1 3OO - C&Th5 GRADUATE STUDIES TEXAS TECH UNIVERSITY

Population Dynamics, Reproduction, and Activities of the Kangaroo Rat, Dipodomys ordii, in Western Texas

Herschel Whitaker Garner

I-

IC r.kl'If

7f 7

No. 7 May 1974 1

TEXAS TECH UNIVERSITY

Grover E. Murray. President

Glenn E. Barnett, Executive Vice President

4 Regents.-Bill E. Collins (Chairman), J. Fred Bucy, R. Trent Campbell, Clint Formby, John J. Hinchey, Frank Junell, A. J. Kemp, Charles G. Scruggs, and Judson F. Williams.

Academic Publications Policy Committee. J. Knox Jones, Jr. (Chairman), Dilford C. Carter (Managing Editor), C. Leonard Ainsworth, Craig C. Black, Frank B. Conselman, Samuel E. Curl, Ray C. Janeway, W. R. Johnson, S. M. Kennedy, Thomas A. Langford. George F. Meenaghan. Harley D. Oberhelman, Robert L. Packard, and Charles W. Sargent.

Graduate Studies No. 7 28 pp. 24 May 1974 $1.00

St4

Graduate Studies are numbered separately and published on an irregular basis under the auspices of the Dean of the Graduate School and Director of Academic Publications, and in cooperation with the International Center for Arid and Semi-Arid Land Studies. Copies may be obtained on an exchange basis from, or purchased through, the Exchange Librarian, Texas Tech University, Lubbock, Texas 79409.

A 1

J 4

Texas Tech Press, Lubbock, Texas

1974 GRADUATE STUDIES TEXAS TECH UNIVERSITY

Population Dynamics, Reproduction, and Activities of the Kangaroo Rat, Dipodomys ordii, in Western Texas

Herschel Whitaker Garner

No. 7 May 1974 A4

TEXAS TECH UNIVERSITY 1

Grover E. Murray, President

Glenn E. Barnett, Executive Vice President

Regents.-Bill E. Collins (Chairman), J. Fred Bucy, R. Trent Campbell, Clint Formby, John J. Hinchey, Frank Junell, A J. Kemp, Charles G. Scruggs, and Judson F. Williams.

Academic Publications Policy Committee.-J. Knox Jones, Jr. (Chairman), Dilford C. Carter (Managing Editor), C. Leonard Ainsworth, Craig C. Black, Frank B. Conselman, Samuel E. Curl, Ray C. Janeway, W. R. Johnson, S. M. Kennedy, Thomas A. Langford, George F. Meenaghan, Harley D. Oberhelman, Robert L. Packard, and Charles W. Sargent.

Graduate Studies No. 7 28 pp. r. 24 May 1974 $1.00

Texas Tech Press, Lubbock, Texas

1974 Population Dynamics, Reproduction, and Activities of the Kangaroo Rat, Dipodomys ordii, in Western Texas

Herschel Whitaker Garner

The kangaroo rat, Dipodomys ordii, is one of the most common rodents of western Texas and adjacent areas. Previous studies have dealt with taxonomy (Setzer, 1949; Desha, 1967; Schmidly, 1968, 1971), reproduction (Blair, 1943a; Day et al., 1956; Duke, 1940 and 1944; Johnston, 1956; McCullock, 1961; McCullock and Inglis, 1961; and Pfeiffer, 1956 and 1960), and movements (Blair, 1943a; Ramsey, 1969) of this species. However, no study has been made that attempted to correlate the dynamics of a local population and diel utilization of area and habitat. My study was initiated in May 1964, and, except for short intervals, was continued until May 1970. The intensive study area was located 6 mi. S and 1 mi. E Kermit, Winkler County, Texas. This area (described in detail by Tinkle et al., 1962) is located in an ecotone between the Chihuahuan and Kansan biotic provinces. According to Setzer (1949) and Hall and Kelson (1959), the subspecies of this region is D. o. medius. The purposes of the study were to determine: 1) seasonal patterns of density; 2) reproductive dynamics; 3) seasonal weight changes; 4) growth rates; 5) activities and movement; and 6) other factors critical to the total dynamics of the population.

METHODS AND MATERIALS The intensive study area was located at approximately 31 47' N latitude and 103 12' W longitude at an altitude of 2800 feet (Tinkle et al., 1962). The study site was in an area of deep, windblown sand deposits, which have their origin from the upper Pecos valley, and are Pliocene to late Pleistocene in age (Wen- dorf, 1961). The study area is owned by Mr. Earl Vest, Monahans, Texas. Collins (1966) and Tinkle et al. (1962) discussed the climatology of the area. Wind has been an important factor in determining the local ecology. The region characteristically has hot, dry summers and mild winters. The average daily temperature in January is 7.8 C (46 F) and the average minimum temperature is - 1.7' C (29 F). Average daily July temperature is 27.8 C (82 F) and average maximum temperature is 35.6 C (96 F). Mean annual rainfall varies from 10 to 15 inches with an average of 11.31 (Tinkle et al., 1962; Collins, 1966). Microclimatic data were recorded from the southeast corner of the study area from September 1965 through October 1966. Temperatures were recorded 6 inches below the surface of the soil on a Honeywell number 602XIF-II-III-87 recording thermometer with a remote sensing probe, and 6 inches above the surface of the soil with a Tempscribe Model SDC recording thermometer with a

3 GRADUATE STUDIES TEXAS TECH UNIVERSITY remote sensing probe. Barometric pressure changes were recorded with a Taylor recording barometer. Soil temperatures ranged from a minimal winter average of 1.6' C (350 F) to a maximum average summer soil temperature of 32.80 C (91 F). Air temperature averages ranged from -3.3 C (26 F) to 44.5 C (112 F). Barometric pressures ranged from 28.8 to 30.7 with an average of 29.8 inches of mercury. The most conspicuous ligneous floral elements were: mesquite, Prosopis glandulosa var. glandulosa; sand sage, Artemisia filifolia; bear grass, Yucca sp.; shin oak, Quercus havardi Rydb.; and broomweed, Gutierrizia sarothrae. Her- 4 baceous plants formed rather extensive ground cover at times, depending upon the presence of moisture. Some of the more common ones were Hymenopappus flavescens; spectacle pod, Dithyraea wislizeni var. Palmeri Payson; and evening primrose, Oenothera spp. Predominant grasses on the area were fluff grass, Erioneuron pulchellum Tateoka; bluestem, Andropogon spp.; and sandbur, Cenchrus incertus. Ord's kangaroo rat was the most abundant mammal on the study area. Other mammals in order of decreasing abundance were: Perognathus flavescens, Onychomys leucogaster, Spermophilus spilosoma, Neotoma micropus, Lepus californicus, and Canis latrans. Burrows of the plains pocket gopher, Geomys bursarius, were observed. Cattle occasionally traversed the area but caused little disturbance. Reptiles noted on the area were: rattlesnakes, Crotalus atrox and C. viridis; Arizona glossy snake, Arizona elegans; rat snake, Elaphe guttatta; long- nosed snake, Rhinocheilus leucontei; western hog-nosed snake, Heterodon nasicus; side blotched lizard, Uta stansburiana; whip-tailed lizard, Cnemi- dophorus tigris; leopard lizard, Crotophytus wislizinii; and the western box turtle, Terrapene ornata. Predatory birds observed in the vicinity were: marsh hawk, Circus cyaneus; red-tailed hawk, Buteojamaicensis; Harris's hawk, Parabuteo unicinctus; sparrow hawk, Falco sparverius; great horned owl, Bubo virginianus; and burrowing owl, Speotyto cunicularia. The kangaroo rat population was studied by a capture-mark-release method using wire-mesh live-traps similar to those described by Fitch (1950). Traps were placed so that each was located in the center of a 50-foot square. Ten rows of these squares were established in 10 columns, forming a 5.75-acre square. Traps were opened in late afternoon and baited with mixed grain (Purina hen scratch) and checked the following morning. Animals were marked using the toe-clip method of Fitch (1952), weighted on a 250-gram Ohaus spring balance, and released at site of capture. Sex, age (as indicated by molt pattern and pelage color), breeding condition, and direction and distance traveled to a refuge burrow upon release were recorded. General condition of health, indicated by vigor and prominence of neural spines was noted. Traps were opened, baited, and examined a minimum of two nights each month from May 1964 until October 1966. In the summers of 1964 and 1965, traps were baited and opened for a maximum of five continuous days per month. Live-traps were set six additional nights in the spring of 1968. GARNER-DIPODOMYS ORDII IN WESTERN TEXAS 5

The population was studied intensively from July to December 1969, with emphasis on activity and movement. Live-traps were placed at every other station the first night, and changed the second night to the remaining stations, thus sampling the entire area. No weights or escape distances were recorded for the animals trapped in this part of the study. A second study area was established about 5 mi. W Tahoka, Lynn County, Texas, to facilitate the collection of specimens without removing kangaroo rats from the Kermit area. Specimens were collected using a .22 calibre pistol with dust shot and by live-trapping (March and April samples only), from October 1966 to September 1967, along fence rows between range land and cultivated fields. Monthly samples contained at least 10 individuals of each sex. Gonads were extracted and placed in Bouin's fixative immediately after death. Gonadal tissue was stained with Galigher's modification of Harris's haematoxylin and counter stained with eosin in 95 per cent ethanol. Permount was used as the mounting medium. Testes were sectioned transversely through the midline of the organ, thereby cutting the seminiferous tubules, the epididymides, and the proxi- mal portion of the vasa deferentia. The presence of spermatozoa in the seminiferous tubules, the epididymides, and the vasa deferentia was noted. Serial sec- tions of the ovaries were prepared; the number and diameter of vesicular follicles and corpora lutea were :ecorded. Diameter of the mature follicles, ova in mature follicles, and the largest follicle without antrum in each ovary were measured with a calibrated ocular micrometer. The remains of the entire reproductive tracts were extracted, dried on a three by five inch filing card for less than five minutes, reference numbered, and then stored in 10 per cent formalin solution. The remains of all carcasses were labeled for future reference and stored in 10 per cent formalin. The pattern of movements of individuals at the Kermit site was obtained from trapping records for the period of July through December 1969. Traps were set two to three nights each month and checked twice each night so that six captures was the maximum number possible for one month. No animal captured five or more times was captured in fewer than two months. Thus, animals captured five or more times were considered residents. Four samples of six animals each were removed from the Kermit area to the laboratory. Two samples were taken from area A and two from area B (see Fig. 1) on 11 and 18 January and 7 and 15 February 1970, respectively. They were housed in the laboratory in cages 16 by 10 by 10 inches, with glass fronts and a sand substrate, an empty, one- pound coffee can providing a den. Animals were housed in the laboratory 24 hours prior to their introduction into the central chamber of the activity re- corder. Two of these recorders were utilized in the study. Recordings were made of the number of times the animal traversed the gate on each of two con- secutive nights. Untested specimens were introduced into the recording chambers every other day until the entire sample had been tested. A female that gave birth to two young before introduction into the chamber was not tested. The chambers were circular so as to reduce any bias caused by directional movement. A swinging door identical to those used in the live-traps rather than an activity .a

4 6 GRADUATE STUDIES TEXAS TECH UNIVERSITY

o 0I o 0 0 0 0 Q 0 0 0

I I 11

SXO 4, 0 o * o 0 0 o o 0 o 0 o

Ai4

0Ioo

q B x 4,r Q X 0 1 0 0 o I o X o 0 1 0

0 50 FIG. 1.-Diagram of study area showing the position of traps (circles) and centers of 1 activity values of animals used in experimental activity studies: first sample (crosses); second sample (squares); third sample (diamonds); fourth sample (triangles). Areas A and B indicate positions where samples were obtained.

wheel was employed in order to reduce bias in the recording of activity data. 4 A second solid baffle was placed a short distance beyond the baffle containing the swinging door. A small amount of mixed grain and a piece of apple were placed between the baffle plates. The entire floor of the chamber was covered r with sand from the Kermit area. Recordings were made by inked pens on a Har- vard 450, 8-inch chart mover. The chart mover was set at its slowest speed (approximately 10 millimeters per minute) and a timing needle was positioned to indicate minutes. A minute was registered on paper at intervals of 8 to 12 millimeters. From this, reasonably accurate hour intervals were recorded. These in- struments and chambers were housed in a laboratory with controlled light, temperature, and humidity. Florescent lights were on from 0730 to 1930 hours; the temperature was maintained at 23.4 C (720 F), and the relative humidity at a constant 55 per cent.

REPRODUCTION 4 At Kermit, reproductive activity in females was determined by the presence of swollen vulvas, lactation, and pregnancies-position of the testes was used to indicate reproductive readiness. Both external and internal evidence of reproductive activity was obtained from specimens at Tahoka as an assay of reliability of reproductive condition recorded solely on the basis of external evidence. GARNER-DIPODOMYS ORDII IN WESTERN TEXAS 7

Adult Moles

25-

20-

Cuv 15-

10-

1964 1965 1966

Adult Females 20-

15-

C a 10- -ur

1964 1965 1966 FIG. 2.-Numbers and reproductive condition of adults captured each month on the Ker- mit study area, 1964 through 1966.

The reproductive cycle of the Kermit population (Fig. 2) was calculated on the basis of the number of individuals captured each month that showed external signs of reproductive readiness. Few data on reproduction were obtained during 1964. In contrast, data on reproductive condition were recorded at each capture in 1965 and 1966. In addition to any month in which captured, animals were 'A

8 GRADUATE STUDIES TEXAS TECH UNIVERSITY

A 4 60 0-

-- Index Estimates 50 ' ------Breeding Condition f E 40

S 30 -

E 20

-+4, I0z I- k TMFTotM F MFlot MuFMFTot MFMFTot MFMFTot MFMFTot Sub Ad Sub Ad Sub Ad Sub Ad Sub Ad Sub Ad Jul Aug Sep Oct Nov Dec FIG. 3.-Numbers and Lincoln Index estimates of individuals captured each month in 1969 on the Kermit area. Active breeding condition is noted in black. assumed to be present on the area if they had been caught prior to and after the '1 month in question. 11 The data suggest a continuous period of reproductive activity in females from August through May, with peak activities in early autumn and late winter. Re- productive inactivity occurred for three to four months in late spring and summer. This same quiescent summer period with increased activity in the autumn was evident also in 1969 (Fig. 3). The presence of scrotal testes in males was noted in every month of the year. There was little indication of reduced activity at times that females were reproductively inactive. 1 4

100- 0

90- v

80- "

70- 0-0 CORPORA LUTEA 0 W 60- 14 A. 0 "-- VESICULAR FOLLICLES 50- w 14 14 " " 40- 0 * 10 301 0 13 * 10 20] " 10 " o -44 10- o 10 10 OCT NOV DEC JAN FEB MAR APR MAY JUN JUL AUG SEP 1966 1967 FIG. 4.-Occurrences of corpora lutea and vesicular follicles in monthly samples of females from the Tahoka study area (sample sizes above plotted points). GARNER-DIPODOMYS ORDII IN WESTERN TEXAS 9

a A Ag e s the f

'~41Ax, A

Fas 7-8-eo of 'vr (7 showin copu luteu (A) and mauegafinfY

FIGS. 5-6.-Recently ruptured follicle showing the point of rupture (5) and a fresh corpus luteum showing the fluid-filled center (6). Fics. 7-8.-Section of ovary (7) showing corpus luteum (A) and mature graafian follicle (B); and section through epididymis (8A) and vas deferens (81B) showing presence of sperm in both structures.

Reproduction in the Tahoka population is shown in Fig. 4. The occurrence of vesicular follicles and corpora was greatest in early autumn, late winter, and early spring. There was evidence of reduced activity in early winter and a period of inactivity in late spring and summer. The diameter of all follicles and lutea was measured from the outer edge of the stratum granulosum. Vesicular follicles ranged from 370 to 800 microns, and largest follicles without antra ranged from 380 to 450 microns. Recently ruptured follicles (Fig. 5) measured 700 to 1000 A

10 GRADUATE STUDIES TEXAS TECH UNIVERSITY -4

microns, and fresh corpora lutea (Fig. 6) 600 to 1450 microns. Pregnancies were recorded for every month except May, June, and July. The average number of embryos per pregnancy was 2.5, with the highest monthly averages in February (3.7) and November (3.0). Five was the greatest number of embryos seen in any one pregnancy. Some ovaries (Fig. 7) contained corpora lutea and vesicular follicles. Scrotal testes were observed in all months of the year. In most instances, males with scrotal testes were reproductively ready as evidenced by the presence of spermatazoa in the epididymides and vasa deferentia (Fig. 8). Ten of 11 males 4 (91 per cent) in May, nine of 10 (90 per cent) in June, and 13 of 14 (93 per cent) in July had scrotal testes. However, of those with scrotal testes, only 30 per cent

in May and June had epididymal spermatozoa. In contrast, only a single July- A taken male (of 13 examined with scrotal testes) lacked epididymal spermatozoa. Some of the specimens without epididymal spermatazoa (Fig. 9) had immature spermatozoa in the seminiferous tubules (Fig. 10). Ord's kangaroo rat has a reproductive cycle that varies throughout its geographic range. Reproductive activity in Oklahoma and the Texas panhandle occurs from August through March (McCullock and Inglis, 1961), in New Mexico from February through July (Johnston, 1956), and in Utah from early January through March and again from early September through October (Duke, 1944). Ramsey's (1969) observations in Lamb County, Texas, support those of Mc- Cullock and Inglis (1961). Reproductive inactivity in females varied slightly for different years at Kermit, but in 1966 it was similar to that reported by McCullock and Inglis (1961). Re- production extended into May in 1965, and reproductive cycles at Tahoka in 1967 were similar to those at Kermit in 1965. Such variation probably occurs from year to year. Johnston (1956) reported a single breeding period for spring and summer only, whereas Blair (1943a) found that two breeding periods obtained in southern New Mexico, one lasting from autumn to early winter, and the other for three months in the spring. Pfeiffer (1960) described pregnant females, or those that had recently ovulated, as having a greatly reduced vulva and a nearly obliterated vaginal orifice. This evidence was used to indicate reproductive activity. Females in the Kermit area were determined to be reproductively active when the vulva was swollen or the vaginal orifice was perforate. Animals were considered pregnant when the fetuses could be felt by palpation. Females were considered nonbreeding when the vulva was not swollen and the orifice obliterated. Based on such evidence, the data from Kermit might have been biased toward nonbreeding (Fig. 2). Duke (1944) reported mature follicles in D. o. columbianus having a mean diameter of 530 microns, recently ruptured follicles 580 microns, and new corpora lutea 590 microns. He suggested that a follicle attains a diameter of at least 550 to 600 microns before ovulation occurs. A similar range in size of follicles and lutea seems to occur in Texan populations as suggested by the presence of corpora lutea larger than 600 microns. However, considerable variation in follicular and luteal diameters was noted in my study. Johnston (1956) GARNER-DIPODOMYS ORDII IN WESTERN TEXAS 11

B111 I:

A Ni /9

;.. .

[ 4' t y B.,..i ' ,y +" 'e ' a A

f >.

P 10

FIG. 9.-Section through epididymis (A) and vas deferens (B) of testis in scrotal position but lacking spermatozoa in epididymis and vas deferens. FIG. 10.-Section through epididymis and vas deferens of a scrotal testis. Note lack of epididymal spermatozoa (A) but presence of immature spermatozoa in tubules (B). used the length of the testes as an indication of breeding readiness. He detected no reduction in size during summer. Males seemed capable of breeding from February through July. No decrease in the percentage of males with scrotal testes in spring and summer was noted in the Tahoka population. Microscopic examination of the testes revealed that the scrotal position of the testes was not a completely valid indication of breeding ability. Samples obtained in May and June revealed 91 and 90 per cent, respectively, with scrotal testes, yet only 30 per cent of each sample showed signs of epididymal spermatozoa. The breeding season at Kermit begins in August or September and ceases in April or May. Nearly all females were reproductively active during the August- September period (see Figs. 2 and 3). In the following month, pregnant females could have been judged as nonbreeding because of reduced size of vulva and an obliterated orifice. Results of histological examination of ovaries suggest that females copulate shortly after parturition. This conclusion is supported by the presence of both mature follicles and corpora in the same ovary (Fig. 7). Mc- Cullock and Inglis (1961) concluded that in Oklahoma estrus occurred immediately following parturition. Their conclusion was based on the presence of near- term fetuses in three females that had enlarged vulvas. However, their data did not suggest a high incidence of second litters in Texas. A postpartum estrus might not occur frequently in the Tahoka population, because only a single ovary was examined in which corpora lutea of two different ages were detected. Females that produce a litter of young in August or September probably do not produce another for three or more months. Six pregnant females on the Ker- mit area in the autumn of 1965 all conceived a second litter the same reproductive season, but were not pregnant again until February or later. The shortest interval A

12 GRADUATE STUDIES TEXAS TECH UNIVERSITY

-4- recorded between pregnancies was from November to February, and the longest, eight months (August to May), with an average of 4.7 months. The high incidence of corpora lutea and mature follicles in ovaries from the Tahoka population in September and the rapid decline in October, November, and December (Fig. 4), suggests an interval between pregnancies similar to that found in the Kermit population. Subadult females did not occur on the Kermit area in the 1965-1966 reproductive season until December. External genitalia of females in subadult pelage indicated that this age category was capable of reproduction. One female born in captivity reached subadult age in 41 days, at which time she was in estrus as evidenced by a swollen vulva. Monthly totals of subadult females showing reproductive activity were: 10 of 11 in December; 11 of 13 in January; 13 of 14 in February; three of nine in March; and none of 14 in April. These data indicate that females in the breeding population during late winter and spring are those that have bred in the previous autumn and females produced from autumn pregnancies. Therefore, females born in late winter and early spring did not contribute offspring to the population during the later spring months. Eight subadult females first taken in April 1966 were in adult pelage and reproductively active in September 1966.

ACTIVITIES The pattern of movements in the Kermit population was studied by periodically resetting traps throughout the night on the study area during the period July through December 1969. Traps were checked at midnight, 0300, and approximately 0730 hours. Recaptures indicated that animals taken at dawn were the same as those taken at 0300. Thus the 0300 trapping period was discontinued and records from that time were incorporated with those obtained at dawn. Animals captured from three to four times were trapped in either the midnight period or the dawn period, but not both (Fig. 11). Some kangaroo rats recaptured four or more times became trap-prone, suggesting that the animals were not at equal risk of being caught (see Fig. 11). Rats usually were captured in a particular time period the first four captures-17 animals, 53 per cent of the resident population, were captured the first four times in the same time period, and 27 animals, 84 per cent, were trapped three out of the first four captures in the same time period. The greatest number of captures of an individual was 19 in 12 capture nights. Capture nights were figured from darkness through dawn regardlessA' of how many times an animal was trapped. Animals that became trap-prone in both time periods (see the rapid decline of number in the late night category in Fig. 11) were captured in the period before midnight, rather than in the late night period (Fig. 11). This pattern possibly resulted from the early emergence of trap-prone animals in anticipation of finding food in the traps. Such early emergence would permit animals to enter traps before they were checked at midnight and would account for captures in both periods of the night. - - y - - - -".--2524 Shel y - - -

NUMBER TREES AND POUNDS OF NUTS IN THE FOLLOWING COUNTIES AS SHOWN BY UNITED STATES CENSUS OF 1920. T 'F

JHERMAN SD aL LA H '.06n ' ) FO DOCH tLrRF LIPSLOMBI 1 A.0MB1D County. trees. pounds. County. tr . pound s. -- -- Anderson------I1 -2,273 , Dickens- -- 48 1,000 Agel-- 1,354.o , init--_-.- ---- 4,----12 3 6 2 2 1 10HA,842Y OE HUTCHINSON . Roe IERrs CPWL Archer ------8 100 Etor Atas nosa -- 1------1,901 7,211 Edwards -_ - ---- 9,269 69,475 - _Aiy--_------496, 37,744 ll ------11,998 167,834cod ail.y------51 ___--_____ Frath------22,460 319,.35 Bandera4------46,581 42,000 Falls-_------2,403 58,904 oLD -C ACA astrop------10,010, 218,379 Fannin- _------3,314 54,36T Baylor ______-- 105 1,00 Fayette------12,570 196,344 Bee ------153 67 Floyd ------157 28 Bell ------14,901 265,6 Fort Bend_-----_---- 6,956 - 129,49 * - ilexar]:anco-----_-----____------_ ____-__ _ - 29.491 619,85 Franklin------177 ,999 67,221 Freestone-______--829 .. 16,326 Borden------16 6 Galveston ______I__-- DEAF 1,817 5,293 RA DA''L-.ARnsr olw . DoNLEY com CSwom Bosque ------12,946 151,970 Bowtes------Gillespie ------O-- 16,334 257,160 1,045 2,955 GlasScock-___--__--___ 9 lBrazoria 2 0 ------3,03 5,555 Goliad ------905 1,o29 Brazos ------1,426 38,780 Gonzales_------_ 29,578 560,390 - d Brewster ------~ - 102 600 Grayson 9,859 236,292 . Brooks ------6------Gregg------412 2,071 ARMER Brown __ _ I C STE LE 50,439 724,65 Grimes-_------1,491 Burleson ------_-- 1,19 4,236 Guadalupe--___-_ 50,503 581,2163,380 -Red Burnet -- 20,535 464,84 Hamilton ------7,310 77,014 '- - .. Caldwell------_------1,347' 141,447 Calhoun -- -__ __------211- _-__Hardin___-----___---Hardenman 189 901 2,161 3,219 .Iver Callh an ______-I __- 6,482 36,1, Hairis ---- 0------Y,,AM2B, Cameron - _-_-_-_--_-- 2,047 1,070 Harrison _--__----L ,t&P 3Y R Camp ------__ -___ -__ -- 7 , Hays 7 A Caso------9 _1 Hndrsn------167_0,74.j IIf__ I. 2---d---so ---- __ __t__ 1_As C a s t------438 Hil----- r -2 , 9 5 4, H i al---- g o - --__-_'- - -12,726- , 1 212.,119 6, 1 Che o- ____-_--,8 5 ,4l Japern -.-.-.- CA Chambers-___------782,3 ------,86 1,968 YW AA 461 i HoodW ------2)5 __ EE"' K// - -OC 7,67J 61 44 o,.t1 2,94I3,59owl-e.I Crokby _-______4,__4572414u A1,73A CollinsworthCallin------ack------eb40g, ,D Jackson------1,333 1,754,10 Cooke- ,22 419800------,795,181047 36,0 JHnt------__Js - ---I111-1I ,192 4 - - -RByD,103.. - -~w~'uX ll ~e~ LA l."fRD iE . -IK~ S - N -

- ,7 7 J i Jas s ,A14,5E T N c------5,68 L C L.HA_ _A CERIpAN4 -75J2Jei100 __._ C-u-brsonr-"-- - - -. Callas------11-__ ------4,875 45,723 _ _Kinny----- A ELKe- ;1 ST RR Y L YN KUNT ALEERRLT 61,852 IS o -ARA ------IpI

aso , 06LP A' ------24 148 76 1E0UaEr- - - -A7DAAL.S0 B D R.\\D \ E H I'4,.tA D O vS

WI--L-R---ECTO- ' b/ ,I 'R LIEOAEAi iLA E RU R D \EA awson -- Laiar.. U FRENcNsAA- _ _ _ /--I A .RTINL~I. - AG 9JS N SA A.~ M A A '. R BEROs$Do L. E P - 1 .

d -I A AMLWLLAR TL.R A,IRI. d

C'ory-l- A ALel -6 co IU i - I L.OA -- -

E ion

yr.

no - -.- -"r,*-vr Is--N4; A3_,L

1 -- - --,----No.-- No." - 0enr e a H. A~ ,E ----- y------,119 0 , 5y a n g e s _ __ - - 6 0 1 , 68 ,1839 Ir" _ w.-- --- " 5",,4a A

y i arrantv«c 11 468o , 517 E4/ N U TAJDKiO\JL9

_ _ _lo n re n - - ' Y 5 S$ OC TERU G u N CL)I/ L

r. -- - -1,9 5. 6289 Sn aricl - - -2 ! - -Wlli ms n 23 1711

/tl - - -- , .- _ _ Smit L__ Wood !A341 "rris- '1 4 - - - -1588R6035So ere - - -,6 2 8d7' Yun - - - - -1 7 10 5a6" Whe 0'B0,1TSVakN 3 __M9,1 terlinV-----7,348 - - - - JEFFo- --. 7! 9.86S oes- -1,796 Le'Lln- . \ o _ .i9n 77K ZvllaEOS \,MJ07 BUR 351 10AM ~ 5 .D/WAKR 8 31 A~ R

b o u n t . t e e s p o n s C o u n t y . - - t r e sTpu n d C o ue n t y t r en I Librty------618 1,055 ale P ____ 164 0--- 144,4erry- ---

LaCale--- - - 262 5 Nolan --- ,0 373 Laa -'--5 142 71,15 liens 92--- 7 3 n----- 1.68Tylerton7-,6-._-RansA-g17,817_0,51---and 1aeon- - -- -3, 35 1,45 drnger------1 4,9733 1 04-al er - - 348 7,2 u m ston e1 Li- Oak-----1427,7 e u i - - -- 4, 0--lo R n l - - - - 35 10 4014 lhroZkmt o- n -- 1 0 1 1,7 IM a na 4,91 Prber--nVctra---10a o -d -_ - - - 3 0 ,0 9 6 , 215 P eck w ar ------, - -W a k e - -j ---- I'iu - - - - - i ,4 O 68 La v b b c k ______7 _ _ 8 ------e e s - - - - 3 0 3 8 a v i sr _ _ _ s s s 5 5 , 6 j3 z A ) 'T " 'G e o K K E C. E" Jedyn------42314 D 9,1 Polk- - - - -2,08, 5,736 D19,84 Trsint n9 3 V AL-. 1

M elena 1,6 2 1075 eaa - - - - -7 2,500 Upshr R-- -3 761 M ultge ______1 , 2 .44 - - e a ------, 8 9 t 2 U am n ----- 21__7,4

Mads -- --- i,2 2,.524 Red R -__6------er,8 a io n - -h e - - --- It -,. 26 ,5 8 Ro e v es l 5. , 1 0, 0 4 7 V a k u V e d3 4 8! 0, 9 7G avrr ------.421 586Stne 1 - - - - -4 ! -V2 an ---- a2ndRv I - -- t---rd -- 7,790 4,047 trok alin - ---- ,307 A 3,15 W a - -- l-er31 1031 '

N l

& GARNER-DIPODOMYS ORDII IN WESTERN TEXAS 13

100 0 Midnight 90 32 E Morning 80 32 Both 32 70 32 6 5 C60 32 26 50 21

40 23 30

20 1 2 I0-

1 2 3 4 5 6 7 8 9 10 II 12 Capture Night FIG. I 1.-Percentages of residents captured in trapping periods in relation to the number of nights. Numerals indicate number of residents.

Samples of marked animals were removed from the study area (Fig. 1) to the laboratory where their activity patterns were tested under controlled conditions. Attempts were made to sample (samples 1 and 3, see Fig. 1) from the period before midnight by collecting only animals trapped before midnight, and after midnight (samples 2 and 4) by not setting and baiting traps until after midnight. Results revealed that 43 per cent (10 animals) had a greater proportion of their movements during the early part of darkness; 35 per cent (eight animals) during the late hours of darkness; nine per cent (two animals) during the middle of the dark period; and 13 per cent (three animals) showed random activity (Fig. 12). French et al. (1966) found for Perognathusformosus that the amount of time spent on the surface of the ground varied throughout the seasons. Activities of the deer mouse, Peromyscus maniculatus, decreased as the light intensity increased (Blair, 1943b). Activities of animals on my study area were recorded only from July through December on nights associated with a new moon. This was done to reduce variation that could result from length of study and varying light intensities. No previous studies have reported circadian activity in Ord's kangaroo rat. Tappe (1941) observed activity of the Tulare kangaroo rat, Dipodomys heermanni, by illuminating the habitat with a spotlight. Because these rats did not emerge from burrows until darkness, he concluded that light intensity was the 14 GRADUATE STUDIES TEXAS TECH UNIVERSITY 4 .4

N=3 30

7 8 9 10 IIl 12 1 2 3 4 5 6 FIG. 12.-Number of movements on the ordinate made by animals in recording chambers in relation to hours of darkness beginning with 7:00 p.m. (on the abcissa). Sample number indicated on respective line. most critical factor affecting emergence and that location of burrow, phase of moon, age of animals, and weather conditions exerted less influence. These findings support those of Dale (1939). The results of my study were influenced by factors similar to those reported by Tappe (1941). Presence of live-traps and bait were additional sources of bias. Tappe (1941) found that the majority of individuals foraged above ground in two periods about four hours apart, in the early part of the night. Yet, some animals tended to forage intermittantly throughout the night. This activity pattern was also revealed in my study both in the field and under laboratory conditions. However, in my study another late night period for certain individuals was evident.Vy

POPULATION DYNAMICS Densities Densities during the period 1964 to 1966 averaged 6.3 rats per acre, from a high of 10.9 in January and March 1965 to a low of 4.0 per acre in June 1964. The greatest number of captures of a single individual was 32 for a male during a period of 18 months. One female was captured 26 times within a 10-month period, and the longest record of a female remaining on the study area was 14 months. Seven males and three females (nearly one-third of the resident population) were captured for periods ranging longer than a year. The resident population during those years had a sex ration of 1.4 males per female. The non- resident segment of the population had a sex ratio of 1.2 males per female and GARNER-DIPODOMYS ORDII IN WESTERN TEXAS 15

Moles 35 p-p Residents Sub Adults

-" Juveniles 30 p "-

25-

20- /A

99515-

E "

M' J'J'A'S'O'N'D J F M A M'J'J'A'S O N D J F'M'A'M J J J"A SO' 1964 1965 1966 5-

Fe m alIes

0) 0 - -- 0 frtoIb

I'%A\

E z z 5- -- -0-

M J J A' S0 N D J F MA M J J A S O N D J F M AM'J J A S 0 1964 1965 1966 FIG. 13.-Numbers of animals captured each month on the Kermit study area. was composed of 35 per cent adult males, 20 per cent subadult and juvenile males, 20 per cent adult females, and 25 per cent subadult and juvenile females. My figures do not differ greatly from those reported by Desha (1967) as being 1.6 males per female in Lynn County, Texas. The adult portion of the Kermit 16 GRADUATE STUDIES TEXAS TECH UNIVERSITY

/ 305 23

a g 25 0

020 / / 0O/0 / /%

5 a/ 0

0 / _o

10- /d0/T

0-0-0

-0.1 00/

5 10 15 20 25 30 35 40 Day FIG. 14.Change in size and weight of five D. ordii of known age. Numerals on the ordinate refer to both millimeters and grams. population from 1964 to 1966 was highest in number each year in late winter and early spring (Fig. 2). Subadults or juveniles (Fig. 13) were recorded in all months, but were most numerous during spring. Growth of young of known age was plotted in standard external measurements (Figs. 14 and 15). Adult proportions were first reached in the ear and hind foot. Individuals attained a weight of 30 A grams in about 35 days. Only one animal, a juvenile, weighing less than 30 grams was taken on the Kermit area. Therefore, animals were at least a month old before being captured in traps. This suggests a nestling period of 30 days, which agrees with that suggested by McCullock (1961). Large numbers of subadults in the spring of 1966 (see Fig. 13) corresponds with the loss (resulting from exposure to reduced temperatures) of 42 adults in live-traps during the previous trapping periods of December and January. Those animals considered resident (Fig. 13) increased from 45 to 47 during the influx GARNER-DIPODOMYS ORDII IN WESTERN TEXAS 17

210

180

150 /

0* 120 -

90 - / /

60 - /

30 *-*

5 10 15 20 25 30 35 40 Day FIG. 15.-Change in size of five D. ordii of known age. Numerals on the ordinate refer to millimeters.

of subadults. Thirty-two animals (Fig. 11) were considered resident during the period July to December 1969. Lincoln Index estimates of population density for the July to December period are higher than the enumeration counts (Fig. 3). Lincoln Index estimates were compared with results from enumeration counts for all months represented in several years (Table 1). Correlation coefficients indicate a significant difference (P .05) between the two methods except in October and December. Change-in-ratio (CIR) estimators, such as the Lincoln Index, are valid only if all animals have equal probability of being captured (Overton, 1969). Eberhardt (1969) concluded that learned behavioral reactions to traps such as trap-shy or trap-prone responses could bias the results of CIR estimates. He further suggested that individuals in populations exhibiting some kind of restricted movement pattern (resulting from population characteristics, such as territoriality or home range) may not be equally trapped, because of different probabilities to trap A

18 GRADUATE STUDIES TEXAS TECH UNIVERSITY

TABLE 1.-Estimation, by Lincoln Index and enumeration (in parentheses), of number of D. ordii on the Kermit area, and correlation coefficient (R) for the two methods.

Year May Sept. Oct. Nov. Dec.

1965 57(51) 38(30) 47(35) 42(39) 78(52) 1966 63(45) 37(27) 30(25) 1968 51(28) " 1969 60(37) 51(41) 50(39) 58(43) r .88 .86 .99 .96 .99

exposure. This differential in trap response may result from the proximity of a trap to the area of familiarity or home range of an animal. As previously shown (Fig. 11), double nightly captures suggest kangaroo rats exhibit little reluctance to reenter traps. Consideration of growth of known-age young (Fig. 14) and minimum weights of field-trapped individuals (Fig. 16) suggested that animals younger than 30 days were not trapped. Kangaroo rats on the Kermit area showed restricted patterns of movement (see section on movements beyond). Thus they were not captured at random (therefore not at equal trap risk), and CIR estimates may not reflect true densities.

Weights McCullock (1961) and Ramsey (1969) determined the age of Ord's kangaroo rats on the basis of weight. McCullock (1961) concluded that immature animals were less than two months old and weighed 48 to 50 grams. Rats that were two to three months old weighed 55 grams and were considered adults. Ramsey (1969) arbitrarily assumed females of less than 50 grams and males of less than 55 grams to be immature. Adult male and female weights (Figs. 17 and 18), have been compared on monthly basis using the methods of Hubbs and Hubbs (1953) and Dice and Leraas (1936). The general condition of health of individuals was noted periodically by visual and tactile means. The prominence of neural spines beneath the dorsal skin was used to indicate loss of weight. Neural spines were noticeably more prominent during the postreproductive summer months. I detected few emaciated animals. McCullock (1961) found mean weights of male D. ordii in Oklahoma to be 5 grams greater during the breeding season and those of females to be 10 grams greater. Weights declined in the period from May to August, following the breeding season. General health of animals on the Kermit area, indicated by weight gain and lesser prominence of neural spines, noticeably improved during autumn, winter, and spring, when reproductive activity was evident and densities were high. The population in the summer was composed chiefly of adults. All individuals tended to weigh less at that time. There were slight indications of a mild degree of emaciation during that period. Availability of food seemed not to limit the population density inasmuch as greater weights occurred at times of high densities. Low densities and reduced body weights occurred during main plant growing and fruiting seasons. GARNER-DIPODOMYS ORDII IN WESTERN TEXAS 19

Weights - Subadults

4 Jan I 6 5 Feb 7 6 Mar 6 13 Apr 4 5 May 6

Jun

Jul

Aug

Sep

Oct

3 Nov

1 5 Dec 5 30 410 50 60 70 80 FIG. 16.-Average and extreme monthly weights of subadults captured from 1964 through 1966 on the Kermit study area. Males are indicated by upper bar and females by lower bar for a given month, with sample size indicated on right. Movements Distribution of the dominant plant species on the study area at Kermit seemed to be random, but distribution of kangaroo rats seemed patterned (Fig. 19). There were no detectable differences in topography or soils on the area. The probability of any trap capturing more individuals than others seemed nil. No differential use of any part of the study area was detected. Because there were no barriers to hinder movements, they were analyzed using the density probability function of Dice and Clark (1953). Reasons for the use of this method have been presented by Dice and Clark (1953), Calhoun and Casby (1958), Jorgensen and Tanner (1963), and Jorgensen (1968), and reviewed by Sanderson (1966). Tinkle (1967) A

20 GRADUATE STUDIES TEXAS TECH UNIVERSITY

Weights - Adult Males

Jan Th =23 -

Feb II

Mar 20 'Iq

Apr __ Z30

May 52 A4

Jun 34

Jul 30

Aug 31

Sep 20

Oct - 32

Nov 14 .4

Dec 20

30 40 50 60 70 80

FIG. 17.-Cumulative monthly weights (average plus or minus one standard deviation, and 95 per cent confidence limits) of adult males captured from 1964 through 1966 on the Kermit study area. Sample size is indicated at right. concluded that home ranges calculated by this method yielded overestimations of his observed ranges for the lizard, Uta stansburiana. Movements of adult male rats in 1969 occurred 95 per cent of the time within a circular area of 21,000 square feet and those of adult females within a circular area of 24,000. Normal curve statistics were used to calculate areas within which 95 per cent of the move- A ments should be found. No skewness was detected in this analysis. Shifts or seasonal expansion of home ranges could bias the calculated length of recapture radii, disrupting a normal distribution. Tests using Fisher's gi statistic (Sokal and Rohlf, 1969) revealed that movements during the 1964-1966 study were nonrandom. Therefore, recapture radii were calculated using areas of the Y standardized Type III function of Carver (1940), as recommended by Dice and Clark (1953), for the distribution. Movements of adult males were within a circular area of 93,000 square feet 95 per cent of the time, and those of adult females within an area of 65,000-much larger areas than those calculated in 1969. GARNER-DIPODOMYS ORDII IN WESTERN TEXAS 21

Weights - Adult Females

Jan a20

Feb III

Mar 24

Apr 22

May 39

Jun 22

Jul 17

Aug 18

Sep 13

Oct 23

Nov 13

Dec 12

3.0 4,0 5.0 6.0 7,0 8,0 FIG. 18.-Cumulative monthly weights (average plus or minus one standard deviation, and 95 per cent confidence limits) of adult females captured from 1964 through 1966 on the Kermit study area. Sample size is indicated at right.

The difference could have resulted from bias introduced by shifts in movement patterns in different seasons and for different densities. Jorgensen (1968) suggested that areas calculated with this method should utilize only movements determined for a short and continuous period of time. Student t-test for significant differences between recapture radii (calculated from animals with two to five recaptures in all combinations) gave values from .005 to 1.4 (P>.01). Individual direction and distance traveled from release point to escape burrow were recorded throughout the 1964-1966 study. No significant differences were found at the 95 per cent level of confidence between mean activity radii calculated from recapture points and escape points for 113 adult males and 75 adult females. Radii of adult males from the center of activity to capture points averaged 71 11 feet and from the center of activity to escape points averaged 61 8 feet. Recapture radii for adult females averaged 64 11 feet, and escape radii 60 x10 feet. 22 GRADUATE STUDIES TEXAS TECH UNIVERSITY A

00 +0 A

+ + A 4

+ + 0

e to 0 0 0 0 ot e

KeOtsud Seron rafO y hOughDeemer+ +eaO9.Ope crcesrprset ras 000Q0Q-+-A.

agsoae sotesi0g h

FiG. 19.-Activity center values of male (crosses) and female (triangles) residents on the Kermit study area from July through December 1969. Open circles represent traps.

The measurement of movements in my study was not designed to evaluate size or shape of the home range, but to provide an index of the social use of space by the population. According to Sanderson (1966), the geometric shape of the home range seems to have little significance. Eisenberg (1963) found that adult kangaroo rats normally were solitary within a burrow system. He concluded that an asocial dispersal pattern was phylogeneti- cally old within the family Heteromyidae. Behavior of kangaroo rats at Kermit Vf suggests that a degree of mutual exclusiveness existed on the surface of the soil as well as in burrows. Home ranges of male kangaroo rats overlapped in southern New Mexico when measured by the inclusive boundary strip method (Blair, 1 943a). When home ranges are calculated using the method I employed (when A there is no evident territorialism), the number of centers of activity per unit area should be random and form a Poisson distribution. This was not observed. Ter- ritorialism should produce activity centers approaching a uniform distributionA (Calhoun and Casby, 1958) with activity center values approximately two standard deviations apart. Adult males on the Kermit area in 1969 had recapture radii that averaged 46 feet with a standard deviation of 22 feet. Therefore, the mean radius plus two standard deviations was used to inscribe a circle within which 95 per cent of an animal's movements could be expected (Fig. 20). Use of an area increases proportional to the degree of closeness of activity centers. - Area use approaches uniformity when activity center values are spaced two stan- GARNER-DIPODOMYS ORDII IN WESTERN TEXAS 23

/ -\

0 0

A 11l" r A

0 0 00 0 0

o / o

0 0 f/

0 070 o 0 0 0

0- 0 0 o A 0 0 A

0 0

0 50 FIG. 20.-Distribution of centers of activity (triangles) of resident males on the Kermit area, July-December 1969. Circles enclose areas in which 95 per cent of the activity radii of males were located. Open dots represent traps. dard deviations apart (Calhoun and Casby, 1958). Twelve of 20 resident males (Fig. 19) in the 1969 Kermit population had activity center values less than two standard deviations (44 feet) apart. This suggests that intensity of area usage is not uniform in the Kermit area. There seem to be three processes (Calhoun and Casby, 1958) permitting densities to increase: 1) development of smaller home ranges; 2) simultaneous coexistence of more than one individual within a single home range; and 3) more uniform distribution of home range centers. Analysis of movements revealed no reduction in area used, even though over half of the resident males had activity center values less than two standard deviations apart. Therefore, distribution of activity centers in the Kermit population was not random. The use of time in a diel cycle is as significant a feature of a living ecological system as is the use of space. Many organisms have accurate internal time- measuring systems on which environment acts to keep the system properly syn- chronized (Menaker, 1969). Further study of movement and activity patterns at Kermit in 1969 suggested that coexistence of more than one animal within a single estimated home range occurred. Animals were most active either in the early hours or late hours of darkness, but not both. Thus the available area and habitat was utilized less intensely during the middle of the night. Such a pattern of differential use of a habitat was first observed by me while collecting rats from the vicinity of Tahoka, Texas. It was noted that collecting success fell sharply 24 GRADUATE STUDIES TEXAS TECH UNIVERSITY

near midnight and remained low for two to three hours. Different periods of nightly activity were exhibited in the first captures among eight of the 12 resident males at Kermit. When any two males were adjacent, their activity periods were different. In eight of the 12 males, where centers of activity values were less than two standard deviations apart, activity periods were different. Such evidence suggests that area and habitat use were stratified with time during the hours of darkness. Environmental resistance (Cockrum, 1955) within a territorialistic species may be caused by intraspecific interactions. Examples are social use of space and time and utilization of available food. Behavioral patterns of the type seen in D. ordii could have resulted from individuals exploiting areas of the habitat and segments of time when less environmental resistance was encountered. Such exploitation of the habitat at times of less environmental resistance might be described as an opportunistic reaction to the environment. The tendency of animals to become trap-prone (to reenter traps frequently) was considered to be an opportunistic behavioral action. Some adult females were recaptured as soon as 30 minutes after release from a trap. Other evidences for such behavior are provided by records of capture position in the field of rats removed to the laboratory. Animals removed from the study area in the first sample (Fig. 1) had centers of activity values in proximity to the sampling area (area A, Fig. 1). Displace- ment of activity center values after the removal of residents suggested that animals outside the study area either expanded or shifted their activities to include the area previously occupied by those removed. Stickel (1946) intensively kill-trapped an area and reported an influx of peripheral animals into the depopulated space. Flake and Jorgensen (1969) suggested that invasion of an area after collection was influenced by proximity to the trapping area of the ranges of peripheral individuals, as well as the diversity of the surrounding population. Both factors probably influenced movements into the sampling area in my study. The presence of vagrant D. ordii on the Kermit area in the 1964-1966 study was most evident in periods of high densities. Little increase was noted at this time in resident density (Figs. 2 and 13). Supporting evidence for such a spatial pattern of residents and vagrants was added when one (an adult female) of five marked animals released one-half mile west of the study area on 14 February 1970 was recaptured on 5 May 1970 on the study area. These were animals that had been captured on 11 and 18 January 1970 on the study area. Movements of such distance did not seem to be normal for established residents. Therefore, such A 4 a movement may be useful as a crude index to the distance that a vagrant may move within a given period of time. Such a long movement was not considered an indication of homing ability because only one of the five kangaroo rats released one-half mile away, and none of the five released 0.8 mile west of the study area, was trapped again. Animals may adjust -their movement and activity patterns depending upon changes in the population demography in any given area. Stratification for time of activity in kangaroo rats into two major periods would permit populations exhibiting territoriality to be about twice as dense as populations with simul- GARNER-DIPODOMYS ORDII IN WESTERN TEXAS 25 taneous activity and similar behavior. Selection might well act to enforce the establishment of two time-active groups in a population, permitting greater density in any given area and allowing greater genetic variation and perhaps an evolutionary advantage. It appears evident that such activity patterns should be considered for studies involving interactions, either interspecific or intraspecific or both. This is par- ticularly so when studies are used to formulate conclusions about population parameters of social use of space, competition, home ranges, and movements.

SUMMARY AND CONCLUSIONS Population studies of Ord's kangaroo rat, Dipodomys ordii, in Winkler and Lynn counties, Texas, from 1964 until 1970 revealed that breeding and parturition occurred from August through May with only slight variations. Young born in the early part of the reproductive season may bear young during the following winter and spring. Litter size averaged 2.5 with up to five successful implantations per pregnancy. Some males had scrotal testes in all months of the year; some with scrotal testes in the summer did not have epididymal spermatozoa. Multiple pregnancies during one year may have occurred, but the low incidence of recognizable corpora lutea of two different ages suggests that some time lapsed after parturition before an implantation occurred again. Studies of activity patterns indicated that a given individual was active only during a portion of the night. Two different activity periods were distinguishable from the field studies, even though there was a tendency for some individuals to reenter traps. Laboratory studies supported findings in the field. Minimum density estimates were obtained by the enumeration method with allowance made for animals known to be alive but not trapped in a particular month. Density increases could be correlated with reproductive activity, but resident increase in density could not be correlated with reproductivity. Twenty- five to 45 (average 34) residents were present on the 5.75-acre study area in all seasons. Age stratification of the population was observed in autumn and spring, but in summer the population was composed of adults. Data from young reared in the laboratory and weights taken in the field indicated that young of less than 30 days of age were not included in the live-trap samples. Weights increased during the winter months when the population was reproductively active and when densities were high. Reduced weights in the summer resulted from mild emaciation of adults rather than misinterpretation of ages. Negative correlation between weight increases against plant growing and fruiting seasons indicated that food may not have been a critical factor in the population dynamics of D. ordii, but the effects of food storage on the population could be an important factor and are in need of further study. Movements were used as a measure of the social use of the habitat by the population. Ninety-five per cent of the time adult movements were within an area of 21,000 square feet for males, and 24,000 for females. Over half of the activity center values of resident males in 1969 were less than two standard deviations apart, suggesting that certain parts of the study area were -4

26 GRADUATE STUDIES TEXAS TECH UNIVERSITY

utilized more intensively than others. Animals with established areas adjacent to or near a depauperate (nonrelease captures) one shifted or expanded their areas of activity into the depauperate one. As a result of this study, the following may be concluded: 1) breeding and parturition in D. ordii in western Texas occur throughout the year except for a short period in late spring and summer; 2) scrotal position of the testes is not a completely valid indication of reproductive readiness in males; 3) females may breed within three months after birth; 4) many D. ordii become trap-prone and individuals are not at equal risk of being captured; 5) young, less than 30 days old, are never live-trapped; 6) population density increases with onset of reproductive activity; 7) densities of residents do not change drastically throughout the year; 8) all hours of darkness are not used with equal intensity; 9) activity during dark hours can be divided into two major periods, one preceding midnight and one in the late hours of darkness; 10) individuals that are active outside their burrows during late hours of darkness usually are not those that are active in the early hours of darkness; 11) residents have restricted areas in which their movements are normally confined (adult males traversed less area than females in 1969); 12) animals displaced from their established areas of residence may travel as much as a half mile in two and one-half months; 13) residents can expand or shift their movements to include depopulated areas; 14) activity center values of residents suggest that D. ordii exhibits a degree of territoriality; 15) distances between activity center values indicate that some areas are utilized more intensively than others; 16) studies of movements and activity patterns suggest that increased densities may be permitted by temporal stratification of activity during hours of darkness; and 17) activity patterns should be considered in studies involving densities or interactions, or both.

LITERATURE CITED BLAIR, W. F. 1943a. Populations of the deer-mouse and associated small mammals in the mesquite association of southern New Mexico. Contrib. Lab. Vert. Biol., Univ. Michigan, 21:1-40. . 1943b. Activities of the Chihuahuan deer-mouse in relation to light intensity. J. Wildlife Mgt., 7:92-97. CALHOUN, J. B., AND J. U. CASBY. 1958. Calculation of home range and density of small mammals. U.S. Pub. Health Ser., Publ. Health Monogr., 55:iv+ 1-24. CARVER, H. C. 1940. Statistical Tables. Edwards Bros., Ann Arbor, Michigan i+206 pp. COCKRUM, E. L. 1955. Mammalian populations. Pp. 122-130 in Laboratory and field manual for introduction to mammalogy. Ronald Press, New York, 160 pp. COLLINS, L. T. 1966. Flora of the Concho Bluff ecotone. Unpublished M.S. thesis. 4 Texas Tech Univ. DALE, F. H. 1939. Variability and environmental responses of the kangaroo rat, Dipo- domys heermanni saxatilis. Amer. Midland Nat., 22:703-73 1. DAY, B. N., J. J. EGOSCUE, AND A. M. WOODBURY. 1956. Ord kangaroo rat in captivity. Science, 124:485-486. DICE, L. R., AND P. J. CLARK. 1953. The statistical concept of home range as applied to the recapture radius of the deermouse (Peromyscus). Contrib. Lab. Vert. Biol., Univ. Michigan, 62:1-15. GARNER-DIPODOMYS ORDII IN WESTERN TEXAS 27

DICE, L. R., AND H. J. LERAAS. 1936. A graphic method for comparing several sets of measurements. Contrib. Lab. Vert. Genetics, Univ. Michigan, 3:1-3. DESHA, P. G. 1967. Variation in a population of kangaroo rats, Dipodomys ordii medius (Rodentia: Heteromyidae), from the high plains of Texas. Southwestern Nat., 12:275-290. DUKE, K. L. 1940. A preliminary histological study of the ovary of the kangaroo rat, Dipodomys ordii columbianus. Great Basin Nat., 1:63-71. . 1944. The breeding season in two species of Dipodomys. J. Mamm., 25:155-160. EBERHARDT, L. L. 1969. Population estimates from recapture frequencies. J. Wildlife Mgt., 33:28-39. EISENBERG, J. F. 1963. The behavior of heteromyid rodents. Univ. California Publ. Zool., 69:1-100. FITCH, H. S. 1950. A new style live-trap for small mammals. J. Mamm., 31:364-365. . 1952. The University of Kansas Natural History Reservation. Univ. Kansas Mus. Nat. Hist., Misc. Publ., 4:1-38. FLAKE, L. D., AND C. D. JORGENSEN, 1969. Invasion of a "trapped-out" southern Nevada habitat by Perognathus longimembris. Great Basin Nat., 29:143-149. FRENCH, N. R., B. G. MAZA, AND A. P. ASHWANDEN. 1966. Periodicity in desert rodent activity. Science, 154:1194-1195. HALL, E. R., AND K. R. KELSON. 1959. The mammals of North America. The Ronald Press Co., New York, 2 vols., 1083 pp. HUBBS, C. L., AND C. HUBBS. 1953. An improved graphical analysis and comparison of series of samples. Syst. Zool., 2:49-56 and 92. JOHNSTON, R. F. 1956. Breeding of the Ord kangaroo rat (Dipodomys ordii) in southern New Mexico. Southwestern Nat., 1:190-193. JORGENSEN, C. D. 1968. Home range as a measure of probable interactions among populations of small mammals. J. Mamm., 49:104-112. JORGENSEN, C. D., AND W. W. TANNER. 1963. The application of the density probability function to determine home ranges of Uta stansburiana stansburiana and Cnemi- dophorus tigris tigris. Herptologica, 19:105-115. MCCULLOCK, C. Y., JR. 1961. Age classification and weight of Ord's kangaroo rat on the southern great plains. Southwestern Nat., 6:149-155. MCCULLOCK, C. Y., JR., AND J. M. INGLIS. 1961. Breeding periods of the Ord kangaroo rat. J. Mamm., 42:337-344. MENAKER, M. 1969. Biological clocks. Bioscience, 14:681-692. OvERTON, W. S., AND D. E. DAvIS. 1969. Estimating the numbers of animals in wildlife populations. Pp. 403-455, in Wildlife management techniques (R. H. Giles Jr., ed.), The Wildlife Society, Washington D. C., vii + 623 pp. PFEIFFER, E. W. 1956. Notes on reproduction in the kangaroo rat, Dipodomys. J. Mamm., 37:449-450. . 1960. Cyclic changes in the morphology of the vulva and clitoris of Dipodomys. J. Mamm., 41:43-48. RAMSEY, P. R. 1969. Analysis of movement patterns in a population of Dipodomys ordii. Unpublished M.S. thesis. Texas Tech Univ. SANDERSON, G. C. 1966. The study of mammal movements: A review. J. Wildlife Mgt., 30:215-235. SCHMIDLY, D. J. 1968. Sexual, geographic, and individual variation in Dipodomys ordii from western Texas. Unpublished M.S. thesis. Texas Tech Univ. . 1971. Population variation in Dipodomys ordii from western Texas. J. Mamm., 52:108-120. SETZER, H. W. 1949. Subspeciation in the kangaroo rat Dipodomys ordii. Univ. Kansas Misc. Publ., Mus. Nat. Hist., 1:473-573. SOKAL, R. R., AND F. J. ROHLF. 1969. Biometry. W. H. Freeman and Co., San Francisco, xxi+ 776 pp. -4

28 GRADUATE STUDIES TEXAS TECH UNIVERSITY

STICKEL, L. F. 1946. The source of animals moving into a depopulated area. J. Mamm., 27:301-307. TAPPE, D. T. 1941. Natural history of the tulare kangaroo rat. J. Mamm., 22:117-147. TINKLE, D. W. 1967. The life and demography of the side-blotched lizard, Uta stansburiana. Misc. Publs. Mus. Zool., Univ. Michigan, 132:1-182. TINKLE, D. W., D. MCGREGOR, AND S. DANA. 1962. Home range ecology of Uta stansburiana stejnegeri. Ecology, 43:223-229. WENDORF, F. 1961. Paleoecology of the Llano Estacado. Mus. of New Mexico Press, Santa Fe, 144 pp.

A Copies of the following numbers of Graduate Studies may be obtained on an exchange basis from, or purchased through, the Exchange Librarian, Texas Tech University, Lubbock, Texas 79409.

No. 1 Pittard, K., and R. W. Mitchell. 1972. Comparative Morphology of the Life Stages of Cryptocellus pelaezi (Arachnida, Ricinulei), 77 pp., 130 figs...... $2.00

No. 2 Shoppee, C. W., ed. 1973. Excited States of Matter, 174 pp...... $5.00

No. 3 Levinsky, R. 1973. Nathalie Sarraute and Fedor Dostoevsky: Their Philosophy, Psychology, and Literary Techniques, 44 pp...... $2.00

No. 4 Ketner, K. L., 1973. An Emendation of R. G. Collingwood's Doctrine of Absolute Presuppositions, 41 pp...... $1.00

No. 5 Wirth, W. W., and W. R. Atchley. 1973. A Review of the North American Lepto- conops (Diptera: Ceratopogonidae), 57 pp...... $1.00

No. 6 Gillis, E. 1974. The Waste Land as Grail Romance: Eliot's Use of the Medieval Grail Legends, 26 pp...... $1.00

No. 7 Garner, H. W. 1974. Population Dynamics, Reproduction, and activities of the Kangaroo Rat, Dipodomys ordii, in Western Texas, 28 pp ...... $1.00 I

STECH

MA Y '