OVERDUE FINES ARE 25¢ PER DAY . PER ITEM Return to book drop to remove this checkout from your record. AN ANALYSIS OF RODENT DISTRIBUTION PAITERNS IN ZACATECAS, MEXICO By John O. Matson A.DISSERIAIION Smeitted to Nfichigan State University in.partial fulfillment of the requirements for the degree of DOCTOR.OF PHILOSOPHY Department of Zoology 1979 ABSTRACI‘ AN ANALYSIS OF RODEN'I' DISTRIBUTION PATTERNS IN ZACATECAS , MEXICO By John O. Matson The purposes of this report are to document the occurrence and dis- tribution of species of rodents in the state of Zacatecas , Mexico , and to examine whether or not the distribution patterns of rodents can be objectively grouped into biogeographical units. Various aspects of the environment of Zacatecas are described, including: physiography, topography , climate , and vegetation. In essence, Zacatecas consists of a low, arid eastern region (desert), a higher, mister, central region (grasslands), a western montane region (pine—oak forests), and a western canyon region (tropical vegetation). Distributional data for 46 rodent species in Zacatecas were analyzed using two numerical taxonanic tecl'miques , Cluster Analysis and Factor Analysis. Initially, the state was divided into a grid system consisting of 73 quadrats for which data were available. I decided that only those quadrats which contained at least ten species would be adequate for analysis, resulting in the use of 57 quadrats. Cluster analysis was carried out using the unweighted pair group method with arithmetic averages on three indices of association, Euclidean distance, Average Faunal Resemblance, and Similarity. These John 0. Matson three indices produced similar dendrograns from which I was able to recognize four major clusters in common to each. The four clusters represented ecogeographic units corresponding to: (1) an eastern desert unit; (2) a central grassland unit; (3) a western montane unit; and (4) a western canyon unit. Differences between the three dendro- graus are discussed in terms of the nature of the calculation of each index. A major drawback of cluster aialysis is that any given quadrat cam be associated with only one cluster. Factor analysis included data for 43 rodents species and five environmental variables. Seven factors , accounting for 60. 9 per cent of the total variation, were extracted frcm a correlation matrix. The factor matrix was rotated by a varimax procedure to simple structure. The seven factors produced generalized patterns that proved to be biologically neau'ngful, defining essentially the sane ecogeographic units recognized by cluster analysis . Three factors defined the sate ecogeographic units; three others refined or subdivided these units. A seventh factor defined a riparian or mesic habitat association. The results of cluster aid factor analyses are synthesized into a classlification of Rodent Faunal Areas in Zacatecas. This classifi- cation and the geographical limits of the Faunal areas corresponds to a testable hypothesis of biogeographical areas. The four Rodent Faunal Areas recognized are: (1) Desert Famal Area; (2) Grassland Faunal Area; (3) Montaue Faunal Area; and (4) Tropical Faunal Area. In addition, a broad region of tra13ition between Desert a1d Grassland Faunal Areas was recognized by factor analysis . Certain quadrats were found to be associated with more than one factor indicating either that John O. Matson they were ecotonal quadrats or that they contained parts of two or more faunal areas. The effects that ma'l has had and is having on the biological resources in Zacatecas is discussed. The status of various rodent species is considered in relation to habitat destruction or modifi- cation. The rareinder of the report concerns the docmentation of species distribution and notes on natural history. WES I thank the members of my guidaice canmittee, Rollin H. Baker (Chairman), Marvin M. Hensley, J. Ala: Holman, and Peter G. Murphy for their helpful suggestions and for critically reading the manuscript. I am particularly grateful to Rollin H. Baker for his invaluable assis- tance, encouragement, and support during all phases of this project. Field work was supported by graits from the Jens Touborg Fund to The Museum ad by Museum Research Assistantships. Travel support to visit various museums was provided by the Department of Zoology and my in-laws, Walter and Fraices Jaffe. I thank J. Paul Abravaya, William David Webster, Janet Hamilton, and Dan and Karen Bemack for their assistance in the field. I especially thaik J. Paul Abravaya for his dedicated work during three summers in Zacatecas. I thank the following curators and institutions for allowing me to examine specimens in their care: Sydney Anderson, American Museum of Natural History; Robert T. Orr, California Academy of Sciences; Robert S. Hoffimam, Museum of Natural History, University of Kaisas; Lani A Lester and Donald R. Patten, Los Angeles County Museum; George H. Lowery, Jr. , Louisiana State University; William Z. Lidicker and James L. Patton, Museum of Vertebrate Zoology, University of California; Philip Myers, Museum of Zoology, University of Michigan; Robert G. Hamun, Private Collection; David J . Schmidly, Texas Cooperative ii iii Wildlife Collections, Texas A 6: M University; Robert J. Baker, The Museum, Texas Tech University; Donald F. Hoffmeister, Museum of Natural History, University of Illinois; J. Keever Greer, Stovall Museum of Science and History, University of Oklahema; Don E. Wilson, Biological Survey Collection, United States Fish and Wildlife laboratory . I thank Sr. Cornelio Sanchez for providing me with infor- mation on specimens housed in the Instituto de Biologia, Universidad Nacional Autonuma de Mexico. I am especially indebted to Lani A. Lester and Donald R. Patten of the 103 Angeles County Museum for allow- ing me to borrow over 800 specimens frem their collections. Collecting permits were provided by the Director General de Fauna Silvestre, Departmento de Consevacion, Mexico, D. F. I am deeply indebted to the people of Zacatecas for their generous hospitality and friendship. Special thanks are due Sr. Alejandro Delgadillo and family for allowing me to use the facilities of Rancho San Marcos during the summers of 1977 and 1978. I thank Frances Jaffe for typing part of the original draft of the dissertation. I thank my wife Natalie and son Jackie for their untiring help in the field, encouragement to pursue this study, and patience to put up with the rigors of field work and graduate studies. TABLE OF CONTENTS LISI‘ OF FIGURES ......................... INTRODUCTION ........................... ENVIRINMENI‘OF ZACATECAS ..................... Phys iography and Topography .................. Climate ............................ Vegetation ........................... MATERIALS AND METHODS ............ fl .......... Species Data .......................... Quadrats ............................ Envirormmental Data ....................... Methods of Faunal Analyses ................... RESULTS ............................. Cluster Analysis ........................ Factor Analysis ........................ DISCUSSION ............................ Cluster Analysis ........................ Factor Analysis ........................ Faunal Areas .......................... Human Impact .......................... Gazetteer ........................... iv oouun U‘I 20 21 28 28 35 56 56 62 68 73 76 Page SPECIES ACCOUNTS ......................... 81 bulleri (J. A. Allen) ............. 82 mexicanus (Erxl e ben) ............... 83 siIosoma Bennett ................. 84 variegatus (Erxleben) ............... 85 Merriam ................... 87 socialis Wagner .............. 88 na aritensis J. A. Allen ......... 88 iBairdS ................. 89 Goldman ................ 90 ................. 91 Baird .................... 98 zacatecae Osgood ............. 100 Dalquest ................. 100 Osgood ............... 101 eremicus Mearns ............ 104 iam .............. 105 ................... 107 ................... 108 ornatus Merriam .............. 110 cratodon Merriam ............. 111 .................... 112 (J A Allen) .............. 114 J A. Allen ............. 115 d) ............... 116 (J A. Allen) ............ 118 Allen) .............. 120 .............. 121 .............. 123 (Wagner) ................ 124 Coues) ................. 126 A.A11en and Chapman .......... 130 Osgood .................. 130 J. A. Allen ............... 132 Osgood ................ 132 A. Allen) .............. 134 Merriam ................ 135 A Allen .......... 136 .............. 137 ............... 138 Allen ................ 139 Merriam ............... 140 ................... 142 Baird ..................... 143 Goldman .................... 144 neotomodon Merriam ............. 144 LITERATURE CITED ......................... 146 Table 1. Table 2. Table 3. Table 4. Table 5. Table 6. Table 7. Table 8. Table 9 . LIST OF TABLES Variables associated with Factor I: Desert Vs Grassland Habitats .................. Variables associated with Factor II: Western Montane Habitats ................... Variables associated with Factor III: Cross Ranges Habitats ................ Variables associated with Factor IV: Canyon Region Habitats ................... Variables associated with Factor V: Desert- Grassland Ecotone ................. Variables associated with Factor VI: Southern Canyons Habitats ................... Variables associated with Factor VII: Si on hispidus—Liomys irroratus Complex .......... Comparison of numbers of species in selected quadrats in Zacatecas ................ Comparative measurements of female Thamomys umbrinus from three localities in Zacatecas .......... Page 37 4O 43 45 50 53 58 95 Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure 10. 11. LIST OF FIGURES Topographic relief in Zacatecas in intervals of 500m ...................... Distribution of Average Annual Precipitation in Zacatecas .................... Distribution of Average Annual Temperature in Zacatecas .................... Distribution of mej or vegetation types in Zacatecas: (1) Desert Scrub; (2) Crassicaulescent Scrub; (3) Grasslands; (4) Montane Forests; (5) Subtropical Scrub; (6) Tropical Deciduous Forests ...................... The quadrat numbering system (top number) and mmber of species per quadrat (bottam number in parentheses) .................... Dendrogramn of quadrats (nurber on right) based upon UPCMA clustering of the distance index and showing the four mej or clusters .......... Dendrogram of quadrats (mnnber on right) based upon UPGMA clustering of the Average Faunal Resemblance index and showing the four major clusters ...................... Dendrogram of quadrats (mmber on right) based upon UPGMA clustering of the similarity index and showing the fonn: major clusters .......... Distribution of the three classes of factor scores for Factor I in Zacatecas. Factor scores ranged fren -2.64 to +1.96 ................ Distribution of the three classes of factor scores for Factor II in Zacatecas . Factor scores ranged frem -1.46 to +2.81 ................ Distribution of the three classes of factor scores for Factor III in Zacatecas. Factor scores ranged from -1.58 to +4.30 ................ vii Page 6 9 10 12 18 29 30 31 39 42 44 Figure 12 . Figure 13 . Figm'e 14. Figure 15 . Figure 16 . Figure 17 . Figure 18. viii Distribution of the three classes of factor scores for Factor IV in Zacatecas. Factor scores ranged from -1.22 to +4.36 ................ Distribution of the three classes of factor scores for Factor V in Zacatecas. Factor scores ranged fram -2.65 to +2.22 ................ Distribution of the three classes of factor scores for Factor VI in Zacatecas . Factor scores ranged fram -1.71 to +5.26 ................ Distribution of the three classes of factor scores for Factor VII in Zacatecas . Factor scores ranged from -2.22 to +2.30 ................ Distribution of the four major clusters of quadrats in Zacatecas based upon the dendrograms in Figures 6 through 8 ............... Distribution of Rodent Faunal Areas recognized in Zacatecas: (1) Desert Faunal Area; (2) Grassland Faunal Area; (3) Montane Faunal Area; (4) Tropical Faunal Area; (5) Desert-Grassland Ecotone ..... Place-names used to identify collecting localities in Zacatecas . Numbers correspond to the locality in the gazetteer .................. nge 47 49 52 54 60 70 77 INTRODUCTION In this day of rapid depletion of our natural resources, much attention is being given once again to surveys of the flora and fauna of various parts of North Amnerica. Usually, these are concerned with environmental impact statements of local areas that are about to be exploited. In the United States fairly good base-line data exist for such surveys. However, much of our knowledge of the biota of latin America still remains meagre or spotty. Mexico is currently faced with one of the greatest rates of popu- lation growth in the world today (Corzo, 1970; Wellhausen, 1976). Much of Mexico' 3 land is rapidly being turned to cultivation to meet the demands of a growing population (Wellhausen, 1976). While this is a necessary consequence of population growth, the environmental changes can have a chrastic effect on the native biota. Without the needed base-line data about the taxa and where they occur, the impact of these environmental perturbations cannot be adequately assessed in future years. The Mexican state of Zacatecas represents an area for which little is know concerning its biota, in particular the mammals . Other than a few scattered records, there exists no comprehensive survey of the flora or fauna from Zacatecas. The recent reports by Baker 31.; a_1_. (1967), Genoways and Jones (1968), Best _e_t a1. (1972), Matson and Patten (1975), Matson (1977), Jones and Webster (1977), and Matson 1 _e_t_ §_]_.. (1978) have surmarized our meagre knowledge of mammals in the state. Other mammals known frum Zacatecas were reported in various taxonamic works most of which were sunnnarized in Hall and Kelson (1959). The purpose of the present study is to help fill in the gaps in our knowledge of one Order of mammals frcm Zacatecas , the Rodentia. The choice of the Order Rodentia as a taxon for study was based on the following considerations: First, it reflects my primary interest in mammals in general and rodents in particular. Second, the Order Rodentia represents a large and diverse group of animals which I believe would be adequate to demonstrate major patterns of distribution within the state. Third, Rodentia represents one of the easiest groups of mammals to sample because of their abundance. This allows for rela- tively complete checklists to be made in an area within a short time period. Fourth, rodents from Zacatecas are well represented in collec- tions held by various institutions, thus facilitating the present study. For three summers (1976, 1977, 1978), I concentrated on study- ing rodents in areas of the state that had not been previously examined. Selection of a study area for a faunal survey can be accomplished in two ways. First, one may select some defined "natural" area, such as the Chihuahuan Desert or Sierra Madre Occidental. This approach was the subject of a recently published symposium on the Chilmahuan Desert (Wauer and Riskind, 1978). Second, one may select a politically defined area. This has probably been the mnost common approach in faunal surveys. Surveys of politically defined areas have been criticized as not leading to any general biological concepts (Mayr, 1971; Findley and Caire, 1978). Both Armstrong (1972) and Anderson (1972) have countered such criticisms on a number of points. First, the limits of the area are precisely defined by the limnits of the political unit. Second, the results of a faunal survey of a political unit are often cemplerentary to other faunal surveys of adjacent units (eventually this would lead to a complete survey of the natural areas as well). Third, surveys of this type often provide more diversity than one of a "natural" area, thus providing the investigator with a broader view of the distri- butional patterns observed. I selected the state of Zacatecas as a study area for a variety of reasons . I becane interested in the mammalian fauna of Zacatecas when in 1972 I had the opportunity to help curate a rather large collection of mammals frum that state at the los Angeles County Museum. It quickly became apparent that many new state records were present (Matson and Patten, 1975; Matson, 1977). Also, there were few other published records of mammals frem Zacatecas. Thus , the need for a com- prehens ive survey of the mammals of Zacatecas seemed warranted. In addition, fairly camplete surveys existed for three contiguous states. These are: San Luis Potosi (Dalquest, 1953); Coahuila (Baker, 1956); and Durango (Baker and Greer, 1962). Also, a fourth contiguous state, Nayarit, is currently being surveyed by personnel of the United States Fish and Wildlife Service (Don Wilson, personal communication). Zacatecas and surrounding states lie in a portion of Mexico that Hershkovitz (1958) considered to be a transition between the Neotropi- cal and Neartic Zoogeographical Regions . Thus , this area would have a theoretical appeal for understanding the distribution and evolution of faunas, although this last reason will not be realized until more survey work is completed. Lastly, the sandy of a state fauna allows me the opportunity to be involved with research in two main fields of biological science, systematics and biogeography. The present study was designed to answer three questions concern- ing the rodents of Zacatecas: (1) What species of rodents occur in Zacatecas? (2) What is the distribution of each species within the state? Both of these questions represent the base-line data that can be used in future studies. (3) Can the distributions of species be objectively grouped into meaningful patterns associated with recog- nizable areas within the state? This question represents a theoretical problem that most earlier works on Mexican mammals have failed to deal with effectively. Only Anderson (1972) and Findley and Caire (1978) have approached this question for Mexican mammals on a more or less objective basis. A note of caution is necessary at this point. The following dis- cussion often take the attitude that the data are complete. This is by no means true. A study of the kind proposed is probably never fin- ished. This is because one can continue to work for finer detail and also because of the dynamic nature of species distributions (Udvardy, 1969) . In addition, man has been a mnajor force in habitat modification for mnany centuries. It is possible, even probable, that the distri- butional pattern of mnany species have been highly mnodified especially since the coming of Europeans. At present we have no direct means by which to assess man's influence. However, the present study is believed to be representative of the mnajor distributional patterns. WOF ZACATECAS Phys iography and Topography Zacatecas covers an area of approximately 72 , 800 sq. km at the southern portions of the Central Mexican Plateau and Sierra Madre Occidental. The Tropic of Cancer essentially bisects the state. The varied topography and climate allows for a complex of environmental situations throughout the state. Phys iographic provinces of Mexico were described and mapped by Raisz (1964) . These provinces more or less correspond to the major areas of topographic relief (Figure 1). Most of the eastern half of Zacatecas is in the Central Mesa Province and corresponds to elevations usually below 2,000 m. It is an area of relatively flat to low hilly terrain that serves as a drainage basin for runoff from the higher lands to the east and west . In fact, the Central Mesa Province corre- sponds to two of the major internal drainage basins of the Chihuahuan Desert described by Henrickson (1978) . These two major basins are the Rio Aguanaval (wlose mnaj or river originates in west central Zacatecas) and E1 Salado (the largest of the Chihuahuan drainage basins). The extrere northeastern portion of Zacatecas is dominated by numerous isolated mountain ranges (areas above 2,500 m in Figure 1). These mountains represent the southern portion of the Cross Ranges Subprovince of the Sierra Madre Oriental Province described by Raisz (1964) . Essentially, these mountains are a western extension of the 5 Topography 125° <1500m 1500-2000m" 1 l [02' IOI' Figure 1. Topographic relief in Zacatecas in intervals of 500 m. Sierra Madre Oriental . They form a connection between the Sierra Mach'e Oriental Province and the Sierra Madre Occidental Province in the area of Nasas, Mango. The highest peak (over, 3,000 m) in the Zacatecas portion of the Cross Ranges occurs at the border with Coahuila in the Sierra Antillero (referred to as the Sierra Encarnacion by Goldman, 1951, and others). The western half of Zacatecas lies at elevations mostly above 2 , 000 m and corresponds to the southern part of the Sierra Madre Occidental Province. Raisz (1964) recognizes two subprovinces of this area, the Eastern Uplands lies to the east of the main mountain mass of the Sierra Madre Occidental and represents an elevated plateau above the Central Mesa. The other subprovince is the lava (Rhyolite) Plateau which represents the main backbone of the Sierra Madre Occidental . In Zacatecas the lava Plateau is represented only by fingerlike extensions or isolated mountains of the sierra. In the extreme south and southwest these mountainous areas are cut deep by river canyons which drain the western and southern slopes of the sierra. The area of deep canyons was considered by Rzedowski and Mcth (1966) to represent a distinct physiographic region they called the "Canyon Region". These canyons form the major western drainage system in Zacatecas . The rivers are all tributaries of the Rio Grande de Santiago and include from west to east: The Rio Atengo; Rio Balai'nos (including the Rio Jérez); and the Rio Juchipila. Climate Climatic data were sunmarized from climatological maps (Secretaria de la Presidencia, Comision de Estudios del Territorio Nacional, 1975) and the published data in Wernstedt (1972) and Soto and Jauregui (1965). These sources provided part of the raw data for the analyses of species distribution (see section on Material and Methods). Only the major climatological patterns are summarized in this section. Most moisture in Zacatecas is precipitated as rain. Some snow falls during winter months and may occur almost anywhere (except in the extrere Southwest) within the state. Rainfall occurs mainly between June and September with little or no precipitation from Novennber through April or May. Average annual precipitation (Figure 2) ranges from below 250 mm in some areas of the northeast to over 800 mm in the extrene southwest (the Canyon Region). The mountains of the northeast receive considerably more rainfall per year than surrounding areas. This is reflected in the different vegetation (see section on Vegetation). Temperatures are moderate over most of the state (Figure 3). The coldest areas are found in the mountain regions, the warmest in the Canyon Region of the southwest. All areas in Zacatecas are probably subject to winter frost. Vegetation The vegetation of Zacatecas is complex and a comprehensive survey of the flora has not been mnade. The recent publication by Mata g _a_l_. (1971) considers the mnaj or vegetation types throughout Mexico. Rzedowski and McVaugh (1966) studied a region defined as "Nueva Galacia" which contained a large portion of southwestern Zacatecas . These two sources provided the primary data base and maps for the description of the vegetation inn Zacatecas . In addition, personal Average Annual Precipitation an. - >600mm 1%“ I. M 400-600mm-4 J I l04’ IO3' l02' IOI' 30' 24° 23' 22' 2|° Figure 2. Distribution of Average Annual Precipitation in Zacatecas. lO Average Annual Temperature I 16°-20° " 0 >20 .— 24’ 23’ 30' 22‘ 2|' IO4° IO3' l02' IOI' Figure 3. Distribution of Average Annual Temperature in Zacatecas. 11 observation and reference to Leopold (1950) , Gentry (1957) , Rzedowski (1957), and Johnston (1978) augmnented the primary data. In the follow- ing descriptions, I have chosen to recognize six major vegetation types in Zacatecas (Figm‘e 4). In essence, my classification most closely corresponds to that of Rzedowski and McVaugh (1966) with the exception of a Desert Scrub vegetation which they did not consider. The map of major vegetation types in Zacatecas should be considered only as a guide to the limnits of the major types. Limits of vegetation are not nearly so clear-cut as the map would indicate. Desert Scrub. -- This vegetation covers almost the entire north- eastern half of Zacatecas . It more or less corresponds to those areas that receive less than 400 mm of precipitation per year. Mata _e_t; 31. (1971) recognized two distinct types of "desert" vege- tation: A Microphyll Desert Scrub domninated by Cresote bush (£31332 tridentata) and tarbush (Flourensia cernua); and Rosetophyll Desert Scrub deminated by lechuguilla (Agave lecheguilla) , tree yuccas (Yucca filifera and _’n_(_. carnerosana) and sotol (Dasylirion sp.). While Mata gt; _a_1. (1971) recognized and mapped these two types, I have combined the two for simplicity. Besides, Jolmnston (1978) has shown that these two, along with other community types , represent a rather complex mosaic that is rot as simple as Mata _e_§ a_l_. (1971) have mapped. Crassicaulescent Scrub. -- This vegetation type lies, for the most part, between the Desert Scrub and Grasslands. It could be considered an ecotone between the two but has sufficiently different plant asso- ciates to be assigned a distinct name (Mata e_t fl- , 1971 and Rzedowski and McVaugh, 1966). Some of its most characteristic plants are nopal cacti (M spp.), cholla (mg spp.), catclaw (Acacia spp.) , and Vegetation a mini ! 0 l 0| 30' 24' 23° 22' 30' 2I' Figure 4. Distribution of major vegetation types in Zacatecas: (1) Desert Scrub; (2) Crassicaulescent Scrub; (3) Grass- lands; (4) l'bntane Forests; (5) Subtropical Scrub; (6) Tropical Deciduous Forests. 13 an understory of short grasses (e.g. , Bouteloua sp.). Leopold (1950) recognized essentially the same vegetation as a subgroup of his Mesquite-Grassland. Grassland. -- Grasslands mostly occupy the western and southern parts of the state. However, in the extreIe north and northeast there are isolated grasslands (Mata _e_t_ g1. , 1971) . These are better con- sidered as a special situation of the Desert Scrub since they occur in the poorly drained interior basins (Johnston, 1978) . The grasses that occur in these isolated areas include Hilaria sp. , Sporobolus sp. , Distichlis spicata, and Bouteloua sp. Also these areas often support mesquite (P_r_gsgp_i_s_ sp. ) . They are not differentiated from Desert Scrub vegetation in Figure 4. The more mesic grasslands of the west and south coincide with isohyets above 400 mm (usually above 500 mm) of average annual precipi- tation. Dominant grasses include grane grass (Bouteloua gracilis) , beard grass (Andropogon sp. ) , muhly grasses (Muhlenbergia sp.) , drop- seed (Sporobolus sp.). In same areas the grasslands take on the appearance of a savannah or open woodland with tree acacias (Acacia tortuosa) . This was observed in the broad open region above the Rio Bolafios south of Monte Escobedo. Montane Forests. -- Mantane forests occur mostly in the western cordillera with a few isolated forests on the mountains in the north- east. For the most part, Montane Forests occur at elevations above 2,300 m. In the northeast (Cerro de Teyra, 24°32'N-102°10'W, and the noun- tains sm'rounding Concepcion del Oro, 24°30'N-101°25'W) , the forests are carposed of pinyon pine (Pinus cembroides) , scrub oak (Mans sp.) l4 and junipers (Juniperus sp.). In the Sierra Astillero (24°32'N- 101°00'N), Johnston (1978) records Douglas fir (Pseudotsuga sp.) and fir (A_b_i_e_§ sp.). I did not observe these in my one visit into that mountain range. I saw pinyon pine, oaks and Madrone (Arbutus sp.) at elevations up to about 2,500 m. Montane Forests in the west and southwest can be pure stands of yellow pine (Plug sp.), pure stands of various oaks (gu_e_rcus spp.) or a mixture of the two (Pine-Oak Forests). I am not aware of any refer- ences to fir forests in that part of the state. Below the Montane Forests there is usually a transition type of vegetation from the lower altitude vegetations to the forests. This is the Chaparral which contain scrub oaks, mnanzanita (Arctostaphylos sp.) and chamise (Adenostoma fasiculatum). Mata 91; 91. (1971) consider this to be a major vegetation type in Mexico. In Zacatecas it is so limited in extent that I have not included it as a mnaj or type. Subtropical Scrub. -- This vegetation, sometimes called dry or arid tropical scrub, is found in the soutlwestern part of the state, usually between 1,500 m and 2,000 m. It is dominated by shrubby or thorny plants less than 5 m in height. Cannon plants include the tree morning glory (Imea sp.), Ma; sp. , Bursera sp. , columnar cacti (Pachycereus sp.) , and an understory of various grasses (Mata _e__t g1. , 1971 and Rzedowski and McVaugh, 1966) . Tropical Deciduous Forests. -- The Tropical Deciduous Forest is limited in Zacatecas to the deepest river valleys of the southwest below 1, 500 m. Climatically it is characterized by occurring in the warmest, wettest areas. Tree species are usually between 8 m and 15 m in height. During the driest months of the year most, if not all, 15 trees loose their leaves. The cannon plants occurring in this forest are the same as in the Subtropical Scrub with the addition of fig trees (Ficus sp.) and bald cypress (Taxodium sp.). MATERIALS AND METHODS Species Data The species of rodents occurring in Zacatecas are the primary units of interest in this study. Data pertaining to each species dis- tribution were obtained __vi_a_ collections housed in various museums , direct field work, or fram published records . I have examined the great majority of specimens directly. Data fram approximately 6,300 specimens (5,900 specimens examined) representing 46 species have been recorded. In the case of the Mexican ground squirrel , Spermophilus medcams, I have treated the widely-separated subspecies as distinct species in the analyses of distributional patterns (see the Species Accounts for justification of this usage). Collecting sites for each species were located on 20'x15' topo- graphic maps (published by the Secretaria de la Presidencia, Camision de Estudios del Territorio Nacional, 1975) . localities were plotted directly as recorded on the specimen labels unless the collectors field notes or other available information indicated deviations fran the stated locality. For instance, specimens in the lACM labeled as "10 mi SW Concepcion del Oro, 7600 ft." were collected by Percy L. Clifton in 1970. In Clifton's field notes, it was noted that this locality was near the small mining village of la Iaja. Now, according to the topo- graphic maps, la laja is only about 5 km southwest of Concepcion del Oro (straight line distance). In 1976 I visited this same area; it is 16 17 about 16 km (10 mniles) by road southwest of Concepcion del Oro. I interviewed local inhabitants and they recalled that Percy Clifton had been in the area several years earlier. Thus, Clifton's locality is actually only about 5 km southwest of Concepcién del Oro. Problems inherent in locality data were discussed by Anderson (1965). I believe that I have identified most of the discrepancies in locality data. Species are treated individually in the section on Species Accounts. Included are specific notes on distribution, ecology, taxon- amy, and a list of specimens examined. In addition to specimens examined, I have also included site records for various sciurid species. These are easily recognized in the field but are saretimes difficult to collect. %drats For the analyses of distributional patterns, I divided the state of Zacatecas into quadrats based upon the same 20'x15' topographic mnaps used to plot the species locality data. This grid system represents sane 115 quadrats or partial quadrats located in Zacatecas. Of these, there were data available fram 73 Quadrats (See Figure 5 for quadrat number system). Since the quadrats are based upon lines of longitude and latitude, there is same variation in size fram north to south. Lines of longi- tude canverge toward the poles; thus, the size of quadrats decreases northward. This difference is only about 3 per cent in total area between the southernmost and northernmost quadrats. It is assured that 18 30. 1 .. mI-\X -25. ‘2 (136) \ KILOMETERS "3 '1. '5 - 4 s o 7 (15) (10) (m (23) ‘30, swayed no ll 12 IR g2) m) (an (12) (12) as) .4 [I4 15 no 17 no / (on (13) (7) (n7) (I2) _24. 19 2o 2: 22 23 24 (15) (n5) (n4) (14) (Ian (1 - V 25 26 27 2a 29 so .421) (12) (4) (13) (II) (2) -30' an 32 33 34 35 (I9) (13) (:4) (II) (127‘ d 36 37 as 39 4o (7) (m (5) (19) (Is) _23. 41 42 43 44 4s 46 (on 9.3) (21) (10) (II) (In) _ 4714a 49 so 51 52 5:“ (mi (9) mi) (13) (no) (19) (19X 430' D 54 ‘55 so 57, so 591 so (II) (H) m go) (133 (14) (no) _ on 42 V 3 63] an (an (2 ——/ o ‘22. 64 as 66 67 m (15) (IL) (I4)‘\ 1 -n I u 69" \ (17) (tom -30. 7o 71 72] (on (5) (n4) 1 9‘ L73 ('3) +2.. 1 l l L L L L l I 104° 103° ID? no:- Figure 5. The quadrat mmbering system (top number) and number of species per quadrat (bottam number in parentheses). 19 this difference does not affect the species canposition of rodents within any quadrat. The presence or absence of each rodent species was recorded for each quadrat. Since the political boundaries of Zacatecas are not con- tained by the quadrat boundaries , I included published data or known specimens fram other states for a few of the quadrats. The quadrats thus augmented and references for the species included are as follows: Quadrat 8 (MSU); Quadrat 13 (Dalquest, 1953) ; Quadrat 48 (Hall and Genoways, 1970; Genoways and Jones, 1973); Quadrat 55 (Schmidly, 1972; Genoways and Jones, 1973); Quadrat 57 (Hooper, 1954, 1955; lee and Hoffmeister, 1963); Quadrat 67 (Dalquest, 1953); and Quadrat 73 (Baker, 1954; KU). Species recorded for these quach'ats are listed in the appropriate Species Accounts. The number of species in each quadrat ranged from 1 to 23 (Figure 5). It is rather incredible to expect an area of any appreciable size to contain only one species. On the other hand, the quadrat with 23 species occurs in an area of high topographic relief, represented by at least three major vegetation types (see section on Environments of Zacatecas). The problem becames: which of these 73 quadrats contain sufficient data to make a meaningful analysis? (he could make a sub- jective decision that a quadrat with at least sane arbitrary number of species be included in the analysis. However, a more objective deci- sion can be made based upon the mean number of species per quadrat minus same multiple of the starndard error. This metlod would assure a normnal distribution of numbers of species amang quadrats. A plot of the mnber of species against quadrats showed that this assumption was not violated by the present data. The mean nurber of species was 12.50 20 per quadrat (SE = .53). I elected to subtract 45E fram the mean to arrive at the number of species necessary per quadrat for subsequent analysis. This resulted in 10.38 species per quadrat which was rounded down to 10. Presumably this would insure that all quadrats with at least 10 species would accurately reflect the species carposi— tion of the quadrats. Thus , of the 73 quadrats with species data, only 57 were used in the final analysis. These 57 quadrats more or less cover the entire state and should be adequate for an analysis of dis- tributional patterns. Environmental Data Data for five environmental variables were included for each of the quadrats analyzed. These included: Average Annual Precipitation (AAP); Average Annual Temperature (AAT) ; Average Terperature Range (TR); Highest Elevation (EH); and Elevation Range (ER). The two variables for elevation effectively define lowest elevation by sub- traction, thus eliminating the redundancy of using a third parameter in the analyses. The climatic variables are rather tenuous since they are taken fram climnatic maps (same source as topographic maps) and represent Isohyets and Isotherms drawn fromn a limited number of climatological stations . Isohyets in each quadrat were averaged to give AAP or data extrapolated fram the nearest isohyets . In like manner , isotherms were averaged to give AAT. The difference between the highest and lowest isotherms gave the average tenperature ranges for each quadrat. These data are believed to reflect only the major general climatic conditions within a quadrat (See Figures 2 and 3). 21 The data for the topographic variables (EH and ER) were recorded directly fram the specimen labels. In many cases they were originally recorded in feet. These were subsequently converted to meters and rounded to the nearest 10 m. Elevation data were checked against the topographic maps (see Figure l for a sumery of topographic relief in the state). In no case did I find any gross difference between recorded elevations and elevations on the topographic maps . Sane col- lectians did not have elevations recorded; these were given the eleva- tion as plotted on the maps. In no case did these appear to deviate from the elevations recorded for other localities within a quadrat. Methods of Faunal Analyses The data matrix upon which the faunal analyses were based con- sisted of data for 46 rodent distributians and five environmental variables for each of the 57 quadrats. In the analyses, quadrats represent operational taxonomic units (OTU ' s) in the sense of Sokal and Sneath (1963) while the species and environmental data are the char- acters. Two different approaches to the analysis of the data were attenpted: Cluster Analysis and Factor Analysis. All calculations were run on the CDC 6500 at the Michigan State University Computer Center. Cluster Analysis. -- Cluster analysis is a means by which OTU's are grouped successively into more inclusive sets (Rohlf, 1970; Sokal and Sneath, 1963; and Sneath and Sokal, 1973). The general method utilized herein is called Hierarchial Clustering (Rohlf , 1970) since it generates a tree or dendrogram which merges the most similar OTU ' s 22 first, adding less similar OTU's in succession. Ultimately, this results in an hierarchial schene of OTU classification. There are several algorithms available for clustering OIU's (for a review, see Sneath and Sokal, 1973). In the present study, the Unweighted Pair Group Method with arithmetic averages (UPQILA) described in Sneath and Sokal (1973) was used. Sneath and Sokal (1973) presented evidence that UPCMA gave the best results when compared with other clustering procedures. The UPGMA was carried out using a program "STRUCI'R" written by R. C. Dubes of the MSU Department of Computer Science. Clustering E the UPGMA can be based upon indices of either dissimilarity or similarity. I have used both kinds of indices to per- form cluster analysis of the quadrats (OI'U's). McIntosh (1973) sug- gested using more than one measure of association since each coefficient differently assesses the relationships. In these analyses only species data were utilized. Dissimilarity between OI'U ' s was estimated using the Euclidean dis- tance measure (d) as described by Sneath and Sokal (1973). It is cal- culated as follows: dij = ‘/’3(Xik - XJ.k)2 where Xik and xjk represent the kth species in OTU's i and j, respec- tively. Clustering based upon Euclidean distances groups OTU's with lowest values first (_i_._e_. , low values indicate closeness). In this index, only mismatches influence the values. The program ' 'ST'RUCI‘R" performs the calculation of Euclidean distances . TWO indices of similarity were utilized: Average Faunal Reserblance (AFR) recamnended by Long (1963); and, the Similarity Index 23 (S) derived by Baroni-Urbani and Buser (1976). The AFR, or some varia- tion of it, is probably the most cannonly used index of similarity in current use. It is calculated as follows: A(ni + nj) AFR where A is the mmber of species held in cannon between OTU's i and j; and mi and nj are the total number of species in OTU's i and j, respec— tively. This index expresses the similarity of two OI'U's as an average of the number of species in cannon between the two being considered. Both the positive matches (species cannon to both OI'U's) and the mis- mnatches influence the values of this index. The Simnilarity Index (S) differs fram the previous two indices by including negative matches (1.53. , the cannon absence of species between two Ol'U's). Baroni-Urbani and Buser (1976) recognize five basic parameters that need to be considered for an index of simnilarity: (l) the nunber of species in cannon between two O'IU's, A; (2) the num- ber of species occurring in OI'U i but not in OTU j, 3;. (3) the number of species in j not in i, C; (4) the number of species absent in both Ol‘U's (negative mnatches) but present in other OTU' s of the area being considered, D; and (5) the total number of species in the fauna under investigation, N. They believe that similarity can be expressed cam- pletely, only if four of these five parameters are used. They propose the following index: 3- Jm+A ” JAD+A+B+C where A, B, C, and D are defined above. 24 Baroni-Urbani and Buser (1976) discussed the cancept of including negative matches recognizing that they can be due to inadequate sampling. They further rote (p. 253) "But the wrong exclusion of a species from a sanple affects both A and D in an equal manner and there are no mathematical methods allowing good conclusions to be drawn fram a bad set of data." To be sure the data I have for at least some quadrats are probably not canplete. I believe, however, that restrict- ing the analysis to quadrats with at least 10 species minimizes this problem. In addition, I wanted to canpare the two methods of esti- mating similarity. Baroni-Urbani and Buser (1976) include a table of the distribution of S corresponding to faunas of different sizes based upon a simulation of an infinite number of randam sets of data. This table (pp. 256-257) sets confidence limits on 8, an advantage that other similarity indices do not have. In other words, the two tails of the distribution can be used to decide whether the S value is significant (3.3. , a "true" esti- mate of the value S) or a randonm value. In the present study, a probability level of .05 at either end of the distribution corresponds to S values of .36 and .62, respectively, for an N of 46. Values between .37 and .61 can be considered as randam values, possibly related to inadequate sampling. A canputer program to compute both AFR and S was written by Edward Rybak. The similarity matrices were then entered into the pro- gram STRUCI‘R for the clustering by UPGMA. Factor Analysis. -- While the various indices of similarity and clustering techniques can shed light on the classification of simnilar faunal areas, they do not identify species with similar distributional 25 patterns. Also, at least for the similarity indices, environmental variables cannot be included. These would have to be superimposed in sane manner upon the final dendrogram. Factor analysis provides a powerful tool to isolate variables that are correlated and to reduce an original multi—variable data matrix to fewer uncorrelated factors. For reviews of the mnathennatical procedures involved with factor analy- sis, see Cooley and lohnes (1971) and Harris (1975). The techniques, assunptions, and uses of factor analyses as applied to biological data were reviewed by Poole (1971), Fisher (1973), and Orloci (1973). Factor analysis is a general term for a fairly large nunber of computational procedures which have in cannon the goal of reducing a large data matrix into a smaller set of cannon factors. In deciding which of the many procedures to use, I followed the guidelines sug- gested by Fisher (1973) . He campared the five most common techniques of factor analysis, methods of selecting the number of factors to be extracted, and two methods of rotation to simple structure. Following Fisher's (1973) recommendations, I used a Principle Factor Analysis (PFA) using the squared multiple correlation coefficient of each vari- able as the diagonal in a correlation (Pearson product-moment correla- tion matrix) mnatrix. The data matrix utilized for PFA consisted of 43 rodent distribution and five environmental variables. Three rodent species had to be deleted fram the analysis since they occurred in only one quadrat each. These were: Liamys pictus, Quadrat 47; Perannyscus Mai, Quadrat 7; and Peranyscus melanotis, Quadrat 42. An initial factor matrix of 48 principle axes (factors) was extracted. I used this initial factor matrix to determine the number of factors to be extracted and rotated according to the procedures outlined by Fisher 26 (1968, 1973). The sun of the eigenvalues for each factor was plotted against the factor number, on log-log paper. The curve is straight until factor VII then the slope of the line changes abruptly. This indicates that seven factors is a good estimate of the relevant common factors in the data set. I also extracted 10 factors and rotated these but too many variables had loadings on the factors near zero, indi- cating that 10 factors were too mnany (Fisher, 1973). On the other hand, five factors did not sunmarize the data sufficiently to be useful. The seven factors accounted for 60.9 per cent of the total variation and 66.1 per cent of the common variation. Rotating the initial factor matrix is a means by which the matrix is simplified for easier interpretation (Cooley and lohnes ; 1971, Davis, 1973). Two main procedures are available: ortlogonal and oblique rotations. Fisher (1973) found no basis upon which to prefer one over the other. He suggested that both be attempted, canpare the results , and select the technique that gives the most easily inter- pretable results . I tried both metlods and each gave biologically meaningful results. In fact, each method gave samewhat similar results. However , I found that the orthoganal rotation gave what I believe to be the most easily interpretable factor matrix. Therefore, I used this method of rotation for the present study. The only other analyses of species distributions that I am aware of used only the oblique rotational procedures (Fisher, 1968 ; Smnith and Fisher, 1970; Stevenson _e_t 31. , 1974; Schnell _e_t_; g1. , 1977). All factor analytic procedures were accanplished using the SPSS routine FACTOR with Varimax rotation described in Nie _e_t_ _a_1_. (1975) . 27 The loadings of variables onto each rotated factor are essentially correlation coefficients of those variables with the factor (Poole, 1971). As such, the loadings can be used as indices of the importance of a variable to a factor. Smnith and Fisher (1970) used loadings of .300 (positive or negative) as limits for inclusion of a variable into a "Factor Group" (_i_._e_. , identifying variables with factors). Stevenson g§_§l, (1974) and SChnell g; g1, (1977), on the other hand, used load- ings of .400 to identify variables associated with a particular factor. As pointed out by Smith and Fisher (1970) , this is an arbi- trary aspect of factor analysis. I considered both .300 and .400 as well as .500 as cut-off points for variables associated with a factor. I found that variable loadings of .400 gave what appeared to be the nmost intuitively satisfying results in the present study. Factor scores for each quadrat were computed using the stan- dardized data matrix and the camplete estimation nmethod described in SPSS (Nie g; 11; , 1975). Each quadrat can then be given a camposite value on each factor. This score allows the quadrats to be ordinated with respect to their species canposition and the environmental vari- ables on each factor. The factors are then naned or identified on the basis of the major habitat areas defined by the loadings of variables and factor scores of the quadrats. RESULTS Cluster Analysis Dendrograms resulting fram the UPCMA of d, AFR, and 8 matrices of species distributions are presented in Figures 6 to 8. Generally, all three indices produced rather similar clusters of quadrats. Most dif- ferences were simple reversals of order of entry into a cluster. For exanple, in the dendrogram of d and S values (Figures 6 and 8), a cluster of Quadrats 8 and 20 was joined next by Quadrat 39. However, in the dendrogram of AFR values (Figure 7), the initial cluster was of Quadrats 20 and 39 joined next by Quadrat 8. This kind of difference between dendrograms is considered a minor one. On the other hand, a few major differences occurred which involved shifts of one or more quadrats between clusters. For exanple, in both d and S dendrograms a cluster of three Quadrats (8, 20, and 39) can be discerned. This sane cluster can be seen in the AFR dendrogram but with the addition of Quadrat 21. This quadrat formed a separate cluster with Quadrats 3 and 17 in the d and S dendrograms (discussed more fully below). Another major shift involved Quadrat 7 which in the d dendrogram is very dis- tinct, not joining with a cluster until far down the tree. In both the AFR and S dendrograms, Quadrat 7 formns a cluster with Quadrats 4 and 6. Reasons for these (and other) mnaj or differences between dendrograms are not readily apparent . Presumably, they would have to do with the 28 29 DISTANCE DENDROGRAM 4; 4.0 3,5 an 2.9; 2.0 1.5 no 7 8 ____r_—{ 20 39 r 4 4H 1 6 15 1 5 l 9 ___{———-E n: L— U l—_3 u can N-‘wNwN-‘N-‘u-‘d wmgmmmmnw‘r“ ”N r—l—i . ..o 57 _ 42 j 4 - 2 65 a U Gib roan l a & Cluster Ill s: on N I C luster IV Coplnenetic Correlation Coefficient: .85 Figire 6. Dendrogram of quadrats (rnunmber on right) based upon UPGMA clustering of the distance index and showing the four major clusters. 30 AVERAGE FAUNAL RESEMBLANCE DENDROGRAM .30 .40 .50 .60 .70 .80 .90 10 g l l L l 39 13 12 17 Cluster l 15 .;_4 11 23 32 F‘L——n:29 35 ‘ 18 —l——t:28 34 22 'E33 40 A q 37? A q B U Cluster l I 68 - l 26 5 L— I U 51 53 60 - 55 t——. 50 UU JIJI Cluster Ill *1 D m 67 46 ___[——I:52 63 25 .L_r—-I 31 69 -——‘ ..____.[ 4 ——m 54 r 61 __ Cluster IV 73 Cophenetic Correlation Coefficient=.82 Figure 7 . Dendrogram of quadrats (umber on right) based upon UPGMA clustering of the Average Faunal Resemblance index and showing the four major clusters. 31 SIMILARITY DENDROGRAM go so go .79 go so 1J0 J——C 1 (DO aumewnm Cluster I buNdewNuNdewdd » owwemmmwuudau ”Mum 5‘5 0‘ I (0" CV l Cluster ll lb b mu C U \l J (D U! CID J) h MI 00 0'1 _e 0301 COD l1 lh Hf acme 0m Cluster Ill CD \I WM ‘01 5 NO: 0'” Cluster IV gamma: N001“ '——L ———-L l J L \IV (0N Coplnenetic Correlation Coefficient:.80 Figure 8. Dendrogram of quadrats (number on right) based npon UPCMA clustering of the simnilarity index and showing the four mnajor clusters. 32 nature of the original similarity or dissimilarity matrices, since the sane clustering technique was used to produce all three dendrograms. Which of the three dendrograms produces the "best" clustering of quadrats? By "best" I mean the gaping of quadrats into a meaningful biogeographic classification (the recognition of fanual areas). Sneath and Sokal (1973) suggested that the best classification would be the one which most closely represented the original similarity (or dis- similarity matrix). They suggested using tte cophenetic correlation coefficient (CC) as a basis for deciding which classification (or dendrogram) was best. In the present study, the CC for the d dendro- gram was .85, while AFR and S dendrograms had CC valLes of .82 and .80, respectively. Thus, using the criteria of Sneath and Sokal (1973) , the dendrogram of d values would be the best of the three. However, both AFR and S are reasonable dendrogranns and the CC's of these are not too different fran that of d. In addition, same of the clusters produced by both AFR and S are intuitively more reasonable for biogeographical classification than those produced by the d dendrogram (e.g. , the rela- tionship between Quadrats 4, 6, and 7). There is no generally accepted method of determining the cut-off points for recognition of distinct clusters. Since all three dendro— grams gave reasonable CC values, I critically inspected and compared the three dendrograms. I used two criteria to define clusters, as fol- lows: (1) within a dendrogram the similarity values were approximately the same for each cluster; (2) the sane clusters were cannon to the three dendrograms. Fran this canparison, I was able to recognize four major clusters in cannon between the three dendrograms. These four major clusters more or less correspend to the major geographic and 33 vegetational areas within the state. Subclusters within these mnajor clusters do not show the same congruence between dendrograms . Certain of these subclusters will be discussed later. The four major clusters and rodent species restricted to then are as follows: Cluster I. -- Cluster I is defined by the following values on each dendrogram: d, 3.6-3.7; AFR, .60; and S, .65. This cluster of quadrats corresponds to the eastern half of Zacatecas . It includes areas of vegetation described earlier as Desert Scrub (including the atypical grasslands), Crassicaulescent Scrub (in part), and the Montane Forests of the north and northeast. Physiographically, it includes the Central Mesa and Cross Ranges. Cluster I has the greatest nnumber of species associated with it. These are Spermophilus mexicanus parvidens, Cyngmys mexicanus, Thanomys bottae, Pappogeanys castanops , Perognathus penicillatus , Dipodanys nelsoni, Peranyscus eremicus , and P. hgoperi. Only four of these species (1:. cast_a_n_ops, P. penicillatus, D. nelsonni, and P. eremicus) are regularly found throughout the area defined by Cluster I. Cluster II. -- This cluster is defined by the following values: d, 3.6; AFR, .61; and S, .66. It corresponds, in part, to the Montane Forests of the west . Other quadrats that have western Montane Forests are defined in other clusters. Only one rodent species is cornfined to this cluster (Perannyscus melanotis) and it is found only in Quadrat 42. Other rodents that would normally be considered part of the Montane Forest habitat cannot be included since they are not restricted to this cluster. This represents one of the problem associated with cluster analysis . 34 Cluster III. -- This cluster is defined by the following values: d, 3.5; AFR, .57; and S, .62. It includes quadrats of the west and south that correspond to Grassland, and in part to Crassicaulescent Scrub and Subtropical Scrub vegetations. It is essentially the Eastern Upland Subprovince of the Sierra Madre Occidental Province. Subclusters within this cluster are more heterogeneous than can be observed in Cluster I. There are no species of rodent completely con- fined to this cluster. Cluster IV. -- The last cluster is defined by the following values: d, 3.9; AFR, .42; and S, .54. The high d value and low AFR and S values indicated that this cluster is the most heterogeneous of the four. It essentially corresponds to areas of Tropical Deciduous Forests, and in part to Montane Forests , Grasslands, and Subtropical Scrub vegetations . The main cannon feature that quadrats have in this cluster is their geographic position associated with the Canyon Region of the south and soutlmest . Thus the cluster heterogeneity can be explained by the great anount of environmental diversity associated with the Canyon Region. Six species of rodents are restricted to this cluster: Spermqphilus mnexicanus mexicanus, Sciurus , au'eogaster, Liamys pictus, Peramyscus gpicilegus, Signodon mascotensis, and Neotama pglatina. An interesting situation occurs with Quadrat 45 which was only weakly linked to Cluster I in all three dendrograms. Because of its geographical locatian, Quadrat 45 could be considered an area of eco- tone between Clusters I and II. Also, in the AFR dendrogram, Quadrat 37 was only weakly linked to Clusters II and III. It would appear to be an ecotone between these two clusters. However, recognition of 35 ecotonal areas are not as easily discernable by cluster analysis as they would be with an ordination technique (see below). Factor Analysis The seven factors extracted fran tte original data matrix group the rodent distributional and environmental variables into what I believe to be meaningful patterns. Factor analysis overcanes some of the inadequacies of clustering techrniques to recognize associated species distributions and environmental variables . However, this is accomplished at the expense of not always being able to distinctly recognize groups of quadrats as faunal areas. Results of the factor analysis are given in Tables 1 to 7 and in Figures 9 to 15. Variable loadings associated with each factor are essentially correlation coefficients indicating the relative contri- butionn of each variable to a factor. Since a variable loading of .40 or greater was arbitrarily selected as the inclusion limit for a vari- able to a factor, it allows for a variable to be associated with more than one factor. Sane rodent species did not have loadings of .40 or above on any factor. These species can be connsidered as representative or independent distributions. They are listed with the factor to which they have the highest correlation. Factor scores computed for each quadrat were grouped into three classes representing low, intermediate, and high score, respectively, as follows: -l.00 or less; -.99 to +.99; and +1.00 or greater. These three classes allowed for easy interpretationn of scores since the scores were computed fromn standardized data matrix. Thus the scores themselves are standardized to have mean zero and standard deviation of 36 one. In all cases, high positive scores were easily interpreted in association with a given factor. With the exception of Factor I , nega- tive scores were interpreted simply as "not being associated" with the positive scores. The distributional patterns explained by the seven factors more or less correspond to major geographical and/ or vegetational areas in the state. Factor VII was the only one that was not associated with a definite geographical or vegetational pattern. The seven factors are as follows: Factor I: Desert versus Grassland Habitats. -- Seventeen species of rodents and three environmental variables were associated with Factor I (Table 1). The four highest loadings are negative and the species involved are usually considered to be desert inhabiting species (Table 1; canpare these four with the species associated with Cluster I in the section on Cluster Analysis). The rodents that have high posi— tive loading are usually associated with grassland though not neces- sarily restricted to it. Peranyscus truei has usually been considered a montane species but one with a highly diverse habitat range (Baker and Greer, 1962). In the present analysis, _P_. _trn_1_e_i_ had its highest correlation with Factor I and does not load at .40 or above on any other Factor. This would indicate that, at least in Zacatecas , P. £91523; might best be considered associated with grassland (it was also found in montane areas). Five species have their highest, although less than .40 loading on Factor I . Both Spermophilus mexicanus parvidens and Perognathus lineatus are known fram only two quadrats each (see Species Accounts). Their low correlations may reflect either insufficient data to 37 TABLEl VARIABLES ASSOCIATED WITH FACIOR I: DESERT VS GRASSIAND HABITATSl’2 -.77 -.75 -.69 -.66 +.62 +.58 (-.54) (+.52) +.52 +.51 (+. 50) +.50 (+.46) -.45 +.43 +.39 +.36 -.35 +.30 +.22 Pappogeomys castanops Dipodannys nelsani Perognathus penicillatus Pergnyscus eremicus Dipodomys phillipsii Peramyscus truei Dipodomys merriamni Thomrxmys umbrinus Elevation High (+.62 on IT) Spermophilus variegatus Perggnathus hispidus Liggys irroratus (+.56 on VII) Average Annual Precipitation (+. 48 on II) Average Annual Temperature (+. 55 on VII) Reithrodontamys nnegalotis Reithrodontagys fulvescens Signodon fulviventer Sgermophilus mexicanus parvidens Perognathus flavus Perognathus lineatus 1loadings associated with a variable are given on the left. Those loadings enclosed with parentheses indicate that a variable is also associated with another factor. Variables followed by parentheses indicate higher loadings on another factor . 2Variables with loadings below .40 are given only when that variable does not load any higher on another factor. 38 characterize their distributions or that their distributions are inde- pendent. On the other hand, Reithrodontannys fulvescens , Sijmodon fulviventer, and Perognathus flavus are known from a number of quadrats throughout the state. This would indicate that the distributions of these species are independent . Four species (Dipodomys merriami , 'Ihcmomys umbrinus , Perognathus hispidus , and Ligyys irroratus) were also associated with other factors. The three environmental variables associated with Factor I were tlnose that more or less correspond to differences between Desert Scrub and Grassland Vegetations . Elevation High and Average Annual Precipitation load positively, thus in the same direction as the grass- land rodent species couplex. Average Annual Terperature loads nega- tively indicating the relatively higher tenperatures associated with the eastern (Desert Scrub vegetation) part of the state. All environ- mental variables had higher associations with other factors. The map of factor scores (Figure 9) indicates the geographic separation of high and low scores . High scores were concentrated in the west and south in areas corresponding in part to Grassland, Crassicaulescent Scrub , and Subtropical Scrub vegetations (compare with Figure 4). Low scores , for the most part, were in quadrats that had Desert Scrub vegetation. Factor II: Western Mnntane Habitats . -- Eight rodent distributionn and four envirionmental variables were associated with Factor II (Table 2). All variable loadings were positive. The four highest loadings were with species that are restricted to areas associated with Montane Forests. Neotama mexicana and Peromyscus difficilis had their highest 39 Foctorl Scores - .99 to +99 3 + 1.00 25. 24' 23' 22’ 2I' Figure 9 . Distribution of the three classes of factor scores for Factor I in Zacatecas. Factor scores ranged from -2 . 64 to +1.96. 40 TABLEZ VARIABLES ASSOCIATED wrm FA R II: WESTERN mNrANE HABITATS ,2 +. 84 Nelsonia neotcmodon +.72 Sigmodon leucotis +.72 Sciurus nayaritensis +.64 Eutamias bulleri (+.64) Neotoma mexicana (+. 62) Elevation High (+.56) Elevation Range (+.56) Percmyscus difficilis +.49 Peranyscus boylii (-.51 on V) (+. 48) Average Annual Precipitation +.48 Thcmcgrs umbrinus (+.52 on I) +.40 Temperature Range (+.45 on III) 1Loadings associated with a variable are given on the left. Those loadings enclosed with parentheses indi- cate that a variable is also associated with another factor. Variables followed by parentheses indicate higher loadings on another factor . 2Variables with loadings below .40 are given only when that variable does not load any higher on another factor. 41 loadings on this factor but were also associated with other factors . Both Peronyscus boylii and Thommys unbrinus had higher loadings on other factors . Of the environmental variables, Elevation High, Elevation Range, and Average Annual Precipitation all had their highest loadings on Factor II. These variables correspond to what one might expect in mon- tane areas. Temperature Range was also associated with this factor but had a higher loading on Factor III . Quadrats with high factor scores are founnd in areas of the west and southwest that have Montane Forests (Figure 10, canpare with Figure 4). Factor III: Cross Ranges Habitats. -- Four rodent distribution and two environmental variables were associated with Factor III (Table 3) . Three of the rodents had their highest loadings on this factor while the fourth had a higher loading on Factor II. Both W b_<_>_t_:_t_a_e and Cynomys nnexicanus have distributions in Zacatecas restricted to the Cross Ranges Subprovince (I. b__o_t__ta__e_ to the mountains specifi- cally end 9. mexicanus to the intermontane valleys). The two environmental variables (Temperature Range and Elevation Range) are indicative of desert mountain areas . Factor scores above +1 . 00 are restricted to the three quadrats in the northeast (Figure 11) that correspond to the Cross Ranges described in the sectionn on Phys iography and Topography. Factor IV: Canyon Region Habitats. -- Five rodent species and two environmental variables were correlated with Factor IV (Table 4). With the exception of Spermophilus spilosoma, all species load positively on this factor. Three species, Neotcma glatina, Sigmodon mascotensis , 42 Factor ll Scores NUMEHEB d 24. Classes -99to+99 u 23. z+100 :- 22. '2" Figure 10. Distribution of the three classes of factor scores for Factor II in Zacatecas . Factor scores ranged from -1. 46 to-HL81. 43 TABLE 3 VARIABLES ASSOCIATED WTI‘H F R III: CROSS RANGES HABITATS ' +.78 Thomcmys bottae +.6l Neotama goldmani (+. 45) Temperature Range +.44 Cynomys mexicanus +.lnl Elevation Range (+.56 on II; +.52 on TV) +.40 Peronnyscus difficilis (+.56 on II) 1Loadings associated with a variable are given on the left. Those loadings enclosed with parentheses indicate that a variable is also associated with another factor. Variables followed by parentheses indicate higher loadings on anotlner factor. 2Variables with loadings below . 40 are given only when that variable does not load any higher on another factor. 30' FoctorlH Scores Figure 11. Distribution of the three classes of factor scores for Factor III in Zacatecas. Factor scores ranged frcm -l.58 to 44.30. 45 TABLE4 VARIABLES ASSOCIATED WITH FA R IV: CANYON REGION HABITATS ' +.78 Neotama palatina +.76 Sigmodon mascotensis (+.52) Elevation Range (+.56 on II) +. 50 Peromyscus pectoralis -.42 Spermophilus spiloscma (+.45 on V) +.4l Spermophilis mexicanus mexicanus (+.4l) Average Annual Precipitation (+.48 on II; +.46 on I) lLoadings associated with a variable are given on the left. Those loadings enclosed with parentheses indicate that a variable is also associated with another factor. Variables followed by parentheses indicate higher loadings on another factor. 2Variables with loadings below .40 are given only when that variable does not load any higher on another factor. 46 and Spermophilus mexicenus mexicanus , are restricted to the Cenyon Regions . Perognyscus pectoralis is widely distributed in Zacatecas but does not load above .40 on any other factor. Of the two environmental variables, Elevation Range would be most indicative of the Cenyon Regionn while Average Annnual Precipitation corresponds to the tropical vegetation of the canyon bottoms . High factor scores were associated with the quadrats of the Cenyon Region described in the sect ion on Phys iography and Topography (Figure 12). Factor V: Desert-Grassland Ecotone. -- Nine rodent distribution and one environmental variables were associated with Factor V (Table 5). Positive loadings were obtained for seven of the nine rodents. Only Dipodomys merriami had a higher, although negative, loading on another factor. These seven species have in cannon distributions in Desert Scrub, Grasslend, end Crassicaulescent Scrub vegetations more or less centered in the central end east central part of the state. The two negative loading species (Peronnyscus boylii annd Baiggys taylori) have distributions nnore to the west end south. The single environmental variable (Average Annnual Precipitation) associated with Factor V had higher loadings on other factors . It would more or less correspond to the increased rainfall of the west as Opposed to the eastern part of the state. Factor scores for each quadrat show the highest values in areas that would correspond to en ecotone between grassland end desert habi- tats (Figure 13). Factor VI: Southern Canyons Habitats . -- Five rodent species but no environmental variables were associated with Factor VI (Table 6) . 47 O Foctorl << (D n O “I (D tn Figure 12. Distribution of the three classes of factor scores for Factor IV in Zacatecas. to +4.36. Factor scores ranged from -l.22 TABLES VARIABLES ASSOCIATED WITH F R V: DESERT-GRASSLAND ECOTONE » +.67 +. 64 +. 57 +. 51 (- . 51) +. 50 - .46 (+1.5) (- .41) +.4O Dipodanys spectabilis Dipgdcmys ordii Onychomys torridus Perognathus nelsoni Peranyscus boylii Dipodomys merriami (-.54 on I) Baianys taylori Spenmophilus spiloscma Average Annual Precipitation (+.48 on IT; +.46 on I) Pergngyscus melanophrys lLoadings associated with a variable are given on the left. Those loadings enclosed with parentheses indicate that a variable is also associated with enother factor. Variables followed by parentheses indicate higher loadings on enother factor . 2Variables with loadings below .40 are given only when that variable does not load eny higher on enother factor . 49 Factor V Scores KILOMETERS - 24- Classes ................... : -.99 to +99 —1 230 z + 1.00 .......... ..... ..... a... ..... ..... ..... .-:-:-:-:-. - 2 2. .......... - 2l‘ Figure 13. Distribution of the three classes of factor scores for Factor V in Zacatecas. Factor scores ranged from -2. 65 to +2 . 22 . 50 TABIE6 VARIABLES ASSmIATED WITH F VI: SOUTHERN CANYONS HABITATS ’ +. 94 Sciurus aureogaster +.82 Peronyscus spicilegus - .43 Peromyscus maniculatus +.4O Neotama mexicana (+.64 on II) -.35 Neotama albigula lLoadings associated with a variable are given on the left. Those loadings enclosed with paren- theses indicate that a variable is also associated with another factor. Variables followed by paren- theses indicate higher loadings on enother factor . 2Variables with loadings below . 40 are given only when that variable does not load any higher on another factor . 51 Three of the five species load positive with the greatest influence from Sciurus aureogaster end Peronnyscus spicilegus. Neotama mexicana loads low on this factor having more influence on Factor II. Peromyscus meniculatus was negatively associated with Factor IV. It is a widespread species but was absent fronn the two quadrats that had high factor scores (Figure 14). Neotoma albigula is onnly loosely asso- ciated with Factor VI. Its widespread distribution in Zacatecas and low loading on all factors would indicate an independent pattern for this species. The high factor scores for two of the southernmost quadrats (Figure 14) reflects the influence of _S_. enreogaster on Factor VI. This species is known only from these two quadrats. Factor VII: Signpodon hispidus-Liam's irroratus Connplex. -- Three rodent species and one environmental variables were associated with Factor VII (Table 7) . Both Signodon hispidus end Lianys irroratus had their highest loadings on this factor. Perognathus hispidus had a higher loading on Factor I. Average Annual Temperature had its highest loading on Factor VII but was also associated with Factor I. Factor scores for the quadrats (Figure 15) do not show any parti- cular geographic associationn. Thus , this factor seens to be associated with the unique distributional patterns of the species involved. One connnron denominator that was found is en associationn with mesic or riparien situations (see Discussions below). Of the 57 quadrats analyzed, onnly six did not score high on any factor. These are considered to be associated with the factor for which they had their highest scores. Quadrats 28 end 32 with scores of +. 60 end +.42 on Factor V, respectively, are considered to be a part of 52 Factor VI Scores KHIMMETERS ~ 24- Classes s-n.oo ...... n.- ..... .......... = -.99 to +99 a 23° 2 +1.00 .et "22’ ................... ............ ................ .......... ............ ............... T Zl' Figure 14 . Distribution of the three classes of factor scores for Factor VI in Zacatecas. Factor scores renged from -l.7l to +5.26. 53 TABLE7 VARIABLES ASSOCIATED WITH FACTOR VII: SIGMODON HISPIDUS-LIOMYS IRRORATUS camel» 2 +. 65 Sigmodon higpidus (+. 56) Lionys irroratus (+. 55) Average Amnual Temperature +.43 Perognathus hispidus (+.50 on I) 1loadings associated with a variable are given on the left. Those loadings enclosed with parentheses indi- cate that a variable is also associated with enother factor. Variables followed by parentheses indicate higher loadings on enother factor. 2Variables with loadings below . 40 are given onnly when that variable does not load any higher on enother factor. 54 Factor VII Scores KILOMETERS - 24- - - .99 to +99 i 30‘ a 22- - 2|. Figure 15. Distribution of the three classes of factor scores for Factor VII in Zacatecas . Factor scores renged frann -2. 22 to +2. 30. 55 the desert-grassland ecotone. Geographically, these quadrats would also fit that pattern. Quadrat 20 scored highest (+.48) on Factor 1, indicating either a grasslend association or en ecotone association. Similarly for Quadrat 49 (+.59) on Factor I. Quadrat 37 scored highest (+.39) on Factor II, the Western Montane habitats but also scored only slightly less (+. 36) on Factor I. This may go along with its position in Cluster II or III of the AFR dendrogren (Figure 7) . Quadrat 29 had its highest score (+.29) on Factor I, indicating only a weak associa- tion with the grasslends. It is probably better associated with the ecotonal area. Quadrat 54 scored highest (+.94) on Factor IV, indi- cating an association with the Cenyon Region habitats. DISCUSSION Cluster Analysis Differences between the various dendrograms of quadrats based on d, AFR, and S matrices were considered to reflect the nature of the original calculation of the indices. The distance index considered only differences (misnnatches) between quadrats. Thus, even if two quachrats had many species in cannon, they may not be close by d-values when the number of differences is also large. For example, Quadrat 7 is only loosely associated with Cluster I in Figure 6. Quadrat 7 had its lowest d—value with Quadrat 6 (d = 2.64) with seven species differ- ences. These two quadrats had 17 species inn cannon. On the average, Quadrat 7 was very different from all other quadrats in Cluster I. This average difference is reflected in its inclusion at a d-value of 3.7. The distinctness of Quadrat 7 is not apparent in the AFR or S dendrograms (Figures 7 end 8). The d dendrogram distorts, to sane extent, the relationship of Quadrat 7 to the other quadrats . While differences should be considered important in determining relationships between quadrats, complete reliance on differences would distort these relationnships. 0n the other hand, d-values as used here provide a check on the other indices. McIntosh (1973) suggested that there is an advantage in canparing the results of more than one index of asso- ciationn. This is true because each index differently assesses 56 57 relationships between the quadrats; thus, each will give slightly dif- ferent results (McIntosh, 1973) . AFR end S matrices produced dendrograms that were , in general , more similar to each other then either was to the d dendrogren. Of course this might be expected since each gives more weight to cannon species. The S index, how-Iver, does include negative matches. There was sane congruence between subclusters of S end d dendrograms that was not found in the AFR dendrogram. For irnstence, both d end S dendro- grens (Figures 6 and 8) have a subcluster cauposed of Quadrats 8, 20, end 39. This sane subcluster is present in AFR but with the addition of Qiadrat 21 (Figure l) , Quadrat 21 formed enother subcluster with Quadrats 3 end 17 in both d end S. This difference cen be attributed to the total number of species that are different between eny two of the quadrats. Table 8 gives the matrix of data upon which these sub- clusters are derived. The We of species in each quadrat is on the diagonal, the species in cannon between eny two quadrats is above the diagonal, end the species differences below the diagonal. By the upper half of the matrix (species in cannon) , it is not at all obvious where Quadrat 21 would be in relation to the other quadrats (_i_._e_. , Quadrats 8, 20, 39 on the one hend end 3, 17 on the other). Because a quadrat must be assigned to only one cluster by the method used here, the intermediate positian (based on cannon species alone) of Quadrat 21 was masked by the AFR index. However, the lower diagonal clearly shows that there were fewer differences between Quadrats 21 end 3, 17 then for 8, 20, 39. A similar situation also occurs for Quach'at 37 in the AFR dendrogren. 58 TABLE8 SELECTEDQUADRATSINZACATECAS CQTPARISONOFNUMBERSOFSPECIESIN 21 17 39 20 8 Quadrats memowmm M5 0898 8 1 3 2 3 l l l l 8 2 4 5 7 l l l l 7 l 2 6 3 1 1 n__ u w... w__ u s u :13: 6 9 8 u: s a u u s m w 3 u mmosaummflo mmwomam 10 21 59 There were a number of quadrats that might also be cendidates for other clusters but due to the dendritic nature of hierarchial c1uster- ing, a quadrat must be assigned only to one cluster (Fisher, 1968). This represents one of the major drawbacks to cluster analysis. A good exenple of this is Quadrat 47, identified with the Canyon Regions . Col- lecting localities in Quadrat 47 included areas that would be classi- cally considered representative of nnontene habitats with the following species Eutenias bulleri, Sciurus nnayaritensis, Peranyscus boylii, P. difficilis, Thanomys umbrinus, end Nelsonia neotomodon. In the cenyon bottans were collected Lia_nys pictus, Neotama palatina, end §_i_gno_d_on_ mascotensis species associated with more tropical habitats. On the flatlends above the Rio Atengo , Qnychanys torridus , Dipodogys phillipsii, end Liaqys irroratus were collected. These are indicative of the grassland habitats to the east. Thus, Quadrat 47 could pos- sibly be a cendidate for Clusters II or III but was clustered with IV, possibly because of this "unique" species association. The supposed "intelligent ignorenous" as defined by Sokal end Rohlf (1970) having no knowledge of the species involved would not recognize this problen. Thus, a classification of faurnal units would be dictated by the defined clusters as given in Figure 16. Although this is a reasonable classification, it does tend to distort the real picture. I believe that the use of objective methodology must be tempered with cannon sense end experience. The methods of numerical taxonany (Sneath end Sokal, 1973) should be used but not necessarily dogmatically accepted. Deviations fran their use, however, should be explained on reasonable grounds . I O KILOMETERS L I l l l L l I l l [04" IO3' l02' IOI' 30' 25° 30' 24‘ 30' 23' 30' 22' 30' 2I° Figure 16. Distribution of the four major clusters of quadrats in Zacatecas based upon the dendrograms in Figures 6 through 8. 61 The four major clusters that were defined by all indices were based solely upon rodent distribution data. The map of these four clusters (Figure 16) shows that they are essentially geographically defined end conform to the major vegetationnal areas in the state. As such, they represent ecogeographic units (_s__e_n_s_u Udvardy, 1969 and Armstrang, 1972). Recognizing four ecogeographic units is intuitively appealing since they are more or less consistent with ecogeographic units of more classical studies (Life Zones of Merriam, 1898; Biotic Provinces of Dice, 1943, and Goldman and Moore, 1946; and faunal areas of Baker end Greer, 1962) . Anderson (1972) recognized three ecogeographic units in Chihuahua (tropical, montene, and desert-grasslend). He was nnot able to differentiate between a distinct desert end grasslend fauna based on distributions of nnarsupials, insectivores, bats, laganorphs, end rodents. Anderson (1972) used the AFR index to assess the relation- ships between a trensect of nine quadrats across Chihuahua. At least three possible explenations could account for Anderson (1972) not dis- tinguishing between desert end grasslend femas. First, it may be as Anderson (1972) suggested, there simply is nno distinct differences between tlnese two (at least in Chihuahua) based on mammalian distri- bution patterns. Second, Anderson (1972) included five orders of men- mals in his analysis. It may be that the similarity values generated fran the increased number of species would not show the distinction that rodents alonne show. Also, many of the species in these other orders are not as restricted in their habitat requirements as are meny rodents. This would tend to increase the similarity values since there would be more species in cannon between desert end grasslend 62 habitats. Using only one index, end that based only on cannon species, might bias the results. Third, Anderson (1972) used quadrats of vary- ing sizes (100 x 30-100 kilaneters on a side). Thus, it may be that the quadrats used were too large to detect the differences between desert end grasslend faunas. A similar argument was given by Murray (1968) against sane of the mammalian faunal mnits recognized by Hagneier end Stults (1964). Factor Analysis Factor analysis allowed for the identification of importent species end environmental variables associated with each factor. The factor scores permitted en evaluation of each quadrat relative to a factor. Restriction of factor scores to these classes (low, inter- mediate, and high), while permitting easier interpretation, defeats part of the objectives of factor analysis. This objective is an ordi- netion of the quadrats with respect to a given factor. However, one of the objectives of the present study was to see if recognizable general patterns of distribution could be objectively determined. I believe that this objective has been realized in both the cluster and factor enalyses. The distortion produced by using only three classes of factor scores is offset by its ease of interpretation. Since the factor scores are stendardized (i3. , mean of zero, standard deviation of one), the intermediate class should contain 66.7 per cent of the quadrats. This meens that only about 16.7 per cent of the cases would have scores higher than +.99 (end likewise for low scores). In other words, I have set a confidence level of approximately .33 in 63 recognizing distinct quadrats. In most cases, the quadrats associated with a given factor had scores far above the .99 cut-off point. 0f the seven factors extracted, three factors (I, 11, end IV) define essentially the sere ecogeographic units defined by cluster enalysis. Three factors (III, V, end VI) subdivide these ecogeographic unnits , thus accounting for the differences between quadrats asso- ciated with Factors I, II, and IV and those in Clusters I to IV. The last factor (VII) represented a distribution unique to gm hispidus and Lianys irroratus. The ecogeographic units defined by the six factors were shown in Figures 9 to 14. Quadrats associated with desert habitats (low scores in Figure 9) were moved somewhat to the east canpared to desert distri- bution of Cluster I (Figure 16). This discrepancy was partly explained in Factor V, the desert-grasslend ecotone. Similarly, high scores on Factor I (Figure 9) do not show the contiguous quadrat distribution shown in Cluster III (Figure 16) . Sane of the quadrats of Cluster III have their highest factor scores on Factor V (Figure 13) . Essentially then, Factors I end V define a desert and grasslend fauna with a rather broad area of transition between the two. This may be a reason for Anderson (1972) not being able to distinguish between these two femal areas, especially with the large size of his quadrats. Factor V (Figure 13) as an ecotonnal area also corresponds geo- graphically to the Saladen Filter Barrier described for the Chilnualmen Desert herpetofauna by Morafka (1977, 1978) . One of the highest scores on Factor V occurs in Quadrat 23. This quadrat fits closely to the northeasterrmost limit of the Saladen Filter Barrier as mapped by Foran (1977, 1978). This filter barrier was described as en area of 64 trensition between the Saladen Subprovince of the Chihual'ma Desert Biotic Province and the Transvolcenic Biotic Province further to the south (out of the present study area). Morafka (1977) described the transition of vegetation as one proceeds south from the desert in Zacatecas as a mixture of nesquite grasslend with patches of desert scrub. To determine whether or not the Saladen Filter Barrier is in fact cannon to both the herpetofaunna end to rodents (or manuals in general) would require further extensive field work in Sen Luis Potosi end south. The suggestion of its importence to rodent distribution pattern in Zacatecas, however, is clear by Factor V. The integrity of a desert rodent fauna defined by Cluster I was also not upheld by Factor I for Quadrats 4, 6, end 7. These three quadrats were associated with Factor III, the Cross Renges habitats (Figure 11). In this instance, the quadrats form a distinct subcluster of Cluster I in both AFR end S dendrograms Figures 7 end 8. Thus, it seems best to consider Factor III as a subunit of the desert faunna. It is not unconnnon in desert areas to recognize isolated nontene situa- tions (Baker, 1956; Findley end Caire, 1978; Packard, 1978 ; Schnidly, 1978). In fact, nontene habitats in desert areas have received nuch attention becense of their importence as biogeographical nodels (Johnson, 1974, 1975; end Brown, 1971, 1978). In this sense, Factor III may be the only cendidate as a factor denending a historical argu- nnent. Smith end Fisher (1970) in their study were able to recognize two factors that denended historical explenations. In the present case, Factor III scores highest on quadrats in the three nountain areas associated with the Cross Renges which are en extension of the Sierra Madre Oriental (see section on Physiography end Topography). The 65 actual factor scores decrease fran east to west in these three quadrats as, 4.30, 3.86, end 1.01, respectively. If these scores are considered indicative of closeness to a source area (the Sierra Madre Oriental) then they would fit a model of islend colonnization patterns as given in MacArthur end Wilson (1967) or more appropriately that of montene islands by Brown (1971). These mountains are isolated fran the source area by nnumerous intermentane valleys which are nnow not suitable for the "uontene" species (Thanamys bottae, Spermophilus variegatus , Peramyscus difficilis, P. pectoralis, end P. truei). These species are not "true" Montene species; instead, they are mostly found in grasslend to pinyon-juniper habitats. Brown (1978) discovered a similar species canposition for IVbntene desert islends of the Great Basin in the United States. All five of these species occur in the Sierra Astillero (the easternmost renge) , the first four occur in the monnntains around Cancepcion del Oro (Quadrat 6), end onnly _S_. variegatus end _P. pectoralis are known fram Cerro de Tayra (Quadrat 4) . Colonnization of these mountains would have to have occurred when climatic connditions were sufficiently mesic to support at least a pinyon-juniper woodland (similar to the habitats they occur in today). This situation could have occurred during the Wisconsin Glacial. Morafka (1977) has pointed out that the Saladan Subprovince (surrounding much of these mountain regions) was most likely entirely pinyon-juniper woodlend at dnat time. This would require only a 200 m downward displacement of the vegetationn. Wells (1978) has documented the downward movenent of pinyon-juniper woodlend by sane 800 m in the Big Bend region of the Rio Grende. These remarks are only tentative, more work is needed especially in the Sierra Astillero end Cerro de Tayra. Since it is 66 possible that the entire Saladen Desert may have been a pinyon- juniper woodlend (Morafka, 1977) , the possibility that the source area was to the west cennot be ruled out. However, the subspecific relationships of at least two species Peramyscus difficilis (see Hoffneister end De la Torre, 1961; end Diersing, 1976) end P. pectoralis (see Schmidly, 1972) end the distributionn of _T_. Page would argue for a relationship with the Sierra Madre Oriental. Hoffmeister (1951) tentatively assigned two specimens of P. _t_r_'u__ei fran Sierra Astillero to the subspecies P. E. gentilis of western affinity but noted that they may prove to be dis- tinct when more specimens became available. Western montene habitats were defined by Factor II end correspond well with Cluster II (Figures 10 end 16). Cluster II had only five quadrats associated with it while Factor II has eight (Quadrat 37 seemed to be intermediate between grassland and montene). Quadrat 31 end 47 both scored high on other factors also. This would indicate the utility of factor enalysis in allowing a samnple to be associated with more then one factor. Both montane and grasslend faunas occur in Quadrat 31, while Quadrat 47 has both montane and tropical femas (it also scored high on Factor V, a result I en unable to interpret). Factor IV (Figure 12) corresponds in part to Cluster IV (Figure 16) . However, Quadrat 54 is onnly loosely associated with Factor TV end Quadrat 72 does not score high on this factor (in fact, the score for Quadrat 72 is high negative, -1.22). In addition, Quadrat 69 is associated with this factor but does not cluster with Cluster IV. QJadrat 69 includes the valley of the Rio Juchipila end from this it seenns reasonable to associate it with the Cenyon Regions. In eny 67 event, the species association of Factor IV is clearly with the tropi- cal cenyon bottanns (Table 4). Factor VI represents an unusual fennnal association in Zacatecas (Table 6). It is the only factor that did not include en environmental variable. It would appear to represent a distinct cenyon association based primnarily upon the presence of Sciurus aureogaster end Pergyscus spicilegus. Both Quadrats 72 end 73 score high on this factor (Figure 14) end are also a subcluster of Cluster IV (Fign‘e l6) . Factor VII has no counterpart in the cluster enalysis. The dis- tribution of factor scores has no geographic continuity . However , there is a cannon denominator for the quadrats with high scores (Figure 15) . This cannon denominator is a relationship to riparien (or at least more mesic) situations. Signodon hispidus end Lignys irroratus cen be found in almost eny of the habitats (except high mon- tene) in Zacatecas. They are found in desert areas but are restricted in such cases to riparien situations (as along the Rio Aguenaval) or in islands of grasslend end mesquite in sane of the internal drainage basins (_e_.g., Quadrats l, 3, 22). Each of the quadrats that had high scores on Factor VII were associated with sane type of riparien habi- tat. Six species in the enalysis did not load high enough to be con- sidered as associated, except loosely, with any factor. The criterion of a .40 loading as the cut-off level may have been too stringent. If the level were lowered to .30 as done by Smith end Fisher (1970), only one species would have been in this category. However, this was en arbitrary decision on my part . It does represent sane intuitive prob- lems. For instence, Signodon fulviventer is usually connsidered a good 68 indicator of grasslend habitats (Baker, 1969 end Petersen, 1973) . Its low loading, whether or not .40 is considered a cut-off point, does not support this idea. Its known distribution in Zacatecas, lnowever, would support its association with grassland habitats (see Species Accounts) . Perognathus flavus, Reithrodontannys fulvescens end Neotama albigula are sufficiently well known in Zacatecas end each is widely distributed so that their low loadinngs on the factors cen easily represent independent patterns . Spermophilus mexicenus parvidens end Perognathus lineatus are known fran only two quadrats each. Their low loadings could be the result of insufficient data on distributionns or independent patterns . However , Sciurus aureogaster end cynagys mexicenus were reported frann only two quadrats. They were both definitely associated with Factors VI and III, respectively. Thus, it would appear that loading is not a function alone of the numbers of quadrats in which a species occurs . Faunal Areas It is now appropriate to synthesize the results of the two types of enalyses into a reasonable system of classificationn of distri- butionnal patterns. Although it was poinnted out inn precedinng sections that the patterns correspond to various vegetation types end other environmental pareneters, I follow the suggestions of Hagmeier end Stults (1964) that patterns based on single groups of enimals be referred to as faunal areas (more precisely in the present study to Rodent Femal Areas). For simplicity, the units here defined will be called Fennel Areas. I have projected the distribution of Faunal Areas in Zacatecas to those quadrats that either were not sampled or did not meet the species 69 number requirenent (i._e_. , a minimum of 10 species for inclusion in the enalysis). For sane quadrats, this presented no problem because they either had indicator species (those that loaded high on the factor analysis) or they were Surrounded by other quadrats that were defi- nitely associated with a faunal area. In other quadrats that were not sempled, the affinity was determined by a knowledge of the vegetation, topography, end climate. In essence then, this syntlnesis represents a best estimate or hypothesis of our current knowledge of the distri- butional patterns of rodents inn Zacatecas. The distributionn of four Fennel Areas is given in Figure 17. The classification scheme represents en objective arrangement of the distributional patternns of rodents. As such, the classification cen be considered a "natural" grouping of species. Natural is used in the content as defined by Goodall (1973) . It represents areas that are relatively hanogeneous internally in their camposition end separated fram other areas by discontinuities. However, the discontuinty between areas is not always a sharp one (Factor V, the desert-grassland ecotone). Various studies have been able to define "natural" fennel areas using objective nnumerical technniques (Webb, 1950; Hagnneier and Stults, 1964; Ryen, 1966; Armstronng, 1972; Anderson, 1972; Findley and Caire, 1978; Morafka, 1977, 1978; and others). The "natnm'alness" of the present classification is also supported by the recognition of similar faunal areas in Chihuahua (Anderson, 1972, see discussian above). Findley and Caire (1978) were able to distinguish a distinct Chihuahuen Desert area based upon cluster enaly- sis of mennalien distributionn patterns . This would corresponnd with my Desert Fennel Area. Possibly the best support for my classification 70 OU Rodent Faunal Areas Figure 17. Distribution of Rodent Femal Areas recognized in Zacatecas: (1) Desert Fennel Area; (2) Grasslend Faunal Area; (3) Montane Femal Area; (4) Tropical Femal Area; (5) Desert-Grasslend Ecotone. 71 canes fram the independent recognition of a Saladen Subprovince of the Chihuhuen Desert (= Desert Faunal Area in the present study) end the Saladen Filter Barrier (= desert-grasslend ecotone in the present study) by Morafka (1977, 1978). His study was based upon the distri- bution patterns of the herpetofauna ; unfortunately, his primary concern was with the Chihuhuen Desert. Therefore, he did not attempt to define other faunal areas. In his analysis, lnowever, he did consider the areas adjacent to the Chihuahuen Desert, especially the Sierra Madres and the grasslands. Morafka (1977, 1978) mapped his herpeto- faunel units and these correspond closely to my mnap based on rodent distributions. Further work on other groups of enimals would provide tests of the hypothesis of four faunal areas in Zacatecas and/or adjacent states. Faunal areas (Figure 17) were named for the general habitat of the geographic area in which they occur. I believe that this general terminology for Faunal Areas in Zacatecas alleviates the confusion that mnight arise with cannparisons of other works. In the following descrip- tions of the Faunal Areas, indicator species were, for the most part, those that showed high correlations with the various axes of the factor analysis. Desert Faunal Area. -— As slnown in Figure 17, the Desert Faunal Area occurs in the nnorthern and eastern portion of Zacatecas. Its dis- tribution corresponds fairly close with that of the Desert Scrub vege- tation (Figure 4), low average annual precipitation (Figure 2), end elevations below 2,000 m (Figure 1). Included in the Desert Femal Area are the subunits defined in Factors III end VII, Cross Renges end 72 riparien habitats, respectively. Also, the ecotonnal area between grasslends end desert habitats (Factor V) is shown. Species indicative of the Desert Faunal Area, include: Pappggeamys castenops , Dipodarys nelsoni, Perognathus penicillatus, Peromyscus eremicus , and Neotama golchneni. Grasslend Faunal Area. -- This Faunal Area corresponds to the Grasslend vegetation end in part to the Crassicaulescent Scrub end Subtropical Scrub Vegetations (Figure 4). Under the vegetation types considered by Leopold (1950), the Grasslend Faunal Area would include almost the entire extent of Mesquite-grasslend in Zacatecas. Species indicative of the Grasslend Faunal Area , include: Dipodomys phillipsii, Peramyscus truei (by Factor I, see Results sec- tion end Species Accounts), Perognathus hispidus , Thamanys unnbrinus (but also associated with the Montene Faunal Area by Factor II), Iigngs irroratus (by Factor I). In additionn, Signodon fulviventer mnight best be considered associated with this Faunal Area (see discussion above). Montene Faunal Area. -- This Faunal Area occnm's as disjunct areas in western and southern portians of Zacatecas (Figure 17). This is so because of the dis junct nature of the Sierra Madre Occidental at its southern extent (see section on Physiography end Topography). The antene Faunal Area is associated entirely with the Montene Forests of the western part of Zacatecas (Figure 4), and usually restricted to elevations above 2 , 300 m (Figure 1). Species indicative of the Montane Fennel area , include: Eutamias bulleri, Sciurus nayaritensis, Peranyscus melenotis, P. b_oylii, P. difficilus, Signodonn leucotis, Neotanne mexicana, and Nelsonnia neotanodon. 73 Tropical Femal Area. -- In Zacatecas , the Tropical Faunal Area is restricted to the cenyon bottans end adjacent slopes of the river val- leys of the west and south (Fignre 17). It coincides with the Subtropical Scrub end Tropical Deciduous Forest vegetation (Figure 4), mostly at elevations below 2,000 m (Figure 1). Species indicative of the Tropical Faunal Area, include: Spernnophilus mexicenus mexicenus , Sciurus aureogaster, Peranyscus spicilegus, Siggodon mnascotensis, end Neotama palatina. Humen Impact The rapidly chenging landscape in Mexico due to increased culti- vation mentionned in the introductory paragraphs and the centuries—old results of overgrezing by cattle end goats (Johnston, 1978), cen drastically affect distribution patterns of mnenmals. A personal experience will emphasize this much better. In Octoer of 1973, on my first trip to Zacatecas, Bob Hennum end I made cenp end set out trap lines at a locality two miles east of Bafnon. This area was in the desert—grasslend ecotone daninated by nopal cactus, creosote bush, tree yuccas, shrub acacias, end a good undergrowth of grasses. Although we did not collect eny bemertail kangaroo rats (Dipodomys spectabilis) , we did observe their mounds end specimens in the IACM taken by Percy Clifton were from this approximate locality. The area at that time was being grazed by a few donnkeys, horses, end goats. We collected 10 species of rodents in two nights of trapping. In July of 1976, I returned to this sene locality but it had been converted to corn fields with no "natural" vegetation over meny hectares. Except along the slnoulder of the gravel road leading to Baf‘non, I observed no sign of 74 kengaroo rats. The conspicuous mounds of bennertail kangaroo rats had vanished campletely. Because of other priorities at that time, we decided not to collect in the area. It would have been interesting to canpare collectionns from the two different years. Although this illustration of the chenges in habitat that cen occur in a relatively short time is a single case, it is indicative of much of Zacatecas end in general Mexico (Wellhausen, 1976). Much of the grasslend end desert-grasslend ecotone in Zacatecas is being turned to agriculture (personal observation and R. H. Baker, personal canmInnication). Johnnston (1978) described the western edge of the Chihuahuen Desert end eastern edge of the grasslends to be in a state of flux due to centuries of lend abuse. In fact, Johnnston (1978) suggests that much of what is now the western portion of the Chihuahuan Desert in Zacatecas may have been grasslands in the past. The same is true for mnany of the fertile river valleys in tropical habitats. They are rapidly being converted to croplends. Especially, this is true of the valley of the Rio Juchipila end Rio Jérez. In addition to the cultivation of land, much of the forested regions of the west are now being exploited for their timber resources. Road building in Zacatecas , especially paving, could also affect the distributional patterns of snnnall mnenmals. The roads thennselves are built high above the surrounding terrain. Vegetation is cleared for same 30-50 m on either side of the road with a depression or gulley adjacent to the road. These gulleys tend to collect end hold rain water for longer periods than the surrounding land. The result is a chenge in vegetation with grasses end herbaceous plents more cannon then in the adjacent more xeric areas. The effects of this habitat 75 disturbence has not been studied. Presumably, species such as cotton rats (Signodon hispidus), spiny pocket mice (Liamys irroratus) , end pocket gophers (Thanamys umbrinus) could use such habitat avenues to invade the desert areas. These three species are already known to occur in desert riparien situations (see discussionn above end Species Accounts). The uses of roadways as dispersal routes for smnall mnemnnals has been documented by Getz eE a1. (1978) and for desert situations by Huey (1941). Sane rodents are used as food items by local inhabitants. These include most all species of squirrels, woodrats (Neotama spp.), cotton rats (Sign-Eon spp.) end the plateau mouse (Peranyscus melanophrys). Though these are taken with regularity in sane areas (personal obser- vations) , the mnain threat to rodent populations remains with habitat perturbatians. The ultimate effect of habitat modification by men on the distri- bution of rodents cennot at this time be objectively evaluated. In eny event, it is obvious that men is actively end potentially affecting the habitats of many mammnals in general and rodents in particular. What the outcame will be depends upon the actions taken in future years by the Mexicen people and authorities. It is hoped that this report will aid wise mnannagenent decisions by the Mexican people. In past years, most concern for Mexicen mannnnals has centered around the gene species (Leopold, 1959; Tinker, 1978). Recently, Baker (1978), Findley end Caire (1978), end Packard (1978) have discussed the status of various memnels in the Chihuahuen Desert Region. Their con- cern was not simply with the gene enimals but included non-game species, including rodents, as well. Findley end Caire (1978) end Baker (1978) 76 list species of rodents that occur in Zacatecas (mostly in the desert areas) which they consider to be rare, endangered, or threatened. These include: Mexicen prairie dog (ginanys mexicenus) , Mexican ground squirrel (Spernophilus mexicenus) , Nelson kangaroo rat (Dipodamxs nelsoni), and Pignny woodrat (Neotanna goldnmani). Other species that occur in Zacatecas which I believe should also be con- sidered threatened due to habitat restriction or lack of knowledge about their status, include: Buller chipmunk (Eutenias bulleri) , Nayarit tree squirrel (Sciurus nayaritensis) , Mexican gray squirrel (S. aureogaster) , bennertail kengaroo rat (Dipodomys spectabilis cratodgn.) , southern bennertail kengaroo rat (2. phillipsii) , and Bole“n'os wood rat (Neotama palatine). Gazetteer The following list of place-names are used in recording localities fran which rodents have been taken or observed in Zacatecas. Each place—name is followed by its approximnate location in degrees and mninutes of latitude and longitude. All place-names were located on and coordinates determined frann 20' x 15' topographic mnaps with the excep- tion of the locality called Sierra Madre. The geographic position of Sierra Madre was taken frcmn Goldman (1951) end is approximate. Place- nanmes are listed alphabetically followed by a parenthetical number identifying each on the mnap (Figure 18) . On the map, place-names are numbered north to south and west to east. Apizolaya. -- 24° 47' N, 102° 15' W (3) Apozol. -- 21° 27' N, 103° 08' W (70) Atolinga. —- 21° 47' N, 103° 28' W (65) 77 30' 25. 24' 23' 22° 30' 2I' .1 .2 .3 KILOMETERS .4 05 6 e7 ..8 .9 .10 .11 012 14 13 015 .16 .13 .17 22 .20 .19 23. .27 .24 26 .25 O. 034 035 '33 37 .38 36. O 040 .39 .41 42 "‘1. . .49 48. 51 52 054 550 53 56 60 .58 .57 . . .61 59 62 063 :4 65 66 67 .68 .69 .70 .71 72. 73 74 .75 l 4 L J l l L 1 l L l l04° I03' [02’ IOI’ Figure 18 . Place-names used to identify collecting localities in Numbers correspond to the locality in the Zacatecas . gazetteer . 78 Baffin. -- 23° 13' N, 102°, 27' W (39) Berriozabal. —— 22° 32' N, 102° 18' W (53) Calera. -- 22° 55' N, 102° 42' W (42) Camacho. -- 24° 26' N, 102° 28' W (11) Cafnitas. —- 23° 33' N, 102° 42' W (32) Caopas. -- 24° 46' N, 102° 15' W (4) Capriote. -- 23° 38' N, 102° 15' W (30) Cedros. -— 24° 38' N, 101° 43' W (5) Chalchihuites. -- 23° 28' N, 103° 56' W (34) Concepcit'xn del Oro. -- 24° 37' N, 101° 23' W (8) Cuenhtémoc. -- 22° 27' N, 102° 21' W (55) E1 Arenal. -- 23° 41' N, 103° 25' W (27) E1 Calabazal. -- 23° 45' N, 103° 50' W (23) El Rosario. -- 24° 26' N, 101° 39' W (12) Fresnillo. -- 23° 12' N, 102° 52' W (40) Gonzales Ortega. -- 23° 56' N, 103° 24' W (18) Guadalupe. -- 22° 45' N, 102° 31' W (46) Huenasco. -- 21° 50' N, 102° 55' W (64) Jalpa. -- 21° 37' N, 102° 58' W (68) Jérez. -- 22° 39' N, 102° 58' W (49) Jimenez del Téul. —- 23° 16' N, 103° 49' W (38) Juen Aldema. -- 24° 19' N, 103° 23' W (14) la Blanca. -- 22° 43' N, 102° 05' W (48) La Colorada. -- 23° 46' N, 102° 31' W (21) la Pendencia. -- 24° 55' N, 102° 29' W (2) Loreto. -- 22° 16' N, 101° 58' W (61) Lulfi (in Sen Luis Potosi). -- 24° 26' N, 100° 52' W (13) 79 Majama. -- 23° 48' N, 101° 41' W (19) Mazapil. -- 24° 38' N, 101° 26' W (6) Mesquituta. -- 21° 14' N, 103° 10' W (73) Milpillas de la Sierra. -- 23° 03' N, 103° 43' W (41) Mannax. -- 21° 55' N, 103° 19' W (63) Monte Escobedo. -- 22° 17' N, 103° 51' W (60) Monte Mariena. -- 23° 18' N, 103° 03' W (36) Morelos. -- 22° 52' N, 102° 38' W (43) Moyahua. -- 21° 16' N, 103° 09' W (72) Nieves. -- 23° 59' N, 103° 01' W (17) Nochistlen. -- ”1° 21' N, 102° 48' W (71) Noria de Angeles. - 22° 22' N, 101° 46' W (57) Ojo Caliente. -- 22° 34' N, 102° 16' W (52) Pinos. -- 22° 18' N, 101° 32' W (59) Plateado. -- 21° 56' N, 103° 08' W (62) Rencho Grende. -- 23° 27' N, 102° 58' W (35) Rio Grande. -- 23° 47' N, 103° 02' W (20) Sabena Grende. -- 24° 28' N, 101° 42' W (10) Sain Alto. -- 23° 34' N, 103° 18' W (31) Saldefia. -- 22° 24' N, 101° 22' W (56) Sen Andres. -- 23° 42' N, 101° 55' W (26) Sen Felipe de Nuevo Mercurio. -- 24° 13' N, 102° 09' W (15) Sen Juen Capistreno. -- 22° 37' N, 104° 10' W (51) Sen Juan de los Charcos. -- 25° 05' N, 102° 33' W (l) Sen Rafael. -- 24° 36' N, 102° 03' W (9) Sen Tiburcio. -- 24° 08' N, 101° 29' W (16) Senta Efigenia. -- 23° 43' N, 101° 52' W (25) 80 Santa Rosa. -- 21° 12' N, 103° 09' W (74) Sarteneja. —- 23° 38' N, 101° 42' W (29) Sierra Madre. -- 22° 37' N, 104° 18' W (50) Sierra Vieja. -- 23° 29' N, 102° 04' W (33) Sanbrerete. -- 23° 38' N, 103° 38' W (28) Tabasco. -- 21° 46' N, 102° 57' W (67) Tecolotes. -- 24° 37' N, 101° 59' W (7) Tepetongo. -- 22° 27' N, 103° 07' W (54) Tetillas. -- 23° 43' N, 102° 58' W (24) Téul de Gonzales Ortega. -- 21° 28' N, 103° 29' W (69) Tlatenengo. -- 21° 47' N, 103° 18' W (66) Trancoso. -- 22° 44' N, 102° 23' W (47) Valparaiso. -- 22° 47' N, 103° 33' W (44) Villa de Cos. -- 23° 18' N, 102° 22' W (37) Villa Insurgentes. -- 23° 45' N, 103° 50' W (22) Villenueva. -- 22° 19' N, 102° 51' W (58) Yahualica (in Jalisco). -- 21° 12' N, 102° 56' W (75) Zacatecas. -- 22° 46' N, 102° 36' W (45) SPECIES ACCDUNI‘S The arrangement of genera in this report follows that of Hall end Kelson (1959) . Species are arrenged alphabetically under their respective genera. When more than one subspecies is considered to be present in Zacatecas these are listed alphabetically under the species. Specimens exemined are listed north to south (west to east at the sanne latitude). Localities are given as recorded on the speci- men label (but see section on Materials end Methods). Other records of occurrence for species are noted. Species records for quadrats in adjacent states end their references are included. Collections frann which specimens have been examined, other than The Museum, Michigen State University, are listed with the following abbreviations: AMNH, American Museum of Natural History; CAS, California Academy of Sciences; KU, Musemn of Natural History, University of Kensas; LAO/I, los Angeles County Museum; LSU, Louisiana State University; MVZ, Museum of Vertebrate Zoology, Unniversity of California; MZ, Museum of Zoology, University of Michigen; 0U, Stoval MuseLmn of Science end History, University of Oklahana; RHG, private collection of Robert G. Hennum; T‘Cm, Texas Cooperative Wildlife Collections, Texas A & M University; TI'U, Texas Tech University; UI, Museum of Natural History, University of Illinois; US, United States National Museum and Biological Sm'vey Collections. Specimens of 81 82 rodents fran Zacatecas housed in the Instituto de Biologia, Universidad Nacional Autonama de Mexico end under the charge of Bernardo Villa- Remirez, have not been exemined. For the most part, they are specimens returned to the Mexicen authorities under the terms of the collecting permnits issued to universities end museums in the United States (Cornelio Sénchez, personal canmnunicetion) . These specimens, then, duplicate the records given in the accounts below. EUTAMIAS BULLERI BULLERI (J. A. ALIEN) Buller's Chipmunk Distribution. -- Pine-oak forests of the Sierra Mach'e Occidental in western Zacatecas. Remarks. -- Buller's chipnmmk seenns to be most common in those parts of the pine-oak forests where rock outcrops are abundent . Preg- nent end lactating individuals were taken in July end August. This is the only species of chipmunk that reaches as far south in Mexico as Zacatecas. Its southernmost distribution is approximately 22° 30' N latitude. Specimens Exenined. -- Total 47, fram: 10 mi SW Chalchihuites, 7200 ft, 5 (IACM); 25 km WSW Milpillas de la Sierra, 2580 m, 3; 9 mi NW Valparaiso, 8350 ft, 7; Valparaiso IVbuntains, 8700 ft, 17 (US); 8.7 mi W Valparaiso, 8400 ft, 2 (0U); 7.2 mi W Valparaiso, 7900 ft, 2 (0U); Sierra Madre, 9 (US); 10 kmn W San Juan Capistrano, 2900 m, 2. Other records: 8 mi 8 Chalchihuites (Jones end Webster, 1977); 8 mi W Milpillas (de la Sierra), (GAS); in Jalisco, Quadrat 48 (Genoways and Jones, 1973). 83 SPERMOPHILUS MEXICANUS (ERXLEBEN) Mexican Grounnd Squirrel Distributionn. -- Disjunct populations; one in the extrene north in desert scrub; one in the southwest in grasslends. Rennarks. -- The few records of this species in Zacatecas do not allow for eny generalizations to be made concerning its ecological distribution. The two females taken in the desert scrub of northern Zacatecas were lactating in July. One female taken in the southwestern grasslend on 3 July contained six embryos. These records document the occurrence of both the northern and southern subspecies of _S_. mexicenus in Zacatecas for the first time. The localities given below narrow the hiatus between the two sub- species to about 200 km. Howell (1938) distinguished the two sub- species an the basis of size and color: A northern subspecies, §_. n_m. parvidens, snmall in size end light in color; end, a southern subspecies, S. n_m. mexicenus, large in size end dark in color. Two adult females here assigned to parvidens and one adult fennale assigned to mexicenus have external measu'ements, respectively, as follows: total length, 38, 36, 44; ear length, 11, ll, 14. Because of the distence between the two populationns and their relatively large size differences, I treated these as separate "species" inn the enalysis of distribution patterns. Specimens Examnined. -- Spermgphilus mexicenus mexicenus (Erxleben), total 2, fran: l6 kmn SSE ante Escobedo, 2010 m, 2. Other records: E Monte Escobedo, (sight); N Téul de Gonzales Ortega, (sight); in Jalisco, Quadrat 55 (genoways end Jones, 1973). 84 Spermgphilus mexicenus pervidens Mearns, total 2, fronm: 10 km ESE San Juen de los Charcos, 1500 m, l; 3 km SE Apizolaya, 1920 m, l. SPERD’DPHILUS SPILOSOMA BENNEI'I‘ Spotted Ground Squirrel Distribution. -- Deserts and grasslands east of the Sierra Madre Occidental in Zacatecas. Remarks. -- Spotted ground squirrels were taken in a variety of open habitats in Zacatecas. A female captured on 8 July contained six embryos. lactatirng individuals were obtained in July end August. Two subspecies of S. spilosama occur in Zacatecas (Howell, 1938). Specimens assigned to S. _S_. pallescens fram the arid northeastern part of the state are lighter in color than S. S. spilosama fram the western end southern parts of the state. AltIough Howell (1938) also records pallescens as being snnnaller in size then spilosama, I cen find no obvious size differences between these two subspecies. Specimens Exemined. -— Spermophilus spilosoma pallescens (A. H. Howell), total 40, fram: 3 km SE Apizolaya, 1920 m, 3; 7 km SE Caopas, 1940 m, 1; 11 mi E Concepcién del Oro, 5300 ft, 2 (LACM); 5 km SW Concepcién del Oro, 2400 m, l; 10 mi SW Concepci6n del Oro, 7600 ft, 2 (lACM); 13 km SE Concepcion del Oro, l; 35 km SSE Concepcién del Oro, 1980 m, 1; 13 mi SW Camnacho, 5900 ft, 2 (LACM); 2 km W Sen Felipe de Nuevo Mercurio, 1740 m, 2; 6 km W Sen Felipe de Nuevo Mercnm'io, 1790 m, l; 20 km NE Rio Grende, 1770 m, 4; 15 mi NE Sen Andrés, 6200 ft, 3 (IACM); 6.5 km S Ia Colorada, 1970 m, 2; 43 mi NE Villa de Cos, 7400 ft, 1; l3 knWNW Capriote, 2100 m, l; Cafiitas, 1 (US); 1 mi W Sierra Vieja, 6100 ft, 2 (0U); 12 km SW Villa de Cos, 1950 m, 1; 45 km NE 85 Morelos Jct, 1 (TI'U); Bafon, 6400 ft, 6 (LACM); 2 mi E Be’ion, 6400 ft, 1 (LACM), l (RGH). Other records: 10 kmESE San Juan de los Charcos, (sight); near Cedros, (sight); 3 km E San Tiburcio, (sight); 15 km NNW Nieves, (sight); in San Luis Potosi, Quadrat l3 (Dalquest, 1953). Sinernnophilus spilosome spilosama Bennett, total 56, fronm: 7 mi SW Sanbrerete, 6800 ft, 2 (LACM), 2 (RGH); 5.6mi ESE Sain Alto, 7700 ft, 1 (0U); Chalchihuites, 7500 ft, 1 (IACM); 2 km 8 Monte Mariana, 2180 m, 1; 4.5 mi E Fresnillo, 8 (MVZ); lOmi NWZacatecas, 1, 1 (KU); 5 mi NW Zacatecas, 7600 ft, 1 (K0); 3 mi SE Guadalupe, 1 (U1); 2 mi ESE Trancoso, 7000 ft, 4 (KU); 7 mi SE Trancoso, 1 (K11); 9 mi SE Zacatecas, 7900 ft, 1 (KU); Berriozabal, 10 (US); 3 mi N Cuauhtémoc, 6600 ft, 8 (IACM); 2 mi NNW Cuauhténoc, 6600 ft, 1 (0U); 5 mi S Ojo Caliente, 1 (U1); 8 mi N Villenueva, 6800 ft, 1 (KU); 7 mi S Pinos, 6800 ft, 10 (IACM) . Other records: 3 mi SW Sannbrerete (Jones and Webster, 1977); 10 mi SW Somnbrerete (CAS) ; 14 km N Fresnillo, (CAS); 40-41 mi W Fresnillo, (Jones end Webster, 1977); 17 mi SW Fresnillo, (CA8); 45 km SW Fresnillo, (sight); 13 km E Jérez, (sight); 10 mi S Pinos, (CA8); 45 km S Pinos, (sight). SPERMOPHILUS VARIEGATUS (ERXLEBEN) Rock Squirrel Distribution. -- Widely distributed in Zacatecas, except for the most arid desert areas. Remarks. -- Rock squirrels are abundent in the mountainous regions of the state usually in arroyos or alonng rock cliffs. In the grass- lends end tropical areas they are often associated with rock 86 outcroppings or rock fences. A fennale taken on 25 July contained four ennbryos. Another fennale captured on 10 July was lactating. Range mnaps in Howell (1938:137) end Hall and Kelson (1949:353) indicate the possibility of three subspecies occuring in Zacatecas. Specimens assigned to S. g. couchii fram northeastern Zacatecas have a dark brown dorsum with a more distinctly blackened head then speci- mens from southern Zacatecas assigned to S. y_. variegatus end which have a more gray dorsum. Specimens fran northwestern Zacatecas assigned to S. _v. gpestris differ fran variegatus in the senne way that couchii does. Fronn couchii, npestris has a lighter brown dor- sum. I was not able to distinguish differences in size between eny of the populations examined. A reevaluation of the subspecies of Spermophilus variegatus is much warranted. Specimens Exemined. -- Spermophilus variegatus couchii Baird, total 14, fran: 12 kmn W Sen Rafael, 2590 m, 3; 10 mi SW Concepcién del Oro, 7600 ft, 3 (LACM); 5 km SW Concepcion del Oro, 2400 m, 8. Other records: 40 km ESE Concepcién del Oro, 2320 m, (sight). Spermgphilus variegatus rupestris (J. A. Allen), total 7, from: 15 km NNW Nieves, 1910 m, l; 9 mi N Nieves, 6000 ft, 1 (LACM); 7 mi SW Sambrerete, 6800 ft, 1 (IACM); 14.5 mi WSW Sannbrerete, 7100 ft, 2 (0U); 14.2 mi N Jimenez del Téul, 7300 ft, 1 (0U); 32 km SE Valparaiso, 2040 m, 1. Other records: 3 mi SW Somnbrerete (Jones and Webster, 1977); 2 km 8 Monte Mariena, (sight); 53 km WSW Fresnillo, (sight); 32 km SW Fresnillo, (sight); Valparaiso, (sight); 13 km E Jérez,_ (sight); 8 kmn SW Jérez, (sight); Zacatecas, (sight). Spermophilus variegatus variegatus (Erxleben) , total 23, fran: Berriozébal, 1 (US); Monte Escobedo, 7300 ft, 3 (LACM); 10 km ENE 87 loreto, 7350 ft, 1 (DU); 7 mi S Pinos, 6800 ft, 1 (IACM); 30 anE Jalpa, 1740 m, 3; 10 mi W Jalpa, 6100 ft, 1 (IACM); 24 km ESE Jalpa, 2590 m, l; 4 mi S Jalpa, 4300 ft, 1 (LACM); 8 mi NW Nochistlén, 6600 ft, 4 (LACM); 6 km S Téul de Gonzales Ortega, 2010 m, l; 16 km SSW Jalpa, l; 10 mi NW Yahualica, 7100 ft, 1 (LACM); .5 mi ENE Mesquituta, 3450 ft, 2 (0U); Santa Rose, 4000 ft, 2 (LACM). Other records: 18 km N Sen Juan Capistrano, 1100 m, (sight); San Juan Capistreno, (sight); 3 mi N Cuauhténoc, (sight); Pinos, (sight); 20 km S Monte Escobedo, (sight); 8 mi S Moyahue, (Jones end Webster, 1977); in Jalisco, Quadrats 48 end 55 (Geoweys and Jones, 1973). CYNOMYS MEXICANUS MERRIAM Mexican Prairie Dog Distribution. —- Known fran only one intermontane basin in extreme northeastern Zacatecas. Remarks. -- Mexican prairie dogs are known fran a few localities in Coahuila, Nuevo Leon, and San Luis Potosi (Hall and Kelson, 1959; Baker, 1956; Dalquest, 1953; Pizzimenti, 1975). The specimens reported upon here represent the first records of this species in Zacatecas. The six specimens were taken in an "islend" of mesquite- grassland in the intermontene basin east of Concepcion del Oro. Extensive searching in this seme valley to the south did not reveal eny other colonies . Specimens Examnined. -- Total 6, from: 11 mi E Concepcion del Oro, 5300 ft, 6 CLAIM). Other records: in San Luis Potosi, Quadrat l3 (Dalquest, 1953). 88 SCIURUS AUREOGASTER SOCIALIS WAGNER Mexicen Gray Squirrel Distribution. -- Tropical to lower montane forests in extrene southern Zacatecas . Rennarks. Three specimens fran two localities represent the first records of the species fran Zacatecas. Sciurus aureogaster has en extensive distribution south of Zacatecas in tropical to lower mon— tene forests (Musser, 1968). Specimens Exemined. -- Total 3, fran: 10 mi NW Yahualica, 7100 ft, 1 (LACM); Senta Rose, 4000 ft, 2 (M04). SCIURUS NAYARITENSIS NAYARITENSIS J. A. AIIEN Nayarit Squirrel Distribution. -- Pine-oak forests of the Sierra Madre Occidental in Zacatecas. Remarks. —- This tree squirrel seems to be most abundent in forests daninated by oaks. Two females taken on 24 and 30 July were lactating. The Nayarit squirrel was taken in one locality in association with S. aureogaster. It would be interesting to study the ecology of these two species in areas where they occnm together. Musser (1968) also records areas where the two species occur in sympatry . Specimens Exemined. -- Total 49, fram: Sierra Valparaiso, 8200 ft, (type locality), 4 (AMNH); Valparaiso Mountains, 8700 ft, 19 (US); 9 mi NW Valparaiso, 8350 ft, 2; 9.5 mi WSW Valparaiso, 8650 ft, 2; Sierra Madre, 1 (US); 36 kmn SSE Valparaiso, 2330 m, 1; Monte Escobedo, 7300 ft, 5 (TAO/I); Plateado, 12 (US); 9 mi WNW Jalpa, 8250 ft, 2; 10 mi 89 NW Yahualica, 7100 ft, 1 (LACM). Other records: 8 mi S Chalchihuites (Jones end Webster, 1977); 41 mi W Fresnillo (Jones end Webster, 1977) ; 17 end 18 mi W Milpillas (de la Sierra), (GAS); 25 kmn WSW Milpillas de la Sierra, (sight); 4-5 mi W Monte Escobedo, (CAS); in Aguascaliente, Quadrat 57 (Lee and Hoffmeister, 1963). PAPPOGECMYS CASl‘ANOPS (BAIRD) Yellow-Faced Pocket Gopher Distributionn. -- Arid to semi-arid areas east of the grasslands in Zacatecas. Remarks. -- This species is most abundent in the interior desert basins of eastern Zacatecas where soils are silty or sandy and deep. Mounds of this species were nunnerous at Rencho San Marcos 7 km W of San Felipe de Nuevo Mercurio. They occurred over en area of several hectares in extent. Two females taken on 5 and 21 August contained one end two embryos, respectively. Russel (1968) recognized four subspecies in Zacatecas based on size differences and geographic locality. I follow his arrengenent of subspecies. Specimens Exemined. -- Pappogeanys castanops goldmani (Merrien) , total 15, fran: 4 kmN Nieves, 1980 m, 3; 6 km SE Tetillas, 2040 m, 1; Ce‘iitas, 5 (US); lmni S Ceiitas, 2 (TI'U); llmni SWCafiitas, 4 (TN). PappogeangLs castenops rubellus (Nelson and Goldman) , total 23, from: Villa de Cos, 6700 ft, 8 (101); 1 mi SW Villa de Cos, 4 (TTU); 20 mi NE Morelos jet, 3 (TI'U); 45 km NE Morelos jct, 8 (TI‘U). Pappogeanys castenops subnnubilus (Nelsonn end Goldnman) , total 30, from: 16.4 mi ENE Concepcién del Oro, 2 (OH); 10 mi E Concepcion del 9O Oro, 5390 ft, 1 (IACM); 11 mi E Concepcion del Oro, 5300 ft, 4 W), 4 (RGH); 15 mi S Concepcion del Oro, 6900 ft, 3 (KU); 35 km SSE Concepcién del Oro, 1980 m, 2; 3 mi N 1.1116, 13 (MVZ); 1 mi N 1.1116, 1830 m, l. Pappogeanys castanops surculus Russell, total 39, fran: 10 kmn ESE Sen Juan de los Charcos, 1500 m, 2; 3 km SE Apizolaya, 1920 m, 1; 7 km SE Caopas, 1940 m, 1; 1 mi S Cedros, 6050 ft, 1; Concepcion del 0rd, 7680 ft, 8 (K11); 3.5 mi E Mazapil, 1 (TN); 2 km SE Sabena Grande, 1945 m, 1; 13 mi SW Cemacho, 5800 ft, 8 (LAQ’I), l (RGH); 7 km SW Sen Felipe de Nuevo Mercurio, 1790 m, 2; 3 km E Sen Tiburcio, 1880 m, l; 25 knn SW Sen Tiburcio, 2030 m, 2; 5 kmn S Ia Colorada, 1960 m, l; 8 mi S Majanna, 7700 ft, 9 (ICU). Other records: 22 mi SW Concepcién del Oro, (GAS). THCMJMYS BOTTAE ANALOGUS GOLDMAN Botta Pocket Gopher Distribution. -- Desert mountain renges of northeastern Zacatecas. Remarks. -- Fresh mounds of Botta pocket gophers were numerous in soils covered with pinyon pine forests in the mountains west of Concepcién del Oro during the sumner of 1976. They were taken in rocky but deep soils of en arroyo. 01d mounds were found in pinyon-oak forests in the Sierra Astillero (SE of Concepcion del Oro) in the sum- mer of 1978. These specimens are clearly assignable to the subspecies _T_. S. enalogn_1_s based upon the characters given by Baker (1953). Specimens Exennined. -- Total 14, fran: 5 km E Mazapil, 2270 m, 1; 3.5 mi E Mazapil, 2 (TTU); 10 mi SW Concepcion del Oro, 7600 ft, 4 (LACM); 5 km SW Concepcion del Oro, 2400 m, 5; 40 km ESE Cancepcién del Oro, 2320 m, 2. Canparative material fran Coahuila, total 15, frann: 91 10mi S, 8miWGeneral Cepeda, 7000 ft, 3 (RH); 10mi S, 7 mi W General Cepeda, 8200 ft, 2 (KU); 11 mi S, 4 mi W General Cepeda, 6700 ft, 9 (1(0); 11 mi S, 6 mi W General Cepeda, l (KU). THOMI/IYS UMBRINUS (RICHARDSON) Southern Pocket Gopher Distribution. -- Widely distributed, except for the most arid parts of Zacatecas. Remarks. -— This species is usually locally abundent but popula- tions are spotty in their distributions. They seem rare in deserts of northeastern Zacatecas being recorded from only two localities. The Pocket gophers becane more abundant in the western parts of the state. In the deep soils of the montene regions of the west, they appear to be most abundant. Two females taken on 27 end 30 July contained five end one embryos, respectively. Two females taken on 1 and 6 August were lactating. The spotty distribution, dependence on particular soil types , and fossoriel habits of pocket gophers has produced considerable geographic variation in this species (Hall and Kelson, 1959; Baker end Greer, 1962; Anderson, 1966, 1972). Four subspecies from a total of five localities were recorded to occur in Zacatecas (Hall end Kelson, 1959) . Recently, Berry end Baker (1971) and Jonnes end Webster (1977) reported upon additional specimens fran Zacatecas. Additional material fromn many localities is now available. Because of smell sample sizes frann most localities, high degree of sexual dimorphism, end high degree of individual variation (Hoffmeister, 1969 and others) assignment of individuals to subspecies is difficult. The original descriptions of 92 the known subspecies fram Zacatecas included but a few individuals; thus, the extent of individual variation was not kown. I have defined adults as specimens with the basioccipital-basisphenoid suture end supraoccipital-exoccipital suture closed (Patton, 1973). A nnumber of the type specimens that I have exemined did not meet these criteria. In the canparisons given below I have considered only females, for the most part, since they exhibit less individual variation than males (Hoffmeister, 1969 and unpubl. data) and they are more numerous in col- lections. In a few instances, I have used males in canparisons when they are the onnly representative adults fran a locality. In most com- parisons only meen values of basilar length of sknnll (BL) are given. Five other cranial measurerents were recorded which, in most cases , substentiate the trends seen in BL. Exterrnal measurements proved to be too variable within populations to be of any value. Notes on general color within populations are also noted. Subspecies fran contiguous states were examined for comparative purposes. Specimens fram two localities in the deserts of northeastern Zacatecas were easily assigned to 'P. g. goldnmani. These are the lightest colored end smnallest (BL = 29.6) of all populations in Zacatecas. Canparative mnaterial fram Coalnuila end the original descrip- tion (Merriem, 1901) indicate goldmnani to be the smallest end lightest colored subspecies on the Mexican Plateau. Specimens assigned to _T_‘. _u_. dura_r1gae frann several localities in grassland end intermontane valleys of northwestern Zacatecas are small (BL = 30.6) although not so small as goldmeni. These are also darker colored then goldnnani. In color end size they correspond to canpare- tive nnaterial fram near the type locality and with a topotype fran 93 Durango. In the original description Nelson end Goldman (1934) remarked upon the simnilarity of dnn'ane to goldmani except in color. In the mountains of west-central Zacatecas, specimens are assigned to I. g. crassidens with the type locality in the Valparaiso mountains. In color, they are only slightly darker then durgggae but with con- siderable variation. They are larger than durggae (BL = 31.5 and 31.2 fram populationns south of Clelchihuites end Sierra Valparaiso, respec- tively). Specimens fran Sierra Madre were assigned to '_I‘_. sheldoni in the original description by Bailey (1915) . later, Nelson end Goldmen (1934) deronstrated that this was only a subspecies of 'P. umbrinus. It is a large subspecies (BL = 32.00) and somewhat darker colored than crassidens. In addition, specimens fram Monte Escobedo agree in size (BL = 31.97) end color with specimens fram Sierra Madre end canparative mnaterial frann the type locality in Nayarit. The Monte Escobedo speci- mens seem best referred to this subspecies, altl'ough a direct route of interbreeding between these pOpulations would not seem likely. The Monte Escobedo populationns may represent en unnened subspecies; low- ever, until more specimens becane available for study, I do not believe it wise to proliferate the already voluminous nanenclature of pocket gophers. Specimens fram near Plateado (in the Sierra Moroni) were described by Nelson and Goldman (1934:112) as P. g. en_i_.xn_ns_ based upon four speci- mens. later, Hall and Kelson (1959) added a second locality in Jalisco. Nelson end Goldman (1934:112) remarked that the skull was very similar to that of _T_. g. zacatecae which they also described, dif- fering only in ". . . color darker end richer." The type specimen of 94 3_‘___an1xus_ (US #90834) was a subadult mnale by my aging criterion. TWo other topotypes had broken skulls end I did not see tte fourth speci- men. For camparative purposes, I used specimens fran the Sierra Moroni northwest of Jalpa. These specimens do not differ appreciably in color fran specimens of zacatecae fran Berriozabal, the type locality. In size, pocket gophers fran the Sierra Moroni do not dif- fer fram topotypes or near topotypes (Table 9). Because of this it scene best to refer all specimens frann the Sierra Pbroni to '_I‘_. _u. zacatecae (Nelson and Goldman, 1934:112) since this neme takes pre- cedence by page position over e_n_ig_m_.ns_. In addition, specimens from near Nochistlan (Sierra Nochistlan) end Sierra Fria (in Aguascalientes) are also assigned to zacatecae. Specimens fran south of Pinos agree in size (BL = 30.10) with zacatecae end are so assigned. Specimens fronm north of Ojo Caliente are larger (BIL. = 31.50) but do not differ in color. These are assigned to zacatecae for convenience until more specimens became available. Dalquest (1951:361) described P. g. nemmeni fram Palma in San Luis Potosi which is near the Zacatecas bor- der. I have exemined his series of topotypes which were all subadult by my criteria. Since Dalquest (1951) remarked upon the similarity in color between newmnani and zacatecae, the smell size he reported for newneni may be due to his using a different aging criteria. In eny event , none of the material fram Zacatecas is referable to nemeni. A single adult mnale frann 25 km ESE of Pinos is, in color, similar to specimens of P. g. arriagensis (Dalquest, 1951:361) that I have examined. Dalquest (1951) considered this subspecies to be one of the largest on the Central Plateau end in this I agree (BL. = 32.20 for topotypes end near topotypes). In size, the single male fran Zacatecas C(MPARATIVE MEASUREMENTS OF FEMAIE IBM? BERINUS CATECAS ’ FRQ’ITHREEIDCALITIESINZA 95 TABLE9 Berriozabal Cuauhtenoc NW of Jalpa (ne2) (n=8) (n=6) Basilar 30.05 30.45 30.42 Length (30.0-30.1) (28.8-31.1) (29.9-30.2) Nasal 11.35 11.417 11.67 Length. (11.3-11.4) (11.1-11.9) (11.0-12.4) Zygomnatic 22 . 30 22 . 84 22 . 78 Breadth, (22.1-22.5) (22.0-23.7) (22.2-23.6) Mastoid 18.40 18.29 17.97 Breadth (18.0-18.8) (17.9-19.2) (17.6-18.4) Rostral 13.15 13.547 13.55 Length (13.0-13.3) (13.1-14.0) (12.8-14.4) Rostral 6. 45 6. 70 6. 93 'Width (6.4-6.5) (6.3-6.9) (6.5-7.5) Maxillary 7.25 8.10 7.67 ToothroW' (7.1-7.4) (7.6-8.4) (7.2-7.8) "Meens with renge in parentheses are given in millimeters. 2Superscript indicates sample size different fram that associated with the locality. 96 does not differ frcm males taken 7 mi S of Pinos (here referred to zacatecae). But then, males from the type locality of arriagensis do not differ to any appreciable extent from males of zacatecae. Since this specimen was taken only a few kilometers from the type locality and it agrees in color with arriagensis, I have assigned it to that subspecies. 1m subadult specimens taken in the tropical valley of the Rio Juchipila north of Moyahua are similar in color and the coarse texture of their pelage to l. g. musculus (Nelson and Goldman, 1934:119). In addition, four specimens fran 9 mi W of Yahualica, Jalisco can also be referred to this subspecies. Nelson and Goldman (1934) and Baker and Greer (1962) considered musculus to be restricted to the deep tropical canyons of the western slopes of the Sierra Madre Occidental . I have assigned these specimens to musculus but note that mnre extensive material , when available, may display characters which will prove them to be different. Specimens Examined. -- Thanomys umbrinus arriagensis Dalquest, total 1, from: 25 km ESE Pinos, 2425 m, 1. Comparative material, total 12, from: 1 km S Arriaga, (type locality) San Luis Potosi, 7 (LSU), 1 (US); 4 mi E Villa Arriaga, San Luis Potosi, 4 (MVZ). Thoumys umbrinus crassidens Nelson and Goldman, total 50, from: 10 mi SW Chalchihuites, 7200 ft, 15 (1AM); 8 mi S Chalchi‘rmites, 8600 ft, 5 (CAS); 8 mi WMilpillas (de la Sierra), 8300 ft, 2 (CAS); 25 km WSW Milpillas de la Sierra, 2580 m, 5; 9 mi NW Valparaiso, 8350 ft, 13; Valparaiso Momtains, (type locality), 10 (US). Thammys umbrinus durangae Nelson and Goldman, total 50, from: 5 km S Gonzales Ortega, 2450 m, 1; 18 km S Gonzales Ortega, 2150 m, 3; 97 3 mi. SW Sombrerete, 5 (GAS); 5 mi SW Sambrerete, 6800 ft, 1 (IACM), 2 (RGH); 7 mi SW Sanbrerete, 6900 ft, 2 (1ACM), 2 (RGH); 14.5 kaSW Sombrerete, 7100 ft, 3 (0U); 10 km S, 2 ka Sambrerete, 17 (MVZ); 14.2 mi N Jimenez de Téul, 7300 ft, 2 (0U); 40 mi W Fresnillo, 7700 ft, 12 (GAS). Comparative material from Durango, total 8, from: Nombre de Dios, 1; 6 mi NW La Pila, 6100 ft, 2; 4 mi SW La Pila, l; 5 mi 8 Dmrango, 6200 ft, 1; 4 mi E, 7 mi S Mango, 6200 ft, 1; Durango, (type locality), 2 (US). Thomamys umbrinus goldmani Merriam, total 27, from: 22 mi S Concepcién del Oro, 1 (GAS); 15 mi NE San Andres, 6200 ft, 26 (IAGM). Comparative material, total 7, from: Mapimi, Durango (type locality), 2 (US); 2 mi S E1 Palmito, Durango, 1; 3 mi NE Sierra ijada, Coal'mila, 4100 ft, 1 (K0); Sierra Mojada, 4150 ft, Coahuila, 1 (K0); 3 mi SE Torreon, 3800 ft, Coahuila, 2 (RU). Thomogrys umbrinus musculus Nelson and Goldman, total 2, from: 2.5 mi N. Moyahua, 4400 ft, 2 (GAS). Comparative material: 9 mi W Yahualica, Jalisco, 4 (KU). Thomays umbrinus sheldoni Bailey, total 42, from: Sierra Madre, 5 (US); 36 km SSE Valparaiso, 2330 m, 3; 3 mi NW Monte Escobedo, l3 (KU); Monte Escobedo, 7300 ft, 21 (IAGM) . Comparative material, total 11, fran: Santa Teresa, 6800 ft, (type locality), Nayarit, 11 (US). Thomomys umbrinus zacatecae Nelson and Goldman, total 145, from: 10 mi SE Fresnillo, 1 (TIU); 5 mi NW Zacatecas, 7600 ft, 1 (K11); 9 mi W Zacatecas, l (GAS); 4 mi W Trancoso, 2 (TN); .5 mi NW Trancoso, 1 (U1); 2 mi 8, 5 mi E Zacatecas, 7700 ft, 1; 3 km N Ojo Caliente, 25 (MVZ); Berriozabal (type locality), 10 (US); 3 mi N Cuauhténoc, 6600 ft, 21 (LACM); .5 mi SE Cuauhténoc, 6600 ft, 5 (DU); 2 km N Noria de 98 Angeles, 2200 m, l; 6 mi NNW Pinos, 7900 ft, 1; 10 km ENE Loreto, 7350 ft, 2 (DU); 7 mi S Pinos, 6800 ft, 9 (IACM); Plateado, 3 (US); 11 mi NW Jalpa, 8000 ft, 15 (KU); 9 mi WNW Jalpa, 8250 ft, 3; 25 km ESE Jalpa, 2590 m, 16; 8 km NW Téul de Gonzales Ortega, 2200 m, 1; 8 mi NW Nochistlén, 6600 ft, 26 (LACM). Comparative material, total 2, from: 3 mi N Cerro de Jaguey, 8200 ft, 2 (MVZ). PEROGNATHUS FLAVUS BAIRD Silky Pocket Mouse Distribution. -- Widely distributed, except for forested regions , throughout Zacatecas. Ranarks. -- Silky pocket mice scanned to be most abundant in the grasslands of western Zacatecas. Burrows of these mice, found in sum- mer in a plowed field, were about 5 to 10 cm deep with a greatest length of about 100 cm. No seed caches were present in any of five burrows excavated. Two females taken on 22 July and 21 August con- tained three embryos each . Baker (1954) , in reviewing the subspecies of Perognathus flavus in Mexico, reported one subspecies from Zacatecas and indicated that two others may also be present. These three subspecies were separated on the basis of size and color . Specimens fran the extreme north are assigned to 1:. f. pallescens based upon their pale color and small size. Specimens assigned to 1:. _f_. medius occur throughout the major portion of the state. They are darker and larger than pallescens. In localities east of Concepcion del Oro specimens appear to be inter- mediate between pallescens and medius. Since the Sierra Astillero and narrow intermontane valleys may act as a partial barrier to 99 interbreeding with populations to the north, these intermediate speci- mens are assigned to me_d_i_ug. Specimens fran the extreme southern part of Zacatecas are assigned to E. _f. parviceps on the basis of small size and light color. Specimens Examined. -- Perognathus flavus medius Baker, total 95, from: 6 km E Mazapil, 2645 m, l; 12 km ENE Concepcién del Oro, 1850 m, 2; 35 km SSE Concepcion del Oro, 1980 m, 6; 3 km NW San Felipe de Nuevo Mercurio, 1770 m, 4; 2 km W San Felipe de Nuevo Mercurio, 1740 m, 2; 1 mi S San Tiburcio, 7000 ft, 2 (KU); 15 km NNW Nieves, 1910 m, 7; 4 km N Nieves, 1980 m, 2; 23 km NE Rio Grande, 1800 m, 5; 12 mi NW Rio Grande, 6800 ft, 1; 18 km S Gonzales Ortega, 2450 m, 1; 5 km S La Colorada, 1960 m, 1; 6.5 km S La Colorada, 1970 m, 1; 13 km WNW Capriote, 2100 m, l; 5 mi SW Somnbrerete, 6900 ft, 1 (LACM), 1 (RG1); 5 km NE Chalchihuites, 2360 m, 1; 3 km SE Rancho Grande, 2190 m, l; Bafion, 6400 ft, 2 (LAGM); 4.5 mi E Fresnillo, 5 (MVZ); 10 km SE Fresnillo, 2250 m, 1; 53 km SW Fresnillo, 2250 m, 5; Valparaiso, 6200 ft, 1 (US); 3 mi SE Guadalupe, 1 (U1); 2 mi S, 5 mi E Zacatecas, 7700 ft, 1; 5 mi SW Zacatecas, l (MVZ); 15 km SWValparaiso, 2250 m, 1; 6 mi W Jérez, 6700 ft, 1; 6 mi E Jérez, 7000 ft, 2 (KU); 8 km SW Jérez, 2030 m, 18; Berriozébal, 1 (US); 5 mi 8 Ojo Caliente, 7 (UI); 2 km N Noria de Angeles, 2200 m, 5; 10 km S Tepetongo, 1950 m, l; 7 mi S Pinos, 6800 ft, 2 (LACM); 25 km ESE Pinos, 2425 m, l; 45 km S Pinos, 2350 m, 1. Other records: 40 mi. W Fresnillo (Jones and Webster, 1977); 10 mi 8 Pinos, (GAS). Perognathus flavus pallescens Baker, total 4, from: 10 km ESE San Juan de los Charcos, 1500 m, 2; 3 km SE Apizolaya, 1920 m, l; 7 km SE Caopas, 1940 m, l. 100 Perognathus flavus parviceps Baker, total 5, from: 16 km SSE Monte Escobedo, 2010 m, 2; 2.5 mi S Momax, 5800 ft, 1; 25 krm ESE Jalpa, 2590 m, 1; 8 mi NW Nochistlén, 6600 ft, 1 (IAQVI). Other records: in Jalisco, Quadrat 73 (Baker, 1954). PEROGNATHUS HISPIDUS ZAGATEGAE OSGOOD Hispid Pocket Mouse Distribution. -- Lower montane, grassland and tropical regions of western Zacatecas. Remarks. -- Hispid pocket mice seem to be most abmndant in open grassland habitats or in riparian areas. An adult female taken on 24 July was lactating. Specimens Examined. -- Total 36, from: 5.5 mi NW Juan Aldama, 6200 ft, 1 (CU); 5 km S Gonzales Ortega, 2150 m, 1; 23 km NE Rio Grande, 1800 m, 1; 10 km SE Fresnillo, 2250 m, 1; 11 mi NE Valparaiso, 7100 ft, 2; Valparaiso, 6200 ft, (type locality), 9 (US); 2 mi S, 5 mi E Zacatecas, 7700 ft, 1; 3 mi SE Guadalee, 1 (UI); 8 mi SE Zacatecas, 7225 ft, 6 (KU); 13 km E Jérez, 2200 m, 1; 8 km SW Jérez, 2030 m, 3; 2 km N Noria de Angeles, 2200 m, 1; 25 km ESE Pinos, 2425 m, 1; 45 km S Pinos, 2350 m, 3; 30 km NE Jalpa, 1740 m, 3; 6 mi SW Jalpa, 4900 ft, 1 (LACM). Other records: 10 mi NW Somnbrerete (Jones and Webster, 1977). PEROGNATHUS LINEATUS DAlQUEST Lined Pocket Mouse Distribution. -- Known from only one locality in southeastern Zacatecas . 101 Remarks. -- Very little is known about this species of pocket mouse. Some natural history observations were reported by Dalquest (1953). In the original description, Dalquest (1951) commented upon the similarity between 1:. lineatus end 1:. nelsoni. Apparently, the major distinction is that E. lineatus lacks the distinct stiff pelage spines that are present on R. nelsoni. Compared to g. penicillatus, which also lacks spines, 2. lineatus has a longer broader skull. In a numer- ical taxonomic analysis in which the species of the subgenus Chaetodipus were considered, 1:. lineatus was found to be distinct from either nelsoni or penicillatus (Caire, 1976) . I compared the two specimens from Zacatecas with topotypes fram Sen Luis Potosi end believe they are best referred to the species B. lineatus. Specimens Exemined. -- Total 2, fran: 1 mi NE Noria de Angeles, 2 (GAS). Comparative material: 1 km S Arriaga, (type locality), Sen Luis Potosi, 2 (LSU) , these specimens also represent the record from Quadrat 67 used in the analysis of distribution patterns. PERCG‘IATHUS NELSONI NELSONI OSGG)D Nelson' 3 Pocket Mouse Distribution. -- Widely distributed in open lands throughout the state. Remarks. -— This species is the most widely distributed end abun- dant pocket mouse in Zacatecas. Apparently, it is absent only from montane forests in western Zacatecas. Ten fenales taken between 5 July end 8 August contained an average of 2.9 (range 1 to 4) embryos. Six females captured between 13 July and 18 August were lactating. Of 102 32 adult females obtained in J1me, none showed eny evidence of breed- ing activity. Prior to the present study, there were relatively few records of _Ij. nelsoni reported from Zacatecas (Osgood, 1900; Jones and Webster, 1977). The large series now available considerably increases our know- lede of the distribution of this species . No appreciable amount of geographic variation was found to occur in Zacatecas. Specimens Exemined. -- Total 511, fran: 10 km ESE San Juan de los Charcos, 1500 m, 5; 6 km W Apizolaya, 1800 m, 39; 3 km SE Apizolaya, 1920 m, 11; 7 km SE Caopas, 1940 m, 3; 8 mi W Cechros, 5650 ft, 1 (LACM), 2 (RGH); 18.6 mi ENE Concepcion del Oro, 1 (0U); 12 km ENE Concepcion del Oro, 1850 m, 57; 4 km W San Rafael, 2140 m, 3; 13 km NE Concepcion del Oro, 1700 m, l; 40 km ESE Concepcion del Oro, 2320 m, 11; 13 km SW Concepcion del Oro, 1900 m, 1; 5 km SW Concepcion del Oro, 2400 m, 2; 5 km SE Concepcion del Oro, 1935 m, 4; 18 km SSW Concepcion del Oro, 2130 m, 5; 2 km SE Sabana Grende, 1945 m, 13; 7 km SE El Rosario, 2100 m, l; 35 km SSE Concepcion del Oro, 1980 m, 10; 13 km SW Camacho, 5800 ft, 2 (LACM); 3 km NW San Felipe de Nuevo Mercurio, 1770 m, 21; 2 km W Sen Felipe de Nuevo Mercurio, 1740 m, 2; 10 mi SE Juan Aldama, 2210 m, 4; 15 km WSW San Tiburcio, 1980 m, 1; 15 km NNW Nieves, 1910 m, 8; 9 mi NE Nieves, 2010 m, 2; 23 km NE Rio Grende, 1800 m, 2; 25 km SW San Tiburcio, 2030 m, 2; 15 mi NE San Andrés, 6200 ft, 2 W); 5 mi SE Rio Grande, 1940 m, 14; 5 km S La Golorada, 1960 m, 3; 6.5 km S La Colorada, 1970 m, 14; 6 km SE Tetillas, 2040 m, 4; 2 mi. 8 Santa Efigenia, 7400 ft, 2 (K11); 9.5 mi w Sambrerete, 1 (0U); 13 km WNW Capriote, 2100 m, 20; 3 km E El Arenal, 2450 m, 2; 9 km NW Sarteneja, 2200 m, 2; 10 mi N Rancho Grende, 6700 ft, 1; 15.5 mi WSW Sambrerete, 103 6400 ft, 1 (OU); Gaf'nitas, 1 (US); 5 anE Chalchihuites, 2360 m, 1; 1 mi W Sierra Vieja, 6100 ft, 1 (OU); 40 km NE Villa de Cos, 2000 m, l (OU); 3 km SE Rancho Grande, 2190 m, 2; 18 km NE Villa de Cos, 2040 m, 8; 3 mi E, 4.6 mi N Villa de Cos, 2 (OU); 2 km S Monte Mariana, 2180 m, 3; 2 mi SE Villa de Cos, 6200 ft, 2; 3 mi NW Fresnillo, 7760 ft, 2; Bafion, 6400 ft, 12 (LACM); 2miEBe’ion, 6400 ft, 1 (LAGM), 2 (RGH); 6 km N Fresnillo, 2250 m, 2; 1.8 mi N Fresnillo, 1 (MVZ); 2.5 km WNW Fresnillo, 7400 ft, 2 (OU); 18 km N San Juen Capistrano, 1100 m, 9; Valparaiso, 6200 ft, 1 (US); 5 mi SW Zacatecas, 4 (MVZ); 2 mi S, 5 mi E Zacatecas, 7700 ft, 1; 8 mi SE Zacatecas, 7225 ft, 14 (KU); San Juen Capistrano, 1 (US); 5 km NE San Juan Capistrano, 1330 m, 4; 3 km N San Juan Capistrano, 1500 m, 7; 5 km E San Juen Capistrano, l (OU); 13 km E Jérez, 2200 m, 5; 8 km SWJérez, 2030 m, 6; Berriozabal, 11 (US); 3 mi N Cuauhtémnc, 6600 ft, 5 (LACM); 3 mi WNW Saldafia, 6850 ft, 1; 5 mi S Ojo Caliente, 13 (U1); 2 km N Noria de Angeles, 2200 m, 2; 10 km ENE Loreto, 7350 ft, 7 (OU); 6 kmENE Ioreto, 6850 ft, 1 (OU); 7 mi S Pinos, 6800 ft, 17 (LACM); 20 km S Villenueva, 1810 m, l; 25 km ESE Pinos, 2425 m, 26; 30 km ENE Jalpa, 1740 m, 10; 2 mi S Tabasco, 2 (UI); 10 mi W Jalpa, 6100 ft, 7 (LACM); r mi S Jalpa, 4300 ft, 5 (IAGM); 13 mi WSW Jalpa, 1 (CU). Other records: 20 mi SW Concepcibn del Oro, (Jones end Webster, 1977); 20 mi S Villanueva (GAS); between Huenasco and Tabasco, (GAS); 7 mi N Jalpa, (Jones and Webster, 1977) ; in Jalisco, Quadrat 55 (Genoways end Jones, 1973); in San Luis Potosi, Quadrat 67 (Dalquest, 1953). 104 PEROGNA'I'HUS PENICIILATUS EREMICUS MEARNS Desert Pocket Mouse Distribution. -- Arid to semi-arid regions east of the grasslands in Zacatecas. Remarks. —— This species seems most abundant in the interior drainage basins where fine soils are deep. The western limit of this species distribution in Zacatecas coincides with the beginning of the grasslend vegetation. Two females taken on 10 July and 5 August con- tained four and five embryos, respectively. Two females taken on 7 and 9 July were lactating. Previous to this report, no specimens of the desert pocket mouse had been reported from Zacatecas , although they were krnown from both north and south of the state (Hoffmeister end Lee, 1967). The speci- mens are assigned to the subspecies E. p_. eremicus based on the charac- ters given in Hoffmeister end Lee (1967). Specimens Examined. -- Total 140, fram: 10 km ESE San Juan de los Charcos, 1500 m, 3; 6 km W Apizolaya, 1800 m, 1; 3 km SE Apizolaya, 1920 m, 23; 8 mi W Geckos, 5650 ft, 1 (LACM); 1 mi 8 Cedros, 6500 ft, 3; 15 mi NE Concepcion del Oro, 8 (OU); 14.5 mi NE Concepcién del Oro, 3 (OU); 12 km ENE Concepcion del Oro, 1850 m, 9; 13 km NE Concepcion del Oro, 1700 m, 2; 11 mi E Concepcion del Oro, 5300 ft, 3 (IAQ’I); 2 km SE Sabana Grende, 1940 m, 6; 3 mi N 1.1116, 7 (MVZ); 35 km SSE Concepcion del Oro, 1980 m, 1; 13 mi SW Camacho, 5800 ft, 8 (LACM), l (RGH); 3 km NW Sen Felipe de Nuevo Mercurio, 1770 m, 37; 2 km W Sen Felipe de Nuevo Mercmrio, 1740 m, 9; 3 km E San Tiburcio, 1880 m, 3; 15 km WSW San Tiburcio, 1980 m, 4; 6.5 krm S La Colorada, 1970 m, 7; 3 mi E, 4.6 mi N Villa de Cos, 1 (CU). 105 DIPODGVIYS MERRIAMI ATRONASUS MERRIAM Merriam' s Kangaroo Rat Distribution. -- Arid to semi-arid regions mostly east of the grasslands in Zacatecas. Remarks. -- This species seems most abundant in level desert situations of northeastern Zacatecas. The distribution of _D. merriami includes the eastern margin of the grasslends. The presence of _D_. merriemi marginally in grassland habitats may be the result of deterio- ration of the grasslands due to overgrazing. Eighteen females taken on 6 July, 1978, in the very arid region near Sen Tiburcio, did not show any sign of reproductive activity. Of 20 females taken on 8 July, 1978, in an area near Capriote with more luxurient vegetation, seven contained en average of 3.0 (renge, 2 to 4) embryos. Of 12 females taken on 14 July, 1978, in the area near Monte Mariana also in green vegetation, four contained an average of 2. 75 (renge, 2 to 3) embryos. These data may be suggestive of different timing of reproduction in different habitats. Lidicker (1960), inhis review of D. merriemi, recognized only one subspecies , D. p. atronasus, in Zacatecas but indicated that enother, D. p. ambiggg, might occur in the extreme northern part of the state. Specimens from the extreme north, now available, show some character- istics of erbiggg (lighter color) but in size all are referable to atronasus . Specimens Exemined. -- Total 557, from: 10 km ESE Sen Juan de los Charcos, 1500 m, l; 2.5 mi N La Pendencia, 5400 ft, 3 (OU); 6 km W Apizolaya, 1800 m, 51; 3 km SE Apizolaya, 1920 m, 26; 7 km SE Caopas, 1940 m, 5; 8 mi W Cedros, 5650 ft, 2 GAIN), l (RGH); 1 mi S Cedros, 106 5600 ft, 1; 4 km W Sen Rafael, 2140 m, l; 15 mi NE Concepcion del Oro, 3 (00); 14.5 mi mm Concepcién del Oro, 6 (OU); 16.4 mi ENE Concepcion del Oro, 5 (OU); 12 km ENE Concepcién del Oro, 1850 m, 22; 13 km NE Concepcion del Oro, 1700 m, 2; 11 mi E Concepcion del Oro, 5300 ft, 3 (LAW) ; 5 km SE Concepcion del Oro, 1940 m, 4; 18 km SSW Concepcion del Oro, 2130 m, 3; 2 km SE Sabena Grande, 1945 m, 9; 3 mi N 11116, 4 (MVZ); 15 mi S Concepcion del Oro, 6900 ft, 8 (KU); 7 km SE El Rosario, 2100 m, 8; 35 km SSE Concepcion del Oro, 1980 m, 8; 5.5 mi NW Juan Aldema, 6200 ft, 3 (OU); 13 mi SW Cemacho, 5800 ft, 1 (LAGM), 2 (RGH); 3 kmNW San Felipe de Nuevo Mercurio, 1770 m, 14; 10 mi SE Juan Aldame, 2210 m, 2; 1 mi SW San Tiburcio, 9 (KB); 3 km E San Tiburcio, 1880 m, 6; 15 km WSW San Tiburcio, 1980 m, 33; 15 kmn NNW Nieves, 1910 m, l; 9 mi NE Nieves, 6050 ft, 1; 4 km N Nieves, 1980 m, 5; 23 km NE Rio Grande, 1800 m, 9; 25 km WSW San Tiburcio, 2030 m, 4; 12 mi NW Rio Grende, 6900 ft, 2; 15 mi NE Sen Andres, 1 (LAO/I); 5 mi SE Rio Grande, 6350 ft, 5; 5 km S La Colorada, 1960 m, 3; 6.5 km S La Colorada, 1970 m, 15; 8 mi S Majoma, 7700 ft, 17 (KU); 6 km SE Tetillas, 2040 m, 4; 13 kaNW Capriote, 2100 m, 39; 9 km NW Sarteneja, 2200 m, 2; 10 mi N Rancho Grende, 6700 ft, 6; Cafiitas, 6; 1 mi N Rencho Grende, 1 (KU); 1 mi W Sierra Vieja, 6100 ft, 14 (OU); 18 kmn NE Villa de Cos, 4; 23 km N Fresnillo, 2140 m, 2; 3 mi E, 4.6 mi N Villa de Cos, 3 (OU); 13 mi NNW Fresnillo, 2 (LAQ’I); Villa de Cos, 6700 ft, 9 (KU); 2 km S Monte Mariana, 2180 m, 34; 2 mi SE Villa de Cos, 6200 ft, 1; 45 kmn NE Morelos Jct, 3 (TI‘U); Bafnon, 6400 ft, 33 (IAGM); 2miE Bafion, 6400 ft, 2 (LACM), 4 (RGH); 4 km ESE Bafion, 6200 ft, 6 (OU); 4.5 mi E Fresnillo, 12 (MVZ) ; 10 km SE Fnesnillo, 2250 m, 4; 24 mi NE Zacatecas, 6700 ft, 2 (LACM), 2 (RGH); 8mi SE Zacatecas, 7225 ft, 16(KU); 2miESE 107 Trancoso, 7000 ft, 6 (K11); 3 mi N Cuauhtemoc, 6600 ft, 13 (IACM); 5 mi S Ojo Caliente, 5 (U1); 3 mi WNW Saldafia, 6850 ft, 2; 2 km N Noria de Angeles, 2200 m, 2; 7 mi S Pinos, 6800 ft, 1 (IACM); 10 mi S Pinos, 7100 ft, 1 (LACM); 25 km ESE Pinos, 2425 m, 1. Other records: 22 mi SW Concepcion del Oro, (Jones end Webster, 1977); 14 mi. S Concepcion del Oro, (GAS); 20 mi S Concepcién del Oro, (GAS); 17 mi NW Fresnillo, (GAS); 16 mi NW Fresnillo, (GAS); 8 mi W Fresnillo, (Jones end Webster, 1977). DIPODCMYS NEISCNI MERRIAM Nelson's Kangaroo Rat Distribution. -- Arid regions of northern end northeastern Zacatecas. Remarks. -- Scattered mounds of this species are conspicuous in desert flatlends end the interior basins where soils are sandy. Its western end southern limits coincide with the desert-grassland ecotone. A female taken on 22 July contained 2 embryos. Alvarez (1960) listed a specimen firm 3 mi N Lulu as Q. spectabilis cratodon. According to Cornelio Sénchez (personal communi- cation), this specimen is, in fact, 2. nelsoni. It was part of a collection made by the MVZ in 1940. Other specimens fran that locality were examined and all are referable to _D. nelsoni. Specimens Examined. -- Total 76, from: 10 km ESE Sen Juan de los Charcos, 1500 m, 2; 6 km W Apizolaya, 1800 m, 2; 3 km SE Apizolaya, 1920 m, 4; 7 km SE Caopas, 1940 m, 6; 8 mi W Cedros, 5650 ft, 1 (RGH); 15 mi NE Concepcibn del Oro, 12 (OU); 14.5 mi ENE Concepcion del Oro, l (OU); 16.4 mi ENE Concepcion del Oro, 2 (OU); 10 mi E Concepcion del 108 Oro, 5300 ft, 1 (LACM); 11 mi E Concepcién del Oro, 5300 ft, 1 (IAGM); 16 km SW Concepcién del Oro, 1900 m, 1:3 mi N Lulu, 4 (MVZ); 7 km SE El Rosario, 2100 m, 1; 35 km SSE Concepcion del Oro, 1980 m, 2; 20 mi SSW Concepcion del Oro, 5 (KU); 22 mi SSW Concepcion del Oro, 6000 ft, 3 (KU); 13 mi SW Camacho, 5800 ft, 5 (LACM); 3 kmNW San Felipe de Nuevo Mercurio, 1770 m, 1; 2 km W Sen Felipe de Nuevo Mercurio, 1740 m, 2; 6 mi NE Sen Tiburcio, 6100 ft, 1 (K11); 3 km E San Tiburcio, 1880 m, l; 1.5 mi SW Sen Tiburcio, 6100 ft, 2 (K11); 3.5 mi SW San Tiburcio, 6100 ft, 3 (KU); 4 mi SW San Tiburcio, 6100 ft, 3 (KU); 7 mi SW San Tiburcio, 6100 ft, 1 (KU); 7.5 mi SW San Tiburcio, 6000 ft, 3 (KU); 8 mi SW Sen Tiburcio, 1 (K11); 9 mi NE Nieves, 6050 ft, 1; 4 km N Nieves, 1980 m, 3; 23 km NE Rio Grande, 1880 m, 1. Other records: 14 mi S Concepcion del Oro, (GAS). DIPODCMYS ORDII WDHOUSE Ord' s Kengaroo Rat Distribution. -- Arid to semi-arid grasslands and deserts east of the Sierra Madre Occidental in Zacatecas . Remarks. -- This species is widely distributed throughout the grasslands end desert of Zacatecas but is less abundant in any given locality than in D. merriamni. It occurs further west into the grass— lends then does 2. merriami. Of forty females taken in June, July, and August , for which data are available , none showed eny evidence of reproductive activity . Based upon characters given in Setzer (1949) , I recognize three subspecies in Zacatecas. In the extrene north specimens are assigned to _D_. _g. idoneus based on their small size and pale color. In 109 northeastern Zacatecas specimens assigned to 12. _o_. _f_us_gs_ are large and pale colored. In central and southeastern Zacatecas , specimens assigned to Q. g. palmeri are medium sized and dark colored. Specimens Exemined. -- Dipodaryys ordii fuscus Setzer, total 43, from: 14.5 mi ENE Concepcion del Oro, l (OU); 12 km ENE Concepcion del Oro, 1850 m, 17; Concepcion del Oro, 7600 ft, 1 (KU); 10mi E Concepcion del Oro, 5300 ft, 1 (IAQ’I); 11 mi E Concepcion del Oro, 5300 ft, 2 (IAQ’I), l (RGH); 16 km SW Concepcién del Oro, 1900 m, 2; 3 mi N Lulu, 5 (MVZ); 35 km SSE Concepcion del Oro, 1980 m, 11; 3 mi E San Tiburcio, 1880 m, 2. Dipodomys ordii idoneus Setzer, total 17, from: 3 km SE Apizolaya, 1920 m, 5; 7 km SE Caopas, 1940 m, 10; 13 mi SW Cemacho, 5800 ft, 1 (IAGM); 3 km NW Sen Felipe de Nuevo Mercurio, 1770 m, l. Dipodomys ordii palmeri (J. A. Allen), total 223, from: 15 km NNW Nieves, 1910 m, 3; 23 km NE Rio Grende, 1800 m, 4; 15 mi NE San Anch'és, 6200 ft, 2 (IACM); 13 km WNW Rio Grende, 2050 m, 1; 6.5 km S La Colorada, 1970 m, 9; Santa Efigenia, 7400 ft, 1 (KU); 5.5 mi SW Sambrerete, 6850 ft, 2 (LAGM); 15.5 mi WSW Somnbrerete, 6400 ft, 1 (OU); 10 mi N Rancho Grende, 6700 ft, 1; Gafiitas, 5 (US); 1 mi S Cafnitas, 2 (Tl'U); 3 km SE Rancho Grande, 2190 m, 1; Villa de Cos, 6700 ft, 3 (KU); 13 mi. NNW Fresnillo, 3 (LAGM); 2 km 8 Monte Mariana, 2180 m, 2; Bafnon, 6400 ft, 22 (IACM); 2 mi E Bafion, 6400 ft, 1 (RGH); 4kmESE Bafnon, 6200 ft, 5 (OU); 4.5 mi E Fresnillo, l (MVZ); 10 mi N Zacatecas, 6200 ft, 1 (KU); 2 mi SE Calera, 7300 ft, 3 (KU); 2 mi S, 5 mi E Zacatecas, 7700 ft, 27; 9 mi SE Zacatecas, 7900 ft, 1 (10.1); 8 mi SE Zacatecas, 7225 ft, 65 (KU); 3 mi SE Guadalupe, 7 (UI); 7 mi SE Trancoso, 1 (K11); 13 km E Jérez, 2200 m, 1; 8 mi N Villenueva, 6800 ft, 1 (K11); 5 mi S 110 Ojo Caliente, 2 (U1); 6 mi NNW Pinos, 7900 ft, 2; 7 mi S Pinos, 6800 ft, 12 (LACM); 10 mi S Pinos, 7100 ft, 1 (RGH); 25 kmESE Pinos, 2425 m, 1. DIPODOVIYS PHILLIPSII ORNATUS MERRIPM Phillip's Kengaroo Rat Distribution. -- Grasslends of western end southern Zacatecas. Remarks. -- This species seems to be most abundent in short grass habitats. In Zacatecas, this kengaroo rat is recorded from open areas in the tropical scrub vegetation of the deep cenyons (1100 m) to the highest grassland plains (2400 m). Three females taken on 19 and 20 July and 3 August contained three embryos each. Other natural history notes on this species cen be found in Genoways and Jones (1971) . All specimens from Zacatecas are assigned to the subspecies 13. p. ornatus following the review by Genoways end Jones (1971) . They found specimens from low elevations to be lighter colored than those from higher elevations. I see the same trend in samples which I have examined. Specimens from near Sen Juen Capistreno (1100 m) are the lightest colored of all populations in Zacatecas. Specimens Exemined. -- Total 99, from: 5 km S Gonzales Ortega, 2150 m end 2320 m, 7; 3 km E El Arenal, l; 5.5 mi SW Sambrerete, 6850 ft, 1 (LAGM); 5 km NE Chalchihuites, 2360 m, 2; 2 km S Monte Mariena, 2180 m, 2; 45 km SW Fresnillo, 2165 m, 1; 53 km SW Fresnillo, 2250 m, 2; 11 mi NE Valparaiso, 7100 ft, 1; Valparaiso, 6200 ft, 11 (US); 18 kmn N Sen Juen Capistrano, 11 m, 7; Zacatecas, 4 (US); 5 mi SW Zacatecas, 4 (MVZ); 2 mi S, 5 mi E Zacatecas, 1; 8 mi SE Zacatecas, 7225 ft, 4 (K11); 5 km NE Sen Juen Capistrano, 1330 m, 2; San Juan Capistreno, 3 (US); 6 111 mi W Jérez, 6700 ft, 1; 13 km E Jérez, 2200 m, 2; 8 km SW Jérez, 2030 m, 4; Berriozébal, 2 (US); 3 mi N Cuauhtemoc, 6600 ft, 15 (LACM); 5 mi S Ojo Caliente, 2 (U1); 2 mi N Villenueva, 1 (K11); 6 mi NNW Pinos, 7900 ft, 1; 7 mi S Fires, 6800 ft, 1 (LAOVI); 25 kmESE Pinos, 2425 m, 2; 45 km S Pinos, 2350 m, 1; Plateado, 5 (US); 30 kmNE Jalpa, 1740 m, 1; 5.5 mi SW Jalpa, 4400 ft, 3 (CU). Other records: 40 mi W Fresnillo, (Jones and Webster, 1977); 12 mi N, 7 mi E Fresnillo, (Genoways end Jones, 1971); 1 mi. S Mcnmax, (GAS); in Jalisco, Quadrat 55 (Genoways and Jones, 1971). DIFOIDMYS SFECIABEIS CRATODON MERRIAM Banner- tailed Kangaroo Rat Distribution. -- Grassland and desert-grassland ecotone of central end southeastern Zacatecas. Remarks. -- This species seems to be most abundent in the desert- grassland ecotone habitats. They do not enter traps readily, most specimens being secured either by digging then from their burrows or shooting them at night. Little is known about the natural history of this southernmost, disjunct population of D. spectabilis. Specimens Examined. -- Total 113, from: 28 mi NE San Andres, 6700 ft, 5 (KU); 6.5 km S La Colorada, 1970 m, 1; 13 km WNW Capriote, 2100 m, 2; 12mi SWSen Andres, 6000 ft, 3 (KU); Caf‘fitas, 3 (US); 40 kmNE Villa de Cos, 2000 m, 1; 3 mi E, 4.6 mi N Villa de Cos, 38 (OU); Villa de Cos, 6700 ft, 3 (K11); 13 mi NNWFresnillo, l (LAGM); Bafnon, 6400 ft, 17 (mm); 4.5mi EFresnillo, 1 (MVZ): Smi SWBafen, 3 (KU); 10mi N Zacatecas, 6200 ft, 3 (K11); 8 mi W Zacatecas, 1 (K11); 8 mi SE Zacatecas, 7225 ft, 3 (K11); 1.5mi E Jérez, 7000 ft, 2 (KU); 112 Berriozabal, 4 (US); 3 mi N Cuauhténoc, 6600 ft, 4 (IAGM); 8 mi N Villenueva, 6800 ft, 2 (KU); 6 mi N Villenueva, 6500 ft, 3 (KU); 4mi N Villenueva, 6400 ft, 4 (KU); 2 mi N Villenueva, 6500 ft, 4 (KU); 5 mi S Ojo Caliente, 2 (UI); Pinos, 4 (AMNH). LICMYS IRRORATUS (GRAY) Mexican Spiny Pocket Mouse Distribution. -- Grasslends to tropical canyons in Zacatecas. Remarks. -- On the Mexican Plateau, this species seems most abun- dant in brushy areas where ropal (quntia sp) is a daminant plant. In tropical cenyons it is also found in heavy brush habitats. In the desert regions of the northeast, it occurred in islands of mesquite- grasslend. Five females taken between 27 July end 18 August contained en average of 4.2 (renge, 3 to 5) embryos. One female taken on 5 July contained five embryos and was lactating. Two subspecies were recognized as occurring in Zacatecas by Genoways (1973). A large sized _I_.._. i. gl._]_._en_i_ occurs on the Plateau and western slopes of the tropical canyons. A smaller L. i. jaliscensis lives in the southernmost trOpical canyon near Santa Rosa. Geroways (1973:102) noted that specimens frcm Zacatecas that be assigned to jaliscensis ". . . evince intergradation between jaliscensis end _al_1£¢__n_i." I have followed Genoways (1973) in the assignment of sub- species, reting that tl’e specimens from near Senta Rosa in the valley of the Rio Juchipila are difficult to identify to one or the other sub- species. Specimens here assigned to £159; from near La Colorada, San Felipe de Nuevo Mercurio, and Apizolaya, all in desert islands of 113 mesquite-grassland, extend the hem renge of I: i. alleni some 100 km east of that given in Geeways (1973). Specimens Examined. -- Lionyp irroratus alleni (Coues), total 249, from: 3 km SE Apizolaya, 1920 m, 1; 2 km W San Felipe de Nuevo Mercurio, 1740 m, 5; 9 mi NE Nieves, 6050 ft, 1; 23 km NE Rio Grende, 1800 m, 1; 5 km S Gonzales Ortega, 2320 m, 2; 13 km WNW Rio Grende, 2050 m, l; 5 mi SE Rio Grande, 1940 m, 2; 5 km S La Colorada, 1960 m, 2; 6.5 km S La Colorada, 1970 m, l; 6 km SE Tetillas, 2040 m, 2; 15.8 mi W Sambrerete, 7050 ft, 2 (OU); 10.4 mi WSW Somnbrerete, 7250 ft, 2 (OU); 14.5 mi WSW Sambrerete, 7100 ft, 7 (OU); 15.5 mi WSW Somnbrerete, 6400 ft, 1 (OU); 10 mi N Rancle Grande, 6700 ft, 11; 14.2 mi N Jimenez del Téul, 7300 ft, 1 (OU); 13 mi NNW Fresnillo, 1 (IAQ’I); 6 km N Fresnillo, 2250 m, 5; 2.5 km WNW Fresnillo, 7400 ft, 1 (OU); Baf'non, 6400 ft, 15 (LAOVI); 2mi E Bafnon, 6400 ft, 1 (LAGM); 4 kmESE Bafnon, 6200 ft, 2 (OU); 11 mi NE Valparaiso, 7100 ft, 5; Sen Juan Capistreno, 2 (US); 6 mi W Jérez, 6700 ft, 3; 13 km E Jérez, 2200 m, 13; 3.5 km W La Blenca, 6650 ft, 13 (OU); 11.7 mi NW Cuauhtémoc, 7100 ft, 2 (OU); 9.7 mi NW Cuauhtémoc, 7100 ft, 5 (OU); 3 mi N Cuauchtenoc, 6600 ft, 27 (IAGM); 3 mi WNW Salde'ia, 6850 ft, 11; 13 km N Villenueva (Ruinas Chicomoztoc), 1 (LACM); 10 km ENE Loreto, 7350 ft, 4 (OU); 7 mi 8 Pinos, 6800 ft, 2 (IAGM); 10mi S Pinos, 7100 ft, 4 (LAG/1), l (RGH); 25 km ESE Fires, 2425 m, 1; 16 km SSE Monte Escobedo, 2010 m, 2; 20 kmn S Villenueva, 1810 m, 2; 45 km S Fires, 2350 m, l; 2.5 mi. S Momax, 5; 30 km NE Jalpa, 1740 m, 18; 3 mi NE Jalpa, 5400 ft, 2; 10 mi W Jalpa, 6100 ft, 7 (LACM); 5.5 mi SW Jalpa, 4400 ft, 22 (OU); 25 kmn ESE Jalpa, 2590 m, 1; 13 mi WSW Jalpa, 6000 ft, 3 (OU); 4 mi. 8 Jalpa, 4300 ft, 6 (LACM); 8 mi NW Nochistlén, 6600 ft, 15 (IAQ’I); 6 km S Apozol, 1170 m, 114 14. Other records: Geeways (1973) unless otherwise noted; 2 mi W Sain Alto, 4 mi NNW Chalchilnuites; 1 mi N Rancle Grande; 5 mi NW Zacatecas; 9 mi W Zacatecas; 2 mi ESE Trancoso; 7 mi SE Trancoso; 8 mi SE Zacatecas; Berriozabal; 1 mi NE Noria de Angeles; 13 mi N Jalpa; 3 mi SW Jalpa; in Durengo, Quadrat 8 (MSU); in Jalisco, Quadrats 48 and 55. Liomys irroratus jaliscensis (J. A. Allen), total 60, from: .5 mi ENE Mesquituta, 3450 ft, 11 (OU); 2.5 mi N Santa Rosa, 3700 ft, 8 (OU); 2 mi N Senta Rosa, 3850 ft, 11; 1.5 mi N Santa Rosa, 3900 ft, 3 (OU); 1.1 mi N Santa Rosa, 3500 ft, 1 (OU); 1 mi N Santa Rosa, 3600 ft, 11; Senta Rosa, 4000 ft, 14 (IAGM); 7 km SE Senta Rosa, 1500 m, 1. LIOMYS PICTUS HISPIDUS (J. A. ALLEN) Painted Spiny Pocket muse Distribution. -- Known fram only one locality in the tropical can- yons in Zacatecas. Remarks. -- A single specimen of this species, from near Sen Juan Capistree, was reported upon by Matson et a1. (1978). At a slightly different locality, L. irroratus has also been recorded (see above account). Genoways (1973) suggested that in the few areas where the two species have been reported in sympatry, L. pic—tug seemed to be associated with more mesic lowlend situations then L. irroratus. The specimen of L. Ripping was taken from a sugar cane field adjacent to the Rio Atengo while the exact conditions from which t‘re specimens of L. irroratus were taken are unknown. This specimen of L. Riggs is assigned to the subspecies hispidus on geographic grounds (Genoways , 1973). 115 Specimens Exemined. -- Total 1, from: 5 km NE San Juan Capistrano, 1330 m, l. REITTRODON'I'CNYS FULVESGENS J. A. ALLEN Fulvous Harvest Mouse Distribution. -- Widely distributed throughout Ion-forested parts of Zacatecas. Remarks. -- In desert areas of rertheastern Zacatecas, this species seems most abundent in grassy or brushy, as opposed to open, habitats. In the western parts of the state there was re apparent habitat preference. Six females taken between 15 July end 2 August contained an average of 4.5 (range, 4 to 6) embryos. Two females taken on 22 July end 6 August were lactating. The fulvous harvest mouse was reported to use old bird nests by Jones and Webster (1977). Hooper (1952) recorded one subspecies as occurring in Zacatecas with the indication that a second may also occur there. Following the characters given by Hooper (1952) which differentiate the two possible subspecies, I recognize R_. _f. w in northern Zacatecas based upon its pale color, small auditory bullae, and narrow frontal region of the skull. Specimens assigned to _R. _f. griseoflavus are darker (redder), with a more or less "distinct dorsal stripe," larger bullae, and broader frontal reg ion. Specimens Exemined. -— Reithrodontamys fulvescens canus Benson, total 26, from: 6 km W Apizolaya, 1800 m, l; 3 km SE Apizolaya, 1920 m, 2; 12 km ENE Concepcion del Oro, 1850 m, 8; 6 km E Mazapil, 2645 m, 1; 18 km ssw Concepcion del Cm, 2130 m, 1; 3 km 81: Sen Felipe de Nuevo Mercurio, 1770 m, 6; 2 km W San Felipe de Nuevo Mercurio, 1740 m, 7. 116 Reithrodontomys fulvescens griseoflavus Merriam, total 116, fromn: 5 km S Gornzales Ortega, 2150 m, 2; 23 km NE Rio Grande, 1800 m, l; 15 mi NE San Andres, 6200 ft, 1 (IACM); 13 km WNW Rio Grende, 2050 m, l; 5 mi SE Rio Grende, 1940 m, 1; 3 mi E E1 Calabazal, 8000 ft, 3; 15.8 mi W Somnbrerete, 7050 ft, 1 (OU); 9 mi W Sambrerete, 7900 ft, 1 (OU); 3 km E El Arenal, 2460 m, 2; 13 km WNW Capriote, 2100 m, 1; 15.5 mi WSW Sombrerete, 6400 ft, 1 (OU); 1 mi W Sierra Vieja, 6100 ft, 1 (OU); 2 km S Monte Mariana, 2180 m, 1; 6 km N Fresnillo, 2250 m, 5; 3 mi NW Fresnillo, 7760 ft, 1; 4 km ESE Bafion, 6200 ft, 2 (G1); 10 km SE Fresnillo, 2250 m, 2; 53 km SW Fresnillo, 2250 m, l; Valparaiso, 6200 ft, 6 (US); 3.5 km W La Blanca, 6650 ft, 3 (OU); 6 mi W Jérez, 6700 ft, 1; 13 km E Jérez, 2200 m, 4; 8 km SW Jérez, 2030 m, 7; 32 kmn SE Valparaiso, 2040 m, 6; 36 km SSE Valparaiso, 2330 m, 3; Berriozabal, 1 (US); 5 mi S Ojo Caliente, 5 (U1); 3 mi WNW Saldaf'na, 6850 ft, 1; 10 kmn S Tepetongo, 1950 m, 1; 10 km ENE Loreto, 7350 ft, 2 (OU); 16 km SSE Monte Escobedo, 2010 m, 4; 20 km S Villenueva, 1810 m, 4; 25 kmn ESE Pinos, 2425 m, 10; 45 kmn S Pinos, 2350 m, 8; 30 km NE Jalpa, 1740 m, 5; 10 mi W Jalpa, 6100 ft, 5 (LACM); 4 mi S Jalpa, 4300 ft, 3 (IACM); 8 mi NW Nochistlén, 6600 ft, 3 (IACM); .5 mi ENE Mesquituta, 3450 ft, 1 (G1); Santa Rosa, 4000 ft, 2 (IAGM); 7 km SE Santa Rosa, 1500 m, 3. Other records: 9 mi. W Zacatecas, (GAS); 8 mi S Pbyahua, (Jones end Webster, 1977). REITHRODONTG’IYS MEGAIOI‘IS (BAIRD) Western Harvest Mnuse Distribution. -- Widely distributed throughout Zacatecas except for the tropical canyons. 117 Remarks. -- This species seems most abundant in brushy to grassy habitats in the desert. However, unlike R_. fulvescens, it was also captured in more open areas. The western harvest mouse was taken most regularly in the grasslends and desert-grassland ecotone habitats of Central Zacatecas. Five females taken between 16 end 24 July contained an average of 3.8 (range, 2 to 5) embryos. A female taken on 18 July was lactating. Hooper (1952) recorded two subspecies occuring in Zacatecas: S. p. megalotis, a light colored subspecies of the Plateau; end, S. p. zacatecae, a dark colored subspecies in the Sierra Madre Occidental. Specimens Exemined. -- Reithrodontomys megalotis megalotis (Baird), total 89, fram: 7 km SE Caopas, 1940 m, 4; 6 km E Mazapil, 2645 m, 3; 2 km SE Sabana Grande, 1945 m, 2; 3 mi N Lulu, 1 (MVZ); 5.5 mi NW Juan Aldama, 1890 m, 1 (OU); 5 km NW Juan Aldama, 1980 m, l; 3 km NW San Felipe de Nuevo Mercurio, 1770 m, 3; 2 km W Sen Felipe de Nuevo Mercurio, 1740 m, 13; 5 km S Gonzales Ortega, 2150 m, 1; 23 km NE Rio Grande, 1800 m, l; 13 km WNW Rio Grende, 2050 m, l; 5 kmn S La Colorada, 1960 m, l; 2 mi S Senta Efigenia, 7400 ft, 2 (K11); 13 km WNW Capriote, 2100 m, 3; 8 mi W, 1 mi N Somnbrerete, 7800 ft, 1; 9 mi. W Somnbrerete, 7900 ft, 7 (OU); 3 km E El Arenal, 2460 m, l; 5 km NE Chalchihuites, 2360 m, 2; 1 mi N Rencl'e Grande, 4 (KU); 3 km SE Rencl'e Grande, 2190 m, 1; 5 mi WSW Sierra Vieja, 6200 ft, 1 (OU); 2 km S Monte Mariana, 2180 m, 4; 6 km N Fresnillo, 2250 m, 3; 3 mi NW Fresnillo, 7760 ft, 2; 2.5 kaNWFresnillo, 7400 ft, 1 (OU); Be’ion, 6400 ft, 2 (LAGM); 4kmESE Baf'non, 6200 ft, 1 (OU); 10 km SE Fresnillo, 2250 m, 2; Zacatecas, 2 (US); 2 mi S, 5 mi E Zacatecas, 7700 ft, 1; 9 mm' SE Zacatecas, 7900 ft, 3 (KU); 13 kmE Jérez, 2200 m, 3; 5 mi S Ojo Caliente, 1 (UT); 2 km N 118 Noria de Angeles, 2200 m, 1; 6 mi NNW Pinos, 7900 ft, 1; 25 km ESE Fires, 2425 m, 5; 45 km S Fires, 2350 m, 3. Other records: Jones and Webster (1977); 22 mi SW Concepcion del Oro; 40 mi W. Fresnillo. Reithrodontomys megalotis zacatecae Merriam, total 24, from: 25 km WSW Milpillas de la Sierra, 2580 m, 2; 9 mi NW Valparaiso, 8350 ft, 1; Valparaiso mountains, 8700 ft, (type locality), 14 (US); 36 km SSE Valparaiso, 2330 m, 1; 25 km ESE Jalpa, 2590 m, 6. Other records: 8 mi S Chalchilnuites, (Jones end Webster, 1977); in Aguascalientes, Quadrat 57 (I-boper, 1955). PERCMYSCUS BOYLII RCMLEYI (J. A. ALLEN) Brush muse Distribution. -- antene regions of western Zacatecas. Remarks. -- The brush mouse seems most abundant at high elevations in pine-oak forests. However, it also occurs at lower elevations in brushy or rocky habitats. Nine females taken between 23 July end 7 August contained an average of 2.67 (range, 1 to 4) embryos. Six females obtained between 23 July end 7 August were lactating. Carleton (1977) derenstrated that _F. spicilegus (see below) was specifically distinct frcm _P_. b_OY_]_.1LIL. He also indicated the need for more taxonomic work to be done on various populations of 1:. M. I have assigned all Zacatecas specimens of L. 1391]}; to the subspecies _F. p. IOWIEE' on the basis of geography (Hall and Kelson, 1959). How- ever, same specimens fram localities in the extrene west end south appear to be larger then specimens from localities further east end north. 119 Specimens Examined. -- Total 292, from: 2 mi E Villa Insurgentes, 8050 ft, 4; 2 mi. E E1 Calabazal, 8000 ft, 2; 14.8 mi W Sambrerete, 7250 ft, 3 (OU); 13.6 mi W Sambrerete, 7450 ft, 1 (OU); 9.5 mi W Somnbrerete, 7900 ft, 11 (G1); 9 mi W Somnbrerete, 7900 ft, 5 (OU); 10.4 mi WSW Sambrerete, 7250 ft, 12 (OU); 10.7 mi WSW Somnbrerete, 7250 ft, 2 (OU); 15.5 mi WSW Somnbrerete, 6400 ft, 1 (OU); 3 km E El Arenal, 2460 m, 2; 21 mi SW Sambrerete, 7800 ft, 2 (IACM), 2 (RGH); 10 mi SW Chalchihuites, 7200 ft, 12 (LAGM); 14.2 mi N Jimenez del Téul, 7300 ft, 18 (OU); 4 km E Jimenez del Téul, 2375 m, 1; 25 km WSW Milpillas de la Sierra, 2580 m, 7; 9 mi NW Valparaiso, 8350 ft, 11; Valparaiso mountains, 8700 ft, 1 (US); 8.2 mi W Valparaiso, 7900 ft, 14 (OU); 6.2 mi W Valparaiso, 7500 ft, 4 (OU); 6 mi W Valparaiso, 7400 ft, 1 (OU); Valparaiso, 13 (US); 15 km SW Valparaiso, 2250 m, 18 ; Sierra Madre, 10 (US); 10 km W San Juan Capistreno, 2900 m, 7; 8 kmn W San Juen Gapistree, 2110 m, 6; 32 km SE Valparaiso, 2040 m, 13; 36 km SSE Valparaiso, 2330 m, 4; 11.7 mi NW Cuauhteroc, 7100 ft, 7 (OU); Monte Escobedo, 1 (US), 7300 ft, 13 (IACM); 16 km SSE Monte Escobedo, 2010 m, 1; 20 km S Monte Escobedo, 1920 m, 6; Plateado, 4 (US); 3.5 mi W Tlatenengo, 6500 ft, 3; 11 mi. NW Jalpa, 8000 ft, 7 (KU); 9 mi WNW Jalpa, 8250 ft, 13; 25 km ESE Jalpa, 2590 m, 2; 8 mi NW Nochistlén, 6600 ft, 16 (LACM); 6 km SW Téul de Gonzales Ortega, 1; 6 kmn S Téul de Gonzales Ortega, 2010 m, 30; 10 mi NW Yahualica, 7100 ft, 1 (LACM). Other records: Jones end Webster (1977) unless otherwise noted; 4 mi NW Chalchihuites; 8 mi S Chalchihuites, 40 mi W Fresnillo; 16 mi W Fresnillo, (GAS); 18 mi W Milpillas (de la Sierra), (GAS); 3 mi NW Téul de Gonzales Ortega, (GAS). 120 FERQVIYSCUS DIFFICILIS (J. A. ALLEN) Rock Mouse Distribution. -- Montene regions of western and northeastern Zacatecas. Remarks. -- The rock mouse seems most abundant in or around rock outcrops or fences in the montene regions of the state. Five females taken between 29 end 31 July, 1970, from northwest of Valparaiso con- tained an average of 3.0 (renge, 2 to 4) embryos, Two females taken during the same time at the same place were lactating. Six females taken on 27 and 28 July, 1977, from southeast of Jalpa contained an average of 3.2 (renge, 2 to 4) embryos. One of these females was lac- tating as well. Of 31 females taken in June from the Sierra Madre Occidental, for which data were recorded, none slnowed eny sign of reproductive activity. Four females taken on 30 June and 1 July, 1978, in the Sierra Astillero, contained three embryos each. It is not possible to determine whether this difference in reproductive timing is due to a geographic phenomeen, climatic variations in differene years , or to small sample size. In reviewing the subspecies of _F. difficilis, Hoffmeister end de la Torre (1961) recognized one subspecies, F. d. difficilis, as occur- ring in Zacatecas. They also indicated that a second subspecies, _P. d. petricola, may live in the rertheastern part of the state. later, Diersing (1976) documented the occurrence of L. _d_. petricola from near Concepcién del Oro. As reported by Hoffmeister end de la Torre (1961) , Etricola is larger in most measurements (smnaller hind foot length) end its color more gray then difficilis. 121 finecimens Exemined. -- Peromnyscus difficilis difficilis (J. A. Allen), total 207, fram: 2 mi E Villa Insurgentes, 8050 ft, 2; 3 mi E E1 Calabazal, 8000 ft, 1; 9.5 mi W Sombrerete, 6 (OU); 10 mi SW Chalchihuites, 7200 ft, 16 (IAGM); 4 km E Jimenez del Téul, 2375 m, 2; 25 km WSW Milpillas de la Sierra, 2580 m, 4; 45 km SW Fresnillo, 2165 m, 2; 9 mi NW Valparaiso, 8350 ft, 28; Valparaiso mountains, 8700 ft, (type locality), 24 (US); Zacatecas, 14 (US); Sierra Madre, 1 (US); 10 km W San Juen Capistreno, 2900 m, 2; 36 km SSE Valparaiso, 2330 m, 1; 11.7 mi NW Cuauhtémoc, 7100 ft, 4 (G1); Monte Escobedo, 7300 ft, 5 (LAG/1); 10 km ENE Ioreto, 7350 ft, 12 (OU); 25 km ESE Pinos, 2425 m, 6; Plateado, 19 (US); 11 mi NW Jalpa, 8000 ft, 8 (KU); 9 mi WNW Jalpa, 8250 ft, 12; 25 km ESE Jalpa, 2590 m, 37; 8 mi NW Nochistlen, 6600 ft, 1 (IACM) . Other records: 8 mi S Chalchihuites, (Jones and Webster, 1977); 1 mi NE Noria de Angeles, (GAS); in Aguascalientes, Quadrat 57 (Hooper, 1955); in Jalisco, Quadrat 55 (Genoways and Jones, 1973). Peromyscus difficilis petricola Hoffmeister and de la Torre, total 76, from: 5 km E Mazapil, 2270 m, 2; 6 km E Mazapil, 2645 m, 8; 7 mi E Mazapil, 1 (U1), 1 (TN); 4mi SEMazapil, 8425 ft, 21 (OU); 6 mi W Concepcién del Oro, 8 (UI); 4 mi W Concepcion del Oro, 10 (U1) ; 3 mi W Concepcion del Oro, 1 (UT); 10 mi SW Concepcion del Oro, 7600 ft, 5 (IAGM); 40 km ESE Concepcion del Oro, 2320 m, 19. FEROMYSCUS EREMICUS FHAEURUS OSGOOD Cactus Mouse Distribution. -- Arid to semi-arid regions east of the grasslands in Zacatecas . 122 Remarks. -- The cactus mouse seems most abundent in brushy habi— tats in the deserts. Its western limit coincides with the western edge of the desert-grasslend ecotone in Zacatecas . Ten females taken between 9 July and 17 August contained en average of 3.1 (renge, 2 to 4) embryos. Six fereles taken between 4 July and 17 August were lac- tating. The subspecies of L. eremicus are in need of a revisionary study. Baker end Greer (1962) considered all specimens of F. eremicus from south of the Rio Nasas, in Durango, to belong to the subspecies L. g. phaeurus. On geographic grounds , then, all Zacatecas specimens are so assigned. Specimens Exemined. -- Total 229, fram: 10 km ESE Sen Juen de los Charcos, 1500 m, l; 6 km W Apizolaya, 1800 m, 2; 3 km SE Apizolaya, 1920 m, 13; 8 mi W Cedros, 5650 ft, 1 (IAGM); 20 km NE Concepcion del Oro, 1910 m, l; 15 mi NE Concepcion del Oro, 1 (OU); 18.6 mi. ENE Concepcion del Oro, 2 (OU); 12 km ENE Concepcion del Oro, 1850 m, 11; 13 km NE Concepcion del Oro, 1700 m, 1; Concepcién del Oro, 7600 ft, 1 (K11) ; 10 mi E Concepcion del Oro, 5300 ft, 1 (LACM); 11 mi E Concepcion del Oro, 5300 ft, 8 (IACM); 2 (RGH); 1 mi WTecolotes, 6150 ft, 1; 5 km SE Concepcion del Ori, 1935 m, l; 12 mi SE Concepcidn del Oro, 7450 ft, 1; 25 km NE Camacho, 1975 m, 3; 2 km SE Sabena Grende, 1945 m, 6; 3 mi NLulCn, 2 (MVZ);1miNLu1fi, 1; SkmNWJuanAldama, 1980m, 1; 13mi SW Cernacre, 5900 ft, 1 (IACM); 3 km NW Sen Felipe de Nuevo Mercurio, 1770 m, 3; 2 km W Sen Felipe de Nuevo Mercurio, 1740 m, 6; 10 mi SE Juen Aldamna, 2210 m, 2; 3 km E San Tiburcio, 1880 m, 2; 1 mi SW San Tiburcio, 7000 ft, 13 (KU); 15 km WSW San Tiburcio, 1980 m, 2; 15 kmn NNW Nieves, 1910 m, 3; 23 km NE Rio Grande, 1800 m, 6; 15 mi. NE Sen 123 Andres, 6200 ft, 1 (LACM); 25 km SW Sen Tiburcio, 2030 m, 16; 13 km WNW Rio Grende, 2050, 2; 5 mi SE Rio Grende, 1940 m, 2; 50 kmn SW San Tiburcio, 2125 m, 1; 8 mi S Majama, 7700 ft, 10 (KU); 6 km SE Tetillas, 2040 m, 2; 2 mi S Senta Efigenia, 7400 ft, 1 (10.1); 13 km WNW Capriote, 2100 m, 7; 9 thW Sarteneja, 2200 m, 3; Cafnitas, 7 (US); 1mi S Cefiitas, 2 (TN); 1 mi W Sierra Vieja, 6100 ft, 1 (OU); 3 km SE Rencle Grande, 2190 m, 2; 18 km NE Villa de Cos, 2040 m, 4; Villa de Cos, 6700 ft, 18 (KU); 13 mi NNW Fresnillo, l (IAOVI); 2 km S Monte Mariena, 2180 m, 14; 6 km N Fresnillo, 2250 m, 2; 3 mi NW Fresnillo, 7760 ft, 1; Be‘ion, 6400 ft, 11 (LACM); 2miE Be’non, 6400 ft, 1(1AGM), 1 (RGH); 4 km ESE Bafnon, 6200 ft, 20 (0U). Other records: 22 mi SW Concepcién del Oro, (Jones and Webster, 1977); 20 mi S Ooncepcién del Oro, (GAS); 8 mi W Fresnillo, (Jones and Webster, 1977). PERQVIYSCUS HGDPERI LEE AND SCHMIDLY Hooper's Fbuse Distribution. -- Known from only one locality in northeastern Zacatecas. Remarks. -- Very little is kmwn about the natural history of this recently described species . Lee and Schmidly (1977) end Sc‘rmidly (per- sonal communication) suggested that L. hgperi my be restricted in its distribution to tree yucca end sotol habitats . This species is very similar in external characters to _P_. eremicus but can be distinguished by cranial and phallic characters (Lee and Sc‘rmidly, 1977) . The specimens listed below were identified by David Schmnidly. 124 Specimens Exemined. -- Total 7, from: .5 mi S Coahuila-Zacatecas border (Mexican Highway 54). 7 (TCWC)- PEROMYSCUS MANICUIATUS (WAGNER) Deer Mouse Distribution. -- Widely distributed throughout Zacatecas. Remarks. -- This species might be expected to occur anywhere in Zacatecas. It is usually less abundant in any given area then most other species of Feronyscus. Eight females taken between 3 and 30 July contained en average of 3.0 (range, 2 to 4) embryos. Three females taken on 18 July, 3 end 5 August were lactating. The subspecific status of populations in Zacatecas are difficult to determine. In general, specimens from the arid central and north- east parts of the state are light colored and correspond to F_. p. blandus as defined by Osgood (1909) . Specimens from the southwestern mountainous regions are, in general , more darkly colored end correspond to L. I_m_. labecula. However, intergradation, at least in color, seems apparent in specimens from near Chalchihuites , Zacatecas , end in the mountains east of Jalpa. Specimens from the valley of the Rio Juchipila are clearly assignable to the light colored blendus while specimens from the mesa above the Rio Bolai’es south of Monte Escobedo seen to be intermediate in color. In assigning specimens from Zacatecas to subspecies , I have used color end geographic locality as best as possible in conjunction with Osgood's (1909) revision. Speci- mens fram near Zacatecas are here considered to be blendus . Specimens Examined. -- Permscus meniculatus blendus Osgood, total 404, from: 2.5 mi N La Pendencia, 5400 ft, 1 (OU); 3 kmn SE 125 Apizolaya, 1920 m, 2; 7 km SE Caopas, 1940 m, 9; 12 km ENE Concepcién del Oro, 1850 m, 8; 4 mi SE Mazapil, 8425 ft, 5; Concepcidn del Oro, 7600 ft, 1 (I01); 11 mi. E Concepcion del Oro, 5300 ft, 5 (LAO/1); 2 km SE Sabana Grande, 1945 m, 2; 3 mi N Lulu, 2 (MVZ); 1 mi N Lulu, 1830 m, 1; 35 km SSE Concepcion del Oro, 1980 m, 2; 5.5 mi NW Juan Aldama, 6200 ft, 4 (OU); 3 km NW San Felipe de Nuevo Mercurio, 1770 m, 7; 2 km W San Felipe de Nuevo Mercurio, 1740 m, 4; 3 km E San Tiburcio, 1880m, 1; 1 mi SW San Tiburcio, 7000 ft, 18; 15 km WSW Sen Tiburcio, 1980 m, 4; 15 km NNW Nieves, 3; 5 knm S Gonzales Ortega, 2150 m, l; 23 km NE Rio Grande, 1800 m, 3; 15 mi NE Sen Andrés, 6200 ft, 1 (LACM); 13 km WNW Rio Grande, 2050 m, l; 45 km SW Sen Tiburcio, 1940 m, 1; 5 mi SE Rio Grende, 1940 m, 4; 5 km S La Colorada, 1960 m, 3; 6.5 kmn S La Colorada, 1970 m, 2; 8 mi 8 Majcma, 7700 ft, 38; 2 mi E Villa Insurgentes, 8050 ft, 1; 6 km SE Tetillas, 2040 m, 2; 2 mi S Santa Efigenia, 7400 ft, 9 (101) ; 13 km WNW Capriote, 2100 m, 15; 13.6 mi W Sambrerete, 7450 ft, 2 (OU); 5 mi SW Sambrerete, 6900 ft, 3 (LAGM), 2 (RGH); 7 mi. SW Somnbrerete, 6800 ft, 4 (LAGM), 1 (RGH); Cafiitas, 3 (US); 5 kmNE Chalchilnuites, 2360 m, 2; 6 mi WSW Sierra Vieja, 6200 ft, 1 (OU); 23 km N Fresnillo, 2140 m, 2; Villa de Cos, 6700 ft, 6 (K11); 4 kmn E Jimenez del Téul, 2375 m, 1; 2 km S Monte Mariena, 2180 m, 14; Be‘ion, 6400 ft, 21 (LAO/I); 2mi E Bafion, 6400 ft, 1 (RGH); 4 kmESE Bafion, 6200 ft, 10 (OU); 4.5 mi E Fresnillo, 3 (MVZ); 24 mi NE Zacatecas, 6700 ft, 1 (IAGM), 1 (RGH); Zacatecas, 4 (US); 2mi S, 5mi E Zacatecas, 7700 ft, 9; 8 mi SE Zacatecas, 7225 ft, 45 (KU); 9 mi SE Zacatecas, 7900 ft, 1 (10.1);3mi SE Guadalee, 10 (UI); 7 mi SE Trancoso, 2 (KU); 13 km E Jérez, 2200 m, 8; Berriozabal, 6 (US); 6 mi NNW Pinos, 7900 ft, 2; 25 km ESE Fires, 2425 m, 11; 10 mi S Fires, 7100 ft, 2 (RGH); 45 km 126 S Fires, 2350 m, 3; 30 km NE Jalpa, 1740 m, 21; 25 km ESE Jalpa, 2590 m, 20; 4mi S Jalpa, 4300 ft, 2 (LACM); 6 kmn S Apozol, 1170 m, 19. Other records: Jones end Webster (1977) unless otherwise noted; 22 mi SW Concepcion del Oro; 10 mi NW Somnbrerete; 4 mi NW Chalchihuites; 16 mi NW Fresnillo, (GAS); 40 mi W Fresnillo; 9 mi W Zacatecas, (GAS). Peromyscus meniculatus labecula Elliot, total 63, from: 11 mi NE Valparaiso, 7100 ft, 1; Valparaiso mountains, 8700 ft, 1 (US); Valparaiso, 17 (US); 32 km SE Valparaiso, 2040 m, 9; 8 km SW Jérez, 2030 m, 19; 16 km SSE Monte Escobedo, 2010 m, 6; 20 km S Monte Escobedo, 1920 m, 8; Plateado, 1 (US); 1 mi. S Atolinga, 7300 ft, 1. Other records: 8 mi W Milpillas (de la Sierra), (GAS); 1 mi S Menax, (GAS); in Jalisco, Quadrat 55 (Geeways end Jones, 1973). FEROMYSCU S MEIANOPHRYS (COUES) Plateau Mouse Distribution. -- Widely distributed, except for forested montane areas, in Zacatecas. Remarks. -- This species cen be expected to occur anywhere on the Plateau end tropical canyons of Zacatecas. It seems to be most abun- dent in dense brush habitats. Dalquest (1953) end Baker end Greer (1962) reported this species being taken in tree yuccas. I have col- lected them in tree yuccas in Zacatecas but have been most successful in collecting them usirng museum special end Sherman live traps placed in dense brush under repal or in tree yuccas. Fourteen females taken between 10 July and 8 August contained an average of 3.1 (range, 2 to 5) embryos. Four females taken between 6 end 19 October contained an average of 3. 75 (range, 3 to 4) embryos. Lactating individuals have 127 been taken as early as 10 July and as late as 10 October. Live individuals, kept in the enimnal colony at The Museum, MSU, are easy to hendle end breed readily. Baker (1952) recorded only one subspecies from Zacatecas but indi- cated that three others might also be present. Specimens assigned to F. p. consobrinus from the Plateau are larger end darker colored then populations frm the north end rertheast. Specimens from the valley of the Rio Juchipila are somewhat smaller then those from the Plateau, approaching L. _m. micropus in size. These same specimens also seem to be intermediate in color. They are here assigned to consobrinus. Specimens fran the valley of the Rio Atengo are darker, reddish on the dorsum compared to consobrinus. They also tend to be smaller in exterrnal characters. These are assigned to the subspecies micrm. Specimens assigned to L. I_r_n. coahuiliensis from the rertheast are lighter colored than consobrinus. Specimens assigned to L. n_m. xenurus frcm the northern part of Zacatecas are also light colored. These dif- fer fram coahuiliensis by having relatively smnaller auditory bullae. Since Baker's (1952) review of the subspecies of L. melenophrys, meny more specimens have become available. It would be interesting to reevaluate the geographic variation in this species . Specimens Examnined. -- Peromyscus melenophnys coahuiliensis Baker, total 42, fram: 18.6 mi. ENE Concepcion del Oro, 1(OU); 16.4 mi ENE Concepcion del Oro, 1 (OU); 12 kmn ENE Goncepcién del Oro, 1850 m, 7; 7 mi E Mazapil, 1 (UT); 6 km E Mazapil, 2645 m, 6; Concepcién del Oro, 7 (K11); Zmi SE Concepcion del Oro, 1860 m, 1; 5 km sw Concepcién del Oro, 2400 m, l; 10 mi SW Concepcion del Oro, 7600 ft, 3 (IACM); 18 kmn SSW Concepcién del Oro, 2130 m, 4; 40 km ESE Concepcion del Oro, 2320 128 m, 3; 12 mi SE Concepcién del Oro, 2270 m, 1; 2 km SE Sabana Grende, 1945 m, 3; 7 km SE El Rosario, 2100 m, 4; 35 km SSE Concepcion del Oro, 1980 m, l; 3 mi NLulfn, 5 (MVZ). Other records: 22 mi SW Concepcion del Oro, (Jones end Webster, 1977) . Peromyscus meleephgs consobrinus Osgood, total 290, from: 15 km SW Sen Tiburcio, 1980 m, 2; 15 mi. NE Sen Andres, 6200 ft, 3 (LAGM); 25 km SW San Tiburcio, 2030 m, 3; 45 knn SW Sen Tiburcio, 1940 m, l; 8 mi S Majama, 7700 ft, 19 (101); 5 mi SE Rio Grande, 1940 m, 8; 6.5 mi S La Oolorada, 1970 m, l; 6 km SE Tetillas, 2040 m, 2; 2 mi S Senta Efigenia, 7400 ft, 3 (101); 13 km WNW Capriote, 2100 m, 7; 12 mi W Somnbrerete, 2 (K11); 4 mi SW Somnbrerete, 7000 ft, 1 (RGH); 10 m N Rancho Grande, 6700 ft, 4; 15.5 mi WSW Sambrerete, 6400 ft, 2 (OU); 14.2 mi N Jimenez del Téul, 7300 ft, 6 (OU); 1 mi. N Renee Grande, 8 (K11); 6 mi WSW Sierra Vieja, 4 (OU); 23 km N Fresnillo, 2140 m, 2; 13 mi NNW Fresnillo, 3 (LAGM); Villa de Cos, 6700 ft, 1 (K11); 2 km S Monte Mariana, 2180 m, 1; 3 mi NW Fresnillo, 7760 ft, 1; 6 km N Fresnillo, 2250 m, 28; 5 km N Fresnillo, 2250 m, 16; Beion, 6400 ft, 1 (LACM); 2 miEBaf'non, 6400 ft, 19 (LAGM), 2 (RGH); 4kmESEBafen, 6200 ft, 12 (OU); 2.5 km WNW Fresnillo, 7400 ft, 9 (OU); 10 km SE Fresnillo, 2250 m, l; 24 mi NE Zacatecas, 6700 ft, 1 (LAGM); 45 km SW Fresnillo, 2200 m, l; 46 km SWFresnillo, 2150 m, l; 2 mi SE Calera, 7300 ft, 1 (K11); 2 mi S, 5 mi E Zacatecas, 7700 ft, 2; 8 mi SE Zacatecas, 7225 ft, 8 (K11); 9 mi SE Zacatecas, 7900 ft, 1 (K11); 2 mi ESE Trencoso, 7000 ft, 1 (10.1); 3.5 kmn W La Blenca, 6650 ft, 4 (OU); 13 km E Jérez, 2200 m, 2; 8 km SW Jérez, 2030 m, 1; 11.7 mi NW Cuauhtémoc, 7100 ft, 2 (G1); Berriozabal, (type locality),12 (US); 3 mi N Cuauhténec, 6600 ft, 7 (IAQ’I); 5 mi S Ojo Caliente, 4 (UT); 3 mi WNW Saldafna, 6850 ft, 2; 13 mi NE Villenueva 129 (Ruinas Chicomoztoc), 2040 m, 2 (LAW), 1 (RGH); 10 km S Tepetorngo, 1950 m, l; 6 mi NNW Fires, 7900 ft, 5; 10 km ENE Loreto, 7350 ft, 14 (OU); 6 km ENE Ioreto, 6850 ft, 2 (G1); ante Escobedo, 1 (US); 10 mi S Pinos, 7100 ft, 2 (LAW); 25 km ESE Pinos, 2425 m, 9; 20 km S Villenueva, 1810 m, 1; 3.5 mi. W Tlatenengo, 6500 ft, 2; 13 mi N Jalpa, 5000 ft, 1 (101); 10 mi W Jalpa, 6100 ft, 5 (LACM); 3 mi SW Jalpa, 4600 ft, 1 (KU); 6 mi SWJalpa, 4900 ft, 4 (LACM); 4mi S Jalpa, 4300 ft, 5 (IAGD’I); 13 mi WSW Jalpa, 6000 ft, 4 (G1); .5 mi ENE Mesquituta, 3450 ft, 1 (OU); 1 mi N Senta Rosa, 3600 ft, 4; Santa Rosa, 4000 ft, 3 (IACM). Other records: 8 mi W Fresnillo, (Jones end Webster, 1977); 9 mi W Zacatecas, (GAS); 1 mi NE Noria de Angeles, (GAS); 20 mi S Villenueva, (GAS); 8 mi S Moyahua, (Jones end Webster, 1977); in San Luis Potosi, Quadrat 67 (Dalquest, 1953). Peromgyscus meleephrys mnicropus Baker, total 16, from: 18 km N Sen Juen Capistreno, 1100 m, 6; 3 km N San Juen Capistree, 1500 m, 1; 5 km NE Sen Juen Capistrare, 1330 m, 4; 5 km E San Juan Capistree, 3350 ft, 3; 6 km E San Juan Capistree, 3300 ft, 1 (OU); San Juan Capistrano, 1 (US). Peraryscus melenophrys xenurus Osgood, total 35, from: 10 km ESE San Juen de los Charcos, 1500 m, 5; 3 km SE Apizolaya, 1920 m, l; 7 km SE Caopas, 1940 m, l; 8 mi WCedros, 5650 ft, 1 (LACM); 4 kan Sen Rafael, 2140 m, 3; 25 km NE Camacho, 1975 m, 9; 10 km NW Juen Aldema, 1860 m, l; 3 km NW San Felipe de Nuevo Mercurio, 1770 m, l; 10 mi SE Juan Aldama, 7250 ft, 1; 9 mi. NE Nieves, 6600 ft, 1. 130 FERCMYSCUS METANOI‘IS J. A. ALIEN AND CHAPMAN Black-Eared Mouse Distribution. -- Known from only two localities in montene forests of the Sierra Madre Occidental of western Zacatecas. Remarks. -- Very little is known about this mouse in Zacatecas. Baker and Greer (1962) restricted its distribution in Durango to high mesic forests of the Sierra Mach‘e Occidental. Specimens Exemined. -- Total 7, fram: 25 km WSW Milpillas de la Sierra, 2580 m, l; Valparaiso mountains, 8700 ft, 6 (US). PERWYSCUS FECIORALIS OSGGJD White-Ankled Mouse Distribution. -- Grasslends, lower montane forests, and tropical canyons of Zacatecas. Remarks. -- This species seered to be most abundant in rocky habi- tats. The only records of L. pectoralis in the desert region of Zacatecas is in the desert mountain renges . However, Baker and Greer (1962) reported it from desert habitats in Durango. Twelve femnales obtained between 1 July and 7 August contained en average of 3.5 (renge, 2 to 4) embryos. Three females taken on 29 end 30 July were lactatirng. Schmidly (1972) recognized two subspecies in Zacatecas . Specimens assigned to L. p. pectoralis have longer tails, shorter skulls , end smnaller auditory bullae compared to L. p. laceienus. The former occurs in western and southern Zacatecas. Sclnmnidly (1972) noted that speci- mens fromn rertheastern Zacatecas are someahat intermediate between typical pectoralis and laceienus but assigned them to the latter 131 because they were more similar to laceienus. I follow Schmidly's (1972) designations for those populations in the rertheast. In addi- tion, specimens from near Sen Rafael (Cerro de Teyra) seen better considered as laceienus than Ectoralis. Specimens Exemined. -- Peromyscus pectoralis pectoralis Osgood, total 137, fram: 5.5 mi NW Juan Aldamna, 1890 m, 4 (OU); 5 km NW Juan Aldama, 1980 m, l; 10 mi SE Juan Aldama, 7250 ft, 7; 13 kmn WNW Rio Grende, 2050 m, 1; 6 kmn SE Tetillas, 2040 m, 3; 12 mi W Somnbrerete, 4 (10.1); 10 mi N Rencl'e Grende, 6700 ft, 1; 2.5 mi WNW Fresnillo, 7400 ft, 1 (G1); 18 km N Sen Juan Capistree, 1100 m, 26; 3 km N San Juen Capistreno, 1500 m, 2; 5 km NE Sen Juen Capistreno, 1330 m, 2; San Juan Capistreno, 2 (US); 5 km E San Juan Capistrano, 3350 ft, 4 (CU) ; 36 km SSE Valparaiso, 2330 m, 2; 13 km NE Villenueva, 2040 m, 5; 2 km N Noria de Angeles, 2200 m, 1; 3 mi WNW Saldaf'na, 6850 ft, 1 ; Monte Escobedo, 7300 ft, 2 (IAGM), 1 (US); 20 km S Monte Escobedo, 1920 m, 22; 10 mi W Jalpa, 6100 ft, 17 (LAW); 6 mi. SW Jalpa, 4900 ft, 2 (IAW); 4 mi S Jalpa, 4300 ft, 7 (LAW); 13 mi WSW Jalpa, 6000 ft, 10 (OU); 6 km S Téul de Gonzales Ortega, 2010 m, 5; .5 kmn ENE Mesquituta, 3450 ft, 1 (OU); 1 mi N Senta Rosa, 3600 ft, 3. Other records: 3 mi SW Sambrerete, (Jones end Webster, 1977); 2 mi S Villenueva, (GAS); in Jalisco, Quadrat 55 (Schmidly, 1972). Percngyscus pectoralis laceienus Bailey, total 66, from: 4 mi W Concepcion del Oro, 14 (UT); 3 mi W Concepcion del Oro, 2 (UT); 1 mi W Concepcion del Oro, 1 (UI); Concepcién del Oro, 14 (101); 5 km SW Concepcion del Oro, 2400 m, 5; 10 mi SW Concepcion del Oro, 7600 ft, 21 (IAGM); 4 km W San Rafael, 2140 m, 2; 1 mi W Tecolotes, 6150 ft, 1; 4O 132 km ESE Concepcion del Oro, 2320 m, 5; 18 km SW Concepcion del Oro, 2130 m, 1. PERGiYSCUS SPIGIIEGUS J. A. ALLEN Gleening Mouse Distribution. —- Known from only three localities in the cenyon region of southwest Zacatecas. Remarks. -- Little is known about the natural history of this mouse. Baker end Greer (1962) end Carleton (1977) suggested that it is restricted to subtropical habitats directly below the pine-oak forests. The localities listed below include tropical deciduous forests end, at most, the lower portion of pine-oak forests. Specimens from Monte Escobedo and northwest of Yahualica were taken with L. b_oy_lii; lewever, it was ret possible to determine if they were taken in the same or different microhabitats. Carleton (1977) suggests that where these two species are taken in close proximity to one eether, F_. spicilegus occurs in more mesic situations. Specimens Examined. -- Total 5, from: Monte Escobedo, 1 (US); 10 mi NW Yahualica, 7100 ft, 2 (LAGM); 8 mi S Moyahua, 5600 ft, 2 (GAS). PERCMYSCUS TRUEI GENTILIS OSGOOD Pifion Mouse Distribution. -- Grasslends to montene forests in Zacatecas. Remarks. -- The pif'on mouse is usually associated with low montane forests, as the cannon neme implies. In Zacatecas, however, they seen most cannon in grasslend end brush habitats, usually in rocky situa- tions. Baker and Greer (1962) reported simnilar conditions for the pifnon mouse in Durengo. Seven females taken between 9 July and 5 133 August contained an average of 3.0 (renge, 2 to 4) embryos. Seven females captured between 30 June and 5 August were lactating. One of the females taken on 5 August was lactating end contained four embryos. Hoffmeister (1951) recognized one subspecies g. _t_. gentilis in Zacatecas. He indicated that a second subspecies, B. p. gr_6_n__, may also occur in the extrene southern part of the state. All specimens from southern Zacatecas seem best referred to gentilis based on the characters given in Hoffmeister (1951) . Specimens Exemined. -- Total 177, from: 40 km ESE Concepcién del Oro, 2320 m, 18; 5 km S Gonzéles Ortega, 2150 m, l; 2 mi E Villa Insurgentes, 8050 ft, 3; 4 mi E El Calabazal, 1 (NZ); 8 mi W, 1 mi N Sombrerete, 7800 ft, 8 (K11); 12 mi W Sambrerete, 5 (10.1); 10.7 mi WSW Sambrerete, 7250 ft, 10 (G1); 4 mi SW Sambrerete, 7000 ft, 1 (LAW); 15.5 mi WSW Sambrerete, 6400 ft, 2 (OU); 3 km E E1 Arenal, 2460 m, l; 2 mi w Sain Alto, 6900 ft, 6 (I01); 5 km NE Chalchihuites, 2360 m, 2; 4 km E. Jimenez del Téul, 2375 m, 2; 25 km WSW Milpillas de la Sierra, 2580 m, 16; 45 km SW Fresnillo, 2165 m, 5; 46 kmn SW Fresnillo, 2200 m, 1; 53 km SW Fresnillo, 2250 m, 2; 11 mi NE Valparaiso, 7100 ft, 1; Valparaiso, 5 (US); Zacatecas, 1 (US); 15 km SW Valparaiso, 2250 m, 7; 13 km E Jérez, 2200 m, 8; 8 km SW Jérez, 2030 m, 5; 32 kmn SE Valparaiso, 2040 m, 11; 35 km SSE Valparaiso, 2330 m, l; 3 mi N Cuauhtemoc, 6600 ft, 4 (LAGM); 2 (RGH); 5 mi 8 Ojo Caliente, 7 (UT); 13 km NE Villenueva (hfinas Clnicomoztoc), 2040 m, 3, l (IAGM); 6 km ENE Loreto, 6850 ft, 1 (G1); 6 mi NNW Fires, 7900 ft, 5; 25 km ESE Fires, 2425 m, 12; 45 km S Pinos, 2350 m, 1; 8 mi NW Nochistlan, 6600 ft, 18 (LAO/1). Other records: Jones end Webster (1977) unless otherwise 134 noted; 3 mi SW Somnbrerete; 4 mi NW Chalchihuites, (GAS); 8 mi S Chalchihuites; 40 mi W Fresnillo. BAICMYS TAYLORI PAULUS (J. A. ALIEN) Northern Pygmy Mouse Distribution. -- Grasslends to tropical canyons in Zacatecas. Remarks. -- The northern pygmy mouse does not seen abundant eny- where in Zacatecas. As noted in Baker end Greer (1962) , it prefers areas of grass or brush habitats that are relatively free of heavy grazing by livestock. Packard (1960) indicated that this species is only locally abundent. Petersen (1978) reted that this mouse was the most abundent small rodent at Atotonilco, Durengo, only a few kilo- meters from the Zacatecas border. Four females taken between 12 and 24 July contained an average of 1.5 (range, 1 to 2) embryos. However, Packard (1960) recorded an average of 2.48 embryos fran 41 females. All specirmens fram Zacatecas are referable to S. p. @1429; by the characters given in Packard (1960). Specimens Examined. -- Total 67, from: 15 km NNW Nieves, 1910 m, l; 5 km S Gonzales Ortega, 2150 m, l; 16 mi NW Rio Grende, 6750 ft, 1; 13 km WNW Rio Grende, 2050 m, l; 3 mi E E1 Calabazal, 8000 ft, 3; Valparaiso, 6200 ft, 10 (US); 15 km SW Valparaiso, 2250 m, 1; 6 mi. W Jérez, 6700 ft, 2; 13 km E Jérez, 2200 m, 3; 8 km SW Jérez, 2030 m, 3; 32 km SE Valparaiso, 2040 m, 9; 36 kmn SSE Valparaiso, 2330 m, 3; 16 km SSE ante Escobedo, 2010 m, 5; 20 km S Monte Escobedo, 1920 m, 1; 20 km S Villenueva, 1810 m, 2; 2.5 mi. S Momnax, 5800 ft, 1; 3.5 mi W Tlatenengo, 6500 ft, 1; 30 km NE Jalpa, 1740 m, 2; 10 mi. W Jalpa, 6100 ft, 8 (LAW); 5.5 mi SWJalpa, 4400 ft, 1 (G1); 25 kmESE Jalpa, 2590 135 m, l; 8 mi NWNochistlén, 6600 ft, 2 (LAGM); 6 km S Téul de Gonzales Ortega, 2010 m, e; 7 km SE Senta Rosa, 1500 m, 3. Other records: Jones end Webster (1977) unless otherwise reted; 4 mi NW Chalchihuites; 40 km W Fresnillo; 8 mi. S Moyahua; in Dutengo, Quadrat 8 (MSU). ONYCHCMY S TORRIDUS GANUS MERRIAM Southern Grasshopper Mouse Distribution. -- Widely distributed, except for montene regions, in Zacatecas. Remarks. -- This species seems most abundant in the desert-grass- land ecotone of central Zacatecas . Usually, only one or two individuals are taken in eny given locality. However, a field party fram the University of Oklahama was able to collect 10 end 15 specimens from near Sierra Vieja and Bafion, respectively. Both localities are in the desert-grasslend ecotone. A female taken on 31 July contained four embryos are was lactatirng. The females taken on 26 July end 21 August were lactating. All specimens from Zacatecas are referable to the subspecies Q. _t_. gapps based Lpon color end size characters given in Hollister (1914) . Specimens from extrene northeastern Zacatecas approach 9. p. su'rufus in having slightly longer enterior palatine foremina but do not possess the spine on the posterior border of the palate. Ven Cura end Hoffinesiter (1966) end Matson and Friesen (1979) denenstrated that color, length of enterior palatirne foramnirna , and the development of a spine on the palate were of questionable utility in separating northern subspecies of Q. torridus. 136 Specimens Examined. -- Total 83, from: 15 mi NE Concepcion del Oro, 2 (G1); 12 km ENE Concepcion del Oro, 1850 m, l; 35 km SSE Concepcion del Oro, 1980 m, 1; 13.5 mi SW Gamacl'e, 5800 ft, 1 (RGH); 3 km NW Sen Felipe de Nuevo Mercurio, 1770 m, 4; 4 kmn N Nieves, 1980 m, 1; 15 mi NE San Andres, 6200 ft, 3 (LACM); 6.5 km S La Colorada, 1970 m, l; 8 mi S Majama, 7700 ft, 5 (KU); 10 mi N Rencho Grende, 6700 ft, 1; 1 mi S Cafnitas, 1 (TTU); 1 mi W Sierra Vieja, 6100 ft, 10 (G1); 40 km NE Villa de Cos, 2000 m, 1; 18 km NE Villa de Cos, 2040 m, 1; 3 mi E, 4.6 mi N Villa de Cos, 1 (OU); Villa de Cos, 6700 ft, 2 (I01); 2 kmn S ante Mariena, 2180 m, 3; 45 kmNE Morelos jet, 1 (TTU); Bafnon, 6400 ft, 2 (IACM); 2miEBe'ion, 6400 ft, 3 (IAGM'), 2 (RGH); 4kmESE Bafen, 6200 ft, 15 (OU); 10 km SE Fresnillo, 2250 m, l; 24 mi. NE Zacatecas, 6700 ft, 1 (IAQVI), 1 (RGH); Calera, 7300 ft, 3 (101); 3 mi SE Guadalnpe, 2 (UT); 8 mi SE Zacatecas, 7225 ft, 1 (K11); 9 mi SE Zacatecas, 7900 ft, 2 (K11); 2 mi. ESE Trancoso, 7000 ft, 1 (I01); 7 mi SE Trencoso, 1 (I01); 13 km E Jérez, 2200 m, l; Sen Juan Capistree, (type locality), 5 (US); 5 mi S Ojo Caliente, 1 (UT); 3 mi WNW Saldafna, 6850 ft, 1. Otlner records: Jones and Webster (1977) unless otherwise noted; 22 mi SW Concepcion del Oro, (GAS); 20 mi S Concepcién del Oro; 40 mi W Fresnillo; 8 mi W Fresnillo. STGVDDOAN FULVIVENTER FULVIVENTER J. A. ALLEN Tawny-Bellied Cotton Rat Distribution. -- Grasslends of western end southern Zacatecas. Remarks. -- This species does rot seen very abundant enywhere in Zacatecas. Baker (1969) considered this cotton rat to be extrenely sensitive to overgrazing. Its spotty distribution end low numbers in 137 Zacatecas may be a reflection of this sensitivity. One fenele taken on 16 July was lactatirng. Other natural history data for a nearby locality in Durengo were given by Petersen (1978) . Baker (1969) reviewed the fulviventer group of cotton rats and referred all speci- mens from Zacatecas to _S_. f. fulviventer. Specimens Examined. -- Total 22, fram: 5 km S Gonzales Orgega, 2150 m, l; 16 mi NW Rio Grande, 6750 ft, 1; 3 mi E El Calabazal, 8000 ft, 1; 9 mi W Somnbrerete, 7900 ft, 1 (OU); 5 km NE Chalchihuites, 2360 m, 1; 12 mi SE Fresnillo, 7000 ft, 1; 45 km SW Fresnillo, 2165 m, 2; Zacatecas, (type locality), 3 (AMNH); 10.9 mi NW Cuauhténec, 7100 ft, 1 (G1); Monte Escobedo, 7300 ft, 8 (IAO’I); 25 kmn ESE Jalpa, 2590 m, 1; 8 km NW Téul de Gonzéles Ortega, 2200 m, 1. Other records: 67 km W Fresnillo, (Baker, 1969); 13 km S Villenueva, (Dalby end Lillevik, 1969); in Durango, Quadrat 8 (MSU). SIG’DDON HISPIDUS BERLANDIERI BAIRD Hispid Cotton Rat Distribution. -- Widely distributed, except for the montane ferests, in Zacatecas. Remarks. -- In the desert regions of Zacatecas, this species occurs in islands of grasslend or near cultivated fields. It seems to be most abundent in brushy habitats in the grasslands. Three females taken on 10, 15 July and 3 August contained 5, 6, and 11 ennbryos respectively. One female taken on 15 July was lactating. The status of various subspecies of _S_. hispidus are in need of review. All specimens fram Zacatecas are assigned to S. _h. berlendieri on the basis of geography (Hall end Kelson, 1959) . 138 Specimens Exemined. -— Total 90, fran: 10 km ESE San Juen de los Charcos, 1.500 m, 1; 3 km SE Apizolaya, 1920 m, 1; 12 km ENE Concepcion del Oro, 1850 m, l; 5 km NW Juan Aldama, 1980 m, l; 13 mi SW Camacho, 5900 ft, 1 (IAGM); 35 km SSE Concepcion del Oro, 1980 m, 1; 1 mi N 1.1116, 1830 m, l; 2 km W San Felipe de Nuevo Mercurio, 1740 m, 14; 5 km S Gonzales Ortega, 2320 m, 2; 23 km NE Rio Grande, 1800 m, l; 13 km WNW Rio Grende, 2050 m, 1; 2 mi W Sain Alto, 6900 ft, 7 (K11); 6 km N Fresnillo, 2250 m, 8; 5 km N Fresnillo, 2250 m, l; 2.5 km WNW Fresnillo, 7400 ft, 2 (G1); 2 mi S Fresnillo, 7420 ft, 1; 5 mi NW Zacatecas, 7600 ft, 1 (K11); Valparaiso, 6200 ft, 3 (US); 8 mi SE Zacatecas, 7225 ft, 3 (KU); 20 km S Villenueva, 1810 m, 1; 25 km ESE Pinos, 2425 m, l; 2.5 mi S Mrnax, 5800 ft, 1; 30 km NE Jalpa, 1740 m, l; 10 mi W Jalpa, 6100 ft, 8 (IACM); 5.5 mi SW Jalpa, 4400 ft, 6 (OU); 6 mi SW Jalpa, 5100 ft, 1 (IACM); 4 mi S Jalpa, 4300 ft, 10 (IAQ’I); 8 mi NW Nochistlan, 6600 ft, 2 (LAW); Senta Rosa, 4000 ft, 8 (IAW). Other records: 9 mi. W Zacatecas, (GAS); in Jalisco, Quadrat 55 (Genoways and Jones, 1973). SIGNDDON LEUCOI'TS IEUCOI‘IS BAEEY White-Eared Cotton Rat Distribution. -- Montane regions of western end southern Zacatecas. Remarks . —- The white-eared cotton rat is fround only in montene areas usually associated with grassy meadows end adjacent brushy areas inn pine-oak forests. Baker (1939:213) reported that they were most cannon in ". . . mixed grass end shrub cover on shallow, rocky soils, 139 ." Three females takenonSend 6August contained 3, 4, and 6 embryos. Ore female taken on 5 August was lactating. The subspecies of S. leucotis were revieved by Baker (1969) . He assigned all specimens from Zacatecas to S. _l_. leucotis. Specimens Exemined. -- Total 19, fromn: 25 km WSW Milpillas de la Sierra, 2580 m, 14; 9 mi NW Valparaiso, 8350 ft, 2; 9 mi WNW Jalpa, 8250 ft, 3. Other records: Baker (1969); 13 kmn S Chalchihuites; 13 km W Milpillas (de la Sierra); 27 kmn W Milpillas (de la Sierra); Valparaiso mountains; 15 km W Zacatecas; 17 km S Pinos; in Aguascalientes , Quadrat 57 . STGNDDON MASCOI'ENSIS J. A. AILEN Mascota Cotton Rat Disbrituion. -- Tropical cenyons of southwestern Zacatecas . Remarks. -- Very little is known about S. mascotensis. It seems cannon along cultivated fields and grassy areas. Two females taken on 21 end 30 July contained six end eight embryos, respectively. Two females taken on 21 July were lactating. Until recently (Zimmerman, 1970) , S. mascotensis was considered to be a subspecies of _S_. hispidus. These two are difficult to identify on external characters. Hmoever, Zimmermen (1970) listed sane cranial characters that are useful in separatirng the two. In addition, The Musemm, Michigen State University, has a large series of known-age specimens of both species which aided in their identification. In a previous report (Matson et a1. , 1978) , a subadult specimen from near San Juan Capistrano was incorrectly identified as S. arizornae. Now 140 with a larger series from a nearby locality, I believe that specimen is better referred to S. mascotensis. Specimens Examined. -- Total 18, from: 18 km N San Juen Capistreno, 1100 m, 6; 5 kn NE San Juan Capistrare, 1330 m, 1; 16 km SSE b’bnte Escobedo, 2010 m, 7; 6 km S Apozol, 1170 m, l; 2 mi N Senta Rosa, 3850 ft, 2; Senta Rosa, 4000 ft, 1 (LAW). NEOIOMA ALBIGUTA LEUCODON MERRIAM White-Throated Wood Rat Distribution. -- Widely distributed throug‘teut the Plateau region in Zacatecas. Remarks. -- White-throated wood rats seem to be most abundant in brushy end rocky habitats of central Zacatecas . Apparently, they are absent from the valleys of the Rio Atengo end other tributaries of the Rio Bolaf'nos where N. palantina occurs. Hall end Genoways (1970) sug- gested that in areas where the renges of these two species approach one eether, N. albigu_la occurs in uplend situations. Five females taken between 16 and 31 July contained an average of 1.6 (renge, 1 to 2) embryos. Five females taken between 6 and 16 July were lactating. All Zacatecen specimens are assigned to the subspecies N. g. leicodon following the recent review of the species by Hall end Geeways (1970) . Previously, populations in western Zacatecas were con- sidered to be a distirnct subspecies N. _a: zacatecae with the type locality at Plateado (Goldmen, 1910) . As shown by Hall and Geeways (1970) , the characters of zacatecae do not hold when compared to the large series of specimens now available of leucodon. Specimens from 141 rertheastern Zacatecas approach N. g. subsolena (Alvarez, 1962) in size end color but are still referable to leucodon. Specimens Exemined. -- Total 262, from: 6 kmn W Apizolaya, 1800 m, 3; 3 km SE Apizolaya, 1920 m, 1; 7 km SE Caopas, 1940 m, 1; 8 mi W Cedros, 5650 ft, 1 (LAW); 12 ka San Rafael, 2590 m, l; 13 km NE Concepcion del Oro, 1700 m, 2; Concepcion del Oro, 7600 ft, 1 (101); 11 mi E Concepcion del Oro, 5300 ft, 5 (LAW); 40 km ESE Concepcion del Oro, 2320 m, 2; 25 km NE Cemacho, 1975 m, 5; 12 mi SE Concepcién del Oro, 7450 ft, 3; 3mi NLulCn, 2 (MVZ); lmi N Lulu, 6000 ft, 1; 7 kmn SE El Rosario, 2100 m, l; 15 mi S Concepcion del Oro, 1 (K11); 13 mi SW Camacl'e, 5800 ft, 2 (LAW); 10 mi SE Juen Aldamna, 7250 ft, 3; 2 km W San Felipe de Nuevo Mercurio, 1740 m, 16; 3 km E Sen Tiburcio, 1880 m, l; 1 mi SW Sen Tiburcio, 7000 ft, 6 (K11); 15 km WSW Sen Tiburcio, 1980 m, 3; 9 mi NE Nieves, 6600 ft, 2; 4 km N Nieves, 1980 m, 3; 5 km S Gonzales Ortega, 2320 m, l; 23 km NE Rio Grande, 1800 m, l; 25 km SW Sen Tiburcio, 2030 m, l; 15 mi NE Sen Anchés, 6200 ft, 1 (LAW); 13 km WNW Rio Grende, 2050 m, l; 45 km SW San Tiburcio, 1940 m, 2; 5 mi SE Rio Grande, 1940 m, 11; 50 km SW San Tiburcio, 2125 m, 1; 6.5 km S La Colorada, 1970 m, l; 8 mi 8 Majomna, 7700 ft, 6 (101); 4 mi E E1 Calabazal, 1 (NZ); 13 km WNW Capriote, 2100 m, 10; 9 km NW Sarteneja, 2200 m, l; 10 mi N Rancle Grende, 6700 ft, 6; 7 km N Rencho Grande, 2190 m, l; 1 mi W Sierra Vieja, 6100 ft, 13 (OU); 40 km NE Villa de Cos, 2000 m, 2; 6 mi WSW Sierra Vieja, 6200 ft, 1 (OU); 10 mi SW Chalchihuites, 7200 ft, 1 (IACM); 3 mi E, 4.6 mi N Villa de Cos, 3 (OU); 23 km N Fresnillo, 2140 m, 1; Villa de Cos, 6700 ft, 6 (101); 2 km S Monte Mariena, 2180 m, l; 2 mi SE Villa de Cos, 6200 ft, 1; 6 km N Fresnillo, 2250 m, l; 3 mi NW Fresnillo, 7760 ft, 1; 2.5 kmn WNW 142 Fresnillo, 7400 ft, 2 (G1); 20 mi NE Morelos jet, 2 (OU); Bafion, 6400 ft, 18 (LACM); 2miE Baf'non, 6400 ft, 3 (IAIN); 4kmESE Bafion, 6200 ft, 19 (OU); 45 km SW Fresnillo, 2165 m, 1; 7.2 mi W Valparaiso, 7900 ft, 3 (OU); 6 mi W Valparaiso, 7400 ft, 1 (OU); Valparaiso, 6200 ft, 5 (US); 10 mi N Valparaiso, 6200 ft, 1 (K11); Zacatecas, 1 (US); 8 mi SE Zacatecas, 7225 ft, 2 (101); 2 mi. ESE Trancoso, 7000 ft, 7 (K11); 8 mi SW Jérez, 2030 m, 1; 11.7 mi NW Cuauhtémoc, 7100 ft, 2 (G1); Berriozabal, 9 (US); 3miNCuauhtémoc, 6600 ft, 13 (LACM);1(RGH); l3kmNE Villenueva, 2040 m, 2; 3 mi WNW Saldaf‘na, 6850 ft, 1; 6 mi NNW Fires, 7900 ft, 1; Fires, 1 (AMNH); bente Escobedo, 7300 ft, 4 (LAW); 8 mi S Villenueva, 6850 ft, 1; 25 km ESE Pinos, 2425 m, 4; Plateado, 1 (US); 13 mi N Jalpa, 5000 ft, 6 (101); 10 mi W Jalpa, 6100 ft, 1 (LAGM); 3 mi SW Jalpa, 4600 ft, 1 (101); 5.5 mi SW Jalpa, 4400 ft, 2 (OU); 6 mi SW Jalpa, 4900 ft, 1 (RGH); 4 mi S Jalpa, 4300 ft, 3 (LAW). Other records: 22 mi SW Concepcion del Oro, (Jones and Webster, 1977); 16 mi NW Fresnillo, (GAS); 6 mi W Fresnillo, (Jones and Webster, 1977) ; in Durengo, Quadrat 8 (MSU); in Jalisco, Quadrat 55 (Hall and Geeways, 1970); in San Luis Potosi, Quadrat 67 (Dalquest, 1953). NEUIDMGOID’IANIMERRIM FigmyWoodRat Distribution. -- Arid regions of northeastern and southeastern Zacatecas. Remarks. -- In Zacatecas, this species seenns to be associated with rocky habitats. Three females taken on 9, 21 July and 8 August con- tained one, two, end one embryos, respectively. A female taken on 29 July was lactating. 143 The geographic distribution of the species was reviewed by Rainey end Baker (1955) . Specimens reported here represent western and southern mnargirnal records for the species. Specimens Exemined. -- Total 15, from: 20 km NE Concepcion del Oro, 1910 m, 2; 4 kmn W San Rafael, 2140 m, l; 6 mi W Concepcion del Oro, 1(UT); Concepcion del Oro, 7600 ft, 1 (K11); 25 km SW Sen Tiburcio, 2030 m, 2; 50 km SW San Tiburcio, 2125 m, 3; 5 mi S Ojo Caliente, 1 (UT); 3 mi WNW Salde“ia, 6850 ft, 2; 6 mi NNW Fires, 7900 ft, 1; 25 km ESE Fires, 2425 m, 1. NEOI‘Q’IAMIGANA BAIRD MexienWood Rat Distribution. -- Montene regions of western and southern Zacatecas . Remarks. -- In the montene forests of Zacatecas, the Mexicen wood rat seems to be associated with rock outcrops. In the tropical vege- tation along the Rio Juchipila, it was taken in association with rock ledges or fences. Two females taken on 30 July were lactating. Hall (1955) recorded mo subspecies from western end southern Zacatecas . Altleugh he also indicated that a third subspecies may be present in the northeastern part of the state, I failed to find them. Specimens assigned to N. n_m. machensis from the western part of Zacatecas are larger end have a more reddish colored dorsunn than speci- mens fram the south assigned to N. p. tenuicauda. Specimens Exemined. -- Neotcme mexicena madrensis Goldman, total 15, from: 4 mi. E El Calabazal, 3 (NZ); 12 mi W Somnbrerete, 2 (101); 14.2 mi N Jimenez del Téul, 7300 ft, 1 (OU); 25 km WSW Milpillas de la 144 Sierra, 2580 m, 4; 9 mi NWValparaiso, 8350 ft, 5. Other records: 8 mi W Milpillas (de la Sierra), (GAS); Sierra de Valparaiso, (Hall, 1955); in Jalisco, Quadrat 48 (Geeways end Jones, 1973). Neotomna mexicena tenuicauda Merriamn, total 14, fram: Fires, 3 (AMNH); Plateado, 6 (US); 10 mi WJalpa, 6100 ft, 1 (IAW); 8 mi NW Nochistlan, 6600 ft, 1 (LAW); .5 mi ENE Mesquituta, 3450 ft, 2 (OU); Senta Rosa, 4000 ft, 1 (IAW). Other records: In Aguascalientes, Quadrat 57 (Hooper, 1955). NEDTOMA FAIATINA GOLEVIAN Bole‘n'os Wood Rat Distribution. -- Known only from the valley of the Rio Atengo in Zacatecas . Remarks. -- This species was collected only in the cenyon bottom of the Rio Atengo in Zacatecas. Matson et a1. (1978) reported two specimens taken in rocky habitats. Two additiornal specimens from north of Sen Juan Capistree were also taken along rock fences in cenyon bottom. A female taken on 3 August contained one embryo. Specimens Exemined. -- Total 5, from: 18 km N San Juan Capistree, 1100 m, 2; 5 km NE San Juen Capistreno, 1330 m, 2; 5 km E Sen Juan Capistrano, 3350 ft, 1 (OU). Other records: in Jalisco, Quadrat 48 end 61 (Hall end Geeways, 1970). NELSONIA NEOIGDDON NEOTCMODON MERRIAM Diminutive Wood Rat Distribution. -- Montene regions of western and southern Zacatecas . 145 Remarks. -- This is a rare species of rodent for which very little is known. Apparently, it is restricted to rocky situations in mesic montene forests of the Sierra Madre Occidental (Hooper, 1954) . All Zacatecen specimens are assigned to N. p. neotarodon following Hooper (1954). Specimens Exemined. -- Total 11, fram: 9 mi NW Valparaiso, 8350 ft, 1; Valparaiso mountains, 8700 ft, 2 (US); Sierra Madre, 4 (US); Plateado, 4 (US). Other records: in Aguascalientes, Quadrat 57 (Hooper, 1954). LITERATURE CITED LITERATURE CITED Alvarez, T. 1960. Sirepsis de las especies Mexicenos del genero W. Rev. Soc. Mexicena Hist. Nat., 21:391-424. . 1962. A new subspecies of wood rat (Neotama) from north- eastern Mexico. Univ. Kensas Fubl., Mus. Nat. Hist. , 14:139-143. Anderson, S. 1965. Sources of error in locality data. Syst. 2001. , 14:344-346. . 1966. Taxonomy of gophers, especially m, in Chihuahua, Mexico. Syst. 2001., 15:189-198. . 1972. Mammals of Chihuahua: taxonamy end distribution. Amer. Mus. Nat. Hist., Bull., 148:151-410. Armstrong, D. M. 1972. Distribution of mammals in Colorado. Univ. Kansas, Mus. Nat. Hist., Monogr., 3:x-415. Bailey, V. 1915. Revision of the pocket gophers of the genus T_h_om_pt_mzs_. N. Amer. Fauna, 39:1-136. Baker, R. H. 1952. Geographic renge of Peromygcus melenophrys with description of new subspecies. Urniv. Kensas Fubl. , Mus. Nat. Hist. , 5:251-258. . 1953. The pocket gopher (genus TM) of Coahuila, Mexico. Univ. Kensas Fubl., Mus. Nat. Hist., 5:499-514. . 1954. The silky pocket mouse (Perognathus flavus) of Mexico. Univ. Kensas Fubl., Mus. Nat. Hist., 7:339-347. . 1956. Mammnals of Coahuila. Univ. Kensas Fubl., Mus. Nat. Hist., 9:127-335. . 1969. Cotton rats of the Si fulviventer group. Misc. Fubl., Univ. Kensas, Mus. Nat. Hist., :177-232. . 1978. Mammals of the Chihuahuan Desert Region - future prospects. pp. 221—225, _in_, Wauer end Riskind. Baker, R. H. end J. K. Greer. 1962. Menmals of the Mexicen state of Durengo. Michigen St. Univ. Fubl. Mus., Biol. Ser., 2:25-154. 146 147 Baker, R. H., R. C. Webb, and F. Dalby. 1967. Notes on reptiles end mammals frcm southern Zacatecas. Amer. Midl. Nat., 77:223-226. Baroni-Urbeni, C. end M. W. Buser. 1976. Similarity of binary data. Syst. 2001., 25:251-259. Berry, D. L. end R. J. Baker. 1971. Apparent convergence of karyo- types in two species of pocket gophers of the genus M (Mamrmalia, Rodentia). Cytogenetics, 10:1-9. Best, T., J. K. Greer, end F. F. B. Elder. 1972. Two bat records fromn Zacatecas, Mexico. Southwest. Nat., 17:97-98. Brown, J. H. 1971. Mammnals of mountaintops: nonequilibrium insular biogeography. Amer. Nat. , 105:467-478. . 1978. The theory of insular biogeography end the distri- bution of boreal birds and mammals. Great Basin Nat. Mem., 2: 209-227. Caire, W. 1976. Fhenetic relationships of pocket mice in the subgenus Chaetodipus (Rodentia: Hetercmyidae). J. Mem., 57:375-378. Carelton, M. D. 1977. Interrelationships of populations of the Permscus boylii species gronp (Rodentia, Muridae) in western Mexico. Occas. Papers, Mus. Zool., Univ. Michigan, 675:1-47. Cooley, W. W. end P. R. lohnes. 1971. Multivariate data enalysis. John Wiley end Sons, Inc. , New York, 364 pp. Corzo, R. H. 1970. Mexico: avifauna end modification of habitat. pp. 63-69, in, The avifauna of northern Latin America. H. K. Bueclnner and—J. H. Buechner, eds., Smithsonien Contrib. Zool., 26: 1-119. Dalby, F. L. end H. A. Lillevik. 1969. Taxonomic enalysis of electro- phoretic blood serum patterns in the cotton rat , §gro_d_on_. Michigan St. Univ. Fubl. Mus., Biol. Ser., 4:65-1 . Dalquest, W. W. 1951. Six new mammnals fram the state of Sen Iuis Potosi, Mexico. J. Washington Acad. Sci., 41:361-364. . 1953. Mammals of the Mexicen state of San Luis Potosi. Louisiana St. Univ. Studies, Biol. Ser., 1:1-229. Davis, J. C. 1973. Statistics end data enalysis in geology. John Wiley end Sons, Inc. , New York, 550 pp. Dice, L. R. 1943. The biotic provinces of North America. Univ. Michigen Press, Arm Arbor, 78 pp. 148 Diersirng, V. E. 1976. An enalysis of Feromyscus difficilus from the mgezggited States boundary area. Proc . Biol. Soc . Washington, Findley, J. S. end W. Caire. 1978. The status of mammals in the northern region of the Chihuahuen Desert. pp. 127-139, ;1I__1_, Wauer end Riskind. Fisher, D. R. 1968. A study of femal resemblence usirng numerical taxonomy end factor enalysis. Syst. 2001. , 17:48-63. . 1973. A comparison of various techniques of multiple factor analysis applied to biosystematic data. Univ. Kensas, Sci. Bull. , 50:127-162. Genoways, H. H. 1973. Systematics end evolutionary relationships of spiny pocket mice, genus LignEs. Spec. Fubl. , The Museum, Texas Tech Univ. , 5:1-368. Genoways, H. H. end J. K. Jones, Jr. 1968. Notes on bats from the Mexicen state of Zacatecas. J. Mamm. , 49:743-745. . 1971. Systematics of the southern banner-tailed kangaroo rats of the Dipodcrrys phillipsii group. J. Memm. , 52:265-287. . 1973. Notes on some mammals from Jalisco, Mexico. Occas. Papers Mus. , Texas Tech Univ., 9:1-22. Gentry, H. H. 1957. 103 pastizales de Durango. Inst. Mexicee Recursos Nat. Renov., Mexico, D. F., 361 pp. Getz, L. L., F. R. Cole, and D. L. Gates. 1978. Interstate roadsides as dispersal routes for Microtus pemnsylvanicus. J. Mamm. , 59: 208-212. Goldman, E. A. 1910. Revision of the wood rats of the genus Neotoma N. Amer. Fauna, 31:1-124. . 1951. Biological irnvestigations in Mexico. Smithsonian Misc. Coll., 115zxiii-476. Goldman, E. A. end R. T. Moore. 1946. The biotic provinces of Mexico. J. Mem., 26:347-360. Goodall, D. W. 1973. Numerical classification. pp. 575-615, in, Ordination end classification of communities. R. H. Whittaker, ed., Dr. W. Junk, Fubl., The Hague, 737 pp. Hagmeier, E. M. 1966. A numerical enalysis of the distributional patterns of North American mammals. II. Re-evaluation of the provinces. Syst. 2001., 15:279-299. 149 Hagneier, E. M. and C. D. Stults. 1964. A numerical enalysis of the distgibutional patterns of North Americen mammals. Syst. Zool., 13 1 5 155 Hall, E. R. 1955. A new subspecies of wood rat from Nayarit, Mexico with new name-combinations for tie Neotama mexicena gronp. J. Washington Acad. Sci. , 45:328-332. Hall, E. R. end H. H. Genoways. 1970. Taxonomy of the Neotoma albigpla group of woodrats in Central Medco. J. Memm, 51: 504-516. Hall, E. R. and K. R. Kelson. 1959. The mammals of North America. The Ronald Press Go. , New York, 2 vols. , 1083 pp. Harris, R. J. 1975. A primer of multivariate statistics. Academic Press, New York, 332 pp. Henrickson, J. 1978 . Saline habitats and halophytic vegetation of the Chihuahuan Desert Region. pp. 289-314, _ip, Wauer end Riskind. Hershkovitz, F. 1958. A geographical classification of Neotropical mermals. Fieldiena, 36:583-620. Hoffmeister, D. F. 1951. A taxonamic and evolutionary study of the pifion mouse, Peromyscus truei. Illireis Biol. lenogr., 21:1-104. . 1969. The species problem in the T1namcnmys bottae-Thememys umbrinus canplex of gophers in Arizorna. Misc. Fubl., Univ. Kensas, Mus. Nat. Hist, 51:75-91. Hoffimeister, D. F. end L. de la Torre. 1961. Geographic variation in the mouse, Feramgscus difficilis. J. Mamm., 42:1-13. Hoffmeister, D. F. end M. R. Lee. 1967. Revision of the pocket mice, Perognathus penicillatus. J. Memm. , 48:361-380. Hollister, N. 1914. A systematic account of the grasshopper mice. Proc. U. S. Nat. Mus., 47:427-489. Hooper, E. T. 1952. A systematic review of the harvest mice (genus Reithrodontamys) of latin America. Misc. Fubl. , Mus. Zool. , Univ. Michigan, 77:1-255. . 1954. A syrepsis of the cricetine rodent genus Nelsonia. Occas. Papers Mus. Zool., Univ. Michigan, 558:1-12. . 1955. Notes on mammals of western Mexico. Occas. Papers Mus. Zool., Univ. Michigen, 565:1-26. Howell, A. H. 1938. Revision of the North American ground squirrels. N. Amer. Fauna, 56:1-256. 150 Huey, 18.. M. 1941. Mammalien invasion via the highway. J. Memm, 22: 3 3-385. Johnson, N. K. 1974. Montene avifaunas of southern Nevada: his- torical chenge in species composition. Condor, 76:334-337. . 1975. Controls of number of bird species on montene islands in the Great Basin. Evolution, 29:545-567. Johnston, M. C. 1978. Brief resure of botenical, including vegeta- tiornal, features of the Chihuahuan Desert Region with special emphasis on their uniqueness. pp. 335-359, _i_n_, Wauer and Riskind. Jones, G. and D. J. Webster. 1977. Notes on distribution, habitat end abundence of sure smnall mammals of Zacatecas, Mexico. An. Inst. Biol., Univ. Anton. Mexico, 47:75-83. Lee, M. R. end D. F. Hoffireister. 1963. Status of certain fox squir- rels in Mexico and Arizona. Proc. Biol. Soc. Washington, 76: 181-189. Lee, M. R. end D. J. Schmidly. 1977. A new species of Per scus (Rodentia: Muridae) from Coahuila, Mexico. J. Mem., 5 : -268. Leopold, A. S. 1950. Vegetation zones of Mexico. Ecology, 31: 507-518. . 1959. Wildlife of Mexico. Univ. California Press, Berkeley, 568 pp. Lidicker, W. Z. , Jr. 1960. An analysis of intraspecific variation in the kengaroo rat, Dippdomys merriemi. Univ. California Fubl. Zool. , 67:125-218. long, C. A. 1963. Mathematical formulas expressing faunal resem- blance. Trens. Kensas Acad. Sci., 66:138-140. MacArthur, R. H. and E. O. Wilson. 1967. The theory of island bio- geography. Princeton Univ. Fress, Princeton, 203 pp. Mata, G. F., J. J. Iopez, X. M. Senchez, F. M. Ruiz, end F. T. Takaki. 1971. Tipos de vegetacion de la Republica Mexicena. Direc. Agrologia, Sec. Recur. Hidraul. , Mexico, 59 pp. Matson, J. O. 1977. Records of mammals from Zacatecas, Mexico. J. Martin , 58:110. Matson, J. O., R. H. Baker, end J. K. Greer. 1978. New records of mammals in the state of Zacatecas, Mexico. Southwest Nat. , 23: 154-156. 151 Matson, J. 0. end D. F. Friesen. 1979. The subspecific status of On c s torridus clarus Hollister 1913 (Rodentia: Cricetidae). . So. California Acad. Sci., 77:116-123. Matson, J. 0. end D. R. Patten. 1975. Notes on same bats from the state of Zacatecas, Mexico. Contrib. Sci. , Ios Angeles Nat. Hist. Mus. , 263-1-12. Mayr, E. 1971. Metl'eds and strategies in taxonomic research. Syst. Zool., 20:426-433. McIntosh, R. P. 1973. Matrix end plexus teclnniques. pp. 157-191, _i_n_, Ordination end classification of ccmmunities. R. H. Whittaker, ed., Dr. W. Junk, Fubl., The Hague, 737 pp. Merriam, C. H. 1898. Life zones end crop zones of the United States. Bull. U. S. Biol. Surv., 10:1-79. . 1901. Descriptions of twenty-three new pocket gophers of the genus M. Proc. Biol. Soc. Washington, 14:107-117. Morafka, D. J. 1977. A biogeographical enalysis of the Chihuahuen Desert through its herpetofauna. Biogeographica, 9 :1-313 . . 1978. Is there a Chihuahuen Desert? A quentitative evalu- ation through a herpetofaunal perspective. pp. 437-454, Ln_, Wauer end Riskind. Murray, K. F. 1968. Distribution of North Americen mammals. Syst. Zool., 17:99-102. Musser, G. G. 1968. A systematic study of the Mexican end Guatemalen gray squirrel, Sciurus aureogaster F. Cuvier (Rodentia: Sciu'idae). Misc. Fubl., Mus. Zool., Univ. Michigen, 137:1-112. Nelson, E. W. and E. A. Goldmnen. 1934. Pocket gophers of the genus 111-5% of Mexican mainland end bordering territory. J . Memm. , :105-124. Nie, N. H., C. H. Hull, .1. G. Jenkins, K. Steinbrernner, end D. H. Bent, eds. 1975. Statistical package for the social sciences. 2nd Ed. , McGraw-Hill Book Co. , New York, 675 pp. Orloci, L. 1973. Ordirnation by resennblence matrices. pp. 249-286, in, Ordination end classification of canmmnities. R. H. Whittaker, ed., Dr. W. Junk, Fubl., The Hague, 737 pp. Osgood, W. H. 1900. Revision of the pocket mice of the genus Perognathus. N. Amer. Fauna, 18:1-72. . 1909. Revision of the mice of the American genus Peromyscus. N. Amer. Fauna, 28:1-285. 152 Packard, R. L. 1960. Speciation end evolution of tie pigmy mice, genus Baig_mys. Univ. Kensas Fubl., Mus. Nat. Hist., 9:579-670. . 1978. Mammals of the southern Chihuahuan Desert: an inventory. pp. 141-153, _i_n_, Wauer end Riskind. Patton, J. L. 1973. An enalysis of natural hybridization between the pocket gophers, Thcmcmys bottae end Thcmcmys umbrinus, in Arizona. . ,5:4 561- 584. Petersen, M. K. 1973. Interactions between the cotton rats, SiMon fulviventer end S. hispidus. Amer. Midl. Nat., 90:319-33 . . Rodent ecology end natural history observations on the men- mals of Atotornilco de Campa, Durango, Mexico. Carter Press, Inc. , end Iowa St. Univ. C0py Center, Ames, 129 pp. Pizzimenti, J. J. 1975. Evolution of the prairie dog genus Ms. Occas. Papers, Mus. Nat. Hist., Univ. Kansas, 39: 1- 73. Poole, R. W. 1971. The use of factor enalysis in modeling natural camenities of plents end enimals. Illinois Nat. Hist. Surv., Biol. Notes, 72:1-14. Rainey, D. G. and R. H. Baker. 1955. The pigmy woodrat, Neotome goldmeni, its distribution and systematic position.Un1v._—'——_Kansas Fubl., Mus. Nat. Hist., 7: 619-624. Raisz, E. 1964. Landforms of Mexico. Office Naval Research, Geog. Branch, Cambridge, Massachusetts. Rohlf, F. J. 1970. Adaptive hierarchical clustering schemes. Syst. Zool., 19:58-82. Russell, R. J. 1968. Revision of pocket gophers of the genus Pappogeomys. Univ. Kensas Fubl., Mus. Nat. Hist., 16:581-776. Ryan, R. M. 1963. The biotic provinces of Central America. Acta. 2001. Mexicena, 6:1-54. Rzedowski, J. 1957. Vegetacion de las partes aridas de los estados de Sen Iuis Potosi y Zacatecas. Rev. Soc. Mexico Hist. Nat. , 18: 49-101. Rzedowski, .1. end R. McVaugh. 1966. La vegetacion de Nueva Galicia. Contrib. Univ. Michigan Herb. , 9:1-123. Schmidly, D. J. 1972. Geographic variation in the white-enkled mouse, Percmyscus pectoralis. Southwest. Nat. , 17:113-138. . 1978. Factors governirng the distribution of mammals in the W Desert Region. pp. 163- 192, _i_n_, Wener end Riskind. 153 Schnell, G. D., P. G. Risser, end J. F. Helsel. 1977. Factor enalysis of tree distribution patterns in Oklahana. Ecology, 58:1345-1355. Setzer, H. W. 1949. Subspeciationn in the kengaroo rat, ms ordii. Univ. Kensas Publ., Mus. Nat. Hist., 1:473-5 . Smith, G. R. end D. R. Fisher. 1970. Factor enalysis of distribution patterns of Kensas fishes. pp. 259-277, _i_n_, Pleistocene end Recent environments of the Central Great Plains. W. Dort,Jr. end J. K. Jones, Jr., eds., Univ. Kansas, Dept. Geol., Spec. Publ., 433 pp. Sneath, P. H. H. end R. R. Sokal. 1973. Numerical taxonomy. W. H. Freeman end Co. , Sen Francisco, 573 pp. Sokal, R. R. end F. J. Rohlf. 1970. The intelligent ignoramus, en experiment in numerical taxonomy. Taxon, 19:305-319. Sokal, R. R. end P. H. H. Sneath. 1963. Principles of numerical taxonamy. W. H. Freeman end Co., Sen Frencisco, 359 pp. Soto, C. end E. Jauregui. 1965. Isotermas extremes e indice de aridez en la Republica Mexicana. Univ. Nac. Auton. Mexico, Inst. Geog., 119 pp. Stevenson, M. M., G. D. Schnell, and R. Black. 1974. Factor analysis of fish distribution patterns in western and central Oklahoma. Syst. Zool., 23:202-218. Tinker, B. 1978. Mexicen wilderness end wildlife. Univ. Texas Press, Austin, 131 pp. Udvardy, M. D. F. 1969 . Dynamic zoogeography with special reference to lend enimals. Ven Nostrand, Reinhold Co. , Nev York, 445 pp. Ven Cura, N. J. end D. F. Hoffinneister. 1966. A taxonomic review of the grasshopper mice, m, in Arizona. J. Memm, 47: 613-630. Wauer, R. H. and D. H. Riskind, eds. 1978. Symposium on the biologi- cal resources of the Chihuahuen Desert Region. Nat. Park Serv. Trens. and Proc., 3:1-658. Webb, W. L. 1950. Biogeographic regions of Texas end Oklahoma. Ecology, 31:426-433. Wellhausen, E. J. 1976. The agriculture of Mexico. Sci. Amer. , 235(3) : 128-150 . Wells, P. V. 1978. Post-glacial origin of the present Chihuahuen Desert less then 11,500 years ago. pp. 67—83, in, Wauer end Riskind. 154 Wernstedt, F. L. 1972. World climate data. Climatic Data Press, P. 0. Boo: 413, Lemnt, Pennsylvenia, 522 pp. Zimmerman, E. G. 1970. Karyology, systematics and chrcmosomal evolu- tion in the rodent genus Sim. Michigan St. Univ. Publ. Mus., Biol. Ser., 4:385-454.