"_Iv wv—v..__lw— w _ ‘ - - u - A PRELEMINARY STUDY OF THE BRUSH mouse, formyscgg m m A CANYON or was SIERRA MADRE OCCIDENTAL m DURANGO, Maxaco Thais For tho Door» oi M. 5. MlCHiGAN STATE UNIVERSITY James iosoph Drake 1958 'L'HtlLfi A PRELIMINARY STUDY OF THE BRUSH MOUSE, Peromyscus boylii, IN A CANYON OF THE SIERRA MADRE OCCIDENTAL IN DURANGO, MEXICO BY ‘ JAMES JOSEPH DRAKE AN ABSTRACT Submitted to the College of Science and Arts, Michigan State: University of Agriculture and Applied Science in partial fulfillment of the requirements for the degree of. MASTER OF SCIENCE Department of Zoology 1958 Approved W ABSTRACT In the summer of 1957, an expedition from Michigan State University spent three weeks at a ranch at 8350 feet in the Sierra Madre Occidental of Durango. While at this ranch, a quadrat was marked off in the head of a mmall canyon for the purpose of studying Peromyscus. The objectives in doing this were (1) to determine which species were in the canyon, (2) to determine their population densities, (3) to determine their home ranges and (4) to gather information on their life histories. A des- cription of the physical features of the canyon is given as well as lists of its conspicuous flora and vertebrate fauna. The methods used consisted of live-trapping in the canyon for nine days to catch, mark, release, and recapture as many mice as possible to obtain home range data, then snap-trapping for three days in an effort to determine the population and obtain Specimens for the Museum. Sherman live-traps and.museum special snap-traps were used, all baited with rolled oats. Small mammals trapped included the brush mouse, Peromyscus bqylii rowleyi (J. A. Allen), the big-cared rock mouse, Peromyscus difficilis difficilis (J. A. Allen), the pifion mouse, Peromyscus truei gentilis Osgood, the western harvest mouse, Reithrodontomys mggalotis zacatecae Merriam, and the Mexican wood rat, Neotoma mexicana madrensis Goldman. It was found that Peromyscus boylii was the most abundant species of Peromyscus in the canyon. Two methods were used to compute the popu- lation density of this species in an area of 4.94 acres. The Lincoln Index.method gave estimates of 30.2 for the males, 23.4 for the females, and 54.0 for the total population, while a removal trapping method des- cribed by Hayne gave estimates of 24.4 for the males, 20.4 for the females, and 44.5 for the total population of this area. Analysis of the estimates obtained from these two methods gave an average of 6.0 males, 4.8 females, or 10.8 Peromyscus boylii per acre. This may be expressed as a biomass of 240 grams per acre for the species. 0f 33 individuals marked 27 percent were transient of which 77 percent were males. Sex-ratio estimates ranging from 1.14 to 1.36 males per female were obtained, the midpoint being 1.25. The average maximum movement of these mice was 43.1 yards for the males and 31.7 yards for the females. The method of calculating home range was primarily that of Haugen. This method gave a home range of from .03 to .16 with an average of .142 acres for the males, and from .06 to .16 with an average of .110 acres for the females. Because of the infrequency of captures at the periphery of the canyon, it is felt that in this area Pcromyscus boylii may be restricted to, or at least prefers, the inclined, more broken, and rocky type of habitat within the canyon to the flatter, less broken environ- ment of the surrounding area. The two other species of Peromyscus, g. difficilis and g. truei, were found along the rim area of the canyon and did not extend their home ranges into the canyon proper. Few captures were made and little information was obtained concerning them. A PRELIMINARY STUDY OF THE BRUSH MOUSE, Peromyscus boylii, IN A CANYON OF THE SIERRA MADRE OCCIDENTAL IN DURANGO, MEXICO BY JAMES JOSEPH DRE I A THESIS Submitted to the College of Science and Arts, Michigan State University of Agriculture and Applied Science in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Zoology 1958 CONTENTS Acknowledgement Introduction Description of the area Location Physical features Vegetation Vertebrates recorded from the area ‘Mammals Birds Amphibians and reptiles Discussion of methods Findings concerning Peromyscus boylii POpulation density Sex ratio Movement Home range Breeding Other speCies of Peromyscus Summary Appendix (Measurements of Peromyscus) Literature cited 10 13 14 15 17 18 23 24 31 32 34 46 49 51 54 ACKE-JOULED EI'IENT The Mexican expedition, during which the data for this report were accumulated, was planned and headed by Dr. Rollin H. Baker, Professor of Zoology and Director of the Museum at Michigan State University, and financed, in part, by a grant from the Michigan State University Develop- ment Fund. Dr. Baker helped plan the study and gave his able assist- ance whenever it was needed. I also wish to thank other members of the expedition for their help in running traps and collecting data in the area. They are: Leslie C. Drew, John K. Greer, Henry L. Short, and Robert G. Webb. Dr. John H. Beaman of the Department of Botany and Plant Pathology, Michigan State University, generously gave his time to the task of identifying the plants collected in the areas studied. Also, I wish to express my appreciation for the hospitality we received from Sefiors Fidel and Emilio Gutierrez, the owners of Rancho Las Margaritas, where the data for this paper were gathered. Sefior Joel Gonzales, manager of the ranch, was most helpful and contributed to the success of this study by providing information not otherwise available. For help in preparation of the manuscript I am gratEful to Dr. Baker, Dr. Roland L. Fischer and Dr. George J. Wallace. INTRODUCTION The field work for this paper was done while on an expedition to Mexico from the Museum at Michigan State University. The primary purpose of the expedition was to study and collect vertebrate animals in little known parts of the Mexican state of Durango. The expedition was in Mexico almost seven weeks from June 24 to August 10, 1957. Six~ teen days of this period, from June 27 to July 13, were spent at a camp in Ciénega de las Burras, a canyon on a ranch belonging to Sefior Fidel Gutierrez, Jr. The ranch, Rancho Las Margaritas, is in a remote area which is 56 miles south of the highway between Durango and Zacatecas on the east slope of the Sierra Madre Occidental. More than five hours were re- quired to make the trip from the highway in a passenger car on an unimproved road. The primary work in this study was the live-trapping and snap- trapping of the small rodents in the canyon. The objectives were (1) to determine which species of mice of the genus Peromyscus inhabited the canyon, (2) to determine the size of the population of these mice 'present, (3) to live-trap, release and recapture as much of the rodent population as possible in an effort to determine home ranges of individ- ual mice, and (4) to gather information on the distribution and life history of the species present. The first objective was met insofar as it was determined that the brush mouse, Peromyscus boylii rowleyi (J. A. Allen), the big-cared rock mouse, g. difficilis difficilis (J. A. Allen), and the pifion mouse, g. truei gentilis Osgood, inhabit the canyon at least during the months of June and July. The second objective was met in that a fairly reliable estimate of the population of P. Egylii was obtained. This species was pre- dominant in the canyon. Peromyscus difficilis and g. truei were .o—uo— comparatively minor Species in the canyon prOper and, as a result, too small a sample of these species was obtained to give reliable population estimates for them. The third objective suffered because of an insufficient number of recaptures. The home range of P. Egylii_is nevertheless comparable to the home ranges of other species of Peromyscus calculated by other workers and will provide useful information when further investigations are made. The home ranges of g. difficilis and g. truei have not been estimated. As to the fourth objective, some information concerning the life history of these small mammals was obtained. DESCRIPTION OF THE AREA recess The Pancho Las Margaritas is located 28 miles south and 17 miles west of Vicente Guerrero, Durango, or more precisely at 23° 19' north latitude and 1040 17' west longitude. This location lies in the Sierra Madre Occidental Biotic Province of Goldman and Moore (1945). This biotic province includes the western, or mountainous portion of Durango, western Zacatecas to the south, and western Chihuahua and eastern Sonora to the north. The province, as a whole, has a rather dry climate although rains occur frequently in the afternoon and evening during the summer months. The higher elevations are rugged, dissected by steep canyons and covered with a mixture of deciduous and coniferous forests. The area in which the canyon is located con- tains a montane zeric forest of the Transition Life Zone. Physical Features The area studied was the upper three—fifths of a short canyon, which drained intermittently to the southeast, near the headwaters of the Rio Chapalagana, into the Rio Grande de Santiagao river system. Boundary markers were placed so that the head and the first 250 yards of the canyon were contained within the study area. The study area encompassed approximately 32,800 square yards or about 6.8 acres. The elevation at the head of the canyon is approximately 8,400 feet. Two hundred and fifty yards down stream from the head of the canyon the elevation of the northeast rim is 25 feet higher than at the head of the canyon while the floor is 102 feet lower. This difference in elevation gives a local relief of 127 feet along the southeastern border of the area where the canyon is about 450 feet from rim to rim. Figure l is a map with 20-foot contours of the area described. The term "canyon" will henceforth be used to refer only to that portion of the canyon described which is the area included in this study. Plates I, II and III illustrate typical scenes within the area. It was estimated that about 33 percent of the total ground area in the canyon was covered by bare rock (see Figure 2). The rock is a rhyolite porphyry which is a fine grained extrusive equivalent of granite. It is composed of feldspar, quartz, and hornblend and showed no signs of having been metamorphosed. It is almost certainly of Cenozoic origin, probably early Tertiary. Rock is present in the canyon in the form of vertical outcrops, weathered horizontal slopes, huge boulders, stair-like shelves up the sides of the canyon, small domes and litter ranging in size down to talus piled below large over- hanging walls. In many areas the rock is fissured where faults have occurred. These faults provide recesses and runways where many animals were observed to go when released from traps. The Mexican wood rat, Neotoma mexicana madrensis Goldman, was taken only in areas of large rock outcrops in which there were large fissures. Approximately 100 yards down from the head of the canyon was a spring flowing out from under a rock shelf, the flow continued as a small stream down the remainder of the canyon. About 50 yards below the origin of the stream and at the base of a vertical rock cliff was a small bog SUpplied with moisture by seepage from the cliff. This bog eventually drained into the stream. The bog, 60 feet long and 15 to 20 feet wide, contained several small pools of open water and sap- ported clumps of moss. The stream, at the point where the bog entered, was 2 to 8 inches deep and from 3 to 6 feet wide. It received no q . I ‘ , , . 4 o .. , ‘ I ‘ ’ , ' ' ' . . , , u , ’ o , . ,- . v ”A _. .. _ . I J . ,A . i . n .. , . ' 7 . . ‘ a. , . v n . — - ‘ 7 , ' . , , - , v , k . , . 4 . 4 , I ,.. . , . , c. _ . _ , , .. V a. ‘4 . , o u , , , A -., - , . .a ‘ >7 a , I 0 210 4}) L L _l__1 I .1 l l at ale in yards Fig. 1. A 20-foot contour map of the canyon. Plate 1. a. The bottom and southwest side about 200 feet below the head of the canyon. b. Up and along the southwest side about 200 feet below the head of the canyon. Plate 11. a. The bottom about 200 feet below the head of the canyon. b. The northeast rim and side about 150 feet below the head of the canyon. Plate III. a. The southwest side about 700 feet below the head of the canyon. b. Looking across the canyon from the southwest rim about 600 feet below the head of the canyon. 20 .//’A ,/ ,/ / / / 3o / / I” //’ | // 40 \, ,/ O 20 40 t" 1‘— :L—f": “I"‘1“_j scale n1 )lnwls Fig. 2. A map showing the percentage of rock along the sides of the canyon. 10 additional water, so did not increase in size while in the canyon. There were several small waterfalls along the course of the stream which created pools as deep as 2 feet; in one of these a larva of a salamander, Ambystoma gp., was observed. Small unidentified fishes also were present. The canyon seemed to be well supplied with moisture from frequent rains. Rain fell either in the afternoon or evening on 8 of the 16 days that our expedition was in the area. Vegetation Almost all of the remaining 67 percent of the surface area was covered by vegetation. The canyoncontained six major plant-types: (l) deciduous trees, (2) pines, (3) bushes, or that which was three feet or more in height, (4) brush, that which was less than three feet in height, (5) grasses, and (6) cacti. The deciduous trees dominated 13 percent of the area, pine 9 percent, bushes 3 percent, brush 10 percent, open grasses 30 percent and cactus less than 1 percent. Approximately one-quarter of the ground in the area dominated by pine was covered by a mat of pine needles and contained no understory vegetational layer. The other three-quarters, as well as most of the area dominated by deciduous trees, had an understory, which usually consisted of grasses. Grasses were also found growing out of crevices in rocks and occurred sporadically over much of the 33 percent which was principally a rock surface. Figure 3 shows the distribution of the plant-types in the canyon. Table 1 includes a partial list of the conspicuous species of plants collected from the canyon. Approximately one percent of the surface area was bare ground scattered with gravel or was covered with water in the bottom of the canyon. /// V v.1“ . \ L L_, L r \ .fi‘- - m \\\ / &/ L, " \ ' , Z :77 \‘ *7 y ’5 [L [/77 \x I l g’ I 1 4' (9/; \‘ssis. \ér‘f‘fivi iiflei/‘l Ill. \ trees O 20 4O Wi’sm‘rd scale in yards Fig. 3. A map showing the distribution of the plant-types in the canyon. Grasses are throughout the entire area. 11 TABLE 1 List of conspicuous plants found in canyon l Arbutus macrophylla Hart. & Gal. Arctostaphylos_pungcns H.B.K. Caenothus buxifolius Willd. Cologania humifusa Hemsl. Eryngium so. Geranium lozani Rose Helianthemum glorneratum Lag. Opuntia_§p. -—- (observed but not collected) Oxalis decophylla H.B.K. Oxalis hernandezi Moc & Sesse. Pingincula caudata Schlecht. Pinus leiophylla Schlecht. & Cham. Pinus ponderosa Dcugl. Pherotrichis balbesii (Decne.) A. Gray Phoradendron bolleanum Eich. guercus arizonica Sarg. Ranunculus macranthus Scheele Sisyrinchium Sp. Str0phostyles ER: Talinum greenmanii Harshberger 1 These specimens have been deposited in the herbarium of Michigan State University. l3 Vertebrates Recorded From The Area Mammals, birds, reptiles, and amphibians which were either pre- served as museum specimens or seen in the vicinity of the canyon are listed in Tables 2, 3, and 4. Names used in these tables follow Miller and Kellogg (1955) for mammals, Friedmann, Griscom, and Moore (1950) and Miller, Friedmann, Griscom, and Moore (1957) for birds, and Smith and Taylor (1945 and 1950) for reptiles and Smith and Taylor (1948) for amphibians. 14 List of mammals from Bancho Las Margaritas, Durango Didelphis marsuaialis mesamericana Oken Myotis velifer velifer (J.A.Allen) Eptesicus fuscus pallidus Young Lepus callotis Wagler Sylvilagus floridanus holzneri (Mearns) Citellus variegatus rupestris J.A.A11en Eutamias bulleri durangae J.A.A11en Sciurus nayaritensis J.A.A11en Thomomys umbrinus sheldoni V.Bailey Reithrodontomys megalotis zacatecae Merriam Peromyscus boylii rowleyi (J.A.Allen) Peromyscus difficilis difficilis (J.A.A11en) Peromyscus truei gentilis Osgood Sigmodon leucotis V. Bailey Neotoma mexicana madrensis Goldman Nelsonia neotomodon Merriam Canis latrans impavidus J.A.A11en Canis lupus baileyi Nelson and Goldman Urocyon cinereoargenteus colimensis Goldman Ursus americanus machetes (Elliot) Procyon lotor mexicanus Baird Mephitis macroura milleri Mearns Conepatus mesoleucus sonoriensis Merriam Felis concolor gzteca Merriam Lynx rufus escuingpae J.A.Allen Odocoileus virginianus covesi (Coves and Yarrow) Opossum Cave Myotis Big Brown Bat White-sided Jackrabbit Eastern Cottontail Rock Squirrel Buller Chipmunk Nayarit Squirrel Southern Pocket Gopher Western Harvest Mouse Brush Mouse Big-eared Rock Mouse Pifion Mouse White-eared Cotton Rat Mexican Wood Rat Dwarf Wood Rat Coyote Gray Wolf Gray Fox Black Bear Raccoon Hooded Skunk Hog—nosed Skunk Mountain Lion Bobcat White-tailed Deer TMEE3 15 List of birds from Rancho Las Margaritas, Durango Cathartes aura teter Fricdmann _Cyrtonyx montezumae montezumae (Vigors) Ieleagris gallopavo mexicana Gould Charadrius vociferus vociferus Linnaeus Columba fasciata fasciata Say Ara militaris mexicana Ridgway Aratinga.§p. Otus asio suttoni Moore Caprimulgus vociferus arizonae (Brewster) Hylocharis leucotus leucotis (Vieillot) Eugenes fulgens fulgens (Swainson) Selasphorus platycercus platycerus (Swainson) Trogpn mexicanus clarus Griscom Colaptes cafer collaris Vigors Melanerpes formicivorus formicivorus (Swainson) Lepidocolaptes leucogaster leucogaster (Swainson) Empidonax fulvifrons rubicundus Cabanis and Heine Mitrephanes phaeocercus tenuirostris Brewster Tachycineta thalassina thalassina (Swainson) Corvus corax sinuatus Wagler Aphelocoma ultramarina wollweberi Kaup Cyanocitta stelleri coronata (Swainson) Parus selateri selateri Kleinschmidt Parus wollweberi wollweberi (Bonaparte) Psaltriparis melanotis iulus Jouy Turkey Vulture Montezuma Quail Turkey Killdeer Band-tailed Pigeon Green Macaw Parrakeet Screech Owl Whip-poor-will White-cared Hummingbird Rivoli Hummingbird Broad-tailed Hummingbird Mexican Trogon Red-shafted Flicker Acorn Woodpecker White-striped Woodhewer Buff-breasted Flycatcher Tufted Flycatcher Violet-green Swallow Holarctic Raven Mexican Jay Steller Jay Mexican Chickadee Bridled Titmouse Black-cared Bush-tit TABLE 3 cont. Certhia familiaris albescens Berlepsch Troglodytes brunneicollis cahonni Brewster Catherpes mexicanus mexicanus (Swainson) Turdus migratorius propinquus Ridgway Sialia mexicana amabile Moore Setophaga picta picta Swainson Icterus EB. Pipilo erythrophthalmus griseipygius van Rossem Aimophila ruficeps simulans van Rossem Junco phaeonotus palliatus Ridgway 16 Brown Creeper Brown-throated Wren Canyon Wren American Robin Western Bluebird Painted Redstart Oriole Rufous-sidded Towhee Rufous-crowned Sparrow Mexican Junco TMEE4 List of amphibians and reptiles from Rancho Las Margaritas, Durangg Amphibia Ambystoma £2. Tschudi Scaphiopus hammondi hammondi Baird Bufo compactilis Weigmann Bufo simus Schmidt Rana pipiens Schreber Reptilia Kinosternon g2, Spix Anolis nebulosus (Weigmann) Phrynosoma orbiculare orbiculare (Linnaeus) Sceloporus grammicus disparilis Stejneger Sceloporus jarrovii jarrovii Cope Sceloporus poinsettii Baird and Girard Sceloporus scalaris scalaris Weigmann Eumeces brevirostris (Gunther) Eumeces lynxe (Weigmann) Barisia imbricata ciliaris (Smith) Conopis nasus Gunther Rhadinaea laureata (Gunther) ThamnOphis egues cyrtopsis (Kennicott) Thamnophis macrostemma (hennicott) Thamnophis melanogaster canescens Smith Crotalus lepidus klauberi Gloyd Crotalus molossus nigrescens Gloyd Crotalus pricei pricei Van Denburgh 17 18 DISCUSSION OF METHODS Live-trapping was used to obtain information on population density and home range. Twenty-six aluminum Sherman folding live-traps, 9 inches long, 3 inches wide and 3% inches high, were available for use in the study. The traps were set and baited with rolled oats. When mice were captured they were first marked by clipping one hind toe and one front toe. ‘The hind toes, from left to right, indi- cated multiples of nine, the left outer toe being nine and the right outer toe being ninety. The front toes were, from left to right, one to eight. By using such a system 98 mice can be marked without clipping more than two toes. After the mice were marked they were sexed and measured for ear length. This information, plus any brief comments about the specimen such as a bobbed tail or torn ear, are recorded in a field notebook with the date of capture and the number of the trap station. The mice were then released at the location where they had been captured. The traps were first set in a line along the bottom of the canyon. They were spaced at stations approximately ten yards apart and were left there for three days. They were then removed and relocated in two transects across the canyon from rim to rim. The first transect was one-third of the distance down from the head of the canyon, and the second was almost one-half of the remaining distance to the down- stream boundary of the canyon. The transects contained nine and seven- teen trap-stations respectively. This second set was also spaced at stations approximately ten yards apart and left for three days. They were then relocated for a third three—day period by spacing them at stations along the sides of the canyon in an effort to cover as much o " V ' ‘ , . .7 , , . . , . - . . l ' VA . ~fi . . , L . 7 . , .. ,, 7i 7‘ v ‘ ._ v . ,. ,' _ I V ,. , . .4 , a l - . . ,. , - l O ‘ ' ‘ - . a — a , .4 . - ,4 , A . . r . - - ' - .2 J _. _ . _ . . , z _, - e , .' . . , . _ ,. 4 a. 4 , - i. a, - VJ , ,a ‘ 7, . 19 of the remaining area as possible with only one more setting. The stations chosen were at twenty-yard intervals down the canyon and ten-yard intervals up from the bottom where the first setting was made. The first three intervals, where the canyon was narrow, had one trap-station on each side ten yards up from the bottom. The next three intervals had two trap-stations on each side, at ten and twenty yards from the bottom. This left only four traps for each side in a large area which had not been previously trapped so at the last two twenty-yard intervals, the trap-stations were spaced thirty yards apart up the sides of the canyon. By setting the traps in this manner, each trap-station was at the center of a square in a ten-yard grid over the entire canyon. Because there were insufficient live-traps to adequately cover the entire area simultaneously, the traps were more or less concentrated along the bottom while the rim areas were neglected. Figure 4 shows the locations of the live-traps for these three periods during the nine days of live- trapping. At the start of the live-trapping the traps were checked in the morning and again in the evening but since only one mouse was found.when traps were visited on the first two evenings, the traps were subsequently checked on mornings only. The traps were inspected just after daylight for the first five or six mornings and many mice were found to be suffering from exposure. Since mice seemed to recover rather quickly when they were warmed, traps during the latter part of the trapping period were left until the sun had been on them for about an hour before they were checked. This delay seemed advantageous for the majority of the mice were then sufficiently ambulatory to make off immediately upon being released. / -/_~\\"/ \ .1 0' ‘0 \ "1 k . . . ’ /’/ 190000796 0 o o o o 0 0/0 "\. [I \ / e/ \‘ ,- I // f / O f , 1 ,/ ’11 1 // \ // X , 0 j j x ,/ oFirst period 0 0 0 f / l/ oSecond period / I, 0 .r . ,, // ‘0 10“. ’ , / ---~.-—-»—.---—~-n ‘3‘“ J .Third period Ink g ‘ _ I; \Lgk {fin-u .“ ‘} w K .“15' ! L 3 l O 20 40 tjerth—IL, -. '1 scale in yards Fig. 4. A map showing the locations of the live-traps in the canyon. 20 21 When the nine days of live-trapping had been completed, three more days were given to snap-trapping the entire canyon in an effort to determine the size of the resident population. Three hundred and twenty-eight museum special snap-traps were set at ten-yard intervals from rim to rim for the length of the canyon. They were baited with rolled oats and checked each morning for three days. The location and date of capture was recorded for each animal taken. All specimens that died or were taken in snap-traps were either prepared as study skins, skeletonized, or preserved in formalin and are now part of the research collections at the Museum at Michigan State University. The Peromyscus were identified to species with the aid of Osgood (1909) and Hoffmeister (1951) and to subspecies with the aid of Osgood (122. SEE-7 and Miller and Kellogg (1955). It is usually impossible to investigate thoroughly a mammalian population throughout the entire range of the species. Consequently, one of the common methods of such a study is to set up a quadrat, where the mammal occurs, and by various methods determine the nature of the population in that portion of its range. The information thus obtained may be useful in analyzing populations occurring in other parts of the mammal's range. One of the assumptions made in doing this is that there is a typical and even distribution of mammals in the area selected for the sample, but this is seldom the case. Any sampling method used has its disadvantages even in an extensive and uniform type of habitat, but where the terrain is not uniform it becomes even more susceptible to error. Murray (1957) discussed small mammal distribution in habitats, which because of their lack of uniformity, create sampling problems. He showed that in three areas of this type there was evidence to indi- 22 cate that the movement and distribution of small mammals was influenced by the habitat so that a sample would be biased, or subject to error. The habitat in the canyon was not uniform; therefore, the results ob- tained from this study are perhaps subject to error because of some unknown factor or factors. The results presented here would in all probability differ in some way from the results of a more extensive study in a larger and more uniform area in which these species are found. Considering the above, it is felt that the data presented here are only approximations or estimates of the nature of the population found in this area at the time the study was made. The results of the trapping in the canyon are as follows: SPECIES OBTAINE NUMBER OF NUMBER OF INDIVIDUALS CAPTURES Peromyscus boylii rowleyi (J.A.A11en) 59 123 Peromyscus difficilis difficilis (J.A.Allen) 9 l4 Peromyscus truei gentilis Osgood 4 5 Reithrodontomys megalotis zacatecae Merriam 4 10 Neotoma mexicana madrensis Goldman 4 7 23 FINDINGS CONCERNING PERONYSCUS BOYLII The species Peromyscus boylii contains 13 subspecies which range from northern California, east to central Missouri, and south through Mexico as far as southern Honduras (Miller and Kellogg: 1955). The subspecies rowleyi, which lives in the canyon, has a geographic distri- bution from southern California, east to Colorado, and south into Mexico, chiefly on the eastern slopes of the Sierra Madre Occidental to central Zacatecas and northwestern San Luis Potosi (Miller and Kellogg, 1955:494). Cahalane (1939:433) states that EQElEZE.iS absent from the Canadian Life Zone whereas it is perhaps the most abundant mammal in the Transition Life Zone. It has been taken at 11,000 feet in the Inyo-White Mountains of California by Bole (1938), and by Cahalane (1941:57), Long (1940:178), and Vaughan (1954:557) from the Upper Sonoran Life Zone in southwestern United States. Rocky areas with considerable local relief are the preferred habitats of this mouse. Svihla (1931:263), however, states that it is characteristic of the southwest deserts. This subspecies, rowle i, was found by Cahalane (1939:433) in moist localities and along streams. Peromyscus boylii has been reported to be somewhat arboreal by several workers. In the canyon there was no correlation between the trapping fre- quency and the trap locations with respect to the stream in the bottom of the canyon. The mice seemed to prefer rocky areas in which there were many faults and crevices. By far the majority of mice captured went into a crevice or under a rock ledge within three to ten feet from where they were released. They did not seem to take cover in grass. 24 POPULATION DENSITY When attempting to estimate the size of the population of a small mammal, the size of the area selected for study, according to Burt (1940: 37), must be larger than the mammal's home range. The size of the area from which a sample of the population is taken is inversely pro- portional to the magnitude of the error resulting from catching individ- uals whose home range lies partly in an adjacent area. This is par- ticularly important when small areas or quadrats of the minimum size are used. Bole (1939:74) believed that the quadrat method provides an accurate means of determining the abundance of small mammals, and that a quadrat 150 feet or 50 meters on a side is large enough to give accurate figures on population density. Bole's method used at least 100 snap-traps in a lSO-foot quadrat and placed them, not in a uniform grid, but where they would be most likely to catch animals. Although Stickel (1946a:158) concluded that live-traps are the only valid means of determining population density, Goodnight and Koestner (1942:437) maintained that live-trapping and snap-trapping are equally reliable. They further stated that whereas six or seven days are required to determine a population with live-traps, three days are sufficient with snap-traps. Bole (1939:74) also stated that a three-day trapping limit should be set, because the resident population is practically exhausted at the end of this time. The area trapped in the canyon was slightly less than 6.8 acres. Since this area is far greater than the expected home range of Peromyscus boylii, it is believed that a sufficiently large area was used in the canyon to give a reliable estimate of the population present when the study was made. , J v 25 The first method of computing the population is with the use of the Lincoln Index (Lincoln, l930:2). This method was not used for the first nine days of live-trapping because the entire canyon was not trapped at the same time and certain portions were not trapped at all. Because of this, an estimate arrived at by the use of the Lincoln Index would, for the canyon as a whole, be rather meaningless. How- ever, a possible application of this method can be used for the data obtained in the last part of the trapping. This was during the time the entire canyon was trapped for three days with snap-traps. By taking the known number of marked animals in the canyon (from the live-trapping) and dividing it by the proportion of marked animals in the samples snap-trapped, an estimate of the population present when the animals were marked can be obtained. Marked mice in canyon a 27 = 54.0 Marked mice snap-trapped first day = 17 Total mice snap-trapped first day = 34 This gives an estimate of 54 individuals for the number of Peromyscus boylii in the canyon during the live-trapping period. After the first day of snap-trapping this calculation may be done again for the re- maining mice. Marked mice in canyon : 10 = 35.0 Marked mice snap-trapped second day = 2 Total mice snap-trapped second day = 7 This gives an estimate of 35 individuals remaining in the canyon at the end of the first-day's snap-trapping. If this figure is correct, by adding it to the number already removed (34), a population of 69 mice was present before any removal. Also, the estimate of 35 individ- 26 uals for the second day can be added to the number theoretically remaining after the first day (54 - 34 u 20) to give a population of 55.0. This calculation cannot be applied to the third-day's snap- trapping because of 4 mice caught, none had been marked. In analyzing these population estimates possible sources of error should be examined. First of all, the population may be over- estimated because of the possibility that during the time lapse between the beginning of the marking period and the completion of the snap-trapping period some of the marked animals may have left the canyon or died without being recovered. This is the result because some of the mice were probably transients and had left the area, and that small mammals such as Peromyscus undergo a rapid turnover with individuals in the population constantly being replaced. Therefore, it is probable that there were not actually as many marked animals present in the population at the time the samples were removed as the number (27) in the equation for the first day indicates. Evidence to support this assumption is the fact that the number of marked animals recovered during the snap-trapping rapidly declined to zero on the third day while eight marked animals were left unaccounted for. In this event, the estimate of 54 individuals would be too high. This same error, if present, would be pr0portionately greater the second day when it was assumed that there were ten marked mice left in the canyon. Consequently, the estimate of 35 mice, and the estimate of 69 mice which was derived from it, would likewise be exaggerated. On the other hand, the possibility of error leading to under- estimation stems from the fact that the mice most liable to be caught were probably taken in the first sample. A high percentage of these 27 trap-liable mice should have been previously live-trapped and marked whereas the more wary individuals may have avoided the live-traps. This would lead to a proportionately high percentage of marked mice in the first sample. If this is the case, the estimate of 54 animals would be too small. Estimates of males and females in the population for the first day are as follows: EELSE Females Marked Mice in Canyon 16 ll Marked Mice snap-trapped first day 9 8 Total Mice snap-trapped first day 17 17 POpulation Estimate 30.2 23.4 The sum of these estimates gives a population of 53.6 mice present before the snap-trapping. It might be pointed out that in gathering these data the area live-trapped was smaller and not as extensively trapped as the larger, superimposed snap-trapping area. This means that the population estimates previously calculated were for a larger area (the entire canyon) than that which was actually live-trapped (see Figure 4). In an effort to apply the data from the area live-trapped to the entire canyon the following hypothetical example is given. If we assume two similar areas, A and B, then live-trap, mark and release animals on A, and then snap-trap both A and B, the correct population may be calculated for the entire area A plus B. The following figures show what would be the results with a population of eight animals on each area and a trapping success of fifty percent. 28 Area A Area B Area A Area AiB 8 Here 8 Here Animals marked 4 4 4 Marked O Marked Marked animals snap-trapped 2 2 % Snap-trapped % Snap-trapped Total animals snap—trapped 4 8 2 Marked O Marked POpulation estimate 8 16 2 Unmarked 4 Unmarked The second method used to calculate the population is one des- cribed by Hayne (1949bz407) for estimating populations by removal trapping. This method can be applied to the three-day snap-trapping period. It is based upon the fact that as animals are removed, fewer animals will be caught with continued trapping until none will be caught when they have all been removed. In using this method, the catch per day to the number of animals previously removed has been plotted graphically in Figure 5. The sexes when taken together give an estimate of 44.5. Taking sexes separately, the males are estimated at 24.4 and the females at 20.4; these together give a p0pulation of 44.8 individuals. This method assumes that no new animals are replacing the ones being removed. This assumption may be made as it is generally thought that little immigration takes place during the first three days plus the fact that since Peromyscus boylii is largely restricted to the canyon in this area, the supply of replacements is limited. The popu- lation estimate, 45 to the nearest whole number, is probably low because there were 45 mice removed during the three days of snap-trapping, mice were still caught on the last day, and mice had previously been removed because some of them had died in live-traps. Weather is a factor which also may be of significance. The first morning was warm v r. a p. ’C __i b it“ a ,- (1' kt. Canyon by ICHHvaI‘trapplng. day - 4O [ ( tit L11 L-t‘;\‘!1(i (1;!) - 7 (,atJI third (fay ~ 7; 3O :.' 'UIQILLLH (.‘Litlfitciti' ' o-o (J 'J .1 i9 20 + C 2 ("J . 10 ' V 4 A L A \ 10 20 3O 40 50 ‘;1zn:i,t 1‘ r): r-x'i L-112,1_xi 1‘C7fiitl\'('(l 21:11 ('5‘ I"{."il(ilL:‘) (at h Irrst day - 17 17 (,ut_r1 nectnul Sn} — j C; L 1h tili I d Gay ‘ .1 ;' jagulaliun estiagtc - ;4_4 CU.4 20 r 20 ' '2‘ 3‘731t‘31 .3 ' f5 10 f 10 a -I; o—J if. I a _‘ L A b1 O 10 20 30 10 20 30 IILJn-t r l I‘l‘CfV’l (:11:;3 }' L‘k???t)\'('oE moHE we Nmpwm%vucw5m>oe moHE mo Ampwmhvucsam>oe moHE mo EDEHxQE emanm>< usnEdz Esaflxde mmwum>m nmmamm ESEmeE owmwm>¢ umaEmm unwmmw monEom mmamz Ham Wmeww somcmo man CH “wakes msomhfiowom mo ucofim>oa EDEmez m magma 34 and that the movement indicated from five captures is subject to some discrepancy. Of the two mice captured five times, both were taken from one location twice and then from a second location on three con- secutive nights. It is believed, as Burt (1940:14) has previously reported for mice in Michigan, that the presence of the traps influ- enced the movement of mice and that they repeatedly came back to the traps for food. Consequently, the best figure appears to be the average maximum movement of those mice caught three, four, and six times, this being 43.1 yards for the males and 31.7 yards for the females. Stickel (1946a:154) believes, for Peromyscus leucopus, that four captures are sufficient to show maximum movement. HOME RANGE A definition and an explanation of home range are given by Burt (1957:15) who states that after an animal is old enough to be on its own, it finds an area in which to make a nest or home. The space around this area which the animal traverses in its normal activities of food gathering, mating, and caring for its young is defined as its home range. In order to determine the home range of an animal, the animal must be either trapped or followed, either visually or by tracking, for a sufficient length of time. For Peromyscus the best method so far developed seems to be live-trapping, marking, releasing and recapturing the animals. This method results in a trap-revealed home range which is not necessarily identical with the true home range. Perhaps the trap-revealed home range is biased by the presence of the traps them« selves. The bait used is usually a new and different source of food in the habitat and may cause some individuals to visit a particular 35 trap night after night and thus not divulge their true home ranges. Contrastingly, Manville (1949) found that some animals may become trap-shy and be captured only once while others may escape capture altogether. A population then may contain an unknown proportion of individuals which may be trap-liable, trap-shy, or unaffected by the traps at all. Because of the difficulties involved, there have been several methods developed for determining home ranges; these are discussed by Hayne (1949a:3). The method used here is primarily that of Haugen (1942) which also was used by Manville (1949) for Peromyscus and other small rodents. Jith this method, the entire area is divided by a grid into many small squares, and a live-trap is located at the center of each. After the trapping has been completed the squares in which a mouse has been caught are connected. The home range of that mouse is then the number of squares times the area of each square. In doing this, the squares were connected by their sides whereas Manville connected them by the shortest possible route. If the shortest route is used, then squares may be joined either by their corners or by their sides. However, an inconsistancy occurs when one linear series of locations connected by its sides is compared with another connected by its corners. For example, five squares connected in a straight line by their sides gives a minimum movement of five times the trap distance whereas a similar line connected by their corners gives a minimum movement of seven times the trap distance. Neither method is without error so the one to be used depends on the situation at hand and the direction of maximum movement through the grid. The qualification that the squares must be connected by their sides results in a larger . . . , . . a - - , . ‘ v . , , . _ . . . , .- , . .; v I ' D _ q , , . - , ,- , . - . C , . . _ ' , , 74 . . ,_. H. H4 _ , . . , , 36 home range, but as Manville felt in his study, the calculated home range is still probably less than the actual home range of the mouse in question. The trap distance used in the canyon was ten yards. Burt (1940:12) tried trap distances of both ten and twenty paces for Peromyscus and got comparable results. He believes that fifteen yards is practical and should catch all mice in five days. Tables 6 and 7 give the calculated home ranges in yards of each Peromyscus boylii taken at least two times in the canyon. Table 8 gives the average home range per number of times trapped (as in Table 5 under Movement). Figures 6 and 7 show the shape and spatial relation- ships of the home ranges for males and females respectively. In choosing a figure, or figures, which best represent the home range of Peromyscus boylii the same assumptions may be made here as were made for the data on movement (see page 34). Therefore, only the data from those mice taken three, four, and six times have been used. This gives an average of 683 square yards (.142 acres) for the males and 533 square yards (.110 acres) for the females. The range in acres is .08 to .23 for the males and .06 to .16 for the females. Storer (1944:181) mentions that the home range of Peromyscus boylii is dif- ficult to measure because this species is prone to wander. He gives figures for the species, from a study of a different subspecies in the Sierra Nevada Mountains of California, which indicate a larger home range for the female than for the male. These figures, along with some others, are given below for comparison; all are in acres. 37 Table 6 Calculated home ranges of male Peromyscus boylii in the canyon Males Times Locations First and last Home rangg caught where caught dates caught (sq. yards) 2 2 July 9-11 200 " " " 5-7* 200 " " " 5-6* 400 " " " 10-11 400 3 2 " 9-12 500 " 3 " 5-11 700 " " " 4-11 400 " " " 2-3* 400 4 3 " 5-11 500 " " " 3-11 1100 " 4 " 3-11 1100 5 3 " 6-10* 400 6 4 " 4-11 800 * Died in live-trap. Table 7 Calculated home ranges of female Peromyscus boylii in the canyon Females Times Locations First and last Home rangg £22825. where caught dates caught (s3. yards) 2 2 July 10-11 400 3 2 " 7-10* 400 " " " 9-11 800 " " " 8-11 400 " " " 5-11 300 " 3 " 6-11 600 " " " 7-11 700 4 2 " 2-6 400 " 3 " 2-12 600 5 2 " 3-10* 300 6 5 " 2-11 600 * Died in live-trap. 39 mmn and nus on oon mm. com com mmn m N ¢ 0H Ha H H N c was a soc NH 0 arsrm ...... ...... am.-- ....... Nam ......... --..N I Z: omN m ass a a on a one a mNn aH 0 cu m can mH mNm sN a on N ..... corsair" 0 as: H can N m com m can n a con 4 ONn oH m com a ONM n. N ooq H muvuoa‘qonvuMduu use: umvuoa< moauamm mafia mo Honauz hookah .dmvuwmmu mafia mo Novuohw.vmvuw¢uu code no mean umwum>< nonasz use: umouo>< nonauz unmade «was: mam“ uuaaa eczema onu aw «HHNOA maomhaouom no «mean use: omnuuhd w 0.3a“. ~\‘ c; «.2 ' \‘Nk - k \ ‘ ‘ 11 t 1 / «_ 1 ‘3‘.“ ~ '\ . ‘:. ‘ v \.‘,3 a. -_.f ‘ 3 0 20 ~10 LmfirT‘T 733:1 sc ale in yards Fig. 6. A map of the canyon showing the home ranges of 13 male Peromyscus boylii. 0 '20 40 W111 sc ale in yards Fig. 7. A map of the canyon showing the home ranges of 11 female Peromyscus boylii. 41 Species g. boylii rowleyi g. boylii g. leuc0pus g. maniculatus g.m.gracilis g.m.gracilis §.m.bairdii MALE 1.17 Ave. .142 .27 .270 .77 3.28 1.41 Ave. .110 .41 .208 .21 42 Source Drake(Present Paper) Storer(1944) Burt(l940) Storer(l944 Manville(l949) Blair(l942) Blair(l942) In all, 123 captures of 59 Peromyscus boylii were made in the canyon. Figure 8 shows the locations and trapping frequency, on a map of the canyon. It may be noted that only one capture occurred beyond the rim of the canyon; this being in a rather intermediate area. occurred in the flatter southern end of the canyon. None These data indi- cate that in the canyon Peromyscus boylii is perhaps restricted to, or at least prefers, the steeply inclined, more broken, and rocky type of habitat within the canyon to the flatter, less broken environment of the surrounding area. An interesting occurrence during the live-trapping was a double capture; a male (- in the same live-trap. U 1238) and a female (MSU 1256) were taken together For this to occur, both individuals must have entered the trap simultaneously, suggesting that the two had been traveling together. Peromyscus leuc0pus do travel in pairs. Burt (1940:24) has evidence to show that at times Evans and Holdenreid (1943) captured 183 Peromyscus boylii of which 24 were involved in double captures. They also obtained double captures involving different species of Peromyscus and one with a Peromyscus and a Perognathus. Blair (1942:31) had ten double captures of Peromyscus maniculatus . 1 capture . 2 captures . 3 captures 410 SC ale in yards Fig. 8. A map of the canyon showing the trapping frequency of the locations where Peromyscus boylii had been taken. 43 44 as well as one of a PEromyscus and a Napaeozapus. In no case did Blair find two females in a trap together. Evidently there is then a considerable amount of tolerance among males of the genus Peromyscus, especially within their own species and perhaps extending to other species in the genus and even to other genera. BREEDING 0f the 21 female Peromyscus boylii that were examined and pre- pared as museum specimens none showed any indication of breeding act- ivity. The literature on Peromyscus boylii rowleyi indicates negative breeding in March from Aguascalientes (Hooper, 1955:16), a pregnant female in November as well as immatures during February and March and again during October and Navember from the San Gabriel Mountains of California (Vaughan, 1954:558), and pregnancy records from Texas to indicate breeding seasons in June and again in August (Davis and Robert- son, 1944:269). For Peromyscus boylii in the northern Sierra Nevadas, Jameson (1953:46) gives a table of the percentage of females with fresh corpora lutea per month. July was the lowest month with 31 percent while the two highest months were May with 86 percent and September with 80 per- cent. The percentage of mice with fresh corpora lutea in this study indicates two breeding periods per year, one in May and the other during August and September with a period of inactivity in July. Storer (1944:180), however, observed breeding in the Sierra Nevadas on April 7 and noted an increase in activity to a peak in June which dropped off gradually in July and did not rise again for an autumn breeding period. In spite of this scant information, it appears that the breed- 45 ing habits of Peromyscus boylii are not uniform throughout its range, probably because of environmental and climatic differences, but that this species generally undergoes two periods of breeding act— 1v1ty per year, one in spring and one in autumn with a period of inactivity during the warmest months. 46 CTUER SPECIES OF PEEGMYSCUS Two other species of gpromyscus, P. difficilis and P. truei, were found in the canyon along with g. boylii. Davis (1944:396) mentions that two species of Peromyscus, one large and one small, will often live together in the same area. He reports that the larger difficilis and the smaller trugi frequently live in association with each other, and further that difficilis is often found in rocks along with bo lii, again a smaller mouse. Hooper 6955:14-16) has taken tgggi in the same trap line with boylii so all combinations of these three species have been taken together in the same area although not necessarily in exactly the same habitat. In the canyon l4 captures of nine Peromyscus difficilis and 5 captures of four Peromyscus truei were obtained. The locations of these 19 captures are not at random in the canyon. Figure 9 (map of stations where captured) shows that these species do not range through- out the canyon, but, with two exceptions, do not seem to extend into the canyon beyond the rim area. Only three captures, all of Peromyscus difficilis, were down into the canyon far enough so as to be away from the rim, one being in a very atypical area of level ground similar to that outside the canyon. The other two may have been caught while on "occasional sallies outside the area (home range), perhaps explora- tory in nature" as mentioned by Burt (1943:351). It is assumed that these 13 mice, if captured within their home ranges, had a home range outside of the canyon, the periphery of which extended to the area of the canyon rim. Because of the small number of individuals, and captures, two caught three times each and one caught twice in the case of Peromyscus . H I , . . . . o , ,7 9 , , ‘ . . ,, i , . . — d , . . i 4 , r a g . _ , , , i - . , . o , . , _ i , : H i <3 Peromyscus difficilis . Peromyscus truei sc ale in yards Fig. 9. A map of the canyon showing the locations where Peromyscus difficilis and Peromyscus truei had been taken. 47 difficilis and one caught twice in the case of Peromyscus truei, no attempt has been made to determine the population density, sex ratio, movement, or home range of these species. 48 49 SUMMAEY In the summer of 1957, an expedition from Michigan State Univ- ersity spent three weeks at a ranch at 8350 feet in the Sierra Madre Occidental of Durango. While at this ranch, a quadrat was marked off in the head of a small canyon for the purpose of studying Peromyscus. The objectives in doing this were (1) to determine which species were in the canyon, (2) to determine their population densities, (3) to determine their home ranges and (4) to gather information on their life histories. A description of the physical features of the canyon is given as well as lists of its conspicuous flora and vertebrate fauna. The methods used consisted of live-trapping in the canyon for nine days to catch, mark, release, and recapture as many mice as poss- ible to obtain home range data, then snap-trapping for three days in an effort to determine the population and obtain Specimens for the Museum. Sherman live-traps and museum special snap-traps were used, all baited with rolled oats. Small mammals trapped included the brush mouse, Peromyscus boylii rowleyi (J. A. Allen), the big-eared rock mouse, Peromyscus difficilis difficilis (J. A. Allen, the pihon mouse, Peromyscus truei gentilis Osgood, the western harvest mouse, Reithrodontomys megalotis zacatecae, Merriam, and the Mexican wood rat, Neotoma mexicana madrensis Goldman. It was found that Peromyscus boylii was the most abundant species of Peromyscus in the canyon. Two methods were used to compute the pop— ulation density of this species in an area of 4.94 acres. The Lincoln Index method gave estimates of 30.2 for the males, 23.4 for the females, and 54.0 for the total population, while a removal trapping method described by Hayne gave estimates of 24.4 for the males, 20.4 for the 50 females, and 44.5 for the total population of this area. Analysis of the estimates obtained from these two methods gave an average of 6.0 males, 4.8 females, or 10.8 Peromyscus boylii per acre. This may be expressed as a biomass of 240 grams per acre for the species. Of 33 individuals marked 27 percent were transient of which 77 percent were males. Sex-ratio estimates ranging from 1.14 to 1.36 males per female were obtained, the midpoint being 1.25. The average maximum movement of these mice was 43.1 yards for the males and 31.7 yards for the females. The method of calculating home range was primarily that of Haugen. This method gave a home range of from .08 to .16 with an average of .142 acres for the males, and from .06 to .16 with an average of .110 acres for the females. Because of the infrequency of captures at the periphery of the canyon, it is felt that in this area Peromyscus boylii may be restricted to, or at least prefers, the inclined, more broken, and rocky type of habitat within the canyon to the flatter, less broken environment of the surrounding area. The two other Species of Peromyscus, g. difficilis and g. truei, were found along the rim area of the canyon and did not extend their home ranges into the canyon proper. Few captures were made and little information was obtained concerning them. 51 Measurements of Peromyscus boylii rowleyi (J. A. Allen) from the canyon M.3.U. Museum catalogue number 1252 1429 1256 1220 1226 1237 1238 1228 1250 1246 1225 1430 1201 1251 1221 1222 1254 1253 2959 1255 +0 4 O +O+O+O+O Total length 159 189 189 175 195 157 190 194 foot 22 20.0 88 20 19 14,3 85 21 21 20.3 91 22 19 13.8 97 23 21 18.6 96 20 21 24.0 21 21 20.9 96 21 20 17.8 105 22 19 20.6 98 21 21 24.6 96 21 21 22.6 20 21.6 95 20 21 20.1 22 21 27.0 21 22 16.2 91 20 21.3 20 22 18.9 22 22 21.6 97 21 24.2 87 20 21.7 104 22 19 26.0 98 22 20 17.4 2960 1227 1223 1230 1243 2958 1234 1248 1231 1242 1240 1259 1258 1257 1260 1261 2962 '1 \l 40 40 +0 +0 178 179 198 191 188 191 176 175 177 167 91 98 95 96 98 78 96 97 90 9O 91 21 22 22 20 21 20 22 21 20 22 22 22 21 20 21 21 22 19 21 21 20 22 19 20 19 h.) b) 25.6 18.0 22.5 22.6 27.2 23.0 24.6 52 2961 , 21 20 25.6 2963 g' 180 94 20 21 19.2 2964 9 21 22 23.3 Average 182.6 93.4 21.1 20.6 22.2 Measurements of g. difficilis difficilis (J.A.A11en) from the canyon 1277 9 210 112 25 28 26.8 1273 c’ 230 121 24 27 25.7 1278 o' 223 121 27 26 30.0 2983 o' 226 127 26 24 32.2 1279 9 222 114 24 35.7 1281 8‘ 223 116 25 27 31.8 1280 9 222 117 22 22 34.6 1282 6‘ 217 111 24 28 30.0 2984 9 233 129 25 29 35.2 Average 222.9 118.7 24.7 26.4 31.3 Measurements of P. truei gentilis Osgood from the canyon 1247 9 177 86 21 25 20.9 1378 9 198 97 26 25 27.2 1377 Q 210 116 24 26 24.5 1379 d 183 92 21 25 22 6 k) L.) O l\. kn L.) {u L») OD Average 192.0 97.8 54 LITERATURE CITED Blair, W.F. 1942. Size of the home range and notes on the life history of the woodland deer-mouse and eastern chipmunk in northern Mich- igan. Jour. Mamm., 23:27-36, 1 fig., February 14. Bole, B. P. Jr. 1938. Some altitude records for mammals in the Inyo-White Mountains of California. Jour. Mamm., 19:245-246, May 14. 1939. The quadrat method of studying small mammal p0pulations. Sci. Publ. Clev. Mus. Nat. Hist., 5:15-77, 17 figs., December 28. Burt, w. H. 1940. Territorial behavior and populations of some small mammals in southern Michigan. Misc. Publ., Mus. Zool. Univ. Mich., 45:1-58, 2 pls., 8 figs., 2 maps, May 8. 1943. Territoriality and home range concepts as applied to mammals. Jour. Mamm., 24:346-352, 1 fig., August 17. 1957. Mammals of the Great Lakes Region. Univ. Mich. Press, Ann Arbor, xvi 3 246 pp., 54 figs. plus unnumbered illustrations and maps. Cahalane, V. H. 1939. Mammals of the Chiricahua Mountains, Cochise County, Ari- zona. Jeur. Mamm., 20:418-440, 2 p1., November 14. 1941. A trap—removal census study of small mammals. Jour. Wildlife Mgt., 5:42-67, 2 pls., 11 figs., January. Davis, W. B. 1944. Notes on Mexican mammals. Jour. Mamm., 25:370-403, 1 map, December 12. Davis, W. B. and Robertson, J. L. Jr. 1944. The mammals of Culberson County, Texas. Jour. Mamm., 25: 254-273, 1 p1., 2 figs., September 8. Evans, F. C. and Holdenried, R. 1943. Double captures of small rodents in California. Jour. Mamm. 24:401, August 17. Friedmann, H., Griscom, L., and Moore, R.T. 1950. Distributional check-list of the birds of Mexico. Pt.I Cooper Ornithological Club, Pacific Coast Avifauna No.29, 202 pp., 2 unnumbered pls., June 30. Goldman, E.A. and Moore, R.T. 1945. The biotic provinces of Mexico. Jour. Mamm., 26:347-360, 1 map, February 12. SS Goodnight, C.J. and Koestner, E.J. 1942. Comparison of trapping methods in an Illinois prairie. Jour. Mamm., 23:435-438, 1 fig., November 14. Haugen, A.O. 1942. Home range of the cottontail rabbit. Ecology, 23:354-367, 6 figs. Hayne, D.W. 1949a. Calculation of size of home range. Jour. Mamm., 30:1-18, 2 figs., February 14. 1949b. Two methods for estimating population from trapping records. Jour. 12mmn, 30:399-411, 3 figs., November 14. Hoffmeister, D.F. 1951. A taxonomic and evolutionary study of the pifion mouse, Peromyscus truei. Illinois Biol. Mono., vol. xxi, no.4: ix 3 104 pp. 5 pls., 24 figs., November 12. Hooper, E.T. 1955. Notes on mammals of western Mexico. Occ. Papers, Mus. Zool. Univ. Mich., 565:1-26, March 31. Jameson, E.W. Jr. 1953. Reproduction of deer mice (Peromyscus maniculatus and P. bgylgi) in the Sierra Nevada, California. Jour. Mamm., 34:44—58, 4 figs., February 19. Lincoln, F.C. 1930. Calculating waterfowl abundance on the basis of banding returns. U.S. Dept. Ag., Cir. 118:1 - 4, May. Long, W.S. 1940. Notes on the life histories of some Utah mammals. Jour. Mamm., 21:170—180, May 14. Manville, R.H. 1949. A study of small mammal populations in northern Michigan. Misc. Publ., Mus. Zool. Univ. Mich., 73:1-83, 4 pls., 6 figs., 1 map, August 16. Miller, A.H., Friedmann, H., Griscom, L. and Moore, R.T. 1957. Distributional check-list of the birds of Mexico, Pt.II. Cooper Ornithological Society, Pacific Coast Avifauna No.33, 436 pp., 7 unnumbered pls., December 20. Miller, G.S. Jr. and Kellogg, R. 1955. List of North Lmerican Recent Mammals. U.S. Nat. Mus. Bull. 205: xii 4 954, March 3. Murray, K.F. 1957. Some problems of applied small mammal sampling in western North America. Jour. Mamm., 38:441-451, 2 figs., November 20. Osgood, W.H. 1909. 56 Revision of the mice of the American genus Peromyscus. N. Amer. Fauna, 28:1-285, 8 pls., 12 figs., April 17. Smith, H. M. and Taylor, E.H. 1945. 1948. 1950. Stickel, L. F. 1946a. 1946b. An annotated checklist and key to the snakes of Mexico. U.S. Nat. Mus. Bull. 187:iv 4 239, October 5. An annotated checklist and key to the amphibia of Mexico. U.S. Nat. Mus. Bull. 194: iv 1 118, June 17. An annotated checklist and key to the reptiles of Mexico exclusive of the snakes. U.S. Nat. Mus. Bull. 199: vi 1 253. Experimental analysis of methods for measuring small mammal pepulations. Jour. Wildlife Mgt., 10:150-159, "I 3 figs., April. The source of animals moving into a depOpulated area. Jour. Mamm., 27:301-307, 2 figs., November 25. Storer, T.I., Evans, F.C., and Palmer, F.C. 1944. Svihla, R.D. 1931. Vaughan, T.A. 1954. Some rodent populations in the Sierra Nevada of Calif- ornia. Ecological Monographs, 14:165-192, 18 figs., April. Mammals of the Uinta Mountain region. Jour. Mamm., 12: 256—266, 1 p1., 1 fig., August 24. Mammals of the San Gabriel Mountains of California. Univ. Kansas Publ., Mus. Nat. Hist., 7:513-582, 4 pls., 1 fig., November 15. :11! ”5‘1 F. w 0011 1335. wt MICHIGAN STATE UNIVERSITY LIBRARIES IIII III III I 3 1293 03062 0649