, Cameron Co., 1 (TTU); 10 mi. W Brownsville, Cameron Co., 2 (LSU); Browns¬ ville, Cameron Co., 116 (33 AMNH, 8 FMNH, 2 KU, 12 LACM, 2 LSU, 8 MCZ, 1 SDNHM, 2 TCWC, 1 UCLA, 47 USNM); 20 mi. E Brownsville on Rt. 4, Cameron Co., 1 (UNM); 17 mi. NW Edinburg, Hidalgo Co., 5 (TNHC); 6 mi. N, 1 mi. W Edinburg, Hidalgo Co., 1 (MWU); Elsa, Hidalgo Co., 2 (ROM); Fort Brown, Cameron Co., 1 (USNM); Ingram, Kerr Co., 1 (UMMZ) (not mapped, see text); Laguna Atascosa, ca. 8 mi. NW Bay- side, Cameron Co., 1 (LSU); Laguna Atascosa, 10 mi. E Rio Hondo, Cameron Co., 1 (LSU); Lomita Ranch, Hidalgo Co., 4 (USNM); 6 mi. S McAllen, Hidalgo Co., 27 (TNHC); 5 mi. S Mission, Hidalgo Co., 1 (LSU); no specific locality, Cameron Co., 2 (SDNHM); 10 mi. NW Raymondville, Willacy Co., 4 (TNHC); San Benito, Cameron Co., 1 (LACM). Additional records. — San Luis Potosi: 10 mi. W Ebano (Dalquest, 1953:121); Tamuin (Dalquest, 1953:122). Tamaulipas: 2 mi. S Ciudad Mante (Ingles, 1959:394); Mesa de Llera (Hooper, 1953:5); 3 mi. W Soto la Marina (Hooper, 1953:5). Veracruz: El Tajin (Ingles, 1959:394). Texas: Bentsen State Park, Hidalgo Co. (Hudson and Rummel, 1966: 346); Noriega Wildlife Refuge, 4.5 mi. NW Brownsville, Cameron Co. (Eads et al„ 1965:18). Marginal records. — Texas: 10 mi. NW Raymondville, Willacy Co.; Laguna Atascosa, 10 mi. E Rio Hondo, Cameron Co., thence southward along the coast. Veracruz: Nautla; 4 km. W Tlapacoyan. Puebla: Pahuatlan, 1100 m.; Metlaltoyuca. Veracruz: Platon Sanchez. San Luis Potosi: 3 km. S Tamazunchale; Apetsco; 30 km. W Valles; Puerto del Lobos; El Salto. Tamaulipas: near headwaters of Rio Sabinas, 8 km. W, 10 km. N El Encino; Sierra Madre Oriental, 5 mi. S, 3 mi. W Ciudad Victoria, 1900 ft.; Rancho Santa Rosa, 25 km. N, 13 km. W Ciudad Victoria, 260 m.; 7 km. SW La Purisima; Villa Mainero, 1700 ft. Nuevo Leon: 7 mi. S, 16 mi. W Linares, 2200 ft.; Rio Ramas, 4 mi. W Allende, 1700 ft.; 3 mi. SW Monterrey; Monterrey, China. Texas: 17 mi. NW Edinburg, Hidalgo Co. Liomys irroratus torridus Merriam, 1902 1902. Liomys torridus Merriam, Proc. Biol. Soc. Washington, 15:45, 5 March. 1911. Liomys irroratus torridus, Goldman, N. Amer. Fauna, 34:55, 7 September. 1902. Liomys torridus minor Merriam, Proc. Biol. Soc. Washington, 15:45, 5 March; holotype from Huajuapan de Leon, Oaxaca. 1903. Heteromys exiguus Elliot, Field Columb. Mus., Zool. Ser., 3:146, 20 March; holo¬ type from Puente de Ixtla, Morelos. Holotype. — Adult female, skin and skull, USNM 69645, from Cuicatlan, Oaxaca; obtained on 14 October 1894 by E. W. Nelson and E. A. Goldman. Skin and skull in good condition. Measurements of holotype. — Total length, 242.0; length of tail, 134.0; length of hind foot, 28.0; greatest length of skull, 30.3; zygomatic breadth, 14.2; inter¬ orbital constriction, 7.9; mastoid breadth, 14.0; length of nasals, 1 1.6; length of rostrum, 13.6; length of maxillary toothrow, 4.5; depth of braincase, 8.9; inter¬ parietal width, 8.9; interparietal length, 3.5. Distribution. — The geographic range of Liomys irroratus torridus lies south of the Transverse Volcanic Belt of Mexico in Puebla, Morelos, northern Oaxaca, and northern and central Guerrero. Comparisons. — When values for external and cranial measurements of samples of torridus (12-14, 18-20) are compared with those for texensis (1-5, 7-10), torridus is found to average smaller than texensis in some cranial measurements, but usually the difference is slight (Table 3). A slight difference is also observed 120 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY when both of these subspecies are compared with samples of jaliscensis, but all three are geographically isolated from each other. Samples of texensis , excepting those from Veracruz (7, 10), have a greater percentage of individuals with the posterior termination of the nasals truncate in shape than do those of torridus (Table 6). From Liomys irroratus alleni, the subspecies torridus is distinguished by its much smaller size, both externally and cranially (compare values of samples 12- 14, 18-20 with 21-24, 29, 30, 33-36 in Table 3). Furthermore, samples of tor¬ ridus have a relatively small percentage of individuals (when compared with samples of alleni ) that have the interparietal bone divided and the posterior term¬ ination of the nasals truncated (Tables 5, 6). For comparison of torridus with bulleri, guerrerensis, irroratus, and jalis¬ censis, see the accounts of those subspecies. Remarks. — The subspecies torridus is composed of individuals of medium to small size for the species. It occurs in the dry areas of Puebla south of the Trans¬ verse Volcanic Belt, in northern Oaxaca, Morelos, and northern and central Guer¬ rero. The distribution in Guerrero is of special interest; in this state, the popula¬ tion is split and one segment is presently isolated from the other. In the uplands north of the Rio Balsas Liomys irroratus is the only species of the genus present and this segment is continuous with other populations of torridus in Morelos. As the elevation decreases in the descent into the valley of the Rio Balsas, L. irrora¬ tus is replaced by L. pictus at about 2000 feet. I have seen specimens of both species from the vicinities of Iguala and Los Sabinos on the north edge of the valley and from near Zumpango on the south edge. The population of torridus in the highlands south of the valley of the Balsas, therefore, is isolated from popu¬ lations to the north of the valley by the intervening populations of L. pictus and is evidently contiguous with other torridus in northern Oaxaca and southern Pueblo along the south side of the Rio Mexcala as is indicated by specimens from Ayotzinapa, Sochi, Tlalixtaquilla, and Tlapa. Specimens from the vicinity of Chilpancingo (sample 17) average somewhat larger than individuals from north of the Rio Balsas (sample 1 8) in some measure¬ ments (total length, greatest length of skull, mastoid breadth, and length of ros¬ trum). In these measurements, specimens from sample 17 tend to be intermediate between irroratus and torridus, but in all other measurements they clearly fall with the northern Guerreran material. Sample 17 has a greater percentage of individuals with the interparietal divided than do other samples of torridus and a slightly smaller percentage than do samples of irroratus. The relationships of Guerreran specimens from south of the Rio Balsas is clearly with other popu¬ lations of torridus. The characters in which this population has diverged from torridus may have resulted from some genetic influence from populations of irroratus in Oaxaca; however, this will be verified only when material becomes available from the mountainous areas of southeastern Guerrero and west-central Oaxaca. Populations of Liomys irroratus torridus and Liomys irroratus alleni inter¬ grade along the southern slope of the Transverse Volcanic Belt in Morelos and GENOWAYS— SYSTEMATICS OF LIOMYS 121 Puebla, and in the lower country of eastern Puebla. As pointed out in the uni¬ variate and multivariate analyses, the sample from the vicinity of Tehuacan in eastern Puebla (11) averages intermediate in size between samples of alleni to the northwest and samples of torridus to the west and south. The intermediate size of specimens from this area was pointed out earlier by Hooper and Handley (1948: 12). The percentage of individuals with the interparietal bone divided in sample 1 1 is similar to that of populations of torridus and unlike that of alleni , but the percentage of individuals with the posterior margin of the nasals truncated is intermediate between the values for torridus and alleni (Tables 5, 6). I have as¬ signed this series of specimens to torridus because it fell in the multivariate anal¬ ysis with the other samples of that subspecies rather than with the group con¬ taining the samples of alleni; in analysis for males and females, sample 1 1 fell relatively near sample 17, which also has been included in torridus. Specimens from 7 mi. S and 3 mi. E Puebla and from San Martin in northern Puebla were also judged to be intergrades between alleni and torridus. These specimens are intermediate in size between the two — values for two males from near Puebla and a female from San Martin, respectively, were total length, 245.0, 234.0, 232.0; greatest length of skull, 33.4, 31.6, 32.0; mastoid breadth, 14.6, 14.1, 14.7 — but only one of six specimens had the nasals truncate in shape posteriorly and only one had the interparietal divided, thus suggesting assignment to torridus. In Morelos, specimens from the vicinities of Tepoztlan and Huitzilac are somewhat larger than those examined from elsewhere in the state, but I do not believe they should be included with populations from the Valley of Mexico as was done by Davis and Russell (1954:73) with specimens from 2 mi. SW Tepoztlan. The largest individual from these localities is a female from 1.5 mi. SE Huitzilac (TCWC 4733) with a greatest length of skull of 32.5, which is clearly below the average, 33.3, of female alleni from the Valley of Mexico; I have assigned these specimens, therefore, to torridus. A subadult from 20 km. NE Cuautla (TCWC 4768) is assigned to torridus on the basis of qualitative cranial characters — undivided interparietal bone and emarginate posterior mar¬ gin of nasals. An adult female (FMNH 55985) from Tenancingo, Mexico, a short distance west of several localities in Morelos, has been assigned to alleni because of its large size (greatest length of skull, 33.4; mastoid breadth, 15.1). The status of Liomys irroratus minor was questioned earlier by Hooper and Handley (1948:13) and my analysis reveals that minor is, indeed, indistin¬ guishable from torridus. In the multivariate analysis (see Figs. 12-15), samples from the vicinity of the type localities of minor (sample 13) and torridus (sample 14) are placed close to each other and I can see no possible justification in this material for recognition of two races. Mice in two samples (12, 19) to the north¬ west of the type locality of minor , in southern Morelos, are progressively smaller than material from near the type locality, and average the smallest of the species with the possible exception of specimens from the vicinity of Ebano, San Luis Potosi. There is a name available, exiguus (Elliot, 1903), for the material from southern Morelos, but I do not feel that the differences between it and typical torridus are sufficient to warrant recognition. Also, the differences are clinal in 122 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY nature, with the material from southern Morelos grading smoothly into samples to the southeast and into samples to the north; in neither of these directions have I been able to detect a sharp break in the trend of variation. Hooper and Handley (1948:13-14) hinted at the possibility that the holotype of Liomys irroratus minor is in reality a specimen of Liomys pictus. This brings up the possibility that the name minor could be the proper name for the race that is herein recognized as Liomys pictus plantinarensis. In a discriminant function analysis that will be discussed more fully in the section on specific relationships, discriminant multipliers were generated for three external and nine cranial meas¬ urements (zygomatic breadth excluded) using a reference sample of Liomys pictus from Michoacan and northwestern Guerrero and a reference sample of Liomys irroratus from south of the Rio Balsas in Guerrero (Fig. 48). The discriminant scores for the pictus reference sample ranged from 40.236 to 43.263 and those for the irroratus reference sample ranged from 44.633 to 49.135. When the meas¬ urements of the holotype of minor were multiplied by the discriminant multi¬ pliers a score of 46.051 was obtained. This value is clearly in the range for Liomys irroratus and I believe the holotype of minor is a representative of this species. The length of ear of six adult males and seven adult females of Liomys irrora¬ tus torridus from western Puebla averaged, respectively, 15.7 (14.0-17.0) and 14.9 (13.0-16.0); weights of these same individuals averaged 44.8 (39.0-51.5) and 37.3 (32.0-43.0). Specimens examined (710).— Guerrero: Vi mi. W Acahuizotla, 3000 ft., 1 (UMMZ); 16 km. N Agua del Obispo, 4400 ft., 1 (KU); 9 km. N Agua del Obispo, 3850 ft., 5 (KU); 5!/2 km. N Agua del Obispo, 3400 ft., 17 (KU); 514 km. N Agua del Obispo, 3350 ft., 3 (KU); 5 km. N Agua del Obispo, 3400 ft., 2 (KU); 21/2 mi. S Almolonga, 5600 ft., 8 (6 TCWC, 2 UNAM); Ayotzinapa, 1 (USNM); 7 km. SW Cacahuamilpa, 1 (UNAM); Cahuilotal, Saca- coyuca, 960 m., 14 (5 KU, 9 UNAM); Cerro Chamilpa, 12 km. ESE Chilpancingo, 1400 m., 6 (4 KU, 2 UNAM); 1 km. NW Chapa, 1470 m., 11 (7 KU, 4 UNAM); Chapa, 1470 m., 11 (5 KU, 6 UNAM); 4 mi. W Chilpancingo, 5800 ft., 3 (TCWC); Chilpancingo, 103 (24 FMNH, 3 MCZ, 7 MVZ, 57 UMMZ, 6 UNAM, 6 USNM); 15 km. S Chilpancingo, 4300 ft., 4 (TCWC); Cocula, 1 (ENCB); 15 km. NNE Iguala, 1025 m., 2 (UNAM); Iguala, 730 m., 8 (UMMZ); 3.2 km. SSE Iguala, 970 m., 5 (2 KU, 3 UNAM); La Cofradia, Teloloapan, 2 (KU); Laguna Honda, 1 Vi km. SSW Yerbabuena, 1840 m., 1 (KU); 2 km. ENE Los Sabinos, 1400 m„ 3 (2 KU, 1 UNAM); Mazatlan, 1 (MVZ); Ojo de Agua de Chapa, 5 km. SE Telo¬ loapan, 1520 m., 12 (UNAM); Sacacoyuca, 1 (KU); Sochi, 1 (USNM); 17 km. S Taxco, 4000 ft., 6 (TCWC); 4 3/10 mi. N Teloloapan, 3 (UNAM); Teloloapan. 13 (UMMZ); Tex- calzintla, 6 km. NNW Teloloapan, 18 (10 KU, 8 UNAM); 1 km. SSE Texcalzintla, 1600 m., 8 (3 KU, 5 UNAM); 2 mi. N Tixtla, 4400 ft., 4 (TCWC); 3 km. E Tixtla. 3 (ENCB); Tlalixtaquilla, 2 (USNM); Tlapa, 5 (USNM); 2 km. W Tomatal, 970 m., 1 (KU); 2 mi. W Xochipala, 4400 ft., 16 (MSU); Yerbabuena, 1850 m., 1 (KU); U/2 mi. SE Zumpango, 4000 ft., 1 (TCWC). Morelos: 3 mi. N Alpuyeca, 4000 ft., 4 (TCWC); 3 km. N Alpuyeca, 1 100 m., 1 (ENCB); Alpuyeca, 4000 ft., 5 (TCWC); 1 mi. E Alpuyeca, 3500 ft., 1 (TCWC); 12 km. NW Axochia- pan, 3500 ft., 2 (TCWC); 20 km. NE Cuautla, 6500 ft., 1 (TCWC); 20 km. NE Cuernavaca, 2100 m., 2 (ENCB); Cuernavaca, 8 (USNM); near Cuernavaca, 2 (CAS); 12 mi. S Cuernavaca, 4500 ft., 2 (MVZ); 150 m. SE Huajintlan, 920 m., 3 (UNAM); 14 mi. SE Huajintlan, 3000 ft., 1 (TCWC); U/2 mi. SE Huitzilac, 8000 ft., 8 (TCWC); 2 km. SW Jonacatepec, 4500 ft., 1 (TCWC); Las Animas, Temixco, 1340 m., 1 (UNAM); 2 mi. SW GENOWAYS — SYSTEMATICS OF LIOMYS 123 Michapa, 5000 ft., 1 (TCWC); Puente de Ixtla, 6 (2 FMNH, 4 USNM); V2 mi. W Tehuixtla, 960 m., 1 (UNAM); Temilpa, 8 (4 TCWC, 4 UNAM); 5 mi. W Tepoztlan, 6000 ft., 3 (TCWC); Tepoztlan, 6000 ft., 2 (UMMZ); 2 mi. SW Tepoztlan, 7000 ft., 6 (TCWC); Teques- quitengo, 900 m., 4 (UMMZ); Tetecala, 4 (FMNH); 6 mi. W Yautepec, 4500 ft., 14 (TCWC); Yautepec, 21 (USNM). Oaxaca: 3 km. NNE Cuicatlan, 600 m., 1 (KU); 1 km. N Cuicatlan, 560 m., 3 (UNAM); 1 km. NNW Cuicatlan, 560 m„ 6 (2 KU, 4 UNAM); Cuicatlan, 600 m„ 45 (2 KU, 5 UMMZ, 1 UNAM, 37 USNM); 1 km. S Ciucatlan, 590 m., 2 (1 KU, 1 UNAM); Huajuapan de Leon, 1600 m., 16 (12 UMMZ, 4 USNM); Teotitlan, 950 m„ 71 (1 1 FMNH, 60 UMMZ); Tlapan- cingo, 1 (USNM). Puebla: 1 mi. NNE Acatlan, 3900 ft., 4 (KU); Acatlan, 2 (USNM); Atlixco, 5 (USNM); 3 mi. S Chila, 5900 ft., 1 (MSU); 5 mi. NNW Izucar de Matamoros, 5 (KU); 3 mi. N Izucar de Matamoros, 4250 ft., 4 (KU); 3 mi. W, 1 mi. S Izucar de Matamoros, 9 (KU); 6V2 mi. SW Izucar de Matamoros, 7 (KU); 8 mi. SE Izucar de Matamoros, 3 (CAS); Mata¬ moros, 2100 m., 2 (1 UMMZ, 1 UNAM); Piaxtla, 19 (USNM); 7 mi. S, 3 mi. E Puebla, 6850 ft., 8 (KU); San Martin, 4 (USNM); 20 km. N Tehuacan, 2110 m„ 1 (UNAM); Techuacan, 31 (30 UMNZ, 1 USNM); U/2 mi. W Tehuitzingo, 3570 ft., 5 (KU); Tepanco, 17 (UMMZ); 1 mi. SSW Tilpa, 3700 ft., 9 KU); 4*/2 km. E Totimehuacan, 2150 m., 1 (ENCB); 6 km. E Totimehuacan, 2150 m., 3 (ENCB). Veracruz: 4 km. W Acultzingo, 7500 ft., 1 (KU); Acultzingo, 2 (UMMZ); 4 mi. SW Acultzingo, 7000 ft., 1 (KU). Additional records. — Morelos: 10 mi. N Cuautla (Davis and Russell, 1954:73). Puebla: Amolac (Goldman, 1911:55). Marginal records. — Puebla: San Martin; 6 km. E Totimehuacan, 2150 m. Veracruz: Acultzingo. Oaxaca: Teotitlan, 950 m.; 3 km. NNE Cuicatlan, 600 m.; 1 km. S Cuicatlan, 590 m.; Huajuapan de Leon, 1600 m.; Tlapancingo. Guerrero: Tlalixtaquilla; 5 km. N Agua del Obispo; 4 mi. W Chilpancingo, 5800 ft.; 2 mi. W Xochipala, 4400 ft.; Texcalzintla, 6 km. NNW Teloloapan; Laguna Honda, 1 1/2 km. SSW Yerbabuena, 1840 m.; Yerba- buena, 1850 m.; 7 km. SW Cacahuamilpa. Morelos: 2 mi. SW Michapa, 5000 ft.; Tetecala; Cuernavaca; U/2 mi. SE Huitzilac, 8000 ft.; 20 km. NE Cuernavaca, 2100 m.; 20 km. NE Cuautla. 124 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY Liomys p ictus Painted Spiny Pocket Mouse Liomys pictus occurs along the west coast of Mexico from a place 23 mi. S and 5 mi. E Nogales, Sonora, southward through Sonora, Sinaloa, Durango, Nayarit, Jalisco, Colima, Michoacan, Guerrero, and Oaxaca. In this area the species is generally restricted to the coastal lowlands and adjacent slopes of the Sierra Madre Occidental and Sierra Madre del Sur; however, in Jalisco, Micho¬ acan, and Guerrero one subspecies is restricted to interior valleys and large river systems. Along the Pacific coast, L. pictus occurs as far south as the vicinity of Tonala, Chiapas, but specimens are known from throughout the central valley of Chiapas and from one locality (Nenton) in Guatemala. The species also is known from several localities in the Isthmus of Tehuantepec and from along the east coast from southern Veracruz northward to San Carlos in central Veracruz. Diagnosis External and cranial measurements medium to small for the genus, although some populations ( annectens ) are relatively large in size; cranium relatively nar¬ row in comparison with length; protoloph of upper premolar generally appearing to be composed of a single cusp; three cusps of metaloph connected by loph so as not to form discrete cones; hypocone largest cusp on metaloph; entostyle always connected to hypocone by loph; re-entrant angle on labial margin of lower pre¬ molar not reaching median valley; baculum long and with a small rounded base, distal end of shaft with ventral keel that is laterally compressed and the shaft dorsoventrally compressed posterior to terminal keel; glans penis short when compared with baculum, tip of glans long; glans but slightly sculptured; urethral lappets trilobed; 2N = 48; FN = 66; head of spermatozoon long and with pointed apex; distinct neck between head and midpiece of spermatozoon; wings of pterygoids narrow; parasitized by the anopluran, Fahrenholzia microcephala\ six plantar tubercles on most specimens, although some individuals of L. p. plantinarensis have only five; upper parts reddish brown; lateral stripe generally ochraceous, but may be rather pale; underparts white; hairs on back not curled upward so as to be conspicuous above spines. Comparisons From Liomys spectabilis, sympatric populations of Liomys pictus (subspecies plantinarensis ) can be distinguished easily by much smaller size. There is no overlap in measurements of adults of the two taxa for total length, length of hind foot, greatest length of skull, interorbital constriction, mastoid breadth, length of nasals, and length of rostrum, and specimens of plantinarensis average signif¬ icantly smaller in all other measurements analyzed. Populations of L. pictus (subspecies pictus) from coastal areas of western Jalisco are again distinguished from L. spectabilis by smaller size, although the difference is not as striking as in the case of plantinarensis. Only in greatest length of skull and length of rostrum is there no overlap in measurements of the two taxa, although pictus averages GENOWAYS— SYSTEM ATICS OF LIOMYS 125 smaller in all measurements except for those of the interparietal bone. A useful character in separating these two taxa externally is length of hind foot, which is rarely more than 30 in pictus and rarely less than 30 in spectabilis. The only subspecies of L. pictus that approaches L. spectabilis in size is annectens from the mountains of Guerrero and Oaxaca. From this race, spectabilis can be distin¬ guished by its slightly larger overall size and proportionately shallower cranium. The morphology of the baculum of L. pictus and L. spectabilis is essentially identical but the ventral keel on the distal end of the shaft is shorter in 1 1 speci¬ mens of pictus (0.85-1.25) than in two specimens of spectabilis (both 1.30). Both species have 48 chromosomes, but pictus has one more pair of metacentrics, giving it a fundamental number of 66 as opposed to 64 for spectabilis. The mor¬ phology of the head and neck regions of the spermatozoon of pictus and spec¬ tabilis is similar, but the head of the sperm of pictus is significantly shorter and narrower than that of spectabilis. The dorsal coloration of the two species is sim¬ ilar, but specimens of plantinarensis are paler than spectabilis, whereas speci¬ mens of annectens are darker. Liomys pictus can be distinguished from Liomys salvini, especially in the zone of sympatry, by its much larger overall size (particularly total length, length of tail, interorbital constriction, length of nasals, and interparietal length). The entostyle on the upper premolar is less distinctly separated from the remaining cusps in pictus than in salvini and re-entrant angle on the labial margin of the lower premolar reaches the median valley in specimens of salvini but not pictus. The baculum of pictus has a distal ventral keel that is laterally compressed, which is lacking in salvini-, both species have a section of the shaft of the baculum that is dorsoventrally flattened. The glans penis of pictus can be distinguished from that of salvini by its shorter length in comparison with the length of the baculum, longer tip, and less external sculpturing. The urethral lappets of pictus are trilobed, whereas those of salvini are bilobed. The karyotype of L. pictus reveals 48 chromosomes with a fundamental number of 66, whereas the karyo¬ type of L. salvini has 56 chromosomes with a fundamental number of 86. The spermatozoon of pictus has a long head with a pointed apex as compared with that of salvini, which has a short head with a blunt, broadly rounded apex. Speci¬ mens of L. pictus are parasitized by the anopluran Fahrenholzia microcephala, whereas L. salvini is parasitized by Fahrenholzia fairchildv, the upper parts of pictus are reddish brown with an ochraceous lateral stripe, but salvini has choco¬ late brown or paler upper parts and lacks a lateral stripe; the hairs on the back of specimens of salvini are curled upward and are visible above the spines, but this is not true of specimens of pictus. Liomys pictus can be distinguished from Liomys adspersus by its smaller over¬ all size except for interorbital constriction and width of the interparietal bone, which are usually actually broader in pictus and always proportionally so. The entostyle on the upper premolar is less distinctly separated from the remaining cusps in pictus than in adspersus, and re-entrant angle on the labial margin of the lower premolar reaches the median valley in specimens of adspersus but not pictus. The baculum, glans penis, and urethral lappets of pictus differ from those 126 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY of adspersus in the same ways as from those of salvini. L. pictus has 48 chromo¬ somes with a fundamental number of 66, whereas L. adspersus has 56 chromo¬ somes with a probable fundamental number of 84. The spermatozoon of pictus has a long head with a pointed apex, whereas the head of the spermatozoon of adspersus is short with a blunt, broadly rounded apex. Specimens of pictus are parasitized by F ahrenholzia microcephala, whereas L. adspersus is parasitized by Fahrenholzia fairchildr, the upper parts of pictus are reddish brown with an ochraceous lateral stripe, but adspersus has chocolate brown or paler upper parts with no lateral stripe; the hairs on the back of specimens of adspersus are curled upward and are visible above the spines, however, this is not true of specimens of pictus. For comparisons of Liomys pictus with Liomys irroratus, see the account of that species. Ecology The habitats occupied by Liomys pictus are extremely variable, ranging from Sonoran desert in northwestern Mexico to arid lowland tropics along both west and east coasts and to cloud forest in the mountains of Guerrero and Oaxaca. The known altitudinal range of the species is from sea level to approximately 7300 feet at Omilteme, Guerrero, and Rio Molino, Oaxaca. Liomys pictus appears to prefer moist habitats along rivers and streams in otherwise relatively xeric inland situations; this is especially evident in areas of sympatry with Liomys irroratus. In all such cases except one of which I have knowledge, pictus occurs in the moist lowlands and irroratus in the drier uplands. The one exception is near La Cima, Oaxaca, in a cloud forest situation, where the two species have been taken in what appears to be the same habitat. The other places where the two have been taken together in Jalisco and Guerrero are areas of close contact be¬ tween upland and lowland habitats. Along the arid coasts, the species appears to be more or less generally distributed, although based upon my own experience and information I have gleaned from field notes of collectors, the painted spiny pocket mouse is, even here, more abundant in moist situations along rivers and streams than in the dry thorn forest. No comprehensive study has been under¬ taken on the ecology of this mouse but additional natural history information may be obtained from several other sources, notable among which are Baker and Greer (1962:104), Burt (1938:43), Hall and Dalquest (1963:283-284), Hooper (1955:9), and Wagner (1961:207). Briefly discussed below are 1 1 local¬ ities from which Liomys pictus has been obtained. These localities are from throughout the range of the species and give some idea of the diverse types of situations under which it lives. Alamos, 54 km. E Navajoa, WOO ft., Sonora [approximately 4 miles by road northwest of Alamos according to field notes of collector ]. — P. L. Clifton, field representative from The University of Kansas, collected at this site between 22 and 29 January 1962. The vege¬ tation in the area was subtropical thorn forest; most trees were about 10 feet tall, and so close together that it was difficult to walk between them. At least two species of tall cactus also grew there. Clifton noted that the vegetation near Alamos-closely resembled that found in the vicinity of San Bias and El Fuerte in northern Sinaloa. Traps were placed under GENOWAYS— SYSTEMATICS OF LIOMYS 127 bushes and cacti along the edge of a weedy cornfield. Forty-five small rodents including four Liomys pictus were taken in 131 traps set the first night. The second night another 100 traps were set in this same area and yielded 19 small rodents of which three were Liomys pictus. On this same night Clifton set 55 mouse traps among the rocks on a low tree-clad hill; these traps did not yield Liomys nor did 82 traps set in the same type of habitat on the third night. On the night of 25 January traps again were placed in the same general area as on the first night and trom 120 museum special traps 40 small rodents (of which three were Liomys pictus) were removed. Burt (1938:43) noted that in southern Sonora Liomys was abundant around cultivated fields and their food “consisted chiefly of seeds of cultivated beans and wheat. Small rodents trapped along with Liomys at Alamos included Perogna- thus artus, P. goldmani, P . pernix , Dipodomys merriami, Reithrodontomys fulvescens, Peromyscus eremicus , P . merriami , Onychomys torridus, Sigmodon arizonae, Neotoma albigula, and N. phenax. 1/2 mi. S Concepcion, 250 ft., Sinaloa. — This locality, which is along the Rio de las Canas, was visited by J. K. Jones, Jr., and R. R. Patterson between 26 and 28 February 1961 and by E. C. Birney, L. C. Watkins, and me between 14 and 16 August 1969. The general vegetation in the area was arid thorn forest. Most of the trees were about 10 feet tall and grew in dense stands with open grassy areas between the stands. Above the bed of the river some of the area is cultivated but along the sloping banks of the river was a relatively dense stand of thorny scrub. Traps were placed along the sandy sloping bank in an area of thorn brush and grass adjacent to a cornfield. From this area only Perognathus pernix and Liomys pictus were trapped, with the Sinaloan pocket mouse being the more abundant of the two species. In the general vicinity, specimens of Neotoma alleni and Sigmodon arizonae were also obtained. 11 mi. S W Autlan, 2000 ft., Jalisco. — This locality is on the Pacific slope of the first major range of mountains inland from the coast. When I visited this location on 1 July 1966, and 11 August 1969, during the rainy season, the area was extremely wet; fog hung along the slopes, at least in the morning, and rain occurred at nearly any time of the day. Clifton worked here in the dry season (4 March 1966) and noted “the tropical deciduous forest is still green on the north slopes and in the arroyos.” Traps were set in a deep arroyo that had a small stream running through it. The arroyo was never more than 50 feet wide and deciduous trees (70 to 100 feet tall) and brush grew in the bottom almost to the edge of the stream. Traps were placed at holes in the bases of trees, along the wall of the arroyo where overhanging rock and brush formed a protective cover, and under the many boulders near the stream. In addition to Liomys pictus , specimens of Peromyscus banderanus and Oryzomys palustris were obtained. 5 km. NNW Barro de Navidad, Jalisco. — M. R. Lee collected mammals at this locality from 24 March to 2 April 1961. Lee made his camp at the base of a cliff in a coco-oil palm grove; vegetation above the cliff was low thorn forest. One hundred and fifty snap traps were set among boulders at the base of the cliff where little or no vegetation was actually growing among the rocks. These traps yielded seven Peromyscus banderanus , one Liomys pictus, and three Neotoma alleni. The following night 150 traps were set on the hill above the cliff where the vegetation consisted of sparse grass, several kinds of cactus, and low shrubby trees and bushes; nearly all of the trees were without leaves. Only two Peromyscus banderanus and one Liomys pictus were obtained in these traps. The remaining time at this locality Lee trapped around the edge of a marsh where cattails and long grass were the major vegetation. No spiny pocket mice were captured near the marsh, although many small rodents including Oryzomys palustris, O. melanotis, Peromyscus banderanus, P. per- fulvus, and Sigmodon mascotensis were obtained there. 2 7/10 mi. WNW Zapotiltic, 5000 ft., Jalisco. — This area, which is near Ciudad Guzman, is heavily agriculturized. Most of the land in the valleys is cultivated (mainly with corn) and the hills are grazed; scattered trees that grew on the hills were mostly deciduous types but conifers were found on some of the upper slopes. In the areas between trees, clumps of brush are obtained, especially in slight depressions. Clifton, who trapped at this place on 5 and 6 128 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY March 1964, placed his traps in the weeds along a rock fence around an old cornfield. Only specimens of Liomys pictus and Liomys irroratus were obtained with Liomys pictus being much commoner than irroratus. At a locality only a short distance away — 3 Vi mi. WNW Zapoltitic, 5100 feet — in essentially the same type of habitat, Liomys irroratus was the much commoner species. 5 km. N Agua del Obispo, 3250 ft., Guerrero. — The town of Agua del Obispo is approxi¬ mately 35 km. S Chilpancingo on the road to Acapulco and is situated on the Pacific slope of the Sierra Madre del Sur. The vegetation in the area is characterized by pine-oak forest on the more arid slopes and dense tropical deciduous trees in the valleys. On 11 June 1964, Clifton placed 50 museum special traps baited with walnuts and oatmeal among the trees in one of the wet valleys. From these traps he removed 14 Liomys pictus, one Megasorex gigas, and one Oryzomys palustris. Clifton revisited this locality on the night of 13 June 1964, when he obtained five Liomys pictus and one Megasorex gigas from 100 traps that were baited with tuna fish and placed under rocks and rotten logs. At a nearby locality 5Va km. N Agua del Obispo, 3350 feet — Clifton obtained both Liomys pictus and Liomys irroratus. Specimens of Liomys pictus all were caught in a dense stand of deciduous trees along a creek, whereas L. irroratus was taken in a more open and drier situation where low brush predominated. 10 mi. S Juchatengo, 5350 ft., Oaxaca. — This locality is situated in dense cloud forest along the Pacific Slope of the Sierra Madre del Sur. On the night of 7 July 1964, Clifton placed 28 traps in 10-foot-high brush around the edge of an old cornfield that was situated in a clearing in the forest. From these traps he removed specimens as follows: Liomys pictus, 4; Peromyscus megalops, 5; P. evides, 3; Neotoma mexicana, 1. On the night of 8 July 43 traps were placed around the cornfield and 30 traps were set under rocks and logs near a stream in the cloud forest. These traps yielded the following specimens: Liomys pictus, 6; Peromyscus megalops, 6; Oryzomys palustris, 1; Reithrodontomys sumichrasti, 1. Puerto Angel, Oaxaca. — R. W. Dickerman worked in the area around Puerto Angel from 3 to 9 September 1954 collecting vertebrates for the Museum of Natural History. He de¬ scribed the general area as arid lowland tropics but in the stream valleys the vegetation was nearly as lush as along the streams higher in coastal ranges. On 4 September Dickerman set traps along a hill facing the bay in low xeric vegetation and along a dry stream bed where the vegetation was slightly more lush. Obtained in these traps were thiee Liomys pictus, one Peromyscus mexicanus, and two species of crabs. On the evening of 6 September traps were placed among rocks and at the bases of trees along a large intermittent stream near the town. The only species obtained was Liomys pictus (seven specimens). Trapping the following night in the same habitat yielded three Liomys pictus and one Peromyscus mexi¬ canus. Finca San Salvador, 15 km. SE San Clemente, 1000 m., Chiapas. — From 3 to 7 August 1965, a field party under the direction of James D. Smith collected mammals and their ectoparasites at this location in the valley of the Rio San Clemente. In the valley along the river were planted corn, bananas, and sugar cane and tall grass grew along the banks of the river. On the hills above the valley grew a pine-oak forest with pine becoming predominant on the upper slopes. Smith described this forest as having the appearance of a “parkland.” Some traplines were set in the grass and cane along the river, whereas others were placed from the river up into the forest. Specimens of Liomys pictus were trapped both in the valley and on the upper slopes; however, they appeared, as did other rodents, to be more abundant in the valley. Other species of small rodents collected at this locality in addition to spiny pocket mice included Oryzomys palustris, O. fulvescens, Peromyscus boylii, P. mexicanus, Baiomys musculus, and Sigmodon hispidus. 3 km. E San Andres Tuxtla, 1000 ft., Veracruz.— At this locality a large series of Liomys pictus was obtained by W. W. Dalquest between 7 and 18 January 1948. The mice were taken in a variety of places, including at the edge of dense tropical forest and near a small lake. Although a large number of traps were placed in the forest, no painted spiny pocket mice were taken there. Most of the specimens were trapped in brush and tall grass at the GENOWAYS— SYSTEMATICS OF LIOMYS 129 Fig. 19. — Approximate geographic areas included in the 30 samples of Liomys pictus. See text for localities included in each sample. edge of fields and pastures. Other small rodents recorded from this locality were Oryzomys palustris, O. melanotis, O. fulvescens, Tylomys nudicaudus, Peromyscus mexicanus, Sigmo- don hispidus, and Mus muse ulus. Rio Blanco , 20 km. WNW Piedras Negras, 400 ft., Veracruz. — This locality is situated on the coastal plains of Veracruz where the vegetation can be described as typical of the arid tropics. There were extensive grasslands broken by thickets of dense thorny shrubs, prin¬ cipally bullhorn acacia. Along the arroyos and the Rio Blanco was extremely dense decid¬ uous forest. W. W. Dalquest trapped this area from 11 to 29 May 1946. Although trapping in the bunchgrass and acacia did yield a few small rodents, Dalquest found the areas of dense vegetation in the arroyos and along the river much more productive. Only one Liomys pictus was taken in the extensive area of bunchgrass; it was taken in a trap set on sandy soil between the bunches of grass and at least 100 yards from the nearest patch of brush and even further from the nearest forested area. Four nights of trapping in the grass and acacia shrub yielded only four specimens of Liomys pictus , whereas two nights of trap¬ ping in the forest along an arroyo yielded six specimens. The ground in the forest was dry and relatively free of underbrush; however, there was considerable dry grass and a cover of dry leaves. Other small rodents taken at this locality were Reithrodontomys fulvescens , Peromyscus mexicanus , Baiomys musculus , and Sigmodon hispidus. 130 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY Geographic Variation Univariate Analysis Nearly all adult specimens from throughout the geographic range of Liomys pictus were grouped into 28 samples for males and 29 samples for females as follows (Fig. 19): sample 1 — Sonora (Alamos, Camoa, El Novillo, Guirocoba, Matape, Nacori, Nogales, Rio Alamos, Rio Cuchahaqui), Chihuahua (labeled with reference to Choix, Sinaloa), and Sinaloa (Choix, El Fuerte); sample 2 — Sinaloa (Agua Nueva, Cosala, Culiacan, El Dorado, Guamuchil, Guasave, Pericos, Sinaloa) and Durango (Chacala); sample 3 — Sinaloa (Camino Real, La Cruz, Mazatlan, San Ignacio, north of Villa Union); sample 4 — Sinaloa (Copala, Hacienda San Jose, Santa Lucia) and Durango (Pueblo Nuevo); sample 5 — Sinaloa (Concepcion, Escuinapa, Isla Palmito de la Virgen, Isla Pal- mito del Verde, La Concha, Matatan, Plomosas, Rosario, Teacapan) and Nayarit (Acaponeta, Huajicori, Playa Novilleros); sample 6 — Nayarit (Plantanares, Rosamorada, Tuxpan); sample 7 — Nayarit (San Bias); sample 8 — Nayarit (Compostela, Ixtlan del Rio, Santa Isabel, Tepic); sample 9 — Nayarit (Ama- tlan de Canas, Ojo de Agua, Rancho Palo Amarillo) and Jalisco (Guadalajara); sample 10 — Nayarit (Las Varas, San Francisco, Valle de Banderas) and Jalisco (La Cuesta, Milpillas, Puerto Vallarta, San Sebastian); sample 11 — Jalisco (Atenqueque, Contla, Jilotlan de los Dolores, Pihuamo, Platanar, Tamazula, Zapoltitic) and Michoacan (Los Reyes); sample 12 — Jalisco (Autlan, Durazno, La Resolana, Limon, Purificacion, Tapalpa, Tecomate); sample 13 — Jalisco (Barro de Navidad, Cihuatlan, Cuitzamala, Melaque, Tenacatita) and Colima (Armenia, La Gloria, Paso del Rio, Santiago, Tlapeixtes); sample 14 — Colima (Cerro Chino, Colima, Hda. Magdelena, Las Juntas [5 km. SE Pueblo Juarez], Pueblo Juarez); sample 15 — Michoacan (Arteaga, La Mira — this sample in¬ cludes only two adult males); sample 16 — Michoacan (Apatzingan, La Huacana, La Salada, Tacambaro, Tumbiscatio, Tzitzio) and Guerrero (El Limon); sam¬ ple 77 — Guerrero (Iguala, Los Sabinos, Rio Balsas); sample 18 — Guerrero (Acapulco, Acapulco Bay, near Ometepec, Sihuatanejo Bay, Zihuatanejo); sam¬ ple 79 — Guerrero (Acahuizotla, Agua del Obispo, Chilpancingo, Colotlipa); sample 20 — Oaxaca (km. 123 Tlaxiaco-Putla Road, La Concepcion, San Vin¬ cente — this sample contains only females); sample 21 — Oaxaca (Llano Grande, Jamiltepec, Pinotepa, Pinotepa de Don Luis, Puerto Angel); sample 22 — Oaxaca (Jamaica Junction, Juchatengo, km. 136 Oaxaca-Puerto Escondido Road, km. 183 Oaxaca-Puerto Escondido Road, km. 198 Oaxaca-Puerto Escondido Road, km. 2\2Vi Oaxaca-Puerto Escondido Road, km. 214 Oaxaca-Puerto Escondido Road, Lachao, Nopala, San Gabriel Mixtepec, Sola de Vega); sample 23 — Oaxaca (Candelaria, Chacalapa, Pluma, Rio Guajolote, Rio Molino); sample — 24 _ Oaxaca (Aguaje las Animas, Aguaje Tres Cabezas, Escuranos, Guiengola, Juchitan, La Ventosa, Nejapa, Salazar, Salina Cruz, San Jose Lachiguiri, Santa Cruz Bay, Santa Lucia, Santa Maria Ecatepec, Tehuantepec); sample 25 — Oaxaca (Reforma, Santa Efigenia) and Chiapas (Madre Mia, Tonala); sample 26 _ Chiapas (Cerro Tres Picos, Chiapa de Corzo, Cinco Cerros, Cintalapa, Comitan, La Trinitaria, Ocozocoautla, Ortiz Rubio, San Clemente, San Gregorio, GENOWAYS— SYSTEMATICS OF LIOMYS 131 Tuxtla Gutierrez); sample 27 — Oaxaca (Ixcuintepec, Lagunas, Santo Domingo); sample 28 — Veracruz (Catemaco, Coatzacoalcos, Jimba, Pasa Nueva, San Andres Tuxtla); sample 29 — Veracruz (Boca del Rio, Carrizal, Otatitlan, Piedras Negras, Presidio, Puente Nacional, San Carlos); sample 30 — Guerrero (Omilteme — this sample contains only females). Dice grams (Fig. 20) have been prepared for several of the measurements in order to facilitate the demonstration of trends in geographic variation. Standard statistics are given for each sample of Liomys p ictus in Table 1 1 . External Measurements Specimens from Sonora and northern Sinaloa (sample 1) have a mean total length (see Fig. 20 and Table 1 1) that is intermediate for the species. Progressing southward along the coast of Sinaloa and Nayarit from sample 1 to sample 6, ex¬ cepting sample 3, there is a relatively smooth cline with means becoming smaller with decreasing latitude. Specimens from the vicinity of Mazatlan, Sinaloa (sample 3), are somewhat smaller than would be expected at that point in the cline. The mean total length of specimens from near San Bias, Nayarit (sample 7), is much greater than that of specimens to the north in the lowlands of Nayarit and more closely resembles that of specimens to the south along the southern coast of Nayarit and on the coast and Pacific slope of the Sierra Madre of Jalisco and Colima (samples 10, 12-14). Specimens from the vicinity of Tepic, Nayarit (sample 8), have a mean total length that is intermediate between that of individ¬ uals in the lowlands to the north (6) and those along the coast to the west (7). This is also true of males from southeastern Nayarit and north-central Jalisco (sample 9), although females from the same area are relatively larger and, there¬ fore, approach specimens to the west. Specimens from the interior drainage basins of Jalisco and Michoacan and along the Rio Tepalcatepec, Michoacan, and Rio Balsas, Guerrero, and their tributaries (samples 11, 16, 17) form a unit char¬ acterized by uniformly short total length. They are, except for some individuals from sample 6, the smallest of the species. Specimens from coastal Guerrero (sample 18) and along the Pacific slope of the Sierra Madre del Sur (sample 19) are comparable with those from Colima. Females from the southwestern coast of Oaxaca (sample 21) resemble those from the adjacent part of Guerrero (18), but males from Oaxaca are much shorter in mean total length than their counterparts in Guerrero. Individuals from the high mountains of Guerrero (sample 30, fe¬ males only) and Oaxaca (samples 22 and 23) have a total length that is (excepting some individuals from sample 25) the longest for the species. Four females from the vicinity of Putla, Oaxaca (sample 20), which is a mountainous area between samples 30, 22, and 23, are smaller than females from adjacent mountainous areas and more nearly resemble specimens from coastal regions of Guerrero and Oaxaca. Specimens from the isthmian region of Oaxaca, Chiapas, and Veracruz (samples 24-29) form a unit that is similar in total length to samples of both sexes from coastal Guerrero (18) and to females from southwestern Oaxaca (21). Fe¬ males from Los Tuxtlas and adjacent areas of Veracruz (28) have a mean total 132 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY length that is relatively short as compared with other females from the region of the Isthmus, but this is not shown by males from the same area. Populations of Liomys pictus exhibit essentially the same pattern of variation in length of the tail (Table 1 1 ) as they did in total length. However, the smooth cline seen in the variation of mean total length along the northwestern coast (samples 1 -6) is not shown by mean tail length, although specimens of both males and females from the southern part of this area do average smaller than those from the northern part. Again, the mean of specimens from sample 7 is large and more nearly comparable to samples to the south (10, 12-14) than to those to the north (4-6). Length of tail of males and females from sample 9 and males from sample 8 average intermediate between those of northern Nayarit (6) and western Nayarit and Jalisco (7, 10). Females from central Nayarit (8) have a mean length of tail that is slightly shorter than that of specimens from northern Nayarit. Specimens from samples 11, 16, and 17 have a mean length of tail that is the shortest of the species. Specimens from the remainder of the samples in western Mexico, the Isthmus in Oaxaca, Chiapas, and Veracruz, with a few exceptions, have a mean length of tail that is essentially the same. Males from sample 21 and females from sample 28 have a shorter measurement than would be expected and this was also shown in the mean total length. Only specimens from sample 23 and the three females from sample 30, of those occurring in the mountains of Oaxaca and Guerrero, have a tail that averages noticeably longer than in speci¬ mens from the coastal lowlands; these specimens are the largest of the species. The variation in length of hind foot (Table 1 1) of Liomys pictus follows closely the pattern of variation shown by the other two external measurements. The pat¬ tern of a smooth cline in Sonora, Sinaloa, and northern Nayarit is present again except that individuals from sample 4, which includes specimens from high ele¬ vations in Sinaloa and Durango, are the smallest of the group. Specimens from the vicinity of San Bias (sample 7), although larger than specimens from northern Nayarit, average smaller than those from samples 10 and 12 to 14 and more closely resemble specimens from samples 8 and 9. These three samples form a group intermediate in size between the smaller specimens to the north and the larger to the south. The specimens from interior Jalisco, Michoacan, and Guer¬ rero (samples 11, 16, 17) have, on the average, small hind feet, with those from sample 1 7 being the smallest of the species. The specimens from the mountains of Oaxaca (22, 23) and Guerrero (30), excepting sample 20, average more than 30 in length of the hind foot and are the largest of the species. The remaining sam¬ ples (1 8-21 , 24-29) include specimens with hind feet that average an intermediate size for the species and agree closely with specimens from western Jalisco and Colima. The males from sample 25 are larger than those from nearby samples as they were in total length, and females from sample 28 are smaller than those from surrounding samples as they were in total length. p,G 20. _ Dice grams for seven selected measurements of males and females of Liomys pictus. See Fig. 9 for method used in constructing these grams. See Fig. 19 and text for key to samples. GENOWAYS— SYSTEM ATICS OF LIOMYS 133 2>0 220 230 240 250 260 270 280 134 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY GREATEST LENGTH OF SKULL MALES + 1-17 17-5 -\ - 1- 2-7 — 3-13 7- 6 8- 6 9-7 10-20 11-20 12-19 13-15 14-6 16-17 18-3 19-12 — 21-6 22-13 E3 23-6 - 24-27 =3 - 25-9 27-3 26-14 28-16 29-8 - H 28 29 30 31 32 33 34 35 Fig. 20. — Continued. GENOWAYS— SYSTEMATICS OF LIOMYS 135 136 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY 12-19 18-5 I mlpm q — 19-14 LENGTH OF NASALS MALES 9.0 10.0 11.0 12.0 13.0 14.0 15.0 16.0 Fig. 20. — Continued. GENOWAYS— SYSTEMATICS OF LIOMYS 137 LENGTH OF NASALS FEMALES 1-16 2-11 3 — 3-14 - 4-21 5-18 6-12 7-9 8-14 - 9-10 10-19 11-27 12-18 13-18 ] 14-9 16-23 17-6 18-17 - 19-11 20-4 21-8 - 22-24 23-6 — i - 1 — 15.0 16.0 9.0 10.0 11.0 12.0 13.0 14.0 Fig. 20. — Continued. 138 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY 1-17 2- 7 3-19 4-12 1 + .3 5-24 - 6-11 7- 6 8-6 9-7 - 10-20 11-21 - 12-19 3 - 13-15 14-6 16-18 ] - 17-5 18-5 — 19-13 ] - 21-6 3 - 22-13 27-1 28-17 29- 7 INTERPARIETAL LENGTH MALES - 1 - 1 - ( - 1 - 1 - 1 - 1 - 1 - 1 - i - > ♦ — >■ ■ i - 1 - 1 - 1- 2.6 3.0 3.4 3.8 4.2 4.6 Fig. 20. — Continued. 5.0 5.4 5.8 GENOWAYS — SYSTEMATICS OF LIOMYS 139 INTERPARIETAL LENGTH FEMALES 1-16 2-11 3-15 4-20 5- 18 6- 12 7-9 8- 13 9-9 11-27 16-25 ■El 10-16 — 12-19 13-17 14-8 18-13 19-12 [J 20-4 21-8 22-20 24-19 23-8 27-6 28-16 29-9 30-3 H - 1 - 1 - 1 - 1 - (- 5.4 5.8 Fig. 20. — Continued. 140 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY 1-13 2-6 3-16 a ZYGOMATIC BREADTH MALES 4-13 5-21 6-10 7-3 8- 5 9- 5 10-16 1 1-19 12-17 - 13-11 ih}- 14—6 16-12 17-4 9 — 13 21-3 2 2-11 ■ 23-2 24-19 25-7 26-11 2 7-2 28-12 29-5 -t - » - 1 - 1 - 1 - 1- 13.0 13.5 14.0 14.5 15.0 15.5 Fig. 20. — Continued. 16.0 16.5 GENOWAYS— SYSTEMATICS OF LIOMYS 141 142 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY MASTOID BREADTH MALES -I - L 1-17 2-7 3-18 -4-13 - 5-25 — 6-11 3 - 7-6 8-6 9-7 3— 10-20 1 1-20 12-19 - 13-15 F(- 14-6 16-18 B - 17-5 18-3 19-14 21-6 2 2-12 23-7 — 24-28 — 25-10 — 26-16 27-3 28-15 29-7 12.5 13.0 13.5 14.0 145 Fig. 20. — Continued. H - 1 - 1 - 1 - 1- 15.0 15.5 16.0 GENOWAYS— SYSTEMATICS OF LIOMYS 143 1-16 2-1 1 3-15 4-21 - 5-19 - 6-12 7-9 8-14 — 9-10 - 10-16 1 1-2 7 12-19 - 13-17 14-8 16-25 17-5 18-14 19-12 20-4 21-8 2 2-20 23-8 24-20 25-9 26-19 2 7-6 28-16 MASTOID BREADTH FEMALES 4 - 1 - 1 - ^ ] - 29-9 30-4 4 - ^ 12.5 13.0 13.5 14.0 14.5 15.0 15.5 160 Fig. 20. — Continued. 144 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY 1-16 g— 2-7 }- 3-18 4-1 1 — 5-21 24-26 2 5-9 26-15 DEPTH OF BRAINCASE MALES — , - — i - - - 1 - 1 - ' - <- 7.0 7.5 8.0 8.5 9.0 9.5 Fig. 20. — Continued. —i — 10.0 GENOWAYS— SYSTEMATICS OF LIOMYS 145 1-16 5- 16 6- 12 7-8 8-13 - 9-9 10-12 1 1-23 12-18 13-16 16- 23 17- 4 18-14 19-12 20-4 f 21-8 2 2-20 - 23-8 24- 17 25- 10 26-16 DEPTH OF BRAINCASE FEMALES — i - 1 - 30-3 7.0 7.5 8.0 8.5 9.0 Fig. 20. — Continued. 10.0 146 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY Table 1 1. — Geographic variation in external and cranial measurements among 30 samples of Liomys pictus. See text for key to sample numbers. Sample _ number N Mean 1 14 239.2 2 8 238.1 3 18 226.8 4 9 234.0 5 24 229.6 6 10 217.6 7 6 239.7 8 4 226.8 9 3 228.7 10 22 242.0 1 1 22 215.5 12 16 238.4 13 10 243.7 14 5 250.2 16 12 218.7 17 4 214.0 18 4 254.0 19 13 241.2 20 21 6 229.3 22 13 263.5 23 5 279.2 24 23 251.1 25 9 262.8 26 14 245.3 27 28 13 237.5 29 9 242.9 30 1 14 119.6 2 8 123.1 3 18 116.3 4 9 121.6 5 24 116.5 6 10 108.7 7 6 119.2 8 4 113.0 9 3 1 13.0 10 22 126.0 1 1 22 106.0 12 16 122.3 13 1 1 125.3 14 5 132.4 16 12 111.5 Males Females Range 2 SE N Mean Range 2 SE Total length 211.0-262.0 7.83 14 233.8 221.0-257.0 5.26 230.0-251.0 4.99 11 221.7 207.0-244.0 6.60 202.0-259.0 5.81 13 217.5 207.0-229.0 3.68 220.0-244.0 5.82 15 224.3 205.0-241.0 4.48 208.0-265.0 4.70 16 219.8 198.0-232.0 4.52 196.0-232.0 6.66 11 214.9 198.0-238.0 7.70 231.0-260.0 8.53 7 229.0 217.0-240.0 6.78 218.0-239.0 9.71 11 220.5 206.0-232.0 4.80 225.0-231.0 3.71 8 226.3 212.0-240.0 6.01 226.0-262.0 4.48 18 232.2 210.0-260.0 5.65 203.0-239.0 4.39 22 207.3 183.0-240.0 4.75 218.0-260.0 6.39 14 225.8 212.0-248.0 5.38 232.0-264.0 7.80 14 232.1 220.0-248.0 4.05 237.0-270.0 13.56 6 231.0 208.0-251.0 13.41 201.0-240.0 6.33 22 211.3 197.0-225.0 3.54 207.0-224.0 7.44 5 203.2 195.0-207.0 4.21 235.0-267.0 14.12 13 239.2 227.0-253.0 4.16 229.0-251.0 3.81 10 239.2 230.0-247.0 3.92 4 237.3 219.0-245.0 12.28 215.0-240.0 8.92 6 232.2 220.0-240.0 6.08 247.0-286.0 5.90 23 249.0 228.0-275.0 5.41 260.0-300.0 12.92 6 256.3 242.0-281.0 13.78 225.0-289.0 5.76 12 234.8 205.0-246.0 6.59 242.0-290.0 10.68 5 239.2 225.0-255.0 10.51 220.0-269.0 8.65 19 234.0 218.0-252.0 4.51 2 232.5 227.0-238.0 1 1.00 218.0-268.0 8.57 13 220.8 214.0-227.0 2.25 221.0-257.0 7.18 9 234.2 216.0-255.0 7.56 3 289.7 283.0-294.0 6.77 Length of tail 91.0-142.0 7.78 14 117.5 110.0-132.0 3.10 116.0-137.0 4.62 11 110.5 97.0-132.0 6.15 98.0-139.0 4.46 13 112.1 103.0-123.0 3.10 113.0-128.0 4.00 15 115.1 105.0-130.0 3.41 101.0-138.0 3.16 16 112.6 97.0-125.0 4.05 101.0-120.0 3.54 1 1 110.2 101.0-125.0 4.69 107.0-132.0 7.56 8 1 18.8 1 10.0-126.0 3.98 106.0-117.0 4.97 1 1 109.7 98.0-118.0 3.67 1 10.0-115.0 3.06 9 113.6 102.0-125.0 4.51 105.0-146.0 3.86 18 122.1 110.0-145.0 4.46 95.0-120.0 3.27 22 102.8 92.0-124.0 3.25 108.0-138.0 4.13 14 116.4 108.0-128.0 3.12 1 17.0-135.0 4.25 15 118.2 1 10.0-127.0 3.01 120.0-152.0 12.14 6 118.8 109.0-132.0 7.03 97.0-128.0 5.50 22 108.2 101.0-115.0 2.08 GENOWAYS— SYSTEMATICS OF LIOMYS 147 Table 11. — Continued. 17 4 106.8 105.0-1 1 1.0 2.87 5 103.0 98.0-105.0 2.61 18 4 135.5 120.0-145.0 11.68 13 126.5 1 19.0-141.0 3.29 19 13 1 17.0 1 1 1.0-123.0 2.35 10 119.0 1 10.0-130.0 3.83 20 4 120.0 103.0-129.0 11.83 21 6 1 13.5 105.0-123.0 5.95 6 1 19.5 1 14.0-123.0 2.77 22 13 135.9 1 13.0-149.0 5.18 23 128.5 1 13.0-147.0 3.66 23 5 153.0 140.0-165.0 8.27 7 134.4 122.0-148.0 7.10 24 23 131.3 95.0-149.0 4.46 12 128.4 108.0-139.0 4.53 25 9 142.7 126.0-163.0 8.46 5 128.2 1 17.0-139.0 8.03 26 14 125.1 100.0-142.0 6.26 19 122.1 106.0-142.0 4.78 27 2 125.5 121.0-130.0 9.00 28 13 1 17.3 106.0-140.0 5.36 13 110.6 105.0-1 17.0 1.93 29 9 125.0 1 12.0-137.0 5.71 9 121.0 111.0-126.0 3.32 30 3 161.3 154.0-168.0 8.1 1 Length of hind foot 1 16 28.6 25.0-32.5 0.93 14 27.7 22.0-30.0 0.97 2 8 27.3 26.0-28.0 0.60 11 27.7 25.0-29.0 0.68 3 20 26.4 24.0-28.0 0.42 15 25.9 23.0-27.0 0.67 4 12 25.8 23.0-28.0 0.86 20 25.8 23.0-28.0 0.56 5 25 26.2 24.0-28.0 0.44 18 26.4 25.0-28.0 0.51 6 1 1 25.9 23.0-29.0 1.13 13 26.0 24.0-29.0 0.96 7 7 27.1 26.0-28.5 0.70 9 26.8 25.0-28.5 0.78 8 6 26.3 25.0-28.0 0.84 14 26.1 24.0-28.0 0.66 9 4 26.5 26.0-27.5 0.71 9 27.1 26.0-29.0 0.75 10 24 28.9 26.0-30.5 0.44 19 28.4 27.0-31.0 0.48 11 22 25.9 24.5-27.0 0.38 28 25.4 24.0-27.0 0.34 12 19 28.6 27.0-30.0 0.35 19 28.1 26.0-30.0 0.48 13 13 28.7 26.0-31.0 0.97 17 28.6 25.0-31.0 0.89 14 5 29.2 28.0-30.0 0.75 9 27.7 24.0-31.0 1.49 15 2 30.0 29.0-31.0 2.00 16 16 25.2 23.0-27.0 0.46 23 24.8 23.0-27.0 0.43 17 5 24.0 23.0-25.0 0.63 5 24.2 22.0-25.0 1.17 18 5 29.2 28.0-31.0 1.17 13 28.6 27.0-31.0 0.66 19 16 28.2 27.0-30.0 0.42 12 28.0 25.0-30.0 0.69 20 4 28.8 27.0-30.0 1.50 21 6 28.3 26.0-29.0 0.96 7 28.6 26.0-30.0 1.17 22 13 30.2 28.0-32.0 0.65 24 30.1 28.0-32.0 0.42 23 7 31.1 26.0-33.0 1.87 7 30.9 27.0-35.0 2.1 1 24 24 28.7 27.0-31.0 0.57 18 28.6 25.0-32.0 0.79 25 10 30.9 28.0-36.0 1.49 8 28.8 26.0-31.0 1.12 26 13 28.7 27.0-31.0 0.65 19 27.5 25.0-30.0 0.55 27 2 25.0 24.0-26.0 2.00 28 17 28.5 26.0-31.0 0.64 16 27.4 25.0-29.0 0.63 29 9 29.0 26.0-30.0 0.88 9 28.9 26.0-32.0 1.31 30 4 34.9 34.5-35.0 0.25 Greatest length of skull 1 17 32.3 30.3-34.5 0.58 16 31.8 30.8-34.0 0.43 2 7 31.7 31.2-32.4 0.33 1 1 31.3 29.9-32.8 0.57 3 17 31.2 30.0-32.7 0.34 14 30.5 29.6-31.5 0.31 4 13 31.0 29.8-32.7 0.49 20 30.6 29.8-31.4 0.19 5 25 30.5 28.6-31.8 0.34 18 30.0 28.1-31.4 0.40 6 11 29.8 27.6-32.4 0.86 12 30.1 27.6-32.1 0.82 148 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY Table 11. — Continued. 7 6 31.8 30.6-32.9 0.64 9 31.6 21.0-32.3 0.35 8 6 31.2 29.7-32.9 1.08 14 30.6 29.4-31.8 0.35 9 7 31.1 30.2-31.8 0.47 9 31.4 30.3-32.8 0.61 10 20 32.1 30.3-33.2 0.33 16 31.7 30.4-32.8 0.40 11 20 30.4 28.9-32.0 0.42 26 29.6 28.0-31.4 0.27 12 19 32.0 30.8-33.5 0.36 18 31.3 30.1-33.0 0.36 13 15 31.9 28.1-34.1 0.74 17 31.7 30.2-32.9 0.29 14 6 32.0 31.1-33.5 0.76 8 31.4 28.5-32.8 1.17 15 2 35.1 34.8-35.4 0.60 16 17 30.0 28.1-33.2 0.63 22 29.0 26.0-31.5 0.57 17 5 29.1 27.1-30.7 1.20 5 28.4 26.9-29.1 0.78 18 3 33.3 33.0-33.5 0.29 13 32.0 30.6-33.4 0.43 19 12 33.0 32.1-34.6 0.44 12 32.7 31.6-33.4 0.30 20 4 33.0 32.2-34.1 0.84 21 6 32.0 31.4-32.9 0.52 8 31.3 30.3-32.1 0.46 22 13 34.3 32.5-36.7 0.53 20 33.6 31.8-36.0 0.50 23 6 33.2 31.4-34.3 0.90 7 33.3 31.2-35.0 0.92 24 27 32.4 30.8-34.5 0.42 19 31.5 30.4-33.0 0.33 25 9 33.2 31.6-35.0 0.74 10 32.3 31.4-33.7 0.52 26 14 32.7 31.4-34.4 0.48 17 32.0 30.2-33.4 0.44 27 3 31.9 31.3-32.4 0.64 5 31.4 30.8-32.3 0.52 28 16 32.8 31.3-34.6 0.39 15 31.6 30.7-32.7 0.30 29 8 32.6 31.5-33.5 0.54 9 32.1 30.5-34.0 0.76 30 3 34.5 33.7-35.3 0.93 Zygomatic breadth 1 13 15.3 14.5-16.5 0.33 16 15.1 14.4-16.1 0.24 2 6 14.8 14.0-15.6 0.51 8 14.8 13.9-15.3 0.33 3 16 14.7 13.9-15.6 0.21 12 14.4 13.7-14.9 0.21 4 13 14.6 14.2-15.2 0.13 19 14.4 13.7-15.0 0.17 5 21 14.5 12.9-15.8 0.28 13 14.4 13.7-15.0 0.20 6 10 14.3 13.5-15.1 0.34 12 14.4 13.6-15.3 0.33 7 3 15.2 14.8-15.6 0.47 3 14.6 14.1-15.0 0.53 8 5 14.7 14.2-15.4 0.42 13 14.7 14.0-15.2 0.20 9 5 14.8 14.4-15.4 0.35 9 14.8 14.1-15.6 0.30 10 16 15.0 13.9-15.9 0.26 16 14.7 13.9-15.4 0.18 1 1 19 14.0 13.4-14.8 0.16 23 13.9 13.3-14.9 0.17 12 17 15.0 14.1-16.3 0.23 16 14.6 14.1-15.2 0.18 13 11 14.8 13.9-15.3 0.26 15 14.8 14.1-15.5 0.21 14 6 15.1 14.8-15.4 0.20 3 14.9 14.5-15.1 0.37 15 2 15.5 15.2-15.7 0.50 16 12 14.1 13.1-15.2 0.41 16 14.0 13.1-14.7 0.26 17 4 14.0 13.2-15.0 0.77 4 13.7 12.6-14.3 0.74 18 11 15.0 14.3-15.8 0.29 19 13 15.6 14.9-16.0 0.18 11 15.4 14.8-16.0 0.21 20 2 15.9 15.7-16.0 0.30 21 3 14.7 13.9-15.3 0.85 4 14.9 14.7-15.3 0.29 22 1 1 15.7 15.3-16.3 0.18 20 15.6 14.6-16.8 0.23 23 2 15.8 15.4-16.1 0.70 4 15.1 15.0-15.3 0.13 24 19 14.9 13.9-15.7 0.20 11 14.8 14.2-15.4 0.26 25 7 15.6 14.9-16.5 0.41 9 14.9 14.4-15.1 0.17 26 1 1 15.0 14.0-15.8 0.31 19 15.0 13.9-15.7 0.20 GENOWAYS— SYSTEMATICS OF LIOMYS 149 Table 11. — Continued. 27 2 14.9 14.5-15.3 0.80 2 15.2 14.9-15.4 28 12 15.6 15.0-16.7 0.27 13 15.3 14.4-16.0 29 5 15.2 15.0-15.4 0.16 5 14.9 14.3-15.4 30 2 15.6 15.5-15.6 Interorbital constriction 1 17 7.7 7.2-8. 5 0.19 16 7.7 7.2-8. 3 2 8 7.4 6. 9-7. 8 0.20 1 1 7.5 7. 1-8.1 3 20 7.6 7. 2-8.0 0.11 14 7.4 7.0-8. 1 4 14 7.6 6.9-8. 1 0.20 21 7.5 7. 0-7. 9 5 26 7.4 7. 0-8.0 0.11 18 7.5 6.7-8. 6 6 1 1 7.4 6. 7-8. 3 0.29 12 7.4 7. 0-7. 9 7 7 7.8 7. 4-8.3 0.27 8 7.8 7. 6-8.3 8 6 7.7 7.4-8. 1 0.20 14 7.4 7. 2-7. 6 9 7 7.8 7.4-8. 6 0.32 10 7.8 7. 5-8. 4 10 22 7.7 7. 1-8.2 0.13 19 7.7 7. 2-8. 5 1 1 21 7.3 6. 7-7. 9 0.12 27 7.1 6. 7-7.4 12 19 7.7 7. 1-8.3 0.15 19 7.6 7. 1-8.3 13 15 7.8 6. 7-8. 7 0.24 18 7.7 6.9-8. 6 14 6 7.8 7.2-8. 2 0.30 8 7.7 7.3-8. 1 15 2 7.8 7. 7-7.8 0.10 16 19 7.0 6. 6-7. 5 0.12 25 6.9 6. 3-7. 5 17 5 6.8 6. 3-7.0 0.25 5 6.5 6. 1-6.9 18 5 7.7 6.9-8. 1 0.44 17 7.6 7. 1-8.2 19 16 7.8 7. 1-8.3 0.16 11 7.8 7.4-8. 5 20 4 8.1 7.9-8. 6 21 6 7.6 7. 4-8.0 0.18 8 7.3 7. 0-7. 8 22 13 8.0 7. 6-8.3 0.15 23 7.8 7. 1-8.3 23 7 7.8 7.4-8. 5 0.27 8 8.0 7. 3-8. 8 24 29 7.8 7.3-8. 7 0.14 22 7.7 7. 1-8.1 25 1 1 7.9 7.3-8. 6 0.25 10 7.8 7. 1-8.5 26 16 7.6 7. 0-8.5 0.20 20 7.6 6.8-8. 1 27 3 7.8 7. 7-8.0 0.20 6 7.8 7. 2-8. 3 28 18 7.8 7. 2-8.4 0.13 16 7.5 7. 1-8.0 29 9 7.8 7. 4-8. 4 0.23 9 7.4 7. 1-7.8 30 4 8.0 7.7-8. 1 Mastoid breadth 1 17 14.4 13.8-14.9 0.16 16 14.2 13.7-14.9 2 7 14.1 13.6-14.8 0.31 1 1 14.0 13.2-14.5 3 18 14.0 13.5-14.5 0.12 15 13.7 13.2-14.1 4 13 13.9 13.6-14.4 0.11 21 13.8 13.3-14.2 5 25 13.7 12.9-14.4 0.15 19 13.7 13.0-14.3 6 1 1 13.8 13.3-14.4 0.21 12 13.6 13.0-14.3 7 6 13.9 13.5-14.7 0.38 9 13.8 13.2-14.7 8 6 14.3 13.9-14.6 0.22 14 13.8 13.0-14.4 9 7 13.9 13.7-14.4 0.17 10 13.9 13.3-14.5 10 20 13.9 12.7-14.5 0.19 16 13.9 13.1-14.7 11 20 13.5 12.8-13.8 0.12 27 13.4 13.0-13.9 12 19 14.3 13.6-14.9 0.16 19 14.1 13.5-14.6 13 15 14.3 13.2-15.0 0.26 17 14.2 13.5-14.8 14 6 14.2 13.2-14.7 0.42 8 14.0 13.2-14.5 15 2 14.8 14.7-14.9 0.20 0.50 0.26 0.42 0.10 0.15 0.21 0.15 0.12 0.20 0.17 0.16 0.07 0.17 0.14 0.07 0.14 0.17 0.24 0.11 0.32 0.15 0.21 0.34 0.19 0.12 0.33 0.13 0.30 0.13 0.35 0.13 0.13 0.19 0.16 0.22 0.12 0.1 1 0.13 0.23 0.32 0.20 0.21 0.22 0.09 0.12 0.17 0.31 18 19 20 21 22 23 24 25 26 27 28 29 30 1 2 3 4 5 6 7 8 9 10 1 1 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 1 2 3 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY Table 11. — Continued. 18 13.5 12.8-14.0 0.16 25 13.3 12.3-14.1 0.17 5 13.5 13.1-14.2 0.39 5 13.1 12.5-13.3 0.30 3 14.2 13.9-14.4 0.31 14 14.2 13.6-14.7 0.18 14 14.6 14.2-15.2 0.15 12 14.6 14.1-15.0 0.17 4 14.6 14.3-14.8 0.21 6 14.1 13.4-14.6 0.33 8 14.0 13.7-14.2 0.13 12 14.4 13.8-15.3 0.20 20 14.4 13.8-15.2 0.16 7 14.4 13.9-14.7 0.21 8 14.5 13.8-15.6 0.49 28 14.0 13.4-14.8 0.14 20 14.0 13.3-14.5 0.15 10 14.3 13.5-14.8 0.24 9 14.0 13.4-14.3 0.21 16 14.1 13.7-14.8 0.15 19 14.2 13.5-15.0 0.19 3 13.9 13.7-14.1 0.24 6 13.8 13.4-14.0 0.19 15 14.5 14.1-15.0 0.16 16 14.3 13.7-14.8 0.14 7 14.1 13.9-14.7 0.20 9 14.1 13.6-14.6 0.22 4 14.3 14.1-14.4 0.14 Length of nasals 17 12.0 10.3-13.0 0.38 16 12.1 11.4-13.2 0.26 8 12.6 11.6-13.3 0.37 11 11.9 11.3-13.1 0.33 19 12.0 11.2-12.9 0.22 14 11.6 10.8-12.4 0.28 14 1 1.7 11.0-12.6 0.27 21 11.5 10.8-12.4 0.20 25 1 1.6 10.5-12.2 0.18 18 11.2 10.3-12.2 0.27 1 1 11.5 10.3-13.4 0.58 12 11.5 10.0-12.7 0.49 7 12.8 11.8-13.5 0.39 9 12.5 10.8-13.3 0.48 6 12.1 11.0-12.8 0.56 14 11.9 10.6-12.5 0.30 7 12.1 11.5-13.1 0.40 10 12.0 11.1-12.9 0.37 23 13.1 11.7-13.7 0.20 19 12.7 11.9-13.6 0.21 22 1 1.7 10.5-12.8 0.25 27 11.2 10.4-12.2 0.19 19 12.9 12.1-13.7 0.25 18 12.5 11.1-13.9 0.32 15 12.9 10.8-14.3 0.43 18 12.7 11.8-13.7 0.24 6 12.4 11.3-13.3 0.59 9 12.4 10.2-13.3 0.70 2 13.7 13.5-13.9 0.40 18 11.5 10.3-12.7 0.33 23 10.9 9.5-12.8 0.33 5 11.1 9.9-11.9 0.72 6 11.2 9.6-11.9 0.69 5 13.0 12.3-13.8 0.51 17 12.6 11.8-13.4 0.24 14 13.4 12.6-14.5 0.32 1 1 13.1 12.8-13.6 0.17 4 13.4 13.1-13.8 0.31 6 12.5 12.0-13.1 0.36 8 12.0 11.2-12.9 0.49 13 14.1 13.3-15.8 0.40 24 13.9 12.7-15.1 0.29 6 13.6 12.7-14.7 0.60 6 13.8 12.6-14.8 0.67 24 12.9 1 1.9-14.2 0.26 21 12.5 11.5-13.7 0.25 1 1 13.5 12.5-14.7 0.36 10 12.9 1 1.9-13.9 0.36 16 13.1 1 1.6-14.0 0.33 18 12.8 12.0-14.0 0.27 3 12.8 12.3-13.1 0.48 6 12.5 11.9-13.2 0.37 18 13.2 1 1.8-14.1 0.24 15 12.6 11.8-13.8 0.30 9 13.2 12.4-13.9 0.34 9 12.7 11.9-14.3 0.50 3 13.5 12.5-14.1 1.03 Length of rostrum 16 14.0 12.8-14.8 0.32 16 13.7 13.1-15.1 0.26 8 13.8 13.0-14.5 0.30 1 1 13.4 12.5-14.5 0.39 18 13.5 12.7-14.6 0.25 14 13.1 12.6-13.9 0.21 GENOWAYS— SYSTEMATICS OF LIOMYS 151 Table 11. — Continued. 4 14 13.4 12.9-14.3 0.25 21 13.2 12.6-13.7 0.13 5 21 13.2 12.3-14.0 0.22 16 12.9 12.1-13.9 0.28 6 1 1 12.7 11.2-14.9 0.61 12 13.0 11.8-14.4 0.48 7 6 14.1 13.4-14.9 0.39 5 13.9 13.5-14.4 0.33 8 6 13.5 12.6-14.4 0.56 14 13.3 12.7-14.2 0.23 9 7 13.4 13.0-13.9 0.22 10 13.5 12.9-14.4 0.31 10 17 14.2 13.7-15.0 0.18 13 13.9 13.2-14.8 0.28 1 1 22 13.0 12.2-13.9 0.20 24 12.6 11.1-13.6 0.24 12 19 14.0 13.0-14.9 0.23 17 13.6 13.0-14.6 0.24 13 14 14.0 12.1-15.5 0.45 16 13.9 13.0-14.6 0.22 14 6 13.9 13.1-14.7 0.46 9 13.8 12.1-14.8 0.63 15 2 15.8 15.5-16.0 0.50 16 14 13.0 12.0-14.4 0.41 19 12.6 10.9-13.8 0.34 17 4 12.7 12.2-13.4 0.56 5 12.6 12.1-13.1 0.35 18 4 14.6 13.8-15.0 0.54 12 14.0 13.0-14.5 0.30 19 14 14.6 13.8-15.7 0.31 11 14.4 13.9-14.7 0.14 20 4 14.4 14.1-14.8 0.30 21 5 14.1 13.6-14.4 0.30 3 13.7 13.3-14.2 0.53 22 13 15.5 14.6-16.9 0.30 24 15.1 13.9-16.4 0.27 23 4 14.6 13.7-15.8 0.88 6 14.9 14.0-16.1 0.58 24 20 14.4 13.4-15.3 0.26 19 14.2 13.2-16.3 0.32 25 1 1 14.9 14.0-16.1 0.39 10 14.4 13.8-15.3 0.36 26 16 14.4 12.7-15.4 0.31 17 14.0 13.2-14.9 0.23 27 2 14.0 13.7-14.2 0.50 6 14.1 13.8-14.9 0.33 28 18 14.8 14.0-15.7 0.21 15 14.2 13.4-14.8 0.23 29 8 14.5 13.8-15.0 0.29 9 14.1 13.0-15.8 0.58 30 3 16.3 16.0-17.0 0.67 Length of maxillary toothrow 1 16 5.1 4.8-5. 5 0.10 16 5.1 4.7-5. 5 0.12 2 8 5.1 4. 9-5.4 0.11 1 I 4.9 4.7-5. 3 0.12 3 19 4.9 4.5-5. 2 0.08 14 4.8 4. 5-5.2 0.12 4 14 5.0 4.7-5. 2 0.09 20 4.9 4. 6-5. 2 0.08 5 24 4.8 4.2-5. 4 0.10 19 4.8 4.5-5. 1 0.07 6 1 1 4.7 4. 5-4.9 0.07 12 4.7 4. 4-5.0 0.1 1 7 6 4.7 4. 5-5.0 0.16 9 4.7 4. 3-4. 9 0.13 8 6 5.1 4. 8-5.4 0.18 14 5.1 4.7-5. 4 0.09 9 7 5.0 4. 5-5. 5 0.25 10 5.1 4. 8-5. 6 0.17 10 22 5.0 4.6-5. 6 0.10 15 4.9 4.4-5. 3 0.12 1 1 22 4.8 4. 6-5. 3 0.08 27 4.8 4.5-5. 3 0.07 12 19 4.9 4. 7-5. 3 0.08 19 4.8 4.4-5. 2 0.1 1 13 13 4.9 4. 5-5.4 0.15 18 4.9 4.6-5. 1 0.07 14 6 4.8 4.5-5. 1 0.22 9 4.8 4.7-5. 1 0.09 15 2 5.2 5. 1-5.2 0.10 16 19 4.8 4. 3-5. 3 0.14 24 4.7 4.3-5. 1 0.09 17 5 4.7 4. 4-5.0 0.23 6 4.7 4.4-4. 9 0.16 18 4 4.8 4.1-4.9 0.10 17 4.9 4.5-5. 2 0.09 19 17 5.2 5. 1-5.5 0.05 12 5.1 4.9-5. 3 0.07 20 4 4.7 4. 3-5.0 0.29 21 5 4.9 4. 7-5. 2 0.18 5 4.7 4. 4-4. 9 0.17 22 12 5.2 4. 7-5. 5 0.13 24 5.1 4. 9-5. 4 0.07 23 6 5.1 4. 8-5. 3 0.20 8 5.1 4. 8-5. 7 0.23 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 1 2 3 4 5 6 7 8 9 10 11 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY Table 11. — Continued. 30 4.8 4.5-5. 3 0.07 21 4.8 4. 2-5. 2 0.13 11 5.1 4. 6-5. 6 0.19 9 4.8 4. 5-5.2 0.12 17 4.9 4.7-5. 1 0.07 19 4.9 4.5-5. 5 0.13 2 4.8 4. 6-4. 9 0.30 5 4.9 4. 6-5.0 0.15 18 4.9 4.3-5. 3 0.13 16 4.7 4.4-5. 2 0.10 9 4.8 4.5-5. 3 0.16 9 4.7 4. 5-4.9 0.11 4 5.3 5. 1-5.4 0.14 Depth of braincase 16 8.5 8. 1-8.9 0.13 16 8.4 7. 9-8. 8 0.12 7 8.3 8. 1-8.6 0.14 11 8.3 7. 8-8.6 0.17 18 8.2 7. 8-8.5 0.10 15 8.1 7. 7-8. 5 0.10 11 8.3 8.0-8. 6 0.12 20 8.2 7. 8-8.7 0.13 21 8.1 7. 6-8. 6 0.10 16 8.0 7. 5-8. 5 0.13 10 8.0 7. 6-8.3 0.15 12 7.9 7.4-8. 5 0.15 5 8.2 8. 0-8.5 0.19 8 8.3 7. 8-8. 8 0.24 6 8.4 8. 2-8. 7 0.15 13 8.2 8. 0-8. 5 0.11 7 8.4 8. 2-8.6 0.12 9 8.3 8. 1-8.6 0.10 17 8.2 7. 8-8. 5 0.08 12 8.0 7. 7-8.4 0.13 19 8.1 7. 5-8.6 0.11 23 7.9 7. 3-8.4 0.11 19 8.4 7. 9-8.8 0.09 18 8.3 8.0-8. 8 0.11 15 8.4 7. 9-8.9 0.15 16 8.3 8. 1-8.6 0.08 5 8.4 8. 1-8.7 0.22 7 8.1 7. 8-8.4 0.16 2 8.8 8.7-8. 9 0.20 17 8.1 7. 7-8.4 0.09 23 7.9 7. 6-8. 2 0.08 5 7.9 7.7-8. 1 0.14 4 8.0 7. 8-8.2 0.17 3 8.6 8. 5-8.8 0.18 14 8.5 8.2-9. 1 0.13 11 8.6 8.3-9. 1 0.17 12 8.8 8. 3-9.6 0.20 4 8.6 8.4-8. 8 0.19 6 8.5 8. 2-8.9 0.20 8 8.3 8.0-8. 5 0.14 13 8.9 8. 6-9.3 0.13 20 8.7 8. 3-9.2 0.11 27 8.8 8. 6-9.0 0.10 8 8.8 8. 4-9. 4 0.29 26 8.8 8. 2-9. 4 0.11 17 8.6 8. 2-9.1 0.11 9 8.8 8. 5-9.2 0.17 10 8.6 8.2-9. 1 0.16 15 8.6 8. 0-9. 4 0.19 16 8.6 8. 1-8.9 0.12 3 8.7 8. 6-8. 8 0.12 2 8.4 8.2-8. 5 0.30 16 8.6 8. 3-9.0 0.10 14 8.5 8. 1-8.9 0.13 7 8.7 8. 3-9.0 0.17 8 8.3 7. 7-9. 3 0.36 3 9.1 8. 9-9. 5 0.37 Interparietal width 17 8.9 8. 2-9.4 0.15 16 8.9 7. 6-9.7 0.26 7 8.7 8. 0-9. 4 0.41 11 8.9 7. 8-9. 9 0.33 19 8.8 8. 1-9.9 0.21 15 8.5 7. 7-9. 2 0.19 12 9.0 8. 5-9. 7 0.23 20 8.9 7. 7-9.6 0.21 24 8.8 8. 1-9.8 0.16 19 8.7 7. 5-9.8 0.22 1 1 8.4 7. 5-9. 2 0.31 12 8.3 7. 6-9.3 0.27 6 8.6 8. 2-9.2 0.29 9 8.4 6. 9-9.0 0.41 6 8.8 8. 5-9.5 0.31 13 8.7 7. 3-9.4 0.30 7 8.8 8. 3-9.0 0.19 10 8.1 6. 7-9. 3 0.59 20 8.9 8. 1-9.7 0.19 15 9.0 7. 8-9.9 0.30 20 8.3 7. 8-8. 8 0.12 26 8.3 7. 5-8.9 0.13 GENOWAYS— SYSTEM ATICS OF LIOMYS 153 Table 11. — Continued. 12 19 8.8 7. 9-9.5 0.18 19 8.9 8. 1-9.5 0.18 13 15 8.9 8.4-9. 6 0.19 17 8.6 1.9-9.5 0.24 14 6 8.6 1.9-9.5 0.56 9 8.8 7. 8-9. 6 0.34 15 2 8.5 8. 3-8. 6 0.30 16 18 8.3 7. 8-9.2 0.21 25 8.2 7. 5-9.2 0.19 17 5 8.2 7. 6-9.0 0.55 6 8.1 7. 6-8. 6 0.28 18 4 8.9 8. 2-9. 7 0.63 13 8.5 7.9-9. 1 0.17 19 13 8.7 7. 9-9.0 0.17 12 8.4 7.2-9. 1 0.30 20 4 9.7 9. 4-9.9 0.24 21 5 8.8 8. 0-9. 9 0.65 8 8.7 8.2-9. 5 0.29 22 12 9.1 1.9-9.9 0.30 20 9.0 8.4-9. 5 0.15 23 7 8.6 7. 8-9.0 0.37 8 8.7 7. 8-9.6 0.42 24 29 8.9 7. 7-9.9 0.21 19 8.9 8. 4-9. 6 0.18 25 10 9.1 8. 5-9.6 0.23 10 8.8 8. 4-9.5 0.20 26 15 9.2 8. 7-9.9 0.20 17 9.1 8. 5-9.9 0.17 27 3 9.0 8. 7-9.3 0.35 5 8.7 7. 5-9.5 0.73 28 16 9.0 8. 5-9. 6 0.14 16 8.8 8. 1-9.8 0.27 29 7 8.7 7.9-9. 1 0.38 9 8.9 8. 2-9.9 0.47 30 3 8.8 8.4-9. 3 0.55 Interparietal length 1 17 4.0 3. 6-4.7 0.14 16 3.9 3. 6-4.3 0.12 2 7 3.7 3. 2-4. 2 0.26 11 3.8 3. 5-4.1 0.13 3 19 4.1 3. 5-4.8 0.15 15 3.9 3. 3-4.3 0.14 4 12 4.0 3. 5-4.7 0.24 20 4.0 3. 2-4.5 0.15 5 24 4.0 3. 6-4.4 0.10 18 4.0 3. 5-4.8 0.15 6 11 3.9 3. 3-4.6 0.28 12 4.0 3.4-4. 8 0.31 7 6 4.7 4. 4-5. 2 0.24 9 4.7 4. 1-5.2 0.25 8 6 3.9 3. 5-4.3 0.24 13 3.9 3. 3-4.5 0.19 9 7 4.2 3. 9-4.6 0.19 9 3.9 3. 4-4. 4 0.23 10 20 4.6 4. 1-5.2 0.12 16 4.6 4.2-5. 2 0.15 1 1 21 3.8 3. 4-4.5 0.13 27 3.7 3. 3-4.2 0.09 12 19 4.4 3.8-5. 1 0.13 19 4.4 4.0-5. 1 0.13 13 15 4.6 4. 1-5.1 0.16 17 4.4 4. 0-5.0 0.13 14 6 4.5 4.1-5. 0 0.27 8 4.4 3. 5-4.9 0.33 15 2 4.7 4. 5-4. 8 0.30 16 18 3.6 3. 2-4.7 0.18 25 3.6 3. 0-4. 2 0.15 17 5 3.7 3. 5-4.0 0.18 6 3.7 3. 4-4.0 0.16 18 5 4.3 3. 9-4.5 0.22 13 4.2 3. 7-4.9 0.20 19 13 4.1 3. 6-4.7 0.17 12 4.1 3. 3-4.7 0.24 20 4 4.6 4. 0-5.0 0.43 21 6 4.4 4. 0-4. 8 0.22 8 4.5 4. 1-5.0 0.19 22 13 4.7 4. 2-5. 3 0.17 20 4.5 3. 5-5.0 0.15 23 7 4.5 4.1-5. 3 0.30 8 4.6 4. 1-5.3 0.29 24 29 4.6 3. 7-5.5 0.17 19 4.5 3. 9-5.0 0.12 25 10 4.5 3. 8-4.8 0.25 10 4.4 3. 8-4.8 0.21 26 15 4.3 3. 8-4.9 0.17 17 4.5 4. 0-5.0 0.14 27 3 4.7 4. 3-5.4 0.68 6 4.5 4.0-5. 3 0.38 28 17 4.6 3. 8-5. 3 0.19 16 4.5 3. 9-4.9 0.15 29 7 4.7 4. 0-5. 2 0.30 9 4.5 4. 2-4. 9 0.16 30 3 5.0 4.6-5. 4 0.46 154 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY Cranial Measurements The 10 cranial measurements will be discussed below in the order they were for Liomys irroratus — 1) measurements of cranial length; 2) measurements of cranial breadth; 3) measurement of cranial depth. A smooth cline is shown in the mean greatest length of skull of specimens from northwestern Mexico (Fig. 20 and Tables 11, 12). In these, there is a decrease in size from the medium-sized specimens of sample 1 in the north to the relatively small specimens of sample 6 in the south. Specimens from sample 7, as they did in total length, average considerably larger than specimens from sample 6 and more nearly approach specimens from samples 10 and 12 to 14 to the south. Males and females from sample 8 and the males from sample 9 have a greatest length of skull that averages intermediate between the smaller animals to the north (6) and larger to the west (7). However, females from sample 9 are rela¬ tively large and more nearly approach specimens from sample 10. The means for specimens from interior Jalisco, Michoacan, and Guerrero (samples 11, 16, 17) become progressively smaller to the south, with specimens from sample 1 7 aver¬ aging the smallest of the species. Mice from the Sierra Madre del Sur of Guerrero and Oaxaca (samples 20, 22, 23, 30) are relatively large; this is especially notice¬ able in the mean greatest length of skull for females. Only males from south¬ eastern Oaxaca and northwestern Chiapas (sample 25) approach them in size. It should be noted that the last sample (25) is the only area where the geographic ranges of Liomys pictus and Liomys salvini are sympatric. Some specimens from the vicinity of San Gabriel Mixtepec, Oaxaca (22), reach the largest size of the species, although the three females from Omilteme, Guerrero (30), average largest. It is interesting to note that specimens from sample 20 more nearly agree with specimens from the montane area rather than the coastal areas as they did in external measurements. Excepting the three rather large males from sample 18 and the relatively small specimens (as they were externally) from sample 21, the other specimens form a group having skulls of medium length that does not appear to be appreciably different from that of specimens in western Jalisco and Colima. One last point of interest is that females from the Pacific slopes of Sierra Madre in Guerrero (sample 19) are somewhat larger than females from adjacent lowlands (18) as might be expected from their intermediate geographic position between the coastal and montane populations. Variation in length of nasals (Fig. 20 and Table 1 1) for populations of Liomys pictus from throughout the range of the species agrees closely with the pattern for greatest length of skull. Samples 1 to 6 show a clinal decrease in size from north (1) to south (6) except that males from sample 1 have, on the average, shorter nasals than do males from sample 2, and females from sample 5 are the smallest of the group. The mean length of nasals for specimens from sample 7 is long and appears to fall into a group characterized by nasals of medium length that includes specimens from samples 10, 12 to 14, 18, 19, 21, and 24 to 29. Mice from samples 11, 16, and 1 7, again, form a group characterized by small size. Specimens from samples 20, 22, 23, and 30 have, on the average, the longest nasals of the species. The mean length of nasals for specimens of both sexes from GENOWAYS — SYSTEMATICS OF LIOMYS 155 sample 19 is intermediate between the mean for coastal and montane populations as is the geographic location. Specimens from sample 21 average smaller than those from surrounding samples but still fall within the lower end of variation for specimens from samples in this group, and males from sample 25 are larger than those from surrounding areas but the females are not. There are only minor differences in the pattern of variation for length of the rostrum (Table 1 1) as compared with that for length of nasals. The rostrum of males from sample 19 is no longer than that of males from coastal areas of Guer¬ rero (18). The mean length of rostrum for females from samples 19 and 20 is intermediate between the smaller specimens from sample 18 and larger specimens from samples 22 and 23; however, they do not approach the mean of specimens from sample 30. The males from the montane sample 23 more closely resemble those of adjacent lowland samples in length of rostrum than they resemble ani¬ mals from the montane sample to the west (22). The length of the maxillary toothrow (Table 1 1) does not display a pattern of variation similar to the three previous measurements of length and, in fact, shows little significant variation at all. Specimens from northwestern Mexico show a slight decrease in mean length of maxillary toothrow from samples 1 to 6 and it should be noted that mean for each sex at sample 7 is the same as those for sample 6. The means for specimens from samples 8 and 9 are longer than for mice from sample 10. The only populations south of sample 10 having a mean length of maxillary toothrow of 5.0 or more are 19, 22, 23, and males from sample 25. Only at sample 17 do males have a mean of 4.7 and only at samples 16, 17, 20, 21, 28, and 29 do females have a mean of 4.7. The remainder of the samples have a mean of either 4.8 or 4.9. The length of interparietal bone (Fig. 20 and Tables 11, 12) averages almost uniformly a medium length for the specimens in samples from northwestern Mexico (1-6), only males from sample 2 being small. Both sexes from sample 8 and females from sample 9 have interparietal bones that average the same length as populations to the north (1-6). Males from sample 9 have a mean length of interparietal that is intermediate between the samples to the west and south (7, 10, 12, 14) with long interparietal bones and the samples to the north (1-6) with interparietals of medium length. Samples 11, 16, and 17 have a short interparietal bone, with specimens from sample 16 having the shortest of the species. The remaining samples including those from montane areas are made up of speci¬ mens characterized by long interparietal bones. The three specimens from Omil- teme, Guerrero (30), are at the upper end of the range of this variation. Speci¬ mens from samples 18 and 19, especially the latter, are somewhat smaller than might be expected in mice from those areas, judging by other mensural data. Although variation in zygomatic breadth follows the same general pattern of variation as the measurements of cranial length, it is not so variable as the length measurements (Fig. 20 and Table 1 1). Zygomatic breadth of males from samples 1 to 6 varies clinally. The males from sample 1 have a zygomatic breadth that is in the intermediate range of variation of the species; mean values become pro¬ gressively smaller to the south until at sample 6 the individuals are on the average 156 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY Table 12. — Results of three typical SS-STP analyses ( greatest length of skull, mastoid breadth, and interparietal length) of geographic variation o/Liomys pictus. Vertical lines to the right of each array of means connect maximally nonsignificant subsets at the 0.05 level. See text for key to sample numbers. Males Females Sample Sample Means Results SS-STP number Means Results SS-STP number 15 35.1 22 34.3 18 33.3 23 33.2 25 33.2 19 33.0 28 32.8 26 32.7 29 32.6 24 32.4 1 32.3 10 32.1 12 32.0 21 32.0 14 32.0 13 31.9 27 31.9 7 31.8 2 31.7 3 31.2 8 31.2 9 31.1 4 31.0 5 30.5 1 1 30.4 16 30.0 6 29.8 17 29.1 15 14.8 19 14.6 28 14.5 1 14.4 Greatest length of skull 30 22 23 20 19 25 29 18 26 1 13 10 28 7 24 27 14 9 2 12 21 4 8 3 6 5 1 1 16 17 Mastoid breadth 20 19 23 22 34.5 33.6 33.3 33.0 32.7 32.3 32.3 32.0 32.0 31.8 31.7 31.7 31.6 31.6 31.5 31.4 31.4 31.4 31.4 31.3 31.3 30.6 30.6 30.5 30.1 30.0 29.6 29.0 28.4 14.6 14.6 14.5 14.4 GENOWAYS— SYSTEMATICS OF LIOMYS 157 22 14.4 23 14.4 13 14.3 12 14.3 25 14.3 8 14.3 18 14.2 14 14.2 29 14.1 21 14.1 2 14.1 26 14.1 3 14.0 24 14.0 4 14.0 10 13.9 9 13.9 7 13.9 27 13.9 6 13.8 5 13.7 17 13.5 16 13.5 1 1 13.5 27 4.7 29 4.7 7 4.7 22 4.7 15 4.7 24 4.6 28 4.6 10 4.6 13 4.6 14 4.5 25 4.5 23 4.5 12 4.4 21 4.4 18 4.3 26 4.3 9 4.2 Table 12. — Continued. 30 14.3 28 14.3 13 14.2 1 14.2 26 14.2 18 14.2 29 14.1 12 14.1 25 14.0 2 14.0 21 14.0 24 14.0 14 14.0 9 13.9 10 13.8 27 13.8 7 13.8 4 13.8 8 13.8 3 13.8 5 13.7 6 13.6 1 1 13.4 16 13.3 17 13.1 rparietal 30 5.0 7 4.7 20 4.6 10 4.6 23 4.6 21 4.5 26 4.5 29 4.5 22 4.5 27 4.5 24 4.5 28 4.5 12 4.4 25 4.4 13 4.4 14 4.4 18 4.3 158 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY 19 4.2 3 4.1 1 4.0 4 4.0 5 4.0 6 3.9 8 3.9 11 3.8 2 3.7 17 3.7 16 3.6 Table 12. — Continued. 5 6 4 1 9 8 3 2 17 11 16 4.1 4.0 4.0 4.0 3.9 3.9 3.9 3.9 3.8 3.7 3.7 3.6 relatively small. Males from the vicinity of San Bias, Nayarit (sample 7), are much larger than males from northern Nayarit (6) and more closely resemble males from southwestern Nayarit and western Jalisco (10, 12-14). Males from samples 8 and 9 have zygomatic breadths that average between the values of the small males from sample 6 and the larger males from sample 7. Females from these areas show a different pattern of variation. Females from samples 1 and 2 are larger than those to the south as were the males, but the females from sam¬ ples 3 to 6 have the same average zygomatic breadth. Although females from samples 7 to 10 have a somewhat greater zygomatic breadth than those to the north (3-6), it is certainly not a striking difference that causes a sharp break in the variation as it does in males. Specimens from samples 11, 16, and 17 form a unit characterized by a narrow mean zygomatic breadth; this is more readily seen in females than in males. Another unit characterized by a broad mean value is formed by the montane samples from Oaxaca and Guerrero (20, 22, 23, 30). The females from sample 23 are relatively small but the sample contains only four individuals. Males from samples 25 and 28 and to some extent the females from sample 28 (Los Tuxtlas, Veracruz) have a mean zygomatic breadth that is broader than that of adjacent populations. The specimens from sample 19 again show a size that is intermediate between coastal and montane samples. The remaining samples (12-14, 18, 21, 24, 26, 27, 29, and females from 25) have a mean zygomatic breadth that is medium in width for the species and most nearly approaches the condition found in specimens from sample 10. This group forms a chain of samples along the west coast of Mexico and across the Isthmus of Tehuantepec into Veracruz and Chiapas in which the zygomatic breadth is of relatively the same size and exhibits no sharp breaks in the variation. Specimens from Sonora, Sinaloa, Durango, and northern Nayarit (samples 1 -6) show a relatively smooth clinal decrease in mean mastoid breadth from north to south; unlike the case in many other measurements the specimens from sample 7 are not appreciably larger than those from sample 6 (Fig. 20 and Tables 11, 12). GENOWAYS— SYSTEMATICS OF LIOMYS 159 The specimens trom this entire area have an average mastoid breadth that is in the middle of the range ot variation tor the species. The means for samples 5 and 6 are the smallest in this northern group, but the mean mastoid breadths for speci¬ mens trom samples 11, 16, and 17 are the smallest of the species. Other samples having a mean mastoid breadth that is in the medium size range for the species include 8 to 10, 12 to 14, 18, 21, 24 to 27, 29, and 30. Specimens from the mountains of Oaxaca (20, 22, 23) and the Pacific slope of the Sierra Madre del Sur of Guerrero (19) have, on the average, the broadest mastoid breadth of the species, although the difference between them and the coastal populations is relatively small. Both males and females from Los Tuxtlas, Veracruz (28), have a mean mastoid breadth that is greater than the mastoid breadth of surrounding samples and falls within the range of values of animals in montane areas of Oaxaca and Guerrero. Variation in the mean breadth of the interorbital constriction along the north¬ western coast of Mexico (samples 1-6) shows a clinal decrease in size from north to south with the exception of males from sample 2 and females from sample 3, which have the same means as specimens of their respective sex from sample 6 (Table 11). The specimens from sample 7 have a broad interorbital region and more nearly resemble specimens from southwestern Nayarit, western Jalisco, and Colima (samples 10, 12-14). Also specimens from samples 8 and 9 excepting fe¬ males from sample 8, which are as small as females from sample 6, have, on the average, an interorbital constriction that more closely approaches that of samples from western Jalisco and Colima (10, 12-14) than those to the north (6). Speci¬ mens from the interior of Jalisco, Michoacan, and Guerrero reveal a cline in mean breadth of interorbital constriction from the largest of the group, sample 1 1 (however, these mice average less than any others except those in this group), to sample 1 7 (smallest interorbital constriction of the species). The remaining sam¬ ples with only a few exceptions have specimens with a mean interorbital con¬ striction that is nearly the same. Females from samples 20, 23, and 30 and males from sample 22 have an interorbital constriction that averages 8.0 or more, which is broader than specimens from other samples. Females from samples 28 and 29 have a mean interorbital constriction that is narrower than for females from adjacent samples, but males from these two samples do not show the narrowing of the interorbital region. Besides the relatively narrow mean interparietal width of specimens from samples 6, 11, 16, and 17 and relatively broad mean interparietal width of speci¬ mens from sample 20, the samples of Liomys pictus have mean interparietal widths that reveal little variation (Table 1 1). Females from samples 7 and 9 have an interparietal width that is small, but males from these samples are similar to males from surrounding samples. The mean depth of braincase varies clinally in specimens from along the north¬ western coast of Mexico, those from sample 1 being deeper than those from sam¬ ple 6 (Fig. 20 and Table 1 1). Only specimens from high altitudes in Sinaloa and Durango (4) are slightly shallower than would be expected at that point in the cline. Specimens from samples 7, 8, and 9 have crania that average deeper than 160 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY Table 13. — Geographical variation in the configuration of the interparietal bone o/Liomys pictus. Figures are given in per cent of total number for each sample. See text for key to sample numbers. Interparietal bone Posterior margin of interparietal bone Sample N Undivided Divided N Notched Slightly notched Unnotched 1 52 48.1 51.9 52 75.0 11.5 13.5 2 39 38.5 61.5 40 80.0 10.0 10.0 3 30 46.7 53.3 31 71.0 22.6 6.4 4 33 42.4 57.6 32 78.1 9.4 12.5 5 70 44.3 55.7 81 81.5 9.9 8.6 6 32 71.9 28.1 33 66.7 12.1 21.2 7 24 100.0 0 24 0 25.0 75.0 8 16 56.3 43.7 24 58.3 12.5 29.2 9 43 62.8 37.2 43 60.5 23.2 16.3 10 52 100.0 0 52 13.5 23.1 63.4 11 72 79.2 20.8 82 85.4 6.1 8.5 12 65 100.0 0 60 18.3 21.7 60.0 13 35 100.0 0 38 7.9 13.2 78.9 14 8 100.0 0 7 14.3 0 85.7 15 3 100.0 0 3 33.3 0 66.7 16 26 96.2 3.8 46 58.7 10.9 30.4 17 9 100.0 0 9 33.3 44.5 22.2 18 10 100.0 0 12 25.0 16.7 58.3 19 28 100.0 0 31 19.4 16.1 64.5 20 3 100.0 0 8 0 37.5 62.5 21 13 100.0 0 19 31.6 21.0 47.4 22 52 98.1 1.9 60 68.3 10.0 21.7 23 23 100.0 0 23 39.1 8.7 52.2 24 13 100.0 0 30 46.7 20.0 33.3 25 7 100.0 0 21 66.7 9.5 23.8 26 22 100.0 0 24 50.0 20.8 29.2 27 9 100.0 0 9 55.6 33.3 11.1 28 38 100.0 0 38 65.8 7.9 26.3 29 22 95.5 4.5 22 13.7 22.7 63.6 30 4 100.0 0 4 0 0 100.0 those from sample 6 and more nearly approach males from samples 10 and 14 and both sexes from samples 12 and 13. The females from samples 10 and 14 have relatively shallow crania and approach specimens from samples 11, 16, and 1 7, which are the shallowest of the species. The mean depth of braincase for speci¬ mens from sample 1 8 is deeper than for specimens to the north and more nearly GENOWAYS— SYSTEMATICS OF LIOMYS 161 approaches that in specimens to the south. The remainder of the samples show little significant variation in the depth of braincase. Specimens from sample 21 average smaller than those from contiguous samples and specimens from Omil- teme, Guerrero (sample 30), have, on the average, the deepest braincases of the species. Qualitative Cranial Characters Condition of interparietal bone (Table 13). — The longitudinal division of the interparietal bone appears to be characteristic of populations from Sonora, Sinaloa, Durango, the lowlands of northern Nayarit (samples 1-6), inland at higher altitudes in the vicinity of Tepic, Nayarit (8), and Guadalajara, Jalisco (9), and the interior drainage basins of Jalisco (11). In these areas as many as 50 per cent or more of the individuals may show the condition. Only three in¬ individuals from other samples were found to have the interparietal bone divided; one of these was from sample 16, which is the contiguous sample to the east and south of sample 1 1 , and the other two were from samples 22 and 29. Notching (large or small) of the posterior margin of the interparietal bone is characteristic of specimens from northwestern Mexico (samples 1-6), inland in the vicinity of Tepic, Nayarit, and Guadalajara, Jalisco (samples 8, 9), and in¬ terior Jalisco, Michoacan, and Guerrero (samples 11, 16, 17); in these areas almost 70 per cent or more of the individuals show some kind of notching. Also specimens from Oaxaca, Chiapas, and southern Veracruz (samples 21-28) show notching at least in 50 per cent of the individuals and usually in more than 70 per cent. Samples from the localities in western Mexico from southern Nayarit through Guerrero (samples 7, 10, 12-15, 18, 19, 30), and in eastern Mexico (central Veracruz, sample 29) have a high percentage of the individuals, more than 60 per cent, that do not have a notch in the posterior margin of the inter¬ parietal bone. Condition of posterior nasal region (Table 14). — Most individuals of Liomys pictus have an emarginate posterior margin of the nasals. Specimens from sam¬ ples 7, 10, 18, and 19 do have a high percentage (46.7-86.2) of individuals that have a truncate posterior margin of the nasals, and samples 20, 23, 27, and 28 have more than 20 per cent of the individuals with truncated nasals. The condition in which the premaxillary bones terminate at the same level as the nasals (no longer than nasals) presents an interesting pattern of variation in northwestern Mexico (samples 1-6). No more individuals from Sonora and northern Sinaloa have the premaxillaries and nasals of equal length than is com¬ monly found elsewhere in the species; southward from sample 1, however, an increasing number of individuals exhibit this condition reaching the highest per¬ centage in sample 4 (86.4 per cent). A high percentage of individuals with the premaxillaries and nasals of equal length is found also in specimens from the vicinity of Tepic, Nayarit, and Guadalajara, Jalisco. In the remaining samples, individuals with the premaxillary bones extending posterior to the nasal bones (longer than the nasals) predominate, although 25 per cent or more of the individ¬ uals from samples 22, 23, and 26 have the bones of equal length. 162 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY Table 14. — Geographical variation in bones of the posterior portion of the nasal¬ premaxillary complex of Liomys pictus. Figures are given in per cent of total number for each sample. See text for key to sample numbers. Sample Shape of posterior margin of nasals Length of premaxillary bones N Emarginate Rounded Truncate N Longer than nasals Equal to nasals 1 51 98.0 2.0 0 51 94.1 5.9 2 39 97.4 2.6 0 39 84.6 15.4 3 31 93.5 6.5 0 30 76.7 23.3 4 35 85.7 14.3 0 34 14.7 86.3 5 73 93.2 6.8 0 82 36.6 63.4 6 33 78.8 15.1 6.1 32 59.4 40.6 7 24 0 37.5 62.5 24 100.0 0 8 24 79.2 12.5 8.3 24 37.5 62.5 9 42 85.7 4.8 9.5 43 18.6 81.4 10 52 15.4 21.1 63.5 52 98.1 1.9 11 81 95.1 4.9 0 84 100.0 0 12 64 60.9 28.2 10.9 64 82.8 17.2 13 36 58.3 25.0 16.7 35 88.6 11.4 14 8 62.5 25.0 12.5 8 87.5 12.5 15 3 100.0 0 0 3 66.7 33.3 16 47 85.1 10.6 4.3 47 100.0 0 17 9 77.8 22.2 0 9 88.9 11.1 18 15 40.0 13.3 46.7 15 93.3 6.7 19 29 6.9 6.9 86.2 32 96.9 3.1 20 8 62.5 12.5 25.0 8 100.0 0 21 19 84.2 15.8 0 19 94.7 5.3 22 61 80.3 8.2 11.5 61 73.8 26.2 23 24 70.8 4.2 25.0 24 62.5 37.5 24 33 87.8 6.1 6.1 36 100.0 0 25 21 90.4 4.8 4.8 21 100.0 0 26 25 92.0 8.0 0 25 64.0 36.0 27 9 66.7 11.1 22.2 9 100.0 0 28 39 56.4 15.4 28.2 39 100.0 0 29 22 72.7 22.7 4.6 22 95.5 4.5 30 4 100.0 0 0 4 100.0 0 Multivariate Analysis Means of the 13 external and cranial measurements and four qualitative cranial characters were used in the NT-SYS multivariate analysis programs. Phenograms diagraming the phenetic relationships of both male and female Liomys pictus were computed from both distance and correlation matrices; the GENOWAYS— SYSTEMATICS OF LIOMYS 163 i _ 2.00 1 2 3 5 4~ 8 1 - 9 7 10 rf 12 _ L 13 14 -29 21 £24 1—27 27 26 18 28 19 22 25 23 LS 6 11 16 -H I - 17 1.65 L L J 1.30 0.95 0.60 025 Fig. 21.— Phenograms of numbered samples (see Fig. 19 and text) of Liomys p ictus (males left, females right) computed from distance matrices based on standardized characters and clustered by unweighted pair-group method using arithmetic averages (UPGMA). The cophenetic correlation coefficient for the phenogram for males is 0.749 and for females is 0.800. The samples labeled LS are composed of specimens of Liomys spectabilis. phenograms based upon the distance matrix are presented herein because they had larger coefficients of cophenetic correlation (Fig. 21). In addition, a map (Fig. 22), which includes values for both sexes, is presented showing the appropri¬ ate distance coefficients between the connected samples; in most cases distance coefficients have been given only for contiguous samples. The first three prin¬ cipal components were computed from the matrix of correlation among the 1 7 characters; these first three components combine to express 82.60 per cent of phenetic variation in males and 81.40 per cent in females. Two-dimensional plots of principal components I-II and I-III and three-dimensional projections of the 30 samples onto the first three principal components based on the matrix of correlation among characters are presented for both sexes (Figs. 23-26). The distance phenogram for males shows the samples falling into four major clusters. The first cluster includes samples 1 to 5, 8, and 9. This, excepting sample 6, represents the specimens from the coast of northwestern Mexico and 164 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY extends inland around Tepic, Nayarit, to the vicinity of Guadalajara, Jalisco. Specimens from northern Nayarit (6) were always found to be small in the uni¬ variate analysis and they are grouped with samples 1 1, 16, and 17, which contain the other small specimens of the species. This grouping, however, is untenable on geographic grounds. Specimens from samples 1 and 2 were usually found to average larger than specimens from other samples in this group and they appear to be phenetically the most distinct. The second cluster includes samples 7, 10, 12 to 14, 1 8, 19, 21, 24, and 26 to 29. This group includes specimens from along the coast of western Mexico from as far north as San Bias, Nayarit, southward to the Isthmus of Tehuantepec and the interior valley of Chiapas and then north¬ ward in the lowland of southern and central Veracruz. Sample 19 shows the greatest distance from other members of this group, which probably can be ex¬ plained by the tendency of specimens from this sample to be intermediate be¬ tween coastal and montane samples in measurements in the univariate analysis. Table 1 5. Factor matrix from correlation among 1 7 characters o/Liomys pictus studied. GENOWAYS— SYSTEMATICS OF LIOMYS 4— * c 13 in >n NO t" — ON o o E'¬ _ ■^r _ 00 C on ■'t o *— « no On — < oo — oo 00 On n aj u a3 X u in >n ON in — NO r-~ _ o «n o o CT) r- ON t"- On m r- in «n r- o m (N ON NO r- ON 00 00 ON OO 00 00 ON ON «n 00 in 00 NO CN (N o o o o o o o o o o o o o o d o o c n ON o r\ M o — (N o O m (N O n O ON NO ON m o _ a — £ o o o o O • o O m (N NO r- NO o o d o o o o d o o o o o o o o o U 1 1 1 1 1 1 1 1 1 1 1 1 o r- c 13 (N o 00 NO NO o m, ON oo ON o W O 3 C/5 rt ■ X o o u o 4—3 >, u 0> X <2 -o o X ♦-3 T3 «3 13 ■4—3 C/3 c o -C *5 C/3 03 E 3 i_ 4—3 >< C/D 03 u c 4—1 rs X ■4— * ao c3 •4-* Na- c X c— ab c J3 1-4 X o o 15 cd > M 4—3 c 1) 1 1 "E ■* -* u a u 13 *— > c 03 4—4 u a i> ■*-» c <4- O c ’5) 03 E a +-> o 03 a u c a3 c n 03 C o c ’5) u 03 E C/5 13 c o X >4 l_ a3 X 03 E 13 13 4— < c/5 o a c o > Q D ♦— c/D o a o X 4-3 00 c 13 J 165 166 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY I nr i Fig. 23. _ Two-dimensional projections of the first three principal components illustrating the phenetic position of the 27 samples of male Liomys p ictus and one of Liomys specta- bilis (LS). Top, component I plotted against component II; bottom, component I plotted against component III. See Fig. 19 and text for key to samples. GENOWAYS— SYSTEMATICS OF LIOMYS 167 • 7 • 10 17® 19# • 20 29* #14 13« 28# ®21 18* .72 • 16 24® *27 • 11 #LS • 25 • 6 • 30 «23 • 26 • 3 • 22 • 1 2#*8 • 5 #9 • 4 I I Fig. 24. — Two-dimensional projections of the first three principal components illustrating the phenetic position of the 29 samples of female Liomys pictus and one of Liomys specta- bilis (LS). Top, component I plotted against component II; bottom, component I plotted against component III. See Fig. 19 and text for key to samples. 168 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY Another group includes samples 22, 23, and 25. Two samples (22, 23) included here are in the high mountains of Oaxaca and the other (sample 25) consists ot specimens from southeastern Oaxaca and northwestern Chiapas and is not con¬ tiguous geographically with the other two samples; these three samples are made up of the largest males of the species. One possible explanation why males from sample 25 are larger than males from surrounding samples is that this is the area where Liomys pictus and Liomys salvini occupy sympatric ranges and some de¬ gree of character displacement may be involved (see section of specific relation¬ ships following the species accounts). The last group includes samples 6, 11, 16, and 17, although, as stated above, sample 6 is out of place here on geographic grounds. The remaining three samples form a contiguous series in interior Jalis¬ co, Michoacan, and Guerrero that is characterized by uniformly small size. The distance phenogram for female Liomys pictus reveals essentially the same four clusters seen in males, in addition to a fifth group composed only of sample 30. The first cluster for females is the same as that for males excepting that sam¬ ple 6 is included with this group as might be expected on geographic grounds. The second cluster differs from that of males only in that sample 19 is farther away from the group and falls almost in an intermediate position between it and the next group. The third cluster of samples includes 20, 22, and 23, which are the three samples from the mountains of Oaxaca. The individuals in this group, except for the three specimens from sample 30, are the largest females of the species. The fourth cluster of samples for females includes only the three samples from interior Jalisco, Michoacan, and Guerrero (11, 16, 17) from which speci¬ mens are characterized by small size. Sample 30 appears to fall some “distance” from the other samples, but it should be remembered that there are only four specimens included in this sample. In the principal components analysis, the amount of phenetic variation repre¬ sented in the first three components for male and female Liomys pictus, respec¬ tively, was 61.23 and 60.48 per cent in component I, 15.40 and 13.84 in com¬ ponent II, and 5.97 and 7.08 in component III. Results of a factor analysis show¬ ing characters influencing the first three components is given in Table 15. From this analysis, it is apparent that the first, and by far the most important, com¬ ponent is heavily influenced by general size. The second component for both sexes is influenced by the qualitative cranial characters. Characters influencing the third component appear to be different between the sexes; for males, there is high positive weighting for length of tail and depth of braincase and high neg¬ ative values for length of maxillary toothrow and the qualitative cranial characters, whereas for females interparietal width has a high positive value and length of maxillary toothrow has a high negative value. Examination of the two- dimensional plots (I II and I-III, see Figs. 23, 24) and the three-dimensional projections (Figs. 25, 26) of the samples reveals a pattern similar to that shown in the distance phenogram. Samples 1 to 6 for males form a chain of samples that represents the clinal variation seen in many of the characteristics in the univariate analysis. Specimens from sample 1 fall at the lower end of the phenetic variation of specimens in group 2 (see below), but the geographic position of this sample GENOWAYS — SYSTEMAT1CS OF LIOMYS 169 i Fig. 25. Three-dimensional projection of 27 samples of male Liomys pictus and one of Liomys spectab ilis (LS) onto the first three principal components based upon a matrix of correlation among the 13 external and cranial measurements and four qualitative cranial characters. Components I and II are indicated in the figure and component III is represented by height. See Fig. 19 and text for key to samples. (Sonora and northern Sinaloa) makes such an arrangement taxonomically un¬ acceptable. Specimens from sample 6 at the small end of this cline have a phenetic position near other samples of small specimens from interior Jalisco, Michoacan, and Guerrero, but again on geographical grounds this is an unacceptable arrange¬ ment. Specimens from samples 8 and 9 also fall into this group, but at a position that is nearly intermediate between the small individuals to the north (6) and the large individuals to the west (7 and 10). The second group for males is composed of samples 7, 10, 12 to 14, 18, 19, 21, 24, and 26 to 29. Sample 21 falls relatively far from its contiguous populations but appears to be phenetically most aligned with this group. A third group of samples is composed of 22, 23, and 25. Sample 25 is discontinuous from the other two, which are from the mountains of Oaxaca, and probably is best thought of as an extreme phenetic variate of the second group (above). The fourth cluster in the analysis of males is made up of the samples from the interior of Jalisco, Michoacan, and Guerrero (11, 16, 17). Sample 6 fits here phenetically but this would not be acceptable geographically as seen above. Females exhibit nearly the same phenetic patterns as males. Sam¬ ple 6 is placed phenetically closer to its geographically contiguous populations in northwestern Mexico and would appear to fit well at the end of this cline. Speci¬ mens from sample 25 are placed with the second group of samples as their geo¬ graphic location would suggest. Females from sample 19 have a phenetic position that lies between the second and third group of samples and possibly represent a transition between the groups. Sample 30 is rather far removed from any other sample but is nearest the group of 20, 22, and 23. Variation in Color The results of my studies of variation in color should be considered as only preliminary, for the reasons enumerated in the section on methods and mate- 170 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY Fig. 26. — Three-dimensional projection of the 29 samples of female Liomys pictus and one of Liomys spectalis (LS) onto the first three principal components based upon a matrix of correlation among the 13 external and cranial measurements and four qualitative cranial characters. Components I and II are indicated in the figure and component III is represented by height. See Fig. 19 and text for key to samples. rials, but they do show some trends in variation (Table 16). Specimens from high elevations are uniformly darker than specimens from adjacent lowlands (see Fig. 27). One striking example of the change in color from montane areas to lowlands is seen along the southern coast of Oaxaca. Specimens from south of Candelaria (4 mi. S Candelaria is the farthest south) are dark but specimens from about 5 kilometers to the south at Chacalapa are pale and resemble other lowland mice. A similar situation is seen when specimens from the vicinity of San Gabriel Mixtepec (dark) are compared with two specimens from Cycad Camp (pale). In this latter case, according to notes by field collectors, the dark mice were found in pine-oak forest, whereas the pale mice were trapped in tropical arid scrub forest. Two other localities from which dark specimens have been collected of which special note should be made are Finca San Salvador (29) and San Andres Tuxtla (32). Finca San Salvador is in the coastal mountains of Chiapas near the type locality of Liomys pinetorum, which was described as a dark species. San Andres Tuxtla is located on the Sierra Los Tuxtlas — an isolated group of mountains on the coast of Veracruz. Particularly pale specimens were from dry areas such as the Sonoran desert (1), along the Balsas River (18), and lowlands of the Isthmus of Tehuantepec (24). Specimens from the highest areas of the Isthmus (25) are darker than those from coastal lowlands but not as dark as specimens from higher elevations. In only one case are lowland specimens darker than those from the adjacent high¬ land; this involves specimens from the vicinity of San Bias, Nayarit, as compared with those from the vicinity of Tepic and Compostela, Nayarit. I feel, as did Hooper (1955:9), that color in and of itself is not a sufficient indicator of major patterns of geographic variation, but that it reflects variation with altitude and the ecological situations in which specimens were obtained. In Table 16. Geographic variation in dorsal coloration of 35 samples of Liomys pictus. Samples differ from those used for univariate and multivariate analyses. See discussion of coloration o/Liomys irroratus. GENOWAYS— SYSTEMATICS OF LIOMYS > o *0 m so o __ in u o rn o sd 00 00 m, o ' ' _ _ ' n q q q 00 C 00 G sd Os oo A 00 ON 63 d q A n A A NO A NO \C U OO o SO r- (N (N m C/5 SO m m m cn n (N o — < m a> A NO A sd A sd G A > ON OO 00 ON o sO u — rn m* A NO A NO A — c CD CD o T3 U CD 60 C 63 06 UJ c/5 n +1 c 63 D A) CL E CS c/5 o in s o o >n o oo G Os 00 L' 00 Os i o o d d q o o i NO «n n A NO r-' sd o ON 00 NO so r-~ «n r*i (N m CN o o o o o o o o +1 +1 +1 +1 +1 +1 +1 +1 (N m o 00 r 63 © ^ ~ E C - 63 (3 ■ - . C/5 — u ^ a. o < as c ” — . -J u l- ~ ~ 63 JO 63 63 63 ^ 3 63 o 63 63 o C/5 C/5 o £° o * G D .2 "a c n 00 o o q m, o o 00* > — a :n ic .£ <« 63 t; CD 63 y za:z JCQZ^hzU^Oh m m nC 00 Table 16. — Continued. 172 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY — t" t" 00 o b vd C N/"i rb b sb sb 00 o o 1 On m r- o r*N m Tj* — 00 r- o SO m •^f m 00 ^r m (N (N (N o o o o o o o o b o o o o +i +1 +1 +i +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 r^) q ON q o ro ON q r- o ON (N 00 00 sb lO b b r- NO b N 6 b b b sb sb o r- m oo q q o On — sD 00 00 o — • — ' 00 ^t m NO r~ m m m r^i o o o o o o o o o o o o o o +1 +1 +1 +i +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 On q q On m q NO NO <0 r- (N i/~> 00 m sb NO sb u 00 b NO NO NO b b sb b b b m q , 00 IO, O r- q — b b d — NO NO ON oo b d as ON b / _ ^ ^ ^ l^i NO, o o o o q i/~> O b rb b NO oo U ro oi b b d b 00 d — — — — — ■ — ■ — — q i i i i/'N i o o q NO o no, o b i o o ri d ON on rb o 00 as r^i b — ’ — ■ — ’ s — '' *— ' — w ' — ' ' — ' On nO o 00 O ON ^ r- r^i 00 r- m so OO q ON «/~i Ci On 00 r- 00 s O 00 — ■ o o o — — d o o o o o o o +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 00 r- NO (N ivo — NO 00 CJ X .2 c -3 rt > - ^ * '03 o 03 o o o o u o c o ^ ^ c <2 < u ^3 i Q t! c “ g > o x c — - C3 C3 c/3 o c £ CO <*> C/3 o J o Oil c o in m © NT, s cr» ON I NO no r- A (A NO NO 00 A A o >A d iA A d 1 1 O co 1 »n 6 q */~l A •t >A n nl in m CO oo m Tf it C/~i c/^i o o o o o d o o d o o o +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 o q r~ o — 00 n , _ s / _ s ^ o CO o o o C^i o o >n O q c^i A A oo 00 A oo ri NO On 00 NO o o o »A »A A d d n o m Tt (N NO >n m nj ^t NO o o o o o o o o o o d o — i +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 00 o O ^ 03 Jrt ► C® ^ H C OfvjrtOuj'—Ojj * .2 '3 * * ^ * .SP 07)jQ2Qottl 03 — 03 c o H c/5 cd a a3 X 3 H a 03 a 03 1) u u ° c 03 c/3 03 03 Q, •- « O 3 03 T3 a a n ^ .2 > <+- . sz a o (j on 03 > O u a A "cd CT3 « q U C/5 N(U u < o 03 .3 c cd cd cd 03 r- CJ cd (U cd a> 03 sor^oo on O — TO > > > > JO -a TO TO > T3 TO T3 c TO C 3 C 3 C 3 3 3 t5 3 3 3 3 ad «— > C, o 3 a u c c ’5) u. 03 E C/5 O D. _ O 3 « z § a 60 c u j3 *: 3 E a ^ i 03 C/D 03 c3 .2 j= 3 a u V 3 00 c sz o ■*— ' o c c 3 1) o r-~ t- 3" 3- 3 SZ 00 c _ D »/~> r- 00 o — Os r- ON C/D -X C/D O o ri *— < <4-. r^i m f*“i ■m 3 O ■— o o o* o t) o* o* o* ■o ■o r- r~ 00 Os 3 On NO r^ oc sO m n ri NO nO NO OX) u sO OC 00 00 oc oc 3 3 3 Q XN oc o aS O _ c*s 3- fN *-« m o £ a — d o o o o o o 4— > o as u OS jz U 00 i/S r- ON o /i O (N NO NO * 00 on SO 00 00 00 so vO NO ON ON r- (N (N r- (N i/'s o o o o o o o o o o o o o o o o o I I I I OX) C O x o £ -a c o X c c 00 c u- X C/5 c3 (U E cd o (U oc u- >> O N c o -*-» u c c/5 c o o a3 x u O £ o o o "O 03 1 — ] C/D aS C C/D O u as E aS u* 15 4—* — ( as « 4—* l 4-4 4-4 <+- J— < JO aS a aS • a OX) OC 00 4—* u 1-4 c d) J c 4> -J c Q Q as c/d aS C 4- o C ’5b u- as E U O u s — aS X aS E 7 I HE Fig. 35.— Two-dimensional projections of the first three principal components illustrating the phenetic position of the 20 samples of male Liomys salvini. Top, component I plotted against component II; bottom, component I plotted against component III. See Fig. 31 and text for key to samples. 228 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY • 2 •12 • #21 • 5 #1 *3 13 • 4 • 11 • 20 • 16 #14 #8 #18 ^•10 #6 • 15 • 7 #19 17# I in Fig. 36. — Two-dimensional projections of the first three principal components illustrating the phenetic position of the 21 samples of female Liomys salvini. Top, component I plotted against component II; bottom, component I plotted against component III. See Fig. 31 and text for key to samples. GENOWAYS— SYSTEMATICS OF LIOMYS 229 Fig. 37.— Three-dimensional projection of 20 samples of male Liomys salvini onto the first three principal components based upon a matrix of correlation among the 13 ex¬ ternal and cranial measurements and four qualitative cranial characters. Components I and II are indicated in the figure and component III is represented by height. See Fig. 31 and text for key to samples. closer to the front than samples 10, 15, and 17. Of note is the fact that sample 5, which occupies an intermediate geographic position between 4 and 6, falls at the lower end of the second group of samples and at a place approximately half way between samples 4 and 6. Sample 21, which was relatively “distant” from other members of the second group in the phenogram, appears to be separated from other members of the group in component III, but it should be remembered that this component represents only 9.92 per cent of the phenetic variation. Sample 1 3 is farthest removed from the group in component I, which represents most of the variation shown in this figure. The projection for females shows samples 1 to 4 forming one group with sample 5 again at a point almost halfway between sam¬ ples 4 and 6. The placement of sample 1 5 with 1 7, as with males, does not appear to be out of the question based on this plot, although these two samples were separated in the phenogram. Sample 18 occupies a place approximately inter¬ mediate between samples 17 and 14 as does sample 19 between 17 and 20. The three-dimensional projection does not reveal as well as does the distance pheno¬ gram a definite division of the second group into two subgroups, but the samples are quite widely scattered on the first principal component where most of the variation is represented; samples 6 to 10, 18, and 19 fall at the middle of the plot and other samples of the second subdivision in the phenogram fall on the left- hand side of the plot. Sample 21 is again most “distant” from other members of 230 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY Fig. 38. — Three-dimensional projection of 21 samples of female Liomys salvini onto the first three principal components based upon a matrix of correlation among the 13 external and cranial measurements and four qualitative cranial characters. Components I and II are indicated in the figure and component III is represented by height. See Fig. 31 and text for key to samples. the second group in the third component, but only 1 1.14 per cent of the phenetic variation is represented in this component. Variation in Color In Liomys salvini, as other species of the genus, variation in color appears to be more closely correlated with habitat than with systematic relationships (see Table 26 and Fig. 39). Specimens from highland areas and from the wet lowlands are dark colored, whereas specimens from dry lowlands are pale. The palest popula¬ tion is from Sacapulas, Guatemala, in the dry valley of the Rio Negro (see Good¬ win, 1934:39, pi. V, fig. 1). This population was described as Liomys anthonyi by Goodwin (1932a: 2), who remarked about the extremely pale coloration of these mice. Other species of mammals from this area also tend to be pale in coloration (Packard, 1958:400; Goodwin, 19326:3). Other pale populations are from along the arid valley of the Rio Motagua (see Land, 1 962: 1 -3, for a descrip¬ tion of this area), Lake Coatepeque in El Salvador, the Cosigiiina Peninsula of Nicaragua, and the vicinity of Tonala, Chiapas, and Reforma, Oaxaca. All of these areas are relatively low and arid. GENOWAYS— SYSTEMATICS OF LIOMYS 231 Fig. 39.— Geographic variation in red reflectance of the middorsal coloration of Liomys salvini. The palest sample is represented by an open symbol, the darkest sample by a com¬ pletely closed symbol, and the remaining samples by symbols that are expressed as a per¬ centage of the difference between these extremes. See Table 26 for areas represented by symbols. The darkest individuals of Liomys salvini are from the highlands of Costa Rica in the vicinity of San Jose. Other dark populations are from Hacienda Bellavista on one of the isolated volcanos in Nicaragua, the highlands of north-central Nica¬ ragua, highlands around Guatemala City, vicinity of San Salvador, El Salvador, two samples from the Caribbean lowlands (San Pedro Sula, Honduras, and vicin¬ ity of Santa Rosa and Villa Somoza, Nicaragua), and three samples along the Pacific coast of Chiapas and southwestern Guatemala. Again, these samples are scattered throughout the geographic range of the species and in many instances are separated from one another by populations of pale mice. Taxonomic Conclusions Based on the foregoing analyses, I recognize three subspecies of Liomys salvini. One occurs along the coast of Oaxaca, Chiapas, and southwestern Guatemala. The oldest available name for this taxon is Liomys salvini crispus Merriam, 1902, with type locality at Tonala, Chiapas. The second subspecies has an ex- Table 26. — Geographic variation in dorsal coloration of 25 samples o/Liomys salvini. Samples differ from those used for univariate and multi¬ variate analyses. See discussion of coloration o/Liomys irroratus. 232 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY 06 > U U U d 1 *0 q o m o 'S'l q A d »o >d d d A NO ■'fr NO un A id nf s"^ s—x ^ w On 'O' oo ON tj- m 't fN r- r~ NO N rn N co G-) m no q NO (N NO CO odd o o o o d o o o o o o +1 +1 +1 +1 +1 +1 +1 +i +i +1 +1 +1 +1 +1 oo rr 'O' o ON «/n NO Tj- q (N m CO id d d a rd iri rd NO »Oj id A h n « co o q cr, NO NO o q un NO d ri oo A i/~i rd Tt rd d 00 ON (N T-* — — 8 8 o © o in >o q vn o VO on 6 d s6 — j K ON r-* 00 NO NO 1 A i 1 1 1 o o ^ i o i o t o id q •A o o o id id 'O' Tf iri iri no" vri A A A A - — ^ 'W' NO N - ^ ^ ^ s-^ ' — ' rr m o o NO >D — ON o r4 ON 00 Tt 'O' c^l nf v~) r~ 00 NO NO (N »o nr odd o o o o o o o o o o o +1 +1 +1 +1 +1 +1 ■H +1 +1 +1 +1 +1 +1 +1 CD NO CO r^i O) oo ra — 00 00 q lo q no d d iri NO 00 NO 00 NO A A A r- r- — ■ q m 00 O in ON o 00 »Oi q — ‘ c d r-* 00 (N o 'i- ri ON — * A A On ^ ^ — /— N o q q q o q o O >r, o q o o A 6 fd ri fN Tl o ID, ID, d A VO A i ir, O O On VO ON 00 cd ON — ' o o ON ON On N - ^ — ^ s— ' ^ N— ^ T—t »— 1— H w w ' - ' ' - ^ 'O' oo oo o NO '1' no o r- o ro •o o t CCS u d U co¬ co O S x — >n Cl3 d of ao cc3 JO o m 00 0 0 ro Nt r^ rn *o m no ■si- r*o n >0 d d d d d d d d d d +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 c-~ on q ON 00 r*s NO ON o\ as >0 d rf no d d d d 0 so _ >0 00 r- — . NO tq O NO ON d d d d d (N d q O O <0 >0 O VT) >0 q 00 1 tq d | NO d d NO NO r-‘ 00 <0 *0 q q d d d 1 q l q »n d d d d d d IT) W w ■w r- »o NO ^3- 00 t" ON O m r^i r*N (N <^3 0 *P) q q q - — - q in >n NO (N d — d d d d d 1 ■ >— *- 1 — 1 *-i — O >0 d di 0 d 1 1 O >n 1 O 1 ON r-‘ Os d d d 00 06 d 00 ’’ ' w r — 00 O m ON 0 0 »o *0 O On 0*1 •—< — t— •H 03 i— i c/3 3 w CO ao C 03 on 03 c =3 _ap . co ^3 O T3 u x c3 03 p CCS CCS q 3 k— 0 on 3 3 00 03 .1 Sf ac cc3 ao cc3 u a = a j5 u ccS u ccj 0 ST. a zs a CJ z W Z > z z s— ^ ' — ' ' — ' r- 00 ON 0 — 0 d &. cc3 3 CCS 3 h- 5 '3 o 3 cd CCS a C c C/5 ■*—* C/5 CJ O CCS cc3 O 0 C/3 U on an £* U UL) 03 ON Higuerones, Pozos, San Francisco Esparta, San Jose, Santa Rosa) _ II 7.2 ± 0.72 (6.0-9.5) 16.5 4.1 ± 0.43 (3.5-5.S) 17.2 4.1 ± 0.46 (3.5-5.51 18.8 234 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY tensive geographic distribution, being found in central and eastern Guatemala, El Salvador, Honduras, north-central Nicaragua, Isla de Ometepe, Nicaragua, extreme southwestern Nicaragua, and in Costa Rica. The oldest available name for this widespread race is Liomys salvini salvini (Thomas, 1893), with type locality at Duenas, Guatemala. The third subspecies of L. salvini is confined to the western coast of Nicaragua to the north and west of Lake Nicaragua and Lake Managua and occurs on Isla de Zapatera in Lake Nicaragua. The appro¬ priate name for this subspecies is Liomys salvini vulcani (J. A. Allen, 1908), with type locality at Volcan de Chinandega, Nicaragua. Liomys salvini crispus Merriam, 1 902 1902. Liomys crispus Merriam, Proc. Biol. Soc. Washington, 15:49, 5 March. 1902. Liomys crispus setosus Merriam, Proc. Biol. Soc. Washington, 15:49, 5 March, holotype from Huehuetan, Chiapas. Holotype. — Adult male, skin and skull, USNM 75105, from Tonala, Chiapas; obtained on 7 August 1895 by E. W. Nelson and E. A. Goldman. Skin in good condition; skull in good condition except both zygomatic arches are broken. Measurements of holotype. — Total length, 210; length of tail, 99; length of hind foot, 27.5; greatest length of skull, 31.8; interorbital constriction, 6.4; mas¬ toid breadth, 13.8; length of nasals, 12.2; length of rostrum, 13.5; length of max¬ illary toothrow, 5.0; depth of braincase, 9.0; interparietal width, 8.6; inter¬ parietal length, 4.3. Distribution. — Coastal areas of southeastern Oaxaca, Chiapas, and south¬ western Guatemala from Reforma, Oaxaca, in the north to vicinity of Mazatenago, Guatemala, in the south (Fig. 40). Comparisons. — From Liomys salvini salvini geographically adjacent to the east and south, Liomys salvini crispus can be distinguished by its shorter total length and length of tail (compare values of samples 1-4 with those of 6-14 in Table 21) and by a unique set of qualitative cranial characters. Samples of cris¬ pus have slightly smaller means than samples of salvini in some cranial measure¬ ments, but in most instances these are at best small average differences. The only possible exceptions are length of nasals and length of rostrum, both of which are relatively short in crispus. No individuals of crispus examined had the inter¬ parietal bone divided, but at least some individuals had the bone divided in all populations of salvini (see Table 23). Samples of crispus have a much higher percentage of individuals with the posterior margin of the interparietal bone deeply notched (more than 64 per cent) than samples of salvini (all except sample 5, which is intermediate in several characters that will be discussed later, have less than 50 per cent with deep notches and most have less than 40 per cent — see Table 23). Comparison of Liomys salvini crispus with Liomys salvini vulcani is given in the account of that subspecies. Remarks. _ Merriam (1902:49) described crispus as a species distinct from salvini and Goldman (191 1:46-47) maintained this arrangement in his revision of the genus. The only difference that Goldman (1911:50-51) noted between the GENOWAYS — SYSTEM ATICS OF LIOMYS 235 two species,” however, was that salvini had a “distinctly tawny” dorsum, where¬ as specimens of crispus from Huehuetan, Chiapas, showed no trace of this color. My analysis of the variation in this group reveals that the relationships between these two taxa is best expressed by considering crispus as a subspecies of the earlier named salvini. L. s. crispus appears to intergrade with the nominal race along the southern coast of Guatemala (see discussion in the account of L. s. salvini). Both Merriam and Goldman recognized specimens from Huehuetan, Chiapas, as a race ot crispus distinct from the nominal subspecies from Tonala, Chiapas. The specimens from Huehuetan were known under the subspecific name setosus and were distinguished from the typical race by their darker color and larger size. Specimens from the vicinity of Tonala are paler than those from southern Chiapas, but color is a highly variable character and is not a good indicator of geographic variation. Males trom northern Chiapas are somewhat smaller than their counter¬ parts in the south, but females from the two areas are essentially the same size. Based on both univariate analysis and multivariate analysis, the relationship of the populations from Chiapas and adjacent areas in Guatemala and Oaxaca is best recognized by assigning them to a single taxon, Liomys salvini crispus. The small size of the males from the vicinity of Tonala, Chiapas, and Reforma, Oaxaca, is of interest because this is the area in which Liomys salvini is sympatric with Liomys pictus. In the account of Liomys pictus pictus, it was pointed out that males from southeastern Oaxaca and northwestern Chiapas were larger than males from adjacent populations of pictus. One explanation for this unusually large size of one species and small size of the other from the only area of sympatry would be that character displacement may be involved. A more detailed discus¬ sion of this situation is given in a following section on specific relationships. An adult male from 3 mi. W Mazatenago, Guatemala (KU 65032), was not in¬ cluded in the univariate and multivariate analyses because the specimen was from a geographical area intermediate between sample 4 and sample 6 and an a priori judgment could not be made as to which sample the specimen should be allocated. However, comparison of the measurements of the specimen from Mazatenago (total length, 211; length of tail, 107; greatest length of skull, 31.4; length of nasals, 11.5; length of rostrum, 13.3) with the values for specimens from sam¬ ples 4 and 6 reveals that it more closely agrees with values for specimens from sample 4, and therefore, should be assigned to the subspecies crispus. Also the specimen has an undivided interparietal bone and the posterior margin of the interparietal has a relatively deep notch; both of these characteristics are typical of crispus. An adult male (AMNH 79240) from San Lucas, approximately 40 kilometers east-northeast of Mazatenago, has large external and cranial measure¬ ments (total length, 255; length of tail, 127; length of nasals, 12.3; length of rostrum, 14.1) and other characteristics that are typical of L. 5. salvini. Based on this evidence it would appear that the two subspecies come into contact in this area somewhere along the slopes of the Guatemalan highlands between Maza¬ tenago and San Lucas. 236 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY Mean values for five adult males and adult (nonpregnant) females from southern Chiapas, respectively, for length of ear were 13.2 (12.0-14.0) and 13.4 (12.0-15.0), and for weight were 45.1 (37.2-55.0) and 37.0 (28.1-46.0). Specimens examined (151). — Guatemala: Hacienda California, San Marcos, 13 (AMNH), 3 mi. W Mazatenago, 1280 ft., Suchitepequez, 1 (KU). Chiapas: 11 mi. NW Escuintla, 100 ft., 4 (2 LACM, 2 UA); Estacion Vieja, 1 km. S Mapastepec, 46 m., 3 (UNAM); Guatimoc, 1 (ENCB); Huehuetan, 14 (USNM); 13.5 km. (Huixtla-Motozintla Road) NE Huixtla, 250 m., 1 (UMMZ); Mapastepec, 45 m., 28 (UMMZ); Pijijiapan, 10 m., 20 (UMMZ); 20 km. SE Pijijiapan, 3 (LACM); 1 mi. SE Puerto Madero, 4 (KU); Talisman, 3 (AMNH); Tapachula, 4 (AMNH); 16 km. SW Tapachula, 35 m„ 11 (KU); 12 mi. S Tapachula, 5 (AMNH); 6 mi. NW Tonala, 9 (KU); Tonala, 19 (USNM); approximately 12x/2 km. SE Tonala, 1 (LACM). Oaxaca: Reforma, 7 (UMMZ). Marginal records. — Oaxaca: Reforma. Chiapas: 6 mi. NW Tonala; Tonala-, approx. 12x/2 km. SE Tonala; Pijijiapan; 20 km. SE Pijijiapan; Mapastepec; 11 mi. NW Escuintla- 13.5 km. (Huixtla-Motozintla Road) NE Huixtla; Guatimoc. Guatemala: 3 mi. W Mazatenago, Hda. California hence northward along the coast. Liomys salvini salvini (Thomas, 1893) 1893. Heteromys salvini Thomas, Ann. Mag. Nat. Hist., ser. 6, 11:331, April. 1911. Liomys salvini, Goldman, N. Amer. Fauna, 34:50, 7 September. 1893. Heteromys salvini nigrescens Thomas, Ann. Mag. Nat. Hist., ser. 6, 12:234, Septem¬ ber; holotype from an unknown locality in Costa Rica (Goodwin, 1946:374, sug¬ gested that “the type came from an accessible part of Costa Rica, probably Escazu ). 1902. Liomys heterothrix Merriam, Proc. Biol. Soc. Washington, 15:50, 5 March; holo¬ type from San Pedro Sula, Cortes, Honduras. 1932. Liomys anthonyi Goodwin, Amer. Mus. Novit., 528:2, 23 May; holotype from Saca- pulas, 4500 ft., El Quiche, Guatemala. 1938. Liomys salvini aterrimus Goodwin, Amer. Mus. Novit., 987:4, 13 May; holotype from Sabanillas de Pirns, about 3730 ft., San Jose, Costa Rica. Holotype.— Adult male, skin and skull, BMNH 75.2.27.35, from Duenas, Sacatepequez, Guatemala; obtained on 31 July 1873 by Osbert Salvin. Skin in fair condition, but no external measurements listed; skull in good condition. Measurements of holotype. — Greatest length of skull, 32.8; zygomatic breadth, 15.0; interorbital constriction, 7.3; mastoid breadth, 14.1 ; length of nasals, 12.5; length of rostrum, 14.2; length of maxillary toothrow, 5.4; depth of braincase, 9.3; interparietal width, 9.9; interparietal length, 4.6. Distribution. — Guatemala (along the southern coast, in the highlands around Guatemala City, and in the valleys of the Rio Negro and Rio Motagua as far as San Pedro Sula, Honduras), most of El Salvador, south-central Honduras, north- central and central Nicaragua, Isla de Ometepe in Lake Nicaragua, extreme southwestern Nicaragua, and western and central Costa Rica (Fig. 40). Comparisons. — Comparison of Liomys salvini salvini with Liomys salvini crispus and Liomys salvini vulcani is given in the accounts of those subspecies. Remarks. — Liomys salvini salvini has an extensive geographic range in Central America. It is characterized by large external and cranial dimensions, at least some individuals in each sample having the interparietal bone divided, and a low percentage of individuals having the posterior margin of the interparietal GENOWAYS— SYSTEMATICS OF LIOMYS 237 Fig. 40. — Geographic distribution of subspecies of Liomys salvini : 1, L. s. crispus ; 2, L. s. salvini; 3, L. s. vulcani. The occurrence of L. s. salvini to the east of Lake Nicaragua is questionable at present. deeply notched. With few exceptions, samples of L. s. salvini exhibited rather uniform values for external and cranial measurements (Table 21); this was not true, however, for color (Table 26). Specimens assigned to L. s. salvini from southwestern Nicaragua and Costa Rica may not be in contact with other populations of the subspecies in central Nicaragua (vicinity of Boaco), because mice from the northwestern edge of Lake Nicaragua at Hacienda Mecatepe and from Isla de Zapatera appear to represent L. s. vulcani, thus causing a break in the distribution of salvini along the western side of Lake Nicaragua. Despite this apparent break in the geographic range, I have assigned specimens from the southern segment to salvini because of their close morphological resemblence to other members of that taxon. In both the univariate and multivariate analyses, specimens from southwestern Nicaragua and Costa Rica exhibited little or no differentiation from samples from central and north-central Nicaragua and south-central Honduras. There are several pos¬ sible explanations for this situation. One of the more likely would seem to be that the two segments of the population of L. s. salvini recently have been in contact along the western side of Lake Nicaragua and agriculturization, recent volcanic action, or some other disturbance has resulted in immigration of vulcani into this 238 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY area (or at least allowed the spread of vulcani-hke characteristics). Another pos¬ sibility is that the two segments are at present in contact around the eastern side of Lake Nicaragua, but the likelihood of this situation is lessened by the fact that a large collection of mammals from La Esperanza, just south of San Carlos at the southeastern edge of Lake Nicaragua, does not contain specimens of Liomys salvini. The possibility exists, however, that during a period when climatic con¬ ditions were different from at present, contact between the populations or coloni¬ zation of Costa Rica could have occurred along the eastern side of Lake Nica¬ ragua. Moreover, it could be that some genetic contact is maintained between the two populations across Lake Nicaragua as might be indicated by the fact that specimens from Isla de Ometepe are assignable to salvini (although in several measurements females from the island had values that were intermediate between those for vulcani and salvini ), although this would seem to be extremely unlikely. In order for this type of contact to be effective, large numbers of mice would need to be transported across the lake because the genetic complement of a small num¬ ber of invaders would be swamped by resident populations. Another possibility that should not be overlooked is that when additional material is available from the northwestern edge of Lake Nicaragua, especially from the vicinity of Grenada, these populations may be found to be properly assignable to salvini, not vulcani. Finally, an explanation that I regard as highly unlikely, but which cannot be com¬ pletely disregarded, is that these two populations have long been separated and that they have come to resemble each other morphologically owing to similar selective pressures. In summary, populations assigned to salvini from southwestern Nicaragua and Costa Rica, and those from central Nicaragua are probably not now in geo¬ graphic contact, but probably were in contact in the recent past. The dis¬ continuity evidently is sufficiently recent that the two segments exhibit little or no differentiation, certainly not enough to warrant taxonomic recognition. If for some reason the population in southwestern Nicaragua and Costa Rica were to be considered a distinct race, the name Liomys salvini nigrescens (Thomas, 1893) is available for it. Two species ( anthonyi Goodwin, 1932, and heterothrix Merriam, 1902) and two subspecies ( nigrescens Thomas, 1893, and aterrimus Goodwin, 1938) formerly considered to be distinct are placed herein as junior synonyms of salvini. Material from the type locality of L. anthonyi (Sacapulas, Guatemala) is included in sample 7 and material from the type locality of L. heterothrix (San Pedro Sula, Honduras) is included in sample 8. When these two samples were compared with one another and with material from the type locality of L. salvini (sample 6), little difference was noted between them in either the univariate or multivariate analysis. Males from the three samples are particularly close in the multivariate analysis. Specimens from these samples do, however, exhibit extreme variation in color. Those from the vicinity of the type locality of salvini are the darkest of this group (10.6 red reading), although specimens from the vicinity of Granados and Salama near the headwaters of the Rio Motagua and just north of the type locality are also dark (1 1.6 red reading) as are specimens from the Caribbean lowlands GENOWAYS— SYSTEMATICS OF LIOMYS 239 at San Pedro Sula, Honduras (1 1.2 red reading). Specimens from Sacapulas, in the dry valley ot the Rio Negro, are the palest of the species with a red reading of 16.5, also, specimens from along the Rio Motagua northeast of the type locality ot salvini are pale (red reading 13.6). There appear to be two disjunct pale and two disjunct dark populations existing in this area; recognition of all four of these populations as distinct would obscure their overall resemblence in mensural data and qualitative cranial characteristics. Liomys salvini nigrescens was named by Thomas ( 1 8936:234) on the basis of a single specimen from an unknown locality in Costa Rica. The holotype of nigrescens was described as differing from the holotype of salvini by its darker color and smaller size. Material from the highlands of Costa Rica where sub¬ sequent authors have supposed the holotype was obtained (see particularly Good¬ win, 1946:374) represents the darkest population of salvini that I have examined, but in mensural characteristics these specimens and the holotype of nigrescens (BMNH 69.7.19.6) fall within the range of variation of samples I have assigned to salvini. On the basis of the above data, I have placed nigrescens in the syn¬ onymy of salvini, although as was discussed earlier in this section, if the disjunct population from southwestern Nicaragua and Costa Rica should prove to be distinct, then nigrescens would be the oldest available name for this taxon. Goodwin ( 1 938:4-5) described Liomys salvini aterrimus on the basis of a single adult female from Sabanilla de Pirns, distinguishing it on basis of dark coloration and long tail. The holotype (FMNH 35211) is extremely dark in color, but it could be matched by other specimens from San Jose in similar pelage. The length of tail of the specimen does fall at the upper extreme for sample 21, but it is not much longer than in other members of the sample and can be matched in other samples (such as 12 and 13). On the basis of the data presently available, even though the specimen from Sabanilla de Pirns exhibits extreme variation in two characters, to accord it taxonomic recognition would obscure the numerous simi¬ larities in other characteristics, such as size and qualitative characters of the cranium, with mice from adjacent areas. Specimens from the southern coast of Guatemala (sample 5) show some char¬ acteristics that are intermediate between populations of crispus to the northwest along the coast and salvini in the highlands to the north and east. In several char¬ acters (length of tail, length of nasals, length of rostrum) of the univariate analysis, females more closely resembled crispus than salvini, although males had values that were similar to salvini in these characters. In the multivariate analyses, the females were more or less intermediate between samples of crispus and salvini , whereas males fell toward the lower limit of the salvini samples. A high percen¬ tage of specimens in sample 5 have the posterior margin of the interparietal bone deeply notched (66.7 per cent), which is characteristic of crispus, but one of the 13 specimens for which data were recorded had the interparietal bone divided, which is characteristic of salvini. The southern coast of Guatemala is an area of intergradation between the races crispus and salvini, but in overall characteristics the sample is somewhat more salvini- like and has been assigned to that race. The specimen from the marginal locality, Tiquisate, Guatemala, is young and sub- 240 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY specific allocation is impossible, but specimens from the nearby locality of Con¬ cepcion del Mar indicate affinities with salvini for specimens from the area. females from 8 km. N Las Maderas, Nicaragua, are intergrades between L. s. salvini in the highlands to the east and north and L. s. vulcani in the lowlands to the west and south. In many cranial measurements, females of sample 1 8 have values that are intermediate between those of samples 14 and 17, but in external measurements, interorbital constriction, and depth of braincase, they are more nearly like females of salvini from sample 14. In the multivariate analysis, sample 18 is near the lower limit of the range of variation for salvini, but it still falls within it and for that reason I have assigned the specimens to salvini. Two adult males from 11 mi. SE Dario, which is about 12 kilometers north of the Las Maderas locality, appear to represent typical L. s. salvini with no indication of intergradation with vulcani. Contact between populations of salvini and vulcani south of the Meseta de las Pueblos is discussed in the account of the latter sub¬ species. The sample from the vicinity of San Salvador, El Salvador (sample 10), is the one exception to the relatively uniform size of specimens of Liomys salvini salvini. In the univariate analysis both males and females (especially males) from the vicinity of San Salvador were uniformly smaller than specimens of the cor¬ responding sex from surrounding areas. The females do fall at the lower edge of the range of variation for samples of salvini in the multivariate analysis, but the males fall well below the limit of other salvini samples. Whether this is a result of a sampling error or a differentiated population in the vicinity of San Salvador will not be known until more material from this area is available. For the present, I believe it is best to assign these specimens to salvini. Specimens from Monte Rey, 22 kn. S San Jose, Costa Rica, represent the southernmost known record for the species. This locality is approximately 300 kilometers northwest of the nearest known locality of Liomys adspersus, the intervening area evidently uninhabited by Liomys. The average length of ear of 10 adult males and 10 adult females of Liomys salvini salvini from northern and central Nicaragua was, respectively, 14.4 (13.0- 16.0) and 14.0 (12.0-16.0); these same animals had mean weights of 54.9 (42.1- 65.2) and 43.5 (38.2-50.8). Specimens examined (914). — Costa Rica; Altos Escazu, San Jose, 3 (1 AMNH, 2 MCZ), i/2 km. S Bagaces, Guanacaste, 1 (TCWC); Boca del Barranca, Puntarenas, 1 (LACM); 1 mi. E Cangrejal, San Jose, 3 (LSU); Cerro de San Juan, 1200 ft., Guanacaste, 4 (UMMZ); Esparta, 300 ft., Puntarenas, 1 (FMNH); 7 mi. SW Filadelfia, Guanacaste, 4 (KU); Finca Coyolar, 5 km. N, 1V2 km. W Liberia, Guanacaste, 5 (LACM); Finca Jimenez, Guanacaste, 2 (MVZ); Vi mi. E Finca Jimenez Headquarters, 30 m., Guanacaste, 7 (UMMZ); 1 8/10 mi by road NE Las Canas, Guanacaste, 1 (LSU); 5 mi. N Liberia, Guanacaste, 1 (LACM); 4 4/10 mi. N Liberia, Guanacaste, 1 (LACM); 9 km. N Liberia, 4 km. E Interamerican Highway, Guanacaste, 2 (LACM); 4V2 mi. S Liberia, Guanacaste, 5 (LACM); Los Higuerones, San Jose, 3 (AMNH); Monte Rey, 22 km. S San Jose, 1100 m„ San Jose, 4 (KU); Playa del Cocos, Guanacaste, 4 (LACM); 3 km. S Playa del Cocos, Guanacaste, 12 (LACM); 1 V2 mi. SE Playa del Cocos, Guanacaste, 5 (LACM); Pozos, ca. 1 km. N Santa Ana, San Jose, 1 (LSU); Rio Bebedero, 2 km. S Bebedero, 5 m., Guanacaste, 1 (KU); R'10 Grande, Villa Colon, San Jose, 13 (10 FMNH, 3 USNM); Rio Tenorio, 3 mi. S, 10 mi. W GENOWAYS— SYSTEMATICS OF LIOMYS 241 Las Canas, 10 ft., Guanacaste, 1 (TCWC); Rio Teopisco, Pan American Highway 30 mi. S Nicaraguan border, 250 ft., Guanacaste, 2 (KU); Sabanilla de Pirns, San Jose, 1 (FMNH); San Francisco Esparta, Puntarenas, 1 (AMNH); San Juanillo, 100 ft., Guanacaste, 1 (UMMZ); 6 mi. SSE San Ignacio, San Jose, 1 (LSU); 1 mi. N Santa Ana, San Jose, 1 (LSU); cY/. 2 km. NW Santa Ana, San Jose, 5 (LSU); 3 mi. N Santa Rosa, Guanacaste, 2 (KU); Tempate, 150 ft., Guanacaste, 3 (2 MCZ, 1 UMMZ); 1 7/10 mi. by road W Tilaran, Guana¬ caste, 1 (LSU); unknown locality, 1 (BMNH). El Salvador: Barra de Santiago, sea level, Ahuachapan, 2 (MVZ); Divisadero, Morazan, 1 (UMMZ); % mi. NE Divisadero, 1000 ft., Morazan, 5 (MVZ); 1 mi. SE Divisadero, 850 ft., Morazan, 11 (MVZ); El Tablon, Lake Guija, 1450 ft., Santa Ana, 3 (MVZ); Hacienda Chilata, 2000 ft., Sonsonate, 10 (MVZ); Lake Coatepeque, Santa Ana, 43 (AMNH); SW edge Lake Olomega, 200 ft., San Miguel, 29 (28 MVZ, 1 UMMZ); 1 mi. S Los Planes, San Salva¬ dor, 2 (KU); Monte Cristo Mine, 700 ft., Morazan, 5 (MVZ); Mt. Cacaguatique, 3500 ft., San Miguel, 9 (8 MVZ, 1 UMMZ); Pine Peaks, 3 mi. W Volcan Conchagua, 3400 ft., La Union, 4 (3 MVZ, 1 UMMZ); Puerto del Triunfo, sea level, Usulutan, 8 (7 MVZ, 1 UMMZ); Rio Goascoran, 13° 30' N, 100 ft., La Union, 1 (MVZ); Rio San Miguel, 13° 25' N, 225 ft., San Miguel, 4 (MVZ); 2 mi. SE San Cristobal (Guatemala), 2950 ft., Santa Ana, 1 (KU), 1 mi. NW San Salvador, San Salvador, 13 (KU); 6 mi. E San Salvador, San Salva¬ dor, 3 (KU); 10 mi. NW Santa Tecla, La Libertad, 3 (KU); Volcan de San Miguel, 1900 ft San Miguel, 17(16 MVZ, 1 UMMZ). Guatemala: Antigua, Sacatepequez, 2 (AMNH); Astillero, 25 ft., Escuintla, 7 (KU); Cabanas, Zacapa, 5 (USNM); 5 mi. S Chiquimulilla, 200 ft., Santa Rosa, 1 (KU); Con¬ cepcion del Mar, Escuintla, 2 (FMNH); 2 mi. N, 1 mi. W Cuilapa, 2980 ft., Santa Rosa, 1 (KU); Duehas, Sacatepequez, 1 (BMNH); El Rancho, El Progreso, 7 (FMNH); 4 mi. W Escuintla, 880 ft., Escuintla, 2 (KU); Finca Cruz, 6 8/10 mi. SW Progreso, 1870 ft., El Progreso, 3 (USNM); Finca El Carnero, Cerro las Flores, 2100 ft., Jutiapa, 1 (UMMZ); Finca Los Arcos, Escuintla, 3 (USNM); Finca Valle-Lirios, Escuintla, 5 (USNM); Gualan’ Zacapa, 14 (ANSP); 4 mi. S Guatemala City, 4700 ft., Guatemala, 1 (KU); 5 mi. S Guate¬ mala City, 4950 ft., Guatemala, 4 (KU); 2 3/10 mi. W, (4 mi. N Iztapa, Escuintla, 6 (KU); Jocotan, 1350 ft., 3 (KU); 3 mi. E Jocotan, 1400 ft., Chiquimula, 24 (KU); Km. 24 on high¬ way S Guatemala City, Guatemala, 48 (18 AMNH, 30 KU); Km. 27 on highway S Guate¬ mala City, 4100 ft., Guatemala, 1 (KU); Km. 33 on highway S Guatemala City, 4000 ft., Escuintla, 1 (KU); Km. 52 on highway S Guatemala City, 1650 ft., Escuintla, 1 (KU); S side Lake Amatitlan, 4200 ft., Guatemala, 12 (USNM); Lake Atescatempa, 2400 ft., Jutiapa, 20 (USNM); La Primavera, Alta Verapaz, 2 (AMNH); 1 mi. SE Monogoy, Jutiapa, 10 (KU); Progreso, El Progreso, 1 (AMNH); Puente Punta Gordo, 2050 ft., El Progreso, 1 (KU); 1 mi. S Rabinal, 3450 ft., Baja Verapaz, 2 (KU); Rio Grande, 3 mi. S, IV2 mi. W Granados, 2000 ft., Baja Verapaz, 7 (KU); Rio Guacalate, 1 mi. W Masagua, Escuintla, 5 (USNM); Rio Santiago, 152 km. NE Guatemala City, 500 ft., Zacapa, 5 (KU); Rio Wite, 5 2/10 mi' E Cabanas, Zacapa, 5 (USNM); Sacapulas, El Quiche, 21 (AMNH); V2 mi. N, 1 mi. E Salama, 3200 ft., Baja Verapaz, 2 (KU); 2V2 mi. W, 2(4 mi. N San Cristobal, 2900 ft., Jutiapa, 4 (KU); San Jose, 15 ft., Escuintla, 9 (USNM); San Lucas, Solola, 3 (AMNH); Tiquisate, Escuintla, 1 (FMNH); Trujillo, Zacapa, 3 (USNM); Volcan San Lucas, Solola 1 (AMNH). Honduras: Catacamas, Olanche, 1 (AMNH); Cerro de los Cuches, Sabana Grande, Francisco Morazan 8 (AMNH); 1 mi. S Comayaguela, 1000 m., Francisco Morazan. 1 (TCWC); El Caliche Orica, Francisco Morazan, 12 (AMNH); Escuela Agricultura Pan- americana, Francisco Morazan, 14 (MCZ); Hatillo, Francisco Morazan, 2 (MCZ); La Cueva Archaga, Francisco Morazan, 6 (AMNH); La Flor Archaga, Francisco Morazan, 10 (MCZ); La Piedra de Jesus, Sabana Grande, Francisco Morazan, 66 (AMNH); 2 mi. S La Venta, 420 m., Francisco Morazan, 2 (TCWC); Monte Redondo, Francisco Morazan, 47 (12 AMNH, 8 FMNH, 20 MCZ, 4 UMMZ, 3 USNM); Sabana Grande, Francisco Morazan, 1 (AMNH); San Pedro Sula, Cortes, 20 (USNM); 12 mi. N Tegucigalpa, 2800 ft.. Francisco Morazan, 1 (TCWC); Toncontin, Francisco Morazan, 1 (MCZ). 242 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY Nicaragua: Wi km. W Alta Gracia, 110 m., Isla de Ometepe, Rivas, 6 (KU); 4 km. S, \Vi km. E Alta Gracia, 40 m., Isla de Ometepe, Rivas, 13 (KU); Finca Amayo, 13 km. S, 14 km. E Rivas, 40 m„ Rivas, 7 (KU); 14 km. S Boaco, 220 m., Boaco, 2 (KU); 8 mi. S Condega, Estefi, 6 (KU); Daraili, 5 km. N, 14 km. E Condega, 940 m., Estefi, 5 (KU); 11 mi. SE Dario, Matagalpa, 7 (KU); 1 km. NE Esquipulas, 420 m., Matagalpa, 20 (KU); 9 mi. NNE Estefi, Estefi, 7 (KU); Hacienda Tepeyac, Matagalpa, 2 (1 KU, 1 USNM); 1 mi. NW Jinotega, Jinotega, 1 (KU); La Danta, 1 km. N, 5 km. E Esquipulas, 780 m., Matagalpa, 2 (KU); 8 km. (by road) N Las Maderas, 380 m., Managua, 20 (KU); Matagalpa, Matagalpa, 1 (AMNH); 2 km. N Merida, 40 m., Isla de Ometepe, Rivas, 2 (KU); 2 km. N, 3 km. E Merida, 200 m., Isla de Ometepe, Rivas, 7 (KU); Moyogalpa, NW end Isla de Ometepe, 40 m., Rivas, 14 (KU); 5 km. E Moyogalpa, Isla de Ometepe, Rivas, 3 (KU); 6 km. E Moyo¬ galpa, Isla de Ometepe, Rivas, 19 (KU); Quilafi, Nueva Segovia, 2 (AMNH); Rio Javillo, 3 km. N, 4 km. W Sapoa, 40 m., Rivas, 3 (KU); 11 km. S, 3 km. E Rivas, 50 m., Rivas, 1 (KU); 3 mi. SE San Pablo, Rivas, 1 (KU); 3 mi. E San Ramon, Matagalpa, 1 (KU); Santa Rosa, 17 km. N, 15 km. E Boaco, 300 m., Boaco, 17 (KU); 1 km. NW Sapoa, 40 m., Rivas, 1 (KU); Savala, Matagalpa, 1 (AMNH); Sebaco, Matagalpa, 2 (AMNH); Uluce, Matagalpa, 6 (AMNH); 1 km. N, 2Vi km. W Villa Somoza, 330 m„ Chontales, 8 (KU). Additional records. — Costa Rica: Canas, Guanacaste (Esquivel et al., 1967:954); Cangre- jal, San Jose (Esquivel et al., 1967:954); Finca Coyolar, 5 km. N, 4 km. W Liberia, Guana- caste (Webb and Loomis, 1971:5); 8.3 km. N Liberia, Guanacaste (Webb and Loomis, 1971:36); 2 km. W Liberia, Guanacaste (Geest and Loomis, 1968:38); 7.3 km. S Liberia, Guanacaste (Geest and Loomis, 1968:38); 7.5 km. S Liberia, Guanacaste (Webb and Loomis, 1971:36); 5 km. NW Tilaran (Geest and Loomis, 1968:25); Santa Ana, San Jose (Esquival et al., 1967:954). El Salvador (Felten, 1957:153-155): Amate de Campo, La Paz; Finca Raquelina, Ahuachapan; Hacienda Nancuchiname, Usulutan; Hacienda San Antonio, Sonsonate; Isla de la Cabra, Santa Ana; Laguna de Guija, Santa Ana; San Salvador, San Salvador. Honduras: El Zapote, Francisco Morazan (Goodwin, 1942:156). Marginal records. — Honduras: San Pedro Sula. Guatemala: 3 mi. E Jocotan. El Salva¬ dor: El Tablon; Lake Coatepeque; 10 mi. NW Santa Tecla; 1 mi. NW San Salvador; 6 mi. E San Salvador; Mt. Cacaguatique. Honduras: Monte Redondo; El Caliche Orica; Cata- camas; Escuela Agricultura Panamericana. Nicaragua: Quilafi; Savala; Santa Rosa, 17 km. N, 15 km. E Boaco; 1 km. N, 1V2 km. W Villa Somoza; 2 km. N, 3 km. E Merida, Isla de Ometepe. Costa Rica: 9 km. N Liberia, 4 km. E Interamerican Highway; 1 7/10 mi. by road W Tilaran; 1 mi. N Santa Ana; Los Higuerones\ Monte Rey, 22 km. S San Jose; Sabanilla de Pirns; Boca del Barranca; San Juanillo hence northward along the coast. Nicaragua: 3 mi. SE San Pablo; 14 km. S Boaco; 8 km. (by road) N Las Maderas; 9 mi. NNE Estefi'. Honduras: La Piedra de Jesus. El Salvador: Rio Goascoran, 13° 30' N; Pine Peaks, 3 mi. W Volcan Conchagua hence northward along the coast. Guatemala: Tiquisate; San Lucas; Sacapulas; La Primavera; Vi mi. N, 1 mi. E Salama; Rio Santiago, 152 km. NE Guatemala City (17 mi. NE Rio Hondo)', Gualan. Liomys salvini vulcani (J. A. Allen, 1908) 1908. Heteromys vulcani J. A. Allen, Bull. Amer. Mus. Nat. Hist., 24:652, 13 October. 1946. Liomys salvini vulcani, Goodwin, Bull. Amer. Mus. Nat. Hist., 87:374, 31 Decem¬ ber. Holotype. — Adult female, skin and skull, AMNH 28315, from Volcan de Chinandega, about 4000 ft., Chinandega, Nicaragua; obtained on 7 May 1909 by W. B. Richardson. Skin in fair condition; skull in poor condition. The skull was broken in half between the orbits and subsequently was glued together; the left zygomatic arch is completely missing, the right zygomatic arch has the middle section missing, the tip of right nasal bone is missing as is the posterior portion of left dentary and the lower left third molar. GENOWAYS— SYSTEMATICS OF LIOMYS 243 Measurements of holotype. — Total length, 220; length of tail, 1 10; length of hind foot, 25; interorbital constriction, 6.8; mastoid breadth, 13.5; length of nasals, 11.1; length of rostrum, 12.5; length of maxillary toothrow, 4.9; inter¬ parietal width, 8.4; interparietal length, 4.0. Distribution. — This subspecies is contined to western Nicaragua — on the volcanos that make up the Cordillera los Marrabios and the lowland to the west of them, west of Lake Managua, on the Meseta de los Pueblos west of Lake Nica¬ ragua, and on Isla de Zapatera (Fig. 40). Comparisons. — From Liomys salvini salvini, the subspecies vulcani can be distinguished by its consistently smaller size in both external and cranial measurements (compare values for samples 15, 17 with those for 13, 14, 16, 20, 21 in Table 21). There does not appear to be any difference between the two subspecies in qualitative cranial characters. Compared with Liomys salvini crispus of Oaxaca, Chiapas, and Guatemala, Liomys salvini vulcani averages longer in total length and length of tail, but smaller in measurements of the cranium such as greatest length of skull, zygo¬ matic breadth, and depth of braincase (compare values of samples 15, 17 with those of samples 1-4 in Table 21). None of 45 specimens of Liomys salvini crispus for which data were recorded had the interparietal bone divided, whereas four individuals of 41 examined from sample 15, and 10 individuals of 81 ex¬ amined from sample 17 (samples of vulcani) had the bone divided (Table 23). The two samples of vulcani have 29.3 and 14.6 per cent of the individuals with the posterior margin of the interparietal bone deeply notched, whereas the lowest percentage for crispus was 64.0 for sample 1 (see Table 23). Remarks. — This small-sized subspecies is confined to lowlands of western Nicaragua and adjacent areas such as Cordillera los Marrabios, Meseta de los Pueblos, and Isla de Zapatera. J. A. Allen described vulcani as a full species in 1908 and it was recognized as such until Goodwin (1946:375) reduced vulcani to a subspecific status under salvini and assigned to it specimens from Guanacaste, Costa Rica. The name vulcani stood in the literature as a valid species for many years not because of its many unique characteristics, but because the type series was in poor condition and spiny pocket mice were unreported from elsewhere in Nicaragua. Not until field parties from The University of Kansas, under the aegis of a grant from the Army Research and Development Command, began collecting mammals in Nicaragua in 1964 did the widespread distribution of the species become apparent and delineation of the geographic range of vulcani be¬ come possible. It now appears that L. s. vulcani approaches the southern end of its distribution in the vicinity of La Calera (near Nandaime), Nicaragua, and that specimens from the Guanacaste region of Costa Rica formerly assigned to vulcani are best considered as representatives of L. s. salvini. Many of the areas, such as northwest of Lake Managua in the vicinity of Lar- reynaga and the lowland between Lake Managua and Lake Nicaragua, in which intergradation between vulcani and salvini is suspected, unfortunately are un¬ represented by specimens of spiny pocket mice at the present time. However, specimens from 8 km. N Las Maderas are clearly intergrades between vulcani 244 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY and salvini and may indicate that vulcani occupies much of the lowland areas around Lake Managua and makes contact with salvini in the foothills of the Cordillera Central. The distribution of L. s. vulcani appears to extend eastward on the Meseta de los Pueblos to the south of Managua in the vicinity of Volcan de Mombacho on the northwestern edge of Lake Nicaragua, thereby interrupting the geographical range of L. s. salvini from east of Lake Nicaragua and southwestern Nicaragua and Costa Rica. (The significance of the interruption of the range of L. s. salvini is discussed more fully in the account of that subspecies.) No adult specimens are available from Hacienda Mecatepe, just to the south of Volcan de Mombacho, but two subadult males (KU 108195-96) that I would judge to be nearly adult size have small external and cranial measurements (total length, 210, 199; greatest length of skull, 31.0, 30.7), and appear to be referable to vulcani. The one specimen from Isla de Zapatera (KU 71222) available for measuring (the other is in alcohol) is a young animal, but it is small (greatest length of skull, 29.8) and I tentatively refer the population from the island to vulcani until adult material becomes available for study. An adult male from La Calera, Nicaragua (KU 108197), has small external and cranial measurements (total length, 220; greatest length of skull, 31.4) and agrees well with typical specimens of vulcani. However, the one specimen from 3 mi. SE San Pablo, Rivas (adult male, KU 71223), only a few miles south of La Calera, is relatively large (total length, 237; greatest length of skull, 32.0) and, accordingly, has been referred to salvini. The zone of contact between these two races in the south is probably in the area between La Calera and San Pedro near the southern edge of the Meseta de los Pueblos. Ten adult males and 10 nonpregnant, adult females from northwestern Nica¬ ragua averaged, respectively, 14.2 (13.0-15.0) and 13.9 (12.0-16.0) for length of ear and 48.6 (36.0-60.5) and 37.1 (32.5-47.7) for weight. Specimens examined (316). — Nicaragua: AVi km. N Cosigiiina, 15 m., Chinandega, 10 (KU); 3 km. N, 4 km. W Diriamba, 600 m„ Carazo, 28 (KU); 3 mi. SE El Crucero, Managua, 2 (KU); El Paraiso, 1 km. N Cosigiiina, 20 m„ Chinandega, 4 (KU); Hacienda Azacualpa, Managua, 7 (KU); Hacienda Bellavista, 720 m., Volcan Casita, 46 (28 KU, 18 UCLA); Hacienda Corpus Christi, Managua, 16 (USNM); Hacienda Cutirre, Volcan Mom¬ bacho, 1400 ft., Granada, 1 (UCLA); Hacienda Las Colinas, 4 mi. WNW Puerto Momo- tambo, Leon, 35 (8 KU, 2 UA, 25 USNM); Hacienda Mecatepe, Granada, 3 (KU); NW side Isla de Zapatera, Granada, 2 (KU); La Calera, Nandaime, 2 (KU); Lake Jiloa, Managua, 20 (AMNH); Leon, Leon, 1 (AMNH); 5 mi. NW Managua, Managua, 14 (KU); 4 mi. W Managua, Managua, 30 (KU); 3 mi. SW Managua, Managua, 67 (KU); 1 mi. SE Masachapa, in Carazo, 1 (KU); 1 mi. ENE Poneloya, Leon, 4 (KU); 2 mi. N Sabana Grande, Managua, 3 (KU); San Antonio, 35 m„ Chinandega, 6 (KU); 6 mi. SE Tamarindo, Leon, 4 (KU); Volcan de Chinandega, Chinandega, 10 (AMNH). Marginal records. — Nicaragua: AVi km. N Cosigiiina; Volcan de Chinandega; Hacienda Bellavista, Volcan Casita; Hacienda Las Colinas, 4 mi. WNW Puerto Momotambo; Hacienda Corpus Christi; 2 mi. N Sabana Grande; Hacienda Cutirre, Volcan Mombacho; NW side Isla de Zapatera; La Calera, Nandaime; 1 mi. SE Masachapa. GENOWAYS— SYSTEMATICS OF LIOMYS 245 Liomys adspersus Panamanian Spiny Pocket Mouse This species occupies a restricted geographic range in central Panama. Speci¬ mens are available from as far west as Guabula and as far east as Chepo. The species occurs mainly in the Pacific drainage, although it is known from several localities in the headwaters ot the Caribbean drainage in the vicinity of the Canal Zone. Diagnosis External and cranial measurements large; premolars similar in structure to those of L. salvinr, baculum with large rounded base, shaft oval to a point just posterior to the slightly upturned tip where it is dorsoventrally flattened; glans penis medium-sized in comparison with length of baculum, tip of glans short, glans highly sculptured and with deeply incised ventral folds; urethral lappets bilobed; 2N = 56; FN probably 84; head of spermatozoon short with bluntly rounded apex and distinct neck between head and midpiece; wings of pterygoids narrow; parasitized by the anopluran species Fahrenholzia fairchildi; six plantar tubercles; upper parts usually chocolate brown (some paler individuals may show grayish tones); no lateral stripe; underparts white; hairs on back curled upward so as to be conspicuous above spines. Comparisons For comparisons of Liomys adspersus with L. irroratus, L. pictus, L. specta- bilis, and L. salvini, see the accounts of those species. Ecology Flandley (1966:778) stated that Liomys adspersus was abundant in the semi- arid savannah country of the Pacific coast of western and central Panama where it was commonest in thorny scrub and weedy fields (see also Bole, 1937:164). An extensive field study of this species was conducted by Fleming (1969, 1970, 1971) in the Canal Zone. He found adspersus to be the commonest rodent in the Rodman forest with its population density ranging from 5.4 to 1 1.0 individuals per hectare. The average home range was 0.56 hectare with no significant dif¬ ference between the size of the home range of males and females. Examination of stomach contents of more than 30 specimens revealed that these pocket mice fed on seeds, other plant material, and insects. Geographic Variation Univariate Analysis Adult specimens from throughout the geographic range of Liomys adspersus were grouped in three samples for analysis. The samples were composed of specimens labeled with reference to the following locations (Fig. 41): sample 1 — Canal Zone (Balboa, Fort Kobbe, Madden Road, Rio Chagres, Summit) and Panama (Cerro Azul, Rio Hato); sample 2 — Panama (Sante Fe, San Fran- 246 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY Fig. 41. — Approximate geographic areas included in the three samples of Liomys adspersus. See text for localities included in each sample. cisco, Tole); sample 3 — Panama (Guanico). Table 27 gives standard statistics for the samples of Liomys adspersus. External Measurements Males from the Azuero Peninsula (3) and central Panama (2) average greater in total length (Tables 27, 28) than do males from the Canal Zone (1). Females exhibit this same pattern, but specimens from sample 2 are somewhat larger than those from 3. For length of tail, (Tables 27, 28) males had two nonsignificant subsets — samples 3 and 2 and samples 2 and 1. Females also had two nonsignifi¬ cant subsets but these are composed of samples 3 and 2 and a nonoverlapping subset of sample 1 . No significant differences were found between the means for males for length of hind foot (Table 27). Females exhibited two overlapping subsets (Table 28) as follows: samples 2 and 3 and samples 3 and 1. Cranial Measurements No significant differences were found between the means of greatest length of skull for both males and females (Table 27). The means for males from samples 2 and 3 are a millimeter or more longer than those from sample 1, but for females the means are not especially different. Males and females both exhibit two over¬ lapping nonsignificant subsets for zygomatic breadth (Table 28). However, for males sample 3 has the largest mean and sample 1 has the smallest mean, whereas for females the reverse is true. Females show no significant differences between the means of the samples for the remaining eight cranial measurements (Table 27). For these measurements, females in sample 1 had the largest mean value for three (interorbital constriction, GENOWAYS— SYSTEM ATICS OF LIOMYS 247 Table 27. — Geographic variation in external and cranial measurements among 3 samples of Liomys adspersus. See text for key to sample numbers. Sample numbei t Males Females * N Mean Range 2 SE N Mean Range 2 SE Total length 1 9 249.6 232.0-264.0 7.30 10 238.6 222.0-264.0 8.18 2 2 265.0 264.0-266.0 2.00 5 253.2 247.0-265.0 6.46 3 18 265.9 248.0-285.0 5.29 6 249.7 244.0-264.0 5.99 Length of tail 1 9 1 19.9 107.0-130.0 6.05 10 118.2 109.0-130.0 3.83 2 2 130.0 127.0-133.0 6.00 5 127.8 124.0-131.0 2.93 3 18 136.6 123.0-148.0 3.37 6 128.8 124.0-138.0 4.54 Length of hind foot 1 9 29.9 27.0-32.5 1.09 14 29.5 28.0-31.0 0.62 2 5 30.6 29.0-32.0 1.36 5 30.8 30.0-32.0 0.75 3 21 31.4 26.0-34.0 0.74 8 30.6 29.0-32.0 0.92 Greatest length of skull 1 8 34.8 32.2-36.3 0.97 15 34.6 32.9-35.7 0.46 2 5 35.9 34.5-37.3 0.91 4 35.0 33.6-36.2 1.06 3 20 35.8 33.4-38.9 0.64 7 34.7 34.3-35.1 0.24 Zygomatic breadth 1 4 15.9 15.1-16.6 0.62 7 16.8 16.3-17.5 0.38 2 3 16.7 16.4-16.9 0.31 3 16.6 16.4-17.0 0.40 3 5 17.0 16.6-17.4 0.33 3 16.0 15.8-16.2 0.23 Interorbital constriction 1 10 7.5 7.1-7. 8 0.14 16 7.6 7.0-8. 2 0.18 2 5 7.6 7.4-8. 1 0.25 4 7.5 7. 1-7.8 0.30 3 21 7.4 7. 0-8.0 0.12 8 7.5 7. 3-7. 8 0.12 Mastoid breadth 1 9 14.5 13.8-15.0 0.24 15 14.5 14.1-14.8 0.12 2 5 15.1 14.7-15.6 0.35 4 14.9 14.8-14.9 0.05 3 20 14.5 14.0-15.3 0.14 8 14.5 13.8-15.2 0.28 Length of nasals 1 10 14.1 12.8-15.0 0.45 16 14.0 13.0-15.2 0.31 2 5 14.5 13.7-15.3 0.57 4 14.2 13.6-15.0 0.61 3 20 15.1 13.9-17.3 0.31 7 14.1 13.5-14.8 0.32 Length of rostrum 1 8 15.7 15.0-16.4 0.44 13 15.5 14.5-16.3 0.26 2 5 16.0 15.2-16.8 0.56 3 15.4 14.6-16.0 0.85 3 14 16.1 15.0-17.9 0.44 5 15.2 14.7-15.5 0.29 248 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY Table 27. — Continued. Length of maxillary toothrow 1 9 5.3 5.0-5. 5 0.11 16 5.3 5.0-5. 5 0.09 2 5 5.3 5. 2-5. 5 0.10 4 5.3 5. 1-5.4 0.13 3 20 5.4 4.9-5. 9 0.11 8 5.3 5. 1-5.6 0.15 Depth of braincase 1 8 9.4 8.8-10.2 0.32 14 9.3 8. 7-9. 9 0.19 2 5 9.8 9.3-10.2 0.30 4 9.4 9.0-9. 8 0.39 3 20 9.5 8.9-10.1 0.14 8 9.2 8. 9-9. 5 0.16 Interparietal width 1 6 8.6 8. 0-9.2 0.36 14 8.5 7. 7-9.2 0.23 2 5 8.3 7. 7-9.3 0.64 4 8.5 7. 8-9.6 0.83 3 15 8.2 7. 2-8.9 0.22 8 8.1 7. 5-8. 5 0.30 Interparietal length 1 6 4.8 4.4-5. 2 0.27 15 4.6 3. 8-5.0 0.18 2 5 4.5 4.0-4. 9 0.29 4 4.3 3. 9-5.0 0.48 3 19 4.4 3. 4-4. 8 0.19 8 4.4 4.0-5. 2 0.27 interparietal width, interparietal length), those of sample 2 had the largest value for four measurements (mastoid breadth, length of nasals, length of rostrum, and depth of braincase), and females from sample 3 averaged largest in one (length of maxillary toothrow). Males exhibit significant differences between the sample means for mastoid breadth and length of nasals (Table 28), but not for the remaining six measure¬ ments. For mastoid breadth, sample 2 forms a subset distinct from samples 1 and 3. Two overlapping subsets are formed for length of nasals. Mice from each of the three samples averaged largest for two of the remaining six measurements (Table 27) as follows: sample 1, interparietal width and interparietal length; sample 2, interorbital constriction and depth of braincase; sample 3, length of rostrum and length of maxillary toothrow. Qualitative Cranial Characters Condition of interparietal bone (Table 29). — None of the specimens examined from the vicinity of the Canal Zone (1) have a divided interparietal bone. A relatively high percentage (27.3) of individuals from the Azuero Peninsula (3) do have the interparietal bone divided, and mice from the geographically inter¬ mediate sample (2) have an intermediate percentage (12.5) with the interparietal bone divided in half. All three samples have a high percentage of individuals with the posterior margin of the interparietal bone notched. Condition of posterior nasal region (Table 30). — All specimens examined had the posterior margin of the nasal bones truncate in shape. All individuals examined but two from sample 3 had the premaxillary bones longer than the nasal bones. GENOWAYS— SYSTEMATICS OF LIOMYS 249 Table 28. Results of SS-STP analyses for those measurements of Liomys adspersus for which theie was found to be a significant difference between the sample means. Vertical lines to the right of each array of means connect maximally nonsignificant subsets at the 0.05 level. See text for key to sample numbers. Males Females Sample Sample number Means Results SS-STP number Means Results SS-STP Total length 3 265.9 2 253.2 2 265.0 3 249.7 1 249.6 1 238.6 Length of tail 3 136.6 3 128.8 2 130.0 2 127.8 1 119.9 1 118.2 | Length of hind foot 2 30.8 3 30.6 1 29.5 Zygomatic breadth 3 17.0 1 16.8 2 16.7 2 16.6 1 15.9 3 16.0 Mastoid breadth 2 15.1 | 1 14.5 3 14.5 Length of nasals 3 15.1 2 14.5 1 14.1 Table 29. — Geographical variation in the configuration of the interparietal bone of Liomys adspersus. Figures are given in per cent of total number for each sample. See text for key to sample numbers. Interparietal bone Posterior margin of interparietal bone Sample N Undivided Divided N Notched Slightly notched Unnotched 1 20 100.0 0 21 81.0 14.3 4.7 2 8 87.5 12.5 8 75.0 25.0 0 3 22 72.7 27.3 27 66.7 22.2 11.1 250 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY Table 30. — Geographical variation in bones of the posterior portion of the nasal¬ premaxillary complex 0/ Liomys adspersus. Figures are given in per cent of total number for each sample. See text for key to sample numbers. Sample Shape of posterior margin of nasals Length of premaxillary bones N Emarginate Rounded Truncate N Longer then nasals Equal to nasals 1 22 0 0 100.0 22 100.0 0 2 8 0 0 100.0 8 100.0 0 3 27 0 0 100.0 27 92.6 7.4 Multivariate Analysis Means of the 13 external and cranial measurements and four qualitative cranial characters were used in the NT-SYS multivariate analysis program. The three samples of Liomys adspersus were analyzed along with the 21 samples of Liomys salvini so that the relationship of these two species could be studied. Distance and correlation matrices were generated for both males and females from which phenograms were prepared. The distance phenograms are presented in Fig. 42. The first three principal components were computed from a matrix of correlation among the 17 characters; these first three components combine to express 79.89 per cent of the phenetic variation in males and 81.99 per cent in females. Distance coefficients between samples of male Liomys adspersus were as follows: between 1 and 2, 0.837; between 2 and 3, 0.735; and between 1 and 3, 1.094. The same values for samples of females were 0.643, 0.687, and 0.860. The distance between these three samples (in numerical order) and the nearest sample of Liomys salvini (sample 21) were for males and females, respectively, 1.489, 1.613, 1.968, 1.762, 1.975, and 1.574. The distance phenogram (Fig. 42) shows the three samples of Liomys adsper¬ sus “distantly” separated from samples of Liomys salvini. Within the cluster of adspersus samples, sample 1 is phenetically most distinct for males and sample 3 for females. In the principal component analysis, the amount of phenetic variation ex¬ pressed in the first three components for male and female Liomys adspersus and Liomys salvini, respectively, was 60.26 and 61.51 per cent for component I, 1 3.90 and 1 3.75 per cent for component II, and 5.74 and 6.72 per cent for com¬ ponent III. On the three-dimensional plots (Figs. 43, 44), it can be seen that the three samples of L. adspersus are separated by a considerable “distance” along the first principal component from the samples of L. salvini. This component expresses most of the phenetic variation and is most heavily influenced by size. Sample 1 of Liomys adspersus is somewhat removed from the other two samples along the first principal component for males, but for females the three samples are grouped near each other. GENOWAYS— SYSTEMATICS OF LIOMYS 251 i _ 2.28 1.88 1.48 1.08 0.68 0.28 Fig. 42. — Phenograms of numbered samples (see Figs. 31 and 41 and text) of Liomys adspersus (labeled Al, A2, and A3) and Liomys salvini (males left, females right) computed from distance matrices based on standardized characters and cluster by unweighted pair- group method using arithmetic averages (UPMGA). The cophenetic correlation coefficient for the phenogram for males is 0.917 and for females is 0.913. Fig. 43. — Three-dimensional projection of three samples of male Liomys adspersus (labeled LAI, LA2, and LA3) and 20 samples of Liomys salvini onto the first three principal components based upon a matrix of correlation among the 13 external and cranial measurements and four qualitative cranial characters. Components I and II are indicated in the figure and component III is represented by height. See Figs. 31 and 41 and text for key to samples. 252 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY Fig. 44. — Three-dimensional projection of three samples of female Liomys adspersus (labeled LAI, LA2, and LA3) and 21 samples of Liomys salvini onto the first three principal components based upon a matrix of correlation among the 13 external and cranial measure¬ ments and four qualitative cranial characters. Components I and II are indicated in the figure and component III is represented by height. See Figs. 31 and 41 and text for key to samples. Variation in Color Only two samples of Liomys adspersus were available for color analysis. Color reflectance readings for the samples from the Canal Zone and the Azuero Penin¬ sula, respectively, were as follows: red reflectance, 9.8 and 10.3; green reflec¬ tance, 5.8 and 5.7; blue reflectance, 5.2 and 5.6. None of the means for the three reflectances were found to be significantly different using the Student’s t-test. These two samples are paler than the sample of Liomys salvini from the vicinity of San Jose, Costa Rica, and more nearly resemble the sample from the Guana- caste of Costa Rica (see Table 26). Taxonomic Conclusions From the results of the univariate and multivariate analyses, and the data presented in the section on specific relationships, it is clear that Liomys adsper¬ sus is a species closely related to Liomys salvini of more northern regions of Cen¬ tral America. Within adspersus there appears to be little geographic variation, and I therefore consider it to be a monotypic species. Liomys adspersus (Peters, 1874) 1874. Heteromys adspersus Peters, Monotsb. preuss. Akad. Wiss., Berlin, p. 357, May. 1911. Liomys adspersus, Goldman, N. Amer. Fauna, 34:51, 7 September. Holotype. — Young adult male, mounted skin with skull, unknown number in Berlin Museum, from Panama. The type locality was restricted to the City of Panama by Goldman ( 1 920: 1 1 8). GENOWAYS— SYSTEMATICS OF LIOMYS 253 Fig. 45. — Geographic distribution of Liomys adspersus in Panama. Measurements of holotype. — (From original description) total length, 240; length of tail, 95; length of hind foot, 30. Distribution. — Central Panama principally on the Pacific versant (Fig. 45). Comparisons. — See above. Remarks. — Although I have not seen the holotype of this species, which is from an unknown locality in Panama, Peters (1874) presented an excellent plate showing the cranium, toothrows, feet, ear, and head of this specimen. Clearly from these figures the specimen is a young adult of the genus Liomys. The geographic distribution of Liomys adspersus appears to lie mainly in the semiarid savannahs of the Pacific drainage of central and western Panama. How¬ ever, a few specimens from Summit, Empire, Buenos Aires, Juan Mina, Rio Chagres, and the Cacao Plantation near Gamboa are from the headwaters of Caribbean drainage in the vicinity of the Canal Zone. What effect construction of the canal has had upon the distribution of these mice can only be a matter of conjecture. The species, Liomys salvini, that is geographically nearest adspersus is also its nearest relative within the genus; the ranges of the two species presently are separated by a distance of approximately 300 kilometers. What little significant variation that is present among samples of this species is mainly for external measurements, which may be the result of variation in tech¬ niques of collectors rather than the mice. In cranial measurements, all taken by myself, significant differences among the means of samples were found only in zygomatic breadth, mastoid breadth, and length of nasals for males, and in zygo¬ matic breadth for females. Basically there is little variation among the mice of this species. 254 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY Mean length of ear for 10 males and 10 females from west-central Panama was, respectively, 15.1 (12.0-17.0) and 14.6 (12.0-16.0). The same individuals (females not pregnant) averaged, respectively, 77.7 (65.0-92.0) and 57.4 (49.0- 66.0) for weight. Specimens examined (194). — Canal Zone: Albrook Air Force Base, 1 (USNM); 2 mi. N Albrook Field, 1 (USNM); Albrook Res., 4 (USNM); Balboa, 3 (FMNH); Chiva Chiva, 1 (USNM); Cocoli, 1 (USNM); Corozal, 1 (USNM); Curundu, 1 (USNM); Empire, 2 (USNM); Farfan, 1 (USNM); Fort Clayton, 6 (USNM); Fort Kobbe, 14 (USNM); Juan Mina, 1 (UA); Madden Forest, 1 (USNM); Madden Road, 6 (USNM); Rio Chagres, 4 (AMNH); Summit, 4 (3 USNM, 1 MCZ). Panama: 2 km. NE Buenos Aires, Panama, 8 (KU); Cerro Azul, Panama, 3 (USNM); Chepo, Panama, 1 (USNM); 1 mi. N Guabala, Chiriqui, 1 (USNM); Guanico, Los Santos, 95 (USNM); Las Cumbres, 7 km. N, 2 km. W Pueblo Nuevo, Panama, 1 (KU); Wi mi. E Montijo Bay [probably about 1 mi. E Paracote], Veraguas, 1 (UMMZ); Paracote, Vera- guas, 1 (UMMZ); 2 mi. E Rio Hato, Code, 10 (USNM); Rio Santa Maria, Santa Fe, Vera¬ guas, 17 (USNM); 2 mi. S San Francisco, 200 ft., Veraguas, 1 (TCWC); 2 mi. NE Tole, 1000 ft., Chiriqui, 3 (USNM). Additional records. — Canal Zone: Cacao Plantation (Brennan and Yunker, 1966:243); Corte Culebra Road (Brennan and Yunker, 1966:254); Experimental Gardens, near Summit (Enders, 1935:451); Nuevo Emperador (Brennan and Yunker, 1966:248); Rodman Naval Ammunition Depot (Fleming, 1970:476; 1971:16). Marginal records. — Panama: Cerro Azul; Chepo. Canal Zone: Balboa; Fort Kobbe. Panama: Rio Hato; Guanico; Paracote; Guabala; 2 mi. NE Tole; Rio Santa Maria, Santa Fe. Canal Zone: Rio Chagres; Juan Mina. Panama: 2 km. NE Buenos Aires. GENOWAYS— SYSTEMATICS OF LIOMYS 255 Status of Liomys centralis The species Liomys centralis was described by Hibbard (1 941 a: 349-350, 1 941 6.277) from the Rexroad Fauna (locality no. 3) of the Upper Pliocene of southwestern Kansas. The material upon which this taxon is based consists of part of the left ramus bearing incisor, p4, ml, and the alveolus of m2 (KUMVP 4589). My study ot this specimen and specimens of the five Recent species of the genus Liomys has led to the conclusion that centralis is not a member of the genus Liomys and almost without a doubt it is not even in the lineage leading to the genus. I have based this conclusion upon the following differences between centralis and Recent members of the genus: 1) the metalophid of p4 of centralis is composed of three fairly distinct cusps, whereas Recent specimens show no more than two cusps in this lophid; 2) in centralis the middle cusp of the metalo¬ phid of p4 is connected across the rather shallow median valley to one of the accessory cusps in the protolophid (see Hibbard, 194 la: 350 for a figure of this tooth), whereas in Recent species the median valley between the protolophid and metalophid is deep and there is never a connection between them across the middle of the tooth; 3) as pointed out by Hibbard (1941a:350, 19416:277) the first wear pattern on the p4 of centralis will be H-shaped, whereas in all Recent species the two lophids first meet at their labial margins giving a C-shaped pattern of wear, but never one of H-shape; 4) the cusps on the molar of centralis must be higher and more distinct than in Recent species of Liomys because on centralis the cusps are still easily discernable with the permanent premolar in place, whereas in Recent specimens the cusps are only discernable in extremely young individuals with unworn molars (by the time the permanent premolar is in place the two lophids have been reduced to crescentic lakes of dentine separated by the enamel median valley); 5) on the molar of centralis there is no trace of an anterior cingulum, whereas this cingulum is visible on specimens of Liomys that are still unworn enough that the cusp pattern is evident; and 6) Hibbard (1941a:350, 19416:277) believed that the molar of centralis would have an H- shaped wear pattern for a short time (this is a point on which I am uncertain based upon my examination of the specimen), whereas in all Recent species of Liomys the two lophids of the molars meet first at their labial margins, resulting in a C-shaped wear pattern (never an H-shaped pattern). In considering trends in the evolution of the Heteromyinae (Wood, 1935), development of a connection between the metalophid and protolophid of the p4 across the middle of the tooth that would give an H-shaped wear pattern pre¬ cludes “Liomys” centralis from the ancestry of the Recent species of Liomys , even though there are some similarities in structure of the protolophid of p4. When more is known about the evolutionary history of the Heteromyidae, I would not be surprised to find that “ Liomys ” centralis is more closely related to the Perognathinae lineage than to the Heteromyinae. However, for the present study it suffices to say that the species described as Liomys centralis is not a Liomys. [Since this manuscript was submitted for publication, Hibbard (Bull. Amer. Mus. Nat. Hist., 148:88-90, 1972) has assigned Liomys centralis to the genus Prodipodomys. He considered it to be a distinct species, Prodipodomys centralis .] 256 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY SPECIFIC RELATIONSHIPS In this section, the relationships among the five species of Liomys recognized in the previous section are discussed based upon 12 morphological and biological characteristics. Also, the relationships of these species to members of the genus Heteromys are examined. In most cases, I have used specimens of Heteromys desmarestianus as representatives of the genus, mainly because these were more readily available and more numerous in collections than were other species of Heteromys. I have used specimens of Heteromys lepturus and Heteromys gaumeri when desmarestianus was not available. Anatomical characters studied include external and cranial measurements, structure of the unworn permanent premolars and molars, wear pattern of molars, bacula, glans penes, karyotypes, head and neck structure of spermatozoa, mor¬ phology of soles of hind feet, structure of pterygoid region, and characters of pelage. Biological characteristics that have been investigated include ecological preferences, reproductive patterns, and ectoparasite faunas. Each of these char¬ acters has been analyzed separately and the relationships as revealed by the char¬ acter presented. In the final section of this work, a picture of the overall evolu¬ tionary and zoogeographic relationships of the species of Liomys and Heteromys desmarestianus will be presented. External and Cranial Morphology The external and cranial morphology of the species of Liomys and of Heteromys have been compared using ratio diagrams and discriminant function analyses. Pair-wise comparisons of the species of Liomys have been made for those pairs that were of interest because their geographic distributions are sym- patric, or nearly so, and for those pairs that were judged to be closely related systematically. The five species of Liomys were compared with Heteromys desmarestianus and Heteromys gaumeri. Crania of the species of Liomys and of Heteromys desmarestianus are shown in Figs. 46 and 47. In the ratio diagrams that follow, a sample that has a smaller mean for a given measurement than the standard will have a negative value and one that has a larger mean than the standard will have a positive value. Theoretically, a species having values that are correlated with the standard, but is either larger or smaller than the standard, will form a line parallel to the standard on the appropriate side. Therefore, fluctuations in the line of the test sample can be interpreted as proportional as well as absolute differences in size from the standard. In the dis¬ criminant function analyses, those characters with the smallest within-groups variance and largest between-groups variance will have the largest discriminant multipliers, and those measurements with the largest within-groups variance and smallest between-groups variance will have the lowest discriminant multipliers. Negative and positive values for the multipliers indicate proportional differences between the measurements. Liomys p ictus and Liomys irroratus These two species occur sympatrically from north-central Jalisco southward to the Sierra Madre del Sur of Oaxaca. In this region, L. pictus occurs mainly in GENOWAYS— SYSTEM ATICS OF LIOMYS 257 Fig. 46.— Dorsal and ventral views of the crania of five species of Liomys and one of Heteromys. The species are as follows: A, Liomys irroratus irroratus (KU 68855, d, 3 mi. W Mitla, Oaxaca); B, Liomys pictus pictus (KU 112276, d, San Sebastian, Jalisco)- C Liomys spectabilis (KU 96051, d, 2 2/10 mi. NE Contla, Jalisco); D, Liomys salvini salvim (KU 65042, d, 2 3/10 mi. W, 1/4 mi. N Iztapa, Escuintla, Guatemala); E, Liomys adspersus (KU 121528, d, 2 km. NE Buenos Aires, Panama); F, Heteromys desmarestianus desmarestianus (KU 95037, d, 7 1/2 mi. NW Pueblo Nuevo, Chiapas). The scale at the lower right is 10 millimeters long. lowland areas of the Pacific coast and Pacific slopes of the coastal mountains, but also occurs inland, to varying degrees, depending upon the geographic area (see Fig. 28). Liomys irroratus, on the other hand, occurs on the Mexican Plateau and adjacent areas of interior Mexico. Although the two species are sympatric over a relatively large area, they are generally microallopatric (Smith, 1965:57). In overall size, specimens of L. irroratus generally are larger than individuals of pictus, although this was not always the case as will be seen in some of the following examples. I have selected four areas where irroratus and pictus were actually or potentially microsympatric to analyze the relationship of the two species based upon their external and cranial morphology and to test for the presence of hybrids. One sample was from southeastern Jalisco in the vicinities of Ciudad Guzman and Contla and consisted of specimens of Liomys pictus plantinarensis and Liomys irroratus jaliscensis. In this area, irroratus appears to be uniformly larger than pictus in all measurements except interparietal width and interparietal length (Table 31); thus, the interparietal bone is proportionally larger in pictus than in 258 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY Fig. 47. — Lateral view of the crania and mandibles of five species of Liomys and one of Heteromys. The species are as follows: A, Liomys irroratus irroratus (KU 68855); B, Liomys pictus pictus (KU 112276); C, Liomys spectabilis (KU 96051); D, Liomys salvini salvini (KU 65042); E, Liomys adspersus (KU 121528); F, Heteromys desmarestianus des- marestianus (KU 95037). The scale at the lower right is 10 millimeters long. irroratus. Mastoid breadth and depth of braincase were the measurements with the largest discriminant multipliers and among those with high discriminant scores. Other measurements with high discriminant multipliers and scores were length of hind foot, zygomatic breadth, and interorbital constriction. There was no overlap in the range of the discriminant scores of the two species from this area of Jalisco, but the ranges are separated only by a score of 0.32. A second set of samples of L. pictus and L. irroratus was drawn from the area between Chilpancingo and Acapulco, Guerrero. In these samples, irroratus averaged larger than pictus in external measurements, whereas pictus averaged larger in most cranial measurements (Table 31). The two cranial measurements in which irroratus averaged larger than pictus were length of rostrum and depth of braincase; both of these measurements have high positive discriminant multi¬ pliers and scores. Greatest length of skull, interorbital constriction, and length of nasals had high negative discriminant multipliers. The nasals of pictus were found to be actually and proportionally longer than in irroratus (actually 0.8 milli¬ meters longer and 39.9 per cent of greatest length of skull in pictus as compared with 37.9 per cent in irroratus ), whereas in the other measurement of this portion of the skull, length of rostrum, irroratus was found to average very slightly larger and therefore proportionally longer (44.1 per cent of greatest length of skull in pictus as compared with 44.7 in irroratus ). The range of discriminant scores of L. pictus from along the Pacific slope of the Sierra Madre del Sur of Guerrero was — 8.71 to —5.49 and for L. irroratus was — 3.51 to 0.82, giving a separation of 1 .98. None of the specimens in this sample indicate hybridization between the two species. GENOWAYS — SYSTEMATICS OF LIOMYS 259 13 12 9 12 6 12 4 11 3 3 4 14 13 6 8 3 3 3 16 12 11 3 1 3 3 3 13 3 10 Fig. 48.— Histogram of linear discriminant scores based on a discriminant function analysis comparing Liomys pictus plantinarensis from Michoacan and Guerrero (sample 16 in analysis of geographic variation) and Liomys irroratus torridus from central Guerrero (sample 17 in analysis of geographic variation). Discriminant scores are indicated along the bottom of the histogram and frequency of individuals is indicated on the left-hand side. Individuals arranged below are from reference samples of pictus, at left, and irroratus, at right. Individuals arranged above are from the test sample; numbers, indicating the place of origin of test individual, are identified as follows: (1) Huajuapan, Oaxaca (holotype of L. irroratus minor)-, (2) 15 km. NNE Iguala, Guerrero; (3) Texcalzintla, 6 km. NNW Telo- loapan, Guerrero; (4) 1 km. SSE Texcalzintla, Guerrero; (5) 4 3/10 km. N Teloloapan. Guerrero; (6) 2 km. ENE Los Sabinos, Guerrero; (7) Los Sabinos, Guerrero; (8) Iguala' Guerrero; (9) Teloloapan, Guerrero; (10) La Cofradia, Teloloapan, Guerrero; (11) 3200 m SSE Iguala, Guerrero; (12) 1 km. NW Chapa, Guerrero; (13) Ojo de Agua de Chapa, 5 km SE Teloloapan, Guerrero; (14) Chapa, Guerrero; (15) 2 1/2 mi. W Mexcala, Guerrero; (16) 1 1/2 mi. SE Zumpango, Guerrero. A discriminant function analysis (Table 31) was performed upon a reference sample of pictus from Michoacan and Guerrero (specimens in sample 16 for analysis of geographic variation) and a reference sample of irroratus from the vicinity of Chilpancingo, Guerrero (sample 17 of geographic variation analysis). Onto the discriminant multipliers generated by this analysis were projected the measurements of specimens of both pictus (sample 17) and irroratus (sample 16) from the vicinity of Iguala, Guerrero, in the Balsas Basin (Fig. 48). Specimens of L. irroratus in the reference sample averaged larger than those of pictus in all 260 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY Table 31. — Multipliers and scores resulting from discriminant function analyses comparing Liomys pictus and Liomys irroratus from four geographic areas. L. pictus L. irroratus Measurements Discriminant multiplier Mean Mean discriminant score Mean Mean discriminant score Total length Southeastern Jalisco -0.069 216.07 - 14.91 231.19 - 15.95 Length of tail 0.054 107.00 5.78 115.82 6.25 Length of hind foot 0.668 25.90 17.30 28.53 19.06 Greatest length of skull -0.243 30.29 - 7.36 31.86 - 7.74 Zygomatic breadth 0.736 14.17 10.43 15.45 11.37 Interorbital constriction 0.823 7.30 6.01 7.90 6.50 Mastoid breadth 1.306 13.57 17.72 14.44 18.86 Length of nasals -0.147 11.66 - 1.71 12.27 - 1.80 Length of rostrum -0.046 13.01 - 0.60 13.98 - 0.64 Length of maxillary toothrow -0.731 4.90 - 3.58 5.10 - 3.73 Depth of braincase 1.709 8.08 13.81 8.69 14.85 Interparietal width -0.393 8.40 - 3.30 8.59 - 3.38 Interparietal length -0.820 3.76 - 3.08 3.61 - 2.96 Total mean discriminant score Range of discriminant scores 36.51 (34.01-38.37) 4U.by (38.69-43.18) Total length Pacific slope of Sierra Madre del Sur of Guerrero 0.014 241.21 3.38 247.48 3.46 Length of tail 0.153 1 19.00 18.21 127.66 19.53 Length of hind foot 0.1 14 27.86 3.18 28.66 3.27 Greatest length of skull - 1.106 32.82 -36.30 32.44 -35.88 Interorbital constriction -2.099 7.91 - 16.60 7.72 - 16.20 Mastoid breadth -0.902 14.60 - 13.17 14.55 - 13.12 Length of nasals -2.714 13.11 -35.58 12.31 -33.41 Length of rostrum 3.564 14.46 51.54 14.50 51.68 Length of maxillary toothrow 0.705 5.08 3.58 5.09 3.59 Depth of braincase 1.669 8.60 14.35 8.75 14.60 Interparietal width 0.020 8.61 0.17 8.29 0.16 Interparietal length 0.096 4.13 0.40 3.86 0.37 Total mean discriminant score Range of discriminant scores -6.84 (- 8.71 to - 5.49) (- — 1.95 3.51 to 0.82) GENOWAYS— SYSTEMATICS OF LIOMYS 261 Table 31. — Continued. Interior Michoacan and Guerrero Total length -0.036 Length of tail 0.084 Length of hind foot 0.469 Greatest length of skull 0.388 Interorbital constriction 0.971 Mastoid breadth -0.087 Length of nasals -0.620 Length of rostrum -0.138 Length of maxillary toothrow 0.233 Depth of braincase 2.827 Interparietal width -0.734 Interparietal length 0.766 Total mean discriminant score Range of discriminant : scores Total length Sierra Madre 0.016 Length of tail 0.051 Length of hind foot 0.203 Greatest length of skull -0.187 Zygomatic breadth -0.200 Interorbital constriction 1.490 Mastoid breadth 1.445 Length of nasals - 1.403 Length of rostrum -0.770 Length of maxillary toothrow -0.969 Depth of braincase 2.866 Interparietal width - 1.328 Interparietal length - 1.751 Total mean discriminant score Range of discriminant scores 216.36 - 7.79 246.90 - 8.89 109.86 9.23 127.63 10.72 25.21 1 1.82 28.53 13.38 29.80 1 1.56 32.40 12.57 6.96 6.76 7.69 7.47 13.46 - 1.17 14.50 - 1.26 11.39 - 7.06 12.38 - 7.68 12.89 - 1.78 13.98 - 1.93 4.83 1.13 5.09 1.19 7.99 22.59 8.74 24.71 8.27 - 6.07 8.30 - 6.09 3.58 2.74 3.88 2.97 41.96 47.16 (40.24-43.26) (44.63-49.14) del Sur of Oaxaca 253.50 4.06 266.89 4.27 130.77 6.67 143.92 7.34 29.98 6.09 32.65 6.63 33.85 - 6.33 33.64 6.29 15.61 - 3.12 15.92 3.18 7.89 1 1.76 8.14 12.13 14.37 20.76 15.00 21.68 13.91 - 19.52 13.00 18.24 15.22 - 1 1.72 15.14 1 1.66 5.16 - 5.00 5.38 5.21 8.83 25.31 9.12 26.14 9.12 - 12.1 1 8.50 1 1.29 4.58 - 8.02 3.99 6.99 8.83 15.33 (6.86-10.85) (12.90- 17.34) measurements, and only in interparietal width did the mean for pictus approach that for irroratus. Depth of braincase had by far the largest discriminant multi¬ plier; other measurements with high positive discriminant multipliers were inter¬ orbital constriction, interparietal length, and length of hind foot. Length of nasals and interparietal width had relatively high negative discriminant multipliers. The range of discriminant scores for the L. pictus reference sample was 40.24 to 43.26 and for L. irroratus was 44.63 to 49.14, giving a separation of 1.37 be¬ tween the two reference samples. Specimens tentatively identified as L. pictus from the Balsas Basin had a range of discriminant scores from 39.30 to 42.13 and those of L. irroratus ranged from 44.41 to 50.29. The specimen with the 262 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY value of 44.41 has extremely small external measurements, which probably are erroneous. The next lowest score for irroratus was 44.72. I find no indication of hybridization between L. pictus and L. irroratus in the region of the Balsas Basin and the two have been taken together at four (6, 8, 1 1, and 16 on Fig. 48) localities there. A specimen of particular interest is the holotype of L. i. minor from Huajuapan, Oaxaca. Hooper and Handley (1948:13-14) suggested that this specimen resembled Liomys pictus in some characters, but my analysis clearly shows it to be a representative of irroratus. The fourth area from which samples of L. pictus and L. irroratus were studied using discriminant functions was the mountains of the Sierra Madre del Sur of Oaxaca (samples 22, 23 for pictus and 16 for irroratus in the analysis of geo¬ graphic variation). In this area, specimens of irroratus averaged larger than those of pictus in external measurements, and in all cranial breadth measurements except interparietal width and depth of braincase. Specimens of L. pictus averaged longer in all measurements of cranial length except for that of maxillary toothrow. Thus, the skulls of pictus in this area are longer and narrower than those of ir¬ roratus , and irroratus has a proportionally longer toothrow and a proportionally narrower interparietal bone. Measurements with high positive discriminant multi¬ pliers are interorbital constriction, mastoid breadth, and depth of braincase, and those with high negative multipliers are length of nasals, length of maxillary toothrow, interparietal width, and interparietal length. The range of discriminant scores for the pictus reference sample was 6.86 to 10.85 and for the irroratus reference sample was 12.90 to 17.34, giving a separation of 2.05. In all four geographic areas analyzed the two species, pictus and irroratus, were separated on the basis of external and cranial measurements using a discrimi¬ nant function analysis. No single or subset of measurements would consistently separate the two, although in all four tests depth of braincase and interorbital constriction had high discriminant multipliers. These data indicate that in order to differentiate these two species in western Mexico on the basis of external and cranial measurements, it is necessary to study specimens in detail in each local area where they occur sympatrically. Liomys irroratus and Liomys spectabilis These two species have never been taken in the same traplines, but their geo¬ graphic ranges do closely approach each other. All of the available adults with complete measurements of Liomys spectabilis and a sample of Liomys irroratus from the vicinity of the range of spectabilis in southeastern Jalisco were compared using discriminant functions (Table 32). Measurements of total length and length of tail were not used because many specimens of spectabilis lacked complete tails. L spectabilis averaged larger than L. irroratus in all measurements except for interparietal width (in which the averages were the same), and depth of braincase (in which irroratus averaged 0.25 millimeters larger). Length of rostrum and interparietal length had high positive discriminant multipliers and depth of brain¬ case, interorbital constriction, and zygomatic breadth had high negative multi- GENOWAYS— SYSTEMATICS OF LIOMYS 263 Table 32. Multipliers and scores resulting from a discriminant function analysis between Liomys spectabilis and Liomys irroratus from southeastern Jalisco. L. spectabilis L. irroratus Measurements Discriminant multiplier Mean Mean discriminant score Mean Mean discriminant score Length of hind foot Greatest length -0.007 30.46 - 0.21 28.53 - 0.20 of skull -0.129 34.16 - 4.41 31.86 - 4.1 1 Zygomatic breadth Interorbital -0.882 15.65 - 13.80 15.45 - 13.63 constriction -0.801 7.98 - 6.39 7.90 - 6.33 Mastoid breadth 0.260 14.58 3.79 14.44 3.75 Length of nasals 0.253 13.63 3.45 12.27 3.10 Length of rostrum Length of maxillary 2.074 15.52 32.19 13.98 28.99 toothrow -0.172 5.25 - 0.90 5.10 - 0.88 Depth of braincase -2.755 8.44 -23.42 8.69 -24.1 1 Interparietal width -0.572 8.59 - 4.91 8.59 - 4.91 Interparietal length Total mean discriminant 2.018 score 4.39 8.86 -5.75 3.61 7.28 - 1 1 .05 Range of discriminant : scores (-6.74 to -4.35) (- 13.75 to -8.81) pliers. The range of discriminant scores for the two species was separated by a total score of 4.35. Liomys irroratus and Liomys salvini These two species are allopatric, their ranges being nearest in Oaxaca, where irroratus has been recorded from Zapotitlan and salvini from approximately 140 kilometers to the east at Reforma. When a reference sample of L. irroratus from the valley of Oaxaca was compared with a reference sample of L. salvini from southeastern Oaxaca and northwestern Chiapas, specimens of irroratus were found to average larger in all measurements except interparietal width. In the discriminant function analysis, interorbital constriction was weighted (high, pos¬ itive value) much more than any other measurement (Table 33). Interparietal width and zygomatic breadth had the largest negative discriminant multipliers. The skull of irroratus was generally larger than that of salvini , with the inter¬ orbital constriction of salvini being proportionately narrower (20.5 per cent of greatest length of skull in salvini as compared with 24.2 per cent in irroratus) and the interparietal proportionately wider. The range of discriminant scores for L. irroratus and L. salvini was separated by more than four points. Liomys pictus and Liomys salvini These two species are sympatric in southeastern Oaxaca and northwestern Chiapas — from the vicinity of Reforma, Oaxaca, to the vicinity of Tonala, Chia- 264 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY Table 33. — Multipliers and scores resulting from a discriminant function analysis com¬ paring Liomys salvini from southeastern Oaxaca and northwestern Chiapas and Liomys irroratus from central Oaxaca. L. salvini L. irroratus Measurements Discriminant multiplier Mean Mean discriminant score Mean Mean discriminant score Total length -0.027 207.10 - 5.59 266.89 - 7.21 Length of tail 0.086 100.32 8.63 143.92 12.38 Length of hind foot Greatest length 0.232 27.03 6.27 32.65 7.57 of skull -0.300 31.37 - 9.41 33.64 - 10.09 Zygomatic breadth Interorbital -0.533 14.67 - 7.82 15.92 - 8.49 constriction 3.143 6.44 20.24 8.14 25.58 Mastoid breadth 0.743 13.80 10.25 15.00 11.15 Length of nasals -0.198 11.73 - 2.32 13.00 -2.57 Length of rostrum Length of maxillary 0.326 13.32 4.34 15.14 4.94 toothrow -0.434 4.76 - 2.07 5.38 - 2.33 Depth of braincase 0.452 8.58 3.88 9.12 4.12 Interparietal width -0.938 8.70 - 8.16 8.50 - 7.97 Interparietal length —0.721 Total mean discriminant score Range of discriminant scores 3.92 - 2.83 24.20 (12.83-17.00) 3.99 - 2.88 15.41 (21.05-26.02) pas. The two species apparently are not ecologically segregated in this area (based upon available field notes). As pointed out earlier, males of pictus from the area of sympatry are larger than those from contiguous samples and males of salvini are smaller than those in contiguous samples. In order to examine this relation¬ ship, a ratio diagram was prepared (Fig. 49) in which the upper diagram is of allopatric populations of the two species that are contiguous with the sympatric populations (sample 24 of pictus and sample 2 of salvini from the analysis of geographic variation). Specimens of L. pictus appear to be proportionally longer in total length and length of tail and have a broader interorbital region. Generally, specimens of pictus were larger than those of salvini , although in greatest length of skull, zygomatic breadth, mastoid breadth, and interparietal width the two averaged the same or nearly so. The lower ratio diagram (Fig. 49) is of the sym¬ patric populations of males from southeastern Oaxaca and northwestern Chiapas (sample 25 of pictus and sample 1 of salvini from the analysis of geographic variation). The mean of the measurements have maintained essentially the same proportional relationship to each other, but the values for pictus have been shifted to the right, indicating a greater difference between the species in the area of GENOWAYS— SYSTEMATICS OF LIOMYS 265 L. SALV IN I L. PI C T US SAMPLE 2 SAMPLE 24 TL — I - 1 - 1 - 1 - 1 - 1 - 1 - 1 - 1 - 1 - 1 - h—H - 1 - 1 - 1 - h -.01 0 .01 .02 03 .0 4 .05 .06 .0 7 .08 .09 .10 .11 .12 .13 .14 )5 Fig. 49. — Ratio diagrams comparing 13 external and cranial measurements of allopatric samples of Liomys salvini and Liomys pictus (upper) and sympatric samples of Liomys salvini and Liomys pictus (lower). 266 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY Table 34. — Multipliers and scores resulting from a discriminant function analysis comparing Liomys pictus and Liomys salvini from southeastern Oaxaca and northwestern Chiapas. L. pictus L. salvini Measurements Discriminant multiplier Mean Mean discriminant score Mean Mean discriminant score Total length -0.001 248.86 - 0.25 206.44 - 0.21 Length of tail 0.084 132.59 11.14 100.39 8.43 Length of hind foot Greatest length 0.096 29.46 2.83 27.22 2.61 of skull - 1.640 32.41 -53.15 31.28 -51.30 Zygomatic breadth Interorbital 0.954 14.87 14.19 14.62 13.95 constriction 1.607 7.82 12.57 6.47 10.40 Mastoid breadth -0.083 14.03 - 1.16 13.77 - 1.14 Length of nasals 1.807 12.85 23.22 1 1.59 20.94 Length of rostrum Length of maxillary 0.323 14.50 4.68 12.26 3.96 toothrow 0.565 4.89 2.76 4.74 2.68 Depth of braincase -2.563 8.73 -22.37 8.61 -22.07 Interparietal width -0.988 8.72 - 8.62 8.77 - 8.66 Interparietal length 2.087 Total mean discriminant score 4.50 9.39 -4.77 3.93 8.20 - 12.21 Range of discriminant scores (-7.17 to -2.74) (- 13.69 to - 10.73) sympatry. These results have been interpreted as indicating that males of both salvini and pictus have undergone character displacement (Brown and Wilson, 1956) in the area of sympatry between the two species. The most likely explanation for this character displacement is competition between these solitary, rather aggressive mice; it may somehow relate to the breeding biology of these animals, because character displacement is observed only in males and not in females. Whatever the answer is, it will not be known until much more information is available on the biology of these species and until they have been studied more intensively in the area of sympatry. A discriminant function analysis was performed also on the specimens from the area of sympatry (Table 34). Large positive discriminant multipliers were generated for interorbital constriction, length of nasals, and interparietal length, whereas large negative multipliers were generated for greatest length of skull, depth of braincase, and interparietal length. The range of discriminant scores for Liomys pictus was —7.17 to —2.1 A and for Liomys salvini was —13.69 to — 10.73, giving a separation of 3.56. Liomys pictus and Liomys spectabilis The restricted geographic range of Liomys spectabilis (southeastern Jalisco) is completely sympatric with the range of Liomys pictus plantinarensis, although GENOWAYS— SYSTEM ATICS OF LIOMYS 267 able 35. Multipliers and scores resulting from discriminant function analyses comparing Liomys spectabilis and Liomys pictus plantinarensis from southeastern Jalisco, and Liomys pictus pictus from southwestern Jalisco. L. spectabilis L. pictus Mean Mean Discriminant discriminant discriminant Measurements multiplier Mean score Mean score Southeastern Jal isco Length of hind foot Greatest length 0.968 30.46 29.49 25.83 25.00 of skull 0.758 34.16 25.89 30.09 22.81 Zygomatic breadth Interorbital 0.479 15.65 7.50 13.97 6.69 constriction 1.305 7.98 10.41 7.18 9.37 Mastoid breadth 1.034 14.58 15.08 13.45 13.91 Length of nasals - 1.291 13.63 - 17.60 1 1.53 - 14.89 Length of rostrum Length of maxillary 0.133 15.52 2.06 12.83 1.71 toothrow 0.049 5.25 0.26 4.70 0.23 Depth of braincase - 1.785 8.44 - 15.07 8.01 - 14.30 Interparietal width -0.127 8.59 - 1.09 8.29 - 1.05 Interparietal length 0.284 4.39 1.25 3.63 1.03 Total mean discriminant score 50.51 58.18 Range of discriminant scores (56.30-59.82) (48.49-52.27) Southwestern Jal isco Length of hind foot Greatest length 0.205 30.46 6.24 28.65 5.87 of skull 0.732 34.16 25.01 31.86 23.32 Zygomatic breadth Interorbital -0.677 15.65 - 10.60 14.89 - 10.08 constriction -2.640 7.98 -21.07 7.67 -20.25 Mastoid breadth 1.134 14.58 16.53 14.21 16.1 1 Length of nasals -2.907 13.63 -39.62 12.86 -37.38 Length of rostrum Length of maxillary 3.845 15.52 59.67 13.94 53.60 toothrow 1.069 5.25 5.61 4.92 5.26 Depth of braincase - 1.837 8.44 - 15.50 8.33 -15.30 Interparietal width 0.734 8.59 6.31 8.79 6.45 Interparietal length -0.444 4.39 - 1.95 4.47 - 1.98 Total mean discriminant score 30.63 25.62 Range of discriminant scores (29.76-32.40) (23.76-27.42) the range of spectabilis is near the eastern limit of the geographic range of pictus at this point. Specimens of spectabilis averaged larger than specimens of L. p. plantinarensis in all measurements (Table 35). Length of hind foot, greatest length of skull, interorbital constriction, and mastoid breadth all had large posi¬ tive discriminant multipliers, whereas length of nasals and depth of braincase 268 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY Table 36. — Multipliers and scores resulting from a discriminant function analysis comparing Liomys salvini /rom Nicaragua and Costa Rica and Liomys adspersus /rom Panama. L. salvini L. adspersus Measurements Discriminant multiplier Mean Mean discriminant score Mean Mean discriminant score Total length -0.017 232.90 - 3.96 257.73 - 4.38 Length of tail 0.035 117.10 4.10 130.60 4.57 Length of hind foot Greatest length 0.161 27.83 4.48 30.80 4.96 of skull -0.595 32.45 - 19.31 35.47 -21.10 Zygomatic breadth Interorbital 0.001 14.95 0.01 16.59 0.02 constriction 1.218 6.83 8.32 7.55 9.20 Mastoid breadth 0.161 13.84 2.23 14.68 2.36 Length of nasals 0.204 12.55 2.56 14.50 2.96 Length of rostrum Length of maxillary 1.425 13.91 19.82 15.72 22.40 toothrow 1.411 4.93 6.96 5.35 8.40 Depth of braincase 1.618 8.45 13.67 9.47 15.32 Interparietal width -0.158 8.42 - 1.33 7.88 - 1.25 Interparietal length 0.362 Total mean discriminant score 4.14 1.50 39.05 4.56 1.65 45.11 Range of discriminant scores (37.45-40.96) (42.95-47.24) had high negative discriminant multipliers. The ranges of discriminant scores for the two taxa were separated by more than four points. Specimens of L. spectabilis also were compared with a sample of L. p. pictus from southwestern Jalisco (Table 35). The average values for the spectabilis sam¬ ple were larger for all measurements except interparietal width and interparietal length; also, there was little difference in the average depth of braincase for the two taxa. Greatest length of skull, mastoid breadth, length of maxillary toothrow, and interparietal width had relatively large positive discriminant multipliers, whereas interorbital constriction, length of nasals, and depth of braincase had large negative multipliers. Range of discriminant scores for L. spectabilis was 29.76 to 32.40 and for L. p. pictus was 23.76 to 27.42. Liomys salvini and Liomys adspersus These two species are not sympatric, Liomys adspersus being isolated in cen¬ tral Panama and forming the southern limit of distribution of the genus. Liomys salvini is known no farther south than central Costa Rica; therefore, the geo¬ graphic ranges of the two species as presently understood are separated by a hiatus of approximately 300 kilometers. Specimens of adspersus averaged larger than those of salvini in all measurements except interparietal width (Table 36). Four measurements — interorbital constriction, length of rostrum, length of GENOWAYS— SYSTEMATICS OF LIOMYS 269 Fig. 50. — Ratio diagram comparing 13 external and cranial measurements of five species of Liomys with Heteromys desmarestianus. Symbols representing each species of Liomys are given at lower left. maxillary toothrow, depth of braincase — had large positive discriminant multi¬ pliers, but only one measurement, greatest length of skull, had a relatively large negative multiplier. The range of discriminant scores for salvini was 37.45 to 40.96 and for adspersus was 42.95 to 47.24. 270 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY Fig. 5i. — Ratio diagram comparing 13 external and cranial measurements of five species of Liomys with Heteromys gaumeri. Symbols for species of Liomys as in Fig. 50. Heteromys and Liomys Samples of the five species of Liomys were compared with Heteromys des¬ marestianus and Heteromys gaumeri using ratio diagrams. In nearly all measurements, Heteromys desmarestianus was larger than specimens of any species of Liomys (Fig. 50). However, interparietal width of four species of Liomys (only adspersus was less) was greater than in Heteromys desmarestianus. Also, Liomys irroratus and L. adspersus had longer maxillary toothrows and L. adspersus had a deeper braincase. Other measurements in which most of the GENOWAYS— SYSTEMATICS OF LIOMYS 271 species of Liomys converge toward Heteromys were greatest length of skull, zygomatic breadth, and mastoid breadth. Measurements of Liomys that appear to be most divergent from those of Heteromys desmarestianus were interorbital con¬ striction, length of nasals, interparietal length, and length of tail. Specimens of Heteromys gaumeri are generally smaller than those of H. desmarestianus and consequently exhibit less size difference when compared with the species of Liomys ; however, the pattern of variation is similar (Fig. 51). Specimens of Liomys adspersus were larger than those of H . gaumeri in greatest length of skull, zygomatic breadth, length of nasals, length of rostrum, length of maxillary toothrow, and depth of braincase. Specimens of Liomys irroratus were larger than H. gaumeri in three measurements (length of rostrum, length of maxil¬ lary toothrow, depth of braincase); L. spectabilis was larger in two measurements (length of rostrum, length of maxillary toothrow); L. salvini was larger in one measurement (interparietal width). Besides length of rostrum and length of maxillary toothrow, other measurements of Liomys that converge toward those of H. gaumeri are greatest length of skull, zygomatic breadth, mastoid breadth, and interparietal width. Those measurements of Liomys that appear to diverge most strongly from Heteromys gaumeri were interorbital constriction and inter¬ parietal length. Specimens of Heteromys (all KU) used in these comparisons are listed below. Heteromys desmarestianus. — 4 mi. S Altamirano, Chiapas, 1; Finca El Para'iso, 4050 ft., Chiapas, 1; 8 mi. NW (by road) Pueblo Nuevo, 5900 ft., Chiapas, 3; Sabana de San Quin- tin, 215 m., Chiapas, 1; Rio Queleya, Vi mi. E Yepocapa, 4300 ft., Chimaltenango, Guate¬ mala, 1; 1 mi. SW Santiago Sacatepequez, Sacatepequez, Guatemala, 2; Bonanza, Zelaya, Nicaragua, 1; 2 km. N, 6 km. E Esquipulas, 960 m„ Matagalpa, Nicaragua, 2; Hda. Tepeyac,' Matagalpa, Nicaragua, 3; 5 mi. S, 2 mi. E Jinotega, Jinotega, Nicaragua, 1; Santa Maria de Ostuma, 1250 m., Matagalpa, Nicaragua, 2; 5 km. SE Turrialba, Cartago, Costa Rica, 2. Heteromys gaumeri.— 3 km. N Piste, Yucatan, 3; 6 km. N Tizimm, Yucatan, 1; 1.5 km. S, 1 km. E Pueblo Nuevo X-can, Quintana Roo, 1; 1 km. SSW Santa Rosa, Quintana Roo, 1; 7 km. N, 51 km. E Escarcega, Campeche, 3; IV2 km. W Escarcega, 65 m„ Campeche, 1; 103 km. SE Escarcega, Campeche, 1. Tooth Structure The unworn upper and lower premolars of Liomys irroratus, L. pictus , and L. salvini, and Heteromys desmarestianus, H. gaumeri, and H. lepturus are here described and compared. The premolars of L. adspersus and L. spectabilis are not figured and are mentioned only briefly in accounts of other species because adequate material at the proper stage of eruption and wear was not available for study. The structure and wear was not available for study. The structure and wear sequences of the molars of all five Recent species of Liomys and the three species of Heteromys listed above also are described and compared. All teeth figured and discussed were in the right toothrow. Dental nomen¬ clature has been modified from Wood (1935:79), Wood and Wilson ( 1 936:388- 391), and Shotwell (1967:1 1). Wood (1935) gave an excellent discussion of the teeth of these two genera. 272 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY A B c D Fig. 52. — Crown patterns of the upper (top) and lower (bottom) premolars of three species of Liomys and of Heteromys (upper and lower from different species). All teeth illustrated are from the right side. For all teeth, anterior is to the top; for the upper premolar, lingual is to the right and for the lower premolar it is to the left. Abbreviations for the upper pre¬ molar are: EN, entostyle; HY, hypocone; ME, metacone; PA, paracone; PC, posterior cingulum; PR, protocone; PS, protostyle. Abbreviations for the lower premolar are: an, anteroconid; hy, hypoconid; me, metaconid; MS, mesoconid; pr, protoconid. The species are as follows: A, Liomys irroratus ( upper, KU 31174; lower, KU 31129); B, Liomys pictus (KU 107667; KU 99792); C, Liomys salvini (KU 71118; KU 83963); D, Heteromys gaumeri (upper, KU 92146) and Heteromys desmarestianus { lower, KU 71255). Upper Premolar Liomys irroratus (Fig. 52 A). — The protoloph of the upper premolar consists of three cusps arranged more or less in a straight line. The cusps are more dis¬ tinct in some individuals than others, but are discernable in all specimens examined. The middle cusp, protocone, is the largest of the three and is located directly anterior to the hypocone of the metaloph. The labial cusp, paracone, is smaller than the protocone, but larger than the protostyle. As pointed out by Wood (1935:199), the cusp called the “paracone” may not represent that cone, but rather a labial style. The metaloph also consists of three cusps, but these are arranged in a crescent with the large middle cusp being the posteriormost. This middle cusp, which represents the hypocone, although largest of the three cusps of the metaloph, is not much larger than the labial metacone. The entostyle, which is noticeably smaller than the metacone and hypocone, is located anterior and slightly lingual to the hypocone. The entostyle is clearly distinct from the hypocone, but is not as widely sepa¬ rated from it as in Liomys salvini and Heteromys. The median valley separating the protoloph and metaloph is reminiscent of the Y-shape found in L. salvini and Heteromys , but the re-entrant angle between the hypocone and entostyle is not continuous with the lingual margin of the tooth. Thus the median valley has a shape of a Y with one arm shorter than the other. As the premolar wears the hypocone and entostyle are quickly united. A well-developed posterior cingulum GENOWAYS — SYSTEM ATICS OF LIOMYS 273 extends from about the middle of the metacone to the lingual edge of the hypo- cone. Between this cingulum and the hypocone is a deep pit of enamel, which persists for some time as an island of enamel surrounded by dentine as the tooth is worn. Liomys pictus (Fig. 52B). — The protoloph of this species appears to consist ot a single cusp, which is probably the protocone. In some specimens, there is a slight development of lateral accessory cusps, but these are extremely weak and only occasionally present. Wood (1935:198) stated in his description of the teeth of Liomys that the anterior loph ot the upper premolar always was composed of three cusps with the central cusp being the largest and “the two lateral ones com¬ pressed almost beyond recognition.” Certainly these lateral cusps are much more difficult to discern on the premolar of specimens of pictus than in irroratus. The metaloph is crescent-shaped and consists of three cusps as in irroratus. However, the cusps are all connected by a loph so that they never form discrete cones as in irroratus. The hypocone is the largest of the three, but the metacone is almost as large. The smaller entostyle is placed anterior to the hypocone and almost completely lingual to the hypocone. The position of the entostyle is more lingual in pictus than in irroratus and it is never separated from the other cusps, always being connected by a loph to the hypocone. The median valley separating the protoloph and metaloph has a shape much as in irroratus. The re-entrant angle between the entostyle and hypocone does not reach the lingual edge of the tooth so that the median valley is Y-shaped, with one of the arms being shorter than the other. A well-developed posterior cingulum extends from the middle of the metacone to near the level of the lingual edge of the hypocone. A ridge extends posteriorly from the hypocone and connects with the cingulum, and the lingual end of the cingulum may be connected with the ridge extending from the entostyle to the hy¬ pocone or it may be free. However, even in the specimen that is figured, slight wear will connect the cingulum with the ridge extending posteriorly from the entostyle. There is a deep valley of enamel between the posterior cingulum and the hypocone, which is divided in half by the ridge extending posteriorly from the hypocone. These two pits of enamel remain as islands surrounded by dentine as the tooth begins to wear. The valley between the hypocone-metacone-cingulum does not persist as long as the one between hypocone-entostyle-cingulum. On the two specimens of Liomys spectabilis that show the least amount of wear on P4, the pattern of cusps appears to be much as in L. pictus, although I am un¬ certain of exact details. Liomys salvini (Fig. 52C). — The protoloph of the upper premolar of this species appears to be composed of a single cusp, the protocone, because the two lateral cusps are so compressed against it that they are indistinguishable (Wood, 1935:198). The crescent-shaped metaloph is always composed of at least three cusps and many times four. The metacone is as large, and in some cases larger, than the hypocone. The extra cusp seen on the premolars of many specimens appears to be the result of a deep re-entrant angle of enamel that divides a loph, extending from the hypocone toward the entostyle. This cusp quickly becomes 274 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY joined with the hypocone as wear to the tooth begins, although the re-entrant angle of enamel is evident for some time (first as an angle of enamel extending from the lingual edge of the tooth into the lake of dentine formed by wear and then as an island of enamel surrounded by dentine). The entostyle is placed anterior and lingual to the hypone and is separated from the other cusps ot the metaloph by a re-entrant angle of the median valley. The median valley has a Y-shape much as in Heteromys, but as the tooth begins to wear the entostyle does become connected with the hypocone, thus giving the median valley a shape as in the other species of Liomys. A well-developed posterior cingulum is present on P4 of L. salvini. The cingulum extends from the metacone to the lingual side of the hypocone and is separated from the hypocone by a re-entrant angle of enamel, which eventually forms an island before disappearing. Although I have not examined extensive material of Liomys adspersus, the pattern of the premolars appears to be essentially as it is in L. salvini. Certainly the isolated entostyle, which is characteristic of salvini and not other species of Liomys, is present in adspersus as shown by Fleming (1969:82-83, 1971:27). Heteromys (Fig. 52D). — The description of the premolar for this genus is based mainly on Heteromys gaumeri because only specimens of that species were available with unworn premolars. Specimens of Heteromys desmarestianus and Heteromys lepturus were also examined, but most possessed teeth that were worn to a degree so as to be of only limited use. The protoloph is composed of three cusps. The centrally located protocone is the largest, but the labially situated “paracone” also is relatively large. The protostyle is small and not evident in some specimens. The metacone and hypo¬ cone are located side-by-side on the metaloph, with the entostyle located almost directly anterior to the hypocone. The metacone is the largest cusp and the ento¬ style is relatively larger than in any of the species of Liomys. The median valley that divides the protoloph from the metaloph extends a branch between the hypo¬ cone and entostyle, giving the valley a Y-shape. The hypocone and entostyle do become connected at their labial margins, but a portion of the re-entrant angle of enamel persists between the two cusps, in some specimens until they are old adults. A well-developed posterior cingulum extends from approximately the middle of the metacone to the middle of the hypocone. It is separated from the hypocone by a re-entrant angle of enamel from the lingual side of the tooth. As the tooth begins to wear, this angle becomes an island of enamel surrounded by dentine. This lake may persist for some time, but certainly not so long as the angle between the entostyle and the hypocone. Lower Premolar Liomys irroratus (Fig. 5 2 A). — The protoloph id of this species is generally composed of three cusps. The two lateral cusps, protoconid (lingual) and mesoconid (labial), are large, but there is variability among individuals as to which is the largest. The anteriorly-placed anteroconid is smaller than the other two cusps and in many instances (as in the specimen figured) this cusp appears to GENOWAYS— SYSTEMATICS OF LIOMYS 275 be divided into at least two cusps (see also Wood, 1935:198-199). However, as wear progresses this division is quickly obliterated. The median valley sepa¬ rating the protolophid and metalophid is connected with a deep re-entrant angle that separates the protoconid and mesoconid at their posteriomedial borders. The enamel valley separating the mesoconid and protoconid extends anteriorly be¬ tween the mesoconid and anteroconid and deeply separates these two cusps. A much shallower branch of this re-entrant angle separates the anteroconid and protoconid. These latter two cusps become joined relatively early in the wear of the tooth, whereas the mesoconid remains separated for a much longer time. The metalophid is composed of two large cusps, which are situated side-by-side but separated by a shallow angle of enamel. The labial hypoconid is generally larger than the lingual metaconid. These two cusps quickly become united into a single loph as wear progresses. No anterior cingulum was observed in the specimens of irroratus. The only possible remnant of a posterior cingulum is a re-entrant angle of enamel that ex¬ tends one-third to one-half the way across the hypoconid from its labial side. This angle may persist for a short time as an island of enamel surrounded by dentine as wear progresses. Liomys pictus (Fig. 52B). — The protolophid of pictus is composed of three cusps as in L. irroratus. The protoconid is relatively larger than in irroratus-, it appears to be expanded labially and has nearly filled the space occupied by the re-entrant angle separating the protoconid and mesoconid. In some specimens (as the one figured), the re-entrant angle of enamel extending between the pro¬ toconid and mesoconid has been completely blocked by the protoconid, but in other specimens a small valley of enamel temporarily separates the two cusps. The mesoconid is smaller than the protoconid. The anteroconid appears to be composed of two cusps as in irroratus-, however, as in irroratus, these quickly become united. The angle of enamel separating the protoconid from the antero¬ conid is much deeper and persistent than is the angle between the mesoconid and anteroconid; therefore, contrary to the condition in irroratus, the mesoconid and anteroconid first become united as wear progresses and then the anteroconid and protoconid. The mesoconid and protoconid are united early in the wear of the tooth at their posteromedial margins, with a deep pit of enamel, the remnant of the re-entrant angle seen in irroratus, forming an island in the middle of the pro¬ tolophid. I did not observe an anterior cingulum on any specimen of this species. The metalophid, which is separated from the protolophid by a deep median valley of enamel, is made up of two cusps, hypoconid and metaconid, that quickly become united into a single loph. The lingual metaconid is smaller than the labial hypoconid. A labial re-entrant angle of enamel extends about one-half the distance across the hypoconid. The area posterior to this angle may represent the posterior cingulum. Unfortunately this re-entrant angle is not well developed in the specimen figured, but on other specimens it is as well developed as in the specimen of irroratus figured. Liomys spectabilis appears to have essentially the same cusp pattern as found in pictus. 276 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY Liomys salvini (Fig. 52C). — The protolophid is composed of three cusps as in the two previous species. However, the cusps are rather weakly defined in salvini because the enamel angles separating them are shallow. The mesoconid and pro- toconid are about equal in size and only slightly separated, if at all, at their posteromedial borders. The anteroconid appears to be composed of two cusps, which are weakly separated from the protoconid and mesoconid. There is a deep pit of enamel near the center of this loph where all of the re-entrant angles coalesce between the cusps. All cusps appear to become united almost simultan¬ eously early in the wear of the tooth leaving an island of enamel near the center of the loph. No anterior cingulum was observed on specimens examined. The configuration of the metalophid is somewhat different than in irroratus and pictus. The re-entrant angle of enamel seen in those species extends to, and is united with, the deep median valley of enamel separating protolophid and metalophid; thus, a large cusp (hypoconid) is isolated on the labial margin of the tooth. The area posterior to this angle of enamel is probably a posterior cingulum as supposed for the two earlier species. In none of the specimens that I have examined is a break evident between the metaconid and the posterior cingulum. The separation of the hypoconid remains longer than the separation of cusps of the protolophid, but eventually the hypoconid and metaconid unite to form a single straight loph. The lower premolars of Liomys adspersus exhibit basic features similar to those seen in L. salvini. Heteromys (Fig. 52D). — The lower premolars of mice of this genus are dis¬ tinct from those of Liomys. Instead of being composed of two lophs as in the latter, the lower premolars of Heteromys are composed of three or four lophs. The three species that I had available for study had only three lophs on the pre¬ molar, but Wood (1935:204-205) described in detail the teeth of some species with four lophs. The anteriormost loph is formed by a well-developed anterior cingulum. This cingulum may consist of two or more cusps. Behind the anterior cingulum is a loph composed of three cusps, although in none of the specimens that I have examined are these three cusps ever separated from each other. The protoconid and mesoconid are united along their medial margins into a single straight loph. The anteroconid is attached to the anteromedial margin of the loph formed by the protoconid and mesoconid. In the species with four lophs, the anteroconid has been forced out of the protolophid and forms the new loph along with a labial and lingual style left behind by the migrating anterior cingulum (Wood, 1935: 204-205). As the premolars of specimens examined begin to show wear, the anteroconid first is connected with the lingual margin of the anterior cingulum. The metalophid, which is separated from the protolophid by a deep median valley of enamel, is composed of a hypoconid and metaconid, which are united into a single straight loph. There is a style developing on the labial margin of the tooth; this style does not appear to represent the same development seen on the labial margin of the metalophid of Liomys where a re-entrant angle developed in this region. There does not appear to be a posterior cingulum in Heteromys. GENOWAYS— SYSTEMATICS OF LIOMYS 277 Specimens used in comparisons of premolars are listed below. Liomys irroratus, — 1 mi. SSE Ameca, Jalisco, 4 (KU 31127-29, 31134); 4 mi. NE Ocotlan, Jalisco, 4 (KU 31156, 31 164, 31174, 31185); 2 mi. WNW Ocotlan, Jalisco, 1 (KU 31187); 3 mi. ENE Santa Cruz de las Flores, Jalisco, 1 (KU 31178); 3V2 mi. WNW Zap- oltitic, Jalisco, 5 (KU 97189, 97196-97, 9720.-01). Liomys pictus. — 2 mi. SSE Autlan, Jalisco, 4 (KU 31205-06, 31214-15); 11 mi. SW Autlan, Jalisco, 3 (KU 107666-68); 2 mi. ESE Tequila, Jalisco, 1 (KU 33459); 7 3/10 mi. ESE Amatlan de Canas, Nayarit, 1 (KU 100496); 10 mi. S Juchatengo, Oaxaca, 2 (KU 98774, 98780); 20 mi. S, 5 mi. E Sola de Vega, Oaxaca, 1 (KU 99792); IV2 mi. N Badira- guato, Sinaloa, 1 (KU 96668); 2 mi. E Costa Rica, Sinaloa, 1 (KU 100495); 8 km. N Villa Union, Sinaloa, 1 (KU 96010); Boca del Rio, Veracruz, 1 (KU 30044). Liomys salvini. — Km. 24 on highway S Guatemala City, Guatemala, 1 (KU 83963); 3 km. N, 4 km. W Diriamba, Nicaragua, 4 (KU 110440-41, 110444, 110446); Finca Amayo, Nicaragua, 1 (KU 104708); 5 mi. NW Managua, Nicaragua, 1 (KU 71118); 4 mi. W Managua, Nicaragua, 1 (KU 71133); 3 mi. SW Managua, Nicaragua, 3 (KU 71204, 71210, 71212), 3 mi. S Managua, Nicaragua, 1 (KU 71188); 2 km. N Merida, Nicaragua, 2 (KU 1 15400-01), 2 km. N, 3 km. E Merida, Nicaragua, 2 (KU 115403-04); Moyogalpa, Nicaragua 2 (KU 97933, 1 15386). ' Heteromys desmarestianus.— Rio Queleya, U2 mi. E Yepocapa, Guatemala, 1 (KU 65069); 1 mi. SW Santiago Sacatepequez, Guatemala, 2 (KU 71254-55); 2 mi. SW Santiago Sacatepequez, Guatemala, 2 (KU 71247, 71252); Sabana de San Quintin, Chiapas, 1 (KU 102653); Hda. Tepeyac, Nicaragua, 1 (KU 104596). Heteromys gaumeri. — 5 km. S Champoton, Campeche, 2 (KU 92146-47); 7 km. W Escarcega, Campeche, 1 (KU 92150); Esmeralda, Quintana Roo, 1 (KU 35196). Heteromys lepturus. — Vista Hermosa, Oaxaca, 3 (KU 99845, 99849, 99852). Molars Liomys (Fig. 5 3 A). — The structure of the upper and lower molars is essen¬ tially the same in all species of Liomys. All molars are composed of two lophs (protoloph or metalophid and metaloph or hypolophid) separated by a median valley of enamel. The three cusps composing each of the lophs are discernable for only a short period after the tooth has erupted. As pointed out by Wood ( 1 935: 199), Goldman’s (1911:32) statement that in members of this genus “loops of molar crowns normally without additional enamel islands even in young” is in¬ correct. These islands, resulting from the presence of a posterior cingulum dorsally and an anterior cingulum below, are present in the youngest individuals of this genus available, especially on the lower molars, although for a short period of time. As the molars wear, the protoloph and metaloph usually are united at their lingual edge and then at the labial edge. The uniting of these lophs usually does not occur until the permanent premolar is in place. One difference among species of Liomys involves the wear pattern of the first upper molar. In irroratus , pictus , and most likely spectabilis, the protoloph and metaloph meet as described above and as a result an island of enamel representing the middle of the median valley is isolated near the center of the tooth. This island is present in almost all, if not all, individuals of these species at the proper stage of wear and probably persists for some time. In salvini and probably adspersus, on the other hand, an island of enamel isolated on the first upper molar is a rare occurrence, although I have seen some specimens with this structure (as did Wood, 1935:199). Either this structure 278 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY Fig. 53. — Crown patterns of the first upper (top) and lower (bottom) molars of one species of Liomys and Heteromys. All teeth illustrated are from the right side. For all teeth, anterior is to the top; for the upper molar, lingual is to the right and for the lower is to the left. Ab¬ breviations used in the illustration are: ME, metaloph; PC, posterior cingulum; PR, pro- toloph; ac, anterior cingulum; hy, hypolophid; me, metalophid. The species are: A, Liomys irrorat us (KU 97189); B, Heteromys lepturus (KU 99845). is formed only in a few individuals or else the island is present for only a short period of time. The explanation for this difference I believe, is that in irroratus and pictus the enamel in the median valley is deepest at the center and shallower on the edges (shallowest on lingual margin), whereas in salvini the enamel at the center of the tooth is nearly the same depth as on the labial margin of the tooth (again being shallowest at lingual margin). On the lower molars, the metalophid and hypolophid are first united at their labial edge and later at the lingual edge. Heteromys (Fig. 53B). — The upper and lower molars of the species of Heter¬ omys appear to be composed of three lophs instead of two as seen in all but the youngest specimens of Liomys. A well-developed posterior cingulum is present on the upper molars and a well-developed anterior cingulum on the lower molars. In the three species ( desmarestianus , gaumeri, and lepturus ) that I studied, the posterior cingulum is well developed and is nearly as long as the metaloph. The cingulum is always connected with the metaloph at their lingual margins and free at early stages of wear at the labial margin (this is clearly evident in a specimen of desmarestianus, KU 71258, from near Jinotega, Nicaragua). As wear progresses, the labial margins of the cingulum and metaloph are united, thus iso¬ lating the enamel valley between them. This island of enamel is present for varying lengths of time; in gaumeri the island disappears much more quickly than in desmarestianus. The anterior cingulum on the lower molars were only one-half or less the length of the metalophid. The labial margins of the two lophs are separate before GENOWAYS— SYSTEMATICS OF LIOMYS 279 wear begins, as wear progresses, the metalophid and cingulum are united at their labial margin isolating one or more islands of enamel in the metalophid. The protoloph and metaloph become united at their lingual margins when the deciduous premolar is still present, but the labial margins do not unite until the tooth exhibits considerable wear. I have the impression that the lingual margins ot the protoloph and metaloph become united much sooner in Heteromys than in Liomys, whereas the labial margins of these are united sooner in Liomys than in Heteromys. In the lower molars, the reverse pattern is seen, with the labial mar¬ gins ot the metalophid and hypolophid being the first to be united and then the lingual margins. Specimens examined in addition to those listed for premolars are as follows. Liomys salvini. — 3 mi. SW Managua, Nicaragua, 3 (KU 71 166, 71196, 71215); 2 mi. N Sabana Grande, Nicaragua, 1 (KU 71128). Heteromys desmarestianus. — 5 mi. S, 2 mi. E Jinotega, Nicaragua, 1 (KU 71258). Heteromys gaumeri. — 4 km. NNE Felipe Carrillo Puerto, Quintana Roo, 1 (KU 92140); Peto, Yucatan, 1 (KU 93640). Size of Teeth Length and width of the upper premolar and first upper molar was measured on selected specimens of all species of Liomys and Heteromys desmarestianus using a binocular microscope with an ocular micrometer. On the average, specimens of Liomys salvini have the shortest and narrowest premolars of species studied (see Table 37). Specimens of Liomys irroratus had on the average the longest and widest upper premolars. When the size of the pre¬ molars is considered relative to the overall size of the animal (greatest length of skull used as a standard), specimens of Heteromys were found to have much shorter and narrower premolars than specimens of Liomys. Essentially these data mean that the premolars of Heteromys are actually about the same size as those of Liomys , but relative to the overall size of the animals they are smaller. The first upper molar is shortest in specimens of Liomys pictus and Liomys salvini and longest in Heteromys desmarestianus. This tooth is widest in Liomys irroratus and narrowest in Liomys salvini. Again, when the size of the tooth is compared to the overall size of the specimen, the length of the first upper molar is somewhat shorter and much narrower in Heteromys than in the species of Liomys. Specimens used for study of tooth size are listed below. Liomys irroratus. — Omilteme, Guerrero, 1 (KU 98761); 2 mi. E Omilteme, Guerrero, 1 (KU 98763); 3 mi. NNW Ameca, Jalisco, 3 (KU 39903-05); Cerro la Caldera, Mexico, 1 (KU 28050); 5 mi. S, 1 mi. W Texcoco, Mexico, 1 (KU 49015); Iturbide, Nuevo Leon. 2 (KU 55950-51); 3 mi. ESE Oaxaca, Oaxaca, I (KU 61602); 4 mi. ESE Oaxaca, Oaxaca, 1 (KU 62633); 1 mi. SSW Tilapa, Puebla, 3 (KU 62617-18, 62620); 8 mi. SW Ramos, San Luis Potosi, 2 (KU 59533, 59536); 7 km. S, 2 km. W San Fernando, Tamaulipas, 3 (KU 36949-50, 36952); Rio Nieves, 1 mi. N Rancho Grande, Zacatecas, I (KU 84697). Liomys pictus. — Finca San Salvador, 15 km. SE San Clemente, Chiapas, 2 (KU 102647, 102650); 3 3/10 mi. NE Contla, Jalisco, 1 (KU 96040); 2 2/10 mi. NE Contla. Jalisco, 3 (KU 96046-47, 96056); 14 km. S Durazno, Jalisco, 4 (KU 87427-28, 87430-31); 10 mi. S Juchatengo, Oaxaca, 1 (KU 98775); 20 mi. S, 5 mi. E Sola de Vega, Oaxaca, 2 (KU 98768- 69); 5 mi. NW Mazatlan, Sinaloa, 4 (KU 85788, 85794-95, 85797); 3 mi. NNW Mazatlan, Sinaloa, 1 (KU 39820); 3 km. E San Andres Tuxtla, Veracruz, 2 (KU 24026, 24028). Table 37. — Measurements of the upper premolar and first upper molar of five species o/Liomys and oj Heteromys desmaiestianus. Liomys Liomys Liomys Liomys Liomys Heteromys 280 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY 3 NO o *o >/~> o NO ON 04 O O NO NO — ; d d d O ''t NO O O — 1 OO — O — — ' — o OO NO r-~ >o 04 04’ +1 +1 +1 CO 3 Co un I II z ■3 ^ o o «o Vi h o — — — o +1 OO ro t" O 00 V i O h - d d d o +1 o On v O " ON On O O o o — o +1 oj t" OO NO 04 04 V O Vi M •'fr m r~ o — — — o NO on Tt o r~ ion — m O o ^ v o m m m o — no — ON 00 — O O On *0 OO — O 2 ^ 04 +1 X +1 +1 X ^ o 3 04 ^Z 3 c/3 o CL x: ON on io m -C ■*-* OX) c 00 c (U _) — o o O ON — On O — 0 — 0 O +1 00 c V 00 ON ON IT) in 2 2 2 2 2 2 2 2 2 - oo ro On © oi o +1 NO ION Tf ON O 04 o +1 ON "3- T 9 d o +1 NO NO v~i 04 O d o +1 NO — 04 © d o +1 r- O ro ro O d o +1 E 3 E »*: UJ GENOWAYS— SYSTEMATICS OF LIOMYS 281 ON m vi o N d o +1 +1 +1 +1 r-~ >n O >n t" NO <0 ON O — ' d o o oo — c rj Tt no n- r- _ , d d d O O'. •O O O cn •O lO t'' O o (N O — < r<"i N /-i O d d d d +1 +1 +1 +1 "Q £ ro U -J CQ < f- — o o — ION m NO o — — ’ d o +1 0. _c w T3 £ N- r~ 0100 — >o tj- o — — ' — ‘ o +1 >o o — O —’—’(NO +1 o o X 3 44 c/3 «4— o x: r~ — t~- N- ON o N- O O +1 NO (N O "O — o O — ’ — ’ — ’ o +1 G\ •4— * OX) c (N s NO OX) c m 1) (N (N x: N- o ^3 (N a> m r- O NO 00 o i— i V3 d C/D d o T3 — ’ d — * o 4— » C/D V3 a> pj aj o 'd CL JO *-» "O £ +1 +1 NO 00 On rf io 0 Tf ON 00 O O d d d O X +1 3 jj C/3 >— o JO On t"- O o +1 (N 00 NO o o On >0 r*N d d d Tf r4 >o >o (N >0 (N I O — ’ — ’ — ’ o 0> *—> a u a i JO 4— > -o £ 0(0 0 OO 00 Tj- d d d +1 +1 NO 00 o o +1 •— 03 c/5 2 £ £ c 3 £ 3 E UU c 3 £ 3 £ UQ c a c >< CS c * 03 aj 03 00 o> 03 00 */■> O o o 00 > £ o u c C/5 c O X +— » C/5 c 0 0 OO t-~ m 10 O) X 00 a 00 iH (N d r<) rn 00 r- u ctf Q 00 3 l/~> l/~> u "5 C/5 C -5 O 0 O 0 £ 5 00 00 m q (N q C4-H (N rn 3- rn rn Q ° O O 0 10 q q 00 m* cn rd m’ rr » o XI E 3 O O 3 00 3 mJ sX X X) *o ^ x c4 oo t"-' d m' d rt r-~ x x x x x 00 rn in ^ 00 oo' 00 00 5 .0 d X 3 ■^4 o §■ >> oo Os o d .3 "3 «5 .0 d o o r-' r-' >3 3 05 V. s 5 o •*-4 a: X 06 m ■'t «o m X m d iri iri 10 in x — O') ^ X 00 d d d d d X 03 00 cn o 03 00 « h 'O 't Os Os O Os X (N OO ON X >. E •ts 3 3 c CJ -o 9 c —J 03 fN x □ d _3 X 0 X >n wN >n in X 0 0 OOOO 0 (N (N N N N N — ^ 10 pericenfric inversions H DESMARESTIANUS 2 N = 60) F N = 82 1 Fig. 60. Hypothetical pathways for chromosomal evolution in five species of Liomys and one of Heteromys. This is the most parsimonious arrangement based upon the as¬ sumption that Liomys irroratus possesses the most primitive karyotype. In order to gain some insight into the possible direction of chromosomal evolution in this group, it is necessary to attempt to identify the primitive kary¬ otype of the genus. The characteristics of a “primitive” karyotype for mam¬ mals have been considered by several authors (Klinger, 1963; Hamerton et al., 1963) to be a high diploid number and numerous telocentrics. This is based upon the assumption that in mammals the main mechanism for changing chromosomal numbers has been centric fusion (Hsu and Mead, 1969; Matthey, 1958; Nadler, 1966, 1969), although there is growing evidence for centric fission in at least some groups (Hoffmann and Nadler, 1968). Following the classical interpre¬ tation, Liomys irroratus would be considered as having the most primitive kary¬ otype for the heteromyine rodents examined. This species has, along with Heteromys desmarestianus, the highest diploid number of the six species studied and has a high number (27) of telocentric chromosomes. (Aside from this evi¬ dence Liomys irroratus seems to have retained more primitive characteristics in other characters studied than have other members of the genus Liomys.) Also, sharing a diploid number of 60 with Heteromys desmarestianus suggests that the primitive diploid number for the Heteromyinae was 60, although final assessment of this possibility must await more data concerning the karyology of other mem¬ bers of the genus Heteromys. I can find no objections to considering Liomys irroratus as having the most primitive chromosomal compliment in the genus. If the assumption of Liomys irroratus having the “primitive” chromosomal compliment is accepted, then possible pathways for derivation of the chromoso¬ mal compliments of other members of the genus can be constructed (Fig. 60). My arrangement should not be construed as being the only possible solution to the problem, but I do believe it is the most parsimonious on the basis of the origi¬ nal assumption. From the karyotype of Liomys irroratus , that of Heteromys 300 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY desmarestianus can be derived by 10 pericentric inversions. To derive the chro¬ mosomal compliments of Liomys adspersus from that of Liomys irroratus would require two centric fusions and 1 1 pericentric inversions and to obtain the chro¬ mosomal compliment of Liomys salvini would require one additional pericentric inversion. On the other hand, to obtain the karyotype of Liomys spectabilis from Liomys irroratus would require six centric fusions and one pericentric in¬ version and to obtain the karyotype of Liomys pictus would require only one additional pericentric inversion. The split into a pictus-Y\nz and salvini-Ymt may not have been directly from irroratus, because a mouse with two centric fusions and one pericentric inversion beyond the chromosomal compliment of irroratus would still be basic to both lines; nevertheless, beyond such a point a split into the two lines would seem to be the most logical explanation. It is of interest to note that the salvini- line appears to have emphasized pericentric inversions in the evolution of its chromosomes, whereas the pictus-Yme appears to have empha¬ sized centric fusions. Specimens examined from which karyotypic material was available are as follows. Liomys irroratus. — La Cienegilla Springs, 46.9 mi. S Jimenez, Chihuahua, 1 (TTU 9868); 5 km. N San Jose, Guanjuato, 3 (WBC 866-868); Omilteme, Guerrero, 1 (KU 120582); 12 km. N Cd. Victoria, Tamaulipas, 1 (WBC 865); 5 mi. SE Brownsville, Cameron Co., Texas, 2 (TTU 9857-58). Liomys pictus. — 4 km. NE Contla, Jalisco, 1 (KU 120590); 2 km. NW Emiliano Zapata, Jalisco, 1 (KU 120587); Rio Cuchajaqui, 8 mi. S Alamos, Sonora, 1 (supplied by J. L. Patton). Liomys spectabilis. — 6 km. NE Contla, Jalisco, 1 (KU 120606). Liomys salvini. — 16 km. SW Tapachula, Chiapas, 2 (KU 120597, 120600); no specific locality, Nicaragua, 1 (supplied by J. L. Patton). Liomys adspersus. — 2 km. NE Buenos Aires, Panama, Panama, 3 (KU 121528-30). Ectoparasites The list of ectoparasites for species of Liomys and Heteromys given in Ap¬ pendix II and the discussion below are based primarily upon two sources of information — a compilation of all records in the published literature and ecto¬ parasites collected during field work under a contract (DA-49- 193-MD-22 15) to Dr. J. Knox Jones, Jr., from U.S. Army Medical Research and Development Command and identified by experts on the various taxonomic groups (listed in Acknowledgments). The latter records are indicated in Appendix II by num¬ bers, which correspond to numbered localities of capture for the host given at the end of the list. Liomys irroratus. — Twenty-two species of mites, of which 17 are trombiculids, three laelapids, and two macronyssids, have been found on specimens of Liomys irroratus. Of these 22 species, eight have been reported also from pictus ( Ectomyx fusicornis, Euschoengastia gagarini , Eutrombicula alfreddugesi, Leptotrombidium panamense, Pseudoschoengastia audyi, P. hoffmannae, P. hungerfordi, and Steptolaelaps liomydis ), two from salvini ( Eutrombicula alfred¬ dugesi and Leptotrombidium panamense ), three from adspersus ( Androlaelaps fahrenholzi, Leptotrombidium panamense, and Ornithonyssus bacoti), and two from Heteromys desmarestianus ( Androlaelaps fahrenholzi and Eutrombicula GENOWAYS— SYSTEMATICS OF LIOMYS 301 alfreddugesi). Two species of ticks have been found on L. irrorat us of which one, Ixodes eadsi’ has also been reported from L. salvini. Two species of lice and three of fleas are known from this spiny pocket mouse; two of the fleas ( Polygenis gwyni and P. martinezbaezi ) also have been reported from Liomys pictus. Liomys pictus. Of the 23 species of mites known to occur on Liomys pictus , 20 are trombiculids and three are laelapids. L. pictus had three species of mites in common with L. salvini { Eutrombicula alfreddugesi, Leptotrombidium pana- mense, Steptolaelaps heteromys), two in common with L. adspersus {Leptotrom¬ bidium panamense and Steptolaelaps heteromys ), and two in common with Heteromys desmarestianus {Eutrombicula alfreddugesi and Steptolaelaps heteromys). Only one species of tick, Ixodes sinaloa, is known from this species; this tick also has been reported from L. salvini. One species of louse and five of fleas are known from L. pictus and of these species, one flea {Polygenis vul- camus) is known also from L. salvini and two fleas are known from L. irroratus (see above). Liomys salvini. — Sixteen species of mites (f 1 trombiculids and five laelapids) are presently known from this Middle American species of Liomys. Of these 16 species, four have been reported also from L. adspersus {Ascoschoengastia dys- crita, Eutrombicula goeldii, Leptotrombidium panamense , and Steptolaelaps heteromys) and five have been reported from Heteromys desmarestianus {Asco¬ schoengastia dyscrita, Eutrombicula alfreddugesi, Eutrombicula goeldii, Trom- bicula dunni, and Steptolaelaps heteromys). Two species of ticks and one flea are known from this species. The one species of louse, Fahrenholzia fairchildi, known from L. salvini also has been reported from L. adspersus and Heteromys des¬ marestianus. Liomys adspersus. — Fourteen species of mites (10 trombiculids, three lae¬ lapids, and one macronyssid) have been reported thus far from this Panamanian pocket mouse. Seven of these 14 species are known also from Heteromys des¬ marestianus {Androlaelaps fahrenholzi, Ascoschoengastia dyscrita, Eutrombicula goeldii, Pseudoschoengastia bulb if era, P. tricosa, Hirst ionyssus microchelae, and Steptolaelaps heteromys). An unspecified species of the genus Amblyomma is the only tick reported from adspersus-, one louse, Fahrenholzia fairchildi, also known from L. salvini and H. desmarestianus has been recorded from adspersus. Of the three species of fleas taken from this species, none has been reported from the other four species of mice presently under consideration, but one {Polygenis dunni) has been taken on Heteromys anomalus and another {Polygenis klagesi) is known from Heteromys australis. Heteromys desmarestianus. — The occurrence of 26 species of mites (16 trom¬ biculids, nine laelapids, and one speleognathid) has been reported from Hete¬ romys desmarestianus. No records of ticks are available for this species. Three species of lice and five species of fleas from three families have been taken from specimens of Heteromys desmarestianus. Conclusions. — Because some of the life stages of trombiculid mites and argasid and ixodid ticks are free-living away from vertebrate hosts, these taxa are of relatively little use in elucidating phylogenetic relationships of their hosts. 302 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY However, when sufficient information becomes available about the ecology of these parasites, differences in the composition of the trombiculid mite and tick faunas may give insight into ecological differences of their hosts. As pointed out by Wenzel and Tipton (1966:699), genera of laelapid mites exhibit a striking host association at the supergeneric level. Members of genus Eubrachylaelaps were found primarily on peromyscine cricetid rodents; the genera Laelaps and Gigantolaelaps were primarily taken from complex-penis- type cricetine rodents. The principal hosts of the genus Tur were caviomorph rodents. Members of the genus Steptolaelaps were restricted to heteromyid ro¬ dents. Two genera, Androlaelaps and Hirstionyssus, have a wider host tolerance in that members of the former genus are found on a wide variety of mammalian hosts and representatives of the latter primarily on heteromyid, cricetid, and sciurid rodents. Because fleas are somewhat more closely tied to their mammalian hosts, it was hoped that members of this group might reveal useful information concerning members of the genera Liomys and Heteromys. At first glance, there appears to be little overlap in the associated species of fleas between these spiny pocket mice. However, further examination reveals that, in most cases, heteromyid rodents are only incidental hosts and that other mammals are the primary hosts of the listed fleas. The only possible exception may be Polygenis dunni, of which more specimens were taken in Panama from Liomys adspersus than from any other host (Tipton and Mendez, 1966:298), and the two species of Wenzella, which are known only from Heteromys desmarestianus. Dr. Robert Traub after reviewing my list of fleas occurring on Liomys and Heteromys made the following statement in a letter (4 May 1971) to J. Knox Jones, Jr.: “. . . it is important to note that various species of Polygenis that are listed for Liomys actually infest many other hosts and are generally far more abundant on such rodents than on Liomys. Of all the fleas listed in my records from Liomys , I doubt that there is one that is specific or even characteristic. My own records of fleas from Hete¬ romys are virtually nil, but the species of Wenzella are the only ones listed. . . that fall in the category of Heteromys- fleas, and these are nest-inhabiting species.” One group that does offer some evidence on possible phylogenetic relationship of their hosts are the anopluran lice of the genus Fahrenholzia. Lice are highly host-specific, passing all life stages on their host and resisting even host-to-host transfer unless there is intimate contact. Species of Fahrenholzia occur only on heteromyid rodents; fortunately, this genus of lice has received recent taxonomic study by Stojanovich and Pratt (1961) for the four species occurring on members of the subfamilies Dipodomyinae and Perognathinae, and by Johnson (1962) for the “ microcephala group” that occurs on members of the subfamily Het- eromyinae. Of the seven species of the “ microcephala group,” five are known from a single species of Liomys or Heteromys (Table 42). The remaining two species, fair- childi and ferrisi , each have been reported from three hosts; specimens of fair- childi are known from L. salvini, L. adspersus, and Heteromys desmarestianus. GENOWAYS— SYSTEMATICS OF LIOMYS 303 Table 42. Distribution of species of the anopluran genus Fahrenholzia on their mam malian hosts of the genera Liomys and Heteromys. 5 ■W- *2 •2? u o 2 Host species 'i) k. k. -sr £ <3 k k k k k k k Liomys irroratus Liomys p ictus Liomys spectabilis Liomys salvini Liomys adspersus Heteromys anomalus Heteromys desmarestianus Heteromys gaumeri Heteromys goldmani + + + + + + + + + + + and specimens of ferrisi are known from three species of Heteromys ( des¬ marestianus , gaumeri , and goldmani ). Because fairchildi is the only species known from L. salvini and L. adspersus and ferrisi is the only species known from H. goldmani and H. gaumeri, it would appear highly probable that these species of lice evolved on one or both of these hosts and subsequently dispersed to Heteromys desmarestianus. Fahrenholzia hertigi is known only from H. des¬ marestianus and is most likely the species that has evolved on this host. Only two hosts are known to harbor more than one species of Fahrenholzia. Liomys irroratus is the only known host of F. ehrlichi and F. texana\ as pointed out above, three species {fairchildi , ferrisi, and hertigi ) have been recorded for Heteromys desmarestianus, although only hertigi is unique to that species. Endoparasites Little information has been recorded concerning the endoparasites of heteromyine rodents; all records available were found coincidental to studies of human diseases or in general surveys for a particular type of parasite. Two species of Coccidia, Eimeria liomysis and Eimeria picti, have been de¬ scribed from Liomys p ictus from Sinaloa and Nayarit and from Liomys irroratus from Jalisco (Levine et al., 1958; Ivens et ai, 1959). These protozoan parasites were recovered from fecal samples of these two species of mice. Esquivel-R. et al. (1967:954) described Trypanosoma zeledoni based on blood smears pre¬ pared from specimens of Liomys salvini from Costa Rica. During a survey for possible reservoir hosts of leishmaniasis in the vicinity of Roaring River, Cayo, British Honduras, Leishmania mexicana was isolated from cutaneous lesions on the tails of six of 58 specimens of Heteromys desmarestianus (Lainson and 304 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY Strangway-Dixon, 1964u:3, 1964/?: 146-147). The only other rodents from which Leishmania was identified were Ototylomys phyllotis and Nyctomys sumi- chrasti, which also had cutaneous lesions on their tails. Leishmaniasis has been reported from the human population of this area of British Honduras, but only from people who lived or worked in heavily forested areas. Disney (1964:581) reported a specimen of Heteromys desmarestianus from Spanish Lookout, Cayo, British Honduras, with a heavy infestation of leishmaniasis on its ears, and upon post-mortem examination was found to have visceral involvement of the liver, spleen, and lungs. Caballero y Caballero (1959) reported specimens of two species of nematodes, Trichuris sp. and Longistriata vexillata, from specimens of Liomys salvini taken in the vicinity of Mapastepec, Chiapas. Hookworms were observed in the duodenum of two specimens of Heteromys desmarestianus from the vicinity of Palenque, Chiapas, by Kuns and Tashian (1954:101). No bacterial or viral diseases have been reported from species of either Liomys or Heteromys, although 5 1 specimens of Liomys salvini were examined during a leptospirosis survey in Nicaragua (Clark et al. , 1966). Reproduction Reproductive data presented herein that were not drawn from the published literature are based upon information recorded on specimen labels or in the field notes of the collector. Although data were recorded from specimens in most museums visited, the bulk of the data are from material housed in the Museum of Natural History at The University of Kansas. Females here recorded as non¬ pregnant were only those that were so noted by the collector. Liomys irroratus. — Of the 281 adult females of this species examined for which reproductive data were available, 52 were gravid and 12 were lactating. Pregnant females were recorded for all months except February, April, and June, and in one of these months, June, a lactating female was taken. The peak repro¬ ductive activity appears to be in the months of August to November when over a third of the females were pregnant in each month (Table 43). The mean num¬ ber of embryos per female was 4.39 (mode four) with a range of two to eight. Males with enlarged testes were collected in each of the eight months (Table 43) for which data are available. Some data reported herein were recorded earlier by Dalquest (1953:122), Hall and Dalquest (1963:285), and Alvarez (1963:433). Davis (1944:389) recorded a female of L. i. guerrerensis from 15 km. SW Chilpancingo, Guerrero, that was found to be gravid on August 17 (no embryo count given), and Koestner (1941: 12) obtained a female on an unspecified date at Ojo de Agua, Nuevo Leon, that carried five embryos. Fleming (1969) reported immature individuals of Liomys irroratus from all months except May; he believed that irroratus breeds through¬ out the year, but with a peak from November to February. Liomys pictus. — Forty-nine of the 284 females with reproductive data were noted as being pregnant and an additional 1 1 were lactating (Table 44). Pregnant females were taken in all months except January and October and a lactating GENOWAYS— SYSTEMATICS OF LIOMYS 305 Table 43. Reproductive data for Liomys irroratus based on information from throughout t te range of the species. Mean and range are given for crown-rump length of embryos and length oj testes when three or more values were available. Females Month of capture Number examined Number pregnant Number lactating Crown- rump length of embryos Males Number examined Length of testes January 5 2 0 3,? 1 26 February 17 0 0 1 22 March 58 4 1 12.5(8-18) 0 April 10 0 0 0 May 18 1 1 8 0 June 31 0 1 5 22.4(20-24) July 40 4 4 11.0(6-15) 1 24 August 39 14 0 9.9(3-15) 1 1 20.5(13-25) September 14 7 1 11.3(5-25) 5 17.2(7-20) October 23 11 3 11.1(3-22) 2 15,20 November 19 7 1 14.3(8-20) 0 December 7 1 0 24 3 20.0(15-23) female was recorded in the latter month. Most of the pregnant females were taken in April, August, September, and November. No pregnant females have been recorded from 2 December to 5 February, although more than 70 females with reproductive data are available from that period. The mean number of embryos per female was 3.80 (mode three) with a range of two to six. Males with enlarged testes were taken in most months (Table 44), but males taken in November through February had testes that averaged much smaller than those of males taken between March and October. Some data reported herein were recorded earlier by Hooper (1955:9) and Hall and Dalquest (1963:284). Wagner (1961:209) reported 17 females of Liomys pictus taken in the second half of March near Villa Flores, Chiapas, that carried the following number of embryos: one female with three embryos, 1 1 females with four embryos, four females with five embryos, and one female with six em¬ bryos. Eisenberg and Isaac (1963:65) and Eisenberg (1963a:l 1) found the mean of six litters of L. pictus born in captivity was 3.5 with a range of two to five. The gestation period for one of these litters was 24 days and for another 26 days (Eisenberg and Isaac, 1963:64; Eisenberg, 1963^:13). Immature individuals were found in each month for which Fleming (1969) had data (no specimens available for September, October, and December), excepting May. Immature specimens formed the largest proportion of the sample in March, 43.6 per cent being immature. Liomys spectabilis. — Only one female of the six (one taken in March and five in September) with reproductive data contained embryos. This female, which was taken on 26 September at a place 2 2/10 mi. NE Contla, Jalisco, carried five 306 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY Table 44. — Reproductive data for Liomys pictus based on information from throughout the range of the species. Mean and range are given for crown-rump length of embryos and length of testes when three or more values were available. Month of capture Females Males Number examined Number pregnant Number lactating Crown-rump length of embryos Number examined Length of testes January 38 0 0 7 8.1(5-14) February 38 3 0 15.7(10-25) 3 9.3(5-18) March 20 2 0 6,7 9 16.3(5-24) April 48 15 3 10.2(4-20) 0 May 19 2 1 6,20 1 20 June 16 1 2 9 8 16.2(12-19) July 15 1 0 ? 10 18.9(6-27) August 23 6 1 15.5(8-23) 11 19.5(10-25) September 30 13 0 15.7(3-24) 27 17.9(4-25) October 4 0 1 5 13.4(5-20) November 14 5 1 13.4(5-20) 2 5,5 December 19 1 2 8 4 9.8(6-17) embryos that measured 4 in crown-rump length. Four males trapped in mid- September had enlarged testes (those of two measuring 21 and the other two, 22). Liomys salvini. — Reproductive data were known for 201 adult females of Liomys salvini of which 33 were pregnant (Table 45); no lactating females have been recorded. Females carrying embryos have been taken in eight different months (those available from April and June were nonpregnant and no data are available from September or October). Females with embryos formed the largest percentage of the total females in November, December, and February, although the total sample from the first two months was small. The mean number of em¬ bryos per female for this species was 3.55 (mode three) with a range of two to six. Adult males with enlarged testes have been taken in the months of January through April and in July and August (Table 45). Goodwin (1946:375) reported that Liomys salvini in Costa Rica bred at all seasons of the year and the usual number in the litter was four. In El Salvador, Burt and Stirton (1961:54) recorded single females taken on 8 and 9 January, that each carried three embryos, and Felten (1957:154) reported a female taken in February with two embryos and one taken in March with three. Liomys adspersus. — Fleming (1969, 1971) studied this species in the Panama Canal Zone; he found the number of embryos ranged from two to four in five females, with a mean of 3.2. Fleming believed this spiny pocket mouse was a seasonal breeder, with females beginning to breed in late November, reaching a peak in March and April, and declining thereafter. Each female was found to produce an average of 1.44 litters per year. The male reproductive cycle was found to follow that of the female; testes of males were largest from October to April. GENOWAYS— SYSTEMATICS OF LIOMYS 307 Table 45. Reproductive data for Liomys salvini based on information from throughout the range of the species. Mean and range are given for crown-rump length of embryos and length oj testes when three or more values were available. No females were examined in which lactation was noted by the collector. Females Month of capture Number examined Number pregnant Crown-rump length of embryos Males Number examined Length of testes January 16 2 14,26 1 15 February 29 11 10.7(3-16) 6 15.7(7-19) March 38 4 10.3(4-21) 15 15.1(4-25) April 5 0 1 25 May 18 2 2,14 0 June 19 0 2 7,9 July 41 9 9.4(3-20) 18 17.901-25) August 28 2 7,4 6 15.2(11-19) September 0 0 October 0 0 November 2 1 10 0 December 5 2 12,24 0 Heteromys desmarestianus. — A female from 2 mi. SW Santiago Sacatepequez, Guatemala, taken on 1 February carried three embryos that measured 4 in crown- rump length and one from Finca El Paraiso, Chiapas, trapped on 21 March carried two embryos that measured 15. Males with enlarged testes were taken in Nicaragua in March (testes 17, 19) and June (testes 15, 20) and in Chiapas in June (testes 14, 18, 19, 19, 20, 20). Juveniles (deciduous premolars present) of H. desmarestianus are present in the Museum of Natural History from the months of January, February, June, July, and August. Murie (1935:25) reported two females of Heteromys desmarestianus from British Honduras with four embryos each that were obtained on 25 February and 1 March, and a female from Guatemala that carried three embryos on 1 1 April. Kuns and Tashian (1954:101), in the course of a parasitological study in the vicinity of Palenque, Chiapas, obtained a female of H. desmarestianus on 21 July that contained three embryos. Goodwin (1934:30) reported the finding of a nest of this species near Puebla, Guatemala, that contained four young. Studying the ecology of mammals in the Panama Canal Zone, Enders (1935:451) found a male with enlarged testes in January and a female taken at the same time that had been lactating recently. Fleming (1970:478), who also studied this species in the Panama Canal Zone, recorded pregnant or lactating females in April, May, July, and November, and juvenile mice in January, February, July, August, October, and December. Based on his data, Fleming (1969, 1970:478) believed that H. desmarestianus breed at all times of the year in Panama. Heteromys gaumeri. — A female in the Museum of Natural History taken on 26 December 1 962 at a place 7 km. N, 5 1 km. E Escarcega, Campeche, contained 308 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY three embryos that measured 9 in crown-rump length. Eleven other females in the Kansas collection taken in January (1), March (1), July (7), and December (2) were nonpregnant. Males with enlarged testes were taken in March (19 in length), April (20), July (15, 16, 21), and August (17). Juveniles (deciduous premolars present) of Heteromys gaumeri were taken on the Yucatan Peninsula in January, April, July, and August. Hatt (1938:336) recorded a female taken at Chichen Itza, Yucatan, on 26 October that contained two embryos. Gaumer (1917:132) in his monograph on the mammals of Yucatan stated females of H. gaumeri gave birth to four young. Heteromys goldmani. — Wagner (1961:209) believed that this species bred from May to December in the vicinity of Villa Flores, Chiapas, and produced three or four embryos. Heteromys lepturus. — At Vista Hermosa, Oaxaca, on 28 June a female was taken that carried two embryos, which measured 7 in crown-rump length; a lactating female was taken at this place on the following day. Four males taken in the vicinity of Vista Hermosa in late June had testes that measured 8, 13, 20, and 22 in length. Also a juvenile was taken at this same time. Heteromys anomalus. — Rood (1963:189) reported that two wild-caught fe¬ males from Venezuela gave birth to young on 2 April (one young) and 17 May (three young). He found that a majority of females trapped at Rancho Grande, Venezuela, in April and May were either lactating or pregnant, whereas those caught in March were not. However, Rood opined that some breeding occurred in winter because immature individuals, probably less than two months old, were taken in April. Rood and Test (1968:94) found no males of H. anomalus with completely retracted testes in the period 24 April to 30 May, although some testes were regressing toward the end of this period. Heteromys australis. — The only reproductive data that I have available for this species are the presence of juvenile specimens in the Darien of Panama on 6 June, 27 June, and 19 July. Heteromys oresterus. — Goodwin (1946:373) in his report on Costa Rican mammals stated that Heteromys oresterus produced from three to five young per litter with four being the usual number. Conclusions. — The results of pairwise comparisons of the embryo counts for four species of Liomys (L. spectabilis not included) and Heteromys des- marestianus are given in Table 46. Liomys irroratus was found to have a signifi¬ cantly higher number of embryos per litter than the other four species (three Liomys and one Heteromys). The number of embryos found in specimens of Liomys pictus was significantly more than in Heteromys desmarestianus. The count for pictus probably will be found to be significantly higher than in speci¬ mens of adspersus when more data are available for the latter species. There was no significant difference between pictus and salvini , although pictus average a slightly higher embryo count. No significant difference was found between the embryo counts of Liomys salvini , Liomys adspersus, and Heteromys des¬ marestianus. Whether the difference in embryo counts between these species is a reflection of interspecific genetic difference or some ecological difference cannot be stated at the present time. GENOWAYS— SYSTEMATICS OF LIOMYS 309 Table 46. Pairwise comparisons of embryo counts for four species of Liomys and one of Heteromys using Wilcoxon two-sample test ( Sokal and Rohlf 1969:392-394). Liomys irroratus Liomys pictus Liomys salvini Liomys adspersus Heteromys desmarestianus Liomys irroratus (52) 4.39 Liomys pictus (49) * 3.80 Liomys salvini (33) *** ns 3.55 Liomys adspersus (5) * ns ns 3.20 Heteromys desmarestianus (7) * ** ns ns 3.28 From the available data, Liomys irroratus and Liomys salvini breed throughout the year. L. pictus appears to reproduce at most times of the year, although repro¬ ductive activity seems to be at a low level or to have ceased in the period from mid-December through January (see Fleming, 1971:67). Fleming (1970:478) believed Heteromys desmarestianus bred at all times of the year in Panama. The only species for which extensive data are available that appears to be a seasonal breeder is Liomys adspersus in which Fleming (1969, 1971:65-69) found preg¬ nant or lactating females only in the period from December through May. Wagner (1961:209) has suggested that Heteromys goldmani may breed only from May through December, but he supplied no supportive data. Why one species of Liomys should be a seasonal breeder while the other species of the genus and Heteromys desmarestianus have some individuals breeding at most seasons of the year is not readily apparent at the present time. It should be pointed out that these comments are based on data for the species from throughout their geo¬ graphic ranges, which is the best that can be done at the present time. However, when more data become available, geographic variation and variation from one year to another may be expected in the timing of the reproductive season within a species. Pterygoid Structure The shape of the pterygoid bones of specimens of Liomys irroratus differs from that of all other species studied (Fig. 61). In Liomys and Heteromys the pterygoid bones extend ventrally and then turn laterally. However, in specimens of Liomys irroratus the bone extends laterally noticeably farther than in other species, giving the pterygoids a teardrop shape when viewed ventrally. In all other species this lateral expansion is minor and the bone is straight in ap¬ pearance when viewed ventrally. Specimens of Liomys adspersus probably have the broadest pterygoids of the remaining four species of Liomys , and some specimens of Heteromys have the tips of the bones expanded, but both are dis¬ tinguishable from irroratus. Shape of the pterygoids has proven to be a useful 310 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY Fig. 61. — Pterygoid structure of five species of Liomys and one of Heteromys illustrating the breadth of the wings of the pterygoids and the shape of the interpterygoid fossa. The species are as follows: A, Liomys irroratus (KU 38373); B, Liomys pictus (KU 96067); C, Liomys spectabilis (KU 96045); D, Liomys salvini (KU 68847); E, Liomys adspersus (KU 121538); F, Heteromys desmarestianus (KU 95036). The scale at the right is three millimeters long. characteristic for separating cleaned crania of Liomys irroratus from crania of Liomys pictus in areas of potential sympatry between the species. The shape of the interpterygoid fossa is rather variable within the species of Liomys-, however, in all it is somewhat U-shaped anteriorly (Fig, 61). In speci¬ mens of Heteromys, on the other hand, the interpterygoid fossa narrows ante¬ riorly into an acute point, giving the fossa a V-shape (see also Goldman, 1911:15). Pelage Variation in the coloration of the middorsal pelage was analyzed with a Bausch and Lomb Spectronic 505 recording spectrophotometer equipped with a visible reflectance attachment (see Bowers, 1956; Selander et al., 1964; Johnston, 1966; Selander and Johnston, 1967, for discussion of colorimetric techniques). Flatness of the 100 per cent line was maintained within limits of 0.5 per cent peak-to-peak, and flatness of the 0 line within limits of 0.25 per cent. Wave¬ length accuracy of the Spectronic 505 is 0.5 millimicrons, and repeatability, 0.2 millimicrons. White standards of 100 per cent reflectance were prepared from magnesium sulfate. From curves of percentage reflectance in the range of wavelengths from 400 to 700 millimicrons, trichromatic coefficients (x, y, z) were derived by the 10 selected ordinate method, which involves a total of 30 readings from each curve. This was done using the trichromatic coefficient computing chart for Illuminant C produced by Bausch and Lomb (catalog no. 33-28-13). From the trichromatic coefficients, the psycho-physical descriptions of the colors were computed in terms of brightness (in per cent), dominant wavelength, and per cent excitation purity (see Judd, 1933). Brightness is simply the comparison of the sum of all the wavelength reflectances with the sum of those reflectances from the standard block and expressed as the per cent of the standard reflectance. Dominant wave¬ length refers to the hue of the sample and purity refers to the percentage of pure color reflected by the sample. The conversion of the trichromatic coefficients to the psycho-physical descriptions was done with the use of a computer program written in Fortran IV by S. A. Rohwer at The University of Kansas. Specimens used for color determination were carefully selected whenever pos¬ sible so as to represent total variation in the species taxonomically, ecologically. GENOWAYS — SYSTEMATICS OF LIOMYS 311 Table 47. Doisal color of five species o/Liomys and two o/ Heteromys. Statistics given are mean, two standard errors of the mean, and range. Species N Brightness Dominant wavelength Purity Liomys irroratus 10 9.36 ± 1.036 (7.50-11.60) 583.0 ± 1.34 (579.5-587.1) 20.86 ± 2.366 (13.58-25.31) Liomys pictus 10 9.57 ± 0.834 (7.25-1 1.25) 583.3 ± 0.50 (582.1-584.3) 26.37 ± 3.686 (15.06-32.32) Liomys spectabilis 10 8.54 ± 0.514 (7.60-10.30) 583.8 ± 1.22 (581.8-588.5) 27.69 ± 30.32 (15.58-34.44) Liomys salvini 10 7.58 ± 1.032 (5.20-1 1.20) 582.9 ± 1.10 (580.4-585.0) 20.58 ± 1.798 (15.49-24.46) Liomys adspersus 5 7.86 ± 0.720 (6.55-8.55) 581.6 ± 1.65 (579.4-583.2) 21.12 ± 2.090 (18.18-24.40) Heteromys desmarestianus 10 6.56 ± 0.322 (5.75-7.35) 584.3 ± 1.48 (580.4-585.3) 20.40 ± 2.280 (15.38-28.40) Heteromys gaumeri 5 8.21 ± 0.302 (7.65-8.55) 582.5 ± 1.16 (581.5-584.3) 23.47 ± 4.156 (19.48-29.09) and geographically. Therefore, values for each species in Table 47 should fairly closely represent the range of variation for the species as a whole. Coloration of dorsum. — The dominant wavelength or hue of the dorsal pelage did not differ to any great extent among the seven species tested. Specimens of Heteromys desmarestianus, with an average dominant wavelength of 584.3 milli¬ microns, revealed the most redness and specimens of Liomys adspersus, with an average dominant wavelength of 581.6 millimicrons, showed the least. The means for the remaining five species fell between 582.5 and 583.8 millimicrons and, considering the total range of variation for all species, there appear to be no significant differences based on dominant wavelength. However, specimens of Liomys pictus and Liomys spectabilis exhibited a much higher per cent purity than the other species, having mean values of 26.37 and 27.69, respectively. The per cent purity for specimens of Heteromys gaumeri with a mean value of 23.47 fell between those taxa with high values and the remaining four species with mean values ranging from 20.40 and 21.12. Specimens of Liomys pictus and Liomys irroratus were the palest of those tested (mean per cent brightness values 9.57 and 9.36, respectively). Specimens of Heteromys gaumeri and Liomys spectabilis had nearly the same mean per cent brightness (8.21 and 8.54, respectively) as was true for specimens of L. salvini and L. adspersus (7.58 and 7.86, respectively). Specimens of Heteromys desmarestianus 312 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY with a mean per cent brightness of 6.56 were the darkest of the seven species under consideration. Typical Ridgway (1912) color designation for the overall dorsal coloration of these seven species are as follows: between Chaetura Drab and Clove Brown ( Liomys irroratus)-, Dresden Brown to Mummy Brown ( Liomys pictus and Liomys spectabilis)-, Clove Brown to Dark Olive ( Liomys salvini and Liomys adspersus)-, near Raw Umber ( Heteromys desmarestianus)-, Saccardo’s Umber to Sepia ( Heteromys gaumeri). Characteristics. — A lateral stripe separating the dorsal coloration from the white belly pelage is absent in specimens of Liomys salvini and Liomys adspersus. In the other three species of Liomys ( irroratus , pictus, spectabilis ) and two species of Heteromys ( desmarestianus and gaumeri ) that I have examined, there is in nearly all specimens a stripe of differently colored pelage that separates the dorsal coloration from that of the venter, although the stripe may vary in color and width. In specimens of Liomys irroratus, the lateral stripe is generally a buffy to a pale pinkish color and usually is relatively narrow (in some specimens of L. i. torridus it may be absent). Never is the stripe in irroratus as strikingly notice¬ able as the ochraceous lateral stripe found in specimens of Liomys pictus and Liomys spectabilis. The lateral stripe in specimens of Heteromys gaumeri is much the same color as in pictus and spectabilis, but in some specimens the stripe is noticeably broad, unmatched in any of the other species under considera¬ tion, whereas in other specimens the stripe is as narrow as in pictus or spectabilis. Specimens of Heteromys desmarestianus have a narrow buffy lateral stripe that is generally faint. On specimens of Liomys salvini and Liomys adspersus, the terminal one-third to one-half of each slender dorsal hair is curved upward so that it stands out conspicuously beyond the spines. A museum study skin of one of these species when viewed laterally has a fuzzy appearance. This characteristic is not seen in the other three species of Liomys or in Heteromys desmarestianus and H. gaumeri. Specimens examined for pelage coloration and characteristics are listed below. Liomys irroratus. — San Gregorio Altapulco, Distrito Federal, 1 (KU 28052); 5 Vi km. N Agua del Obispo, Guerrero, 1 (KU 99783); Omilteme, Guerrero, 1 (KU 98761); 2.6 mi. E Etzatlan, Jalisco, 1 (KU 96019); Wi mi. N Zaragoza, Nuevo Leon, 1 (KU 98635); 4 mi. ESE Oaxaca, Oaxaca, 1 (KU 62633); 1 mi. SSW Tilapa, Puebla, 1 (KU 62623); 10 mi. NE San Luis Potosi, San Luis PotosI, 1 (KU 39898); Soto la Marina, Tamaulipas, 1 (KU 55880); 4 km. W Tlapacoyan, Veracruz, 1 (KU 24010). Liomys pictus. — Los Sabinos, Guerrero, 1 (KU 28458); 0.9 mi. NE Contla, Jalisco, 1 (KU 96033); San Sebastian, Jalisco, 1 (KU 1 12277); 7.3 mi. ESE Amatlan de Canas, Nayarit, 1 (KU 98942); 10 mi. S Juchatengo, Oaxaca, 1 (KU 98775); 20 mi. S, 5 mi. E Sola de Vega, Oaxaca, 1 (KU 98768); 10 mi. W, 2 mi. N Tehuantepec, Oaxaca, 1 (KU 54698); 5 mi. NW Mazatlan, Sinaloa, 1 (KU 85792); Alamos, 54 km. E Navajoa, Sonora, 1 (KU 89461); Puente Nacional, Veracruz, 1 (KU 19371). Liomys spectabilis. — 8 km. N Contla, Jalisco, 3 (KU 96035-37); 3.3 mi. NE Contla, Jalis¬ co, 4 (KU 96042-45); 2.2 mi. NE Contla, Jalisco, 3 (KU 96050-52). Liomys salvini. — 22 km. S San Jose, Costa Rica, 1 (KU 39252); 1 mi. NW San Salvador, El Salvador, 1 (KU 71079); 3 mi. E Jocotan, Guatemala, 1 (KU 84125); 1 mi. SE Monogoy, Guatemala, 1 (KU 71068); 16 km. SW Tapachula, Chiapas, 1 (KU 120602); 6 mi. NW Tonala, Chiapas, 1 (KU 68839); Hda. Beliavista, Nicaragua, 1 (KU 106378); La Danta, 1 GENOWAYS— SYSTEM ATICS OF LIOMYS 3j3 km. E Esquipulas, Nicaragua, I (KU 1 15352); 4 mi. W Managua, Nicaragua, I (KU 71139); Santa Rosa, 17 km. N, 15 km. E Boaco, Nicaragua, 1 (KU 1 10403). Liomys adspersus. 2 km. NE Buenos Aires, Panama, 4 (KU 121527-30); Las Cumbres, 7 km. N, 2 km. W Pueblo Nuevo, Panama, 1 (KU 121535). Heterc>mys desmarestianus. — 1 mi. SW Santiago Sacatepequez, Guatemala, 1 (KU 71256); VM1'^co,antia80 SacateP^uez’ Guatemala, 1 (KU 71247); Finca El Paraiso, Chiapas, 1 (KU 66583); 8 mi. (by road) NW Pueblo Nuevo, Chiapas, 1 (KU 95034); 7.5 mi. (by road) ^ W Pueblo Nuevo, Chiapas, 2 (KU 95037-38); Sabana de Quintal, Chiapas, 1 (KU 102653); 2 km. N, 6 km. E Esquipulas, Nicaragua, 2 (KU 115410-11); Santa Maria de Ostuma, Nic¬ aragua, 1 (KU 106523). Heteromys gaumeri, — 3 km. N Piste, Yucatan, 3 (KU 92123-24, 92128); 2 km N Piste Yucatan, 1 (KU 92130); Piste, Yucatan, 1 (KU 92133). Distribution Although the distribution and ecology of the various species of Liomys were discussed in the systematic accounts, I have drawn this information together here for comparative purposes. Only a brief summary of the information from the systematic accounts is given below. Liomys irroratus. — This species occurs on the Mexican Plateau and adjacent areas, from southern Texas and Chihuahua in the north to the Sierra Madre del Sur of Oaxaca in the south. The vegetation in the areas where irroratus occurs is mainly steppe, thicket, and scrub desert and montane formation series (Wagner, 1964:223); the montane formation series contains xerophytic plants, mainly pine and oak. Liomys irroratus occurs sympatrically only with Liomys pictus , although it approaches the range of Liomys spectabilis. In those areas where irroratus and pictus are sympatric, which is in a band from central Jalisco to the mountains of the Sierra Madre del Sur of Oaxaca, irroratus is generally found in the drier areas and pictus in the moister areas. Generally, therefore, they are microallopatric in these areas. Liomys pictus. — This species is known along the western coast of Mexico from Sonora to northwestern Chiapas, in the central valley of Chiapas, and in southern Veracruz. Liomys pictus occurs most often in seasonal formation series type of vegetation, mainly lowland dry forest, although some specimens are known from the xerophytic montane formation (Wagner, 1964). Liomys pictus occurs sympatrically with Liomys irroratus (see above), Liomys salvini (see below), and Liomys spectabilis (see below). Liomys spectabilis. — This species is known only from the relatively moist interior basins of southeastern Jalisco. Because the area has been significantly altered by agricultural practices, it is difficult to determine the exact habitat occupied by the species. At three locations pictus and spectabilis were obtained in the same trapline in such habitats as the edge of a cornfield and around clumps of brush in a heavily grazed pasture. Liomys salvini. — Occurring along the Pacific coast and adjacent mountains of Central America from southeastern Oaxaca to central Costa Rica, L. salvini also is known from dry river valleys of the Caribbean drainage such as along the Rio Motagua in Guatemala. Vegetation of these areas is generally lowland dry forest and lower montane dry forest, which is largely pine-oak (Stuart, 1966). 314 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY This species is known to occur sympatrically with Liomys p ictus, the area of sympatry being in southeastern Oaxaca and northwestern Chiapas. No data are available concerning the ecological condition under which the two species have been taken in the area of sympatry, but males of the two species do show character displacement there. Liomys adspersus. — This geographically restricted species is known only from Pacific lowlands of central Panama and the headwaters of the Caribbean drainage in the vicinity of the Panama Canal; this is an area of savanna. The nearest species geographically and phylogenetically is L. salvini, which occurs as far south as central Costa Rica, but the two species are separated by the low¬ land rainforest occurring around the Golfo Dulce of Costa Rica and western Panama. Heteromys. — Most of the species of Heteromys (except gaumeri ) occurring in Mexico and Central America inhabit areas of lowland rainforest and quasi¬ rainforest, lower montane wet forest and cloud forest, and to some extent the lower montane dry forest (Stuart, 1966). However, even in the latter areas they are found in very mesic situations. At only a few places have Heteromys and Liomys been taken together (Hda. Tepeyac, Nicaragua, for example); when more precise data are available, it likely will be found that microallopatry is the case (even at Hda. Tepeyac), as in the vicinity of Esquipulas, Nicaragua, where H. desmarestianus occurs in the cooler, moister uplands and Liomys salvini in the hotter, drier lowlands. The one exception to Heteromys occurring in mesic situa¬ tions is H. gaumeri, which occurs in lowland dry forest and thorn forest on the Yucatan Peninsula (Stuart, 1966) in the absence of Liomys. Other Studies Several studies by other researchers have dealt with Liomys and Heteromys and should be mentioned; these could not be used for analysis of relationships, however, because some species were not included. Eisenberg (1963 a, 1963 b, 1967) has presented some interesting studies on the behavior of heteromyid rodents, which included Liomys pictus, Heteromys anomalus, H. desmarestianus, and H. lepturus. Behavioral studies such as these and other types such as huddling experiments certainly would be useful in understanding the relationships of members of the genera Liomys and Heteromys, especially in areas of actual or potential sympatry. Two studies of anatomy, one by Hatt (1932) on post-cranial skeletal anatomy and the other by Quay (1965) on sebaceous and suboriferous glands, have included comments on some species of Liomys and Heteromys. The study by Hudson and Rummel (1966) on water metabolism and temperature regulation of Liomys salvini and Liomys irroratus is the only physiological work on members of the genus of which I am aware. A great deal still remains to be learned about the biology and systematics of species of Liomys and Heteromys. One important aspect of the biology of these animals has been virtually unstudied — their ecology, where only the work of Fleming on Liomys adspersus is the exception. Good autecological studies of any other species would be of real value, but of even more interest would be GENOWAYS— SYSTEMATICS OF LIOMYS 3|S studies of species in areas of actual or potential sympatry. Three areas in which the ecological relationships of Liomys irroratus and L. pictus could be success- u ly studied are in 1 ) central Jalisco in the vicinity of Guadalajara; 2) to the north ot Iguala, Guerrero, southward through the Balsas basin to Chilpancingo and on to Acapulco; and 3) southward from Sola de Vega, Oaxaca, over the Sierra Madre del Sur to Puerto Angel. The ecological relationships of Liomys pictus to L. salvini would need to be studied in the area from Reforma, Oaxaca, to Tonala, Chiapas. The ecological preferences of members of genus Liomys could be com¬ pared with those of Heteromys in a place such as Esquipulas, Nicaragua, where salvini occurs at lower elevations and desmarestianus at higher elevations. Studies of ecological physiology would be extremely useful in understanding habitat pref¬ erences of the species. More systematic work in such areas as serology and immunology still remains to be done. A much more extensive survey of karyological differences among the species will need to be done before these characteristics can be fully understood. Many anatomical systems, such as muscular and gastric, have not been studied. It is my hope that the present study has resulted in a workable systematic ar¬ rangement of members of the genus Liomys so that studies such as those de¬ scribed above can be carried out in a more meaningful way. 316 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY EVOLUTIONARY AND ZOOGEOGRAPHIC RELATIONSHIPS Removal of “Liomys” centralis from the genus limits the known fossil material of Liomys to late Pleistocene and sub-Recent cave deposits, thereby eliminating the possibility of deducing evolutionary relationships within the genus based up¬ on the paleontological record. Therefore, in order to obtain an objective estimate of the evolutionary relationship among the five species of Liomys and their rela¬ tionship with Heteromys (characters used in this analysis based upon desmares- tianus, gaumeri, and lepturus but mainly the first), I have employed cluster analyses based upon both correlation and distance matrices and a principal com¬ ponents analysis, both of which are part of the NT-SYS programs, and a Prim Network. Characters used for analysis and methods for scoring them are given in Appendices III and IV, respectively; these are mainly characteristics analyzed and discussed in the previous section. These characters were used in the NT-SYS programs to obtain a representation of the phenetic relationships of the species. The Prim Network gives an estimation of the cladistic and patristic centers for the group of OTU’s (species) being analyzed and thereby provides a useful and objective framework for inference of primitive character states. Fifty-three characters were contained in the original matrix for analyses (see Appendix III). However, those characteristics dealing with ectoparasite faunas and morphology of spermatozoa had to be eliminated because data were not avail¬ able for all species being analyzed. The remaining characteristics were submitted to a character correlation analysis and those characters found to be highly corre¬ lated with others in the matrix were eliminated. Most characters eliminated ap¬ peared to be measures of size (the only measure of general size remaining in the matrix is greatest length of skull). The final matrix used in all analyses in this section, therefore, contained only 30 characters (marked by an asterisk in Ap¬ pendix III). Examination of the correlation matrix (Table 48) and correlation pheno- gram (Fig. 62) based upon this matrix (coefficient of cophenetic correlation is 0.955) reveals that Liomys pictus and Liomys spectabilis are the most highly cor¬ related species. The next most highly correlated pair of species and the only other pair that possesses a positive correlation coefficient is Liomys salvini and Liomys adspersus. Liomys irroratus exhibits only negative correlation coefficients with the other five species included in the analysis; based upon these values it appears that irroratus is only slightly closer to L. pictus and L. salvini than to Heteromys. The characteristics of irroratus show less correlation with those of spectabilis than with these three taxa and the least correlation with those of adspersus. He¬ teromys is relatively far removed from species of Liomys in the matrix, being closest to L. irroratus and farthest from L. pictus-, it is grouped with irroratus in the phenogram, but it must be remembered that they are negatively correlated. Liomys pictus and L. spectabilis show little correlation with L. salvini and L. ad¬ spersus with all correlation coefficients between the species being over a —0.3. The distance matrix (Table 48) and distance phenogram (Fig. 62) based upon this matrix (coefficient of cophenetic correlation 0.985) reveal nearly the same pattern as their correlation counterparts. In the phenogram, the five species of GENOWAYS— SYSTEMATICS OF LIOMYS 317 Correlation and distance matrices based upon 30 characters comparing the five species of Liomys and a composite of three species o/Heteromys, mostly desmarestianus. Liomys Liomys Liomys Liomys Liomys Heteromys irroratus pictus spectabilis salvini adspersus species Liomys irroratus 1.000 Liomys pictus -0.119 1.000 Liomys spectabilis -0.220 0.713 1.000 Liomys salvini -0.115 -0.350 -0.432 1.000 Liomys adspersus -0.301 -0.335 -0.363 0.442 1.000 Heteromys sp. -0.152 -0.525 -0.404 -0.307 -0.295 1.000 Liomys irroratus 0.000 Liomys pictus 1.187 0.000 Liomys spectabilis 1.257 0.594 0.000 Liomys salvini 1.160 1.249 1.306 0.000 Liomys adspersus 1.348 1.335 1.367 0.855 0.000 Heteromys sp. 1.679 1.864 1.812 1.740 1.800 0.000 Liomys form a cluster separate from Heteromys-, examination of the distance matrix reveals that all species of Liomys are approximately equidistant from Heteromys, with Liomys irroratus being phenetically the closest with a distance coefficient of 1.679 and Liomys pictus being the most distant with a coefficient of 1.864. Within the species of Liomys on the phenogram, L. salvini and L. ad¬ spersus form a cluster as do L. pictus and L. spectabilis, whereas L. irroratus forms a cluster by itself. From the phenogram, it appears that irroratus is slight¬ ly closer to the pictus- spectabilis group than the salvini-adspersus group; how¬ ever, values in the matrix show salvini to be closest to irroratus, but only slight¬ ly more so than pictus. L. spectabilis is more distant from irroratus than the latter two species and adspersus is the most distant of all the species within the genus. The pictus-spectabilis group and salvini-adspersus group are phenetically dis¬ tant from each other with the lowest distant coefficient between the members be¬ ing 1 .249. The lowest phenetic distance coefficient between a pair of species is 0.594 between pictus and spectabilis and the next lowest is 0.855 between sal¬ vini and adspersus. In the principal components analysis, 100 per cent of the phenetic variation based upon the matrix of correlation among the 30 characters was explained by the first five components as follows: component I, 44.79; component II, 28.75; component III, 16.47; component IV, 6.65; component V, 3.34. Results of the factor analysis are given in Table 49 and two-dimensional plots of component I against components II to V are given in Fig. 63. Along the first principal com¬ ponent Heteromys is widely separated from the five species of Liomys and characters with high weighting in this component are those that tend to dis¬ tinguish the two genera. Among the highly weighted characters were: 1 9, presence or absence of accessory enamel islands on the molars (positive value); 20, number 318 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY _ HOMY S IRRORATUS _ _ HETEROMYS _ LIOMYS SALVINI — _ LIOMYS ADSPERSUS _ LIOMYS PICTUS _ LIOMYS SPECTABI L IS I - 1 - 1 _ i _ I _ i -0.40 -0.15 0.10 0.35 0.60 0.85 _ LIOMYS IRRORATUS LIOMYS PICTUS LIOMYS SPECTABILIS _ LIOMYS SALVINI _ LIOMYS ADSPERSUS HETEROMYS i - 1 - 1 _ i _ i _ i 1.80 1.55 1.30 1 05 0.80 0.55 Fig. 62. — Correlation (upper) and distance (lower) phenograms for five species of Liomys and a composite of values for Heteromys computed from correlation and distance matrices, respectively, based on standardized characters and clustered by unweighted pair-group method using arithmetic averages. The cophenetic correlation coefficient of the correlation phenogram is 0.955 and for the distance phenogram is 0.985. of lophids on the lower premolars (negative); 22, position of entostyle on upper premolar (negative); 34, length of baculum divided by length of hind foot times 100 (positive); 45, shape of interpterygoid fossa (negative); 47, soles of hind feet (negative); and 49, ecological relationships (negative). Other characters receiving relatively high values but not as high as the previous set were: 1, greatest length of skull (negative); 14, condition of the posterior margin of the interparietal bone (negative); 23, length of P4 divided by greatest length of skull times 100 (posi¬ tive); 30, length of glans divided by length of baculum times 100 (negative); 32, morphology of baculum (positive); and 52, per cent brightness of dorsal pelage (positive). It should be noted that along this component, which tends to separate GENOWAYS — SYSTEMATICS OF LIOMYS 319 Table 49. Factor matrix from correlation among 30 characters of the five species of L.omys and a composite of three species of Heteromys, mostly desmarestianus. Character 1 14 15 16 17 18 19 20 21 22 23 27 28 30 31 32 33 34 35 36 37 45 46 47 48 49 50 51 52 53 Component 1 -0.731 -0.701 0.497 -0.688 -0.418 0.665 0.969 -0.969 0.665 -0.840 0.745 0.542 -0.204 -0.733 0.341 0.731 0.094 0.943 -0.679 -0.419 -0.531 -0.969 -0.039 -0.969 0.422 -0.969 -0.204 -0.489 0.748 0.568 Component II 0.020 0.510 -0.353 -0.428 0.665 -0.735 -0.177 0.177 -0.735 -0.542 0.153 -0.635 0.931 -0.239 0.874 0.343 0.155 0.189 -0.513 -0.758 -0.763 0.177 -0.036 0.177 0.481 0.177 0.931 0.767 0.380 0.735 Component III 0.230 -0.142 -0.377 0.046 0.357 0.095 -0.165 0.165 0.095 -0.018 -0.616 0.521 -0.269 -0.623 0.241 0.576 0.639 0.268 -0.515 0.442 -0.364 0.165 0.966 0.165 -0.739 0.165 -0.269 0.003 -0.431 0.361 Component IV 0.586 -0.450 0.362 0.584 -0.060 -0.082 0.020 -0.020 -0.082 -0.028 -0.074 -0.179 0.136 -0.131 0.246 -0.124 0.688 0.013 0.096 -0.223 0.040 -0.020 -0.242 -0.020 -0.068 -0.020 0.136 -0.392 0.198 0.015 Component V 0.265 0.157 -0.596 0.012 -0.502 0.049 0.054 -0.054 0.049 -0.004 0.190 0.008 -0.028 -0.017 -0.031 -0.033 0.293 0.059 0.038 -0.070 0.042 -0.054 -0.073 -0.054 0.201 -0.054 -0.028 0.137 -0.268 0.084 the two genera, Liomys irroratus, although far removed from Heteromys , is the species nearest to it and that Liomys pictus is farthest from Heteromys. The second principal component, which accounts for more than a quarter of the total phenetic variation, appears to distinguish the species groups within the genus Liomys. The three characters with the highest loading (all positive) in this component were 28 (number of lobes on urethral lappets), 31 (length of tip of glans penis divided by total length of glans times 100), and 50 (presence or absence of lateral stripe). Other characters with relatively high positive loading in 320 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY I I Fig. 63. — Two-dimensional projections of the first five principal components illustrating the phenetic position of the five species of Liomys and a composite of values for Heteromys. Superimposed upon the projection of component I plotted against component II is the least interconnected network based upon the distance matrix. Symbols used are as follows: A, Liomys adspersus\ H, Heteromys', I, Liomys irroratus', P, Liomys pictus', S, Liomys salvini', SP, Liomys spectabilis. this component were 51 (dominant wavelength of dorsal coloration) and 53 (per cent purity of dorsal coloration). Characters with high negative values in this component were 18 (wear pattern of molars), 21 (re-entrant angle on labial margin of metalophid), 36 (fundamental number of chromosomes), and 37 (type of X-chromosome). In this component, Liomys pictus and Liomys spectabilis form a group at the top of the plot and Liomys salvini and Liomys adspersus form another group at the bottom of the plot. Liomys irroratus occupies an interme¬ diate position between these two groups. It should be noted that although Heteromys is widely separated from irroratus in the first component, it is near that species in the second component. Superimposed upon the plot of component I against II (Fig. 63) is the least interconnected network of species based upon the distance matrix. The network shows Heteromys connected only to Liomys irroratus. Liomys irroratus, aside from being connected with Heteromys , also is connected with Liomys pictus and GENOWAYS— SYSTEMATICS OF LIOMYS 321 LIOMYS ADSPERSUS 14 39 LIOMYS SALVINI 25 89 HETEROMYS 4108 _ LIOMYS IRRORATUS 15 63 LIOMYS PICTUS 9 63 LIOMYS SPECTABILIS Fio. 64. A Prim Network illustrating relationships between the five species of Liomys and a composite ot values for Heteromys. Characters used and methods for scoring them are given in appendices III and IV, respectively. Liomys salvini. L. pictus in turn is connected with Liomys spectabilis and L. sal¬ vini with L. adspersus. In the third component, which accounts for slightly more than 15 per cent of the total phenetic variation, the characters with the high loading were those that distinguished L. irroratus from the other species, especially of Liomys. Characters with high positive loading were 27 (amount of sculpturing on glans penis), 32 (morphology of baculum), 33 (length of baculum), and 46 (number of plantar tubercles); characters with high negative loading were 23 (length of P4 divided by greatest length of skull times 100), 30 (length of glans divided by length of baculum times 100), 35 (number of chromosomes), and 48 (reproductive rate). Characters 46 and 48 had the highest loading values. In component III, Liomys salvini is the species nearest to irroratus, which is located at the bottom of the plot, and Liomys adspersus is the species farthest from irroratus. The fourth component, which accounts for less than seven per cent of the total phenetic variation, is evidently heavily influenced by those characters that distin¬ guish Liomys salvini and Liomys adspersus, which are located at opposite edges of the plot with the other species located near the center. Characters with high positive loading in this component were 1 (greatest length of skull), 16 (shape of termination of nasal bones), 31 (length of tip of glans penis divided by length of glans times 100), and 33 (length of baculum), whereas 14 (condition of posterior margin of interparietal bone), 36 (fundamental number of chromosomes), 46 (number of plantar tubercles), and 51 (dominant wavelength of dorsal color¬ ation) had high negative values. Generally for factor component IV and V loading values were much lower than for the first three components. The fifth component, which accounts for less than four per cent of the total phenetic variation, is generally influenced by those characters that distinguish Liomys pictus and Liomys spectabilis, which are located at opposite edges of the plot with the other species clustered near the middle. Characters with high posi¬ tive values for this component were 1 (greatest length of skull), 33 (length of baculum), and 48 (reproductive rate) and those with the highest negative values were 15 (division of interparietal), 17 (length of nasals in comparison with pre- maxillaries), and 52 (per cent brightness of dorsal coloration). The Prim Network shows Liomys irroratus at the cladistic and patristic 322 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY center of the six species under study (Fig. 64). It seems plausible (see also below) that L. irroratus occupies such a position within the genus Liomys and that the other members of the genus were derived from a stock similar to it. However, based upon study of the paleontological record presented by Wood (1931, 1935), there may have been separate origins of the genera Liomys and Heteromys from the early heteromyine stock ( Peridiomys-Proheteromys complex). This would mean that on the Prim Network the cladistic and patristic centers for the two genera would lie somewhere between Liomys irroratus and the representatives of Heteromys studied. If Liomys irroratus is the living species most nearly resembling the ancestral form of the genus, it would be expected to share more characters with the “sister- group” Heteromys than would other Liomys-, characters shared between the genera can be presumed to have probably been present in their common ancestor. Liomys irroratus and Heteromys lepturus exhibit similarities in the structure of the glans penis such as the amount of sculpturing, number of urethral lappets, ra¬ tio of length of glans to length of baculum, and amount of tip of glans that is ex¬ posed. The morphology of the baculum of Liomys irroratus and Heteromys des- marestianus and H. gaumeri is nearly identical. Both L. irroratus and H. des- marestianus have a chromosome number of 60 and a large submetacentric X- chromosome (the latter is also true of L. adspersus and L. salvini). A lateral stripe is present in most specimens of L. irroratus and H. desmarestianus as well as Liomys p ictus and Liomys spectabilis. It seems most likely to me (in view of what is known about Perognathus ) that the spermatozoa of Heteromys will be found to be similar to those of irroratus, pictus, and spectabilis and unlike those of salvini and adspersus. The upper premolar of Liomys salvini and Liomys ad¬ spersus more closely resembles those of Heteromys than do other species of Liomys, but some differences also were noted. These two species of Liomys have a high fundamental number, which is near the value for Heteromys desmare¬ stianus. The anopluran louse, Fahrenholzia fairchildi, has been reported from L. salvini, L. adspersus, and H. desmarestianus, but it seems probable to me that this louse evolved on the salvini-adspersus group and has recently spread to des¬ marestianus. All species of Liomys, except irroratus, and Heteromys desmare¬ stianus have six plantar tubercles; L. irroratus has only five plantar tubercles, which I believe is the derived condition, six being the primitive number. A number of characters serve to distinguish members of the genus Heteromys from Liomys-, probably the most important of these deal with the dentition. The lower premolars of Liomys are composed of but two lophids, whereas those of Heteromys are comprised of either three or four lophids. Accessory enamel is¬ lands are persistent in Heteromys but are found only in the youngest specimens of Liomys-, also the median valley on the molars is persistent into adulthood in most species of Heteromys, whereas it is worn away in Liomys before adulthood is reached. Other characters that are unique to Heteromys are presence of the anopluran lice species Fahrenholzia ferrisi, F. hertigi, and F. schwartzi, naked soles on hind feet (except H. gaumeri), V-shaped interpterygoid fossa, and occu¬ pancy of areas of moist forests in Central America and Mexico (except H. gaumeri). GENOWAYS— SYSTEMATICS OF LIOMYS 323 PERIDIOMYS- PROHETEROMYS COMPLEX Fig. 65.— Hypothetical representation of the phylogenetic relationships among the five species of Liomys and their relationship to the genus Heteromys. My concept of the phylogenetic relationships within the genus Liomys and their relationship to Heteromys based upon the preceding analyses is presented in Fig. 65. From early heteromyine stock of which the Proheteromys-Peridiomys complex is the most likely (see Wood, 1931, 1935), the genera Liomys and Heteromys had separate origins. The ancestor of the Liomys lineage probably most closely resembled the Recent species Liomys irroratus, but almost certainly differed from it in a number of ways. This basic stock underwent differentiation into three lines giving rise to the precursors of irroratus , the salvini-adspersus group, and the pictus-spectabilis group. If the degree of differentiation between the groups can be used as a gauge, it would appear that the three originated at approximately the same time. Within the salvini-adspersus and the pictus- spectabilis groups, there was a much more recent splitting to give rise to the Recent species. It seems most likely that these last differentiations occurred in relatively late Wisconsin time, but it is impossible to actually date any of these events because of the almost complete lack of a fossil record. The distribution patterns of the two groups that have recently undergone speciation suggest that these may have come about by what Brown (1957) termed centrifugal speciation. If this is true, adspersus and spectabilis would be expected to more nearly re- 324 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY semble the precursors of their respective groups, the species salvini and pictus, with their larger gene pool and consequently more opportunities for mutations and recombination, having essentially “evolved away from” this precursor. I have little evidence to present on either side of this question, but adspersus and spect- abilis both are large-sized species as were many Pleistocene taxa and their chro¬ mosome compliment is slightly less differentiated from the ancestral condition (if parsimony is envoked) than is that of pictus and salvini. On the other hand, in most of the analyses above, when all characters were considered, pictus and salvini, within their respective groups, revealed more similarities to irroratus than did either spectabilis or adspersus. Although Jackson and Crovello (1971) have pointed out pitfalls in using some of the numerical analyses employed here¬ in to deduce evolutionary relationships, these are presented along with my con¬ cept of phyletic relationships within the genus Liomys as the best estimate pos¬ sible at this time based upon the available data. However, if fossil material be¬ comes available in the future and additional biosystematic data are obtained, these results should be compared with the present model and, if necessary, the model modified in accord with new findings. Zoogeographically, Recent members of the genus Liomys occur mainly in the Neotropical Region and the transition zone between the Neotropical and Nearctic regions, but portions of the ranges of L. pictus and L. irroratus are included in the Nearctic Region (Hershkovitz, 1958, 1969; Hooper, 1949). Members of the genus Heteromys occur exclusively in the Neotropical Region. In the classifi¬ cation of major units of North American rodent fauna proposed by Hooper (1949), Liomys would be included in both the Sonoran and tropical faunas, whereas Heteromys would be placed only in the tropical fauna. Biotic provinces are much less useful in describing the geographic range of these two genera because individual species range through several provinces (for example Liomys irroratus is found in Chihuahua-Zacatecas, Sierra Madre Occidental, Tamaulipas, Sierra Madre Oriental, Veracruz, Transverse Volcanic, and Sierra Madre del Sur — see Goldman and Moore, 1946) and the provinces of Central America (Ryan, 1963) are poorly defined because the ranges of many species are at best ill defined. Essentially the picture that emerges is that members of the genus Liomys occur within, and just to the north of, the Neotropical Region and Heteromys is ex¬ clusively Neotropical in distribution (also see Baker, 1963). The next question that arises is what was the past zoogeographic history of these genera. Unfortun¬ ately, only an outline of this history is possible at present because the fossil record of these groups is poor and our knowledge of vegetational and geological history of Mexico and Central America is only now emerging. It is evident that the evolu¬ tionary history of the family Heteromyidae has been tied to the developing aridity in the Great Plains and Southwest during the Miocene and Pliocene (Wood, 1935:230-236). Members of the subfamily Heteromyinae are known from as far north as Oregon and Nebraska, and from Florida during this time. However, the evolution of Liomys and Heteromys probably occurred to the south in Mexico and Central America. GENOWAYS— SYSTEMATICS OF LIOMYS 325 The ancestors ot Liomys probably evolved on the southern Mexican Plateau in much the same areas as the southern part of the range of Recent Liomys irroratus (Fig. 66). Whether these animals evolved as the Madro-Tertiary Geo¬ flora developed in this area (Axelrod, 1958, 1960) or simply evolved in the already developed flora cannot be stated at this time, but the distribution of Liomys is closely correlated with this flora. On the other hand, I believe Hete¬ romys evolved in isolation from Liomys in Nuclear Central America (Lloyd, 1963, Stuart, 1966); members ot this genus were evidently undergoing their evolution in areas of the Neotropical-Tertiary Geoflora. The exact reasons for the shitt of Heteromys into a more mesic habitat than other members of the family can only be a matter of conjecture, but one reason may be that this was essentially the only habitat available in Nuclear Central America at the time. With the closing of the Panamanian Portal in the late Pliocene, Heteromys was able to enter northern South America (Hershkovitz, 1969:23). Sometime probably in the early Pleistocene the ancestor of the salv ini-adspersus group extended along the Pacific coast of Central America where the Madro- Tertiary Geoflora had developed as the result of increasing aridity caused by the uplift of Nuclear Central America beginning in the Pliocene (Fig. 66). Isolation of this group from other Liomys most likely was accomplished across the Isthmus of Tehuantepec, if conditions as suggested by Duellman (1960) occurred in the region during the Pleistocene. He believed that the Isthmus was never completely submerged during the Pleistocene, but that isolation was ac¬ complished by a combination of fluctuations in sea level and climatic conditions. During glacial periods the lowlands of the Isthmus were probably more exten¬ sive and had more semiarid environments than at present because of the depressed sea level; these times would allow passage of lowland animals, but would result in isolation of highland taxa. Interglacial periods, on the other hand, would have been characterized by warmer temperatures, higher sea level, and moister conditions across the portion of the Isthmus that remained emergent; in these times lowland kinds would have been isolated with highland faunas being able to cross the Isthmus. This set of conditions easily explains the passage of Liomys across the Isthmus to the south, their subsequent isolation and speciation (Fig. 66), and the fact that members of the genus Heteromys were able to pass through the Isthmus to the north. During a glacial period, probably the Wisconsin, members of the salv ini-adspersus group, because of the depressed sea level re¬ sulting in more extensive lowlands and arid conditions, were able to extend their Fig. 66. — Some possible events resulting in speciation in the genus Liomys. A _ Hypothetical distribution of precursors of Liomys in Mexico and Heteromys in Chiapas and Central America. B — Increased geographic range of precursor of Liomys across Isthmus of Tehuantepec in response to increasing aridity. C — Isolation of precursors of Liomys pictus- group in western Mexico, Liomys salv ini- group in Central America, and Liomys irroratus on the Mexican Plateau probably in response to increasing moist conditions and rising sea level in vicinity of Isthmus of Tehuantepec. D — Distribution of Recent species of Liomys. See text for additional discussion. — » 326 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY GENOWAYS— SYSTEMATICS OF LIOMYS 327 104 Fig. 66. — Continued. 328 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY range into the savannas of the Pacific coast of Panama. The return of warmer and moister conditions and the restriction of the lowland by rising sea levels in the subsequent interglacial has resulted in isolation and speciation giving rise to Liomys salvini in most of Central America and Liomys adspersus in the savannas ot Panama. These species are separated today by tropical rainforest and quasi- raintorest in the vicinity of the Golfo Dulce region of Costa Rica and western Panama (Wagner, 1964; Stuart, 1966). This distributional pattern is exhibited by other vertebrate groups, especially the herpetofauna (Duellman, 1966; Savage, 1966). About the same time the precursor of the salvini- adspersus group entered Cen¬ tral America, the ancestors of the p ictus- spectab ills group entered the Pacific Coast of Mexico (Fig. 66) in that region termed the North Subtropical Sector by Baker (1967). What served to isolate this group is not clear, but most likely it was the mountain masses, Sierra Madre Occidental and Sierra Madre del Sur, of western Mexico. Evidently spectabilis was isolated in interior drainage basins of Jalisco from the much more widespread pictus and subsequent to speciation pictus has reinvaded that area to become sympatric with spectabilis. The exact circumstances of this isolation and reinvasion are not clear at present, but prob¬ ably they were associated with climatic fluctuations in the late Wisconsin and post- Wisconsin times. Another possible explanation that should not be disre¬ garded is that spectabilis was isolated from pictus by volcanic activity associated with the Transverse Volcanic Belt. The geographic range of spectabilis lies just to the northeast of the Sierra Nevado de Colima and Volcan de Colima, and its iso¬ lation from pictus on the coast could have resulted from activity of these volcanos. At the time other species were undergoing differentiation in western Mexico and Central America, Liomys irroratus was evolving from the ancestral stock of the genus, more or less in situ, on the Mexican Plateau (Fig. 66). During my study of this group, two factors have seemed to me to be most im¬ portant in determining the distribution of Recent species of Liomys (Fig. 66). First, moisture appears to be a significant factor; as stated above several times, in areas of high moisture Liomys is replaced by members of the genus Heteromys. On the other hand, members of the genus are limited also by extreme aridity, because no species occur in areas receiving less than 250 millimeters of rain annually and most do not occur in areas receiving less than 500 millimeters of rain annually (data from Vivo Escoto, 1964). Second, the presence of a congeneric species appears to form a significant barrier to dispersal. This would account for the restricted areas of sympatry be¬ tween species and certain unusual patterns of distribution (such as only L. pictus occurring in the central valley of Chiapas where it appears to be a barrier to the entry of L. salvini). Essentially, the first species present would serve to block the entry of the second. A similar situation was noted by Russell (1968:758-760) among members of the closely related family Geomyidae. GENOWAYS— SYSTEMATICS OF LIOMYS 329 LITERATURE CITED Aldrich, J. W., and B. P. Bole, Jr. 1937. Introduction. Pp. 5-8, in The birds and mam¬ mals of the western slope of the Azuero Peninsula [Republic of Panama] (J. W. Aldrich and B. P. Bole, Jr., eds.), Sci. Publ. Cleveland Mus. Nat. Hist., 7:1-196. Allen, J. A. 1881. List ot mammals collected by Dr. Edward Palmer in north-eastern Mexico, with tield-notes by the collector. Bull. Mus. Comp. Zool., 8:183-189. 1891. Notes on new or little-known North American mammals, based on recent additions to the collection of mammals in the American Museum of Natural His¬ tory. Bull. Amer. Nat. Hist., 3: 263-310. - 1897. Further notes on mammals collected in Mexico by Dr. Audley C. Buller, with descriptions of new species. Bull. Amer. Mus. Nat. Hist., 9:47-58. - 1903. List of mammals collected by Mr. J. H. Batty in New Mexico and Durango, with descriptions of new species and subspecies. Bull. Amer. Mus. Nat Hist 14:587-612. - 1906. Mammals from the states of Sinaloa and Jalisco, Mexico, collected by J. H. Batty during 1904 and 1905. Bull. Amer. Mus. Nat. Hist., 22:191-262. - 1908. Mammals from Nicaragua. Bull. Amer. Mus. Nat. Hist., 24:647-670. Alston, E. R. 1879-82. Biologia Centrali-Americana, Mammalia, xx + 220 pp., 22 pis. (introduction by P. L. Sclater). Alvarez, T. 1963. The Recent mammals of Tamaulipas, Mexico. Univ. Kansas Publ., Mus. Nat. Hist., 14:363-473. - 1968. Notas sobre una coleccion de mamiferos de la region costera del Rio Balsas entre Michoacan y Guerrero. Rev. Soc. Mexicana Hist. Nat., 29:21-35. Atchley, W. R. 1970. A biosystematic study of the subgenus Selfia of Culicoides (Dip- tera: Ceratopogonidae). Univ. Kansas Sci. Bull., 49:181-336. Axelrod, D. I. 1958. Evolution of the Madro-Tertiary Geoflora. Bot. Rev., 24:433-509. - 1960. The evolution of flowering plants. Pp. 227-305, in Evolution after Darwin (S. Tax, ed.), Univ. Chicago Press, l:viii + 1-629. Bailey, V. 1905. Biological survey of Texas. N. Amer. Fauna, 25:1-222. Baker, R. H. 1954. The silky pocket mouse (Perognathus flavus) of Mexico. Univ. Kansas Publ., Mus. Nat. Hist., 7:339-347. - 1963. Geographical distribution of terrestrial mammals in Middle America. Amer. Midland Nat., 70:208-248. - 1967. Distribution of Recent mammals along the Pacific coastal lowlands of the Western Hemisphere. Syst. Zool., 16:28-37. Baker, R. H., and J. K. Greer. 1962. Mammals of the Mexican state of Durango. Publ. Mus. Michigan State Univ., Biol. Ser., 2:25-1 54. Baker, R. H., R. G. Webb, and P. Dalby. 1967. Notes on reptiles and mammals from southern Zacatecas. Amer. Midland Nat., 77:223-226. Berry, R. J., and A. G. Searle. 1963. Epigenetic polymorphism of the rodent skeleton. Proc. Zool. Soc. London, 140:577-615. Birney, E. C. 1970. Systematics of three species of woodrats (genus Neotoma) in central North America. Unpublished Ph.D. dissertation, The University of Kansas, Lawrence, 277 pp. Blair, W. F. 1949. Extensions of the known ranges of three species of Texas mammals. J. Mamm., 30:201-202. _ 1952. Mammals of the Tamaulipan Biotic Province in Texas. Texas J. Sci., 4:230- 250. Bole, B. P., Jr. 1937. Annotated list of mammals of the Mariato River District of the Azuero Peninsula. Pp. 140-188, in The birds and mammals of the western slope of the Azuero Peninsula [Republic of Panama] (J. W. Aldrich and B. P. Bole, Jr., eds.), Sci. Publ. Cleveland Mus. Nat. Hist., 7:1-196. 330 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY Booth, E. S. 1957. Mammals collected in Mexico from 1951 to 1956 by the Walla Walla College Museum of Natural History. Walla Walla College Pub!., Dept. Biol. Sci. and Biol. Station, 20:1-19. Bowers, D. E. 1956. A study of methods of color determination. Syst. Zool., 5:147-160, 182. Brennan, J. M. 1952. Two new Venezuelan chiggers (Acarina:Trombiculidae). J. Parasit., 38:143-146. - 1957. Revival of Crotiscus Ewing, 1944, and descriptions of four new species of the genus from Peruvian rodents (Acarina: Trombiculidae). J. Parasit., 43:673-680. - 1960a. Ectonyx, a new neotropical genus of chiggers (Acarina, Trombiculidae). Acarologia, 2:88-91. - 19606. Eight new species of Pseudoschoengastia from Mexico and Panama with a revised key to the species (Acarina: Trombiculidae). Acarologia, 2:480-492. - 1962. Four new chiggers from Mexico. J. Parasit., 48:618-620. - 1965. Two new species and other records of chiggers from Texas (Acarina: Trombiculidae). Acarologia, 7:79-83. - 1968a. Endoparasitic chiggers: IV. Revision of Kymocta Yunker and Brennan, 1962 (Acarina: Trombiculidae), with description of a new species. J. Parasit, 54- 613-615. - 19686. A collection of chiggers (Acarina: Trombiculidae) from rodents in south¬ western Colombia. J. Parasit., 54:679-685. - 1969. Three new species of subgenus Parasecia Loomis (genus Fonsecia) from northeastern Brazil and a key to the included species (Acarina: Trombiculidae). J. Parasit., 55:662-666. Brennan, J. M., and H. T. Dalmat. 1960. Chiggers of Guatemala (Acarina: Trombi¬ culidae). Ann. Ent. Soc. America, 53:183-191. Brennan, J. M., and E. K. Jones. 1959. Pseudoschoengastia and four new neotropical species of the genus (Acarina: Trombiculidae). J. Parasit., 45:421-429. - 1960. Chiggers of Trinidad, B.W.I. (Acarina:Trombiculidae). Acarologia, 2:493- 540. - 1961. New genera and species of chiggers from Panama (Acarina: Trombiculidae). J. Parasit., 47:105-124. Brennan, J. M., and C. E. Yunker. 1966. The chiggers of Panama (Acarina: Trombicu¬ lidae). Pp. 221-266, in Ectoparasites of Panama (R. L. Wenzel and V. J. Tipton, eds.), Field Mus. Nat. Hist., Chicago, xii + 861 pp. - 1969. Endoparasitic chiggers: V. New genera, species and records from Vene¬ zuela and Brazil (Acarina: Trombiculidae). J. Med. Ent., 6:299-304. Brown, W. L. 1957. Centrifugal speciation. Quart. Rev. Biol., 32:247-277. Brown, W. L., and E. O. Wilson. 1956. Character displacement. Syst. Zool., 5:49-64. Buchanan, O. M., and T. R. Howell. 1965. Observations on the natural history of the thick-spined rat, Hoplomys gymnurus, in Nicaragua. Ann. Mag. Nat. Hist., ser. 13, 8:549-559. Burt, W. H. 1936. A study of the baculum in genera Perognathus and Dipodomys. J. Mamm., 17:145-156. _ 1938. Faunal relationships and geographic distribution of mammals in Sonora, Mexico. Misc. Publ. Mus. Zool., Univ. Michigan, 39:1-77. _ 1960. Bacula of North American mammals. Misc. Publ. Mus. Zool., Univ. Michigan, 1 13:1-76. Burt, W. H., and F. S. Barkalow, Jr. 1942. A comparative study of the bacula of wood rats (subfamily Neotominae). J. Mamm., 23:287-297. Burt, W. H., and R. A. Stirton. 1961. The mammals of El Salvador. Misc. Publ. Mus. Zool., Univ. Michigan, 1 1 7: 1-69. GENOWAYS— SYSTEMATICS OF LIOMYS 331 Caballero y Caballero, E. 1959. Estudios helmintologicos de la region oncocercosa de Mexico y de la Republica de Guatemala. Nematoda. 10 a. parte. An. Escuela Nac. Cien. Biol., Mexico, 9:61-76. Choate, J. R. 1970. Systematics and zoogeography of Middle American shrews of the genus Cryptotis. Univ. Kansas Publ., Mus. Nat. Hist., 19:195-317. Clark, G. M. 1967. New Speleognathinae from Central and South American mammals (Acarina, Trombidiformes). Proc. Helminthological Soc. Washington, 34:240-243. Clark, L. G., V. M. Varela-Diaz, C. R. Sulzer, R. R. Marshak, and C. J. Hollister. 1966. Leptospirosis in Nicaragua: preliminary report on the first year of study. Amer. J. Trop. Med. and Hyg., 15:735-742. Costa Lima, A., and C. R. Hathaway. 1946. Pulgas bibliografia, catalogo e animais por elas sugados. Monogr. Inst. Oswaldo Cruz, Brasil, 4:1-522. Cova Garcia, P., and E. Tallaferro D. 1959. Pulgas mas comunes de Venezuela. Arch. Venezolandos Patol. Trop. y Parasit. Med., 3:327-348. Cushing, J. E„ Jr. 1945. Quarternary rodents and lagomorphs of San Josecito Cave, Nuevo Leon, Mexico. J. Mamm, 26:182-185. Dalquest, W. W. 1953. Mammals of the Mexican state of San Luis Potosi. Louisiana State Univ. Studies, Biol. Sci. Ser., 1:1-229. Dalquest, W. W., and E. Roth. 1970. Late Pleistocene mammals from a cave in Tama- ulipas, Mexico. Southwestern Nat., 15:217-230. Davis, W. B. 1944. Notes on Mexican mammals. J. Mamm., 25:370-403. - 1960. The mammals of Texas. Bull. Texas Game and Fish Comm., 41 : 1-252. Davis, W. B., and J. R. Dixon. 1959. Snakes of the Chilpancingo region, Mexico. Proc. Biol. Soc. Washington, 72:79-92. Davis, W. B., and R. J. Russell. 1954. Mammals of the Mexican state of Morelos. J. Mamm., 35:63-80. Disney, R. H. L. 1964. Visceral involvement with dermal leishmaniasis in a wild-caught rodent in British Honduras. Trans. Roy. Soc. Trop. Med. and Hyg., 58:581. Duellman, W. E. 1960. A distributional study of the amphibians of the Isthmus of Tehuantepec, Mexico. Univ. Kansas Publ., Mus. Nat. Hist., 13:19-72. - 1961. The amphibians and reptiles of Michoacan, Mexico. Univ. Kansas Publ., Mus. Nat. Hist., 15:1-148. - 1966. The Central American herpetofauna: an ecological perspective. Copeia, pp. 700-719. Dunnigan, P. B. 1967. Pocket gophers of the genus Thomomys of the Mexican state of Sinaloa. Radford Rev., 21:139-168. Eads, R. B. 1950. The fleas of Texas. Texas State Health Dept., Austin, 85 pp. Eads, R. B., and G. C. Menzies. 1949. A preliminary list of the Siphonaptera of Texas. Texas J. Sci., 1 (4): 33-39. Eads, R. B., H. A. Trevino, and E. G. Campos. 1965. Ectoparasites of the spiny pocket mouse, Liomys irroratus texensis Merriam. Southwestern Nat., 10:17-21. Eisenberg, J. F. 1963a. The behavior of heteromyid rodents. Univ. California Publ. Zool., 69:iv+ 1-100. _ 1963 b. A comparative study of sandbathing behavior in heteromyid rodents. Behaviour, 22:16-23. _ 1967. A comparative study in rodent ethology with emphasis on evolution of social behavior, 1. Proc. U.S. Nat. Mus., 122(3597): 1-51. Eisenberg, J. F., and D. E. Isaac. 1963. The reproduction of heteromyid rodents in captivity. J. Mamm., 44:61-67. Elliot, D. G. 1903a. A list of a collection of Mexican mammals with descriptions of some apparently new forms. Field Columbian Mus., Zool. Ser., 3: 141-149. _ 1903/). Descriptions of apparently new species of mammals of the genera Heteromys and Ursus from Washington and Mexico. Field Columbian Mus., Zool. Ser., 3:233-237. 332 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY - 1904. The land and sea mammals of Middle America and the West Indies. Field Columbian Mus., Zool. Ser., 4(Part I):xxi + 1-439 + xlix. - 1905. A check list of mammals of the North American Continent, the West Indies, and the neighboring seas. Field Columbian Mus., Zool. Ser., 6:iv+ 1-761. - 1907. A catalogue of the collection of mammals in the Field Columbian Museum. Field Columbian Mus., Zool. Ser., 8: viii + 1-694. Emerson, K. C. 1971«. Records of a collection of Mallophaga and Anoplura from Nic¬ araguan mammals. J. Kansas Ent. Soc., 44:332-334. - 19716. New records of Anoplura from Mexico. J. Kansas Ent. Soc., 44:374- 377. Enders, R. K. 1935. Mammalian life histories from Barro Colorado Island, Panama. Bull. Mus. Comp. Zool., 78:385-502. Esquivel, R. R., J. A. Zuniga, M. Alfaro, and E. Kotcher. 1967. Trypanosomes of Costa Rican feral mammals. J. Parasit., 53:951-955. Fain, A. 1967. Nouveaux hypopes vivant en association phoreteque sur des rongeurs et des marsupiaux (Acarina:Glycyphagidae). Acarologia, 9:415-434. Fairchild, G. B., and C. O. Handley, Jr. 1966. Gazetteer of collecting localities in Pan¬ ama. Pp. 9-22, in Ectoparasites of Panama (R. L. Wenzel and V. J. Tipton, eds.), Field Mus. Nat. Hist., Chicago, xii+ 861 pp. Fairchild, G. B., G. M. Kohls, and V. J. Tipton. 1966. The ticks of Panama (Acarina: Ixodoidea). Pp. 167-219, in Ectoparasites of Panama (R. L. Wenzel and V. J. Tipton, eds.), Field Mus. Nat. Hist., Chicago, xii + 861 pp. Farris, J. S. 1966. Estimation of conservatism of characters of constancy within bio¬ logical populations. Evolution, 20:587-591. - 1967. The meaning of relationship and taxomonic procedure. Syst. Zool., 16:44-51. - 1970. Methods for computing Wagner Trees. Syst. Zool., 19:83-92. Felten, H. 1957. Nagetiere (Mammalia, Rodentia) aus El Salvador. Teil 1. Sencken- bergiana Biol., 38:146-155. Ferris, G. F. 1922. Contribution toward a monograph of the sucking lice, Part III. Stanford Univ. Publ., Univ. Series, Biol. Sci., 2:135-178. Fleming, T. H. 1969. Population ecology of three species of Neotropical rodents. Un¬ published Ph.D. thesis, Univ. Michigan, Ann Arbor, 231 pp. - 1970. Notes on the rodent faunas of two Panamanian forests. J. Mamm., 51: 473-490. - 1971. Population ecology of three species of Neotropical rodents. Misc. Publ. Mus. Zool., Univ. Michigan, 143:1-77. Forman, G. L. 1968. Comparative gross morphology of spermatozoa of two families of North American bats. Univ. Kansas Sci. Bull., 47:901-928. Fox, I. 1947. Notes on ectoparasites from Venezuela (Siphonaptera and Acarina). Zool- ogica: Sci. Contrib. New York Zool. Soc., 32: 117-119. Furman, D. P. 1955. Steptolaelaps (Acarina: Laelaptidae) a new genus of mites parasitic on neotropical rodents. J. Parasit., 41:519-525. Furman, D. P., and V. J. Tipton. 1961. Acaros parasitos laelaptine (Acarina:Laeiaptidae) de Venezuela. Mem. Soc. Cien. Nat. La Salle, 21:166-212. Gabriel, K. R. 1964. A procedure for testing the homogeneity of all sets of means in analysis of variance. Biometrics, 20:459-477. Gardner, A. L. 1966. A new subspecies of the Aztec mastiff bat, Molossus aztecus, Saussure, from southern Mexico. Contrib. Sci., Los Angeles Co. Mus., 111:1-5. Gaumer, G. F. 1917. Monografia de los mamiferos de Yucatan. Depto. de Talleres Graficos de la Secretaria de Fomento, Mexico, xii + 33 1 pp. Geest, J. C., and R. B. Loomis. 1968. Chiggers of the genus Pseudoschoengastia (Acarina: Trombiculidae) from Costa Rica. Contrib. Sci., Los Angeles Co Mus 150:1-49. GENOWAYS— SYSTEMATICS OF LIOMYS 333 oways, H. H. 1971. A new species of spiny pocket mouse (genus Liomys) from Jal¬ isco, Mexico. Occas. Papers Mus. Nat. Hist., Univ. Kansas 51-7 Genoways H. H„ and J, R. Choate. 1972. A multivariate analysis of systematic relation- ships among populations of the short-tailed shrew (genus Blarina) in Nebraska Syst. Zool., 21:106-1 16. Genoways, H. H„ and J. K. Jones, Jr. 1971, Systematics of southern banner-tailed kangaroo rats of the Dipodomys phillipsii group. J. Mamm., 52:265-287. Glass, B. P. 1947. Geographic variation in Perognathus hispidus. J. Mamm., 28: 1 74- 1 79. oldman, E. A. 1904. Descriptions of five new mammals from Mexico. Proc. Biol. Soc. Washington, 17:79-82. 1911. Revision of the spiny pocket mice (genera Heteromys and Liomys). N. Amer. Fauna, 34:1-70. - 1920. Mammals of Panama. Smithsonian Misc. Coll., 69(5): 1-309. - 1951. Biological investigation in Mexico. Smithsonian Misc. Coll., 115-xiii+l- 476. Goldman, E. A., and R. T. Moore. 1946. The biotic provinces of Mexico. J. Mamm 26:347-360. Goodwin, G. G. 1932a. Two new mammals from Guatemala. Amer. Mus. Novit., 528:1-2. - 19326. Three new Reithrodontomys and two new Peromyscus from Guatemala. Amer. Mus. Novit., 560:1-5. - 1934. Mammals collected by A. W. Anthony in Guatemala, 1924-1928. Bull. Amer. Mus. Nat. Hist., 68:1-60. - 1938. Four new mammals from Costa Rica. Amer. Mus. Novit., 987:1-5. - 1942. Mammals of Honduras. Bull. Amer. Mus. Nat. Hist., 79:107-195. - 1945. Specialized claw in Liomys. J. Mamm., 26:308-309. - 1946. Mammals of Costa Rica. Bull. Amer. Mus. Nat. Hist., 87:271-474. - 1954. Mammals from Mexico collected by Marian Martin for the American Museum of Natural History. Amer. Mus. Novit., 1689:1-16. - 1956a. A preliminary report on the mammals collected by Thomas MacDougall in southeastern Oaxaca, Mexico. Amer. Mus. Novit., 1757:1-15. - 19566. Seven new mammals from Mexico. Amer. Mus. Novit., 1791: 1-10. - 1969. Mammals from the state of Oaxaca, Mexico, in the American Museum of Natural History. Bull. Amer. Mus. Nat. Hist., 141:1-270. Grant, C. D. 1947. North American mites of the genus Laelaps (Arachnida: Acarina: Parasitidae). Microent., 12:2-21. Gray, J. E. 1868. Synopsis of the species of Saccomyincie, or pouched mice, in the col¬ lection of British Museum. Proc. Zool. Soc. London, pp. 199-206. Griscom, L. 1932. The distribution of bird-life in Guatemala. Bull. Amer. Mus. Nat. Hist., 64:ix + 1-439. Hall, E. R. 1941. Revision of the rodent genus Microdipodops. Field Mus. Nat. Hist., Zool. Ser., 27:233-277. - 1946. Mammals of Nevada. Univ. California Press, Berkeley, xi + 710 pp. Hall, E. R., and F. H. Dale. 1939. Geographic races of the kangaroo rat, Dipodomys microps. Occas. Papers Mus. Zool., Louisiana State Univ., 4:47-62. Hall, E. R„ and W. W. Dalquest. 1963. The mammals of Veracruz. Univ. Kansas Publ., Mus. Nat. Hist., 14:165-362. Hall, E. R., and K. R. Kelson. 1959. The mammals of North America. Ronald Press, New York, l:xxx + 1-546+ 79. Hall, E. R„ and B. Villa-R. 1948. A new pocket gopher (Thomomys) and a new spiny pocket mouse (Liomys) from Michoacan, Mexico. Univ. Kansas Publ., Mus. Nat. Hist., 1:251-255. _ 1949. An annotated check list of the mammals of Michoacan, Mexico. Univ. Kansas Publ., Mus. Nat. Hist., 1:431-472. 334 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY Hamerton, J. L„ H. P. Klinger, D. E. Mutton, and E. M. Lang. 1963. The somatic chromosomes of the Hominoides. Cytogenetics, 2:240-263. Hamilton, W. J., Jr. 1946. A study of the baculum in some North American Micro- tinae. J. Mamm., 27:378-387. Handley, C. O., Jr. 1966. Checklist of the mammals of Panama. Pp. 753-795, in Ecto¬ parasites of Panama (R. L. Wenzel and V. J. Tipton, eds.), Field Mus. Nat. Hist., Chicago, xii+ 861 pp. Hardy, L. M., and R. W. McDiarmid. 1969. The amphibians and reptiles of Sinaloa, Mexico. Univ. Kansas Publ., Mus. Nat. Hist., 18:39-252. Harris, W. P., Jr. 1943. A list of mammals from Costa Rica. Occas. Papers Mus. Zool., Univ. Michigan, 476:1-15. Hatt, R. T. 1932. The vertebral columns of ricochetal rodents. Bull. Amer. Mus. Nat. Hist., 63:599-738. - 1938. Notes concerning mammals collected in Yucatan. J. Mamm., 19:333-337. Hedges, S. R. 1969. Epigenetic polymorphism in populations of Apodemus sylvaticus and Apodemus flavicollis (Rodentia, Muridae). J. Zool., 159:425-442. Hershkovitz, P. 1958. A geographic classification of Neotropical mammals. Fieldiana: Zool., 36:581-620. - - 1966. South American swamp and fossorial rats of the scapteromyine group (Cricetinae, Muridae) with comments on the glans penis in murid taxonomy. Z. Saugetier., 31:81-149. - 1969. The evolution of mammals on southern continents. IV. The Recent mam¬ mals of the Neotropical Region: a zoogeographic and ecological review Quart Rev. Biol., 44:1-70. Hibbard, C. W. 1941a. New mammals from the Rexroad Fauna, Upper Pliocene of Kansas. Amer. Midland Nat., 26:337-368. - 1 941 h. Mammals of the Rexroad Fauna from the Upper Pliocene of southwestern Kansas. Trans. Kansas Acad. Sci., 44:265-313. Hoffmann, A. 1963. Contribuciones al conocimiento de los trombiculidos mexicanos (Acarina: Trombiculidae) (10 a. parte). An. Escuela Nac. Cien. Biol, Mexico 12:101-109. Hoffmann, R. S., and C. F. Nadler. 1968. Chromosomes and systematics of some North American species of the genus Marmota (Rodentia: Sciuridae). Experientia 24- 740-742. Hooper, E. T. 1947. Notes on Mexican mammals. J. Mamm., 28:40-57. - 1949. Faunal relationships of Recent North American rodents. Misc. Publ. Mus. Zool., Univ. Michigan, 72:1-28. - 1952. A systematic review of the harvest mice (genus Reithrodontomys) of Latin America. Misc. Publ. Mus. Zool., Univ. Michigan, 77:1-255. - 1953. Notes on mammals of Tamaulipas, Mexico. Occas. Papers Mus. Zool., Univ. Michigan, 544:1-12. - 1955. Notes on mammals of western Mexico. Occas. Papers Mus. Zool., Univ. Michigan, 565: 1-26. - 1957. Records of Mexican mammals. Occas. Papers Mus. Zool., Univ. Michigan 586:1-9. - 1958. The male phallus in mice of the genus Peromyscus. Misc. Publ. Mus. Zool., Univ. Michigan, 105:1-24. - 1959. The glans penis in five genera of cricetid rodents. Occas. Papers Mus. Zool., Univ. Michigan, 613:1-1 1. - 1960. The glans penis in Neotoma (Rodentia) and allied genera. Occas. Papers Mus. Zool., Univ. Michigan, 618:1-21. - 1961. The glans penis in Proechimys and other caviomorph rodents. Occas. Pa¬ pers Mus. Zool., Univ. Michigan, 623:1-18. GENOWAYS— SYSTEMATICS OF LIOMYS 335 1962. The glans penis in Sigmodon, Sigmomys, and Reithrodon (Rodentia, Cricetmae). Occas. Papers Mus. Zool., Univ. Michigan, 625:1-1 1. Hooper, E. T„ and C. O. Handley, Jr. 1948. Character gradients in the spiny pocket mouse, Liomys irroratus. Occas. Papers Mus. Zool., Univ. Michigan, 514:1-34. Hooper, E. T., and B. S. Hari. 1962. A synopsis of Recent North American microtine rodents. Misc. Publ. Mus. Zool., Univ. Michigan, 120:1-68. Hooper, E. T., and G. G. Musser. 1 964. The glans penis in Neotropical cricetines (family Muridae) with comments on classification of muroid rodents. Misc. Publ. Mus Zool., Univ. Michigan, 123:1-57. Hsu, T. C., and R. A. Mead. 1969. Mechanisms of chromosomal changes in mam¬ malian speciation. Pp. 8-17, in Comparative mammalian cytogenetics (K. Benirschke, ed.), Springer-Verlag, New York, xxi + 473 pp. Hubbard, C. A. 1958. Mexican jungle and desert fleas with three new descriptions. Entom. News, 69:161-166. Hudson, J. W., and J. A. Rummel. 1966. Water metabolism and temperature regulation of the primitive heteromyids, Liomys salvani [57’c] and Liomys irroratus. Ecology 47:345-354. Ingles, L. G. 1959. Notas acerca de los mamiferos mexicanos. An. Inst. Biol, Mexico 29:379-408. Ivens, V. F„ J. Kruidenier, and N. D. Levine. 1959. Further studies of Eimeria (Pro- tozoa:Eimeriidae) from Mexican rodents. Trans. Illinois State Acad Sci 51(1-2)- 53-57. Jackson, R. C., and T. J. Crovello. 1971. A comparison of numerical and biosystematic studies in Haplopappus. Brittonia, 23:54-70. Jakway, G. E. 1958. Pleistocene Lagomorpha and Rodentia from the San Josecito Cave, Nuevo Leon, Mexico. Trans. Kansas Acad. Sci., 61:313-327. Johnson, P. T. 1957. A classification of the Siphonoptera of South America, with de¬ scriptions of new species. Mem. Ent. Soc. Washington, 5:1-299. - 1962. The species of Fahrenholzia Kellogg and Ferris from spiny pocket mice (Anoplura: Haplopleuridae). Ann. Ent. Soc. Amer., 55:415-428. Johnston, R. F. 1966. The adaptive basis of geographic variation in color of the Purple Martin. Condor, 68:219-228. Jolicoeur, P. 1959. Multivariate geographical variation in the wolf Canis lupus L. Evolution, 13:283-299. Jones, J. K., Jr. 1953. Geographic distribution of the pocket mouse, Perognathus fasci- atus. Univ. Kansas Publ., Mus. Nat. Hist., 5:515-526. - 1966. Bats from Guatemala. Univ. Kansas Publ., Mus. Nat. Hist., 16:439-472. Jordan, K., and N. C. Rothschild. 1923. On the genera Rhopalopsyllus and Parapsyllus. Ectoparasites, 1:320-370. Judd, D. B. 1933. The 1931 I. C. I. standard observer and coordinate system for colori¬ metry. J. Optical Soc. America, 23:359-374. Keirans, J. E., and E. K. Jones. 1972. Description of the immature stages of Ixodes (/.) sinaloa Kohls and Clifford (Acarina: Ixodidae), from rodents in Mexico and Nicaragua. Acarologia, 13:471-475. Klinger, H. P. 1963. The somatic chromosomes of some primates ( Tupaia glis , Nyctice- bus coucang, Tarsius bancanus, Cercocebus aterrimus, Symphalangus syndactylus). Cytogenetics, 2: 1 40- 151. Kluge, A. G. 1969. The evolution and geographical origin of the New World Hemi- dactylus mabouia-brookii complex ( Gekkonidae , Sauria). Misc. Publ. Mus. Zool., Univ. Michigan, 138:1-78. Kluge, A. G., and J. S. Farris. 1969. Quantitative phyletics and the evolution of anurans. Syst. Zool., 18:1-32. Koestner, E. J. 1941. An annotated list of mammals collected in Nuevo Leon, Mexico, in 1938. Great Basin Nat., 2:9-15. 336 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY Kohls, G. M. 1953. Ixodes venezuelensis , a new species of tick from Venezuela, with notes on Ixodes minor Neumann, 1902 (Acarina: Ixodidae). J. Parasit., 39:300- Kohls, M and C M. Clifford. 1964. Ixodes eadsi , a new species of tick from ro¬ dents in southern Texas (Acarina: Ixodidae). J. Parasit., 50:466-470 -• I966- . Three new species of ,xodes from Mexico and description of the male of . auritutus auritulus Neumann, /. cone pal, Cooley and Kohls, and /. km, He, Men- dez and Ortiz (Acarina: Ixodidae). J. Parasit., 52:810-820. Koopman, K. F„ and P. S. Martin. 1959. Subfossil mammals from the Gomez Farias region and the tropical gradient of eastern Mexico. J. Mamm 40- 1-12 ’ MV1L’.AN“ E Tash,an' I954- Notes on mammals from northern Chiapas, Mexico. J. Mamm., 35:100-103. Lainson, R. and j. Strancways-Dixon. I964U. Leishmania mexicanm epidemiology of dermal leishmaniasis in British Honduras. Trans. Roy. Soc. Trap. Med and Hyg 58:3. -r£”’ 19646. The epidemiology of dermal leishmaniasis in British Honduras. Part II. Reservior-hosts of Leishmania mexicana among the forest rodents. Trans Rov Soc. Trop. Med. and Hyg., 58:136-153. Land, H. C. 1962 A collection of birds from the arid interior of eastern Guatemala Auk, 79:1-11. Lawlor, T E. 1965. The Yucatan deer mouse, Peromyscus yucatanicus. Univ. Kansas Publ., Mus. Nat. Hist., 16:421-438. - 1971. Evolution of Peromyscus on northern islands in the Gulf of California Mexico. Trans. San Diego Soc. Nat. Hist., 16:91-124. Lawrence B. and W H. Bossert. 1969. The cranial evidence for hybridization in New England Canis. Breviora, Mus. Comp. Zool., 330:1-13. Lee, M. R 1969. A widely applicable technic for direct processing of bone marrow for chromosomes of vertebrates. Stain Tech., 44: 155-158. Levine, N_ D„ V^Ivens, and F. J. Kruidenier. 1958. New species of Eimeria (Protozoa: lmerndae) from Mexican rodents. Trans. Illinois State Acad. Sci., 50:291-298 Lidicker, W. Z„ Jr. I960. An analysis of intraspecific variation in the kangaroo rat Dipodomys merriami. Univ. California Publ. Zool., 67:125-218 - 1962 The nature of subspecies boundaries in a desert rodent and its implications tor subspecies taxonomy. Syst. Zool., 1 1:160-171. Ling, J. K. 1970. Pelage and molting in wild mammals with special reference to aquatic forms. Quart. Rev. Biol., 45:16-54. Lloyd, J. J. 1963. Tectonic history of the south Central-American orogen Pp 88-100 in Backbone of the Americas (O. E. Childs and B. W. Beebe, eds.), American Assoc' Petro. Geol., Memoir, 2:1-320. Long, C. A. 1968. An analysis of patterns of variation in some representative Mammalia. A^V S^' * e^Q |^^2 2^7 ^ m at eS variability in selected measurements. Trans. Kansas 1969. An analysis of patterns of variation in some representative Mammalia. Part II. Studies on the nature and correlation of measures of variation. Pp. 289- 302, in Contributions in mammalogy (J. K. Jones, Jr., ed.), Misc. Publ. Mus Nat Hist., Univ. Kansas, 51:1-428. - 1970. An analysis of patterns of variation in some representative Mammalia Part III. Some equations on the nature of frequency distributions of estimated’ vari¬ abilities. Acta Theriol., 15:517-520. Loomis, R. B. 1969. Chiggers (Acarina, Trombiculidae) from vertebrates of the Yucatan Peninsula, Mexico. Misc. Publ. Mus. Nat. Hist., Univ. Kansas, 50:1-22. -• I97L The ®enus Euschoengastoides (Acarina: Trombiculidae) from North America. J. Parasit., 57:689-707. GENOWAYS— SYSTEMATICS OF LIOMYS 337 Loomis, R. B„ and D. A. Crossley, Jr. 1963. New species and new records of chiggers (Acarina. Tiombiculidae) from Texas. Acarologia, 5:371-383. Makino, S. 1953. Chromosome numbers of some American rodents. Science, 1 1 8:630. Matthey, R. 1958. Les chromosomes des mammiferes eutheriens: liste critique et essai sur evolution chromosomique. Archiv der Julius Klaus-Stiftung, 33:253-297. Mavr, E^ 1966. Animal species and evolution. Harvard Univ. Press, Cambridge, Massachusetts, xiv + 1-797 pp. Merriam C. H. 1902. Twenty new pocket mice (Heteromys and Liomys) from Mexico. Proc. Biol. Soc. Washington, 15:41-50. Mones, A. 1971. Restos oseos de mamiferos contenidos en regurgitaciones de lechuza del estado de Oaxaca, Mexico. An. Inst. Biol., Mexico, 39:169-171. Murie, A. 1935. Mammals from Guatemala and British Honduras. Misc. Publ. Mus Zool., Univ. Michigan, 26:7-30. Musser, G. G. 1964. Notes on geographic distribution, habitat, and taxonomy of some Mexican mammals. Occas. Papers Mus. Zool., Univ. Michigan, 636:1-22. 1969a. Results of the Archbold Expeditions. No. 91. A new genus and species of murid rodent from Celebes, with a discussion of its relationships. Amer. Mus. Novit., 2384:1-41. 1969 b. Results of the Archbold Expeditions. No. 92. Taxonomic notes on Rattus do 1 1 man i and Rattus hellwaldi (Rodentia, Muridae) of Celebes. Amer Mus Novit 2386:1-24. - 1970. Species-limits of Rattus brahma , a murid rodent of northeastern India and northern Burma. Amer. Mus. Novit., 2406:1-27. Nadler, C. F. 1966. Chromosomes and systematics of American ground squirrels of the subgenus Spermophilus. J. Mamm., 47:579-596. 1969. Chromosomal evolution in rodents. Pp. 277-309, in Comparative mam¬ malian cytogenetics (K. Benirschke, ed.), Springer-Verlag, New York, xxi + 473 pp. Packard, R. L. 1958. New subspecies of the rodent Baiomys from Central America. Univ. Kansas Publ., Mus. Nat. Hist., 9:397-404. Patton, J. L. 1967. Chromosome studies of certain pocket mice, genus Perognathus (Ro¬ dentia: Heteromyidae). J. Mamm., 48:27-37. Peters, W. 1874. Die Taschenmause, Nager mit ausseren taschenformigen Backen- taschen, und eine neue Art Derselben, Heteromys adspersus , aus Panama. Monatsb. preuss. Akad. Wiss., Berlin, pp. 354-359. Power, D. M. 1970. Geographic variation of Red-winged Blackbirds in central North America. Univ. Kansas Publ., Mus. Nat. Hist., 19:1-83. Prim, R. C. 1957. Shortest connection networks and some generalizations. Bell Syst Tech. J., 36:1389-1401. Quay, W. B. 1965. Comparative survey of the sebaceous and suboriferous glands of the oral lips and angle in rodents. J. Mamm., 46:23-37. Reeder, W. G. 1953. Age variation in enamel patterns of the spiny pocket mouse, Liomys pictus sonoranus. J. Mamm., 34:59-64. Ridgway, R. 1912. Color standards and color nomenclature. Privately published, Wash¬ ington, D. C., iii + 43 pp. Rising, J. D. 1970. Morphological variation and evolution in some North American orioles. Syst. Zool., 19:315-351. Rohlf, F. J. 1968. Stereograms in numerical taxonomy. Syst. Zool., 17:246-255. Rood, J. P. 1963. Observations on the behavior of the spiny rat Heteromys melanoleucus in Venezuela. Mammalia, 27:186-192. Rood, J. P., and F. H. Test. 1968. Ecology of the spiny rat, Heteromys anomalus, at Rancho Grande, Venezuela. Amer. Midland Nat., 79:89-102. Rowley, J. S. 1966. Breeding records of birds of the Sierra Madre del Sur, Oaxaca, Mexico. Proc. Western Foundation Vert. Zool., 1:107-204. 338 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY Russell, R. J. 1960. Pleistocene pocket gophers from San Josecito Cave, Nuevo Leon, Mexico. Univ. Kansas Publ., Mus. Nat. Hist., 9:539-548. - 1968. Revision of pocket gophers of the genus Pappogeomys. Univ. Kansas Publ., Mus. Nat. Hist., 16:581-776. Ryan, R. M. 1963. The biotic provinces of Central America. Acta Zool. Mexicana, 6 (2-3): 1-54. Savage, J. M. 1966. The origin and history of the Central American herpetofauna. Copeia, pp. 719-766. Schaldach, W. J., Jr. 1963. The avifauna of Colima and adjacent Jalisco, Mexico. Proc. Western Foundation Vert. Zool., 1:1-100. - 1966. New forms of mammals from southern Oaxaca, Mexico, with notes on some mammals of the coastal range. Saugetierk. Mitt., 14:286-297. Schnell, G. D. 1970. A phenetic study of the suborder Lari (Aves). I. Methods and results of principal components analyses. Syst. Zool., 19:35-57. Schmidly, D. J. 1971. Population variation in Dipodomys ordii from western Texas. J. Mamm., 52:108-120. Selander, R. K., and R. F. Johnston. 1967. Evolution in the House Sparrow. I. Intra¬ population variation in North America. Condor, 69:217-258. Selander, R. K., R. F. Johnston, and T. H. Hamilton. 1964. Colorimetric methods in ornithology. Condor, 66:491-495. Setzer, H. W. 1949. Subspeciation in the kangaroo rat, Dipodomys ordii. Univ. Kansas Publ., Mus. Nat. Hist., 1:473-573. Shotwell, J. A. 1967. Late Tertiary geomyid rodents of Oregon. Bull. Mus. Nat. Hist., Univ. Oregon, 9:1-51. Simpson, G. G. 1941. Large Pleistocene felines of North America. Amer. Mus. Novit., 1 136:1-27. Simpson, G. G., A. Roe, and R. C. Lewontin. 1960. Quantitative Zoology. Revised edition, Harcourt, Brace, and World, Inc., New York, 440 pp. Smith, H. M. 1965. More evolutionary terms. Syst. Zool., 14:57-58. Sokal, R. R., and F. J. Rohlf. 1969. Biometry: the principles and practice of statistics in biological research. W. H. Freeman and Co., San Francisco, xiii + 776 pp. Sokal, R. R., and P. H. Sneath. 1963. Principles of numerical taxonomy. W. H. Freeman and Co., San Francisco, xvi + 359 pp. Speth, R. L. 1969. Patterns and sequences of molts in the Great Basin pocket mouse, Perognathus parvus. J. Mamm., 50:284-290. Stojanovich, C. J., and H. D. Pratt. 1961. Fahrenholzia texana, new species, with a key to the United States species of Fahrenholzia (Anoplura: Hoplopleuridae). Ann. Ent. Soc. America, 54:693-696. Strandtmann, R. W., and C. E. Yunker. 1966. The genus Hirstionyssus Fonseca in Pana¬ ma (Acarina: Dermanyssidae). Pp. 105-124, in Ectoparasites of Panama (R. L. Wenzel and V. J. Tipton, eds.). Field Mus. Nat. Hist., Chicago, xii + 861 pp. Stuart, L. C. 1950. A geographic study of the herpetofauna of Alta Verapaz, Guate¬ mala. Contrib. Lab. Vert. Biol., Univ. Michigan, 45:1-77. _ 1954. Herpetofauna of the southeastern highlands of Guatemala. Contrib. Lab. Vert. Biol., Univ. Michigan, 68:1-65. _ 1966. The environment of the Central American cold-blooded vertebrate fauna. Copeia, pp. 684-699. Sumner, F. B. 1927. Linear and colorimetric measurements of small mammals. J. Mamm., 8: 1 77-206. Thomas, O. 1893a. Description of two new “pocket-mice” of the genus Heteromys. Ann. Mag. Nat. Hist., ser. 6, 1 1:329-332. _ 18936. On two new members of the genus Heteromys and two of Neotoma. Ann. Mag. Nat. Hist., ser. 6, 12:233-235. GENOWAYS— SYSTEMATICS OF LIOMYS 339 ' i927’ A selectlon of leclotypes of American rodents in the collection of the British Museum. Ann. Mag. Nat. Hist., ser. 9, 19:545-554. Tipton, V. J., R. M. Altman, and C. M. Keenan. 1966. Mites of the subfamily Lae- aptinae in Panama .(Acarina: Laelaptidae). Pp. 23-82, in Ectoparasites of Panama (R L. Wenzel and V. J. Tipton, eds.). Field Mus. Nat. Hist., Chicago, xii + 861 pp. ipton, V. J., and E. Mendez. 1966. The fleas (Siphonaptera) of Panama. Pp. 289-385, in Ectoparasites of Panama (R. L. Wenzel and V. J. Tipton, eds.), Field Mus. Nat. Hist., Chicago, xii + 861 pp. Traub, R. 1953. Wenzella obscura a new genus and new species of fleas from Guate¬ mala (Siphonaptera). J. Washington Acad. Sci., 43:77-85. Traub, R„ and P. T. Johnson. 1952. Fleas collected during a plague survey in Vene¬ zuela. Bol. Ofic. San. Panamericana, 32:1 1 1-135. Twente, J. W„ and R. H. Baker. 1951. New records of mammals from Jalisco, Mexico, from barn owl pellets. J. Mamm., 32:120-121. Villa-R., B. 1949. Mamlferos del Soconusco, Chiapas. An. Inst. Biol , Mexico 19- 485-528. - 1953. Mamlferos silvestres del Valle de Mexico. An. Inst. Biol., Mexico 23- 269-492. Vivo Escoto, J. A. 1964. Weather and climate of Mexico and Central America. Pp. 187- 215, in Natural environment and early cultures, vol. 1, Handbook of Middle Ameri¬ can Indians (R. C. West, ed.), Univ. Texas Press, Austin, vii + 570 pp. Vogelsang, E. G., and T. Santos Dias. 1953. Nueva contribucion al estudio de la fauna ixodologica en Venezuela. Rev. Med. Vet. y Parasit., Caracas, 12:63-89. Wade, O., and P. T. Gilbert. 1940. The baculum of some Sciuridae and its significance in determining relationships. J. Mamm., 21:52-63. Wagner, H. O. 1961. Die Nagetiere einer Gebirgsabdachung in Sudmexiko und ihre Beziehungen zur Umwelt. Zool. Jb. Syst., 89:177-242. Wagner, P. L. 1964. Natural vegetation of Middle America. Pp. 216-264, in Natural environment and early cultures, vol. 1, Handbook of Middle American Indians (R. C. West, ed.), Univ. Texas Press, Austin, vii + 570 pp. Webb, J. P., Jr., and R. B. Loomis. 1971. Trombiculid mites of the genus Microtrom- bicula (Acarina) from Costa Rica. Contrib. Sci., Los Angeles Co. Mus., 207:1-15. Wenzel, R. L., and P. T. Johnson. 1966. Checklist of the sucking lice of Panama. Pp. 273-279, in Ectoparasites of Panama (R. L. Wenzel and V. J. Tipton, eds.), Field Mus. Nat. Hist., Chicago, xii + 861 pp. Wenzel, R. L., and V. J. Tipton. 1966. Some relationships between mammal hosts and their ectoparasites. Pp. 677-723, in Ectoparasites of Panama (R. L. Wenzel and V. J. Tipton, eds.). Field Mus. Nat. Hist., Chicago, xii + 861 pp. Werneck, F. L. 1952. Contribuigao ao conhecimento dos Anopluros. I. Rev. Brasil. Biol., 12:69-78. Wood, A. E. 1931. Phylogeny of the heteromyid rodents. Amer. M us. Novit., 501:1-19. _ 1935. Evolution and relationship of the heteromyid rodents with new forms from the Tertiary of western North America. Ann. Carnegie Mus., 24:73-262. Wood, A. E„ and R. W. Wilson. 1936. A suggested nomenclature for the cusps of the cheek teeth of rodents. J. Paleont., 10:388-391. Wrenn, W. J., and R. B. Loomis. 1967. Otorhinophila, a new genus of chiggers (Acarina: Trombiculidae) from western North America. Acarologia, 9:152-178. Yunker, C. E., and J. M. Brennan. 1962. Endoparasitic chiggers: II. Rediscovery of Doloisia synoti Qudemans, 1910, with descriptions of a new subgenus and two new species (Acarina: Trombiculidae). Acarologia, 4:570-576. Yunker, C. E., and F. J. Radovsky. 1966. The dermanyssid mites of Panama (Acarina: Dermanyssidae). Pp. 83-103, in Ectoparasites of Panama (R. L. Wenzel and V. J. Tipton, eds.). Field Mus. Nat. Hist., Chicago, xii + 861 pp. 340 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY APPENDIX I— GAZETTEER Names of geographic and topographic features listed below are those used in the text of this report. The primary sources for spellings and coordinates of lo¬ calities were the gazetteers of the United States Board on Geographic Names (prepared by the Office of Geography, Department of Interior) for British Hon¬ duras (no. 16 issued March 1956), Costa Rica (no. 18 issued March 1956), Guate¬ mala (no. 94 issued September 1965), El Salvador (no. 26 issued October 1956), Honduras (no. 27 issued October 1956), Mexico (no. 15 issued February 1956), Nicaragua (no. 25 issued October 1956), and Panama and the Canal Zone (no. 1 10 issued June 1969). If place-names were not found in the above-listed gazet¬ teers, other sources used to locate them included: 1) The American Geographic Society’s “Map of Hispanic America. ...” scale 1:1,000,000, and its accom¬ panying Index (1944); 2) Estados Unidos Mexicanos, Comision Intersecretarial Coordinadora del Levantamiento de la Carta Geografica de la Republica Mexi- cana, 1:500,000 (1957-1958); 3) a collection of maps entitled “Caminos de Mexico,” published by Compania Hulera Euzkadi, S. A., third edition (1967); 4) Mexico, Mapa Turistico de Carreteras elaborado por la Secretarfa de Obras Publicas con la colaboracion de Secretarfa de Comunicaciones y Transportes, Departamento de Turismo, y Petroleos Mexicanos (1966); 5) set of maps covering most of the country of Nicaragua, 1:50,000, available from Direccion General de Cartografia, Ministerfo de Fomento, Managua, Nicaragua; 6) Mapa de la Republica de Guatemala, compilado por Federico P. de Torroella, lito- grafia Arimany, Guatemala (1948); 7) Mapa de carreteras de Venezuela pro¬ duced by the Creole Petroleum Corporation in 1966; 8) and various road maps prepared by petroleum companies, most notable of which are those for the Cen¬ tral American countries prepared by Esso Standard Oil, S. A. Any entry listed from one of these secondary sources is marked with an asterisk (*). Places that could not be located directly, but that were located indirectly with reference to some other known feature, can be recognized because the word “approximately” is used in their designation. Names in brackets refer to other names or spellings of a given place that are frequently encountered in the litera¬ ture or on maps. All latitudes are north of the equator and longitudes west of Greenwich. Sequence of countries, of states within Mexico, and individual local¬ ities within the countries and states is alphabetical. Each locality listed for Texas is followed by the name of the county in which it is located, and localities in Cen¬ tral American countries are followed by the name of the departamento in which they are found. Other general gazetteers dealing with the region covered herein are those of Choate (1970:312-317), Davis (1944:371-374), Hooper (1947:40-42, 1952: 220-249), and Musser (1964:1-5). Several of these contain brief ecological de¬ scriptions of localities. Other useful gazetteers that deal with individual states or countries are cited under the appropriate subheading below. GENOWAYS— SYSTEMATICS OF LIOMYS 341 British Honduras Roaring River, Cayo- 17° 16', 88°48'* Spanish Lookout, Cayo - 17° 13', 88°59' Costa Rica Altos Escazu, San Jose — approximately 9° 56', 84°09' (heights above the town of Escazu) Bagaces, Guanacaste- 10°3 1', 85° 15' Bebedero, Guanacaste - 10°22\ 85° 12' Boca del Barranca, Puntarenas - 9°58\ 84°44'* Congrejal, San Jose — approximately 9°46', 84° 12' (this locality is 5 mi. SSW San Ignacio, according to A. B. McPherson) Esparta, Puntarenas - 9°59', 84°40' Filadelfia, Guanacaste - 10°26', 85°34' Finca Jimenez, Guanacaste — approximately 10° 20', 85° 12' (according to E. T. Hooper, Finca Jimenez is a part of the larger Finca Taboga and straddles the Rio Hig- ueron, 6 mi. S, 6 mi. W Las Canas) Las Canas, Guanacaste - 10°25', 85°07' Liberia, Guanacaste - 10°38', 85°27' Los Higuerones, San Jose - approximately 9° 56', 84°09' (small farms about Escazu) Monte Verde, Puntarenas — approximately 10" 19', 84°49' (a Quaker settlement near the junction of the provinces of Alajuela, Guanacaste, and Puntarenas, according to A. B. McPherson) Pandora, Limon - 9°45', 82°57' Playa del Cocos, Guanacaste - 10°33', 85°43' Puerto Viejo, Heredia- I0°26\ 83°59' Rio Teopisco, 30 mi. S Nicaraguan border, Guanacaste — approximately 10°47', 85° 33' Sabanilla de Pirn's, San Jose — 9°44', 84° 1 6' San Francisco Esparta, Puntarenas— a town adjoining Esparta (see that locality for coordinates) San Ignacio, San Jose - 9C48', 84°09' San Jose, San Jose - 9°56', 84°05' San Juanillo, Guanacaste- 10°02', 85°44' Santa Ana, San Jose - 9°56', 84° 1 1 ' Santa Cruz, Guanacaste — 10° 16', 85° 36' Santa Rosa, Guanacaste — 1 0° 5 1 ', 85°38' Tempate, Guanacaste - 10 24', 85 44' Tilaran, Guanacaste— 10c28', 84°59' Turrialba, Cartago- 9°54', 83 41' Villa Colon, San Jose - 9 55', 84 14' For additional information on localities in Cost Rica see Harris (1943:1-6) and Goodwin ( 1 946:454-458). El. Salvador Amate de Campo [Laguna Limpia], La Paz- approximately 13°24\ 89°07' Barra de Santiago, Ahuachapan — 1 3°47', 90°03' Divisadero, Morazan- 13°36', 88°04' El Tablon, Santa Ana— approximately 14" 15', 89°35' (this locality is on SE shore Lake Guija) Finca Raquelina, Ahuachapan — approxi¬ mately 1 3°52', 89°48' Hacienda Chilata, Sonsonate - 1 3°40', 89° 32' Hacienda Nancuchiname, Usulutan- ap¬ proximately 13°25', 88°41' Hacienda San Antonio, Sonsonate — approx¬ imately 1 3°42', 89°45' Isla de la Cabra, Santa Ana-13°51', 89°34'* (in Lake Coatepeque) Laguna de Guija, Santa Ana - 14° 13', 89° 34' Lake Coatepeque, Santa Ana— 13°52', 89° 33' Lake Olomega, San Miguel - 1 3° 1 9', 88°04' Los Planes [Planes de Renderos], San Sal¬ vador- 13°39', 89° 11' Monte Cristo Mine, Morazan - approxi¬ mately 13°36', 88°05' (this mine is Wi mi. W Divisadero) Mount Cacaquatique, San Miguel and Mora¬ zan- 13°46\ 88° 13' Pine Peak, La Union — approximately 13° 17', 87°55' (this peak is 3 mi. W Volcan Cochagua) Puerto del Triunfo, Usulutan— 13° 17', 88° 33' Rio Goascoran, La Union — approximately 1 3 0 3 1 ', 87°44' Rio San Miguel, San Miguel - approxi¬ mately 1 3°25', 87°44' San Salvador, San Salvador — 1 3°42', 89° 1 2' Santa Tecla [Nueva San Salvador], La Libertad- 13°41', 89° 17' Volcan San Miguel, San Miguel - 1 3°26'. 88° 1 6' For additional information on localities in El Salvador see Burt and Stirton (1961: 8-17). 342 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY Guatemala Antigua, Sacatepequez- 14°34', 90°44' Astillero, Santa Rosa- 13°51', 90°21' Cabanas, Zacapa- 14°56', 89°48' Cerro las Flores, Finca El Carnero, Jutiapa- 14° 17', 89°59'* Chiquimulilla, Santa Rosa- 14°05', 90°23' Concepcion del Mar, Escuintla - approxi¬ mately 14°00', 91°30' (information about this locality was supplied by Joseph C. Moore) Cuilapa, Santa Rosa - 1 4° 1 7', 90° 1 8' Duehas, Sacatepequez — 14° 31', 90° 48' El Rancho, El Progreso - 14°55', 90°00' Escuintla, Escuintla- 14° 18', 90°47' Finca El Progreso, Santa Rosa-14°13', 90°26' Finca Los Arcos, Escuintla— 14° 14', 90° 51'* Finca Valle-Lirios, Escuintla — approxi¬ mately 13°56', 91u00' (information about this locality was supplied by C. O. Handley, Jr.) Granados, Baja Verapaz- 14°55', 90°31' Gualan, Zacapa - 15°08', 89°22' Guatemala, Guatemala- 14°38', 90°31' Hacienda California, San Marcos- 14°33', 92° 1 0' Iztapa, Escuintla - 13°56', 90°43' Jocotan Chiquimula- 14°49', 89°23' Km. 24 on highway S Guatemala City, Guatemala - approximately 14°28', 90° 38' Km. 27 on highway S Guatemala City, Guatemala - approximately 14°27', 90° 39' Km. 33 on highway S Guatemala City, Escuintla — approximately 14°25', 90° 40' Km. 52 on highway S Guatemala City, Escuintla - approximately 14° 19', 90° 46' Lake Amatitlan, Guatemala — 1 4°27', 90° 34' Lake Atescatempa, Jutiapa - 14° 1 2', 89°42' La Primavera, Alta Verapaz - approxi¬ mately 15°25', 90°29' (see Stuart, 1950: map 2, for location of this finca) Masagua, Escuintla- 1 4° 1 2', 90°51' Mazatenango, Suchitepequez — 14°32', 91° 30' Monogoy [MongoyJ, Jutiapa- 14° 15', 89° 43' Nenton, Huehuetenango - 1 5 0 48', 91 45' Progreso, El Progreso - 14° 51', 90° 04' Puebla, Izabal - 15° 18', 89°00'* Puente Punta Gordo, El Progreso — approxi¬ mately 14°48', 90° 12' (this bridge is lo¬ cated at km. 51 on the highway from Guatemala City to Puerto Barrios, ac¬ cording to Jones, 1966:469) Rabinal, Baja Verapaz - 15°06', 90°27' Rio Grande, 152 km. NE Guatemala City, Zacapa- approximately 15°06', 89°24' (this locality is approximately 17 mi. NE Rio Hondo) Sacapulas, El Quiche - 15°20', 91°04' Salama, Baja Verapaz - 15°06', 90° 16' San Cristobal [Frontera], Jutiapa- 14°1 1', 89°40' San Jose, Escuintla — 1 3° 55', 90° 49' San Lucas, Solola- 14°38', 91°08' Santiago Sacatepequez, Sacatepequez - 14° 38', 90°41 ' Tiquisate, Escuintla - 14° 1 7', 9 1 °22' Trujillo, Zacapa- 14°57', 89°49' Volcan San Lucas, Solola- 14°37', 91°07'* Yepocapa, Chimaltenango - 14°30', 90°57' For additional information on localities in Guatemala see Griscom (1932:413- 425), Stuart (1954:36-41), and Jones (1966:467-469). Honduras Catacamas, Olanche- 14°54', 85°56' Cerro de los Cuches, Francisco Morazan — approximately 13°48', 87° 14' (this locality is 4 km. SE Sabana Grande) Comayagiiela, Francisco Morazan— 14° 05', 87° 1 3' El Caliche Orica, Francisco Morazan — approximately 14°31', 87°03' (this locality is W Orica) El Zapote, Francisco Morazan - approxi¬ mately 1 3°45', 87° 15' (this locality is 7 km. S Sabana Grande) Escuela Agricultura Panamericana, Fran¬ cisco Morazan - 1 3°59', 86°59' (this school is on the Tegucigalpa-Danli road on the Yeguare River)* Hatillo, Francisco Morazan - 14°88', 87° 10' La Cueva Archaga, Francisco Morazan - approximately 14° 17', 87° 13' (this locality is near Cerro Archaga) La Flor Archaga, Francisco Morazan - approximately 14° 15', 87° 15' (this is a village on Talanga road E Cerro Archaga) GENOWAYS— SYSTEMATICS OF LIOMYS 343 La Piedra de Jesus, Francisco Morazan - approximately 13°47', 87° 1 5' (this place is 5 km. S Sabana Grande) La Venta, Francisco Morazan - 14° 1 8' 87° 10' Monte Redondo, Francisco Morazan — approximately 14° 16', 87° 15' (this locality is 15 mi. NNW Tegucigalpa, according to Hooper, 1952:227) Sabana Grande, Francisco Morazan— 13° 50', 87° 15' San Pedro Sula, Cortes- 15°27\ 88°02' Tegucigalpa, Francisco Morazan - 14°06\ 87° 13' Toncontin, Francisco Morazan - 14°02', 87° 13'* For additional information on localities in Honduras see Goodwin (1942:108-1 1 1). Mexico Aguascalientes Aguascalientes- 21°53', 102°18' Calvillo — 21°51', 102°43' Chicalote — 22°02', 102°16' La Labor- approximately 21°58', 102°42' (according to the specimen label, this locality is 9 mi. by road N Calvillo) Campeche Champoton- 1 9° 2 1 ', 90°43' Escarcega- 18°37', 90°43' Chiapas Altamirano— 16°53', 92°09' Arriaga- 16°14', 93°54' Berriozabal — 1 6°48', 93° 1 6' Bochil- 1 6°59', 92°55' Cacahuatal - approximately 16°29', 93°59' (this locality is 3 km. eastward from Rizo de Oro or Los Amates, 16°28', 94°00'*, on the road to Tuxtla Gutierrez, accord¬ ing to information supplied by A. L. Gardner) Cerro Tres Picos — 1 6° 1 2', 93° 37'* Chiapa de Corzo [Chiapa] — 1 6°42', 93 00' Cinco Cerros — approximately 16°31', 93 57' (this place is approximately 6 to 8 km. from Rizo de Oro toward Tuxtla Gutier¬ rez, according to information supplied by A. L. Gardner) Cintalapa- 16°44', 93°43' Ciudad Cuauhtemoc - 15°37', 92 00' Comitari — 1 6° 1 5', 92°08' Escuintla- 15°20', 92°38' Finca El Paraio- 17°07', 91°54'* Guatimoc [Cuatimocl - 15°02', 92°09' Huehuetan- I5°0I', 92°22' Huixtla - 1 5°08', 92°28' Jaltenango - 15°55', 92°43' LaTrinitaria- 16°07', 92°03' Madre Mia- 16°08', 93°39' Maspastepec- 15°26', 92°54' Ocozocoautla- 16°46', 93°22' Ortiz Rubio- 16°22', 93°52'* Palenque — 1 7°3 1 ', 91°58' Pan-American Highway, 16 mi. N Guate¬ malan border - approximately 15° 50', 91° 58' Pijijiapan - 15°42', 93° 14' Pueblo Nuevo [Pueblo Nuevo Solista- huacan] - 17°06', 92°53' Puerto Arista- 15°56', 93°48' Puerto Madero- 14°44', 95°25' Sabena de San Quintin - 16°24', 91 °20' San Bartolome [Venustiano Carranza] - 1 6°2 1 ', 92°33' San Clemente - 16°26', 93°43' San Gregorio- 15°55', 92° 10' San Miguel — approximately 16° 14', 93°34' San Vincente- 16°00', 91°46' Talisman- 14°58', 92°09' Tapachula - 14°54', 92° 17' Tonala- 16°04', 93°45' Tuxtla Gutierrez— 16°45', 93°07' Union Juarez - 1 5° 04', 92° 05' Valley of Jiquipilas — approximately I6°40', 93°35' (according to Goldman, 1951:116, specimens labeled in this manner were taken along a tributary of the Rio de la Venta at Rancho San Ricardo) Villa Flores - 16° 14', 93° 14' Chihuahua Batopilas- 27°01', 107°44' Jimenez- 27°08', 104°55' Parral IHidalgo del Parral ]- 26°56', 105° 40' Santa Rosalia [Ciudad Camargo] - 27°40', 105° 10' Colima Agua Zarca - 19° 13', 103°55' Armeria- 18°56', 103°58' Cerro Chino - 1 9° 12', 104°00'* Colima- 19°14'. 103°43' Comatlan [de Miraflores] - 19° 1 3', 104° 14'* 344 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY Hacienda Magdelena [Pueblo Juarez] - 19° 10', 103°55'* Hacienda San Antonio - 19°27\ 103°44' (this locality is 26 mi. N Colima, accord¬ ing to Goldman, 1951:136) La Gloria- approximately 19° 10', 104° 12' (this place is 29 km. W Pueblo Juarez, according to specimen label) Las Juntas- approximately 19°10', 104°10' (this locality is 26 km. W Pueblo Juarez, according to specimen label) Las Juntas- approximately 19°07', 103°53' (this locality is 5 km. SE Pueblo Juarez, according to specimen label) Las Lomas- approximately 19° 11', 103°45' (this place is 3 or 4 mi. S Colima on highway to Manzanillo, according to A. L. Gardner) Manzanillo- 19°03', 104°20' Paso del Rio- 18°57', 103°56' Pueblo Juarez - see Hacienda Magdelena Queseria- 19°22', 103°35' Santiago- 19°07\ 104°21' Tlapeixtes - approximately 19°04', 104° 18' (this locality is 4 km. ENE Manzanillo, according to Gardner, 1966:4-5) Trapichillos- 19°08', 103°30'* For additional information on localities in Colima see Schaldach (1963:7-12). Distrito Federal Cerro Xaltepec - 19° 1 9', 99°02'* Ciudad Universitaria - 1 9°20\ 99° 1 1 '* Contreras - 19° 18', 99° 17' Mexico - 1 9°24', 99°09' Pedregal de San Angel - 1 9° 1 9', 99° 1 1 ' (this is the pedregal to the south of Ciudad Universitaria)* San Geronimo [San Jeronimo] - 19°20', 99° 14'* San Gregorio Altapulco — 99°03', 1 9° 1 5'* San Mateo Xalpa [San Mateo Jalpa] - 19° 14', 99°07'* Tlalpan [Tlalpam] - 19° 17', 99° 10' Xochitepec — 19° 16', 99°07'* Zapotitlan - 19° I 8', 99°02' Durango Canatlan - 24° 31', I04°37' Chacala— 24c50', I06°45' Chorro — 24° 1 6', 104°27' Durango - 24°02', 104°40' Inde - 25°54', 105° 13' Navarro -26°42', 106° 12' Pueblo Nuevo- 23°23', 105°23' Rio Sestin - approximately 26°07', 105°40' (this is the location of the town of Sestin, which is probably near the collecting site of J. H. Batty, see J. A. Allen, 1903:591) Rodeo- 25° IT, 104°34'* Rosario -26° 31', 105°40' Tepehuanes [Santa Catarina de Tepe- huanes] -25°21\ 105°44' For additional information about local¬ ities in Durango see Baker and Greer (1962:54-55). Guanajuato Acambaro — 20 02', 100°43' (railroad sta¬ tion) Moroleon — 20°08', 101° 12' Salvatierra- 20° 13', 100°53' San Jose [San Jose Iturbide] -21°00', 100° 23' Silao - 20°56', 101 °26' Guerrero Acahuizotla— 17°23', 99°27' Acapulco- 16°51', 99°55' Acapulco Bay- 16°50', 99°53' Agua del Obispo - 17° 13', 99°31'* Almolonga- 17°36', 99° 16' Apaxtla- 1 8°09', 99°52' Ayotzinapa [Ayusinapa on specimen label] - 17°33\ 98°45' Cacahuamilpa- 18°41', 99°30' Chapa- 18° 19', 99°49' Chilpancingo- 17°33', 99°30' Cocula- 18° 14', 99°40' Colotlipa — 17°25', 99°09' El Limon- 18°05', 101 °40' El Rincon- 17° 1 9', 99°28' Iguala — 1 8 ° 2 1 ', 99°32' Infiernillo — 18°09', 102°09' Los Sabinos- 18°22', 99°46'* Mazatlan- 17°27', 99°29' Mexcala [Mezcala] - 17°56', 99°37' Ojo de Agua de Chapa — approximately 18 1 9', 99°51' (this locality is 5 km. SE Teloloapan, according to specimen labels) Ometepec- 16°41', 98°25' Omilteme [Omiltemi] - 17°30', 99°40' Pie de la Cuesta- 16°54', 99°58' Rio Balsas [Balsas] - 17°59', 99°47' Sacacoyuca- 1 8° 14', 99° 33' GENOWAYS— SYSTEMATICS OF LIOMYS 345 Sihuatanejo Bay [Bahia de Zihuatanejo] — 17°37\ 101 °34' Sochi [apparently an abbreviation for Xochihuehuetlan] - 17°55', 98°26' Taxco [de Alarcon] - 18°33\ 99°36' Teloloapan- 1 8° 2 1 99°51' Texcalzintla- approximately 18°25', 99°54' (this locality is 6 km. NNW Teloloapan, according to specimen labels) Tierra Colorada- 17° 10', 99°35' Tixtla — 17°35', 99°26' Tlalixtaquilla — 17°21', 98°28' Tlapa— 1 7°33', 98°33' Tomatal - 18° 19', 99° 30' Xochipala — 17°49', 99°37' Yerbabuena- 18°27', 99°54'* Zacatula- 17° 59', 102°09' Zihuatanejo- 17°38', 101°33' Zumpango [del Rio] - 17°39', 99°30' For additional information on some local¬ ities in Guerrero see Davis and Dixon (1959: 79-80). H idalgo Actopan— 20° 16', 98°56' Epazoyucan - 20°02', 98° 38' Ixmiquilpan — 20°29', 99°14' Maguey Verde - approximately 20°49', 99° 16' (this place is 8V2 mi. NE Zimapan, according to specimen label) Marques - 20° 19', 99° 35'* San Agustin - 20°04', 99°05' Tula [de Allende] -20°03', 99°21' Zacaultipan - 20°39', 98° 36' Zimapan -20°45', 99°21' Jalisco Acatlan [de Juarez] -20°26', 103°38' Amatitan — 20°50', 103°43' Ameca — 20°33', 104°02' Arroyo de Gavalan — approximately 20°48', 104°29' (supposedly 20 mi. W San Mar¬ cos, according to J. A. Allen, 1906:237; however, this may not be correct) Arroyo de Plantanar - not exactly located, but is near Amatlan de Canas, Nayarit (see J. A. Allen, 1906:237) Atemajac [del Valle] — 20°44', 103°20'* Atenqueque- 1 9 0 3 1 ', 103°30' Autlan [de Navarro] — 19°46', 104 22' Ayoel Chico- 20°32', 102°21' Barranca de Tule, Sierra Nayarit -not exactly located, but may be in vicinity of 2 1 °57', 1 03°56' Barro de Navidad [Barra de Navidad and Navidad] - 19° 12', I04°41' Belen de Refugio - 2 1 °3 1', 102°25' Bolaiios — 2 1 °41 ', 103°47' Chamela Bay - 19°33', 105°07' Chapala - 20° 18', 103° 12' Cihuatlan- 19°14', 104°35' Ciudad Guzman [Zapotlan] - 1 9°4I ', 103° 29' Contla - 1 9°45', 103°05' Comanja de Corona - 2 1 0 1 9', 1 0 1 °42' Cuitzamala [Cuitzmala] - 19°23', 104°59'* Durazno-not exactly located, but approxi¬ mately 19 32', 104° 16' (based on M. R. Lee’s field notes) El Zapote — 20° 3 1', 103° 18' Emiliano Zapata — approximately 1 9 0 2 1 ', 104°58' Encarnacion [deDiaz] -21°31', 102° 14' Estancia-not exactly located, but is near Amatlan de Canas, Nayarit (see J. A. Allen, 1906:237) Etzatlan — 20°46', 104°05' Guadalupe de Victoria- 21 °37', 101 °38' (given as Matanzas on current maps, ac¬ cording to information supplied by R. G Webb) Guadalajara- 20°40', 103°20' Huascato — 20° 32', 102° 14' Huejucar — 22°21', 103°13' Huejuquilla [el Alto] - 22°36', 103°52' Ixtapa— 20°42', 105°15' Jalostotitlan - 21° 12', 102°28' Jazmin- 19°39', 103°48' Jilotlan de los Dolores- 1 9° 1 2', 103°13' Jocotepec - 20° 1 8', 103°26' La Cuesta— 20° 10', 104°51' La Cumbre - approximately 19°42', 104°22' (this place is 9 mi. SSW Autlan de Navarro, according to Schaldach, 1963: 12) Lagos de Moreno - 2 1 °2 1 ', 101 0 5 5' La Laguna, Sierra de Juanacatlan - 20°39', 1 04°47' La Primavera - 20°40', 103°32'* La Resolana- 19° 34', 104°31' Las Canoas- 19°37', 103°32' (this locality is 40 mi. W Tuxpan, according to J. A. Allen, 1906:238; however, I feel that this is more likely the small town with the same name that is located 10 km. NW Tuxpan) La Venta- 20°44', 103°33' Limon- 1 9°52', 104°07' 346 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY Magdelena— 20°55\ 103°57' Mascota- 20°32\ 104°49' Mazamitla- 19°55', 103°02' Melaque — 19°13', 104°43' Mezquitic— 22°23', 103°4L Milpillas — 20°44', 104°56' Navidad — see Barro de Navidad Navidad Bay- 19° 12', 104°43' Nevado de Colima- 19°33', 103°38' Ocotlan - 20°21', 102°46' Piguamo- see Pihuamo Pihuamo [Piguamo] - 19° 15', 103°23' Platanar [Plantinar] - 19°28', 103°27' Puerto Vallarta — 20°37', 105° 15' Purificacion - 19°43', 104°38' Rio Ameca- not exactly located, but is near Amatlan de Canas, Nayarit (see J. A. Allen, 1906:237) Rio Santa Maria- not exactly located, but is near Amatlan de Canas, Nayarit (see J. A. Allen, 1906:237) Sal se Puedes-not exactly located, but is in mountains of west-central Jalisco, possibly near Talpa de Allende (see J. A. Allen, 1906:237) San Gabriel [Venustiano Carranza] — 19° 44', 1 03°47' San Marcos- 20°47', 104° 11' ( Liomys irroratus locality) San Marcos- 19°25', 103°28 ' (Liomys pictus locality) San Sebastian - 20°47', 104°51' Santa Cruz de las Flores- 20°29', 103°30' Soyatlan del Oro- 20°20', 104° 1 5' (informa¬ tion about this locality was supplied by A. L. Gardner)* Tala — 20°40', 103°42' Tamazula - 19° 38', 103° 15' Tapalpa- 19°57', 103°46' Tecalitlan — 19°26', 103° 15' Tecomate- 19° 35', 104° 28' Tenacatita- 19° 17', 104°47' Tenacatita Bay — 19° 17', 104°50' Tepatitlan [de Morelos] — 20°49', 102°44' Tequila- 20°54', 103°47' Teuchitlan — 20°41', 103°52' Tizapan [Tzipan on specimen labels] -20° 10', 1 03°04' Tonala — 20°37', 103° 14' Tuxpan — 19°33', 103°24' Union de San Antonio - 2 1 °06', 101 0 58' Villa Guerrero- 21 °59', 103°36' Wakenakili Mountains - not exactly located, but is in the mountains of west-central Jalisco, possibly near Talpa de Allende (see J. A. Allen, 1906:238) Yahualica— 21 °08', 102°51' Zapotiltic — 19°37\ 103°26' Zapotlan - see Ciudad Guzman Zapotlanejo- 20°38', 103°04' Mexico Atlacomulco [de Fabela] - 19°48', 99°53' Ayotla- 19° 19', 98°55' Barrientos- 19°35', 99°1 1'* Cerro la Caldera - approximately 19°20', 98°57' Chilpa- 1 9°38', 99°10'* Ecatepec [Morelos] - 19° 35', 99°04' Hacienda Cordoba - approximately 19°20', 98°43' (this hacienda is 35 mi. ESE Mex¬ ico City on Mexico City-Puebla highway, according to Hooper, 1947:42) San Rafael - 19° 14', 98°45' Temascaltepec [de Gonzalez] - 19°02', 100°03' Teotihuacan [San Juan Teotihuacan] — 19°41', 98° 52' Tenancingo [de Degollado] - 18°58', 99°35' Texcoco [de Mora] - 19°31', 98°53' Tlalpitzahuac [Tlalpizahua on specimen label] - 19°20', 98°56' Tlalmanalco [de Velaquez] — 1 9° 1 3', 98° 48' Tlalnepantla [de Galeana] - 19°33', 99° 12' Valle de Bravo- 1 9° 1 1 ', 100°08' Zoquiapan - 1 9° 1 9', 98°51 ' Michoacan Agua Blanca- not exactly located, but ac¬ cording to W. H. Osgood's field notes this locality is “near Jungapeo f 1 9° 27', 100° 29'] and Zitacuaro [19°26\ 100°21']” (information supplied by Joseph C. Moore) Apatzingan- 19°05', 102° 15' Arteaga- 18°28', 102°25' Carapan- 19°52', 102°03' Coalcoman- 18°47', 103°09' Cuitzeo [del Provenir] - 19°59', 101 °09' Dos Aguas- approximately 18°45', 102°55' (this lumber camp is on the eastern slope of Cerro de Barolosa, located about 22 km. WNW Aguil ilia, according to Duell- man, 1961:135) El Atuto - approximately 18° 10', 102°09' (this locality is 3 km. NE Infiernillo, Guerrero, according to specimen label) GENOWAYS— SYSTEMATICS OF LIOMYS 347 Jacona- 19°57\ 102° 16' La Huacana- 18°58\ 101 °49' La Mira- 18°01', 102°20' La Palma- 20°09', 102°46' La Salada- approximately 18°51', 101 °55' (this ranch is 40 mi. S Uruapan, according to Goldman, 1951:190) Los Reyes - 19° 35', 102°29' Melchor Ocampo- 17°59\ 102° 1 1' Morelia- 19°42\ 101°07' Patzcuaro- 19° 31', 101° 36' Querendaro- 19°53', 100°54' (railroad station) Quiroga- 19°40', 101°32' Tacambaro- 19°14', 101°28' Tancitaro- 19°20', 102°22' Tarecuato— 19°51', 102°29' Tinquindi'n [de Argandar] - 19°45', 102°29' Tumbiscatio- 1 8 ° 3 1 ', 102°21' Tzitzio- 19°34', 100°55' Zamora [de Hidalgo] - 19°59', 102° 16' For additional information on localities in Michoacan see Hall and Villa (1949:438) and Duellman (1961:129-141). Morelos Alpuyeca- 18°44\ 99° 16' Axochiapan- 18°30', 98°46' Cuautla- 1 8°48', 98°57' Cuernavaca— 18°55', 99° 15' Huajintlan — 18°36', 99°25' Huitzilac- 1 9°02', 99° 16' Jonacatepec- 18°41', 98°48' Michapa- 18°42', 99°29'* Puente de Ixtla- 18°37', 99°20' Tehuixtla— 18°33', 99°16' Temilpa- 18°41', 99°06' Temixco — 18°50', 99° 14' Tepoztlan — 18°59', 99°06' Tequesquitengo — 18°36', 99° 16' Tetecala — 18°43', 99°24' Yautepec- 18°53', 99°04' Nayarit Acaponeta— 22°30', 105°22' Aticama - 2 1 °29', 105°13' Amatlan de Canas - 20' 52', 104°27' Compostela — 2 1 0 1 4', 1 04° 55' Huajicori - 22° 39', 105° 18' Ixtlan del Rio-21°02', 104°22' Jalisco- 21 °27', 104°54' Jalisco-Nayarit border on Mexican Highway 1 5 - 21 °03', 104° 1 3'* La Cuchara — approximately 22°30', 104° 49' (according to specimen label, this locality is “approximately 40 mi. E Acaponeta”) Las Varas- 21° 10', 105° 10' Navarrete- 21°39', 105°07' Ojo de Agua— not exactly located, but is near Amatlan de Canas (see J. A. Allen, 1906:237) Pedro Pablo - approximately 22°29', 105° 05' (according to Goldman, 1951:202, this village is 22 mi. E Acaponeta along the western slope of the Sierra de Tepona- huasta) Platanares- 21 °57', 105°01' Playa Novilleros - approximately 22°22', 105°39' Rancho Palo Amarillo- not exactly located, but is near Amatlan de Canas (see J. A. Allen, 1906:237) Rosamorada [Rosa Morada] - 22°08', 105° 12' Ruiz- 21°57', 105°09' San Bias — 21 °31', 105°16' San Francisco- 20°57', 105°21' San Jose del Conde - 21 °05', 104°44' San Miguel- 20°20', 105° 17' Santa Isabel - 21° 10', 104°37' Santiago- 21 °49', 105°13' Tepic — 2 1 °30', 104°54' Tuxpan — 21 °57', 105° 18' Valle de Banderas- 20°49', 105° 17' Nuevo Leon Allende — 25° 17', 100°01' Aramberri - 24°06', 99°49' Ascension — 24°20', 99°55' Cerro de la Silla - approximately 25°39', 100° 14' (this mountain is 6 mi. E Mon¬ terrey, according to Goldman, 1951:206) China- 25°42\ 99° 14' General Teran — 25° 1 6', 99°4I' Ibarilla — 24°28', 99°4I'* Iturbide— 24°44', 99°44' Linares- 24° 52', 99° 34' Montemorelos - 25° 1 2', 99°49' Monterrey - 25°40', 100°19' Ojo de Agua-24°55', 100° 11'* Pablillo — 24°36', 99°59' San Francisco (Javier] -24°59', I00°2I' San Josecito Cave - near Aramberri San Pedro Santiago- 25°22', 100°07' Zaragoza - 23°58', 99°46' 348 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY Oaxaca Agua Blanca- 16°4T, 95°45' Aguaje Guayabo [Portillo Guayabo] - approximately 16°61\ 95°33' Aguaje las Animas - approximately 16° 15', 95°22' Aguaje Sapotal - approximately 16°21\ 95° 32' Aguaje Tres Cabezas— approximately 16° 21', 95°27' Angeles Bay - 15°40\ 96°28'* Buena Vista [San Dionisio del Mar] — 16° 20', 94°45' Candeleria— 15°54', 96°30'* Cerro Calderona- 16°23', 95°21'* Cerro de Tigre - approximately 16°21', 95° 14' Cerro Ocotepec - approximately 16°25', 95° 43' Cerro Palma de Oro - approximately 1 6° 1 8', 95°32' Cerro San Felipe - 17° 10', 96°41'* Cerro San Pedro — approximately 16° 18', 95°28' Cerro Sombrerito — approximately 16° 16', 95°27' Cerro Verde- approximately 16°35', 95°36' Chacalapa- 15°49', 96°28' Chahuites - 1 6° 17', 94° 1 1 ' Chicapa- 16°26', 94°49' Chivaguela- 16°28', 96°12' Chontecomatlan - 16°15', 96°01' Cuicatlan- 17°48', 96°58' Cycad Camp [km. 233 Oaxaca-Puerto Es¬ condido Road] - approximately 15°59', 97°06' Ejutla [de Crespo] - 16°34', 99°44' El Bambita- approximately 16°20', 95°35' El Campanario- 16°45', 95°07'* Escondido Bay— 1 5 ° 5 1 ', 97°05'* Escurano- approximately 16°25', 95°27' Gueladu - approximately 1 6°29', 95° 1 8' Guichicovi- 16°58', 95°06' Guichina- approximately 16°27', 96°06' Guiengola- 16°21', 95°20'* Huajuapan de Leon [Huajuapam] - 17° 48', 97°46' Huilotepec- 16° 14', 95°09' Isthmus of Tehuantepec, 8 mi. S Veracruz- Oaxaca state line — approximately 17° 15', 95°03' Ixcuintepec - 16°56', 95°42' Jalapa - 16°30', 95°28' Jamaica Junction [km. 212 Oaxaca-Puerto Escondido Road] - approximately 16°09', 97°06' Jamiltepec— 16°17', 97°49' Jicotlan - 17°48', 97°28' Juchatengo- 16°21', 97°06' Juchitan — 16°27', 95°01' Km. 136 Oaxaca-Puerto Escondido Road [Turpentine Camp] — approximately 16° 24', 97°05' Km. 177.2 Oaxaca-Puerto Escondido Road — approximately 16° 17', 97° 07' Km. 183 Oaxaca-Puerto Escondido Road — approximately 16° 16', 97°07' Km. 198 Oaxaca-Puerto Escondido Road — approximately 16° 12', 97°06' Km. 212V2 Oaxaca-Puerto Escondido Road - approximately 16°09', 97°06' Km. 214 Oaxaca-Puerto Escondido Road [Rio Ranas] — approximately 16°07', 91° 06' Km. 123 Putla-Tlaxiaco Road — approxi¬ mately 17°05', 97°51' Km. 135 Putla-Tlaxiaco Road — approxi¬ mately 1 7°03', 97°53' La Chacahua- approximately 16°23', 95° 59' Lachao — see San Juan Lachao La Cima [km. 184 Oaxaca-Puerto Escon¬ dido Road] - approximately 16° 16', 97° 07' La Concepcion - approximately 17°02', 97° 52' Lagunas - 16°48\ 95°04' La Mata- 16°38', 95°00' La Parada- approximately 17°08', 95°30' (this localtiy is 15 mi. NE Oaxaca, ac¬ cording to Goldman, 1911:54) Las Cuevas- approximately 16°26', 95°2L Las Lagunas de Sol y Luna — approximately 1 6°28', 94° 1 2' Las Minas - approximately 16°21', 94°07' Las Tejas- approximately 16°20', 95°22' La Ventosa- 16°33', 94°57' Limon - approximately 16°20', 95°29' Llano Grande- 16°28', 98° 18'* Matatlan [Santiago Matatlan] - 1 6° 52' 96°53' Mesones- 16°55', 97°58' M iahuatlan [de Porfirio Diaz] - 16°20' 96°36' Mitla— 1 6°55', 96°24' M ixtequilla - 16°22', 95° 15' Monte Alban - 17°02\ 96°45'* GENOWAYS— SYSTEMATICS OF LIOMYS 349 Mount Zempoaltepec [Zempoaltepetl or Cempoaltepetl] - 17° 10', 95°59' Nejapa— 16°37', 95°59' Nizanda- 16°40', 95°02' Nopala [Santos Reyes Nopala] - 16°06\ 97° 10' Oaxaca [de Juarez] - 17°03\ 96°43' Ocotita- approximately 16°26', 95°23' Pinotepa- see Pinotepa Nacional Pinotepa de Don Luis- 16°25\ 97°55' Pinotepa Nacional [Pinotepa] - 16° 19', 98°01' Pluma [Pluma Hidalgo] - 15°55', 96°24' Portillo Guayabo— see Aguaje Guayabo Portillo Zapote - approximately 16° 18', 95°36' Pozo Rio - approximately 16°20', 95° 18' Puerto Angel - 15°40', 96°29' Puerto Escondido- 15°51', 97°06'* Putla — 17°02', 97°56' Reforma - 16°24', 94°28' Rio Chile— approximately 16°25', 95°37' Rio Grande— not exactly located, but may be near 16°21', 95°59' Rio Guajolote — not exactly located, but near the Rio Jalatengo and Rio Molino (see Schaldach, 1966:286-287) Rio Jalatengo [km. 178 on Oaxaca-Puerto Angel Road] - approximately 15°58', 96°27' Rio Molino - approximately 16°02', 96°29' (see Schaldach, 1966:286-287) Salazar- 16°25', 95°20 Salina Cruz — 16°10', 95°12' San Andres Chicahuaxtla — 17° 10', 97°50' San Antonio- 16° 15', 95°28' San Bartolo — see San Bartolo Yautepec San Bartolo Yautepec - 16°26', 95°59' San Gabriel Mixtepec— 16°05', 97°06' San Jose Lachiquiri — 1 6°23', 96°20' San J uan Acaltepec - 1 6° 1 6', 95° 56' San Juan Lachao [Lachao] — 16° 14', 97° 09'* San Juan la Jarcia- 16° 31', 95° 56' San Lucas Ixcotepec — 1 6° 1 8', 95 55'* San Miguel Caja de Agua- 16 09', 95°23' San Miguel Suchixtepec— 16°05', 96°28' San Pedro Jilotepec - 16°39', 95°44' San Vincente— 17°05', 97 53' Santa Catalina Quieri— 16°20', 96 15'* Santa Cruz Bay - 16°45', 96°08 * Santa Efigenia — 1 6°27', 94° 12' Santa Lucia - approximately 16 18', 95°28' Santa Maria Candelaria- 16° 12', 95°57'* Santa Maria Ecatepec- 16° 17', 95° 53' Santiago Lachiguiri- 16°41', 95°32' Santo Domingo - 16°49', 95°09' Santo Tomas Ocotepec - 1 7°08', 97°46' Santo Tomas Quieri - 16°21', 96° 10' Santo Tomas Teipan [Teipam] — 16° 15', 95°58' Sierra de San Felipe Lachillo — approxi¬ mately 1 5°59', 96° 12' Sierra Juarez - 17°23', 96°29'* Sierra Madre [15 mi. N Zanatepec] — approximately 16°41', 94°21' Solo de Vega- 16°31', 96°58' Tapanatepec- 16°21', 94°12' Tehuantepec- 16°21', 95° 14' Tenango- 16° 16', 95°36' Teotitlan [del Camino] - 18°08', 97°05' Tequisistlan — 16°25', 95°37' Tlacolula [de Matamoras] - 16°57', 96°29' Tlapancingo [Tlapacingo] - 17°28', 98° 14' Vincente- 18°31', 96°31'* Vista Hermosa — approximately 17° 36', 96° 22' Y aitepec [Santiago Yaitepec] - 1 6° 14', 97° 14' Yalalag- 1 7° 1 1 ', 96°1 1' Zacatepec- 16°46', 98°00' Zanatepec- 16°29', 94°21' Zapotitlan- 16°08', 95°43'* Zarzamora - approximately 16°21', 95°48' For additional information on localities in Oaxaca see Duellman (1960:33-35), Rowley ( 1 966: 1 09-1 1 2), and Goodwin ( 1 969: 256-265). Puebla Acatlan — 18° 1 2', 98°03' Amolac — 18°03', 98°23' Atlixco - 1 8°54', 98°26' Chila- 1 8°07', 98°29' Izucar de Matamoros - 1 8°36'. 98°28' Metlaltoyuca - 20°44', 97°51' Pahuatlan - 20° 1 7', 98°09' Piaxtla- 18° 12', 98° 15' Puebla [de Zaragoza] - 19°03', 98° 1 2' San Martin [San Martin Texmelucan] - 19° 17', 98°26' Tehuacan- 18°27', 97°23' Tehuitzingo- 18°21', 98° 17' Tepanco- 18° 34', 97° 33' Tilapa - 18°35', 98°33' Totimehuacan [San Francisco Totime- huacan] - 18°58', 98° 1 1' Zihuateutla - 20° 1 6', 97°53' 350 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY Queretaro Amoles [Pinal de Amoles] -21°09', 99°39' Cadereyta [de Montes] - 20°42', 99°49' Jalpan — 21° 14', 99°29' Pinal de Amoles — see Amoles Queretaro- 20° 36', 100°23' Tequisquiapan [Tequisquiapam] -20°31', 99°52' Toliman- 20°55', 99°56' Quintana Roo Felipe Carrillo Puerto- 19°35', 88°03' Pueblo Nuevo X-can - 20°50\ 87°43' Santo Rosa- 19°57', 88°23'* San Luis Potosi Ahualulco- 22°24', 101° 10' Ajinche - approximately 22°09', 98°25' Alvarez -22°03', 100°37' Apetsco - approximately 2 1 °24', 99°02' Arriaga- 21°54', 101°23' Axtla- approximately 21°28', 98°51' Bledos — 2 1 °5 1 ', 101°07' Cerro Campanario - 21 °54', 100°25'* Cerro Penon Blanco - 22°33', 101 °39'* Ciudad del Mafz-22°24', 99°36' Ciudad Valles [Valles] -21°59', 99°01' Ebano — 22° 1 3', 98°24' El Abra— 21 °58', 98°56' El Banito- not exactly located, but probably in the vicinity of Ciudad Valles (informa¬ tion supplied by Joseph C. Moore) El Naranjo [Naranjos] — 22°31', 99°20' El Sabinito - approximately 22°3 1 ', 99°22' El Salto - 22°36', 99°24'* Hacienda Angostura — not definitely iden¬ tified, but may be small railroad station 33 km. NNE Rio Verde at 22° 14', 100° 04' Hacienda Capulin - approximately 21°30', 99° 2>1' Hacienda La Parada- 22°21', 101° 13' Huichihuayan — 21 °30', 98°57' Jesus Maria- 21 °55', 100°54' Paso de San Antonio - 22°01 ', 100°24' Platanito- approximately 22°30', 99°27' Puerto de Lobos - approximately 22°29', 99°34' Ramos- 22°50', 101 °55' Rio Verde [Rioverde] - 21°56', 99° 59' San Carlos- 23°01', 100°21' San Luis Potosi - 22°09'. 100° 59' Santa Maria del Rio - 21°48'. 100°45' Tamazunchale- 21°16', 98°47' Tamuin — 2 1 °59', 98°45' Taninul - 2 1 °58', 98°54' Valles - see Ciudad Valles Villar - 22°33', 100°29' For additional information on localities in San Luis Potosi see Dalquest (1953:215- 223). Sinaloa Agua Nueva- 24°45', 107°14' Badiraguato — 25°22', 107°31' Camino Real-23°52', 106°39' Chele — 23° 1 3', 105°53' Choix — 26°43', 108°17' Comanito— 25°08', 107°39' Concepcion [La Concha] - 22°32', 105°28' Copala— 23°23', 105°56' Cosala — 24°23', 106°41' Costa Rica- 24°35', 107°25'* Culiacan - 24°48', 107°24' Culiacancito - 24°50', 107°32' El Cajon- 23°21', 106°02' El Dorado- 24°22', 107° 19 El Fuerte — 26°25', 108°39' Elota— 23° 58', 106°42' Esquinapa- 22°50', 105°46'* Guamuchil- 25°28', 108°06' Guasave — 25°34', 108°27' Isla Palmito de Verde- 22°40', 105°50' Isla Palmito de la Virgin- 23°00', 106° 10' Km. marker 1206 on Mexican Highway 15 -not exactly located, but in vicinity of Mazatlan La Concha — see Concepcion La Cruz — 23°55', 106°54' Matatan- 23°02', 105fiw45' Mazatlan -23° 13', 106°25' Panuco — 23°25', 105°55' Pericos — 25°03', 107°42' Piaxtla - 23°52', 106°39' Plomosas- 23°04', 105°29' Presa Sanalona — 24°48', 107°09'* Rosario - 23°00', 105°52' San Ignacio - 23°55', 106°25' San Juan- 23°51', 106°20' Santa Lucia - 23°26', 106°5 1 '* Sierra de Choix - approximately 26°50', 108° 11' (this locality is approximately 15 mi. NE Choix instead of 50 mi. NE Choix as stated on the specimen labels, according to Goldman, 1951:251) Sinaloa - 25°50', 108° 14' Teacapan- 22°33', 105°45' GENOWAYS— SYSTEMATICS OF LIOMYS 351 Vaca — 26°48', 108°25' Villa Union- 23° 12', 106°14' For additional information about local¬ ities in Sinaloa see Hardy and McDiarmid (1969:221-225). Sonora Alamos- 27°0r, 108°56' Camoa- 27° 13', 109° 17' Chinobampo- 26°59\ 109° 18' El Novillo — 28°57', 109°39'* Guirocoba - 26°54', 108°42'* Hermosillo- 29°04', 110°58' Matape — 29°08', 109°58' Nacori [Grande] -29°04', 110°03' Nogales - 31 °20', 1 10°56' Rio Alamos- approximately 26°55', 109° 09' (according to R. J. Baker, who col¬ lected the specimens, this locality is 11 mi. by road from Alamos in a more or less SSW direction) Rio Cuchajaqui - approximately 26° 54', 108° 57' (according to R. J. Baker, who collected the specimens, this locality is 8 mi. S Alamos) Tecoripa — 28°37', 109°57' Tesia — 27° 10', 109°23' Ures — 29°26', 110°24' For additional information about local¬ ities in Sonora see Burt (1938:map 26). Tamaulipas Acuna — 23° 1 2', 98°26' Altamira— 22°24', 97°55' Antiguo Morelos- 22°33', 99°05' Aserradero del Paraiso— 22° 59', 99° 1 5'* Bagdad -25°57', 97°09' Ciudad Mante [El Mante] - 22°44', 98°57' Ciudad Victoria [Victoria] - 23°44', 99°08' Cruiilas — 24°45', 98°31' Cueva del Abra- approximately 22°37', 99°02' (10 km. NNE Antiguo Morelos) Ejido Santa Isabel - approximately 23° 14', 99°00' El Barretal — 24°05', 99° 16'* El Carrizo - 23° 15', 99°07' El Encino- approximately 23 08', 99 07' El Limon — 22°50', 99°00' Gomez Farias - 23°03', 99 09' Hacienda Santa Engracia - approximately 24°01', 99° 12' Hidalgo -24° 15', 99°26' Jaumave - 23° 25', 99° 23' La Pesca - 23°46', 97°47' La Purisima- 24° 17', 99°28' Marmolejo - 24°38', 99°01 ' Matamoros - 25° 5 3', 97°30' Mesa de Llera - approximately 23°20', 99° 01' (see Hooper, 1953:3, for additional information about this locality) Miquihuana- 23°34', 99°47' Mulato [El] — 24°53', 98°56' Nicolas- 23°21', 100°04'* Padilla- 24°01', 98°47' Palmillas — 23° 1 8', 99°33' Pano Ayuctle - approximately 23°06', 99° 07' Piedra - 23°29', 98°06' Quintero- 22°41', 99°02' Reynosa— 26°07', 98° 18' San Fernando - 24° 51', 98° 10' San Jose- 24°42', 99°04' Soto la Marina- 23°46', 98° 13' Tampico -22° 13', 97°51' Tula— 23°00', 99°43' Victoria — see Ciudad Victoria Villa Mainero - 24°34', 99°36'* For additional information on localities in Tamaulipas see Alvarez (1963:386-387). Tlaxcala Tlaxcala [de Xicohtencatl] — 1 9° 1 9', 98° 14' Veracruz Achotal - 1 7°44', 95° 10' Acultzingo - 1 8°43', 97° 1 9' Alvarado- 18°46', 95°46' Boca del Rio - 1 9°06', 96°06' Carrizal - 19°21', 96°38' Catemaco- 18°25' 95°07' Cerro Gordo - approximately 19°25', 96° 40' Coatzacoalcos - 18°09', 94°25' Coyutla — 20° 1 5', 97°39' El Tajin — 20° 27', 97°23' Gutierrez Zamora - 20°27', 97°05' Ixcatepec — 21 0 1 3', 98°00' Jacales — 20°25', 98°27' Jimba- 1 7°56', 95°24' Lago Catemaco - 1 8°25', 95°05' La Mar- 21 °32', 97°40' Miahuapan [Miahuapa or Miahuapam] - 20°37', 97°37' Nautla — 20° 13', 96°47' Orizaba - 1 8°5 1 ', 97°05' Otatitlan - 18° 12', 96°02' Ozuluama - 2 1 °40', 97°51' 352 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY Papantla - 20°27', 97° 19' Paraje Nuevo - approximately 18°52\ 96° 52' Pasa Nueva- approximately 17°59\ 95°11' (see discussion of this locality in Hall and Dalquest, 1963:184) Paso de San Juan - 19° 12', 96° 19' Piedras Clavados — approximately 21° 11' 97°59' Piedras Negras- 18°46', 96° 1 1' Plan del Rio - approximately 19°23', 96°36' Platon Sanchez- 21°17', 98°22' Potrero Llano [Potrero del Llano] - approx¬ imately 21 ° 10', 97°43' Presidio - approximately 18° 39', 96°46' Puente Nacional - 19°20', 96°26' San Andres Tuxtla- 18°27', 95°13' San Carlos [Ursulo Galvan] - 19°24', 96° 21' San Juan de la Punta [Cuitlahuac] - 18°49' 96°43' San Marcos- 20° 12', 96° 57' Santa Maria- approximately 19° 19', 96°35', (this is a village at 1800 ft. near the river of the same name, in the lowlands about 20 mi. NE Mirador, according to Gold¬ man, 1951:281) Santiago Tuxtla- 18°28', 95° 18' Tihuatlan — 20°43', 97°32' Tlapacoyan- 19°58', 97°13' Tuxpan — 20°57', 97°24' Veracruz - 19° 12', 96°08' For additional information on localities in Veracruz see Hall and Dalquest (1963: 180-186). Yucatan Chichen Itza- 20°40', 88°34' Peto — 20°08', 88°55' Piste- 20°42', 88°35' Tizimin- 21°09', 88c09' Zacatecas Berriozabal — 22°33', 102° 19' Chalchihuites - 23°29', 103°53' Hacienda San Juan Capistrano — 22°37', 1 04°05' Jalpa — 21 °38', 102°58' Moyahua — 2 1 ° 1 6', 103° 10' Noria de Angeles - 22°26', 101°54' Rancho Grande - 23°27', 102°58' Rio Grande- 23°50\ 103°02' Sain Alto - 23°35', 1 03° 1 5' Santa Rosa-21°13', 103°09' Trancoso- 22°44', 102°22' Valparaiso- 22° 46', 103° 34' Zacatecas- 22°47', 102°35' Nicaragua Alta Gracia, Isla de Ometepe, Rivas— 11° 34', 85°35' Boaco, Boaco- 12°27', 85°43' Bonanza, Zelaya- 13°57', 84°32' Condega, Esteli- 13°21', 86°25' Cosiguina, Chinandega- 12°55', 87°30' Daraili, Esteli— 13°24', 86°16' (this finca is 5 km. N, 14 km. E Condega)* Dario, Matagalpa- 12°42', 86°08' Diriamba, Carazo- 1 1°53', 86° 15' El Crucero, Managua - 12°00', 86° 19' Esquipulas, Matagalpa- 12°37', 85°49' Esteli, Esteli - 13°05', 86°23' Finca Amayo [Hacienda Amayo], Rivas - U°19', 85°43' (this finca is 13 km. S, 14 km. E Rivas)* Hacienda Azacualpa, Managua- 12°02', 86°32' (this hacienda is 5 km. N, 2 km. W Villa El Carmen)* Hacienda Bellavista, Volcan Casita, Chinan¬ dega- 12°41', 86°58' Hacienda Corpus Christi, Managua - 12° 16', 86°23' (this hacienda is 8 km. NE Mateare)* Hacienda Cutirre, Volcan Mombacho, Granada- 11 °50', 85°56' (this hacienda is 1 1 km. S, 3 km. E Granada)* Hacienda Mecatepe, Granada - 1 1 °46', 85° 57' (this hacienda is 2 km. N, 11 Vi km. E Nandaime)* Hacienda Tepeyac, Mataglapa- 13° 11', 85° 56', (this hacienda is 10 Vi km. N, 9 km. E Mataglapa)* Isla Zapatera, Granada - 1 1°45\ 85°50' Jinotega, Jinotega- 13°06', 86°00' La Calera, Granada- 1 1°44', 86°06' (this hacienda is 3 km. S, 5 km. E Nandaime)* Lake Jiloa, Managua - 12° 13', 86° 19' Las Maderas, Managua - 12°28\ 86°03' Leon, Leon - 12°26', 86°54' Managua, Managua- 12°09', 86° 17' Masachapa, Managua- 1 1°47', 86°31' Matagalpa, Matagalpa- 12°53', 85°57' Merida, Isla de Ometepe, Rivas- 1 1°26' 85°33' Moyogalpa, Isla de Ometepe, Rivas- 11° 32', 85°42' GENOWAYS— SYSTEMATICS OF LIOMYS 353 Poneloya, Leon - 12°22\ 87°02' Puerto Momotombo, Leon — 12°24', 86°37' Quilali, Nueva Segovia - 13° 32', 86°00' Rivas, Rivas - 1 1°26', 85°51' Sabana Grande, Managua - 12°07', 86° 10' San Antonio, Chinandega- 12°32', 87°03' San Pablo, Rivas - 1 1°36\ 85°57' San Ramon, Matagalpa- 12°53', 85°49' Santa Maria de Ostuma, Matagalpa- 12° 57', 85°58' Sapoa, Rivas - 1 1 0 1 5', 85°38' Savala, Matagalpa - approximately 13°03', 85°3 1 ' (this locality is 45 km. ENE Mata¬ galpa, according to Buchanan and Howell, 1965:549) Sebaco, Matagalpa- 12°51', 86°0C Tamarindo, Leon - 12° 12', 86°46' Uluse [Uluce], Matagalpa- 12°53 85°37' Villa Somoza, Chontales- 12°08', 84°58' Volcan de Chinandega [Volcan San Crist¬ obal] , Chinandega - 12°43', 87°03' Panama and Canal Zone Albrook Air Force Base, Canal Zone - 8°59', 79733' Balboa, Canal Zone - 8°57', 79°34' Buenos Aires, Panama - 9° 1 O', 79° 37' Cacao Plantation, Canal Zone-9°06', 79° 41' Cerro Azul, Panama - approximately 9° 14', 79°21' Chepo, Panama - 9° 1 O', 79°06' Chiva Chiva, Canal Zone - 9°02', 79°35' Cocoli, Canal Zone— 8°59', 79°35' Corte Culebra Road, Canal Zone -coordi¬ nates not given by Fairchild and Handley (1966:14) Curundu, Canal Zone - 8°59', 79°33' Empire, Canal Zone-9°04', 79°40' (this is the old administrative center of the canal, located on west bank about halfway be¬ tween Paraiso and Gamboa) Farfan, Canal Zone-8°56', 79°35' Fort Clayton, Canal Zone - 8°59', 79°36' Fort Kobbe, Canal Zone - 8°55', 79°35' Guabala, Chiriqui - 8° 1 3', 8 1 °44' Guanico [Guanico Arriba and Las Palmi- tas] , Los Santos - 7° 1 8', 80°26' Juan Mina, Canal Zone — 9° 10', 79 39' Madden Forest, Canal Zone - 9°06', 79°37' Madden Road, Canal Zone-C-25 Road, extending from Paraiso to Madden Dam, passing through Madden Forest Montijo Bay, Veraguas - near Paracote (see Aldrich and Bole, 1937:7-8) Nuevo Emperado, Canal Zone-9°00', 79° 44' Paracote, Veraguas - approximately 7°40', 8 TOO' (see Aldrich and Bole, 1937:7-8, for additional information about this lo¬ cality) Pueblo Nuevo, Panama - 9°01 ', 79° 3 1' Rio Chagres, Canal Zone — 9°08', 79°41 ' Rio Hato, Code - 8°23', 80° 10' Rodman Naval Ammunition Depot, Canal Zone - approximately 8°57', 79°37' (8 km. W Balboa, according to Fleming, 1970:473; 1971:7) San Francisco, Veraguas - 8° 15', 80° 58' Santa Fe, Veraguas - 8° 31', 81°05' Summit, Canal Zone- 9°04', 79°39' Tole, Chiriqui - 8° 14', 81 °41' For additional information on localities in Panama and the Canal Zone see Goldman (1920:4-18) and Fairchild and Handley (1966:9-22). United States Texas Alamo, Hidalgo County - 26° 1 1 ', 98°07' Bayside, Cameron County -not exactly located, but approximately 26°07', 97° ] 5' Bentsen State Park, Hidalgo County -ap¬ proximately 26° 10', 98°21' (4 mi. SW Mission) Brownsville, Cameron County - 25°54', 97° 30' Edinburg, Hidalgo County - 26° 1 9', 98°09' Elsa, Hidalgo County - 26° 1 8', 97°59' Fort Brown, Cameron County - 25°54', 97°30' Ingram, Kerr County - 30°05\ 99° 14' Lomita Ranch, Hidalgo County - approxi¬ mately 26° 10', 98°22' (based on informa¬ tion in files of U.S. Biological Survey, Washington, D.C.) McAllen, Hidalgo County- 26° 1 1', 98° 14' M ission, Hidalgo County - 26° 1 2', 98° 1 9' Raymondville, Willacy County - 26°29', 97°46' Rio Hondo, Cameron County - 26° 1 5', 97° 35' San Benito, Cameron County — 26°08', 97° 37' All localities in Texas were plotted from the Monterrey quadrat (NG-14) of the 354 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY American Geographical Hispanic America. . . information on localities (1952:231-232). Society’s “Map of For additional in Texas see Blair Venezuela Agua Santa, Trujillo - 9°32', 70°38'* Carupano, Sucre - 10°40', 63° 14'* Maracaibo, Zulia- 10°40\ 71°37'* Rancho Grande, Aragua— 10°22', 67°41'* Valera, Trujillo - 9° 1 9', 70° 37'* GENOWAYS— SYSTEMATICS OF LIOMYS 355 APPENDIX II Ectoparasites known from Liomys and Heteromys are listed below. Numbers in source column indicate unpublished records from the files of the following specialists: Richard B. Loomis (Trombiculidae); Russell W. Strandtmann (Lae- lapidae), Burruss McDanial (Listrophoridae); John B. Kethley (Cheyletidae); Glen M. Kohls and Eleanor K. Jones (Ixodides); K. C. Emerson (Anoplura); Robert Traub (Siphonaptera). Localities of collection corresponding to each number are listed following the list of ectoparasites. Ectoparasite Family Species Source Trombiculidae Liomys irroratus Ectonyx fusicornis Brennan ( 1960u:88); 6 Euschoengastia gagarini Brennan (1962:618) Euschoengastoides bigenuala Loomis and Crossley tt Euschoengastoides lacerta (1963:379); Eads et al. (1965:17) Loomis and Crossley tt Euschoengastoides loomisi (1963:380); Eads et al. (1965:17) Loomis and Crossley ft Eutrombicula alfreddugesi Fonsecia ( Parasecia ) universitatis (1963:380); Eads et al. (1965:17) Eads et al. (1965:17) Hoffmann (1963: 108); ft Hexidionis allredi Brennan (1969:666) Eads et al. (1965:17) n Hexidionis jessiemae Brennan ( 1 965:83) tt Leptotrombidium panamense Loomis and Crossley tt Odontacarus cayolargoensis (1963:378); Eads et al. (1965:17) Loomis and Crossley ft Pseudoschoengastia audyi (1963:381); Eads et al. (1965:17) Brennan and Dalmat tt Pseudoschoengastia farneri (1960:191); Eads et al. (1965:17) Loomis and Crossley tt Pseudoschoengastia hoffmannae (1963:381); Eads et al. (1965:17) 5 tt Pseudoschoengastia hungerfordi 5, 6 tt Trombicula bakeri 8 tt Trombicula n. sp. Eads et al. (1965:17) tt Xenacarus plumosus Loomis and Crossley Laelapidae Androlaelaps fahrenholzi (1963:382); Eads et al. (1965:17) Eads et al. (1965:17) 356 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY tt Hirstionyssus neotomae Eads et al. (1965:17) tt Hirstionyssus n. sp. Eads et al. (1965:17) tt Steptolaelaps liomydis Grant (1947:8); Furman (1955:525); Eads et al. (1965:17) Macronyssidae Ornithonyssus bacoti Eads et al. (1965:17) ft Ornithonyssus sylvarium Eads et al. (1965:17) Listrophoridae Listrophorus n. sp. Eads et al. (1965:17) Argasidae Ornithodoros talaje Eads et al. (1965:17) Ixodidae Ixodes eadsi Kohls and Clifford (1964:466); Eads et al. (1965:17) Hoplopleuridae Fahrenholzia ehrlichi Ferris (1922:161); Johnson (1962:417); Eads et al. (1965:17); Emerson (1 971 b: 374) ft Fahrenholzia texana Ferris (1922:161); Stojanovich and Pratt (1961:693); Johnson (1962:426); Eads et al. (1965:17) Pulicidae Hoplopsyllus ( Euhoplopsyllus ) glacial is 10 Rhopalopsyllidae Polygenis gwyni Eads and Menzies (1949:37); Eads (1950:62); 1, 2 ft Polygenis martinezbaezi Liomys pictus 2, 3, 4, 7, 9 Trombiculidae Anahuacia 41 ft Ectonyx fusicornis 8, 13, 14, 18, 20, 21, 22, 25, 26, 27, 36 ft Euschoengastia gagarini 26, 28 ft Euschoengastoides arizonae Loomis (1971:699); 1 ff Euschoengastoides expansellus Loomis (1971:700); 8 ft Euschoengastoides turn id us Loomis (1971:703); 7, 8, 27, 31 It Eutrombicula alfreddugesi 41 ft Eutrombicula sp. 28 tf Fonsecia ( Parasecia ) sp. 25, 26, 28, 37 tt Hexidionis allredi 2,6 rt Leptotrombidium panamense potosinum 2, 3, 4, 8, 9, 10, 19, 20, 21, 22, 24, 26, 28, 37,41 tt Leptotrombidium n. sp. “c” 8, 11, 12, 15, 16, 20, 21, 22, 23, 24, 25, 26, 29 rt Neotrombicula sp. 20, 24 it Odontacarus sp. 3, 28 tt Otorhinophila intrasola Wrenn and Loomis (1967:160); 3, 9 tt Otorhinophila sinaloae Wrenn and Loomis (1967:164); 8 GENOWAYS— SYSTEM ATICS OF LIOMYS 357 n Pseudoschoengastia aberrans 38 n Pseudoschoengastia audyi Brennan and Dalmat (1960:191) n Pseudoschoengastia guatemalensis Brennan and Dalmat (1960:191) n Pseudoschoengastia hoffmannae Brennan (1960 b: 486), 34 ii Pseudoschoengastia hungerfordi 13, 33, 34, 37 n Pseudoschoengastia scitula Brennan and Jones (1959:427) n Pseudoschoengastia n. sp. “e” 20 n Pseudoschoengastia n. sp. “j” 20, 24, 25, 26, 28, 29 n Pseudoschoengastia sp. 41 n Sasacarus whartoni 26 Laelapidae Androlaelaps leviculus 11 II Steptolaelaps heteromys 14, 41 II Steptolaelaps liomydis Furman (1955:525); 11, 14, 17, 18, 41 Ixodidae Ixodes sinaloa Kohls and Clifford (1966:81 1, 813); Keirans and Jones (1972:474) II Ixodes sp. 11, 14 Hoplopleuridae Fahrenholzia microcephala Ferris (1922:161); Johnson (1962:416); Emerson (1971 6:375); 11, 14, 17, 18, 37, 41 Ceratophyllidae Jellisonia wisemani 13, 14 Rhopalopsyllidae Polygenis gwyni 32 II Polygenis martinezbaezi 4, 30, 35 II Polygenis vazquezi Hubbard (1958:165) II Polygenis vulcanius 38, 39, 40, 41 II Polygenis sp. Liomys salvini 5 Trombiculidae Anahuacia n. sp. “t” 19, 21, 23 II Ascoschoengastia dyscrita 19 II Cordiseta mexicana 6, 9, 18, 20, 21, 23 II Euschoengastia sp. 15 II Euschoengastoides sp. 6 II Eutrombicula alfreddugesi 3, 6, 9, 1 1, 12, 17, 20, 21, 22, 23 II Eonsecia ( Parasecia ) sp. 7,9 II Leptotrombidium panamense panamense 6, 7, 15, 19, 20 II Leptotrombidium panamense potosinum 1 II Microtrombicula perplexa Webb and Loomis (1971:5) II Pseudoschoengastia costar ice nsis Geest and Loomis (1968:38, 40) II Pseudoschoengastia guanacastensis Geest and Loomis (1968:36, 38) II Pseudoschoengastia hoguei Geest and Loomis (1968:25, 28) 358 SPECIAL PUBLICATIONS MUSEUM TEXAS TECH UNIVERSITY n Pseudoschoengastia sp. 6, 7, 9, 11, 15, 17 n Trombicula dunni 9 Laelapidae Androlaelaps fenilis 7, 12 n Eubrachylaelaps ? circular is 7 it Hirstionyssus galindoi 7 it Hypoaspis lubrica 7 ii Hypoaspis sp. 7 it Steptolaelaps heteromys 7, 8, 9, 10, 11, 12, 13, 14, 16, 17 Listrophoridae Listrophorus sp. 11 Cheyletidae Eucheyletia n. sp. 7 Ixodidae Amblyomma sp. 4 If Ixodes eadsi (or near) 9, 11 II Ixodes sinaloa Keirans and Jones (1972:474) Hoplopleuridae Eahrenholzia fairchildi Johnson (1962:419); Emerson (1971a: 333) Rhopalopsyllidae Polygenis vulcanius Liomys adspersus 2, 5, 24 Trombiculidae Ascoschoengastia dyscrita Brennan and Jones (1961:107); Brennan and Yunker (1966:227) it Colic us liomys Brennan and Jones (1961:121); Brennan and Y unker ( 1 966:254) II Crotonasis fissa Brennan and Yunker (1966:231) II Eutrombicula goeldii Brennan and Yunker (1966:236) II Leptotrombidium panamense Brennan and Yunker panamense (1966:238) II Odontacarus field i Brennan and Yunker (1966:224) II Polylopadium kramisi Brennan and Jones (1961: 1 12); Brennan and Yunker ( 1 966:243) It Pseudoschoengastia b ul b if era Brennan (1960 6:483); Brennan and Yunker (1966:245) II Pse ud oschoengastia tricosa Brennan and Jones (1961:123); Brennan and Yunker (1966:257) n Pseudoschoengastia zona Brennan (19606:490); Brennan and Yunker ( 1 966:248); Geest and Loomis ( 1 968:25) Laelapidae Androlaelaps fahrenholzi Tipton et al. (1966:34) It H irstionyssus rn icrochelae Strandtmann and Yunker (1966:1 14) It Steptolaelaps heteromys Tipton et al. (1966:39) Macronyssidae Ornithonyssus bacoti Yunker and Radovsky (1966:89) Ixodidae Amblyomma sp. Fairchild et al. (1966:21 I) GENOWAYS— SYSTEMATICS OF LIOMYS Hoplopleuridae Fahrenholz ia fa irch ild i Johnson (1962:419); Wenzel and Johnson (1966:275) Pulicidae Ctenocephalides felis felis Tipton and Mendez (1966:292) Rhopalopsyllidae Polygenis dunni Tipton and Mendez (1966:298) It Polygenis klagesi Liomys sp. Tipton and Mendez (1966:298) Ceratophyllidae Jellisonici hay e si 3 Rhopalopsyllidae Polygenis gxvyni 2 II Polygenis nmrtinezbaezi Heteromys anomalus 1 Trombiculidae Acomatacarus tubercularis Brennan ( 1952: 146) II Anomalaspis ambiguus Brennan (1952: 146) II Colic ns floe hi Brennan and Jones (1960:504) It Crotiscus desdentatus Brennan (1957:675) II Eutrombicula alfreddugesi Brennan and Jones (1960:506) It Eutrombicula goeldii Brennan and Jones (1960:508) It Fonsecia ( Parasecia ) manueli Brennan and Jones (1960:520) It Kymocta faitkeni Brennan ( 1 968