VAGILITY AND LOCAL MOVEMENTS OF POCKET GOPHERS (THOMOMYS AND PAPPOGEOMYS) IN AN ÁREA OF SYMPATRY by STEPHEN LORY WILLIAMS, B.S.
A THESIS IN ZOOLOGY
Submitted to the Graduate Faculty o£ Texas Tech University in Partial Fulfillment o£ the Requirements for the Degree o£ MASTER OF SCIENCE
Approved
December, 1973 73 • 1973 l\)o.7j\S ACKNOWLEDGMENTS Cl-õp • ^ I extend my sincere appreciation to my major advisor, Dr. Robert J. Baker, for his advice, encouragement, and assistance with many aspects of this study. I am indebted to Dr. Hugh H. Genoways for generously providing time and suggestions applicable to this study, and to Dr. Joe R. Goodin for providing laboratory equipment for soil analyses. I am also grateful to Drs. Robert J. Baker, Hugh H. Genoways, Robert L. Packard, and Joe R. Goodin for their criticai reviews of this manuscript. Finally, I acknowledge with gratitude financial assistance provided by the International Center for Arid and Semi- arid Land Studies, the Theodore Roosevelt Fund, and the Organized Research Fund of Texas Tech University.
11 TABLE OF CONTENTS
ACKNOWLEDGMENTS 11 LIST OF TABLES iv LIST OF ILLUSTRATIONS v I. INTRODUCTION 1 II. METHODS AND MATERIALS 3 III. RESULTS 13 IV. DISCUSSION 33 V. SUMMARY 43 LITERATURE CITED 45
111 LIST OF ILLUSTRATIONS
Figure Pa;.
1. Aerial photograph of the study área ....
2. Data form used in recording captures ... 6
3. Individual movements of Pappogeomys castanops and Thomomys bottae in the study área 16 4. Distribution of Pappogeomys castanops and Thomomys bottae in the study área . . 2 2 5. Distribution of Pappogeomys castanops and Thomomys bottae in the vicinity of the study área 24
6. Graphical representation of changes in mean soil moisture between June 1971 and May 1972 26
7. Extent of flooding in the study área on 23 August 1971 30 8. Burrow systems eroded by floods 32
V LIST OF TABLES
Table Page 1. Sample sizes, means, ranges, and standard errors of movement data, grouped according to age, sex, and species .... 14 2. Leveis of significance and F values derived from comparisons õT movement data grouped according to age, sex, and species 15
IV INTRODUCTION
Vagility, the potential or ability to disperse, is a characteristic influencing many biological aspects of a species. The rate and degree of population growth, gene flow, and morphological divergence are some features affected by vagility. Vagility, in turn, may be affected by gross morphology, behavior, physiology, natural history, and population dynamics of the species (Udvardy, 1969). Organisms with limited vagility encounter problems such as range expansion, maintenance of heterozygosity, and complete dependence on local environment. A group of mammals with low vagility are pocket gophers of the family Geomyidae (Patton ejt ai., 1972). These animais are specifically adapted to a herbivorous, fossorial habitus (Patton, 1972) and are restricted to a burrow system. Confinement to a burrow system obviously restricts the vagility of a species. The morphological adaptations of pocket gophers cause them to be awkward and vulnerable outside of their specific habitus. The location selected for this study was an área, described by Reichman and Baker (1972), where Thomomys bottae limpiae and Pappogeomys castanops pratensis occur sympatrically. This área is particularly applicable to this study because Reichman and Baker (19 72) had already noted movements occurring between these populations. In addition this locality was favorable because the habitat was relatively undisturbed, the áreas occupied by pocket gophers are restricted, and both genera could be studied simultaneously under similar environmental conditions. Because members of the family Geomyidae are not known to have subdivided the fossorial habitus (Russell, 1968), the role of vagility is particularly criticai when pocket gopher species are sympatric. METHODS AND MATERIALS
This study was conducted 9.0 km N, 9.5 km E Fort Davis, Jeff Davis Co., Texas, between May 1971 and May 1973. Intensive live-trapping (Baker and Williams, 1972) was conducted in an área which extended 250 meters on either end of the zone of sympatry. Trapping was occasionally extended to other áreas to determine the local distribution of both species. Traps were set in áreas where recent mounding activity was indicated. An aerial photograph (Fig. 1), obtained from the Department of Interior Geological Survey, was used to map the zone of sympatry and plot individual capture sites. The distance and direction of any movement could be determined by plotting capture sites on the aerial photograph. This method of recording localities proved to be reasonably accurate (+^ five meters) and easily utilized. In addition, it minimized the problems of working large áreas with irregular terrain. Each pocket gopher captured in the study área was coded according to species (P = Pappogeomys; T = Thomomys), sex (d*;?), sequential collecting number, and age (A = adult; S = subadult, juvenile). Specimens were " p. Cd B o w p a> p ti Cd • o 4-> iH to CO t*^ 3 Oi to O Cd -P.£3 to ^ -H Ü PJ u 0) ,£3 D Cd +J •H
P E r y •H MOL T EVIDE f SUBA D JUVEN I e ADUL T < -—' o 2-, O (T ( - ^ UJ 5 z ^—2 — o s d u_ ^—, ' • o J m o •^ ^ z —^ identified as being a juvenile, subadult, or adult. More specific age determinations were not possible. Weights were not used because of their questioned validity (Hansen, 1960). The home range of pocket gophers is e^sentially restricted to the burrow system (Turner ejt ai., 1973; Wilks, 1963). Because this study is primarily concerned with dispersai tendencies beyond the burrow system the actual área of the home range, as treated by other investigators (Howard and Childs, 1959; Turner ejt ai. , 1973; Wilks, 1963), is not criticai. However, the size of the burrow system, or home range at any specific time, is important in differentiating movements within a single burrow system and movements into new áreas. Movements within a single burrow system could be eliminated if the maximum length for a burrow system was known. An estimate of the size of a burrow system was obtained for each species by determining the maximum distance between capture sites of any individual during a 24 hour period. This time limit was used because it is unlikely that an individual pocket gopher will move into a new área, establish its territory, and build a new burrow system within such a short period of time. Therefore, distances less than the length of a burrow system, as determined above, are considered to be movements within a single 8 burrow system, whereas greater distances are treated as movements into new áreas. This value is arbitrary and subject to error because the actual size of a burrow system cannot be accurately determined without excavating the burrow system or using radio-telemetry.
Because pocket gopher habitat, in Limpia Canyon, is a narrow zone paralleling the creek, nearly ali movements could be recorded as either upstream or downstream from previous capture sites. A Chi-square goodness of fitness test (Sokal and Rohlf, 1969) was used to determine if direction of movements, either upstream or downstream, was random. If multiple movements of an individual showed opposing directions, the directional trend was determined as the difference between distances traveled upstream and downstream. Statistical analysis of individual movements were conducted using Power's UNIVAR program (Power, 1970). Individual movements were grouped according to age, sex, and species. Comparisons were made using average distance of ali movements, total distance of movements exceeding the prescribed burrow length, total distance of ali movements, and greatest distances between consecutive capture sites. The average distance indicates the possible size of the home range (Davis, 1953), whereas the total distance of movements exceeding the burrow length is intended to be an indication of actual dispersai tendencies. The total distance combines the movements in and out of the burrow system and provides a general evaluation of vagility. The greatest distance between consecutive trap sites, which is sometimes used to indicate home range size (Davis, 1953), was used as an indication of the distance that might be expected in a single movement. Population movements were determined by noting geographic location and numbers of captured individuais along the respective margins of either population. These observations were made on an annual basis because periods of extensive trapping were not sufficient to provide more detail. ^ Assuming the size of a burrow system equals the size of actual home range, it is possible to estimate the geographical limits of the population by establishing a theoretical home range for individual pocket gophers. The capture site is designated as the center of the home range and the length of the burrow system represents the diameter of the área.' Because the burrow may go any direction, the área of habitation is represented as a circle. Using this method for each capture site, an estimate of the habitat and distribution of both species was made and compared. These estimates indicate the 10 general distributional trends for the duration of the study. Changes in distribution were not determined because the trapping periods did not furnish sufficient detail.
Monthly precipitation records were examined from 1931 to the present to determine what effect rainfall may have on the local ecology. Records of precipitation were obtained from the statistical reports of the United States Weather Bureau (United States Department of Commerce, 1931-1972) for the weather station at Fort Davis
A monthly determination of soil moisture provided a means of testing selection of soil types by T^. bottae and £. castanops. Soil samples were collected for 12 months, beginning June 1971, at three depths (15, 30, and 45 centimeters) in ten localities distributed throughout the trapping área. In order to reduce sampling error, five samples were taken at each depth at each locality. The localities were selected according to capture sites of five P. castanops and five T. bottae and were assumed to represent suitable habitat. The following procedure was used to collect soil samples and determine moisture content. Five holes, exceeding 45 centimeters in depth, were dug at each locality. At each depth the exposed surface of the hole was scraped away in order to remove any soil which may 11 have already lost moisture. Immediately afterward, fresh soil was collected in a 15 x 45 mm glass vial and capped. The capped vial was sealed with plastic tape and kept as cool as possible in an air-tight container to reduce the loss of moisture. Within 24 hours the vials containing the soil were weighed to the nearest .001 gram. Subtraction of the vial weight provided the wet weight of the soil sample. Samples were then stored in an oven set at 110°C for 72 hours, allowed to cool, and were weighed again. Dry weights were obtained by subtracting vial weights from the resulting values. With the wet and dry weights determined, the moisture content was derived by dividing the wet weight into the difference of both weights. Percent moisture was obtained by multiplying the resulting value by 100. Care was taken to give the same treatment to ali samples. The values used in analyzing the soil moisture represent an average of the five samples collected from a specific site, depth, and month. Samples from a given set of variables of one species were compared statistically to corresponding samples of the other species using Power's UNIVAR program (Power, 1970). The effects of floods in the zone of sympatry were studied by actual observation and by examination of signs left by floods. Photographs were taken to document the 12 extent of flooding. During the floods soil permeability was tested by systematically digging a series of holes at various distances from the flood waters. These holes were dug to a depth below the water levei. The frequency and magnitude of the floods, recorded by a flood stage meter located in the study área, were obtained from the Water Resources Division of the United States Department of Interior, Geological Survey in San Ângelo, Texas. RESULTS
During this study 77 P. castanops and 122 T. bottae were marked in the zone of sympatry. Of these, 71 individuais were recaptured up to six times, resulting in 317 captures. About 65 other pocket gophers were collected outside of the study área. Movement data, based on age, sex, and species are presented in Table 1. The F value and levei of significance between groups are presented in Table 2. Data for subadult male Pappogeomys and total distance of the prescribed burrow length for subadult Thomomys were omitted because of insufficient data. Individual movements are plotted on a map of the study área (Fig. 3). The greatest distance measured for P. castanops in a 24-hour period was 50 meters (P ? 49) , whereas the greatest distance for T. bottae was 40 meters (T $ 23). These measurements were used as the prescribed burrow length for each species. Records of time required to make greater movements were more difficult to ascertain. Individuais of P. castanops made movements of 270 meters in 20 days (P d" 12) , 65 meters in eight days (P ? 12), and 100 meters in 21 days (P ? 27).
13 14
ArfP A$P S5P A<íT A9T S«/T S9T
e N 8 13 7 11 29 4 9
H5 CO 217.81 46.22 435.71 31.83 28.36 8.32 20.56 Q 7
RANGE S-495 5-100 5-270 10-81.7 0-120 0-20 0-50 LVERAG E ^ S.E. •69.22 •^8.91 •136.68 • 8.44 ^5.29 • 4.25 •4.44
u ((•ZX ou N S 6 5 4 9 0 0 (U (UZ
H (U % 7 391.00 85.00 185.00 112.50 74.44 0.00 0.00 CO o »H CO aí QHoí RANGE 13S-59S 60-100 65-340 55-215 45-120 0 0 VEME N H E B U OTA L HgH S.E. •82.88 • 7.30 •52.58 •36.08 ^7.97 • 0.00 • 0.00
N 8 13 7 11 29 4 9
7 256.88 60.38 137.14 66.36 44.14 8.32 20.56 5TANC E H4 Q ^ RANGE 5-625 5-130 5-350 0-245 0-150 0-20 0-50 < H O H S.E. •^85.87 •^10.35 •49.30 •21.32 •^7.39 • 4.25 M.44
(U Z N 8 13 7 11 29 4 9 >J cu CO u eu wtu Z ^ >H 7 253.12 55.00 135.71 45.45 34.48 8.32 20.56 S T I ECUT I : E BE l IR E S I RANGE 5-595 5-100 5-340 0-130 0-120 0-20 0-50 CON S STAN C CAPT l IREAT E Q S.E. +^83.63 +^9.00 •^48.28 •12.62 •,5.75 • 4.25 M.44
TABLE 1. Sample sizes (N) , means (7) ranges, and standard errors (S.E.) of movement data of pocket gophers grouped according to age (A = adult; S = subadult), sex (d',9) , and species (P = Pappogeomys; T = Thomomys). Means and ranges are given m meters 15
AJ9 A9P S9P A^T AíT S/r S?T
kiff 9.87^ 2.194 9.86^ 27.06, 4.364 9.16^
A?P 9.87^ 15.S9g 1.34j 3.264 S.225 S.O75
S«P 2.194 15.59, 14.09j 39.40, S.335 12.09g STANC E o A/r 9.86^ l.34j 14.09g 0.12^ 2.623 1.222 8 í A?T 27.06g 3.264 39.40, 0.12^ I.9O3 0.622 sar 4.364 5.22^ S.S35 2.623 I.9O3 2.783 AVE I sn 9.I67 5.075 12.09g 1.222 0.622 2.783
^54g 4.4O4 13 A((P 7.895 27.23, X X A«P 16.S4g 4.3I4 0.842 0.842 X X
S?P 4.4O4 4.3I4 I.IS2 7.87j X X •ANC E O F .XCEEDIN I r LENGT H A^ 7.89^ 0.842 I.IS2 ^"s X X ATT 27.235 0.842 7.87^ 2.I83 X X BURRO » KL DlS I MENT S E SA X X X X X X TH E TOT i MOVE I s?r X X X X X X
A^ s^.46y I.3S2 6.l7j 21.72, 3.994 8.S6 S«P I.3S2 4.O64 ^"j 11.33, 3.724 7.24^ TANC E AíTT 6.17^ 0.07j 2.2Sj 2.563 3.644 DI S >^"s A«T 21.72, I.SS3 11.33g I.S93 3.144 3.OI4 TA L g SfT 3.994 7.34^ 3.724 Z.S63 3.144 2.783 S«T 9.3O7 7.24^ 3.644 S.OI4 2.783 »S*6 A^ 9.14^ 1.372 8.30j 25.07, 4.O84 8.74^ A«P 9.14^ 4.82^ 0.40j 3.834 7.77^ 8.97y s«p 4.82^ 4.«35 IS.7S, 3.794 7.36j A^ 8.30g 0.40j 4.83^ 0.822 2.963 2.923 A«T 2S.07, 3.SS4 15.7S, 0.822 2.763 I.7O3 CAPTUR E SITE S SrfT 4.O84 7.77^ 3.794 2.963 2.763 2.783 ETWEE N CONSECUTIV E TES T SINGL E DISTANC S«T 8.74^ 8.97^ 7.36^ 2.92j I.7O3 2.783 TABLE 2. Leveis of significance (subscript) and F, values derived from comparisons of movements grouped according to age (A = adult; S = subadult) , sex (cf,^) , and species (P = Pappogeomys; T = Thomomys). Coding used in this table is as follows : O = 0.0; 1 = 0.25; 2 = 0.50; 3 = 0.75; 4 = 0.90; 5 = 0.95; 6 = 0.975; 7 = 0.99; 8 = 0.995; 9 = 0.999; "X" = unknown values due to insufficient data. 16 (D O •p to 4-> •H O CO 42 >> to 6 a> . O OHl (D C OC (U i^ O ÍH O ex O p-m to ctí M-l P. C-, •H 3 n3 O M-l fH O d) t>o +-> to O TJ 4J 2 (D C -M <6D Ctí- ^o > CJ t/) O ÍH -H :s rt x.r: Cd 13 +-> n33 +-3> 13 •H to C > Cd •H <Ü Ti p^í to tí "P Individuais of T. bottae moved 95 meters in 75 days (T $ 42), 65 meters in 29 days (T $ 54), 45 meters in 82 days (T $ 55), and 80 meters in 84 days (T ? 68). This study indicated that P. castanops makes greater movements than T. bottae (Fig. 3). Adult male Pappogeomys made the longest total movements (maximum 625 meters) followed by subadult female Pappogeomys and adult female Pappogeomys. The greatest total distance moved by T_. bottae was also by adult males (maximum 245 meters). The order of magnitude of movements following the adult male Thomomys was adult female Thomomys, subadult female Thomomys, and subadult male Thomomys. Some individuais of Thomomys did not show any movement during a period of several months. Statistical analyses of average distances show that adult male Pappogeomys have significantly larger (P > .95) movements than any other group except subadult male Thomomys. The home ranges of adult female Pappogeomys were significantly greater than male and female subadult Thomomys and significantly less than subadult female Pappogeomys. There were no significant differences between the Thomomys groups. Notable differences (P = .90) were also noted between adult male Pappogeomys and subadult female Pappogeomys, adult male Pappogeomys and subadult male Thomomys, and adult female 18 Pappogeomys and adult female Thomomys. The greatest similarity of movements occurred between adult female Pappogeomys and adult male Thomomys. Analyses of the total movements exceeding the burrow length indicate adult male Pappogeomys have significantly greater movements than ali other groups except subadult female Pappogeomys. Subadult female Pappogeomys moved significantly greater distances than adult female Thomomys. No other criticai differences were indicated. Major differences (P = .90) in distances moved were noted between adult male Pappogeomys and subadult female Pappogeomys, and subadult female Pappogeomys and adult female Pappogeomys. Analyses of the total distance moved indicate that adult male Pappogeomys have significantly greater movements than any other group with the exception of subadult female Pappogeomys and subadult male Thomomys. The total distances of subadult female Pappogeomys are significantly greater than the adult and subadult female Thomomys. Movements of adult female Pappogeomys were not significantly greater than any group except subadult male and female Thomomys. The only significant differences found among the Thomomys groups were between adult and subadult females. In this case the adults had greater movements than did the subadults. In addition to these 19 results, several cases of notable differences (P = .90) were observed (see Table 2 for detail). The greatest similarity (P = .25) of total movements was found between adult female Pappogeomys and adult male Thomomys. Analyses of the greatest distance between consecutive capture sites indicate adult male Pappogeomys move significantly greater distances than do the other groups with the exception of subadult female Pappogeomys. The movements made by subadult female Pappogeomys were significantly greater than those of adult and subadult male Thomomys. Adult female Pappogeomys moved significantly greater distances than male and female Thomomys. There were no differences between the Thomomys groups. High leveis of difference (P^ = .90) in movements occurred between adult female Pappogeomys and adult female Thomomys, and subadult female Pappogeomys and subadult male Thomomys. The greatest similarity (£ = .25) occurred between adult female Pappogeomys and adult male Thomomys. In many áreas, movements of pocket gophers allowed replacement of individuais of the same species and reutilization of burrow systems. This replacement was first noticed when a Pappogeomys (P ? 16) was killed in a trap. One week later this burrow was occupied by another Pappogeomys (P ^ 12) which had previously been 20 captured 270 meters away. Eight months later another Pappogeomys (P $ 5) moved 45 meters to the same site. During this study there were at least three other cases of replacement in Pappogeomys. At one site a Thomomys (T $ 32) was trapped in July 1971. One month later another Thomomys (T ? 35) was collected at the same site. It was discovered in March 1972 that T ? 32 had moved 85 meters upstream. However, by this time the collecting site of both specimens was occupied by another Thomomys (T $ 58) . At least three other conspecific replacements occurred between individuais of Thomomys. Some trap sites produced both species at different times during the study but there was no evidence to suggest that either utilized the burrow system of the other. A Chi-square goodness of fitness test was used to determine if the direction of movement, upstream or downstream, was random. A majority of the Thomomys (N = 46, X^ = 0.82) moved downstream, but the difference in direction was not significant. Individuais of Pappogeomys (N = 25, X = 4.84) showed a significant (P > .95) directional trend downstream. Throughout the zone of sympatry there were regions which were never occupied by either species. Other áreas were inhabited by either one or both species. In some instances groups of individual pocket gophers were 21 isolated from the main population (Fig. 4). Fluctuations in population boundaries were noted but revealed no unusual directional trend in either population. There was no notable gain or loss of territory by £. castanops in 1971 or 1972. A recession of approximately 130 meters in the northern boundary was noticed in January 19 73. The Southern boundary of T. bottae remained basically the same throughout 1971, 1972, and possibly 1973. In January 19 73 an unmarked individual (T d* 2 8) was collected from a site which represented the southernmost record since 1968. This individual represents either a southward movement of 245 meters, or is a specimen which escaped previous trapping efforts. This indicates that the southern boundary of T. bottae has not changed since 1968. Aside from this individual, the southern boundary of T^. bottae receeded approximately 45 meters between 1971 and 1973. Distributional investigations in the Davis Mountains revealed that £. castanops had a more extensive distribution than was previously believed (Reichman and Baker, 1972). Pappogeomys was collected above the cliffs bordering the zone of sympatry, thus increasing the chorology of this species 180 meters in this área. This species was also collected at the northern boundary, 19.0 km N, 13.4 km E Fort Davis, of the T. bottae ^iW^pr 22 tfl >> B O B o ,cHf W O Td P: >N rt "p •H t/) í3 P- d o 'H c P ca e: +j 0 t/) Ü CtJ U) Ü iH TS to >>lO 6 -a o 4-> o tsOiO) o •P pJO CtpfJ e V)• eu 0 • «H m ctj u o (D (D ?-• P. cí rt tn o •H >s^ +J'T3 +-> ^30 Xi •^'^ •H tfl U tí P (D 'H t/) .C • H "P t/) Q tí c 0 • H -H P ^ (D Cd zi 1—( 0 4-> :3 ÍH P P. 3 0 0 taous P. •H PU 23 population (Fig. 5). It is estimated that during the two years of this study, T. bottae has inhabited 72.6 percent of the suitable fossorial habitat in the zone of sympatry, whereas ?_, castanops has occupied 59.6 percent. The difference between these values indicates that the actual zone of sympatry between the two species is only 13.0 percent of the suitable habitat. Although the área occupied by £. castanops is notably less than that of T^. bottae, P^. castanops occupies more área (approximately 50 percent greater) per individual than T. bottae. During this study the population size of T. bottae remained relatively stable. Sex ratios always favored females with ratios of 2.4:1 in 1971, 3.4:1 in 1972, and 2.8:1 in 1973. Pappogeomys showed a drop in population density. Sex ratios in this species also favored females with ratios of 1.8:1 in 1971, 2.6:1 in 1972, and 2.7:1 in 1973. The ratio of individuais of T. bottae to individuais of £. castanops increased as the population size of £. castanops dropped. In 1971 and 1972 the ratio of Thomomys to Pappogeomys was 1.2:1. This ratio increased to 3.0:1 in 1973. Examination of local ecological conditions indicated that precipitation during 1971 resulted in 24 104'50 Pappogeomys castanops . •i 1111IIII« Thomomys bottae 1 234 56789 méy 3040' Figure 5. Distribution of Pappogeomys castanops and Thomomys bottae in the vicinity ot the study área. Arrow indicates location of study área. 25 one of the driest years since the middle 1950»s. In 1971 there was no substantial amount of precipitation until July. Most rains came in August and progressively decreased until the end of the year. Annual precipitation in 1971 was 27.99 centimeters. Annual precipitation during 1972 was more characteristic of a normal year (mean annual precipitation obtained from available records since 1931 is 38.4 cm), with the largest rains occurring in May, August, and September. Annual precipitation was 47.40 centimeters. Generally, annual precipitation patterns corresponded to annual fluctuations in soil moisture. Monthly means of soil moisture at ali depths were usually higher in áreas inhabited by T^. bottae than in áreas inhabited by JP. castanops (Fig. 6). The degree of difference between representative sites decreased as depth increased. At the 15 centimeter depth, high leveis (P > .90) of difference were found during every month except September. At the 30 centimeter depth significant differences (P_ > .90) in soil moisture were noted in ali months except October, December, January, and May. At 45 centimeters only samples taken in June, November, and February were significantly different (£ > .90). The mean burrow depth (Thomomys = 18.6 centimeters; Pappogeomys = 25.7 26 'd d) c: Cd • •9 P to to B •H to 0 u O P -H o P T $ 23, T $ 26, T $ 37, T $ 55) with burrow systems in the flood zone survived the floods of 1971 and 1972. Fourteen other individuais (T cí 10, T T $ 45) revealed similar burrow systems. The magnitude of flooding was measured by the discharge of water (cubic meters per second = cms). It is estimated that floods exceeding a discharge of 1.35 cms would threaten many pocket gophers inhabiting the banks of Limpia Creek. This estimate is based on observations of flood waters on 23 August 1971 (6.91 cms), 24 August 1971 (1.73 cms), 28 May 1972 (.73 cms), and signs left from floods which were not witnessed. Flooding of this magnitude (1.35 cms) occurred 20 times during the two year period of this study. Flooding was almost entirely restricted to the months of July, August, and September. The discharge reported on 23 August 1971 (Fig. 7) was exceeded three other times the following year on 15 August (7.53 cms), 28 August (8.64 cms), and 21 September (10.42 cms). The latter discharge has been exceeded only three times since records were started in 1962. The greatest discharge (14.82 cms) was reported on 22 August 1966. Larger floods prior to these records were reported in 1932 and 1946. Effects of floods differed during the two years of this study. Following the floods of 1971 several áreas occupied by pocket gophers showed extensive erosion of burrow systems. Occasionally several square meters and as many as three leveis of tunnels were exposed by this 30 1^ CTi f-i •P to 3 to 3 < ro CM C O Cd 0 h Cd ÍK t3 3 -P to 0 X p P . •H /^—\ to MX P P. •H Cd 13 U O ÜO roH po MH o X m P4 o e p o p ÍH 0 MH P X13 W 0 o 3 » T3 r^ ^o 0 P. ^t 0 D CcS taOv- ^ IX, 31 erosion (Fig. 8). In 1972, little or no erosion was noticed after the floods. The primary difference between the two years was the amount of vegetation on the banks of the creek. Although there were larger and more numerous floods in 1972, there was actually less erosion because of increased vegetation in the flood zone. This vegetation was probably brought about by the earlier rains in May 19 72, which were lacking in 19 71. Burrow systems, exposed by erosion, were usually not reoccupied by pocket gophers until revegetation occurred. 32 V) •d o o iH MH >s 'd 0 Tá o VI 0 (O 0 4.» (O >s V) O ^1 ^1 pq 00 0 U •H \íi DISCUSSION The vagility of P_. castanops is indicated by its ability to travei great distances in short periods of time. Because of the large distances between capture sites and the irregular terrain of the study área, it is assumed that £. castanops moves into new áreas by traveling overland. This study showed adult males and 0 subadult females have the greatest dispersai tendencies of ali the groups studied. It is suspected that movement patterns of subadult males, which are presently unknown, will be comparable to movements of adult males and subadult females. In addition to extensive movements, adult male and subadult female Pappogeomys have exceptionally large home ranges as indicated by their average movements. The sizes obtained in this study may be exaggerated by the small sample size of these groups. The cause for directional movements in Pappogeomys is not known. This trend may result from population pressure further up the canyon. Intraspecific behavior in Pappogeomys may cause individuais to be more successful in moving into áreas occupied by Thomomys than into conspecific áreas. Because Thomomys is more 33 34 successful in moving into áreas occupied by other Thomomys, it is suggested that Pappogeomys may be interspecifically more aggressive, and therefore behaviorly superior to Thomomys. It is unlikely that a decrease in population pressure would cause the loss of área by Pappogeomys which was noted in 1973. Such a recession might be expected when the population density in an área is not large enough to absorb the loss of individuais by death of dispersai. This study suggests that vagility in T. bottae is relatively low. There was a remarkable similarity in the limited dispersai tendencies among different age and sex groups. It might be expected that one sex would show greater vagility as a result of rutting, but this was not the case. Also, it would seem likely that subadults would show greater vagility if they are to find suitable habitat for themselves. Data indicate that subadults had shorter movements*than adults, and therefore may remain in the parental burrow system until they mature. The burrow length of 40 meters for Thomomys approximates the expected length of the home range. This length included home range sizes determined by average distances in this study and home range sizes found in other studies (Howard and Childs, 1959; Inglês, 1952; Turner e^ ai., 1973). Howard and Childs (1959) 35 suggest most movements, outside of the home range, by T. bottae in Califórnia do not represent movements into entirely new áreas. Instead, they represent boundary shifts by employing reutilization and abandonment of portions of burrow systems. Their study noted movements, or boundary shifts, up to 120 meters. Vaughan (1963) reported mean maximum movements of 135 meters in T. bottae introduced into new áreas. The study in Limpia Canyon revealed dispersai patterns of T. bottae to be comparable to studies by Howard and Childs (1959) and Vaughan (1963). The greatest single distance recorded from !_, bottae was 130 meters (T d" 5) . Because extensive movements required considerable time, burrowing activities may represent the primary means of dispersai for T. bottae in Limpia Canyon. If 1^. bottae does not actually make movements into entirely new áreas, then the isolating tendencies, noticed in this study are more readily explained. Not only were groups isolated from each other and the main population (Fig. 4), but also there was no exchange of individuais between these groups (Fig. 5). Apparently these isolating tendencies are common in Thomomys populations (Hansen, 1960; Patton and Dingman, 1968; Thaeler, 1968; Vaughan, 1967). With such tendencies so apparent on a small scale it is conceivable how a lack 36 of gene flow would eventually result in a high degree of subspeciation as seen in the genus Thomomys (Hall and Kelson, 1959). Because individuais of T. bottae replace each other and the local distribution of this species is occasionally discontinuous and confined to áreas around Limpia Creek, it is suggested that many individuais are restricted by the availability of suitable habitat. Habitat selection in pocket gophers is well documented. Inglês (1949) noted seasonal changes in pocket gopher distribution due to ground water and snow. The distribution of species and local populations have been explained on the basis of soil types (Davis e^ ai. , 1938; Davis, 1940). Judd and Reichman (1972) suggested the necessity of succulent vegetation for some geomyid rodents. In Limpia Canyon soil moisture was a criticai factor to a depth of 30 centimeters for T. bottae. Because soil moisture is affected"by several variables such as precipitation, temperature, runoff, vegetation, and soil type, consistent records of high moisture leveis stresses the possible importance of this factor in Thomomys. It is not certain whether Thomomys was selecting soil with high moisture or selecting associated vegetation or soil types. It is assumed that soil moisture may be the most important factor because similar vegetation and soil types occurred 37 in áreas not inhabited by Thomomys. The success of £. castanops in replacing T. bottae is primarily due to its broader habitat requirements and greater vagility. Previous to this study very little was known about these characters in P. castanops. Judd and Reichman (19 72) pointed out that V_. castanops is capable of successfully utilizing less succulent foods than T. bottae. Reichman and Baker (19 72) noted P. castanops occurred in a wide range of habitats, ranging from the mesic habitat of T. bottae to more xeric habitats away from the creek. The local distributions of V_. castanops outside of the study área and soil analyses of characteristic habitats further suggests the ubiquitous nature of this species. ' Pappogeomys castanops is capable of replacing T. bottae because of its greater vagility and wider selection of habitats. The attributes of P. castanops would probably allow replacement of Thomomys by direct competition with sufficient time. Range extension is probably achieved when P. castanops invades áreas left unoccupied by T. bottae (Reichman and Baker, 1972). Because the distribution of both species has been relatively stable, other factors must be promoting the replacement of Thomomys. Reichman and Baker (1972) discussed how differential reproduction and the 38 association of floods and drier climates might contribute to this trend. Another possible mechanism occurs when > 11 burrow systems on the banks of Limpia Creek are exposed by flood erosion. Because T. bottae generally has shallower burrow systems it is probably affected more by the floods than P^. castanops. Even though most individuais survive the floods, T^. bottae is not apt to reclaim eroded áreas, thus subsequent occupation of £.* castanops is possible. Such erosion is characteristic of years without spring rains to establish vegetation cover to curtail the forces of the floods. Precipitation records indicate that years with low spring precipitation are relatively frequent. Another important factor in the success of P^. castanops replacing T. bottae is the population size and structure at times when the opportunity to claim more territory exists. Because most movements of ?_. castanops are made by adult males and subadult females, success of colonization not only depends on the population size but also the number of individuais included in these groups. Colonization by £. castanops would probably be most successful when a large population with a high ratio of adult males and subadult females exists. Such a combination of factors would provide expansion pressure and a means for colonization. The degree of success 39 would probably vary according to the total amount of dispersai pressure in the Pappogeomys population working against the total amount of dispersai pressure in the Thomomys population. Since T. bottae does not demonstrate high vagility as a species, or individually by age or sex, its population size and structure is not important when considering its impact on P. castanops. Although the suggested mechanisms may have a definite effect on the ability of P. castanops to replace T. bottae, their impact on past displacement is questionable. It appears that there was some degree of sympatry between both species as early as 1937. At this time, P. castanops was collected 1.6 km N Fort Davis (Blair, 1940) and T. bottae was reported from 1.6 km N Fort Davis (Blair, 1939), 1.6 km NW Fort Davis, and 3.2 km NW Fort Davis (Blair, 1940). Blair (1940) pointed out the "stream-bed association" of T. bottae and "short grass association" of £. castanops which still persists today. Data taken from study specimens at the Texas Wildlife Research Collection (Texas A § M University, College Station, Texas) suggest the geographical distribution of both species was unchanged at least until 1941. Davis and Beuchner (1946) reported collecting topotypes of T. bottae and indicated, from distributional descriptions of ?_, castanops, that both species were still sympatric. 40 However, in 1968 the geographical distribution of T. b. limpiae was reported to have changed to 9.0 km N, 9.5 km E Fort Davis, indicating a recession of approximately 13 kilometers. In spite of this change, T. bottae and £. castanops are still sympatric and maintain the same ecological associations (Reichman and Baker, 1972)\ The present study has indicated no major changes in the geographical distribution or ecological associations of either species. To explain the present location of T. bottae in Limpia Canyon it would be necessary for this population to receed over 400 meters each year. Because this has not been the case between 1937 and 1941, and 1968 and 1973, it is doubtful that the described mechanisms have played a major role in past displacement. The population of T. bottae has probably been sporadically distributed along Limpia Creek in the past. Great distances between colonies would facilitate greater recession rates if isolated populations die out. The occurrence of such colonies is supported by the isolated groups of pocket gophers in the study área. Assuming that isolated colonies once existed, the cause for their disappearance is still not explained. The current distribution of Thomomys in Limpia Canyon has probably been the result of one or more periods of severe environmental conditions. Several events have 41 occurred since 1941 that may have contributed to the present conditions. Although there were no significant fluctuations in population boundaries in 19 71, xeric conditions may become more criticai if the condition persists. This may have been the case during the drought of the early 1950's. Annual precipitation was exceptionally low for six consecutive years. During this time it is likely that T. bottae suffered greatly from the prolonged lack of suitable vegetation and habitat. Similarly, the effect of normal flooding may become more crucial by larger floods. It is feasible that floods of a much greater magnitude could easily eliminate individual pocket gophers and possibly small isolated populations. In 1946 there was extensive flooding in Limpia Canyon. This extensive flooding was repeated, to a lesser degree, in 1966. Undoubtedly, there are other factors which make it possible- for Pappogeomys to replace Thomomys. The success of replacement depends on the importance of individual factors or combination of factors. In áreas where Pappogeomys occurs with, or has replaced, other pocket gopher populations (Baker ejt ai., 1973; Hall and Villa, 1949; Russell, 1968; Williams and Baker, 1973), the importance of these factors may differ considerably. 42 The área of sympatry between £. castanops and T. bottae in Limpia Canyon is still of considerable interest Further investigation could provide more Information about this unique situation. If analysis of this problem is continued, care should be taken so that the future of these pocket gophers is not influenced or jeopardized. I 4 \ SUMMARY A two year live-trap study in the Davis Mountains of Texas was conducted in a zone of sympatry between £. castanops and T^. bottae. Specimens were marked, released, and recaptured. Specific ecological factors affecting the distribution of both specie.s were studied. The primary findings of this study are: 1. Thomomys bottae demonstrated low leveis of vagility. There were no unusual differences in dispersai ability attributed to sex or age. The lack of vagility may contribute to the isolating tendencies, thus the high levei of subspeciation, observed in this species. 2. Pappogeomys castanops demonstrated high leveis of vagility which were primarily expressed by adult males and subadult females. 3. Pappogeomys castanops is able to adapt to a wider range of habitats. Thomomys bottae is generally restricted to áreas of high soil moisture. 4. Individual movements of Pappogeomys castanops showed a significant directional trend toward the population of Thomomys bottae. Individual movements of Thomomys bottae showed no distinct directional trends. 43 44 5. The population size and structure of Pappogeomys and flood erosion of Thomomys habitat may contribute to the replacement of T. bottae by £. castanops. 6. There are probably several factors allowing the replacement of these pocket gophers, but the main cause of major displacement is believed to have been severe environmental conditions in the form of phenomenal flooding or aridity. LITERATURE CITED Baker, R. J. and S. L. Williams. 1972. A live trap for pocket gophers. J. Wildlife Mgt., 35: 1320-1322. Baker, R. J., S. L. Williams, and J. C. Patton. 1973. Chromosomal variation in the Plains Pocket Gopher, Geomys bursarius major. J. Mamm., 54: 775-779. Blair, W. F. 1939. New mammals from Texas and Oklahoma, with remarks on the status of Thomomys texensis Bailey. Occ. Papers Mus. Zool., Univ. Michigan 403: 1-7. . 1940. A contribution to the ecology and faunal relationships of the mammals of the Davis Mountains Region, southwestern Texas. Misc. Publ. Mus. Zool., Univ. Michigan 46: 1-39. Davis, D. E. 1953. Analysis of home range from recapture data. J. Mamm. 34: 352-358. Davis, W. B. 1940. Distribution and variation of pocket gophers (genus Geomys) in the southwestern United States. Texas Ag. Expt. Sta. Buli. No. 590: 1-38. 45 46 Davis, W. B. and H. K. Buechner. 1946. Pocket gophers (Thomomys) of the Davis Mountains, Texas. J. Mamm. 27: 265-271. Davis, W. B., R. R. Ramsey, and J. M. Arendale, Jr. 1938. Distribution of pocket gophers (Geomys breviceps) in relation to soils. J. Mamm. 19: 412-418. Hall, E. R. and K. Kelson. 1959. The mammals of North America. Ronald Press, New York, 1: xxx + 1-546 + 79. Hall, E. R. and R. B. Villa. 1949. An annotated check list of the mammals of Michoacan, México. Univ. Kansas Publ., Mus. Nat. Hist., 1: 431-472. Hansen, R. M. 1960. Age and reproductive characteristics of mountain pocket gophers in Colorado. J. Mamm. 41: 323-335. Howard, W. E. and H. E. Childs, Jr. 1959. Ecology of pocket gophers with emphasis on Thomomys bottae mewa. Hilgardia 29: 277-358. Howard, W. E. and L. G. Inglês. 1951. Outline for an ecological life history of pocket gophers and other fossorial mammals. Ecology 32: 537:544. 47 Inglês, L. G. 1949. Ground water and snow as factors affecting the seasonal distribution of pocket gophers, Thomomys monticola. J. Mamm. 30: 343-350. . 1952. The ecology of the Mountain Pocket Gopher, Thomomys monticola. Ecology 33: 87-95. Judd, F. W. and 0. J. Reichman. 1972. A comparison of the water requirements of the pocket gophers, Pappogeomys castanops and Thomomys bottae. Southw. Nat. 17: 302-307. Miller, M. A. 1946. Reproductive rates and cycles in the pocket gopher. J. Mamm. 27: 335-358. Patton, J. L. 1972. Patterns of geographic variation in karyotype in the pocket gopher, Thomomys bottae (Eydoux and Gervais). Evolution, 26: 574-586. Patton, J. L. and R. E. Dingman. 1968. Chromosome studies of pocket gophers, genus Thomomys I. The specific status of Thomomys umbrinus (Richardson) in Arizona. J. Mamm. 49: 1-13. Patton, J. L., R. K. Selander, and M. H. Smith. 1972. Genetic variation in hybridizing populations of gophers (Genus Thomomys). Syst. Zool., 21: 263-270. 48 Power, D. M. 1970. Geographic variation of red-winged blackbirds in Central North America. Univ. Kansas Publ., Mus. Nat. Hist. 19: 1-83. Reichman, 0. J. and R. J. Baker. 1972. Distribution and movements of two species of pocket gophers (Geomyidae) in an área of sympatry in the Davis Mountains, Texas. J. Mamm. 53: 21-33. Russell, R. J. 1968. Revision of pocket gophers of the genus Pappogeomys. Univ. Kansas Publ., Mus. Nat. Hist. 16: 581-776. Sokal, R. R. and F. J. Rohlf. 1969. Biometry. W. H. Freeman and Company. San Francisco, xxi + 1-776. Thaeler, C. S. 1968. An analysis of the distribution of pocket gopher species in northeastern Califórnia (genus Thomomys). Univ. Califórnia Pub. Zool. 86: 1-46. Turner, G. T., R. M. Hansen, V. H. Reid, H. P. Tietjen, and A. L. Ward. 1973. Pocket gophers and Colorado mountain rangeland. Buli. Colorado State Univ. Expt. Sta. 554S: ix + 1-90. U. S. Department of Commerce. 1931-1972. Climatography of the United States. U. S. Government Printing Office, Washington, D. C. 49 Udvardy, M. D. F. 1969. Dynamic zoogeography. Van Nostrand Reinhold Company. New York, xviii + 1-445 Vaughan, T. A. 1963. Movements made by two species of pocket gophers. Amer. Midland Nat., 69: 367-372. . 1967. Two parapatric species of pocket gophers Evolution 21: 148-158. Wilks, B. J. 1963. Some aspects of the ecology and population dynamics of pocket gophers (Geomys bursarius) in southern Texas. Texas J. Sei. 15: 241-283. Williams, S. L. and R. J. Baker. In press. Geomys arenarius. Mammalian Species.