SOCIAL FACTORS INFLUENCING SPATIAL DISTRIBUTION IN POPULATIONS OF PRAIRIE DEERMICE By C. RICHARD TERMAN AN ABSTRACT Submitted to the School for Advanced Graduate Studies of Michigan State University of Agriculture and Applied Science in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Zoology 1959 Approved_____ _____________________________________ _____ C. Richard Terman ABSTRACT Factors influencing the geographical distribution of a species are effective through the local population. There­ fore, the study of spatial distribution within local popu­ lations offers a "dynamic" approach to problems associated with the spread of a species over its broad geographical range. This study combined laboratory and field techniques in order to examine the importance of social factors to the spatial distribution of local populations of Prairie Deer— mice (Peromyscus maniculatus bairdii). The objectives were: 1. To study the ©ffect of social manipulation of deer— mice populations in the laboratory upon their sub­ sequent spatial distribution in a semi—natural environment. 2. To examine behavioral patterns effecting spatial distribution in free ranging, semi-natural popu­ lations. 3. To measure the individual awareness of,and pref­ erence for, the home area as well as the stability of such an area. Successive populations of deermice were raised from weaning age (21 days) in isolation or in groups. At ten weeks of age, four bisexual pairs from each social treat­ ment were systematically released into different one—half C. Richard Terman acre "mouse—proof" plots. There were 8 experimental periods between June 6 and November 21, 1958, during which 8 suc­ cessive populations of four bisexual pairs were living in each of the two plots. Following release into the plots, the daily occurrence of the mice in nest boxes was recorded. Seventeen days after release, all mice were taken to the laboratory for 36 hours. Each was then reintroduced into its home plot at a point distant to its previously established home area. The location of each mouse for the next three days was re­ corded. In the last 3 experiments, young alien mice were retained in one-half the nest boxes during the first night after reintroduction of the residents. At the conclusion of each three week experimental period, all field experi­ m e ntal and their laboratory controls were killed, weighed, and the adrenals removed and also weighed. Other dynamics of the populations were measured by two periods of live trapping, by recording the time and fre­ quency of feeding activity, and by direct observation of social interaction between residents and aliens. The "mouse-proof" field proved to be an effective method for the study of population dynamics under semi-natural conditions while maintaining a measure of the controlled conditions which are possible in the laboratory. Phenomena which may be specific to populations of prairie deermice, to mice of the species Peromyscus maniculatus; or which may have general significance to small mam­ mal populations were suggested by this study. 0. Richard Terman a. Between 55 and 60 per cent of the mice moved to different nest boxes each night. Less than half of these moves were to boxes previously unvisited by the mouse mov­ ing. These data indicate that each mouse maintained several refuges and/or nest sites rather than a single one around which its activity centered. b. During the breeding season, the prairie deermice in this study were generally found alone or in bisexual pairs and rarely occurred in a nest box with an animal of the same sex. c. Mice of the opposite sex succeeded each other in nest boxes on successive days significantly more often than those of the same sex. Females followed males into boxes significantly more often and males followed females sig­ nificantly less often than expected by chance. d. No reliable evidence for territoriality was ob­ tained. The data indicated that animals of the same sex were spatially segregated as a result of a negative re­ pulsive force, hypothetically, avoidance. e. Mice of both social treatments homed significantly more often than expected by chance. Such performance indi­ cated that the individual mice recognized both their "own” nest boxes and those of their neighbors. Thus, a spatial distribution framework or "positional stability" may be a characteristic of local populations. f. Mice of both social treatments homed significantly less often to nest boxes temporarily occupied by young aliens than they did in the experiments prior to the introduction C. Richard Terman of aliens. This decrease in homing supports the hypothesis that the spatial distribution of prairie deermice may be achieved through mutual avoidance of individuals. Extension of the range of the local population as well as of the geo­ graphical distribution of the species may be largely by a diffusion-like process. Young mice, upon leaving the home nest site may move into an occupied area or into a temporar­ ily empty nest site rather than moving long distances until an unoccupied area is found. Such behavior could cause a partial displacement of the residents due to avoidance and result in a gradual extension of the range at the periphery. The differential effects of the social treatments were as follows: a. The isolation raised mice combined with others less often than the group raised mice; were slower in combining; and generally maintained a greater distance from their fellows. b. The isolation raised mice homed significantly more often than group raised during the experiments when homing was established as a phenomenon. The introduction of aliens into one-half the nest boxes had a more adverse affect upon the homing performance of isolation raised mice than of group raised. c. The isolation raised mice appeared to be less soc­ iable and more spatially oriented than group raised. d. The data suggest that spatial patterns of distrib­ ution existent in the plots were largely determined by social interaction and were of greater significance to isolation raised mice than to group raised. Isolation raised mice G. Richard Terman adapted less easily to changes in the social and related spatial stimuli than the more sociable group raised mice and, thus, more frequently returned to the earlier estab­ lished spatial patterns. The introduction of aliens dis­ rupted the social-spatial equilibrium existent in the plots. This disruption had a more severe and longer lasting affect upon isolation raised mice than upon group raised due to the inability of the former to quickly adapt to the environ­ mental changes. e. Differences in social behavior have been shown to be important factors determining spatial patterns of dis­ tribution within local populations. The importance of social factors to the evolution, genetics, and dynamics of popu­ lations is, therefore, evident. SOCIAL FACTORS INFLUENCING SPATIAL DISTRIBUTION IN POPULATIONS OF PRAIRIE DEERMICE By C o RICHARD TERMAN A THESIS Submitted to the School for Advanced Graduate Studies Michigan State University of Agriculture and Applied Science in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Zoology 1959 ProQuest Number: 10008580 All rights reserved INFORMATION TO ALL USERS The quality of this reproduction is dependent upon the quality of the copy submitted. In the unlikely event that the author did not send a complete manuscript and there are missing pages, these will be noted. Also, if material had to be removed, a note will indicate the deletion. uest. ProQuest 10008580 Published by ProQuest LLC (2016). Copyright of the Dissertation is held by the Author. All rights reserved. This work is protected against unauthorized copying under Title 17, United States Code Microform Edition © ProQuest LLC. ProQuest LLC. 789 East Eisenhower Parkway P.O. Box 1346 Ann Arbor, Ml 4 8 1 0 6 -1 3 4 6 ACKNOWLEDGMENTS The author wishes to express his sincere appreciation to Dr* John A. King, Roscoe B. Jackson Laboratory, Bar Harbor, Maine, who provided facilities, invaluable suggestions and criticisms without which this study would have been im­ possible# Grateful acknowledgments are also due to Dr. Don W. Hayne, Institute of Fisheries Research, Michigan Depart­ ment of Conservation, advisor during the major part of the author’s graduate work, who stimulated the author’s interest in population phenomena and with whom many of the ideas in­ corporated in this research were discussed? and to Dr. James C. Braddock, Department of Zoology, who was chairman of the writer's graduate committee during the research and the preparation of the dissertation. Dr. Braddock offered numerous valuable suggestions during the course of the re­ search and provided many helpful criticisms and ideas during the editing of the manuscript. The author is greatly indebted to Dr. Philip J. Clark, Department of Zoology, and to Dr. Walter C. Stanley, Jackson Laboratory, for their considerable aid in the statistical analyses. Sincere thanks are due to Dr. John Paul Scott, Director of the Behavior Division of the Jackson Laboratory and to Dr. John L. Fuller, Assistant Director for Training of the Jackson Laboratory, for their interest in the research and their many helpful suggestions; to other members of the re­ search staff of the Behavior Division and to the general ii staff of the Hamilton Station, for aid in many ways; to Dr. Charles D. Richards, Department of Botany and Plant Pathology, University of Maine, who made final identifi­ cation of the plants collected. The writer is grateful to the graduate committee mem­ bers, Dr. Karl A. Stiles, Head, Department of Zoology; Dr. George J. Wallace, Department of Zoology; and Dr. John E. Cantlon, Department of Botany and Plant Pathology; for numerous suggestions and advice. He wishes to thank Mrs. B. R. Henderson, secretary, Department of Zoology, for kindnesses and aid in many ways. The financial support provided by a graduate teaching assistantship in the Department of Zoology for most of the author's graduate training is gratefully acknowledged. This research was made possible by the R. B. Jackson Laboratory, Bar Harbor, Maine, which provided funds through grant C. R. T. 5013 while the author was a Pre— doctoral Research Fellow at the laboratory. Deepest gratitude is due the author's wife, Phyllis McAdam Terman, who assisted in the collection of data, aid­ ed in the typing and preparation of the manuscript, and pro­ vided encouragement throughout the study. TABLE OF CONTENTS ACKNOWLEDGMENTS LIST OF TABLES vii ................................. ix .................................... 1 Biological Importance of Spatial Patterns to Populations ............ •.......... Species evolution ........ 1 2 Factors Affecting Spatial Distribution ........ 12 ........................ 15 ....................................... 18 STATEMENT OF THE PROBLEM Experimental Animals ......................... ............... Laboratory Facilities Experimental Field Design 20 ............. 20 ................ Vegetation analysis 20 ............... 29 ...................................... Laboratory Procedures Social treatments ........... Group raised mice .................. ....... Preparation of the micefor release Field Procedure 38 39 39 ...••••. ...................... Preparation of the experimental field iv 38 38 ......................... Isolation raised mice 18 19 •••«.••......... General description PROCEDURES 1 ......................... Population control MATERIALS PAGE ii .................................. LIST OF FIGURES INTRODUCTION ................................. ..... 41 44 44 Introduction of the mice into the field Period of population establishment PAGE 44 .... ......... 46 Combination of Laboratory and Pield Procedures *. Homing 48 48 Establishment of the homing phenomenon ••• 48 Homing with aliens temporarily in the field 49 ......... ... *. ........... Lata recorded Population removal General Pro cedures ............... ..... . 51 *......... Measurement of feeding activity Observations 50 •........ 51 51 ........ 51 Weather recording .... •.••••.............. ... 52 Predation control ...... 54 ........ 54 Recapitulation RESULTS ......... ................................ Descriptive Measurements Use of nest boxes .... 56 .................. Distance moved between nest boxes ....... .... Bisexual combinations .................... .................... Latency of combining with the opposite sex Establishment and termination of bisexual pairs .............. v 61 61 ........................ . • Mono sexual combinations 58 60 .......................... Gregarious behavior Pair combinations 56 .... Moves between nest boxes Social Relationships 56 64 64 65 .• 71 71 Successive occupancy ofnestboxes Observations ..... ....... Spatial Patterns Homing ........... 76 ..... ...................... .. 77 .............................. .......... 80 Establishment of the homing phenomenon ...... Homing with aliens temporarily in the field Differential homing with aliens in the Comparison of Social Treatments . DISCUSSION ....... Movement data ...... 97 100 ....... ..... Social Behavior 101 105 .......................... ........................................ LITERATURE CITED 96 97 ...... Attraction and repulsionbetween mice Homing 90 •«•................... Gregarious behavior 84 90 ....................................... Population Phenomena 80 field • 85 Occurrence in Empty or in Alien Occupied Nest ........... Boxes SUMMARY PAGE 73 . ................................. 109 116 120 LIST OF TABLES TABLE PAGE I* Variability of population asymptotic levels within populations................. 7 II* Plants recorded in experimental field..... 30 III* Tabulation of occurrence in quadrats of all species recorded in at least five quadrats ............................. IV. The distribution of plant species in the experimental field.................. V. Average number of different nest boxes occupied per mouse during the period of population establishment............... VI. VIIo VIII. IS. 32 34 57 Analysis of variance of the proportion of nest box records in which mice were alone............ 63 Comparison of the observed frequencies of pairs with that expected by chance..... 64 Records of mice in nest boxes with the same sex or otherwise................. 64 Bisexual combinations..................... 65 S. Comparison of numbers of bisexual combinations.............................. 69 XI* Comparison of the duration of bisexual combinations.............................. 70 XII. Originator and terminator of bisexual pairs..................................... 73 XIII. Frequency and sex of mice occupying nest boxes on successive days.................. 74 XIV. Average distance to nearest neighbor....... 77 XV. Comparisons of the average daily distance to the nearest neighbor during the period of population establishment............... 78 vii TABLE XVI. XVII. XVIII. XIX. PAGE Homing behavior in experiments prior to alien introduction......... <>. 83 Comparisons of homing performance before and after alien introduction.............. 85 Homing to alien occupied or empty nest boxes.......................... Occurrence in empty or in alien occupied nest boxes................................ viii 88 94 LIST OF FIGURES FIGURE PAGE 1* Interaction of factors affecting distribution within local populations.................... 9 2o Design of the experimental field............. 22 3o North-east view of the experimental field.... 4* North-east view of the experimental field along the partition between plots........... 23 5* Diagram ofa nest box.............. ....... 25 6. Diagram ofa feeding station...... 28 7* The distribution of plant mass.............. 37 8. Mouse boxes used in the two social treatments............... 40 9. Hypothetical experimental procedure......... 42 10. Patterns of introduction of the mice into the plots ..................... «... 11o Tethering of alien mice for observational studies........... 12* 45 53 Movements to previously unvisited nest boxes and to other nest boxes....... 59 Occurrence of mice alone, in pairs and in groups of three..... 62 Proportion of mice found with the opposite sex and the mean duration of each bisexual combination................................. 67 The number of different bisexual combinations and days in combinations.................... 68 16. Latency of combining with the opposite sex... 72 17. Percentage of mice homing before and after aliens — combined data ° 81 13« 14. 15* 18. . Percentage of mice homing before aliens — social treatments. ........... ix 82 23 PAGE FIGURE 19* Percentage of mice homing to alien occupied or to empty nest boxes - combined data...... 86 20o Percentage of mice homing to alien occupied or to empty nest boxes — social treatments... 87 21. 22. Percentage of mice found in alien occupied or in empty nest boxes - combined data....... 91 Percentage of mice found in alien occupied or in empty nest boxes - social treatments... 93 x INTRODUCTION The distribution of an animal species in space may be geographical or local in scope• Geographical distribution of a species is measured in terms of the range of occurrence of its naturally existing populations. This treatment of distribution is essentially historical and has great de­ scriptive and comparative value. A species, however, is a "dynamic system” (Blair, 1956), whose basic components are local populations (Mayr, 1942) existing in equilibrium with the ever-changing environment. Therefore, the study of factors influencing spatial distribution within local populations offers a "dynamic" approach to many problems associated with the spread of a species over its broad geo­ graphical range. The Biological Importance of Spatial Patterns to Populations Species evolution. Local populations must not only be considered as important effectors of the distribution of a species, but they are also the operational units of evolution. The relationship of spatial distribution within local popu­ lations to the evolution of the species is a complex one. The amount of inbreeding and the. rate of spread of genetic factors are functions of the mobility and dispersal tendencies of individuals of the population. The mobility and dispersal of individuals may be influenced by physical factors of the environment such as food, nest sites, barriers, etc., and 1 2 by sociobiological factors such as territoriality, dominance— subordination, dominance-submission, intra— specific social tolerance, and assortative mating# The spatial patterns of distribution within local populations reflect, partially at least, those forces affecting the evolution of a population# Population control# Populations characteristically grow in a manner described by the logistic curve (Allee, et al, 1949, p# 304; Andrewartha and Birch, 1954, p# 348). That the logistic curve of population growth merely fits the data and cannot be considered a law of population growth for extrapolation and prediction, has been pointed out by numerous workers, e.g., Cole, 1957; Southwick, 1956; Wilson and Puffer, 1933* It follows, therefore, that the arti­ ficially derived constants of this curve should not be in­ terpreted* as the constants of nature with relation to popu­ lation biology (Allee et al, 1949, p# 304). As a population approaches the upper asymptote of the logistic curve per­ mitted by the environmental conditions, growth is arrested by a reduction in the contribution of progeny to the popu­ lation# With the passage of time, populations tend to re­ main at equilibrium with the environment# This equilibrium has been defined as "the average size held by a population over a considerable period of time”*(Allee, 1949, p# 315) and as such exhibits a range of variability. In general, factors controlling animal populations have been divided into two groups: those that are depend­ ent upon population density and those that are independent 3 of population density (Dice, 1952, p.344; Smith, 1935). Density-dependent controlling agencies are mostly biotic in nature and are believed to be the ones that principally determine the equilibrium density of populations. Density- independent agencies are mostly climatic. It should not be thought that the above two groups of factors affecting populations are mutually exclusive. Nicholson (1957) states that the same factor may be densitydependent in one situation and density-independent in an­ other. Much of the confusion surrounding these two classi­ fications of factors is due to the ambiguous use of the term "density-dependent factor" as originally defined by Smith (1935)* He used the term to designate factors which were so influenced by population density that they opposed with greater intensity at higher densities than at lower ones the innate tendency of populations to grow. Most prominent theories of the control of natural popu­ lations vary in the degree of importance assigned to the density-dependent factor of intraspecific competition. Andrewartha and Birch (1954) stressed a comprehensive approach in which all types of factors are involved in population control. In Milne’s words (1957) this theory may be summarized as follows: "Natural control is a matter of numbers in­ creasing and decreasing just so long as the en­ vironment permits; environmental conditions fluc­ tuate and the requisite conditions do not endure long enough either for unlimited increase or for decrease to zero; the ruling components of environ­ ment in this respect are multifarious and by no means confined to competition (intraspecific or interspecific), parasites, predators and pathogen." 4 Another widely accepted theory of population control stipulates that the only factors able to control are those whose actions increase in severity as density rises (Cole, 1957; lack, 1954a, 1954b; Nicholson, 1933, 1954a, 1954b, 1957; Nicholson and Bailey, 1935; Smith, 1935). Milne (1957) in his recent review outlined this theory as having the following two tenets: "1. 2. That the environment is comprised of densitydependent and density-independent factors; that (a) controlling factors must be densitydependent, and (b) the chief of these are (i; competition within the population itself and (ii; the affect of enemies.” Lack (1954a, 1954b) emphasized the importance of food shortage, predation, and disease as factors causing higher death rates at higher population densities. Nicholson (1954a, 1954b, 1957) goes further by postulating (1954b) that "the mechanism of density governance is almost always intraspecific competition, either amongst the animals for a critically important requisite, or amongst natural ene­ mies for which the animals concerned are requisites". Most theories of population control (of which the above are examples differing widely in their emphasis) suggest that intraspecific competition, while differentially evalu­ ated as to its importance, may be effective in controlling populations. Requisites promoting intraspecific competition among the individuals of a population include food, water, nest sites and any protected associated area, mates, and other non— sexual social relationships (Scott, 1956). Competition 5 for these requisites is not only a function of the number of animals present but is also determined by what Frank (1954, 1957) termed the "Condensation Potential" of the species. The condensation potential consists of behavioral mech­ anisms that enable many "cyclic" species to live at an un­ commonly high population density. "It is based on all in­ traspecific and especially social behavior that favors the increase of density" (Frank, 1957). Frank (ibid) further postulated that "the condensation potential is normally limited by intrinsic behavior, especially by territoriality, to a saturation point which is approximately adapted to the carrying capacity of the environment". Davis (1958) refers to the same relationship in a reverse way as the "individual distance tolerance limit" ]?\rhich is a species characteristic and may be so low that individuals can never associate to­ gether. This is apparently the "individual distance" con­ cept explored by Marler (1956). It should be noted that the expression of condensation potential is most noteworthy during the breeding season. This is true because many species aggregate during the win­ ter, supposedly as a heat conservation measure, and then segregate during the breeding season (Howard, 1949; Nichol­ son, 1941)* The concept of the condensation potential is also applicable to species which do not show the great fluc­ tuations in numbers charactEristic of "cyclic" forms. Fur­ ther, it is indicative of the interaction between individuals 6 of a population related to the area they occupy. As such, the condensation potential is a measure of the relationship between behavior and spatial patterns of distribution. Calhoun (1952) suggested a behavior - space relation­ ship following an experiment in which a colony of freeranging rats (Rattus norvegicus) did not increase to over 200 animals while living in the same amount of space in which 50,000 individually caged rats could have been main­ tained. He concluded that under free—ranging conditions, the rats expressed behavioral potentialities which were im­ possible in the caged conditions. Recent studies of laboratory populations in which the physical environment was controlled have shown wide vari­ ations in the asymptotic level for several populations in the same amount of space. Table I is a tabulation of several studies showing the range in maximum numbers in the several experimental populations of each study# The data in the table cannot be compared between studies because methods of considering young as members of the total population varied. Also, the environmental conditions under which each study was carried out were different. The variations in maximum population size among the several experimental populations of each study appear to be due only to behavioral differences in the social structure of each population. 7 TABLE I VARIABILITY OF POPULATION ASYMPTOTIC LEVELS WITHIN EXPERIMENTS TTumber of Enclosures Number of Animals at Asymptote 61 x 41 4 3 - 1 9 " (29” x 74") x 2 decks 4 Crowcroft " & Sowe (1957) " 6 1 x 6* 4 Southwick (1955) " 6* x 25* 6 25 500 sq. ft. 2 48 - 115 (approx.) "large openair cages" 2 41 - 58 3 28 - 67 Study Brown (1953) Animal Mus mus cuius Christian (1956) " " Strecker & " Emlen (1953) Enclosure Size 21 - 120 150 - 120 " Clarke (1955) Microtus agresiis Louch (1956) Microtus 6 1 x 251 penns.yivanicus Evidence contrary to the above was provided by Crow— croft and Rowe (1957) studying four freely growing popu­ lations of Mus muscuius in pens 6' x 6*. They found a marked difference in the rate of growth of their populations but a~I1 contained comparable total numbers after eighteen months One of the populations reached the amazing total of 150 an­ imals or 4.1 mice per square foot! The previously mentioned wide range of population size between several populations maintained under identical en­ vironmental conditions indicates that population density must be defined in terms of units of social pressure which are at 8 present poorly known. Christian (1957) while discussing this concept suggested that the "social structure of a popu­ lation in terms of aggressiveness of the individual members, their equality or lack of equality and other less well known behavior factors would determine the maximum population density...". Thus, in different populations varying numbers of animals may comprise each unit of social pressure as a result of individual differences in behavior (Brown, 1953; Southwick, 1955; Christian, 1957). The total units of social pressure for asymptotic populations under identical environ­ mental conditions would be the same although the numbers of animals in each population might differ markedly (Table I). Southwick (1955b), working with laboratory populations of Mus mus cuius, found that fighting varied between popu­ lations irrespective of the numbers of animals present in the same size area* Population growth ceased or was greatly impaired through excessive litter mortality when his popu­ lations built up to a point at which fighting occurred at the rate of one aggressive encounter per hour per mouse. This point was independent of density within the limits set by the physical environments of the experiments. Christian (1955a, 1955b, 1956, 1957) studying the effects of increase in population density upon adrenal size, reproductive func­ tion, and litter survival in house mice found that the phys­ iological alterations were quantitively and qualitatively similar for asymptotic populations irrespective of the actual numbers of animals present. 9 With the ahove tacts in mind, the question may now be asked how units of social pressure, condensation potential, and population density relate to spatial distribution• How does spatial distribution relate to each of the others? Figure 1 shows the hypothetical interrelationship of these factors. The interactions are complex. They are influenced, and may be controlled by, the omnipresent environment com­ posed of density independent factors and such density de­ pendent ones as disease, predation, and parasitism. ENVIRONMENI SOCIAL PRESSURE SPATIAL \ DISTRIBUTION CONDENSATION POTENTIAL Pig. 1. Interaction of factors affecting distribution within local populations. 10 Within the environmental complex, social pressure may be the fundamental factor influencing population growth (Christian, 1957; Prank, 1957)V It is an intrinsic attri­ bute of each local population and is the sum of the individ­ ual differences in social behavior which affect population growth (Scott, 1956). Social pressure effects population control chiefly through increased stress, fighting, and mortality, and through decreased reproduction, maternal care, and survival of young (Brown, 1953; Clarke, 1955; Christian, 1957; Southwick, 1955a, 1955b). Social pressure units are independent of density within limits and are basically in­ involved in the condensation potential of species populations. Por each species population, the condensation potential, as Prank (1957) defined it, has a characteristic value which is related to the carrying capacity of the environment. It is related to distribution in space as well as to the in­ dividual behavioral potentialities of the members of the population.1 Social pressures operate within and upon the condensation potential. Indeed, in experimental laboratory populations, social pressures have been shown to control population growth irrespective of the number of animals liv­ ing in the available space (Brown, 1953; Southwick, 1955b). Both the condensation potential and social pressure are effective in determining the spatial distribution of the population. The most important known social factor operat­ ing as a component of the condensation potential to control spatial distribution is territoriality. 11 Density is a term of descriptive value in a broad ecological sense but of questionable value in the study of spatial distribution within local populations and of the biological importance of such distributions. Density is defined as "the number of organisms per unit of space” (Allee, et al, 1949, p« 266) end is found by the following formula: Density = (Absolute numbers of organisms in area). (Number Vf^s'patxaT units in that area") Nicholson (1957) and Cole (1957) suggested that in order to be useful in the study of population dynamics, the above definition of density must be expanded and an implication drawn from it specifically for population application. Cole (1957) stated that density "in this sense, (the population sense) is not the number of organisms per unit area or vol­ ume but is the difference between this actual density and that which would prevail at carrying capacity. The con­ cept includes both the effects of crowding as a governing factor...and the effects of environmental inadequacies”. Thus, density may mean at least two different things. The relationship of density as defined by Allee to spatial distribution is also not clear. Density, by def­ inition, is independent of spatial distribution. However, in the broad sense, the pattern of spatial distribution de­ termines the number of organisms in the area. At a constant density, the distribution of individuals may vary from ex­ treme clumping to the maximum equal distribution possible in the unit of area defined by the density measurement. Spatial distribution is dependent upon density since an in- 12 crease or decrease in the latter will affect spatial dis­ tribution when the denominator of the above equation is unity. Such changes in d:ensity may not, however, affect spatial patterns in the density defined area when the num­ ber of spatial units of the denominator is more than one* Since density used in the descriptive sense obliterates the biologically important concepts of spatial distribution within local populations, a description of animal assoc­ iation and abundance relative to distances separating in­ dividuals would appear to offer the best method of meas­ uring these social-spatial-biological aspects of population dynamics. Within any confined area, density of animals is con­ trolled by social pressures acting through and upon the condensation potential. The experiments of Southwick (1955); Christian (1957) and Brown (1953) suggest that there is no direct relationship between density and social pressure since their populations stopped growing at widely different densities while at similar social pressures. Factors Affecting Spatial Distribution Spatial distribution of populations of small mammals living under natural or semi-natural conditions is related in a oomplex way to the influence of the physical as well as of the biotic environment. The importance of character­ istics of the habitat in determining spatial distribution has been indicated by numerous studies. 13 Brand (1955) studying the White-footed Mouse (Perom.yscus leuconus n oveb or a censi s) has shown a direct seasonal re­ lationship between the spatial distribution of the popu­ lation and tree density, degree of slope, and density of fallen trees* The significant factors of the habitat in this relationship were probably occurrence of food and po­ tential nest sites. Orgain and Schein (1953) effected a decrease in numbers of rats in city blocks for a short period of time by removing harborage sites which existed in excess of need. Davis (1958) changed the spatial dis­ tribution within laboratory populations of house mice through introduction of baffles and additional nest sites to the study area. Blair (1951) and Provost (1940) showed the re­ lationship of distribution and abundance to habitat differ­ ences. The location of food as a factor influencing spatial distribution was demonstrated by Calhoun (1949, 1950) with populations of rats (Rattus norvegicus) and by Strecker (1954) with house mice under semi-natural conditions. Under natural conditions this has been shown by Orgain and Schein (1953) with rats and by Brown (1953) with house mice. That population size and competition are important in determining spatial distribution is evident from Calhoun1s (1950) study of a freely growing rat population in a quarter acre enclosure structured to provide a gradiant of availabil­ ity to the food. Rats born near the source of food were able to maintain more effectively their home areas and force an­ imals living in submarginal areas to remain there. Stability 14 of spatial distribution related to numbers of animals in an area is evidenced by experiments in which part of the . population was removed. Stickle (1946) and Blair (1940) with Peromyscus, Orgain and Schein (1953), with rats, and Calhoun and Webb (1953) with several species of small mam­ mals demonstrated the tendencies of animals living in sur­ rounding areas to move into vacated areas following the re­ moval of the residents. Animals released into areas al­ ready populated by members of the same species rapidly dispersed (Blair, 1940; Calhoun, 1948) or disrupted the population and this resulted in a temporary decrease in the population numbers (Davis and Christian, 1956). Purely social factors influencing spatial distribution will be discussed later. It should be remembered, however, that the separation of the discussion of such factors does not mean that they operate independently of those mentioned above. All factors influencing spatial distribution in local and geographical populations probably operate simul­ taneously although at any point in time they may control populations singly (Leopold, 1933, p« 38). 15 STATEMENT OP THE PROBLEM The nature of the social factors influencing spatial distribution in local populations of small mammals has not been clearly demonstrated,. Individual differences in social behavior may influence the establishment and maintenance of patterns of distribution, Calhoun (1949, 1950, 1952) studied rats (Rattus norvegicus) in a 100 foot square area surrounded by a rat proof fence. The field was structured to produce a grad­ ient of availability to the food. Rats living in the alleys which adjoined the food were in a more favorable food sit­ uation than those rats living farther away. To get food, the rats living at a distance from the food source, were forced to pass through the home areas of the rats living closer to the food. Rats living close to the food grew more rapidly, and since weight is an important factor in attaining high social rank, they were more favored in attain­ ing higher social status than their peripheral neighbors who got less food and grew more slowly. The social status of the peripheral rats was passed on to the later generations. Smaller adults were forced into the peripheral areas of the enclosure. Their young, because they were born distant from the food source, grew slowly and thus were relegated to a low social status. Calhoun (1956) studied the effects of behavioral differ­ ences upon population dynamics where genetic factors were 16 controlled* He studied freely growing populations of Mus muscuius, using the two genetic strains DBA/2 and C57B1/10 which differ in physiological and behavioral character­ istics. His study was drastically curtailed since there was extremely poor reproduction among DBA/2 mice. He show­ ed, however, that physiological disturbances resulting from the poor adjustment of the mice to new situations, including relations with other individuals, were more prevalent in the DBA/2 mice. His results indicated that as population density increased, social differences due to genetic factors may have been differentially expressed in the physiology of the individual mice. King and Eleftheriou (1957) raised Peromyscus maniculatus bairdii in the laboratory in isolation or in groups of six individuals from weaning until 60 days of age. Following, this social treatment, the mice were released into an isolated field containing nest boxes and live traps in an effort to discover any differences in their ability to adapt to the natural environment. Subsequent to release, isolates moved about the field more and at greater distances than group raised mice. G-roup raised mice were found together more frequently than the isolation raised mice. These re­ sults were difficult to evaluate, however, since the popudeclined rapidly after release. One social factor offered to explain spatial patterns of distribution in small mammals is territoriality, defined as defense of an area (Noble, 1939). While territorial behavior is of widespread occurrence in birds (Howard, 1920; 17 Nice, 194l), such behavior in small mammals is a matter of controversy (Blair, 1 9 5 3 ^ Burt, 1940; 1943, 1949; Crowcroft, 1955; Scott, 1944). If territoriality does exist as a' characteristic mechanism maintaining spatial distribution relative to the physical environment and the biology of most species, what social factors are important in the es­ tablishment and maintenance of such behavior? If territor­ iality Is not characteristic of all species, what social mechanisms effectively regulate spatial patterns of dis­ tribution in non-territorial species? Do differences in the social behavior of the individual animals of the popu­ lation affect spatial distribution? This study was designed to combine laboratory and field techniques in order to examine the significance of social factors in the spatial distribution of Prairie Deermice within local populations. The objectives of this study were: 1. To study the affect of social manipulation of deermice populations in the laboratory upon their subsequent spatial distribution in a semi-natural environment. 2. To examine behavior patterns effecting spatial distribution in free ranging, semi—natural populations. 3. To measure the individual awareness of and preference for the home area as well as the stability of such an area. 18 MATERIALS Experimental Animals The Prairie Deermouse (Peromyscus maniculatus bairdii# Hay and Kennicott), inhabits prairies, open fields, beaches and dense grass along fence rows in the midwestern United States* It does not occur in Maine, where this study was conducted, although, another subspecies, Peromyscus manicu— latus abietorum, is found in the Spruce-Pir forest surround­ ing the study area on Mount Desert Island# P# m# bairdii was selected as the experimental animal because quantities were available from a laboratory colony$ and a grassland form was considered better suited to the study contemplated# No native grassland species of Peromyscus occur in Maine# An attempt was made to correct for any ecological imbalance between the maritime environment and the adaptability of this subspecies by providing food, nest sites and nesting material in excess of need# The biology of the experimental mice in terms of reproduction, weight, general body condition, and survival indicated that this attempt was successful, at least for the short periods under study# All of the mice were born in the laboratory and most were born in a single colony of 60 breeding pairs# These were descendants of twelve original pairs whose offspring had been in the laboratory for approximately 15 generations# Although natural populations of Prairie Deermice have been extensively studied (Blair, 1940; Dice, 1932; Howard, 19 1949), the factors involved in the spatial distribution of local populations are unknown* Burt (1949) suggested that territoriality is a part of the behavioristic pattern of many kinds of animals and results in population dispersal. On the basis of live trapping and nest box studies of wild populations, Blair (1940, 1953) and Howard (1949) concluded that P. m* bairdii does not show antagonistic behavior and ,thus is nonterritorial in the defined sense# Howard (ibid), however, postulated that Prairie Deermice exhibit some kind of natural negative response to crowding beyond a certain density within a limited area, but he was not able to de­ scribe, measure, or demonstrate the existence of any such negative force*t Laboratory Facilities The laboratory phase of these experiments involved no manipulation other than keeping the mice in specific social situations under conditions of light, temperature and humidity similar to those to which the complete laboratory breeding colony was subjected# The social situations and materials used will be discussed in the appropriate procedural sections# All of the mice were kept in basement rooms in the Be­ havior Division of the Roscoe B. Jackson Memorial Laboratory/ Lights were on in these rooms from approximately 7*30 A.M. to 4:30 P.M. each day. Each mouse was housed in a single compartment of a two compartment mouse box described in the discussion of the social treatments. Purina laboratory pel­ lets and water in excess of need were provided by means of 20 a food hopper and water bottle in the lid of each box* The floor.was covered with a thin layer of wood shavings, and the mice were placed in other, clean nest boxes every two weeks* Experimental Field General Description The experimental area was a 0*9 acre field located approximately 150 yards north of the Hamilton Station of the B. Bo Jackson Laboratory. The field was 250 feet long and 165 feet wide with the length having a compass bearing of NE — SW. A gradual slope in the longitudinal direction from SW to NE caused runoff of water toward the NE end dur­ ing heavy rains* The vegetation was predominantly herbaceous, although forbs were found in quantity in parts of the field. The experimental area was one of three one—acre fields. In 1945, all trees of the Spruce-Eir forest were cleared from the area, and the soil smoothed. to level or dump fill in the area. Ho attempt was made The northeast end of the field was soggy due to poor internal drainage and the runoff of water from the elevated southwest end. Design Enclosure* When originally constructed for dogs, the experimental field was surrounded by a seven feet high wooden fence to which wire fencing was attached at the bottom and buried in the ground* fox proof. This made the enclosed area dog and The northern part of this fence suffered severe 21 wind damage prior to the start of the study reported here and approximately thirty feet were replaced by a wooden fence four feet high* A map of the experimental field is shown in Figure 2. The area was divided into two 0*44 acre plots, each 240 feet by 80 feet and surrounded by corrugated aluminum partially buried on edge in the ground (Figure 3). The rolls of alum- inum measured 100 feet by 28 inches and were buried to a depth of 8"* Flat pieces of aluminum placed horizontally were wired to the enclosure to prevent escape where two pieces of aluminum fencing met (Figure 4)o Since the two plots were adjacent, a single piece of aluminum served as a common inside boundary for each* The eastern plot was designated as Plot A and the west­ ern one as Plot B* Nest boxes and live traps were placed in an identical pattern in each plot as shown in Figure 2. Due to an observation booth previously built into the wooden fence, files A and B were 18 feet and 6 feet shorter respect­ ively than the other files* Thus, in file A there was one less trap station than found in the others. Nest boxes* sented by circles Twenty—four subterranean nest boxes repre— * in Figure 2 were placed in each plot along three files and eight ranks* D, F, and in Plot In Plot A these files were B, B they were T, W, and Y. The even numbered ranks indicated nest boxes in both plots* Each nest box was 30 feet from its nearest neighbor in the same file or rank* Nest boxes in files B and F in Plot A and T and Y in Plot B were 10 feet from the longitudinal aluminum fence nearest P L OT B C D P L OT E F H S T U W FSI FS5 FS2 FS6 FS3 FS7 FS4 FS8 B X Y Z 2 40* A A 10' 3 0 '— >04-^ 80' P ig . 2. O - NEST BOX X “ TR A P STATI ON F S ~ FEEOI NG STATION D e s ig n o f th e e x p e rim e n ta l f i e l d . 23 Pig* 3* North-east view of the experimental field/ Fig/ 4. North-east view of the experimental field along the partition between plots/ 24 them* The end nest “boxes in each plot were 15 feet from the end fence. Each nest box (Figure 5) was built of one-half inch lumber, and had the following outside dimensions: x 5-3/4” deep x 7” high. 6i ” wide The nest chamber was floored by a piece of 3/8” hardware cloth supported li” above the in­ side floor of the box, and roofed by a ceiling of hardboard 3” above its own floor. 3”* Thus, it measured 5i” x 4— 3/4” x An air space of 1” for purposes of insulation was pres­ ent between the ceiling of the nest chamber and the inside surface of the lid of the nest box. A 6” piece of white rubber laboratory hose of li” outside diameter was connect­ ed to an opening centered 2i" above the floor of the nest chamber. This arrangement provided a pliable entrance tun­ nel of 1” inside diameter which was small enough to prevent larger animals from entering the box and was completely water­ proof when snugly attached. Eabh nest box was thoroughly saturated with paraffin and the hose entrance tunnel was heavily coated at the point of connection with the box. This method of preparing the boxes was effective in pro­ viding a waterproof nest site for the mice. Each nest box was completely buried in the ground. The top was covered by a piece of heavy roofing material and a large piece of sod. These provided additional insul­ ation. Cotton batting was used as nesting material. Six dif­ ferent colors were used in order to obtain a measure of the transportation of nesting material and were placed in the Diagram of a nest box 25 26 nest boxes in the following pattern Piles Boxes 2, 6, 10, 14’ 4, 8, 12, 16 Red White Blue Green Purple Yellow T & B W & I) Y & E There was, however, little evidence of such transportation during the period reported here. The mice used nest boxes almost exclusively and there was little indication of nesting in the field and subsequent sporadic use of the nest boxes as reported in some other studies (Nicholson, 1941; Howard, 1949; King, 1957). Live traps. There were four files of trap stations (X in Figure 2) alternating with the files of nest boxes in each plot. Due to the previously mentioned variation in the eastern outside boundary of Plot A, there were only 35 trap stations in this plot while Plot B contained 36. Two live traps were placed at each trapping station within the area enclosed by the aluminum fence. At each marginal trap sta­ tion, one additional trap was located outside the aluminum fence and was left set continually. Thus, a total of 169 live traps were used in the experiments, of which 27 were located outside the enclosure. Each live trap was a rectangular box, 9” long x 3" wide x 3” high, the sides and top of which were made of galvanized sheet metal and the floor of 3/8” plywood. The hardware cloth door was supported by a wire attached to a treadle. When a mouse stepped on the back half of the treadle, the support was removed and the door fell, releasing a lock 27 which prevented the trapped mouse from opening the door or entry hy other mice* Feeding stations* Food was provided in excess of need in the field and was placed in a small hardware cloth food hopper to prevent hoarding. One food hopper was placed at each of four feeding stations (FS in Figure 2) in each plot located at the intersections of files D and W with ranks 3, 7, 11, and 15* The feeding stations remained in the field at all times and were distributed equidistant from the near­ est nest boxes. Each feeding station (Figure 6) was provided with a mechanism for recording time and frequency of visits by the mice* When a mouse traveled to or from the food hopper, a pen attached to a treadle, which was part of the ramp lead­ ing to the food, was brought into contact with a continuous I1' wide paper tape. A spring-wound clock powered a large spool which wound tape from a reservoir spool containing enough for 48 hours of recording. Two different colors of Esterline—Angus ink were used so that tapes could be exchang­ ed between recorders and a single tape used for a longer period of time. Lights. Light for nocturnal observations was provided by seven 150 watt flood lights attached twelve fdet above the ground to posts located along file PI at ranks 3, 5, 7, 9, 11, 13, and 15 (Figure 4). A headlight powered by 4 standard size flashlight bat­ teries was used when the floodlights were not desired. Diagram of a feeding station 28 29 .Weather station. A small weather station (Figure 4) was attached one foot above the ground to the light post at H—9 in Plot A. The station was a wooden "box 24” wide x 8” deep x 18" high with a roof covered with asphalt roofing. The box was painted white and the front which faced the south­ west was louvered. A 24-hour recording thermograph, a wet- dry thermometer, and a maximum—minimum thermometer were placed inside the weather station. Barometric recordings and addi­ tional temperature recordings were made on apparatus set up a short distance from the experimental field. Vegetation Analysis Species list. A total of 66 species of plants from 28 families were identified from the experimental field and are listed in Table II. Plant distribution. Plant distribution in the plots was ascertained by recording species present in meter square areas near the nest boxes in each plot. A meter square quadrat was systematically placed to the east of nest boxes in files I) and F in Plot A and files W and Y in Plot B and to the west of file B in Plot A and file T in Plot B. The difference in sampling direction was necessary because of the proximity of files B & T to the aluminum fence. At each nest box the investigator took one stride in the proper direction, held the quadrat at arm's length and then dropped it. All plants emerging from the soil within the quadrat-circumscribed area were recorded. Table III lists those species recorded in at least 5 of the 24 quad- 30 TABLE II PLANTS RECORDED IN EXPERIMENTAL FIELD Family Scientific Name Musci Sphagnum sp. Equisetaceae Equisetum arvense Common Horsetail Grramineae Agrostis alha Calamagrostis canadensis Festuca rubra Holcus lanatus Phleum pratense Redtop Cyperaceae Carex sp. Eleocharis obtusa Eleocharis tenuis Scirpus rubrotinctus Sedge R * Br * Spike—Rush Spike-Rush Bulrush Juncaceae Juncus Juncus Juncus Juncus Luzula Rush Toad-Rush Soft Rush Rush Woodrush Iridaceae Iris versicolor Si syrinchium angusti folium Blue Flag Orchidaceae Spiranthes romanzoffiana Hooded Ladies*— Tresses Corylaceae Alnus crispa G-reen Alder Polygonaceae Polygonum per sic aria Rumex acetosella Lady* s-thumb Sheep— sorrel brevicaudatus bufonius effusus tenuis multiflora C aryophyllac eae Stellaria graminea Common Name Blue-joint Fescue-grass Velvet grass Common Timothy Blue— eyed grass Common stitchwort Ranunculaceae Ranunculus acris Tall Buttercup Droseraceae Drosera rotundifolia Round-leaved sund< Rosaceae Fragaria virgin!ana Potentilla canadensis Rubus hispidus Rubus idaeus Rubus sp. Strawberry Five-finger Dewberry Rasberry Bramble 31 TABLE II — Continued Rosaceae Rosa virginiana Spiraea latifolia Rose Meadow Sweet Leguminosae Trifolium agrarium Trifolium pratense Yicia cracca Yellow Clover Red Clover Tufted Vetch Oxalidaceae Oxalis europaea Wood— sorrel Rhamnaceae Impatiens capensis Spotted Touch—me—not G-uttiferae Hypericum canadense Hypericum perforatum St." John*s—Wort Common St. JohnfsWort Cistaceae Lechea intermedia Pinweed Violaceae Viola sp. Violet Onagraceae Epilobium angustifolium Willow-herb Ericaceae Vaccinium angustifolium Low Sweet Blue­ berry Labiatae Galeopsis tetrahit Lycopus uniflorus Prunella vulgaris Hemp—nettle Water Horehound Heal-All Solanaceae Solanum dulcamara Bittersweet Scrophulariaceae Rhinanthus crista-galli Yellow Rattle Common Speedwell Veronica officinalis Plantaginaceae Plantago lanceolata Plantago major Ribgrass Common Plantain Caprifoliaceae Diervilla lonicera Bu sh—hon ey su ckl e Compositae Achillea millefolium Aster lateriflorus Aster simplex Aster umbellatus Chrysanthemum leucanthemum Cirsium G-naphalium uliginosum Hieracium aurantiacum Hieracium canadense Hieracium flagellare Hieracium pratense Leontodon autumnalis Solidago bicolor Solidago graminifolia Common Yarrow Aster Aster Aster Taxonomy after Pernald (1950) White Daisy Common Thistle Low Cudweed Orange Hawkweed Hawkweed Hawkweed King Devil Pall Dandelion White G-oldenrod G-oldenrod 32 TABLE III TABULATION OF OCCURRENCE IN QUADRATS OF ALL SPECIES RECORDED IN AT LEAST FIVE QUADRATS Plant Species Sphagnum sp* Agrostis alba Festuca rubra Holcus lanatus Carex sp* Juncus sp • Luzula multi flora Sisyrinchium angustifolium Stellaria graminea Fragaria virginiana Potentilla canadensis Spiraea latifolia Trifolium agrarium Trifolium pratense Vicia cracca Oxalis europaea Viola sp* Prunella vulgaris Rhinanthus crista-galli Veronica officinalis Achillea millefolium Aster lateriflorus Hieracium sp* Solidago graminifolia Leontodon autumnalis Plot. A 7 20 24 5 7 9 13 8 8 2 15 8 10 20 6 0 1 7 9 7 6 13 9 5 5 Plot B 1 17 24 2 15 2 10 8 3 12 13 8 4 8 0 5 9 2 2 6 3 18 18 12 3 33 rats sampled in either Plot A or B, The number of quad­ rats in which each species was found is also given. There was a wider distribution of the plant species in Plot A than in Plot B. Table IV shows the distribution of the plants recorded in 5 or more quadrats in at least one of the plots. Since each sample was taken in conjunction with a nest box, the presence of each species is recorded by nest boxes. The pattern of distribution shows a similarity between plots with the moisture loving plants occurring in the wetter northeast end of the plots in the areas of nest boxes 12, 14, and 16. Standing crop. A measure of the distribution of plant mass in each plot was taken late in September, 1958, (Exper­ iment 6). One cubic foot samples of plant material were taken every 15 feet along each of the three files of nest boxes in each plot. Samples were taken in the same direction from files of nest boxes as was done when sampling the dis­ tribution of plant species. All plant material within the one cubic foot of volume enclosed by the quadrat was clipped at a height of 1*' above the ground, placed in a paper sack and transported to the laboratory. The sampled plant material from each quadrat was kept in the laboratory for a minimum period of one week. Sub­ sequently, all samples from one transect in each plot were placed in an electric oven for 48 hours at a temperature of 50 degrees Centigrade. The samples remained in the paper sacks during drying and this procedure was repeated for all 34 Ph VO MMM MM X X R VO MMM XX txj MM rH H pq co Pq -st- MMM R MMM FIELD pq ^ IN THE EXPERIMENTAL i —t M XX MM M X M M X M X M M X MM M M HI M M M M MMM M M M R CM MMM M M M M M M MM M MM MMM pq o H MM R O H MM rH pq o M pq CO MM EH RCO PS PS r pqco MM MM MM M R MD MM MM M M OE PLANT pq 'H* M M MM M pq CM MM p CM rS PS pq CO R ■ ‘H- p CM M M MM MM MMM M M 02 cd cd p • ,q p, -p RH p G d m cd p p M M MMM 03 ds Fj 03 Cd H *30Sh H -P H O a> ra O m R R CQ EH MM M M M o !> R ftj M M M M M M H M M M M M o cd ® a) . H cd cd •H H P h ® p03 R 2 H H o3 p O-P S 02 H R 02 cd H r cd ,3 U CG P *H cd H H cd •H H o 02 1 pi -— 1 . -p H 02 Cd H =J 03 h+= < D- H ■■ 02 r l fl H H §>"£ § § M S H 5) H qh C S O O (3d oj © 9rl cfl o - p o «) o ^H ® Fh ch ch - H H H pi Pl cd a> cd-p H H H O cd O p H **0 R tib c D o c d p p H U P in h -H M r!fn fl Ph to M M M M MMM H cd 02 SH^I«h -P H VO I —I VO M EH VO MM rH rH M EH Hi —I MM MM M MM O H MM EH O MM U n rS M M >H CO M M M MM MMM M M M MM MM MMM M >H VO ^ VO pq pi EH VO eh ^H H^ H* EH H * >H CM ter CM EH CM M M M M M M M M MM M M o CD PllH rQ CD M M M M M M M M M MM MM M M MM M M rS PH MM M M I—I fi-i T1 § § a) H Ei *H P cd rH 3 i—! M M MM rS RS 0 0 0 -H P cd 0 cd cd Pi O 0 *H ttf) rH P M M M 0 •H H Cd •H rH cd Pi •H O H •H .. H C H cd CH 0 tto o H ^ *H cd H °P *ri ?> cd cd p i> h • pip ,d-H 05 rH rH Ej JEl Cd p * p I o pi 0 g O P -H c«h ■_ a j j h 0 p,cd^JcdcdcdCH •H*H C d*rl *H O p O • rH *H CH M M iii SSI 8 MM M M M M M *rl 03 M M - m U MM MM M M M M M M MM o3 0 cd od ^ M M MM M MM M M MM M M M M M M M M M MMM M M M M M M M rS rS KJ* M MM M M MM M M M M MM M MM MM M M MM M M EH 0 0 MM MMM MM CO M M >H O H ^ K s*K t rS PJ S MM M •H P Pi MMMM M M i —t i-q M M rH M M MM CM EH CM > H M M M iH EH O M M M ^ pq M MM MM rH o o M M &s H- >H CM pj M M iH P M M M >H H* H 0 P M M K 0 cd p p •H rH P P O 0 H H 0 i— 1P 0 *H i—I i— i 0 cd s o 03 p* 2 X P H fi 0QrH. ffl d d O O (Jt I cj © 8 r i f l r l H P P cd •H p P O rH ^0 0 p CfHP *H rH i— 1 P p p 0*H»H 0 cd P *rH K H 0CdP-H*H-HOCdOp*H O § P rH 0 rH ) g 5 p u o p, p U -H X -H P Jp Pm 0 0 ■H O w < J P m K o '- o P 0 co £h fij co eh eh > o > JH PP x i co <*i 0O 0O >• 10* 10 — 30*— >0-*-* 30' V E GE T A T I ON 2-25 Pig# 7. □ WE I GH T / GR A MS 26 -48 III 49-72 The distribution of plant mass, 38 PROCEDURES This research was designed as a study of the affects of social manipulation in the laboratory upon the spatial distribution of Prairie Deermice subsequently living in semi-natural field conditions# There were eight experimental periods, each of three weeks duration between June 6 and November 21, 1958# During this time 8 populations of four bisexual paris raised either in isolation or in groups were living in each of the two plots of the experimental field (Figure 2)# Thus, the total number of mice released into the plots was 128 of which 64 were raised under each of the two social treatments* The procedures utilized generally may be placed into three categories designated as laboratory, field, and a com­ bination of the two* A discussion of general procedures also includes others not readily placed in the above.categories# Laboratory Procedures Social Treatments All mice were born in the laboratory and remained with their parents until weaning at three weeks of age# Each fam­ ily occupied one compartment of a standard two compartment mouse box* Each compartment was 6” x 11” x 6" in size, and food and water were provided through a food hopper and water bottle in the cover. Litters were selected for social treat­ ment by the following criteria: birth must have occurred with­ 39 in a nine day period centered around a day 70 days prior to the scheduled day of release into the experimental field; each litter must have contained at least one male and one fe­ male; and hoth parents must have survived until the date of weaning* At three weeks of age, the young mice were separated from their parents and randomly placed either in a group or in isolation* Isolation raised mice* Isolation, as used in this study, refers to a single mimse living in one compartment of a stand­ ard mouse hox (Figure 8) from weaning at three weeks until ten weeks of age* Visual and tactual contact with other mice during the treatment period was thus prevented* In all cases male and female sibs lived in the different compartments of the same box* G-roup raised mice* The group social treatment differed from the isolation treatment in two ways* First, the parti­ tion separating the two compartments in each mouse box used in the group treatment was made of i n hardware cloth instead of wood as in the isolation boxes (Figure 8). Secondly, a male and female were placed in each compartment of each box* All four mice assembled in each mouse box were strangers to each other and were young from four different litters. A male and female from one litter were cross-paired with a male and female from another litter and the two cross-litter pairs thus formed were placed in two different mouse boxes. Each mouse in the group situation had tactual contact with one other mouse, its pair-mate, and visual contact with the three 40 Fig.* 8*1 ’ Mouse boxes used in the two social treatments;'7 Left - G-roup Raised; Right - Standard box — Isolation Raised® 41 mice in its box. Figure 9 illustrates the procedure for placing mice in the social treatments. The group situation was devised to assure, as nearly as possible, similar social development of successive populations of group raised animals at the time of their introduction in­ to the field. The unpredictability and difficulty of meas­ uring social hierarchies, dominance-submission, etc., which exist in larger groups, were considered sufficient to nec­ essitate as rigid control of social interaction as possible. During the treatment period, food and water were supplied in excess of need and the animals were moved to other, clean mouse boxes every two weeks. Both treatment populations were kept in their respective social situations until release into the field at ten weeks of age. At this age the mice were phys­ ically and reproductively mature (Clark, 1938; Dice and Bradley, 1942). Preparation of the Mice for Release At ten weeks of age, (Figure 9) eight mice from each social treatment were released into different one-half acre plots of the experimental field. Animals to be released were selected according to two specifications. females were acceptable. spection and palpation. First, no pregnant Pregnancy was ascertained by in­ Second, populations released in the field were composed of a male and a female from each of four litters. For the isolation raised mice, this was accomplished by random selection of four mouse boxes each containing a male and female. As indicated previously, group raised males and females were raised as cross-litter pairs. Thus, random EXPERIMENTAL PROCEDURE o» CD m w •o •o o» •o •o •o •o CD K> I o ^ •o c o o» o •o cw CD Ll Q CD *o 41 if* JC JC e s * at v> c o CL o a M Ui o > £ Of c M *o K> . Hypothetical experimental procedure. All fitters did not contain 2 males and 2 females. Thus, the same litters were not represented in "both treatment groups. 42 cr\ • P-H H 43 selection of one pair necessitated the selection of the re­ ciprocal mating involving the sihs of the first pair select­ ed* Only one of the two pairs living in a single group mouse box was released into the field* Henceforth, in this discussion, those mice which were in­ troduced into the field shall be referred to as experimentals and those remaining in the laboratory in their original social treatment as controlsOn the day preceding scheduled introduction into the field, both the experimentals and controls of each social treatment were etherized, weighed, and numbered by toe clip­ ping. In addition, all experimental mice were tagged with a numbered fingerling tag which was used to attach a colored celluloid disc to one ear. Immediately after weighing and tagging, the isolation raised experimentals were returned to their boxes where they remained in isolation until release into the field. Following weighing and tagging, each pair of group raised experimentals was placed in a different grouptreatment mouse box. The individuals of each pair were sep­ arated and placed, one in each compartment. Since only one pair of group raised mice was selected for the field from any single group box, recombinations of the control mice were made to fill the vacant compartments. This was done by moving pairs in from other partially vacated group boxes. 44 Field Procedure Preparation of the Experimental Pield Prior to the introduction of each experimental population, the nest boxes in each of the plots were cleaned and clean cotton nesting material and five pellets of Purina laboratory mouse food were placed in each nest box* Introduction of the Mice to the Field The experimental mice were released in the field at least one hour before sunset on the same day that the pre­ vious populations were removed* This was also true for the first experiment since preliminary populations preceded it* Release of populations raised in each social treatment alter­ nated between plots A and B for successive experiments. The experimental populations were introduced into the plots as systematically as possible since preliminary work indicated that releasing the populations at the center of each plot re­ sulted in high mortality. Thus, a pattern of introduction in­ to nest boxes was followed for each plot and was repeated for the eight experimental periods. This pattern of introduction is illustrated in Figure 10. Assignment of the experimental mice to nest boxes was done by using the most closely associated pairs as the basic units for distribution* The isolation treatment was repre­ sented by 4 sib—pairs of mice with the individuals of each pair having been raised in separate compartments of the same mouse box. The most closely associated group raised pairs were not sibs as in the case of the isolates, but were mice 45 P LO T plot FS5 FS 2 FS6 FS3 FS7 FS4 FS8 240* FS b 10* 30 *— >0 ^ f A->0<— 30' A 17 C o l o r = C l o__ s e ly a s s o c ia te d O - N E S T BOX X - T R A P STATION FS - FEEDING STATION 10, P a tte rn s o f in tro d u c tio n o f t h e m ic e i n t o t h e p l o t s . 80‘ F ig . p a ir 46 living together since weaning. Males and females of each closely associated pair were released into nest boxes sep­ arated from each other by a distance of 67*1 feet. In addi­ tion the pattern of introduction of the animals of each popu­ lation was such that the nearest neighbors of the same sex were separated by a distance of 42.4 feet and those of the opposite sex by 30 feet. This procedure assured that: (a) Naive animals gained experience with one nest box* (b) Original distribution patterns were con­ stant for each experimental population no matter into which plot it was placed. (c) Individuals of all closely associated pairs were initially equidistant from each other. (d) All mice were equidistant from the near­ est neighbor of the same sex. (e) All mice were equidistant from the near­ est neighbor of the opposite sex. Period of Population Establishment Following introduction into the field, a period of 17 days was considered sufficient for the mice to establish themselves in a relatively stable manner in the field. The 17 day period was set arbitrarily and was considered adequate after inspection of data collected in preliminary experiments. Daily records. During the "period of establishment", daily checks of all nest boxes in each plot were made and the whereabouts of each mouse recorded. No time schedule was followed for checking the nest boxes in each field, and since all mice wore a celluloid disc of a different color on one ear, it was not necessary to handle them for identification. 47 The procedure was to open the box, push the cotton aside, and record the mice present® nest and tried to escape® Rarely, mice jumped from the They were caught, if possible, and returned to the nest box. It was interesting to note that escape behavior most frequently occurred among animals recently released in the field or among those that had es­ caped or tried to escape at previous inspections* If animals were prevented from escaping during the first two nest box checks following introduction, very few made escape attempts during subsequent inspections. Trapping periods* Two periods of live trapping were conducted during each experiments "period of establishment". Each trapping period was of four nights* duration and on each night, traps remained set until two hours after sunset. The first trapping period began on the second night the mice were in the field and continued through the fifth night after in­ troduction. The second trapping period was conducted on the 14th, 15th, 16th, and 17th nights of each period of establish­ ment. On the night preceding the first trapping period, the traps were baited and turned upside down permitting the mice to enter and obtain the food without capture. A single pel­ let of Purina laboratory mouse food was used as bait in each trap during the prebaiting as well as in the regular trapping period. Traps were set at various times during the day but were checked and unset at two hours after sunset each trap night. Mice were released by turning the traps over and allowing them to escape or by shaking them into a plastic sack and then releasing- them. Each mouse was recorded by 48 its number as indicated by the colored ear disc and combina­ tion of toes clipped. Subsequent to each trap check the traps were turned over, baited with a single pellet of food, and left open so that mice could enter and not be caught for the remainder of that night. On the day following each trapping period, the bait was removed from each trap and the traps were unset, although they remained open and accessible to the mice. Combination of Laboratory and Field Procedures Homing Homing was defined in this study as the ability of in­ dividual animals to return to previously occupied nest boxes in an experimental plot after being absent from the plot and nest boxes for approximately 36 hours* Decision was made to use this homing test to measure the ability of the mice to home; measure the significance of previously established spatial distribution patterns; and provide basic data from which to make comparisons of homing ability when the home nest boxes were empty and also when aliens were in the home nest boxes* Establishment of the homing phenomenon* On the morning of the 17th day following the introduction of the populations into the field, all mice were removed from the experimental plots. Each was taken from a nest box at approximately 8:00 A.M., placed in a live trap, transported to the laboratory, and kept in isolation in a standard mouse box until sunset of the following day. Thus, each animal was in the laboratory 49 for approximately thirty— six hours during which food and water were supplied in excess of need* Rarely, all mice known to he in the field were not found on the morning they were sched­ uled to he removed and thus could not he taken to the labora­ tory for the total isolation period* In this event, the live traps in the plot were set that evening and the animals were taken to the laboratory when caught. Following the thirty— six hours of isolation, each ex­ perimental animal was reintroduced to its home plot at a point distant to the area of its most frequent previous occurrence. Following the technique of Hayne (1949) for calculating the center of activity from trap data, a "Resi­ dence Center" was calculated for each animal, using nest box records obtained during the last 10 days preceding re­ moval from the field. The field was bisected transversely and mice whose residence centers occurred in one—half of the field were reintroduced at the release point in the opposite half. In Plot A> the release points were feeding stations 1 and 4 and in Plot B they were feeding stations 5 and 8 (Figure 2). Reintroduction to the field was made at sunset of the day following removal. The mice were taken from the labora­ tory in live traps and at the appropriate release points the traps were turned over and left on the ground. The investi­ gator immediately left the area and the mice could leave the traps at any time. Homing with aliens temporarily in the field. A single alteration in the homing test was made for Experiments 6, 7, 50 and 8# During previous tests all nest boxes in each plot were empty at the time of the reintroduction of the experi­ mental mice. Beginning with experiment 6, however, the plots were transversely bisected, and a single young male deermouse 20-30 days old was placed in each of the 12 nest boxes located in half of each plot. The section of each plot which was selected to receive these aliens was the area in which the residence centers of most homing animals were located. The young alien males were placed in the nest boxes approxi­ mately two hours before reintroduction of the residents and were retained there until the following morning. Retention of the aliens in the nest boxes was achieved by putting a tiny collar on each animal and attaching a fine wire leader to each collar. The leader was then snapped on a wire loop in the cover of each nest box. All other techniques and pro­ cedures for this homing test were similar to those of the previously described test. Data recorded. Following reintroduction of the resi­ dent populations, the location of each mouse was recorded during the next three days. In those experiments in which aliens occupied the home nest boxes of residents, the pres­ ence or absence of the residents in the alien occupied nest boxes and the condition of the aliens were recorded. As in­ dicated previously, all aliens were removed from the plots on the morning following reintroduction of the residents.' 51 Population Removal On the third morning of the homing phase of each experi­ ment, all experimental mice were taken from the nest boxes, killed with chloroform, weighed, and the adrenal glands re­ moved and weighed* laboratory control animals were likewise sacrificed, weighed and dissected within approximately 24 hours before or after the experimentals. In the case of pregnant females, the number of young, and the weight of the uterus plus young were recorded in addition to the other measurements# The adrenals were removed from each animal as soon as possible after its death. Cleaning of the adrenals preparatory to weighing was done with a scalpel using a dis­ section microscope at 20X magnification. The adrenal glands rested on a wet paper towel on the stage to prevent desicca­ tion. Weights to the nearest .2 milligram were obtained on a Boiler—Smith Precision Balance. General Procedures Measurement of Feeding Activity Excess food was available at the feeding stations at all times during each experiment. The feeding activity recording mechanism of each feeding station was activated during the eight nights intervening between the two trapping periods of each period of population establishment. These data will not be discussed here. Observations The nocturnal activities of the mice were observed with a flashlight or with the seven flood lights in the field. 52 Most observations were made immediately after release of the mice from the traps at two hours after sunset and while alien mice were on tethers near the feeding stations or in the vicinity of nest boxes known to be occupied by resident mice. Alien mice were used to promote social interaction and were removed from the field immediately at the close of the observation period. The aliens wore a collar made of small fishing tackle snaps. A fine six inch wire leader was attach­ ed to a swivel on the collar and then to a stake in the ground (Figure 11 )♦ Since there were swivels at both ends of the leader, each mouse was free to move within a circular area of radius equal to the length of the leader. This con­ finement did not appear to impede the mice from moving and manipulating their bodies* In addition, the collars were apparently not detrimental to the mice since they wore them in the laboratory for several months with no ill effects. Weather Recording Weather data were recorded when possible at 8:00 A.M. each day. Rata recorded included maximum and minimum temp­ eratures, wet and dry bulb thermometer temperatures, and ob­ servational data on cloud cover and wind conditions. Also, a 24 hour thermograph recorded temperatures in the weather station in the field. Air pressures were indicated on a "Taylor Cyclo—Stormo— graph11 barometer which was located approximately 150 yards from the experimental field. A Taylor thermograph recorded outside temperatures at the same distance from the field. m ^ m u F i g . 1 1 1 .' T e t h e r i n g o f a l i e n m ic e f o r o b s e rv a tio n a l s tu d ie s . 54 Additional weather information was obtained at an Official TJ. S. Weather Bureau Station at Acadia National Park, a distance of eight miles from the experimental area. Predation Control A steel trap was continuously set near each of five feeding stations to insure against predation by mammals. Predation was not an important factor in this research; for only seven out of 128 experimental animals disappeared during the periods of population establishment. Recapitulation Eight successive populations of Prairie Deermice were raised from weaning (21 days) in two social situations, either in isolation or in groups. The groups were composed of two mated pairs living on either side of a wire screen partition dividing a standard mouse box into two compartments. At ten weeks of age, four bisexual pairs from each social treatment were systematically released into different onehalf acre "mouse proof" plots. Each experimental population remained in the plots a total of three weeks. During this time the location of each animal was recorded daily. Seven­ teen days after release, all mice were removed from each plot and kept in isolation in the laboratory for 36 hours. Each was then reintroduced into its home plot at a point distant to its previously established home area and the location of each for the next three days was recorded. In the experiments following the establishment of homing as a phenomonon, alien mice were in one—half of the nest boxes during the first 55 night of the homing test. At the conclusion of each three week period, all field experimentals and their laboratory controls were killed with chloroform, weighed, and the ad­ renals removed and also weighed. The laboratory controls were the same age as the experimentals but had continued in the original social arrangements in the laboratory rather than being released in the plots. During each experimental period other dynamics of the population were measured by two periods of live trapping, each four nights in duration; recording time of feeding activity; and direct observation of social interaction be­ tween residents and aliens. 56 RESULTS Descriptive Measurements A prerequisite to meaningful analysis and interpretation of the data is an understanding of the adaptation of the pop­ ulations to the experimental situation. The following de­ scriptive measures were taken as a means of gaining this un­ derstanding. Use of the nest hoxes. For the eight experiments re­ ported here, a total of 1909 occurrences of mice in nest hoxes was recorded during the periods of population establishment. One population of 8 animals was in each of the plots of the experimental field during every experimental period. The period of population establishment for 5 of the 8 populations of each social treatment was 16 days instead of the originally planned 17 days. This occurred because a few mice avoided capture and could not be removed from the plots on the day scheduled. Further, 11 mice (5 isolates and 6 group raised) of the 128 released in the plots died (4) or disappeared (7) during the period of population establishment. Adjusting for these deviations in the design, a total of 2023 recordings of mice in nest boxes was possible during the experiments. Thus, the mice were found in nest boxes 94*9$ of the times possible. Table V lists the mean number of different nest boxes occupied per mouse during the period Of population establish­ ment and during the last 10 days prior to removal from the 57 O • VO H CM 0 vo co • 'H' • ■H- • in CM e •'vh CO • vo co • ■d* 03 & r• co CO 00 • in in in • -M" co • -d- in • vo CO 0 CM in 0 ■'d* CO • in * in in • vo cO • -d* CO • 'M* co • in co • c— -d* OO • ■st* u o s & m i=> o 0 H- vo EH Pm P CQ LT\ • 'M- vo RH O• CO • -H- o\ vo -H- CO '• "d- CO • •ed* LCV CO ♦ 'H* in 0 co in • co 'H* CO • CM in * O CTi 05 P 0 M R O EM O CO O R Pi in • in in c- in 0 vo in • in in • *d~ co • CO CO • in vo in in • ^d* co 0 in CO • co in • co -d* CO • -d- CO • t— • co Pi H vo in • ■'H* t— « VO in • vo LTV CO • in CO <> CM co 0 c— co • CO CO ■ in P O HIH O EM R > CD 'CJ 0 rH CD 3 0 Pm 0 rH CD S O rH -P 0 CD R 0 rM CD £ 0 Pm 0 H cD in -p Pi 0 3 C i i H TO PH O 0 *H *d rH 0 R •H CD Pi -P 0 03 E- ° £ Pi 'd 0 1— 1 cD 0 Pm CD-P cD •rD O 1— 1 0 rH CD Jgj *P 0 cD Pm P3 R P -P P (D 0 Qi 0 03 CD J-M R O »H Pi CD P R H 0 O *H o 0 Pi O *H 0 O H p$ -P r O f pi 05 Pi •H 0 rH CD £ 0 Pm 0 —1 CD ^ 1 ^ Pi >> 0 r- t -p Pi r s C OT) K CD W X 03 O ® >» .®o§ ci a ^i ^ J” V- in > &- > to E x: CD ® o *~ > to O o g* 2 H Z O o C ^ 3 CO CO CL d and X unvisited. nest boxes CO to other :£H-z>z: CO .00 X X X Movements CO CO Q. o -H •-* »-I a cc -rH rei to IO IP Of K fO F-m id o in iueuju8dx3 j 9P^.05). Figure 15 illustrates the number of bisexual combina­ tions and the total number of days mice of each social treat­ ment were found in a bisexual combination of any type* The number of events (# in Table IX) in each category was ad­ justed by dividing by (N), the total number of animals of the sex being measured or by (0), the number of mice which combined in bisexual combinations* This second method re­ moved the effect of differential rates of combination and compared the two sexes and the social treatments on the basis of the number of animals which were found with the opposite sex. The analyses of the differences between social groups are discussed below. Different bisexual combinations. Comparisons of these data were made with the t test and although no differences were significant at the 5$ confidence level, there were several which showed a consistent low level of probability of occurrence by chance. These comparisons and the P value for each are listed in Table X. 67 CD CO * o cr c o X) CD V) 6 q c JC o o o o Ui c : •••••• Ql o co •••■■a ■■■••a aaaaai 3 O *r -O ° E c o □ o o to •H £ > cu i < o Pi •cH -I—' d§ o 3 Q s o lx i OJ S<» S d ML < W 0> A d q J■■■■■■■■ ■■■■P ■■a i:■■■ ■ ■ ■ II■■■■■■ ■a ■■■■■■■P " r r 1 * * m f“ mz f“ " ■" r“ r * i i * r r I L -__ _ _ _ L *II I I■■■■ n ■ ■ ■ n ■I _ ■ ■ ■ ■ ■ ■ ■ ■ ■ ■ mr a a■■aa a a a■a ■Pa a B ■ ■ n* I r r "" r * Hm r " " ■ jj r tn *“ " p’ r“ z E ii Z ■■■ ■ B ■ i ■ ■ ■ ■ ■ ■m i■ ■ a ■ ■ a n a n a ■ ■ ■ a ■ ■ a ■ ■ £ a a a 3 3 D d H i ^ CD CD o C w O o ^ CO O ft Ph o CL a O CD ■H o CD d | pi o JC *o — <+H P O ft o •H llfll— 1WI CD CO uoi|D|ndod ui e o i j o jaquunfsj jad Aouanbajj s o o "rro to esnow J3d CD a 3 -E S' O ^c CL § r .O .E U_ x x a> CO gq —f oo IS C o (A (A a >> ~o —o uj e e «4— c c o < < >» H— <+o o o c w k. a> a > a> 3 JD -O cr 1 E 0 w5 3 3 Li. 2 2 ti II i i # 2 O ■■r— CD o IO c o Aouenbe-tj o Q. O corfb O in c aJ d.s,ys vt o W—* ' 1 •H coi-ibinb. t i o l. s end OC

«/> — o E xi 6 o o of o N # tt JC Tc.0 nuiiibei Of ^-5» » o m 69 TABLE X COMPARISON OP NUMBERS OP BISEXUAL COMBINATIONS Comparison larger Number of Combinations Signifi cance Level #/0 #/rr Isolation vs* Group All mice G> I (.2>P>.l) (.3>P>.4) Pemales G>I (•3>P>.2) (.4>P>.3) Males G>I (o2>P>.l) (P>.5) Males vs* Pemales Group raised N 0 G 1 c1 $ (.4>P>.3) Isolation raised d’*> $ (.2>P>.l) All mice d* > o -r (.05>P>.02) = Number of animals in population. =s Number of animals combining with the opposite sex* = Group raised mice, = Isolation raised mice. Group raised animals occurred in consistently, but not significantly, higher numbers of bisexual combinations than the isolation raised mice* Purther, males of both social treatments combined with more females than females did with maleso When data for like sexes from both social groups were combined, males were found in a significantly greater number of bisexual combinations than females („05>P>.02). Duration of bisexual combinations* Differences between social treatments in the total number of days mice were found in bisexual combinations (Pigure 15) were measured by t tests. Group raised mice were in combinations (#/N) a longer period of time than isolates, although, not at the 70 P=.05 level of significance (Table XI). No differences be­ tween the two sexes of each social treatment were evident from these analyses. TABLE XI COMPARISON OP THE DURATION OP BISEXUAL COMBINATIONS Comparison_________ Larger Number of Days Significance Level # / N _____ #/0 Isolation vs* Group All mice G> I (.3>P>.2) (P>.5) Pemales G> I (o3>P>.2) (P>.5) Males G> I (o3>P>.2) (P>.5) Males vs. Pemales :(p>.5) Group raised # N 0 G 1 Isolation raised o>> $ (.5>p>.4) All mice 6*>£ (.02>P>.01) = = = = = frequency of event. — — Number of animals in populations. Number of animals combining. Group raised*mice. Isolation raised mice. When the results concerning the total number of animals of the same sex were combined and then compared, males were found to have been in bisexual combinations (#/0) a signif­ icantly greater number of days than were females (.02) P) .01). The influence of the social treatments upon bisexual combinations may be summarized by saying that a higher pro­ portion of group raised mice combined; they formed a greater number of different combinations; and they continued in com- 71 binations for a longer period of time than isolation raised mice# Consistent differences between isolation raised and group raised mouse populations have not been demonstrated to be statistically significant by individual test compari­ sons# A significant difference between the social treat­ ments has been shown, however, by the 3—way analysis of variance shown in Table VI. There, the units of measure­ ment were the proportion of nest box recordings in which mice were either alone or combined with other mice# In that analysis, group raised mice were in combinations for a sig­ nificantly higher proportion of nest box records than isolates* Latency of combining with the opposite sex* The number of days that elapsed following introduction to the field be­ fore each mouse was found with a mouse of the opposite sex was recorded# The percentage of mice of each social treat­ ment found for the first time with an animal of the opposite sex for each day is indicated (Figure 16) as an accumulative percentage curve. The distributions for the social treat­ ments were compared by the Kalmogorov-Smirnov test and group raised mice were found to combine with the opposite sex significantly sooner (P<*00l) after introduction to the field than the isolation raised mice* No comparison was made of the latency of combining with the same sex because of the low frequency of such events. Establishment and termination of bisexual pairs* Mice were not recorded as bisexual pairs unless they were found together in a nest box at the time of the daily check. 72 “O Q> Ui - . — CM £ .. 9 “2 “O c•H C. O Pt o &• CM CO -CO < o ID P>.05) Iy G (.02>P>.0l) I y q. (.05)P > .02) ^ (*3>P>o2) C*> £ (.3>P>.2) I> G (.05>P>.02) Kales Or) I I> G (P<.001) I> G (P>.5) Males vs. Pemales Group Raised Isolation Raised cf (.02>P>.01) cf> £ (r>.5) (.3>P>.2) &> (P <.00l) f = 'Isolation raised mice* G = Group raised mice. There were no significant■differences between isolation raised and group raised females and between isolation raised and group raised males in the distance to the nearest neigh­ bor of the same sex when compared for the total period of establishment* Only five populations of each social treat­ ment had a 17 day period of population establishment. Thus, a radical difference in the distance to the nearest neighbor on day 17 for one experiment could greatly affect the average 79 for all experiments* This happened in the measurement of distance to nearest neighbor of the same sex* Therefore, the social treatments were compared using only data from the first 16 days# Isolation raised females were signifi­ cantly further from other females,than group raised females (.05) P^ *02)* Conversely, group raised males were further from other males in the populations than isolation raised males (*1^p) .05). The average daily distance to animals of the opposite sex was significantly greater for isolation raised females (#02^-p V * 01) and males (*05)P/ *02) than their group raised counterparts* Comparison of the distance to nearest neighbors of either sex showed that isolation raised females were signif­ icantly more dispersed than group raised females (P^*00l) while isolate males did not differ from group males (P } *5). When the sexes were compared within social treatments, isolation raised females were found at greater distances from females than males were from males (.02) P) *0l). This was not statistically demonstrated for group raised females (*3>P> .2). There was no significant difference for either social treatment in the distance to nearest neighbor of the opposite sex. Isolation raised females were significantly further away than males from animals of either sex (P P> .2). The above analyses indicate that isolation raised fe­ males maintained a greater distance to nearest neighbors 80 "than group raised females* as females, however* Males were not as consistent The only significant difference found when comparing males showed isolate males further away from females than group raised males* Group males occurred further away from other males than isolate males, although not significantly so* These differences in male and female hehavior may he indicative of a differential affect of the social treatments upon the two sexes*' Homing Establishment of the homing phenomenon* Homing was established as a phenomenon in a series of 9 experiments involving 4 populations of isolation raised mice and 5 populations of group raised mice. Following the 36 hour period in the laboratory, a total of 30 isolation raised and 35 group raised mice were reintroduced into the field* The results of these experiments will be discussed as com­ bined data, incorporating material from both social treat­ ments, and separately with regard to each social treatment. Figure 17 shows the percentage of mice homing by the first and third day after reintroduction to the field* Out of 65 animals released, 45 homed by day 1 and 52 by day 3* Since each mouse occupied an average of 4*3 nest boxes dur­ ing the last 10 days preceding removal from the field for the homing test (Table V), a homing success required a rigorous discrimination of nest boxes. The homing behavior of populations of isolation raised and group raised mice is shown in Figure l80 Thirty isola— Gl tti M M M CD in CVJ CD < CD U CM IB ■ IB ( ■ ■ ■ ■ ■ ■ ■ ■ I “ “ ■ ■ ■ ro >% o o o; O 0 1 C O r m b i M Pi I—i a u ■B. i ..r. .r.w.nylf.ir.g.T..Mli, a ■■■iB-n ra 9 'yi b r i * ■ mm ■ ■ ■ ■ ■ ■ ■ ■ ■ ■ • ■ ■ ■ M B ■ ■ ■ ■ ■ ■ ■ ■ ■ ■ ■ ■ ■ ■ N o o •rH C CD • wmm < Q> w. O HQ> m mmfmm o mmfm O CD a 0 } 1AI CD JO o a> c o jD Q> o: -1^ < o (/) or CM UJ tn 00 ro 3DIH 30 3 9V l N 3 0 U 3 d CM Fig. 18. Percentage of mice homing before aliens - social treatments. E (p < . o o i) 0> E ( P < . 0 0 l) pVE (.0 1 > P > .0 0 1 ) I^ G ( „ 0 1 > P > .0 0 l) E = Homing expected. I = Isolation raised mice. G = Group raised mice. A comparison of the homing performance of the mice of the two social treatments (Table XVI) showed that a signif- 84 icantly higher proportion of isolatinn raised than group raised mice homed on both day 1 (P<.05) and day 3 (P--<.Ol). with aliens temporarily in the fieldo As men­ tioned in the procedures, the homing phase of experiments 6, 7, and 8 differed from those of the preceding experi­ ments in which homing was established as a phenomenon. During the later experiments, young male aliens were re­ tained in one-half the nest boxes during the first night following reintroduction of the residents. Figure 17 shows the combined homing performance for all mice released during these experiments. Of the 42 mice released, 11 homed by day 1 and 26 by day 3* With regard to the social treatments, a total of 19 isolation raised mice were released while aliens were in the plots, of which 5 homed by day 1 and 13 by day 3* Of the 23 group raised mice reintroduced to the plots, 6 homed by day 1 and 13 by day 3* Table XVII shows the results of tests comparing the homing performance before and after aliens were temporarily, in the plots. Exact probabilities were calculated in those comparisons in which expected values less than 5 occurred. On day 1, homing performance was significantly poorer than that recorded in establishing the homing phenomenon. Thus, introduction of aliens for one night significantly reduced the homing performance on that night. 85 TABLE XVII COMPARISONS OP HOMING PERFORMANCE BEFORE AND AFTER ALIEN INTRODUCTION Larger Frequency and Significance Day 1 Day 3 Comparison Comparing homing before and after aliens Combined data B)A (P<.00l) B>A (.1>P>.05) Isolatioh raised B> A (P<.00l) B> A (P=.0116) B> A (.05>P>.02) ND Group raised a T= Homing with aliens^temporarily in field# B = Homing rate when establishing homing as a phenomenon. ND = Difference in frequencies small or none. The presence of aliens significantly reduced the hom­ ing performance during the first night for both isolation raised and group raised mice. Homing performance by day 3 was significantly poorer than before aliens for isolation raised mice only. Differential homing with aliens in the field. As in­ dicated in the procedure, the plots were divided transversely and aliens were placed in the 12 nest boxes in one—half of each plot. Thus, only half the mice of each reintroduced population were homing to nest boxes into which aliens had been placed. Figure 19 illustrates the combined homing behavior un­ der these differential conditions. The performance of iso­ lation raised and group raised mice is shown in Figure 20. 86 l or E ri"f w i r i ' T f i t i i w,B"W"g w w i g w aa B a a a a aa aa aa aa aa aB aa ae aB aaaBaBaaaaBa a a a B A sa a aa a p ia a a a a iaaaaaaaa aaaa a a a a a a a a a a a a a a a B a a a a a as as a a a a a a a a a a a a a a a a a a a B a a a a a a a a a a a a B a B C\J a a a a a a a a B a a a a a B a a a a a a a a a a a a a i- a,aaaaaaiaaaaaaaaai o of mice homing to alien occupied nest hoxes - combined data. Ll I £ rfBaABakal T3 0> a. ' WITWW 3 o o O — C .2 _ CM CVJ i■a■a■a■a »a u ai a II a a a» ama a _ a a a a a a s a B a a a a B ^ e a e a a a s a a a laaaaaaaaai a a e a a a a a B B l a a a a a a a a a i a a a a a a a a l a a s a a a a a a a a a a B a a a e a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a aa a aa aa oal aaa aa ae aa a aa a a a a a a a a a a a a a a a a a a a a i a a a a a a a a a a m . a j ^ a m. a. n « - . i « a H CD i/) a CD m Percentage Q to empty fO O'N •H "I" O O o o 00 (£> 8 0 ! 1^1 JO O O CVJ 8 6 0 |U30J8d 87 I... > h~ CL 5 “O UJ CD CM to O q : GL IO O 3 O W111T111n 1111 ^ W 1■11!i!i!i1 UJ q - CD >- < Q ITT ilU r L O A o Q CD O I < r» 03 •H aC wD a Pi -p CD pm P 4 a © * O O ■ 1:i ' i ‘ ‘ ' - - -P CvJ ■ ttfj PR LUO. •H =■ 3DIW JO 39VJLN3Db3d 88 Table XVIII shows the results of statistical comparisons of homing to each type of nest situation with the perform­ ance prior to alien introduction. TABLE XVIII HOMING TO ALIEN OCCUPIED OR EMPTY NEST BOXES Larger Frequency and Significance Day 1___________ Day 3______ Homing after aliens compared with that before aliens To empty nest boxes Combined data B >A (,1>P>.05) ND Isolation raised B> A (.02}P>.0l) By A (.05>P>.02) ND ND B> A (P^.OOl) B> A (.05>P>.02) Isolation raised B> A (P=.00007) B> A (P=.0136) Group raised B> A .01>P).00l) B> A (.5)P).3) Group raised To occupied nest boxes Combined data A = doming with a l i e n s - 'temporarily in field. B = Homing rate when establishing homing as a phenomenon. ND = Difference in frequencies small or none. When the data for both social treatments were com­ bined, homing to empty nest boxes was not significantly different than that observed in populations before the aliens were placed in the field (Table XVI). Of the 20 mice released, 9 had homed to empty boxes by day 1 and 14 89 "by 3ay 3« Only 2 of the 22 mice released which were homing to occupied nest "boxes did so by day 1 while 12 had homed by day 3* The homing success to occupied nest boxes was significantly less on both day 1 (P ('.OOl) and on day 3 (.05) P > .02). Thus, while the combined homing rate to empty boxes was not different from the established rate on day 1 or 3j homing to alien occupied nest boxes was significantly less on both days. Isolation raised mice homed significantly less after aliens than before aliens no matter whether they were re­ turning to empty or occupied nest boxes. Of the 8 mice re­ turning to empty nest boxes (Figure 20), only 3 homed by day 1 (.02)P^.0l) and 6 by day 3 (.05,)P}.02). Only 2 of the 11 mice homing to alien occupied nest boxes did so on day 1 (P = .000.07) and 7 by day 3 (P = .0136). Thus, the presence of aliens in the field for the first night after reintroduction of the residents disrupted the homing per­ formance of isolation raised mice returning to empty, as well as occupied, nest boxes. This was true on both day 1 and day 3° Subsequent to alien introduction, group raised mice did not significantly differ in their homing to empty boxes from the rate observed prior to introduction of aliens. Of the 1} mice returning to empty nest boxes, 6 did so by day 1 and 8 by day 3. There were 12 group raised mice homing to occupied nest boxes. Of these none homed by day 1 and 90 only 5 did so by day 3° Comparison of the proportion of homing successes observed with that in establishing homing as a phenomenon shows a significant difference on day 1 only (.01^ P ^ .001)o The deleterious affects of introduction of aliens upon homing performance was evident only upon these mice homing to alien occupied nest boxes. This was true only for day 1 and did not have the longer lasting effects as observed for the isolation raised mice. Differences in homing behavior following introduction of aliens were evident for both isolation raised mice and group raised mice when compared with homing performance established before aliens. G-enerally, isolation raised animals showed greater differences in homing following alien introduction than group raised mice. Comparison of Social Treatments. Following alien introduction there were no significant differences between social treatments in homing per se, homing to empty nest boxes, or homing to occupied nest boxes. Occurrence in Empty or in Alien Occupied Nest Boxes Figure 21 bhows the percentage of reintroduced residents which were found in nest boxes in which an alien had not been placed or in nest boxes in which aliens had been pres­ ent. A total of 42 mice were reintroduced into the field as part of the homing tests. Of these 6 were found in occu­ pied and 26 in empty nest boxes on day 1. By day 3 twenty mice had been recorded in alien occupied and 2/7 in empty nest boxes. The number of mice recorded for day 3 exceeded 91 OJ Q. £ LU O c CVJ O 3 O o C — - O OJ ww w w mmw w'u w * w wrvrw m a a a a a a a a a a a a a a a l a i a l a i a a a i a a a i a a a i a a a i a a a i a a a a B a a i a a ten had been in alien occupied boxes and 16 had been in empty boxes. Since one-half of the nest boxes were empty and onehalf were occupied by aliens, half of the mice could be expected to be in each type of nest box. Mice occurred in empty nest boxes significantly more often than expected by chance for both the combined data and for the populations of each social treatment. Comparison of the total numbers of mice found in each of the two types of nest boxes revealed that a significantly higher proportion of the mice were found in empty boxes on d:ay 1 than were found in occupied boxes. 93 >H CD S CD rO CL ui TJ <1> CO o 111111i.i.i O Q 2 ^ O r?1 11 ii o ild d CD CL 3 O w o UJ CL < o cr $ O PO CM •H 9 O . 03 O P oa CD pi a CD P •H 03 rH CD £d-f H P •H rH OJ d •H O O o 03 >»- CL s UJ “O in O cc I CD O 03 •H ID ax o o p CD 03 t*Q CD ctf pi P Pi CD P O Ph p) a CD CD « •H CM CM tiD •H £4 301W JO 3 9 V l N 3 0 « 3 d 94 Table XIX shows the results of statistical comparisons of these data. TABLE XIX OCCURRENCE IN EMPTY OR IN ALIEN OCCUPIED NEST BOXES Comparison Larger Erequency and Significance Day 1 Day 3 Number occurring in each type of nest box Compared with chance (P=.5) Ey0 (P<.00l) Combined data Isolation raised E> 0 (.05>P>.02) Group raised E> 0 („02>P>.0l) Comparing numbers found in each type of nest box. M> A (P<„001) Combined data Isolation raised M >A (.Ol)P>.001) Group raised Comparing social treatments M> A (P<.Q0i) CP/*5; E = Expected frequency of mice 0 = Observed frequency of mice M Ereouency of mice found in A = Ereouency of mice found in ND = Difference in frequencies M> A (.3>P>.2) HD (P>.5) M> A (.2>p).i) ND (P>.5) In "each type of nest box. in each type of nest box. empty nest boxes# alien occupied nest boxes small or none# 95 When the social treatments are evaluated separately, the proportion of mice occurring in empty boxes was signif­ icantly larger than the proportion in occupied nest boxes for day 1 in both isolation and group raised mice. There was no difference in nest box occurrence by day 3# These data, therefore, indicate that mice occurred in empty nest boxes significantly more often than expected by chance* On day 1, a significantly greater proportion of mice occurred in empty nest boxes than in boxes occupied for one night by a young, male, alien mouse. By day 3, two days after removal of the aliens, this preference for empty nest boxes had disappeared. 96 DISCUSSION Factors affecting populations in the wild are diffi­ cult to ascertain experimentally due to lack of control of the experimental situation* Loss of animals due to dis­ persal (Blair, 1940; Howard, 1949? King and Eleftheriou, 1957) and death from many causes make it difficult to control numbers while studying related variables* This study was an attempt to combine the laboratory and field approach with a minimum loss of their individual ad­ vantages (Schneirla, 1950; Scott, 1950)* Thus, the biology of populations, both physical and social, may be studied under semi—natural conditions while maintaining, partially at least, a measure of the controlled conditions enjoyed in the laboratory. Accordingly, the "mouse—proof " experimental field has proven an effective method for studying populations. It appears that this technique has almost unlimited applica­ bility to the study of the population dynamics of small mammals. The results of these experiments naturally separate into two major frames of reference and will be discussed accordingly. The first area for consideration concerns those data suggesting the existence of phenomena which may be specific to populations of prairie deermice, to mice of the species Peromyscus mani culatus, or that have general significance with regard to small mammals. The second category concerns the differential behavior of the populations 97 as a result of the two social treatments, in which mice were raised in isolation or in groups* Population Phenomena Care should he taken in the extrapolation to natural populations of conclusions based upon data obtained under artificial situations* The results discussed here, and the hypotheses suggested by them, may be functions of the experimental situation only* Movement data* The frequency of moves, and the dis­ tance moved to nest boxes during the last 10 days of the period of population establishment, showed that between 55 and 60 per cent of the mice moved to different nest boxes each night* Although more than half of the mice moved to different nest boxes each night, less than half of these moves were to boxes previously unvisited by each mouse* Therefore, while mice moved at a relatively constant rate, they moved less frequently to unvisited boxes, even though there were many available* Thus, each mouse localized to the use of a few nest boxes among which it continued to move* The commonly accepted concept that an animal maintains one nest site around which its activity centers was contra­ dicted by the evidence cited above* The idea of a single nest site has conceptual value in the explanation of terri­ torial behavior in mammals (Burt, 1940, 1948, p® 21)• It may, however, lead to error when dealing with species in which such defense behavior is questionable (Blair, 1940, 1951, 1953b). 98 The fact that each mouse used an average of 4*3 nest boxes during the last 10 days and subsequently homed to one of these boxes raises two questions. Were these multiple residence sites only a function of the large number of boxes available or was each a home site or a refuge? tions cannot be conclusively answered here® These ques­ Howard (1949) studying natural populations of prairie deermice, found that mice living in areas with a large number of nest boxes changed their hom§4ites more frequently than those living where the nest boxes were more widely distributed. Since he visited his nest boxes only once a week, his experiments were not designed so that the number of nest boxes used per mouse could be ascertained. Nicholson (1941) studying the White­ footed Mouse (Peromyscus leucopus noveboracensis) in the wild, found that mice left the nest boxes after a "short period of residence". During his study, 174 mice were captured more than once in nest boxes. Of these, only 16 lived in a box for more than four consecutive weeks. His data are of questionable significance in this comparison since heJ too.* checked his boxes only once a week. Blair (1940), studying prairie deermice found that in­ dividuals fled to a number of different holes after their release from live traps. He concluded that each mouse had six or more refuges within its home range, of which an un­ known number were permanent homes. His later study (1951) of the Beach Mouse (Peromyscus polionotus leucocephalus), indicated that the home range of each mouse contained an average of about 20 holes, of which a mean of 5 were entered 99 by each after release from the traps® The fact that mice stored food and nested there for short periods indicated that these holes were more than refuge holes* In this study, the mice moved among approximately 5 nest boxes during the last 10 days of each experiment. In those experiments in which some mice disappeared early in the experimental period, no increase was noted in the aver­ age number of nest boxes occupied by the remaining individ­ uals. This may indicate that each mouse selected a rela­ tively constant number of boxes to be used as refuges and/ or nest sites. Since nest boxes were checked only during the day, there was no way of distinguishing between nest sites and refuges. General nocturnal observations, however, indicated that nest boxes were used for both nest sites and refuges, depending upon the circumstances. As suggested by Burt (1940), it would be of distinct survival value for a mouse to have several nests distributed over the area. It could retreat to them and lessen the chance of death from exposure and predators. An important difference between the conditions main­ tained in this study and wild conditions is that the exper­ imental mice did not have young while in the field, although most of the females were pregnant at the end of each exper­ imental period. Certainly, care of a litter and perhaps pregnancy, may cause sedentary behavior resulting in the occupancy of fewer nest boxes<> The possibility should not be overlooked that the daily nest box check caused the mice to move to other nest boxes 100 as frequently as they did. Checking of the nest boxes seem­ ed of little significance, since only half of the mice moved each day. This indicated that movement was a matter of in­ dividual behavior. The nest box checks could not have caus­ ed the movement unless each examination affected each mouse differently each day. If mice move between temporary residence sites, home ranges and artificial distribution measurements (Burt, 1940, 1943; Blair, 1940, 1941; Hayne, 1949; Holenreid, 1940; Mohr, 1947; Stickle, 1946) calculated on the basis of a few nights sampling may be greatly in error. Gregarious behavior. The experimental mice were found alone significantly more often than in combinations. Mice were found in bisexual pairs significantly more often than in any other type of combination. Mice were in bisexual pairs significantly more often than expected by chance or than found in mono sexual pairs. Thus, a preference was shown for occurrence alone in nest boxes and for combina­ tions only as bisexual pairs. The above data illustrate the expected attraction be­ tween animals of the opposite sex. The fact that mice of both social treatments occurred alone so frequently, however, was unexpected. Howard’s (1949) nest box study suggested that deermice were infrequently found alone in nest boxes, but were usually found in bisexual pairs. Howard (ibid) further noted that when the sex ratio was unequal, several mice of the more numerous sex were simultaneously found with the opposite sex. He does not, however, give frequen­ cies or the exact time of year for these data. 101 Nicholson*s (l94l) nest box study indicated that whitefooted mice were usually found living singly. Associations with other mice were of short duration and combinations were usually as bisexual pairs. Nicholson*s study further reveal­ ed that very few associations composed of the same sex were formed during the breeding season. Of those formed, all occurred at the extreme limits of the breeding season. The data reported here were similar to Nicholson*s findings. Out of 23 combinations of the same sex, 14 occurred during the last experiment (October 31 - November 2l). These data indicate that during the breeding season, prairie deermice generally are found alone, combine rarely in other than bisexual pairs, and very rarely occur in a nest box with an animal of the same sex. These data may differ from Howard’s because of the semi—natural conditions under which these experiments were carried out. Most groups found by Howard were composed of both parents with their offspring. This was impossible under the experimental procedure utilized here and may have resulted in the differ­ ences. Attraction and repulsion between mice. little information is available regarding the responsibility for estab­ lishing or terminating bisexual pair combinations among deermice. Trapping records have been used as evidence that males have a larger home range than females (Burt, 1940; Blair, 1940; 1942). Peromyscus males are thought to move more and to greater distances than females (Burt, 1940; 102 Nicholson, 1941; Blair, 1942)* Thus, males would pass through more home ranges of females than females would of males# It might be supposed that males would combine in home sites with females more frequently than the reverse# In this study, however, no significant difference was found as to which sex originated the bisexual pair or which left it first# When the data measuring succession of mice in nest boxes were examined, it was determined that females succeeded males significantly more often than expected by chance. Males succeeded females less than expected. It is doubt­ ful that these data indicate an avoidance between opposite sexes since the number of bisexual combinations was larger than that of the other combinations. Also, general avoid­ ance of the opposite sex would not have survival value to the species. The special significance of the behavior noted above is that females succeeded males significantly more than ex­ pected. These data indicate that females follow males into nest boxes on successive days more often than males do fe­ males. There was no evidence, however, that the females initiated the formation of bisexual pairs. Evidence on this point is obliterated due to the fact that of the 116 records of the first day pairs were found together, 55 occurrences were in nest boxes in which neither mouse was recorded on the preceding day. As indicated earlier, the frequency of the occurrence of two animals of the same sex in a nest box was significantly 103 less than expected by chance. Monosexual pairs were record­ ed with significantly less frequency than bisexual pairs. In addition, there were significantly fewer records of ani­ mals of the same sex succeeding each other in nest boxes than were recorded for the opposite sex animals. These data suggest that some type of negative force was operating to segregate mice of the same sex spatially. That this separ­ ation of like sexes is not merely a result of the lack of sexual attraction between such animals, but operates through some spatially or socially directed behavior pattern, is evidenced by the low frequency with which animals moved in­ to nest boxes occupied on the previous day by like sex ani­ mals. These frequencies were significantly less than ex­ pected by chance. No conclusive information is available concerning the territorial behavior of prairie deermice, ?either in the de­ fined sense of "defense of an area" (Noble, 1939; Greenberg, 1947) or in the less rigid and more dynamic sense discussed by Emlen (1957) and described by Davis (1958), Marler (1956), and Jenkins (1944). No studies of natural populations of the species P. maniculatus have reported active defense of a home area by the mice (Blair, 1940, 1941, 1942, 1943, 1951, 1953^ Dice, 1932; Howard, 1949). Territorial behavior has* however, been inferred from live trapping studies of Peromyscus leucopus (Burt, 1940, 1949). Some evidence of a "dynamic" non-aggressive type of behavior leading to apportioning of the area among animals 104 of the same population was collected through direct obser­ vations and by the homing test# Observations made during the preliminary experiments as well as during the experimental periods discussed here, indicate that when animals met in the field, both jumped and ran away from each other. Occasionally one mouse chased another mouse a short distance. Observations of social in­ teraction between residents and aliens while aliens were teth­ ered at the feeding stations or at the home nest sites of residents, indicated that residents predominately avoided or ignored the aliens. In the few cases (5) when the res­ idents attacked the aliens, the attacks were of short dur­ ation and were terminated by the residents leaving the area. Laboratory studies (King, 1957) of P. m. bairdii showed single males to be relatively non-aggressive toward other males. Terman (1958) found single females to be slightly more aggressive than single males when aliens were placed in their home mouse boxes. The incidence of attacks was very low for both sexes, however. When these same males and females were paired and then aliens placed in the mouse box, the number of attacks by residents greatly in­ creased and the males were usually the attackers. gression by the male was usually of short duration. Such ag­ In these laboratory experiments, neither the resident or the alien could leave the mouse box, so fighting was generally terminated by the alien assuming a defense position and the resident hesitating to attack. In the experiments in the field, however, the residents were free to leave the area and did so. 105 The results of the homing experiments with aliens in the field further elucidate this problem. Homing to alien- occupied nest boxes while young aliens were in the field, was significantly less than that observed before aliens were in the field. ments. This was true for both social treat­ Avoidance of aliens was further demonstrated by the occurrence of the mice in empty nest boxes significantly more often than in alien occupied boxes. These data indicate that spatial distribution may be affected through avoidance or some unmeasured negative force between animals (Howard, 1949). Prior occurrence (Braddock, 1949) or merely pres­ ence in the area may be the important factors. With regard to the question of territoriality, then, no reliable evidence for active defense of an area was shown. Animals of the same sex may be spatially segregated as a result of a negative repulsive force, hypothetically, avoidance. Homing. Homing ability in species of Peromyscus has been demonstrated in many studies (Burt, 1940; Johnson, 1926; Kendiegh, 1944; Murie and Murie, 1931, 1932; Stickle, 1949). Murie and Murie, (1931) reported a few mice return­ ing from distances of two miles and many returning from shorter distances. Homing success in the above mentioned studies was measured by capture of mice in traps in their inferred home areas subsequent to their release at varying distances. As such, the above studies have measured the return of Peromyscus to a home area. 106 Homing was used in this study to measure the signifi­ cance to the mice of the previously established spatial distribution patterns, as shown by their ability to home to a few specific nest boxes. Also, the homing perform­ ance before aliens were in the field was compared with the homing performance after aliens were temporarily in the field. Both socially treated populations homed significantly more often than expected by chance# This indicated that the mice were able to "traverse at least half the length of the experimental plot, bypass numerous nest boxes, and return to one of the few (4 or 5) boxes previously occupied. Further, mice that did not home by day 1 frequently did so by day 3. Such performance indicated that the individual mice recognized both their "own" nest boxes and the nest boxes of their neighbors* These homing data may be indicative of the existence of some sort of spatial framework or "positional stability" (Orr, 1955) as a characteristic of each population. Howard (1949) and Dice and Howard (1951) found little tendency for prairie deermice to move from home areas once they had bred there. Studies of natural populations of Peromyscus have shown that removal of all individuals living within a spec­ ified area was followed by immigration to the vacated area by mice living in adjacent areas (Blair, 1940; Calhoun and Webb, 1953; Stickle, 1946). Such behavior is an indication of the part social interaction may play in the distribution 107 of mice. Orr (1955) suggested that animals were aware of neighboring individuals living around the periphery of their home range and when these neighboring animals were removed by one cause or another, the social stimuli were also removed, with the result that the remaining animals expanded their ranges. Introduction of 3 week old, alien males into the home nest boxes of residents disrupted homing per se for both social treatments. There was no significant difference between the homing performance of mice from either social treatment following the introduction of aliens. Examination of the homing data following introduction of aliens revealed that the decrease in homing success was due to differential homing to nest boxes occupied by aliens and to boxes in which aliens had never been. The data indicate that the difference in homing performance after aliens were intro­ duced was primarily due to the poor homing success of mice returning to alien occupied nest boxes. The importance of social behavior as a determiner of spatial distribution is, therefore, evident. The influence of social factors was further demonstrated by the few occurrences of resi­ dents on day 1 in any nest boxes occupied by young aliens. What significance does the avoidance of aliens have in relation to the spatial distribution within local populations and to the geographical distribution of prairie deermice? The data collected in this study suggest that spatial dis­ tribution may be achieved by mutual avoidance between ani— 108 mals of like sex. Statements regarding heterosexual inter­ action should not he made on the basis of the evidence avail­ able. Further, such interactions affecting spatial dis­ tribution undoubtedly vary. Estrus and the care of young, no doubt, affect female distribution patterns. A population of prairie deermice may space itself as a result of some intrinsic mutual repulsion mechanism which is adjusted to the physical as well as the biotic environment, and is not a function of territoriality in the defined sense. Such a mechanism was discussed by Frank (1957) for Microtus arvalis and Microtus -agrestis and termed the "condensation potential". The importance of individual behavioral differ­ ences within this concept was discussed in the introduction of this paper. Spatial distribution through mutual avoidance of in­ dividuals would appear to be of at least equal adaptive significance to the biology of the population as that suggested for territoriality by Burt (1949). Mutual avoid­ ance behavior would achieve distribution in accordance with the carrying capacity of the environment without the wound­ ing and deleterious effects of fighting (Clarke, 1955; Calhoun, 1950, 1952) and social stresses (Christian, 1956, 1957)« Such lack of overt aggressinn among prairie deermice has been previously pointed out in this study and in a lab­ oratory study by King (1957). King mentioned that in con­ fined P . m . bairdii males used in his tests "frequent nos­ ing and grooming behavior suggested dominance without a fight". 109 Howard (1949) studying dispersal from "birth place to breeding site showed that 119 or 76/^ of 150 young deermice moved less than 500 feet from the site where they were born before breeding and nesting* These data were biased by the fact that the greater the distance the mice moved, the less chance there was of their being recorded in nest boxes a— gain* The proportion of mice moving this short distance from birth site to breeding site, however, is indicative of a short dispersal pattern (Dice and Howard, 1951; Blair, 1953b)* It is of further interest that evidence exists for a major range extension of P • m* bairdii and of the species P. maniculatus in recent times (Blair, 1953a, 1953b). Extension of the range of the local population of prairie deermice as well as the geographical distribution of the species may be largely through a diffusion—like process rather than by long individual moves. Young animals, upon leaving the nest may move into an occupied area or into a temporarily empty nest site rather than making long moves until an unoccupied area is found. Such behavior could cause a partial displacement of the residents due to avoid­ ance behavior, and result in a gradual extension of the range at the periphery* Social Behavior In the previous section, it has been shown that social interaction between animals is important in determining spatial distribution within local populations. This study 110 was designed to manipulate two social variables and measure their differential affects upon the spatial distributions of mouse populations under semi—natural conditions. While some measurements revealed behavioral patterns which were unaffected by the social manipulations and apparently were characteristic phenonomena of the populations, other meas­ urements showed a consistent difference between the popu­ lations raised in groups and in isolation. The measurements of gregarious behavior indicated that isolation raised mice were recorded alone in nest boxes proportionately more often than group raised mice. The reciprocal of this significant result is that the pro­ portion of nest box records in which one mouse was combined with at least one other mouse was higher for group raised mice than for isolation raised mice. No significant differences were found between social treatments when comparing various data having to do with bisexual combinations, although consistent trends were noted. G-roup raised mice were consistently different from isolation raised mice in,that a higher proportion combined with the opposite sex, they were found in a larger number of different combinations, and they continued in combina­ tions for a longer period of time. Group raised mice combined with the opposite sex significantly sooner after introduction into the plots than did isolation raised mice. Daily comparisons of the average distance between the nearest neighbors of the same, opposite, or either sex showed Ill "that isolation raised females were consistently and sig­ nificantly further away from their neighbors than group raised females* Isolate male mice were significantly fur­ ther from females than group raised males* G-roup raised males were further from other males than isolate males were, but not significantly so (.1 > P X05)« It is possible that there was a differential sex affect of the social treatment which an eventual analysis of the data will re— ' veal. The homing performance after aliens were introduced to the nest boxes showed no difference between social treat­ ments in homing per se, to nest boxes in which aliens had never been placed, or to alien occupied nest boxes. Mice of both social treatments homed to alien occupied nest boxes significantly less frequently than they did when homing was established as a phenomenon. Group raised mice showed no significant difference between their established rate of homing and their homing performance to alien-free nest boxes after the aliens were in the field. During and following the time aliens were in the field, however, iso­ lation raised mice homed significantly less often to boxes which never received aliens than was the case in the homing establishment experiments. The significant difference be­ tween pre and post-alien homing performance to empty nest boxes for isolation raised mice, was due to their better homing performance as compared to that of the group raised mice during the experiments before aliens were in the field. 112 Thus, the introduction of aliens to half the nest boxes had a more adverse affect upon isolation raised mice than group raised whether they were homing to empty or alien occupied nest boxes. Group raised mice homed to empty nest boxes signifi­ cantly more than to alien occupied on day 1 only. Isolation raised mice did not home with a significantly different frequency to either type of nest box. There were no sig­ nificant differences between social treatments in the num­ ber of mice found in empty or alien occupied nest boxes. The comparison of social treatments during the experi­ ments when homing was established as a phenomenon showed that although mice of both social treatments homed by day 1 significantly more often than expected by chance, the iso­ lates homed significantly better than the group raised. Not only did isolates home more often than group raised mice but the difference in homing performance between the two social groups was greater and was significant at a higher level of probability on day 3 (.01^ on day 1 (.05^ P^.02). .001) than Thus, it appears that isolate ani­ mals which did not succeed in homing on day 1 sought to re­ turn to a previously occupied box by day 3 more often than did the group raised mice. A summary of the behavioral characteristics of iso­ lation raised mice as opposed to group raised mice is as follows: the isolation raised mice combined with others less often than the group raised; were slower in combining; generally, with few exceptions, maintained a greater distance 113 from their fellows; and homed significantly more often.• Differential homing behavior after aliens were in the field, showed isolation raised mice to be more adversely affected by the introduction of aliens than were group raised mice# Isolation raised mice thus appeared to be less sociable and more spatially oriented than group raised mice* Dew studies of the affect of isolation upon social be­ havior have been made among mammals (Beach and Jaynes, 1954)* Zing and Gurney (1954) and Zing (1957) reported that male C57BL/10 mice, raised in isolation from weaning, were less aggressive than their controls raised in social groups. Kahn (1954) found that male Mus raised in isolation from weaning were more aggressive than males raised with their mothers. King and Eleftheriou (1957) raised Peromyscus in isolation and in groups and then released them into the wild in an effort to ascertain the affects of social experience upon adaptation to the natural environment. They were greatly hampered by a precipitous decline in the populations by the end of the first week, but observed that isolation raised mice were found together in nest boxes less frequently and moved about the field more and to greater distances than group raised. With the above facts in mind, the greater homing per­ formance of isolation raised mice may be hypothetically ex­ plained in the following manner: spatial and social pat­ terns of distribution were of greater significance to the isolation raised mice. As was pointed out earlier, mice and perhaps other small mammals, may maintain a positional 114 stability or equilibrium. Isolation raised mice, after once establishing themselves in a few nest boxes, seek to return to this social and spatial equilibrium to a greater extent than the more sociable group raised mice. This is undoubtedly not an active seeking but rather that the bal­ ance of social and spatial stimuli is not similar to earlier adapted levels until the mice are back in their home areas. Since group raised mice are socially better oriented, ad­ justment to different social and related spatial stimuli may be more easily made. The fact that there was no difference in homing behavior between isolation raised and group raised mice after aliens were in the field, while there was a significant difference before aliens, signifies that the introduction of aliens had a more adverse affect upon the homing of the former. The lack of homing difference between social treatments after aliens indicates that the avoidance of aliens or alien occupied nest boxes may be a natural population phenomenon, and that the basic spatial equilibrium pattern existent in the study area before introduction of the aliens had been disrupted. As was mentioned earlier, homing with aliens in the field took place in a series of experiments immediately following those establishing homing as a phenomenon. The difference in the time when each series of experiments was performed cannot be considered influential in the poorer homing performance while aliens were in the field, since 115 the group raised mice showed no reliable differences in homing to empty nest boxes before or after aliens. Social behavior, therefore, may be an influential force shaping spatial distribution patterns within popu­ lations. Social forces may operate as general behavioral characteristics, e.g., territorial behavior, the conden­ sation potential of the species, or they may be effective through individual intrinsic behavioral differences (Christian, 1957; King, 1957; Southwick, 1955; Wellington, 1957). It is not inoonceivable that wild, free—living mammals may experience behavioral manipulation no less rigorous than the techniques employed here. Such behavioral varia­ tions would have great importance in the genetics, evolu­ tion, and dynamics of populations. 116 SUMMARY Successive populations of Prairie Deermice were raised in the laboratory in isolation or in groups* The affect of the social treatments upon the subsequent spatial distribu­ tion of the mice in a semi—natural environment was studied* There were eight experimental periods, each of three weeks duration between June 6 and November 21, 1958* During this time, 8 populations of four bisexual pairs, raised either in isolation or in groups, were living in each of the two 0*44 acre plots of the experimental field. Por 17 days following the release of the mice into the plots, their daily occurrence in nest boxes was recorded* At this time all mice were removed from the plots and kept in isolation in the laboratory for 36 hours* Each was then reintroduced into its home plot at a point distant from its previously established home area* The location of each mouse for the nest three days was recorded. In the last 3 experiments, young alien mice were retained in one—half the nest boxes during the first night after reintroduction of the residents* The "mouse—proof" field proved to be an effective method for the study of population dynamics under semi-natural con­ ditions while maintaining a measure of the controlled con­ ditions which are possible in the laboratory* The following are phenomena noted in this study which may be specific to populations of prairie deermice, to mice 117 of the species Peromyscus maniculatus; or which may have general significance to small mammal populationsi a* Between 55 and 60 per cent of the mice moved to different nest boxes each night* Less than half of these moves were to boxes previously unvisited by the mouse mov­ ing* Therefore, each mouse localized and used a few nest boxes among which it continued to move* These data indicate that each mouse maintained several refuges and/or nest sites rather than a single one around which its activity centered* Such behavior should be considered when measurements are spatial made of the/patterns of individual animals in the wild* b. During the breeding season, the prairie deermice in this study were generally found alone or in bisexual pairs, combined rarely in other than bisexual pairs, and very rarely occurred in a nest box with an animal of the same sex* c* No conclusive evidence was obtained as to which sex originated or terminated bisexual pairs* d* Mice of the opposite sex succeeded each other in nest boxes significantly more often than those of the same sex* Pemales followed males into nest boxes on successive days significantly more often and males followed females significantly less often than expected by chance. e. No reliable evidence for territoriality as defense of an area was obtained* The data suggest that animals of the same sex may be spatially segregated as a result of a negative repulsive force, hypothetically, avoidance* 118 f* Mice of both social treatments homed to previously occupied nest boxes significantly more often than expected by chance* This indicated that the individual mice recog­ nized both their "own" nest boxes and those of their neigh­ bors* Thus, some sort of spatial distribution framework or a "positional stability” may be a characteristic of each population* g* Mice of both social treatments homed significantly less often to nest boxes temporarily occupied by young aliens than they did in of aliens* the experiments prior to the introduction This significant decrease in homing supports the hypothesis that the spatial distribution of prairie deer­ mice may be achieved through a mutual avoidance of individ­ uals* Extension of the range of the local population as well as of the geographical distribution of the species may be largely through a diffusion-like process rather than by long individual moves* Young animals, upon leaving the nest may move into an occupied area or into a temporarily empty nest site rather than making long moves until an unoccupied area is found* Such behavior could cause a partial displace­ ment of the residents due to avoidance and would result in a gradual extension of the range at the periphery. The differential effects of the social treatments were as follows: a. The isolation raised mice combined with others less often than the group raised mice; were slower in combining, and generally, maintained a greater distance from their fellows. 119 b« Isolation raised mice homed significantly more often than group raised during the experiments when homing was established as a phenomenon* The introduction of aliens into one—half the nest boxes had a more adverse affect upon the homing performance of isolation raised mice than upon that of group raised* c* Isolation raised mice appeared to be less sociable and more spatially oriented than group raised mice* d* The data suggest that spatial patterns of distrib­ ution existent in the plots were largely determined by social interaction and were of greater significance to iso­ lation raised mice than to group raised* Isolation raised mice adapted less easily to changes in the social and re­ lated spatial stimuli than the more sociable group raised mice and, thus, more frequently returned to the earlier established patterns of spatial and social stimuli* The introduction of aliens disrupted the social— spatial equil­ ibrium existent in the plots. This disruption had a more severe and longer lasting affect upon isolation raised mice than upon group raised due to the inability of the former to quickly adapt to the environmental changes* d* Differences in social behavior have been shown to be important factors determining spatial patterns of dis­ tribution within local populations* The importance of social factors to the evolution, genetics, and dynamics of populations was also suggested. 120 LITERATURE OITED Allee, W. C., A* E. Emerson, 0* Park, T. Park and K. P. Schmidt. 1949. Principles of animal ecology. Saunders, Philadelphia. 837 pp. Andrewartha, H. G. and L. C. Birch. 1954. The distribu­ tion and abundance of animals. The University of Chicago Press. Chicago. 782 pp. Beach, P. A. and J. Jaynes. 1954. Effects of early ex­ perience upon the behavior of animals. Psychol. Bull., 51s 239-263. Blair, W. P. 1940. A study of prairie deer-mouse popu­ lations in southern Michigan. Amer. Midi. Nat,, 24s 273-305. ______ • 1941. The small mammal population of a hardwood forest in northern Michigan, Contr, Lab. Yer. Biol. Univ. Mich., 17: 1-10. . 1942. 'Size of home range and notes on the life history of the woodland deer-mouse and eastern chip­ munk in northern Michigan. Jour. Mamm., 23: 27-36. . 1943* Populations of the deer-mouse and assoc­ iated small mammals in the mesquite association of southern New Mexico. Contr. Lab. Vert. Biol. Univ. Mich., 21: 1-40. . 1951. Population structure, social behavior, and environmental relations in a natural population of the beach mouse (Peromvscus polionotus leucocephalus). 1953a. Factors affecting gene exchange between populations in the Peromyscus maniculatus group. The Texas Jour. Sci., 5: 17-33. . 1953b. Population dynamics of rodents and other small mammals* Advances in Genetics, 5: 2—37. . 1956. The species as a dynamic system. 'Southwestern Naturalist, 1 : 1-5* The Braddock, J. C. 1949* The effect of prior residence upon dominance in the fish Platypoecilus maculatus. Physiol. Zool., 22: 161-169. 121 Brand. R. A* 1955* Abundance and activity of the woodmouse IPeromyscus leucopus noveboracensis) in relation to the characteristics of its habitat. Ph. D. Thesis, Univ. of Michigan,. Brown, R. Z. 1953. Social behavior, reproduction, and population changes in the house mouse (Mus musculus L.). Ecol. Mono., 23: 217-240. Burt, W. Ho 1940.: Territorial behavior and populations of some small mammals in southern Michigan. Misc. Pub. Mus^ Zool. Univ. Mich., 45: 1—58. ______ * 1943. Territoriality and home range concepts as applied to mammals. Jour. Mamm., 24: 346- 352. . 1948. The mammals of Michigan. Ann Arbor. 288 pp. ______ . 1949. Territoriality. Univ. Mich. Press, Jour. Mamm., 30: 25-27. Calhoun, J. B. 1948. Mortality and movement of brown rats (Rattus norvegicus) in artifically supersaturated populations. Jour7^Kldl. Mgt., 12: 167-172. ______ . 1949. A method of self-control of population growth among mammals living in the wild. Science, 109: 333-335. . 1950. The study of wild animals under controlled conditions. Annals New York Academy Sci., 51: 1113—1122. . 1952. The social aspects of population dynamics. Jour. Mamm., 33: 138—159. 1956. A comparative study of the social behavior of two inbred strains of house mice. Ecol. Mono., 26: 81-103. Calhoun, J. B. and W. L. Webb. 1953. Induced emigrations among small mammals. Science, 117: 358—360. Christian, J. J. 1955a. Effect of population size on the weights of the reproductive organs of white mice. Amer. J. Physiol*, 181: 477—480* 1955b. Effect of population size on the adrenal glands and reproductive organs of male mice in popula­ tions of fixed size. Amer. J. Physiol., 182: 292-300. . 1956. Adrenal and reproductive responses to popuXation size in mice from freely growing populations. Ecology, 37: 258-273. 122 « 1957* A review of the endocrine responses in rats and mice to increasing population size including de­ layed effects on offspring# Naval Medical Research Institute# Lecture and Review Series, No. 57: 443-462. Clark, Po 1938. Age of sexual maturity in mice of the genus Peromyscus. Jour. Mamm., 19: 230-234. Clark, P. F* and F. C. Evans, 1954. Distance to nearest neighbor as a measure of spatial relationships in populations. Ecology, 35: 445-453. . 1955. On some aspects of spatial pattern in bio­ logical populations. Science, 121: 397- 398. Clarke, J. R. 1955. Influence of numbers on reproduction and survival in two experimental vole populations. Proc. Roy. Soc., London, s. B, 144: 68—85# Cole, L. C. 1957. Sketches of general and comparative demography. Cold Spring Harbor Symposia Quant. Biol., 22 : 1-15 . Crowcroft, P. 1955# Mus musculus L. Territoriality in wild house mice, Jour. Mamm., 36: 299—301. Crowcroft, P. and F. P. Rowe. 1957. The growth of confined colonies of the wild house-mouse (Mus musculus L.) Proc. Zool. Soc. Lond. 129: 359—37^ . Davis, D. E. 1958. The role of density in aggressive be­ havior in house mice. Anim. Behav., 6: 207— 211. Davis, D. E. and J. J. Christian. 1956. Changes in Norway rat populations induced by introduction of rats. Jour. Wild. Mgt., 20: 378-383. Dice, L. R. 1932. The prairie deer-mouse. brook Inst. Sci., 2: 1-8. . 1952. Natural communities. Ann Arbor. 547 pp<» Bull. Cran- Univ. Mich. Press, Dice, L. R. and R. M. Bradley. 1942. Growth in the deermouse, Peromyscus maniculatus♦ Jour. Mamm., 23: 416— 427. Dice, L. Rv, and W. E. Howard. 1951. Distance of dispersal by prairie deermice from birthplaces to breeding sites. Contr. Lab. Vert. Biol. Univ. Mich., 50: 1-12. Emlen, J. T. 1957. Defended area — A critique of the territory concept and of conventional thinking. Ibis, 90: 352. 123 Fernald, M. L* 1950* Gray*s manual of botany. Amer. Book Co., New York. 8th Edition. Frank, F. 1954. Die Kausalitat der Nagetier— Zyklen im Lichte neuer populations— dynamischer Untersuchungen an deutsohen Microtinen. Zeit— schr. f. Morphol. u. Oekol., 43: 321-356. -- - • 1957. The causality of Microtine cycles in Germany. Jour. Wild. Mgt., 21: 113-121. Greenberg, B.^ 1947. Some relations between territory, social hierarchy, and leadership in the green sunfish (Le-pomis cyanellus). Physiol. Zool., 20: 267-299. Hayne, D. W. 1949. Calculation of size of home range. Jour. Mamm. 30: 1—18. Holdenried, R. 1940. A population study of the long-eared chipmunk (Eutamias quadrimaculatus) in central Sierra Nevada• Jour. Mamm., 21: 4o5-4li. Howard, E. 1920. Territory.in bird life* Press, London. 224 pages. John Murray Howard, W. E. 1949. Dispersal, amount of inbreeding, and longevity in a local population of prairie deermice on the George Reserve, southern Michigan* Contr. Lab. Vert. Biol. Univ. Mich., 43: 1-50. Jenkins, D. W. 1944. Territory as a result of despotism and social organization in geese. Auk, 61: 30-47. Johnson, M. S* 1926.Activity and distribution of certain wild mice in relation to biotic communities. Jour. Mamm., 7: 245-278. Kahn, M. V. 1954. Infantile experience and the mature aggressive behavior of mice: some maternal influences. Jour. Genet. Psychol., 84: 65—75. Kendeigh, S. C. 1944, Homing of Peromyscus maniculatus gracilis. Jour. Mamm., 25: 405:4(56. King, J. A. 1957. Intra- and interspecific conflict of Mus and Peromyscus. Ecology, 38: 355-357. King, J. A. and B. E. Eleftheriou. 1957. Effects of social experience in the laboratory upon adaptation of Peromyscus to a natural environment. Anat. Rec®, 128: 5i f Tab str ac t). 124 King, J* A# and N. 1 . Gurney. 1954. Effect of early social experience on adult aggressive behavior in C57BL/10 mice . Jour. Comp, and Physiol. Psych., 47: 326— 330. Lack, D. 1954a. 18: 25— 38. --___ Cyclic mortality. Jour. Wildl. Mgt., • 1954b. The natural regulation of animal numbers. Oxford Univ. Press, London. 343 pp. Leopold, A. 1939. Game management. Charles Scribner's Sons, New York and London. 423 pp* Louch, D. D. 1956. Adrenocortical activity in relation to the density and dynamics of three confined popu­ lations of Microtus pennsy1vanicus. Ecology, 37: 701-713 * Marler, P. 1956. Studies of fighting Chaffinches. (3) Proximity as a cause of aggression. Brit. Jour. Anim. Behav., 4: 23-30. Mayr, E. 1942. Systematics and the origin of species. Columbia Univ. Press, New York. Milne, A. 1957. Theories of natural control of insect populations. Cold Spring Harbor Symposia Quant. Biol., 22: 253-271. Mohr, C. 0. 1947. Table of equivalent populations of North American small mammals. Amer. Midi. Nat., 37: 223-249. Murie, 0. J. and A. Murie. 1931* Jour. Mamm., 12: 200— 209* Travels of Peromyscus. • 1932. Further notes on travels of Peromyscus. Jour. Mamm., 13: 78—79. Nice, M. M. 1941. The role of territory in bird life. Amer. Midi. Nat., 26: 441—487. Nicholson, A. J. 1933. The balance of animal populations. Jour. Anim. Ecol., 2: 132-178. . 1941. The homes and social habits of%the woodmouse (Peromyscus leucopus noveboracensis) in southern Mi chi gan. Amer. Midi* Nat., 196— 223. 1954a. Compensatory reactions of populations to Stresses and their evolutionary significance. Aust. Jour. Zool., 2: 1—8 . 125 * 1954b. An outline of the dynamics of animal populations* Aust* Jour* Zool., 2: 9— 65* ° 1957. The self-adjustment of populations to change. Cold Spring Harbor Symposia Quant. Biol., 22: 153-173. Nicholson, A. J.'and V. A. Bailey, 1935. The balance of animal populations. Proc. Zool. Soc. Lond.. 3: 551-598. Noble, G-. K. 1939* The role of dominance in the life of birds. Auk, 56: 263-273* Or gain, H. and M. W. Schein. 1953* A preliminary analysis of the physical environment of the Norway rat. Ecology, 34: 467-473* Orr, H. D. 1955. Ranging activity of the northern White­ footed mouse, Peromyscus leu copus noveboracensis Fisher. Ph. D. Thesis' tf. Pi'ttsburgh. "96 numb. leaves. Provost, M. W. 1940. Dynamics of Peromyscus populations. M.S. Thesis. Univ. N. H. 80 numb, leaves. Schneirla, T. C. 1950. The relationship between obser­ vation and experimentation i n .the field study of behavior. Annals New York Academy Sci., 51: 1022-1045. Scott, J. P. 1944. Social behavior, range and territor­ iality in domestic mice. Proc. Ind. Acad. Sci., 53: 188-195. . 1950. The social behavior of dogs and wolves: an illustration of sociobiological systematics. Annals New York Academy Sci., 51: 1009—1021. ______. 1956. The analysis of social organization in animals. Ecology, 37: 213-221. Smith, H. S. 1935. The role of biotic factors in the determination of population densities. Jour. Econ. Entom., 28: 873—898. Southwick, C. H. 1955a. The population dynamics of con­ fined house mice supplied with unlimited food. Ecology, 36: 212-225. 1955b. Regulatory mechanisms of house mouse populations; social behavior affecting litter survival. Ecology, 36: 627— 634. 1956. The logistic theory of population growth past and present attitudes. Turtox News, 34. 126 Stickle, L* P. 1946. The source of animals moving into a depopulated area. Jour. Mamm., 2 1i 301-307. • 1949. -An experiment on Peromyscus homing. Midi. Nat., 41: 659-664. Amer. Strecker, B. L. 1954. Regulatory mechanics in house— mouse populations; the effect of limited food supply on an unconfined population. Ecology, 35: 249— 253. Strecker, R* L. and J. T. Emlen. 1953. Regulatory mech­ anisms in house mouse populations — the effect of limited food supply on a confined population. Ecology, 34: 375-385. Terman, C. R. 1958. A laboratory study of area defense in Peromyscus maniculatus bairdii. Unpublished. Wellington,' W. G. 1957. Individual differences as a factor in population dynamics; the development of a problem. Canadian Jour. Zool., 35: 293—323. Wilson, E-. B. and R. R. Puffer. laws of population growth. and Sci., 68: 285- 382. 1933. least squares and Proc. Amer. Acad. Arts