SOCIAL INTERACTIONS BETWEEN CONFINED JUVENILE AND ADULT PEROMYSCUS MANICULATUS BAIRDI: EFFECTS OF SOCIAL FACTORS 0N JUVENILE ' SETTLEMENT AND GROWTH Dissertation for the Degree of Ph. D. MICHIGAN STATE UNIVERSITY JOHN E ENDERS 1977 I I‘ u: . I‘.’ '1 H ' human I .. . LIBRARY .‘J (31‘ '-‘ This is to certify that the thesis entitled SOCIAL INTERACTIONS BEIWEEN CONFINED JUVENILE AND ADULT PERCMYSCUS MANICUIATUS BAIRDI: EFFECTS OWW SEI'I'IJEI‘IENI‘ AND W presented by John E. Fnders has been accepted towards fulfillment of the requirements for Ph. D. degree in Zoology 4mb‘- , I- Major professor Date August 11, 1977 0-7 639 ABSTRACT SOCIALINI‘ERACTIOI‘BBEIWEENWINEDJUVENIIE AWADULTPERQ’IYSCUSMANICUIA’IUSBAIRDI: m mom SEITLWI'ANDWH-I By John E. Ehders The objective of the present study was to investigate the social factors that may influence the outccmes of juvenile-adult interactions in Perouyscus nanimilatus bairdi using field (3.8 m2) or laboratory (1.2 m2) enclosures. It was hypothesized that‘the sex of juveniles and the sex of adults and/or the reproductive stage of adult fanales would affect the settlenent and growth rates of juveniles mder two conditions of space quality. A juvenile Hale or fanale was placed in an enclosure containing: (1) an adult male, (2) a mnlactating adult female, (3) a lactating female, or (4) no adult as a control. Five replications of each of the eight social treatment carbinations were used in the field and labora- tory experiments . The juveniles were free to interact with the adults over. a four-day period. Juveniles could move through a two-way tunnel to an adjacent unoccupied enclosure, but adults could not, and all animls had access to nest boxes. Juvenile settlanent was evaluated experimentally by determining the proPensity of juveniles: (l) to nest in enclosures occupied by adults, (2) to occupy the same nest with adults, and (3) to retrain, during the nighttime active period, in enclosures occupied by adults. Significant differences annng the social cmbinations were judged to Join E. Ehders indicate relative effects of adults in limiting juvenile settlement. The differemes between the initial and final juvenile weights were examirwd to determine the effects of the experimtal treatments on juvenile growth. The settlenent of juvenile fanales , carpared with juvenile males , was limited by nonlactating fermles . Settlement differences between the juvenile sexes were not significant when juveniles interacted with adult unles or lactating females or encountered the control condition. Lactating farales, coupared with adult males, nonlactating females, and the control condition, limited the settlenait of juvenile males. The settlement of juvenile famles was not significantly different annng the adult classes and the contzo1. Similar outcanes of social inter- actions in free-ranging populations would result in nonlactating and lactating females , but not adult males , limiting the settlanent of dis- persing juveniles depending upon their sex. The contradictions in the results of previous studies on Percuyscus (3g. , Keeley, 1967) and the present research may be due to differaices in the social behavior of different taxa and/or to differences in exper- inental designs since several biological determirmts of social rela- tionships were examined in the present investigation. lactating _I_’_. m. _b__a__ir_<_1_i_ exhibited maternal aggression, but adult males and nonlactating females were not aggressive. It is proposed that the settlenent of juvenile male _13. g. M within populations is inversely related to the proportion of adult famles that are lactating. 0n the other hand, juvenile fanale settlement my be inversely related to the population density of adult females irrespective of their reproductive condition. John E. Ehders Juvenile females, carpared with juvenile males, had significantly lower growth rates when adult males were encountered . Differences in growth rates between male and fennle juveniles that interacted with nonLactating fanales or lactating fenales , or amountered the control condition, were mnsignificant. There were no significant differences in juvenile growth rates when the juvenile sexes were coupared sepa- rately for the adult social factors , but control juvenile Hales gained significantly less weight than those sales that interacted with adult neles. These results my also reflect differences in the social behavior of different Peronyscus taxa. lbwever, social interactions involving behaviors other than aggression, such as sexual behavior, may affect juvenile growth by altering juvenile activity and/or food intake, at least on a short-term basis. The effects of the social factors on juvenile settlement were not similar mder field and laboratory conditions possibly due to differ- ences in the frequencies of behavioral interaction and/or avoidance. The effects of the social factors on growth rates were similar under the two conditions of space quality. But juveniles in the field expel:- inmt gained significantly more weight than the juveniles in the labor— atory experimt. SOCIAL INTERACTIONS BETWEEN OONFINED JUVENILE Aflfl)£fllfifl?PEROMYSCUS‘MANICULNTUS BAIRDI: EFFECTS OF SOCIAL.ENCTORS ON JUVENILE SETUUflfiflflfAmEIGROWTH by John E. Enders A.DISSERDKEKI§ Submitted to MiChigan.State university in.partia1 fUlfillment of the requirements for the degree of DOCTOR.OF PHIDOSOPHY Department of Zoology 1977 WW I wish to express my appreciation to Dr. Rollin H. Baker, Dr. Richard W. Hill, and Dr. John A. King for their advice, moouragement, constructive criticism, and enthusiasm during this study. I also wish to thank Dr. Harold H. Prince for his constructive criticism of the manuscript. The late Dr. Stanley C. Ratner provided valuable advice and encom'agemait during the planning of these experiments. I am grate- ful for the statistical advice of Dr. John L. Gill. Finally, I want to express my gratitude for the competent assistance and encouraganent of my wife, Joyce, throughmrt the entire study. Teaching Assistantships fran the Department of Zoology and the Biological Science Program and Research Assistantships fran The Miseun provided financial support for this project. A Sigra Xi grant is also gratefully acknowledged . The Natural Science Department at Michigan State University provided space for the laboratory aperiment . ii ‘* F. ‘ .- “s g. 7’ I J .53 ’U’ .‘ y c 'Y‘ T I is TABLEOF CONTENTS LISTOFTABLES .......................... LISI‘OFFIGURES .......................... INTRODUCTION ........................... Literature Review ........................ Social Relationships ..................... Determinants of Social Relationships ............. Social Relationships Affecting the Settlement of Juvenile Rodents ........................ Hypotheses Concerning Dispersal in Small Rodents ........ Evolution ............................ Departure From the Natal Site ................. Crossing Unfamiliar Space and Settlement ........... Sumnry ............................. Adult Male-Juvenile ...................... Adult Nonlactating Fannie-Juvenile .............. lactating Female-Juvenile ................... The Purpose of This Study .................... Tunnels ............................ Recording Equiment ...................... Nest Boxes .......................... ’ tal Procedure ..................... Procedure Used in the Field Ecperinent ............. Procedure Used in the Laboratory Experiment ........... t Variables ....................... Juvenile Settlement ...................... Juvenile Weight Change .................... Descriptive Variables ..................... Statistical Procedures ..................... RESULTS .............................. Juvenile Settlement ...................... Juvenile Weight Change .................... Descriptive Measures ..................... iii Page 59 119 127 iv Page DISCUSSION ............................ 135 Juvenile Settlement ...................... 136 Juvenile Weight Change .................... 150 $111me ............................. 155 APPENDIX A. ANALYSIS OF THE DISAPPEARANCE OF JINENILE‘S DURING THE FIELD ECPERIMFNI‘ ................. 158 APPENJH B. JUVENILE ACTIVITY WITHIN AREAS ............ 159 APPENDRC. ANALYSESOFIl-IEDAEYNE‘STDEAREAFRHIUEI‘ICIES. . . . 169 LISTOFREFERENJES ........................ 173 succeed Herschel HIDE 0:5... 3 0.th magnum ONES“. . . II. , .IJ A J. .\ .\J 6 H . I.» A. SI Table l . 2 . 10. 11. LIST OFTABLES Theemerimental designusedinthissmdy ......... 'Ihe outcares possible for the three measures of settle- nnent used in this study ................... Sumnry of the daily and four-day total group mean per- centages of the juveniles that nested in the resident Chi square analysis of the four-day total nesting area frequencies of juveniles for the eight treatment caIbi- nations used in the field experiment ............ Chi square analysis of the four-day total nesting area frequencies of male and fenale juveniles during the field eqaerirnett. The data were pooled for the adult classes ........................... Chi square analysis of the four—day total nesting area frequencies of juveniles during the field experimt . The data were pooled for the juvenile sexes to examine differences annng the adult classes ............. Chi square analysis of the four-day total nesting area frequencies of juveniles for the eight treatment carbi- nations used in the laboratory experinent .......... Chi square analysis of the four-day total nesting area frequencies of juvenile males aid felnles for adult males during the laboratory experiment ........... Chi square analysis of the four-day total nesting area frequencies of juvenile males and fenales for non- lactating fenales during the laboratory experiment ..... Chi square analysis of the four-day total nesting area frequencies of juvenile males and fenales for lac- tating fenales during the laboratory ecperiment ....... Chi square analysis of the four-day total nesting area frequencies of juvenile nnales and fenales for the con— trol with no adult during the laboratory experiment ..... Page 30 54 62 62 63 63 64 64 65 65 Vb .- 4.: “MW My . Table 12 . 13. 14. 15. l6. l7. 18. 19. 20. 21. 22. 23. 24. Chi square analysis of the four—day total nesting area frequencies of juvenile Hales during the laboratory experiment ......................... Chi square analysis of the four-day total nesting area frequencies of juvenile fenales during the laboratory experiment ......................... Snmmry of the daily and four-day total group nean per- centages of the juveniles that nested with adults ...... Chi square analysis of the four-day total nesting arrange- ment frequencies of male and fenale juveniles for adult males during the field experiment .............. Qni square analysis of the four-day total nesting arrange- ment frequencies of nnale and female juveniles for non- lactating females during the field experiment ........ Chi square analysis of the four-day total nesting arrange- nent frequencies of male and fenale juveniles for lac- tating fenales timing the field experiment ......... Chi square analysis of the four-day total nestirng arrange- nent frequencies of juvenile neles for adults during the field espe'iment ...................... 33:33am analysis of the four-day total nesting arrange- ies of juvenile fenales for adults during the t:field experiment .................... Chi. s e analysis of the four day total nesting arrange- ment gun-macs of male and female juveniles for adult males dnnting the laboratory experiment ........... Chi square analysis of the four-day total nesting arrange- ment frequencies of male and fenale juveniles for non- lactating fenales during the laboratory experiment ..... Chi square analysis of the four-day total nesting arrange- nent of juvenile males for adults during the laboratory experiment ......................... Chi square analysis of the four-day total nesting arre-ang tent frequencies of juvenile fenales for adults during the laboratory experiment .................... Chi square analysis of the four-day total nesting arrange- ment frequencies of pooled juvenile males and fenales for adults during the laboratory experinent ........... Page 66 66 71 72 72 73 73 74 74 76 76 78 78 Table 26. 27. 29. 30. 31. 32. 33. 35. 36. 37. Chi square analysis of the proportions of juveniles that retained in the resident area during all four niglnts of the expaimental test period . ............... Summry of the passage tunnel and/or feeder turnnel use by juveniles prior to the placenent of restraining grids during the pretest period .................. metric analysis of the four-night mean pa cent tinne juveniles were in the resident area during the field enpaitnent ......................... The pa cent tinne juveniles spent in the resident area during the active paiod (N -= 5) .............. Analysis of variance of the four-night nean pa cent time juveniles spent in the resident area during the laboratory eqnaiment .................... Mean pa cent time juveniles spent in the resident area for each of the adult classes during the laboratory experimnt (N = 10) ..................... Analysis of variance of the pa cent tine juveniles spent in the resident area on the first night during the laboratory eqnaiment .................. Analysis of variance of the pa cent tine juveniles spentintheresidentareaonthesecondnightmring the laboratory expainent .................. Analysis of variance of the pa cent time juveniles spentintheresidentareaonthethirdnightduring the laboratory expaiment .................. Analysis of variance of the pa cent time juveniles spent inttneresidentareaantlnefonmthniglnten'ing the laboratory expaimnt .................. Sumary of the analyses of the three settlerent variables .......................... of the analyses ~of the significant statistical intaactions between juvenile sexes and adult classes for the three settlenent variables ............. Nonparametric analysis of the four—night mean juvenile activity rate in the passage camel during the field 82 82 83 85 86 92 93 94 95 103 105 C. :4 .3 n...“ .2 .& an .2 L4 be the are s. . .. .(J. I. 4 v {o ;.,~ I .\ Table 38 . 39. 41. 42. 43. 45. 47. 49. 50. viii The nmba of passage tunnel trips per hour of the active period (N = 5) ....................... Analysis of variance of the four-night mean juvenile activity rate in the passage tunnel during the labora- tory experiment ....................... Mean (i l S.E.) juvenile activity rate in the passage tunnel for each of the adult classes during the labora— tory expaiment (N = 10) .................. Student-Neman-Keuls test for differences amng the four-nignt adult class means of juvenile activity rate in the passage tunnel during the laboratory expainnent . . . Analysis of variance of the juvenile activity rate in the passage tunnel on the first night during the laboratory expaiment .................... Analysis of variance of the juvenile activity rate in thepassage tunnelonthesecondnight during the laboratory experiment .................... Analysis of variance of the juvenile activity rate in thepassagetmnelonthethirdnightdnmingthe laboratory enqnaiment .................... Analysis of variance of the juvenile activity rate in thepassage tunnelon the fourthnight during the laboratory ecpaiment .................... Student-Neman-Kenls test for differences amng the first night adult class means of juvenile activity rate inthepassagetnmnelmringthelaboratoryecpaiment. .. Groupmeans (+ 18..E)of theweiglnt changeofjuve- Negative values indicate a mean loss of weight within a group (N = 5) ................... Analysis of variance of juvenile weight change ....... Sumary of the juvenile sex x adult class interaction gromuneans(tlS.E.) ofthewe tchangeofjuve— niles (N = 10). Negative values cate a mean loss ofveightwithinagroup .................. Student—Neman-Keuls test amng adult class means (i l S.E.) for the weight change of juvenile males (N = 10). Negative values indicate a nean loss of weight with a group ............................ 108 109 110 111 114 114 115 115 116 120 121 124 2.2.6 ’9 J.- g. 4-. Table 51 . 52 . 53. 55. 56. 57. 59. 60. 61. 62. 63. Treatment group sunmnary of juvenile wounding (N =5) ..... The total number of chases pa treatment grow. Each cell represents 300 lS-second intavals ........... The total rumba of avoidance and nutual avoidance behaviors per treatment grow . Each cell represents 300 lS-second intavals ................... Treatment grow snmnary of the mnba of juveniles involved in pregnancies (N = 5) ............... Chi square analysis of the proportions of juveniles that tripped the feeda tunnels in the resident area during the field and laboratory expaiments ......... Chi square analysis of the pr0portions of juveniles that tripped the feeda tunnels in the dispersal area during the field and laboratory ecpaiuents ......... The four-night group means of the nunber of feeda tun- nel trips per hour of the active paiod during the field experinent(N=5) ..................... Nonparanetric analysis of the four—night nean juvenile activit;t rate in the resident area during the field expaimnent ......................... Nornparanetric analysis of the four-night nean juvenile activity rate in the dispasal area during the field expaimnt ......................... The back transformed four-night group neans of the nun- baof feedertmnel tripspahouroftheactivepaiod duringthelaboratoryenperiment(N=5) .......... Analysis of variance for tlne four-night mean juvenile activity rate in the resident area during the laboratory expainent ......................... Analysis of variance for the four-night nean juvenile activity rate in the dispersal area during the laboratory expainent ......................... Chi square analyses of the daily nesting area fre- quencies. These canparisons show the lack of joint independence of location of the experinent, juvenile sex, and adult class ....................... Page 130 132 133 134 160 160 161 162 162 163 165 168 169 C :. 1. C C. E . s. 5:... fl 3 2. 3 2 2:1 3 35K Q T5,... Q Ravi an a. a Q 3.. me. a; . . . . 3.. . n. r. .. .. I a. .. K... «a... .mw a 1.. .35. .....J. Table 64 . 65. 66. 67. 68. 69. 70. 71. 72. 73. Chi square analyses of the daily nesting area fre- quencies of juveniles for the eight treatment carbi- nations used in the field expaiment ........... Chi square analyses of the daily nesting area fre- quencies of male ad fenale juveniles during the field experiment. The data were pooled for the adult classes. . . Chi square analyses of the daily nesting area fre- quencies of juveniles during the field expaiuent. The data were pooled for the juvenile sexes to examine dif- faences amng the adult classes ............. Chi square analyses of the daily nesting area fre- quencies of juveniles for the eight treatment conbi- nations used in the laboratory experiment ......... Chi square analyses of the daily nesting area fre- quencies of juvenile males ad fenales for adult males during the laboratory eqneriuent ............. Chi. square analyses of the daily nesting area fre- quencies of juvenile tales and fenales for nonlactating fenales during the laboratory experiment ......... Chi square analyses of the daily nesting area fre- quencies of juvenile males and fenales for lactating fenales during the laboratory experiment ......... Chi square analyses of the daily nesting area fre- quencies of juvenile nnales and fenales for the controls with no adult during the laboratory experiment ...... Chi square analyses of the daily nesting area fre- quencies of juvenile males during the laboratory expainent ........................ Chi square analyses of the daily nesting area fre— quencies of juvenile fenales during the laboratory eqnerinent ........................ Page 169 170 170 170 171 171 171 172 172 172 .v :gIE LISTOFFIGURES The field enclosures (A) and laboratory enclosures (B) usedinthisstudy ..................... Arrangement of the field enclosures anddescription of habitats. The locations of the resident and dispersal areas (1. e. north- south enclosure pairs) were systenatically altaed during the expaiment to remve habitat selection as a factor in detennining tlne unve- nent of juveniles ..................... Arrangenent of the laboratory enclosures . The locations oftheresident (R) addispasal (D) areaswaethesane for all tests ....................... Thetnmelapparatususedinthisstudy (A=operating am, B-microswitch, C-tunnelcova, DandD' = restraining grids, ad E - feeding station) ........ Dmle of two ecpaimental enclosure pairs slewing the t of passage tunnels (X), feeding stations (F), arrangemen and nest boxes (0) ..................... Four-day total nesting area frequencies of juveniles (N = 20 juvenile days). The designations for adults (6" 8 male, 5’ - nonlactating female, IAC 3 =- lactating fenale, andCIJN-controlwithno adult) willbeusedforotha figuresinthispapa ................... Four-day total nesting arrangennnt frequencies. Only days on which juveniles nested in the resident aeas are The four-night grow means for the per cent time juve- nile tales spent in the resident area during the laggra- tory experiment. The mean marked the astaisk is significantl diffaent firm the 0 means at the .01 levelCN-S .............. . ......... Graph slnowing the significant diffaence (P = .005) in the four-night mean pa cent tine juvenile males ad fenales spent in the resident area occwied by nonlac- tatingfemlesinthelaboratoryeqnaimentm-S) . . .. xi Page 34 37 39 42 44 69 90 18. Figure 10. 11. 13. 14. 15. 16. 17. 18. Nightly neans of the pa cent tine juvenile males spent intheresidentareadnn'ingthelaboratoryeqnaiment . The astaisk (*) indicates a significantly diffaent neanatthe .01 level (N-S) ............... Graph slaving the significant diffaences (.025 > P) in the nightly canparisons of the nean pa cent tine juve- nile neles ad fenales spent in the resident areas occwied by nanlactating fenales during the laboratory expaimnt (N-Sforeachnight) ............. Nightly canparisons of the uean pa cent tine juvenile males and fenales spent in the resident areas occupied bCN lact):ating fenales during the laboratory experiment - 5 .......................... Four-night adult grow neans for the juvenile activity rateinthepassagetnmeldaingthelaboratory t. Thevertical lines indicate t lS.E. (N-lO) .......................... First night adult grow means of the juvenile activity rateinthepassagetmnelchrringthelaboratory t. The vertical lines indicate 1 l S.E. erpainen (N- 10) .......................... Graph of the significant diffaence between the unean weight changes for juveniles used in the field expai- nent and the laboratory expainent. The vertical linesirdicatetlSE. (N840) .............. Graph of the s icant diffaence between the mean weight changes rjuvenile males ad juvenile fenales tested “EN adnfigr males. The vertical lines indicate 2*: l S.E. = ..................... of the nean weight changes for juvenile males. The vertical lines indicate 3: l S.E. (N - 10) ....... Graph showing the significant diffaence (.01 > P > .005) in the four-night mean juvenile unale ad female activity rate in the resident area occupied by non- lactating fenales in the laboratory iment . The verticallirnesirdicatetlS.E (N-S .......... Page 98 100 102 113 118 123 126 129 nine: 1"! £9 I. AID “asses . I a. ‘L' tlga IN'I'ROIIJCI'ION Dispasal, the mvenent of an organism fran its natal site to another locality where breeding nay occur (I-bward, 1960), is being given increasing attention by ecologists studying small rodent popula- tions . Sadleir (1965) ad l-Iealey (1967) indicated that Pagnyscus naniculatus austaus inmigrants were prevented fran settling by aggres- sive adult males. Honeva, in tlnese stndies, adult fenales were largelyignored, adtleiwortanceoftlesecoftlejuvedlesin detamining the antenna of social intaactions was not evaluated. Metzgar (1971) reported that the populatien density of adult fenale P. m was invasely related to the nunba of juvenile fenales that settled within a local population, but males did not exhibit this relationship . The importance of different reproductive stages of the adult fenales was not evaluated by Metzgar, but diffaences in behavior towards strange juveniles nay be expected because lactating Pacuygcus camonly exhibit aggression or nest defense whereas nonlactating females typically do not (Eisenberg, 1963) . Dispasal innorthernPeranyscus populations occurs during tie breeding season (see Stickel, 1968) . Dispasing juveniles my encounter ad intaact with any of three adult ser/reproductive-stage classes: tales, nonlactating fenales, ad lactating fenales. This investigation was designed to obtain quantitative measnmes of the out- cane of these juvenile-achilt intaactions in 1:. n_n_. M, the prairie l dea miss. The objectives of this study were to determnine if: (1) the sex of juveniles affects the outcane of social intaactions, (2) the sex of adults ad/or reproductive stage of adult fenales affects the outcane of social interactions, ad (3) tlne outcane of social intaactions are similar unda field enclosure conditions ad controlled laboratory conditions. The impetnis for this research steamed from the fact that intra- specific social intaaction has been hypothesized to lead to the regu- lation of small mamml populations (_e_.g. , Chitty, 1967; Healey, 1967; ad mtzgar, 1971) . The basic argunent in the proposed population regulation models is as follows. When an individual dispases, the localpopulationdensityisdeaeasedadthedensityoftrepopulation intowhichtheirdividual iumigrates is increased. Sothen, population densities can be dependent won the behavioral intaactions influencing the rates of departure fran the natal site ad settlenent in a new site. Systematic study of the social behavior of anall mnamnals is required in ordatoudastadtheroleofbehavioralmechanians indispasalad consequent population regnlaticn. Savidge (1970) operationally partitianed dispersal into three pro- cesses: (l) departure from the natal site, (2) crossing a barrier, usually unfaniliar space, ad (3) settling in a new area. Each process can be investigated separately by controlling irdepedent variables thatmnaybeinportantindetaminingtleoutcareof socialbehavior. Savidge (1974) reported that juvenile P. g. ba__i_.n_:_c_l_i_ crossed a bar- ria (an electrified grid connecting a natal cage with another cage) whenthe fatha crossedor ifthenotherwas aggressivewonthebirth of a subsequent litta. Nicholsm (1941) suggested that _P. M neles also attracted littas, and juveniles of other rodent species Tevebeenreportedtodepartfrcmthenatal areainresponsetoaggres- sive mthas (Bradt, 1938, for Castor canadensis ad Errington, 1939, for gnd_at_r_a_ zibethica). Dispasing juvenile rodents, that have departed the natal site ad are crossing unfamiliar space, encounta strange resident (Lg. , located inaharerange) achnlts, interact, adnnayremainnear theadultorcon— tinue to dispase. Social intaaction between resident adults and dis- persing juveniles has been reported to affect the settlenent process in sciurids (9g. , Drabek, 1973, for Spannphilus richardsonii ad Carl 1971, for S. undulatus), P. g. anstaus (Sadleir, 1965, ad Healey, 1967) , ad _P_. 13m (reagent, 1971) . Detailed accounts of the social variables that may affect the antennas of intaactions in ground sciurids are available because this group is diurnal ad easily obsaved in the field (3.5., am, 1964, ad Betta, 1976). m, studies on the effects of encomtas between strange adult and dispasirng juve- nile Paomyscus are rare. Such studies cannot be based on direct obsavation due to the difficulty in obsavirg dea mice in the field. Rather, they should focus on the evaluation of the social factors that may influence spatial relationships while the mice are confined. When social conditions are controlled, the resultant spatial relationships can be inferred to be detamnined by social factors. Social conditions were controlled in the present study by 1min- taining prairie dea mnice in enclosures (l) unda field conditions ad (2) under controlled laboratory conditions where space quantity ad canplexity were decreased relative to the field conditions. The effects of social intaaction won juvenile settlement were evaluated ”H C a m t" «53' emerinentally by neasuring the spatial relationships of juveniles and adults. Healey (1967) reported that aggressive nele g. g. austerus retarded the growth rate of juveniles in field ad laboratory experi- ments. The growth rates of all juveniles in the present investigation weremeasured inanattenpt to reecanineandexpanduponthese find- ings. Also, several other measures (_e_.g. , activity ad wounding) were taken. literature Review Dispersal is a poorly understood denographic factor in suall rodent populations. Pinch of the lack of understanding can be attri- buted to the difficulty in studying dispersal in free-ranging pepula- tions (Myers ad Krebs, 1971). Also, dispersal is a dynamic process; a variety of biotic ad abiotic factors may act alone or in caninination to determine an observed dispersal rate or pattern (Lidicker, 1975). This literature review is restricted, for the most part , to pertinent informtion of the intraspecific social factors governing the dispersal process of settlenent. The reviev is organized in tlne following fash- ion. Firstly, a general definition of social relationships ad deter- minants of social relationships will be given, and secondly, the social relationnships that may affect the settlement of juvenile rodents will be presented. Ihirdly, hypotheses conncernixg dispersal in suall nam- nals will be discussed, but only those dealing with the settlerent ad growth rates of dispersing juveniles will be analyzed in detail. Fourthly, a general presentation of evolutionary considerations will be given followed by, five, a general summary of the predictions of the onztconnes of social interactions between strange juveniles ad adults. Social Relationships Wnentwoindivimnalscmnetogetheradreacttooneanother, their behavior is no longer indepedent, ad a social relationship may be £013an (Scott, 1973) . Scott developed the concept of social relation- ship to pramte the investigation of social organization in group-living species suchasprimtes. The saneconceptual franeaorknaybeusedto stuiy social encounters between strange juvenile ad adult Pmscm eventtnnghthisgennsusuallydoesnctoccm‘ingmmsotherthana mncganouspairadtheiroffspring, at leastmringthebreedingsea- son (Tennan, 1968). The outccne of a social encounter, in terms of settlenentandsubsequentuseof spacebythejuvenile, mybedeter- mined by the factors that deternnine the formation of a social relation- ship. Deternflnants of Social Relationships According to Scott (1973), two cmpenents are necessary for the differentiation of behavior. into a social relationship. Firstly, ani- mls nust have ability to discriminate between individuals, ad secondly, behavioral patterns nustbepresent that canbeuodifiedby experiential ad biological factors . Individual recognitian in Peromyscus is probably acccuplished by usingavarietyofcues, adthisgenushaswelldevelopedreceptor organs for auditory, tactile, visual, ad olfactory stiuuli (King, 1968a). Pair bonding (Howard, 1949) ad stable daninance hierarchies over time (Sadleir, 1970) provide evidence for the existence of indiv- idual recognition in Peranyscus. thneriential factors my be subdivided into two categories: (1) early experience and (2) experience with an area of faniliarity with another animl. The mount ad quality of infantile stimflation lnas been sham to affect behavior at a later developmental stage in rodents amine gt _a_l_., 1967), but the processes and mechanisms associated with the effects of early experience are not known (Russell, 1971) . Terman (1963) reported that 1:. m. M, raised in isolation, were less inclined tobe social thanweremice raised insocial groups. Anunnder- stading of the preoeanling biotic ad abiotic environments is required prior to evaluating tlne significance of early experience in the subse- quent social behavior of Peranyscus. However, the available evidence indicates that early experience nust be controlled in experimental stud- ies of social behavior. 'nne outcone of social encounters in Peromyscus, ad other rodents, candepednpanthelocationoftheenconmterandthedegreeof fad- liarity between animls. An animal in familiar surroundings usually becams dadnant over an intruder, if a dcminance-subordinant relation- ship ensues (Eisenberg, 1962, ad Grant, 1972), admice ted to ednibit less aggressien towards familiar connspecifics than towm'ds strange conspecifics (Poole and l‘brgan, 1975). These conncepts are integral to the settlemnt process of juveniles for several reasons. The activity of adult Peggyscus is usually restricted to a localized areaorhmnerange; ithssbeenclaiuedthatacmltsarefaniliarwith the area encanpassing tlneir home range (Furrer, 1973). Secondly, the encounter between a dispersing juvenile ad strange resident adult typically takes place in the adult's hane range. Finally, agonistic behavior or overt aggression (fighting) involved in the establisl'nent of the daninance-subordinant relationship may force the juvenile to leave the presence of the adult and continue to disperse. No points trust be considered prior to accepting the apparent role of daninance- subordinant relationships, as determined by agonnistic behavior, in influencing juvenile settlenent (Healey, 1967): (1) there is no evi- dence of territoriality or defense of mitually exclusive areas in Perayscus (Hill, 1970) ad (2) behaviors other than those that are agonistic my be involved in the establishment of a social relationship. Thelatterpointwillbediscussedindetail inthesectionnonthetypes of social relationships that may influence juvenile settlenent. Agonis- ticbehavior, asusedinthispaper, refers toanybelnaviorwhichis cennnected with a centest or conflict between two ccnspecifics whether fighting, escaping, or "freeing" (Scott, 1972). Biological factors my be subdivided into four categories: (1) genetic constitution, (2) age, (3) sex, and (4) ranoductive condi- tion/physiological state. A genetic basis for behavioral differences in Permyscus, beyond species and subspecies couparisons of habitat selectionn ad social behavior (Harris, 1952, ad see Eisenberg, 1968), has not been reported. Male (Healey, 1967) ad lactating fenale (Savidge, 1974) _11. nenicnflatus have been fonrd to vary mrkedly in aggressive tedencies. These reported differences may have a genetic basis, but selection eqneriments are required in order to determine heritability. Radanization of experimental animls can be used to avoid biasing results in studies on social behavior in which possible genetic differences are of no interest. Agecanbeaninportantfactor influencing socialbehavior. In general, adult animals are usually daninant over inmatures. Grant (1970) stated that body size, which is directly related to age in most rodents including 2. mniculatus (Dice, 1932 ad 1936) , is a reliable correlate of dandnance in agonnistic encounters. Field studies have indicated that maturing rodents undergo gradual changes in responsive- nesstoconspecificsadviceversa, whichinturn, cansechanges in social relationships (_e.g. , Barash, 1973, for Marmta olyugus and Betts, 1976, for §_. colunbianus) . These changes, while certainly correlated with growtln and increased body size, are also correlated with physio- logical changes associated with tlne onset of sexual natunrity (Poole ad Pbrgan, 1975). There is a nneed for detailed study of the developnent ofsocialbelnaviorinPermysensasmsdeneforbflcrofinngestis (Wilson, 1973). Until this information is available, age classes slnculd be stadardized in studies examining other determinants of social behavior. This procedure would result in a reduction of confounding variation due to changes in age-specific respansiveness of the experi- mental animls. The senofnemnals isofparamunt inportance inthedetermination of their social behavior ad social relationnships . Adult uale Perenyscushavebeenclainedtobemreagonnisticthanadnlt fenales, especially toward other mles, although both sexes are capable of the sauna aggressive behavioral patterns (Eisenberg, 1968) . Ibwever, only a few species have been critically examined. It is interesting to note that bothmale and fenale P. polionnctus, _P_. 1%, ad _P_. n_n_. ba_i£__di 28-day olds were similarly agonistic towards Microtus weanlings (Baennninger, 1973) . Therefore, differences in levels of agonistic behaviormyappearas dee'ndcemun'e toadults. 'nne social relationships of resident adult and dispersing juvenile Peromyscus fenales have been critically examined in only onne study (lbtzgar, 1971, for 1:. 1m), ad repulsion of juvenile fenales was reported. Fenales have been largely ignored in other investigations of mallumnnl dispersal inwhichbehavioralmechanismswere examined (Krebs, 1970, and Myers and Krebs, 1971) . A systenatic study of the effects of social factors on the settlement patterns of dispersing juve— niles should evaluate differences due to the sexes of the interacting juvenile-adult canbinations . Changes in reproductive condition ad conncomitant physiological state of manuals give rise to changes in social relationships. Male Peranyscus renain in breeding condition throughout the breeding season adhavebeenreportedtobeaggressivechring this tine (Sadleir, 1965, ad Healey, 1967, for P. g. austerus), probably due to high tes- tosterone levels (Bronson ad Desjardins, 1971). (h the other had, femle Perenyscus exhibit behavioral chages within a single breeding season. Prairiedeermicecaneintoestrusabouteverythreedaysupon emsure tonnales (Bronson andMarsden, 1964), and the behavior of fenales may shift in response to chages of the estrus cycle. For enanple, Payne ad Swanson (1970) fonrd that feels Mesocricetus aruatus were less agonistic towards male conspecifics during peak estrus com- pared to other stages in the cycle. Maternal aggression can be elicited bythepreseneoftheoffspringadsouethreateningagent (Mayer, 1971). King (1958) ad Scudder _e_§ 51. (1967) reported that lactating P. g. m exhibit nest defense, and fenale Mericnes unguienlatus were found to greatly increase aggressiveness following parturition CDniessen, 1973). Other rodent fenales, in the lactation stage, have 10 been inferred to be agressive towards conspecifics (Blair, 1942, for m striams and Broadbooks, 1970, for Eutamias amenus). The effects of mnlactaticn ad lactation stages on juvenile-achlt social relatimships between strage aninnls have not been critically mined in Permyscus , but differences in juvenile settlenent may be related to these different adult reproductive stages . Social Relationships Affecting the Settlemnt of Jivenile Rodents The spatial organization of individuals in Peronnyscus populations has been described in capture-recapture ad nest box studies. In general, the novenent of resident and probably breeding adults is local- izedwithinalrrnerangewhere theanimlnorually travels inpursuit of its routine activities . Hana ranges are usually maintained for the life of the individnal, but sane adults leave established here rages. The neasnn'edspatialarragenentofhmnerangesisvariableandseensto dependuponthetanonmlcgroupsbeing studied, geographical locationn, ad possibly the methods used in the measurenent. During the nonbreed- ing season, the exclusiveness of bane ranges decreases (see Stickel, 1968, ad Ternnan, 1968). The mechanisms leading to the emchnsiveness of lune ranges, the formtion of social relationships, and the incor- poratian or settlenent of juveniles into the mtrix of here rages is unclear. I-bwever, Hill (1970) fonrd that neither mle nor fennle 1:. 3. Mad; Le_nc_opu_s_defededafamiliar areafransanneoropposite- sen connspecifics. There are nanny types of social relationships that can be famed depeding uponn the experiential ad biological characteristics of the animls involved in interactions (Scott , 1973) . No relationnships hints 11 between strange resident adults and dispersing juveniles are central to the settlenent process . Firstly, daninannce-subordinate relationships my be established. I-bwever, there are serious methodological diffi- culties with the cbminnance concept (see Richards, 1974, for review). Several definnitions of dominance have been proposed, but direct obser- vatian of interacting animls in order to obtain the required measure- ments is iupossible for Peranyscus in the field. That daninnance in uurid rodents is established by agonnistic beha- vior is well Imam (g.g., Poole ad Fbrgan, 1973, and Bovet, 1972), ad dominance hierarchies resulting from agonnistic encounters have been reported for Perennyscus (Sadleir, 1970) . The works of Sadleir (1965) and Healey (1967) are often cited to indicate that aggressive inter- action is virtually universal in altering the settlenent pattern of juveniles (3.3. , Barash, 1977). It my be sutmitted that agonnistic innteracticn and subsequent escane of the juvenile my be sufficient for influencing the settlenent of juveniles, but it is not necessary. Orr (1959) and Termn (1962) reported that Percgyscus simly avoid conspeci- fics upon encounters, though it was nnot known if elements of agonnistic behavior were involved in the avoidance. Also, it has been fonrd that subordinate Permscus frequently nest with daninants (Hill, 1970) . IBnce, subordirnate individuals in free-raging pOpuLations my not necessarily leave the presence of daninnant individuals. Sexual relationships are the second type of relationship that my be involved in the settlenent process. In the absence of agonnistic interaction ad/ or avoidance, opposite sex individuals my be attracted . Tl'et juvenile Per'anyscus usually disperse at about tlne time they attain semal mturity aboard, 1949, for P. g. ba__irdi__, ad see Blair, 1953, 12 and Stickel, 1968, for other species) suggests that sexual behavior my beinportantindeterminingtheoutcaneofsocialenconmtersasthey relate to settlenent. The social connditions influencing social relatienships ad cense- quent settlenent patterns in dispersing deer mice are poorly understood. The lack of ederstardinng can be attributed to the failnnte of past workers to censider the factors that determine the formtien of social relatienships . Sex and reproductive condition are important factors in determining social relatienships ad can be manipulated in a systematic study on the social factors innfluencing settlenent in juvenile deer mice. Other factors can be stadardized or randanized. Itpotheses Cencerning Dispersal in Smll Rodents An array of proposals cencerning the causes ad effects of social behaviorinthedispersalphencuenonadsubsequentpopnflationad individual consequences have been published. Howard (1949) proposed that dispersal ability enables a species to reinnvade depopulated areas or newly created favorable habitats . Further , a genetically determined tedency to disperse was suggested by lbward (1960) . That is, sane vertebrates are claimd to have an innherited propensity to disperse ad avoid readily available and suitable habitat to enter novel, ad prob- ably unfavorable, areas where their chances of survival are potentially low. Presmably, these individuals disperse irrespective of the social ceditions in the natal and settlenent sites . Conversely, other indiv- iduals are claimed to disperse due to social ad physical parameters such as mate selection, territoriality, parental ejection, or lack of suitable hannesites . According to I~Ioward, the innate dispensers are 13 often sacrificed for the "good" of the species because they facilitate geneflow, enlargednerangeofthespecies, adpermitthespeciesto mintain a discentinnuous distribution and reinvade depopulated areas. This proposal necessitates the acceptance of group selection. I-bwever, tlne role of group selectien cannnot be supported by current evolutienany theory (William, 1966) . C'hitty (1967) hypothesized that selection for aggressive dominant individuals increases as population density becomes higher, and that dispersal of genetically subordinnant individuals is greatest at high population densities. Myers ad Krebs (1971) found genetic differences between dispersed ad resident Microtus, but these authors could not determine whether the different genes caused different levels of aggressien. Further, no mrked differences in aggressive behavior between dispersing and resident 'mle Microtus were found, though it was suggested that dispersing g. pennsylvaticus mles were none aggressive than resident mles (Myers and Krebs, 1971). Other linnes of centra- dictory evidence my be used to argue against acceptance of Chitty's idea concerning the papulation density-dispersal ad population density- aggression relatienships. Myers ad Krebs (1971) ad Rose ad Gaines (1976) found that dispersal in Microtus papulations was not greatest during periods of highest populatien densities. However, Joule ad Caneron (1975) reported that Sim ad Reithrodontenys dispersal was positively correlated with populatien densities , though the causative nechanism were nnot evaluated inn tlneir study. Lidicker (1973, for n. californicus) ad Rose ad Gaines (1976, for lg. ochrogaster) reported that wmnnding, ad therefore aggressien, was not greatest during peak population densities, but seemed to vary with seasenal changes. In .8 N r:- 0; [_nu 14 cenflict with these reports, Krebs (1970) observed high levels of agressien in laboratory encounnters between wild-caught male 111, ochrogaster andlfi. pennsylvanicus captured during theperiodof peak population density, and Conley (1976) described similar findings for mle ad femle 1‘_i_. legicandus. Also, Christian (1971) nnoted that warding in mle n__n_. pennsylvanicus was positively correlated to popu- lationn density. Clearly, the relatienship between agressien and popu- lation density has not been firmly established ad requires fuunre surly. Theaboveinvestigatiensweredesignedtoenaminethecansead effect of dispersal at the population level. The primary purpose of these studies ms to describe attributes of dispersing individuals, establish the relatienships between populatien density-dispersal-aggres- sionn correlates, or to explore alternative relationships between dis- persal ad enviremntal canpenents. Anotter line of research has been directed towards specific questiens cencerning the social relatienships between resident adults and dispersing juveniles in reference to the process of juvenile settlenent within local populations. These investi- gatienshaveadirectbearingenthepresent studyandwillbeanalyzed in detail. Sadleir (1965) proposed that resident adult 2. m. ansterus mles were aggressive during the breeding season ad innhibited the survival of juvenile conspecifics. Healey (1967) cenfirned the claim that mle 1:. m. austerus in breeding cediticn were aggressive. This antler also reportedtinat thegrowthrates ofjuveniles wereretardeddue to social interaction with resident adult mles in confined laboratory enclosures . In addition, the growth rates ad survival (i._e_., retaining within an I .0. '0' n n OOIQO an. .‘U kg‘é 15 area) of released juveniles were lower in field plots containing aggres- sive adult mles cenpared to those juveniles released onto plots con- taining docile mles. Healey proposed that local Perenyscus papulations were regulated by aggressive adult mles forcing surplus juveniles to settle elsewhere. Flowerdea (1974) renoved adult mle Apodems ylvaticus from a sandy plot ad found that the survival of subadults and juveniles was significantly better cenpared to a centrol plot on which achnlts were allowed to rennin. Further, laboratory ennperinents indicated that sene aggressive adult mles decreased juvenile growth rates ad killed juve- nniles, thoughthesupportive evidencewasweak. Thisauthor concluded that papulation control wwwas sinnilar to that claimd for W by-Healey (1967). Metzgar (1971) hypothesized that 1:. w populations are regu- lated by intraspecific behavioral interactions . Juveniles were released onto a study plot centainning resident adults ad settled in areas unoccupied by adults of the sane sex. When resident adult fenale den- sity increasa, fenale inmigrants were less likely to settle, but mles did not exhibit this pattern. Metzgar (1971) sugested that social interaction between fenales results in nore repulsien of juveniles cen- pared to interaction between mles. That is, resident fenales limit the settlenent of transient fenales , but resident mles do not limit the settlenent of transient mles . The cenflicting cenclusiens in de above sudies can be attributed to two factors. First, the social organizatiens of the different taxa my be different, althougln there is no experimental verification. May (1967) suggested that 1:. I_n_. austerus are organized into units of l6 familiar animals which repulse the settlement of strange deer mice. Brown (1969) suggested that A. sylvaticus occur in groups in which a single dominant mle is surmuded by a few females and several immatures. Dendnant mles have been found to control the activity of group members in Am (Bovet, 1972) ad other mnrids (Crowcroft and Rowe, 1962, for Q and Calhoun, 1963, for Ra_t_tn_i_s), but aggression within groups is low. Strange conSpecifics are repulsed fren organized colenies of 39% (Calhoun, 1963) . Myten (1974) recently reported that _P_. m are arranged innto "family groups" containing a single adult fenale, several mles, ad a number of juveniles. However, the relatedness of these wood mnice was not kmwn. Secedly, theconclusiensmyhavebeeninfluencedbytheenperi- mental designs. The field experiments suffered frcm a lack of control of test animls. Adults and juveniles disappeared for unknown reasons, effects of social interaction between ad among resident adults ad immigrating juveniles could nnot be eliminnated, ad Healey (1967) meas- uredgrowthrates enlyenthosejuveniles thatremainedonthestudy site ad were recaptured. Also, some docile mles becane aggressive and same aggressive mles became docile during Healey's (1967) study, but sinnce results from these animals were nnot inncluded in his analysis, the cenclusiens were not influenced. The functimal significance of the effects of adult mle P. n_m_. austerus en juvenile growth is not clear primrily because the cause- effect relationship between aggressien ad delayed mturatien in rodents is not understood. The physiological mechanism respensible for the inhibition of juvenile yowth ad possible retardation of semal manna- tien is often postulated to be the pitnnitany-adrenocortical-gonadal triad 17 which responds to stress stemming fren social conditiens (King, 1973). However, the supportive evidence for the role of this physiological mechanism has been obtained primrily frcm laboratory experiments where animls were exposed to supemormal social stresses such as extremely high cage population densities, long periods of exposure to social treatments, ad high levels of aggressien canpounded by severe confine- nent (see Christian 93 31., 1965) . A seced ad more general enqnlana- tion my be that short-term adult-juvenile interactions affect juvenile growth by altering the activity, energy expenditure ad/or food intake of juveniles . These alterations could be mediated directly by social interactions, involving aggression or sane other behavior, if adults restrict juvenile movenent ad the amount of food juveniles contact ad eat ad/or somehow cause differences in activity ad therefore energy expedimre of tie juveniles. Data frem laboratory studies do nnot provide conclusive support of either the existence of inhibition or the ’specific social correlates of inhibition that can be extrapolated to free-raging populations . Termn and Gardner (1970) investigated the effect of cage population density on the growth of _P. n_m_. ba__i_r___di. Only a few significant body weight differences were noted in prairie deer mice that were 40, 60, 80, or 100 days of age. These differences along mles or fenales were either due to age differences or could nnct be explained in relatien to any independent variable in the experimental design. In a recent study, Tomas ad Terman (1975) reported that mle, but nnot fenale, 1:. g. hai__rd_i_had significantly higher body weights at 100 to 140 days of 888 whim reared with bisexual pairs than mles reared in populationns of four bisexual pairs ad offspring. Whether cemunnal nursing, social 18 interaction, or some other population experience cansed the difference in growth was not known. The laboratory portions of the studies by Ihaley (1967) and Flowerdew (1974) can be severely criticized because the juveniles could not escape the presence of aggressive adults, as is the case for free- rangingpopilationsifopenspaceisavailable. 'Ihereisaneedto investigate the effects of adult- juvenile behavioral interaction on juvenile growth rates in conditions where the juvenile can escape the presence of the adult. Adult fenales were largely ignored in the studies of Sadleir (1965) , Healey (1967) , and Flowerdew (1974) . lbwever, pilot studies conducted by Healey (cited in Metzgar, 1971, and Fordham, 1974) indi— cated that aggressive female _l_’_. g. austerus determined the umber of immigrating fenales that settled in an area. Metzgar's (1971) work on g. M supports this tentative finding for fenales, but general conclusions concerning fenales must be tenpered until an evaluation of the role of reproductive stage (Lg. , nonlactation vi. lactation stages) in determining the outcane of social encounters is conducted. Finally, the sexes of juveniles in the research conducted by Sadleir (1965) , Healey (1967) , ad Flowerdew (1974) were not reported, but mtzgar (1974) found differences in the settlenent patterns of juve- nile mles ad fenales. The juveniles used by Healey were sexually inmate, but Phtzgar uSed 40 to 60 day-old young. Clark (1938) reported that g. M go into their first estrus at about 46 days of age. Therefore, it is possible that Metzgar's deer mice were sexually nature. In light of tlne role of set in determining social relation- Ships, ready acceptance of generalizations onthe importance of resident £5.15 if. at? ,e e' 19 adults in affecting dispersing juveniles is prenature until a quanti- tative evaluation of the effects of the sex of juveniles has been made. Evolution Dispersal, at tines, results in animals leaving an optinun area andentering a different area in time ad space inwhich the probability of mrtality is high (9g. , Carl, 1971). Sane autl'nrs have indicated that dispersing individuals are sacrificed for tlne "good" of the species sinnce ranges my be expanded ad/or population regulation my be effected (Howad, 1960, and Christian, 1970) . Van Valen (1971) veri- fied thatgroup selectioncould leadtodispersalbyusingacanputer sinulation with the assnnptions of haploid organism ad asexual repro- duction. I-bwever, there is no aspect of malian dispersal that cannot be explained on tlne basis of individual selection. Natural selection is the nonrandom differential perpetuation of varying gene frequencies (Mettler ad Gregg, 1969) . That is, genes that enhance adaptiveness by increasinng the probability of reproductive success of the individual are selected for over time so that their fre- quencies inncrease. Lidicker (1962 ad 1975) proposed that the prob- ability of reproductive success of individuals that disperse frun a population before the habitat becanes saturated with conspecifics could beincreasedduetofonn'advantages: (l) dispersershaveaninereased opportmity for mating because they contact are potential mtes , (2) with outbreeding there is an increased chance for prodncmg favorable genetic recanbinants, (3) disperses my avoid population crashes ad Predator baildups, ad (4) during periods of high population density renaming resources becane scattered ad resource utilization becomes Lexie ages an trim a 325565 0 $59.3 fiirin @st 20 inefficient; dispersers avoid this inefficiency. These proposed advan- tages are logical ad deserve experimental evaluation. Ibwever, the topic of the present research is the social factors affecting dispersal withinapopulation. 'Ihebasicpreniselhavetakenis that thepro— cesses of dispersal (departure fren the natal site, crossinng unfamiliar space, ad settlenent) occur when individuals are likely to increase their innclusive fitness. This increase can be achieved by increasing the probability of (l) a disperser's reproductive success ad/or (2) the reproductive success of animls related to a dispersing individual (Hamilton, 1964, and Mnrray, 1967) . The following presentation of evolutionary considerations will be divided into two sections: (1) departure fran the natal site and (2) crossing unfamiliar space and settlenent . Departure From the Natal Site that 1:. E!- M (Howard, 1949, ad Dice and Howard, 1951) and other rodents (_e_.g. , Drabek, 1970, for §_. tereticaudus, Rengstad, 1965, for _S_. tridecenlinneatus, and Layne, 1954, for Tannias striatus) disperse as juveniles. However, fall-born 3. g. _b_a_i_rg._i_ litters rennin with their parents until the following spring (Howard, 1949). Sane sciurids do not disperse until they are two or three years old (Barash, 1974). On the other hand, King (1955) reported that Cyrmgs ludovicianus juve- nilesdonotdisperse, butrathertheparentsdo. Further, 38percent of the juvenile fenales and 28 per cent of the juvenile male 1:. g. M stLdied by Howard (1949) did not disperse (Dice and Ibward, 1951). Similar contradictory reports can be foud for other mammlian Orders (_e_.g., Ever, 1973, for carnivores). Diffs minim min 21 Differences between taxa can be partially attributed to different evolutionary histories. Barash (1974) used the couparative approach to ascertain behavioral-environmental correlates of dispersal in three species of Marmta. Marmta unnax live at low elevations ad disperse during their first year; 11. flaviventris occur at median elevations and disperse as one-year olds; 11. 1123115 are found at high elevations and disperse as two-year olds} The food supplies of the local environment vary inversely with elevation, but the age to sexual maturity vary dir- ectly with elevation. Barash suggested that dispersal was delayed at higherelevations inorder toallowtheyoungtoattainadultbody size prior to dispersing. The social determinant of dispersal seened to be aggressive anthers, but fenales exhibited little aggression towards their unturing young until tlney attained adult body size. According to Barash, these anther-offspring relationships would increase the prob- ability of the reprodnctive success of dispersing juveniles, which in turn, would inncrease the inclusive fitness of the parental adults. mile the angestions above are speculative ad require Innch mre sup- portive data before they can be accepted, the systenatic coupaative approach has great potential in identifying salient features of the env'irement that my exert selective pressures on the dispersal pro- cesses. It is apparent that the evolutionary histories of taxa mist be censidered prior to maldng generalizations across groups. Agonistic interaction between the fennle parent and weaned off - springuponthebirthofasubsequent litterhasbeenreportedor sug- gestedtocauseimnaturestodepartthenatal siteing. g. M (Savidge, 1974), g. canadensis (Bradt, 1938), Q. zibethica (Erringten, 1939), and sciurids (e.g., Michener ad Michener, 1973, ad Snith, n n if 1:285 5%) t F. can t} 22 1968) . The situation where a fenale actively expels his offspring frcm a suitable living space seens paradoxical. According to Trivers (1974) , parental-offspring conflict is to be eqnected in sexually reproducing species. Parents are assured to etpedenergyadincu'physicalrisksoastomaxindzethemnberof offspring that survive ad reproduce. However, individual offspring can be thaught of as trying to maximize their reproductive success by attempting to secure more investment than the parent is selected to give. It mist be uderstood that the parent ad offspring are related by one-half, as are siblings. Any increase in the parental investment for one genetically (Lg. , a genetically determined behavior that eli- cits mre mternal care than is necessary) favored offspring will ted to decrease the nmber of siblings that survive, ad consequently reduce the number of like genes (in reference to the favored offspring) located in the siblings. Then, the gene in the favored offspring will be selected against, if it extracts too great a cost fran the parent, even if it confers sane benefit on tlne favored offspring. Benefit and cost are definned by potential reproductive success or fitness (Hamilton, 1964). In order for an act (behavior governed by a gene) to be selected for, the benefit/ cost ratio times the probability that the recipient of a favorable act has the gene mist be greater than one. Specifically, for an act to have survival value, its benefit/cost ratio mist be larger than the inverse of the actors degree of related- ness to the recipient. Asyoungnmmlsgrowtheybeccuemreecpensiveforthemtherin terms oftheenergetic cost ofmilkproductionad intermofthe delay in producing another litter of offspring. Trivers argued that as 2 r. ”.1 E a.» .mu am. new in. e. we were ”we we. .mw bi new h... an D'. U. m ,3 a.-. 23 offspring age they benefit less fran nursing and becane increasingly capableof indepedency, hitthecosttothenntherincreasesastime goeson. Ifcostadbenefit are ennpressedinthesamemits, thecost totheuntherwillexceedthebenefittoheryougatsouepoint. Then the inclusive fitness of the nnther will be reduced. Ibwever, the off- spring sl'nuld attenpt to rnurse until tlne cost to the anther is more thantwicethebenefittohimelfbecanseheisrelatedtofutnme sibs byone-half. Sothen, atsomeperiodthereisaconflictbetween individual mximization of the mother's ad offspring's innclusive fit- rnesses. ' 'llnennthercaildbeecpectedtoexpelheroffspringvtnenthebene- fit/cost ratio of the act to the anther exceeds the inverse of her relatedness to her offspring, if production of anther litter is pos- sible . It follows that the aggressive interaction observed between fenales ad their offspring which results in the departure of offspring fran the natal site can be emlained on the basis of the inclusive fit- nesses of the anther ad the future litters of offspring. 'nneabovediscussionwasnntmeant tapertainto all rodent species. King (1955) reported that Q. 1Ldovicianus parents were rnt aggressive towards their offspring, ad usually deserted weaned litters. Adults established new burrows in available open areas, ad the young usually retained in central areas of the colony where predation was lower canpared to peripheral areas. Iayne (1954) nggested that T. striatms litters gradually wander away frcm the natal area. The unchan- isms associated with this "passive" deparmre were Int stated, but the possibility exists that exploratory behavior, uder selective control, can lead to departure fran the natal area. This tentative suggestion 24 hasbeeninvestigated. Saneg. 1.9.- Mjuvenilesdonnt leavethe natal area (Dice and Hward, 1951). Failure of these juveniles to departthenatal areamnayhavebeenduetodisappearanceofttemnthers of variation in the propensity of fenales to be aggressive towards their offspring as was reported by Savidge (1974) . Afinalpointtoconsider fordeparturefrantlenatal siteis that inntensive inbreeding could lead to a decrease in reproductive per- fonnance due to decreased heterozygosity (Crow ad Kimura, 1970) . A precise neasure of innbreeding has nnt been published for free-raging anall rodent populations, tintgh Howard (1949) estimated that 4 to 10 per cent of the litters produced during his study were from parent-off- spring or sib matings. This antlnr pointed out that juvenile dispersal fran the netal site would reduce the probability of parent-offspring matings. The trend for fenale deer mice to nature earlier than sales may serve toreducebreedingamnng litternntes (see Layne, 1968), if litters in free-raging populations disperse together as was found by Savidge (1974) in a laboratory study. Given the proposed selective advantages of juveniles that leave the natal site, other factors must be Operating to limit tre dispersal of juveniles because Rasuussen (1964), using I-bmrd's data for 2. En. M, estimated the size of interbreeding units within continuous habitat to be 12 to 99 individ- uals. Crossing Unfamiliar Space and Settlenent It is often stated that dispersing or wadering rodents have a high probability of death (5.3. , Lidicker, 1975), but the documentation forthis claimcanes frenfield stndiesvdnereendgrationfromthestndy 25 plot ad realized mnrtality cannot be distinguished (3.5., Watts, 1969, and Myten, 1974). Phtzgar (1967) ad Ambrose (1972) reported that experience within an area was a significant factor in avoiding preda- tion uder controlled laboratory conditionns for g. M and M. gemsylvanicus , respectively. Therefore , predation is probably rela- tively high for juveniles wadering through unfamiliar terrain. Presunably, the chance of reproductive success is inversely related to the time during which the individual continues to wader in unfamiliar terrain. Unfortunately, this time dependent measure cannnt be obtained for small mammals in free-raging populations, and various indices of dispersal distance have been used instead. Reports rage frun anecdotal accounts to detailed analyses . For example, layne (1954) simply snggested that youg T. striatus left their natal woodlot ; whereas Godfrey (1953) reported that only one of 21 g. pennsylvanicus, first capturedas immatures, was takenmnrethanabout27mfromthe innitial capture point when mature. Inward (1949) was able to determine the breeding site of 155 g. _m_n. ba__i._r__diout of 675 individuals for whom the birthplace was also know. One limited and nineteen individuals bred within about 152 m of their natal site. The renaining mnice nnved uptoabouthOOmfrantl'eirbirthplace. Males tendedtomnvefurther than fenales (Dice and Phaard, 1951). Murray (1967) proposed that the relative daninance of individuals in procm'ing a breeding site determine dispersal distances in verte- brates. A breeding site can be considered to be an area conntainning the necessary habitat features such as suitable space, food, nesting sites, predator escape routes, etc. According to Mirray, selection will favor individuals that aggressively procure breeding sites, but move any 26 frendeminant animals, evenifcontinueddispesal entailsariskto the individual. The probability of reproductive success of the sub- ordinnantanimalisnandmizedbyleavingtlepresenceoftledominnant; to stay and "fight to tie death" muld reduce its chances of repro- ductive success to zero. There are two stnrtcenings in this proposal pertaining to dis- pesal in g. g. ba_ir_d_i_. Firstly, the assunption that aggressive inte- action is the only behavior innvolved in the establisl'ment of social relationships may nnt be valid. Secondly, as was the case in past studies of the settlenent process in juvenile Peanyscus , failure to consider the role of the biological detenninnants of social behavior and consequent social relationships may result in inaccurate caise ad effect explanations of the settlenent process in juveniles. I-bweve, the conceptual framework of Mnrray' s proposal can be used to formulate snmnary predictions concerning the potential outcomes of social inte- action between a resident adult ad a dispesing juvenile. The follow- ing sumnsry predictions are jointly based on previous studies ad con- siderations of evolutionary arguments. In each case tl'e resident adult is nnt the parent of the strange, dispesing juvenile and is in breeding condition. ' S‘flfl Adult bale-Juvenile The works of Sadleir (1965) ad Healey (1967) lead to the predic- tion that an adult male should aggressively repulse all juveniles. The ultimate reasons for this prediction follow tlne argunent of Murray (1967). 01 the otlner had, Metzgar's (1971) findings sugest that 27 juvenile fenales ahead settle near adult males, but juvenile males sl'nuld nnt. The ultimate reasons for this prediction may be that a sensual relationship is formed between adult males ad juvenile fenales ad juvenile males are repulsed. Annther possibility is that juvenile males stnuld avoid adult males ad continue to dispese until tley encounter mates . Adult Nonlactating Fenale-Juvenile l-lealey (1967) largely igrnred adult fenales because preliminary laboratory expeinents indicated that they were rnt aggressive and did Int affect juvenile survival or growth rates. The work of Metzgar (1971) leads to tie prediction that juvenile males settle rear nonlac- tatirg fenales, but juvenile fenales do nnt . The ultimate reasonn for this prediction could be that a sexual relationship is formed between nonlactating fenales ad juvenile males , and juvenile fenales are repulsed. I-bweve, there is little evidence to indicate that nnnlac- tating fenales are aggressive. Annther possibility is that juvenile females should avoid nnnlactating fenales ad continue to dispese until they encomter a mate. lactatingFemale-Juvenile Social innteactions between tlnese social combinations have Int been previously examined. A tentative prediction, based on the occur- rence of rest defense ad matenal aggressien in Percmyscus, is that both male ad fenale juveniles should be repulsed (King, 1958, ad Scudder 31; gl_. , 1967). The ultimate reasen for this prediction may be that lactating fenales maintain a resource supply for milk production and/or for the offspring as they becane indepedent. Howeve, tie area . . I 333.: L..' ': image 3.. :1 J" ~65 28 ovewhichtherest defense is exteded isrnt knom. Also, theaggres- siveness of female Peenyscus with littes appears to diffe among tamnmic groups (Eisenbeg, 1968). The Purpose of This Study The purpose of tie present study was to obtain quantitative uneas- ures of the outcome of social interactions between an adult 2. g. M ad a strage juvenile conspecific. This investigation was designed to control or randomize the confounding variables of expe- ience, genetic constitution, age, ad population density. The sex of juveniles ad the sex and/or reproductive stage of adults were used as the indepedent variables. Ebcperimnents were repeated under two environ- mental conditions -- field-enclosure and laboratory -- in order to detemineiftlesocialvariableshadtlesamerelative influencein two conditions of space quality. A dispersing juvenile in a free-raging population may encounter a strange adult ad remain with that adult or continue dispesing. Tie adult presumably remains in its home range. This condition was simu- lated by placing a juvenile male or fenale in an enclosure containing an adult male, a nmlactating ad nonpregnant adult fenale, a lactating fenale witin he litter, or nn adult as a control. Juveniles could move through a two-way tunnel to an adjacent uncoupled enclosure, but the tunnels wee designed to prevent the passage of adults. All dee mice had access to nest boxes. The effects of social inteaction upon juvenile settlenent wee evaluated ecperimentally by measuring the spatial relationships of juve- niles ad adults within the enclosures. Sinnce it has been claimed that is! (r‘ Y E “I if}: ‘1 29 tie growth rates of juvenile Peomyscus are retarded by same adults, the weight changes of juveniles were measured. Also, seveal othe measures wee taken (9g, activity and wounding) during this study. The social conditions were altered independently. Theefore , the rela- tive diffeences in the dependent variables in respect to the juvenile sexes ad in respect to the adult classes could be determined. It was expected that the outcome of juvenile-adult social inteactions would be affected by tl'e biological classes of inteacting animals. The present investigation evaluated the following working hypo- theses concerning the outcenes of juvenile-adult social inteactions in E. g. bairdi: (1) The sex of juveniles ad the sex of adults ad/or the repro- ductive stage of adult females affects juvenile settlement. (2) Tie sex of juveniles ad the sex of adults ad/or the repro- ductive stage of adult fenales affects juvenile growth. (3) The relative effects of the biological classes will be simi— lar under field-enclosure conditions and controlled laboratory condi- tions . 5s?“ [fittat “mm MEIHODS seas: 'llneeqneimentrequiredamaleor fenale juvenileto encounte: (1) an adult male, (2) an adult fenale that was neither lactating rnr pregnant, hereafte referred to as a nonlactating fenale, (3) a lac- tating fenale and her litte, or (4) 1n adult as a central. Expei- mnents were repeated using field and laboratory enclosures with five replications for each treatment canbinatien (3.3. , location it juve- nile 1: adult, Table l). Oveall, a total of 80 juveniles and 60 adults were used. Juveniles were allowed to interact with adults over four consecutive nignts. 'fierefore, the total nunbe of test nights was 320. Tablel. Theexpeimental designusedinthisstudy. 1.— Iocation: Field Enclosure laboratory Enclosure Adult Juvenile Male Fenale Pale Fenale Male 5 5 5 5 Nonlactating Fenale 5 5 5 5 lactating Fenale 5 5 5 5 Control (No Adult) 5 5 5 5 30 31 Experimental Animls The prairie deer muse, Perogyscus maniculatus bairdi, was chosen as the experinental animl for the following reasons: (1) A vast aunt of background information is available for this gems and especially this species (King, 1968b) . (2) Previous workers have described dispersal in free-ranging populations and have evaluated the influence of social interaction on thedispersalpheuneminthisgenus Ohmd, 1949; DieearxiHovmd, 1951; Healey, 1967; Metzgar, 1971; and Savidge, 1974) . (3) This species is easily unintained in captivity. Allndceused inthis studywere theFl, F2, or‘n“3 generations of wild caught parental stock collected in the vicinity of East lansing andAnnArbor, Michigan. While in the colony, udcereceived food (Wayne lab-Blunt) and water _a_d_ M3 wood shavings served as bedding. Mice usedinthe field testsweremaintainedintheoolonynmderanatural photoperiod, andmiceusedinthe laboratory testswereunintainedand tested under a 14-hour light; lO-hour dark cycle. Mice were dram at randan for testing and were used only once. Food (Wayne Iab-Blox) was scattered ad 1.413.132. in the enclosures during tests. The water source for mice in the field enclosnures was dew and rain, but water bottles with sippermbeswereprovidedduringthelaboratory experiment. All juveniles were unintained under similar conditions in the col— onyuntil testinginordertoreducevariationinthedepedentvari- ables due to differences in early experience (see Russell, 1971). The brood cage population densitywas the same for all juveniles: the par- ent pair and four litter mates, usually two mles and two fenales. Culling was done by the time litters were five days old. The bedding 32 of brood cages ms not charged unless absolutely necessary in order to keepthehandlingofjuveniles to aminiunnn. Litterswereweanedat 21 to 24 days. Subsequent to weaning, juveniles fran a single litter were housed in side—by-side, 130 x 150 x 280 cm cages; mles were placed in one side and the fenales in the other. A section of 6.7-nnm—tnesh hardware cloth wsplacedoveratmnelconnectingthetwosides. 'Iherefore, visual, olfactory, and anuditory contacts between the male and fenale sibs were possible, but copulation ms prevented. A section of 1.3-en—mesh hard- ware clothwas placedbetween thewater andfood sources to insure that alljuvenilestadecperiencesqueezingtm‘onghthissizeopeningsince theywere required to do so during the experinents. Juveniles were 30 to 36 days old at the start of testing. Breeding pairs provided a constant supply of litters and served as enperinentaladults. fiepairsweremintainedinlBOxlSOxZBOen or150x125x280cmcagesandwerenot separateduponthebirthofa litter. Adult mles were drawn for the experiants irrespective of the breeding condition of their paired fenales. Adult fenales that were to be used in the eqnerinnents as nonlactating fenales were separated from theirpairedneles topreventpregnancybyusingthe sane typeof cage used to separate opposite sex sibs. These fenales were judged not to bepregnant iftheyfailedtogainnweightdnringalO-dayperiod inund- iately prior to being used in the eu-nperimts. lactating fenales had a onetoeightdayoldlitteratthestartofanexperimntalrm. A11 adults were tested with strange juveniles, were in breeding condition at the tine of the eucperiuents, and were successful breeders. 33 mes. Enclosures The field and laboratory enclosures differed in respect to size, space canplexity, and environmental vicissitudes. The field enclo- sureswereabourninetimes larger inareathanthe laboratory enclo- sures in grassland habitat, annd were exposed to the weather. Sixteen 3.8 x 3.8 m enclosures located in grassland habitat were used for the field equeiment (Figure 1). Figure 2 gives the spatial arrangement of the field enclosures annd describes the surrounding habitat. True walls of the field enclosures were abouut 66 cmhigh; the bottomle was 6.7- mn-nneshhe'dm'eclothandthetopmsalundnmflashing. Inorderto inhibit burrowing, the hardware cloth was buried and linestone chips wereplacedbythehardware clothatthegrounlsurface. Tarpaperwas placedoverthehardwareclothtopreventmicefranseeingothermce in adjacent enclosures and fran viewing the surrounding ground-level landscape, bunt the treeline of a nearby woodlot was possibly within View of the mice. 'Ihe grass. in the enclosures was periodically clipped sothat sindlarhabitatwasmintaineddringthecourseofthesmdy. Anelectric livestockfencewasplacedaroundtheperiphery ofthe enclosures to discourage terrestrial predators. Sane juveniles dis- appearedfromthefieldeuqnerimntsduetounknmreasons. Anemi- nationn of differences in the tendencies of mle and fennle juveniles to disappear is given in Appendix A. Twelve 1.2 x 1.2 m nascnite enclosures with 0.66 m high walls were used for the laboratory experiunent (Figure 1). These enclosures were placedonnadle-coveredconcretefloor. FigureBgivesthespatial arranngenent of the laboratory enclosures. fine "day" period light Figure 1. The field enclosures (A) annd laboratory enclosures (B) used in this study. 35 Figure l. 36 2 Figure 2. Arregement of the field enclosures anddescription of sur- rounding habitats. The locations of the resident and dis- persal areas (i.e. , north-south enclosure pairs) were systemtically altered during the experiment to remve habitat selection as a factor in determining the unvenent of juveniles. gras 37 grassland 2 road grassland #— wood lot grassland Figure 2, L - grassland 38 Figure 3. Arrangemnt of the laboratory enclosures. The locations oftheresident (R) anddispersal (D) areaswerethesane for all tests. Figure3. 39 40 intensity was abouut 3 candlepower white light (a loo-watt bulb over each grouup of four enclosures) and about 0.25 candlepower red light (a 7.5- mtt buulb over each of the enclosures). 'nne "night" period lighting conn- sisted of the red lights alone. We Tum-my tunnels (4.3 x4.8 x 17.5 cm), constructed from 7.8-nnn- thick, uurnpainted Plexiglas, were used to treasure the location and acti- vity of juveniles. An operating arm activated a micromitch when a juvenile passedtl'rougnatunnel (Figurelu, seeAandB, respectively). A sectian of 6.7-nm-unesh hardmre cloth covered with clear plastic sheetirgmsplacedovertl'emltoprotecttheMcroswitchfrantle weatherinthefieldandtodiscourageudcefrunactivatiugtheoper- atinngarmfranouutsidetheumnel (Figurelu, seeC). 'l'ne tunnelswere set to be operated by a weight of 7.0 gun; juveniles reely weighed less thann9.0 gn. Inordertorestrictthemvenentoftheadultswhileallmingthe juveniles tomve freelythroughthe tunnels, 1.3-en-ueshhardwarecloth rectangular structures hereafter terued restraining grids were placed overtheendsoftheturnnels (Figure4, seeD). Juvenilescouuldsqueere thrmghthe 1.3-cnn-squareopening innthehardvare cloth, but tlneadults couuld nnot (Savidge, 1974). A tunnel, tanned the passage tunnel, allowed juvenilestounvebackandforthbetween: (l)anenclosurecontaininng anadultormadultasacontrol,hereaftertermedtheresidentarea and(2)anadjacentenclosuretowhichtheadultdidnnothaveaccess, hereafterternnedthedispersalareawigureS). Thepassagetumels hadrestraininnggridsonbothends (Figure4, seeDanndD'). 41 Figurelu. 'nnetunnelapparatususedinthisstnudym-operatingarm, B-unicroswitch,C-tunnelcover,DandD'-restraining grids, annd E - feeding statian). 42 I l';;:.. l I l 7.: fli: .- a nil. ’ .- Figure 4. 43 FigureS. Bumpleoftwoeuquerimeralenclcsurepairsshovdngthe anrangenentof passage tunnels (X), feeding stations (F), anndnestbounes (0). DISPERSAL AREA RESIDENT AREA Figure 5. Could ' 1 Eltmr 45 Other tunnels were used to measure the activity of juveniles within both the resident annd dispersal areas. These tunnels, termd feeder tunnels, hadarestraininggridononeendandafeeding stationnonthe other end (Figure 4, see D and B, respectively). The feeding stations were constructed from 6.7-nnnn-mesh hardware cloth and had a 7.5 drann plasticvialattachedtooneside. Unnhuuskedsunflower seedswereplaced inthevialanndcouldberemvedbythejuvenilesviaanopeningcut in thebottannof thevial. Feeding stations were placed at ranndomintwo of the three available corners of the dispersal annd resident areas, respectively. 'lhe arrangenent of the passage tunnels annd feeding sta- tions was the sane for the field annd laboratory experiunents (Figure 5). Recordinquuipmmt 'lhe mcr'oswitches were connected to a 20-channnel Esterlinne—Angus event recorder powered by a lZ-volt battery in the field annd a lZ-volt powersupplyinthelaboratory. Tnepaperspeedms1524anperhmr, andtheresolutionwasaboutlSsecndnnimmtimebetweensignals. 'Ihe activity of juveniles was detennined by examining the chart paper. 'Ihennunnberoftripsmadetlroughapassageorfeedertumel cenldbecouunteddirectlyfrounthepaper. 11netinejuvenilesspenton eithersideofthepassageturmel,thatis,withintheresidentareaor dispersalareas,cmuldbeascertainedbysumdngthetineperiodsdring whichthetreadlewaspoinntedtowardsagiven area. For earple, the unlcrowitchndghtbeoffwhenajuvenilewasintheresident area. unenthejuvenilepassedttroughthetunneltothedispersalarea,the positionofthetreadlechargedandcausedtheoperatinganmtoturn thendcroswitchon. 'I‘hen,themieroswitchmuldbeturnedoffwhenthe 46 juvenilepassedbackintotheresidentareaagain. Inthisexanple, thetotaltinethejuvenilespentintheresidentareacoufldbecalcu- latedbysumdngthetmperiodsdringwhidnthenflcroswitchwasoff. NestBoxes Alluricehadaccesstonestbonnesduringtheenperinents. Two nestboxeswereavailableineachresidentarea;onewasplacedineach dispersalarea. 'l‘hearrangenentofdnenestboxeswasthesameforthe fieldandlaboratory experinnnts (FigureS). Claydrainnage tiles were used fornestboxesdnringthefield experiment. Thesetiles: (l) were ofacylinndrical shape, (2) wereopenat both ends,and(3) were placed verticallyinthegromdabout three-quarters of theirlegth. Circu- lardisksof sheet metal, placedonplastic Petri dishes, wereput insidetlnetilestoserveasthebottomofthenestbonnes. One-half- gallon plastic containers were slipped over the tops of the tiles to preventsumlightfranenteringthenestbouues,and22.6en2plymod sectionns were attached to the plastic containers to protect the nnest boucesfrantheweather. A2.6enentrytnlewasdrilledintl'esideof thetiles. Asection of 6.7—nm—neshhardm’e cloth, placedonthe insideofthetileandextendirgfruntheentryholetothebottanof thenestboau,allowedmicetoclinbinandout. Plasticfreezerboxes (l.5pinnt)servedasnestbouuesdringthelaboratoryeuperiment. These nestboxeswereplacedontheirsidesandweretapedtothefloor. A l.9x2.6enaccessholewascutinthecovers. Allnestboxeswere suupplied with cotton Nestlets (Ancare Corporation) annd several lab chow pellets. 47 Eugerimental Procedre An experimental enclosure pair consisted of a resident area end an adjacent dispersal area . Adults wee always restricted to resident areas, and juveniles always had access to both resident and dispersal areas. ‘llne juveniles and adults were maintained in separate experi- nental enclosure pairs for a three-night pretest period. For enemle, ajuvenilemayhavebeenmaintainedinenclosurepaierfFigureS anndann adultunyhavebeenmaintainned in the resident area of enclosure pair B of Figure 5. Bring the pretest period juveniles couuld nave tlrouughpassageturnnelstoadispersalareaandhadaccesstofeeding stations. The restraining grids wee placed on the tunnels for the last night of the pretest period. Adults were restricted to a resident areaardhadaccess tonestboxes. 'Ihepresentperiodwas intededto allow adults to establish residency inn a resident area. Healey (1967) indicated that a two-day period was sufficient for residency establish- nent. Also, allmicecouldbecaneaccustanedtotheexperinental environments and the test apparatus (tunnels and/or nest boxes) during pretesting. The euperinents wee conducted over four-night periods innediately following the pretesting, and wee initiated by placing the juveniles intheresidentareasoccupiedbytheadultsorcontainingnoadultsas controls (g.g., a juvenile might have been rennved fran enclosure pair Aandplacedintheresident ereaof enclosurepairB, Figure 5). The four-day period of tinne for the experiments was chosen for the follow— ingreasauns: (1) alongerperiodintheconfinedareasmyhaveledto social habituation between the adults and juuveniles (Hill, 1970) , (2) Healey (1967) nnoted a drauatic weight channge in juveniles after four 48 days exposure to aggressive adults, (3) _li. n_n_. air}; fenales come into estrus inat leastthreedaysuponeuposuretonales (Bronsonand Mesden, 1964) , and it was of interest in the present study to determine if juveniles were reproductively coupetent , and (4) the four-night period allowed a robust measure of consistent differences between the juvenile-adult social ccubinnations. Procedure Used in the Field ngeriment The sequence of testing the juvenile-adult canbinnations was ran- domly chosen to insure that weather conditians in the field did nnot bias the experiment in a systenatic fashion. Juveniles wee weighed to theneeest 0.1 gnannatriplebeanbalance, examined forwouunds, and fur-clipped for identification. Adults couuld be identified by toe- clips made when they were paired for breeding. All mice were placed in plastic cages, driven to the field, {and restricted to nest boxes in separate ranndanly chosen resident areas . 'Ihese operations wee cen- pnetedbynooon'sn. Nestboxeswereopenedintheevenirgpriortothe first pretest night by 1900 ET, and all mice could nave about in their respective experiunental areas during the three-day pretest period. After the three-day pretest period was canpleted (anne juvenile tested in the field experiment as a conntrol was in the pretest condition for four days), juveniles were hand-carried to a resident area containing annadult orno adult as a control andweerestricted to annenptynnest box by 1000 PST. All tunnels, restraining grids, and feeding staticnns weeremvedfranthepretest equeiunentalareas, cleanedofurineand feces, annd placed in the experimental areas occuupied by resident adults 49 or containing nno adults. The deer mice were free to interact usually byl900mmenthejuvenileswerereleasedfrunthenestboxes. The sleeping sites of the mice were determined each day by quuietly looking into nest boxes or searching for surface nnests usuually before 0900 EST. I-bwever, no recordings of activity in the passage and feeder tunnelswereeuannlnedindetailunntiltheexpeimentshadbeencan— plated to avoid introspection of the data. Tunnels wee checked for propefunnctioningevenyeveningbyl900ESI, butscuedatawerelost due to equipumt malfunctions. At theedofthefour-night testperiadallnniceweeplacedin plastic cages and driven to the laboratory usually by 0930 ESL Juve- niles wereweighedto thenearest 0.1 gnona triplebeenbalanceand eeminedforwaunds. lnordertoassure that juvenileswee capableof going throughtherestraining yids, theywereusuallynmintainedalane incages inwhichthewateranndfoodsourceswere separatedby 1.3-cm- mesh hardware cloth for at least onne week subsequent to testing. Juve- nile fenales tested with adult males and nonlactating fenales tested with juvenile males were mintained in individual cages for at least threeweeks subsequent tothe experiments. 'l‘hesedeernndceweechecked periodically for the birth of litters to determine if they had been inpregnated during the experiments . All othe adults wee returnned to their respective antes. 'Ilunnnels, tunnel caves, feeders, restraining yids, andnestboxeswerecleanedorreplacedattheendofeachtest. At leastonedaywas sldppedbetweenequerimentalruns toallawdeyada- tionn of scents deposited by mice in tlne previauus ruunn. Wecker (1963) shmed that _13. g. b9;i_r_'_c_l_i_ selects yassland aver woodland habitats. Therefore, since the mice ccuuld possibly see the 50 treelinne in the woodlot located to the south of the enclosures (Figure 5), the positions of the resident and dispersal areas were reversed for every replication of juvenile-adult ccnbinatians . For exenple, one eqneinentalrumusingajuvenilemaleandanadultmalemighthavehad the resident area to the south, and the next replicate of the juvenile male-adult male social canbinationwould have had the resident area to the nnorth. This remved habitat selection (_i_._e_., mving away from the woodlot) as a possible consistent influence on the mvenent of juveniles between the resident and dispersal areas . Field testing was conducted during June to October, 1975. Procedre Used in the laboratory Eugeinent Treatment of mice used in the laboratory and field tests was the seas, except that the animals were hannd carried in plastic cages to the laboratory. Also, sinnce behavioral observations were nude in the laboratory, juveniles were randonnly chosen by another individual to insure that the sex of juveniles was unknown to the observe. Mice were placed inn separate resident areas and restricted to nest boxes by 1315 EST. All mice were released from the nest boxes between 1850 end 1900 EST innnediately prior to the lO—haur “night" period (1900 to 0500 F31“) . 'Ihe sleeping sites of the mice were observed by 1200 EST, and the correct opeation of equipnent was usually veified between 1700 and 1830 EST. laboratory pretest annd test procedres were the sanne as in the field experiment, except that juveniles in one replicationn within each of three treatment ccnbinatianns had mo days experience with the restraining yids instead of onnly one day. Mice were returned to the 51 colanyby0930ESTfollawirgthelast test night. All equipment, includingthepens,wasthoruughlycleenededanotlereupeimental rumwasstartedthesanedayblelSEST. Micewereobsewedfranaplatformarstepladderaverlookingthe residentareacnthefirstandlast night ofthetest periods. The observations usuually startedabount 10to 15minnutes into the"night" period. Eachresident areawas observed for a 15-second interval every nnlnuute for 60 consecutive nninnuutes. Observations were ccntinnuuous in nnst casesbecausecodedbehaviorswerenotedbyspeakingsoftlyintoatape recorderwhiletheattentianoftheobserverenainedcnthendce. The tapesweretrannscribedatalatertime. Duringanenightthecodes were sinplywrittenonpape, andtheobservehadtobriefly look away franthemice. 'nnefollawingdefinniticnsofbehaviarsreporteduponin thispaperwereused: (l) Muonnemusemshingatamtherandthelattenming may (2) wulockedfigm (3)Avoidanr:e--annusennvingauayasanotherapproachedwit1uut rushing (4) Mrualavoidancenbothmicenavingauayafteraneapproached 'nneavoidencebehaviorwasnotedonlywhenndceweremnnrethmnBOto 45cmapartatthetinecneorbothnnved. Abehaviorwasscorednnre thannanceinasingle 15-second interval ifannuse stoppedthebehavior, engagedinadiffeentbehaviar,edtheninitiatedthefornebehaviar. 'lhe lab testingwas conducteddringJanuan-y toApril, 1976. 52 Depeidext Variables The basic objective of this research was to obtain quantitative masures of the outcone of social interactions between juvenile and adult prairie deer mice. In order to achieve this objective, several dependent variables were measured. The variables represented three dif- ferelt categories: (1) juvenile settlenent, (2) juvenile weight change, and (3) descriptive variables . Juvenile Settlenmt The effects of social factors on the settlenent of juveniles were evaluated eqaerimentally by measuring three aspects of the spatial relationships of juveniles and adults within the enclosures. If the social factors exandned in this atuiy affect the settlenent of juve- niles, then there should be differences in the propensity of juveniles to (l) nest, daring the daytime inactive period, in enclosures occupied by adults, (2) share a nest with adults if juvexiles nest in enclose-es occupied by adults, and/or (3) ramin in enclosures occupied by adults during the night-time active period. Nesting Area. The area in which juveniles nested (hiring daytime was observed on 315 out of the 320 days of the experimts. (11 five days juveiiles could not be found during the field experimt, and the position of the passage ttml treadle was used to indicate their loca- tion.Inafevcaseachn'ingbod1d\efieldandlaboratoryecperimmts, juvemileswerelocatedinthepasaagemlandwerescoredasbeing inthedispersal area simetheyhadmvedfrantheresidentarea. Nestinngrraggmt. 'lhe nesting arrangemnt of adults and those juvmiles that retained in the resident area was determined. The 53 arrangenitmsscoredasmmaalnestingornesting separately. Inall cases,enceptonevhetthemcesharedasurfacenest,juvetilesam adiltsthatnestedtogetl'erdidsoinnestboxes. Mt ofTimeintheResideit Area. Anestinate of time that juvmilesspeitintheresideitarea,expressedastheperce1tofthe totalactiveperiod,msmdefrantherecordingaofactivityinthe passage tumel. 'Ihismeasuremistbecmsideredanestinnte because juveniles omldgothrmghthepassagettml,h1tstill couldrenain intherestraininggridratherttmtrulyeitertl'eresideutarea. 'merefore, activity in the passage unnel did not necessarily measure thetinejuveiileswerefreelymvirgmtheresideitarea. 'meactiveperioddm'iigfieldtestsmstakmastheperiodfran mettosmriseasreportedbytheU.S.NationalWeatherServicefor EastLansing, Michigan. 'meactiveperiodchrringlaboratorytesting mscmsideredtobethelO-hmm'night"period. Theremsscnejuve— nileactivitychringthedayperiods,h1tthiswasigmredinthe analysis. Samdatawerelostduetoequipmtnnlfimtimvhenelec- tricalcomectimsbecanelooseortumlsweretrippedinthewrong direction. Inthesecases,theactiveperiodwastakminthetotal tinedmingdxenigitperiodvtmtreumlmsoperatingcorrectly. Iclnsetotakethesemasmesbecmseeachprovidedanestiuate ofthemmaluseoftheresidentareabyjuvexilesarflachlts,“ the incorporation of three measures allowed greater precision in estab- lishing the juvenile-adult spatial relationships coupared to the ham range overlapmeasures ccnnrnlytakminfield studies (5.3., Metzgar, 1971) . A significant differ-ewe in one or more of the settlenent measures annng social coubinations chosen for the present study was 54 interpreted as indicating a relative limitation of juvenile settle- Mt. 'nne neasures of juvenile settlenent could be expressed either as daily and nightly values or total and nean scores. Table 2 gives a snmmry of the outcomes possible for all three variables. Table 2. The outccmes possible for the three measures of settlenent used in this study. Nesting Arrangenent for Per Cent Time Nesting Area Juveniles that Nested Juveniles Spent in of Juveniles in the Resident Area the Resident Area Resident area With Adult 100% or to Outeane or Dispersal area Alone 07. Juvenile Weight Change The weight change of juveniles was measured by determining the dif- ference between their initial and final weights. This dependent vari- able was clnsen because the effects of social factors on the saturation of juvenile small rodents has not been critically enshrined under experi- mental conditions, such as those of the present study, that allowed juve- niles to avoid interactions with adults and that maintained all juve- niles for nneasurenent . Descriptive Variables In order to W quantify the effects of the social factors and to gain insight into the behaviors and social relationships involved in interactions between juveniles and adults , several descriptive measures were taken. These treasures were related to juvenile activity, wounding of 55 juveniles, direct observation of social interactions during the labor- atory ecperinent, end reproduction. Juvenile Activity. Tho aspects of juvenile activity were measured: (1) activity between the resident end dispersal areas end (2) activity within the resident and dispersal areas. The activity in the passage tumel provided a measure of the frequency of juvenile unvenents between the resident end dispersal areas. These data were expressed as the nun- beroftnrnnneltripspertnrroftheactiveperiod. Analysis ofthe passage tunnel activity indicated whether the rate of mvenent between areas was influenced by social factors . Tl'e activity in the feeder tunnels provided en estimate of the frequency of juvenile mvenents within the resident end dispersal areas. Tnesedatawere expressedas thenunberof tunnel tripsperl‘nurduring which feeders were available to the juveniles. For exeple, if a juve- nilewas intheresident areaforthreehours duringtheactiveperiod, totalnmberof feederl'mrrswas six since therewere two feeders avail- able in the area. This measure unst be considered a gross estimte of activity becense of interfering variables . Differences in mvenent could havebeenchnetodifferencesinindividualpreferences forsmflower seeds or in propensities of juveniles to escape adults by seeking refuge in the feeding stations. These data will be presented in Appendix B. _Wg_nr_n£1__ing. The wounding of juveniles by adults was determined by flaking a direct count. Care was taken to avoid double counts (33., wounds inflicted by upper endlowerincisors). Therefore, a wound was considered to be the result of one bite. Direct Observations . Behavioral interactions between juvenile- adult social conbinaticns were observed. 56 Reproduction. The mnber of adult end juvenile fenales that became pregnant during the enqneriments was determined. Statistical Procedures The dependent variables chosen for evaluation provided data that were discrete end eelyzed with chi square tests or continuous end analyzed with nonparametric or parenetric procedures . The four-day total end daily nesting area frequencies were analyzed with multi-way chi square tests (Gill, 1976) , but only the four-day total enalyses are presented in the results section. The results of tests on the daily values are given in Appendix C. The data were treated in this manner because for all comparisons except one, the total end daily enalyses had the sene outcomes for at least three of the four days. Also, the chi square approximation may be poor for the daily tests because of the smell expected frequencies in some cells. The four—day total frequencies represent en artificial four-fold increase in sample size since the sens juveniles were scored over the four-night periods. Ferguson (1966) pointed out that as sample size end differences between samples increases by a constant multiplicative fac- tor, the xz-value will also increase by tie sane constent factor. There- fore, since many juveniles remained in the same area on all four days (64 of 80 tested), the calculated xz—values for the four-night total fre- quencies were divided by four to arrive at a very conservative adjusted critical value taking the lack of statistical independence into account. The four-day total nesting arrengenent frequencies were analyzed with Imlti-way chi square tests , but the daily values were not enalyzed due to the ectrerely low expected frequencies in mneny cells. The S7 calculated xz-values were divided by four to adjust for the lack of statistical independence across days because 27 out of the 37 juveniles that remained in the resident area on all four days did not change nesting arrangenents. It should be pointed out that the degrees of freedom for the multi- way chi square tests are based on the number of criteria of classifi- cation rather then the number of rows end columns (Gill, 1976). For example, prior to testing for differences in the individual mnain effects of location, juvenile sex, arnd adult class (Lg. , pooling data for other main effects), it was necessary to determine if differences in the proportions of juveniles that nested in the resident area end/or nested with adults existed mg the 16 treatment groups (1.3., 2 locations x 2 juvenile classes x 4 adult classes) by using a 16 x 2 contingency table. In these tests, df = (2-1) (2-1) (4-1) (2-1) = 3 since there were 2 locations, 2 juvenile classes, 4 adult classes, end 2 areas or 2 pos- sible nesting arrangements. Two-way contingency tables were used for some comparisons of nest- ing area, nesting arrangement, end tunnel utilization data. Yates' correction for continuity was not used as suggested by Conover (1974) . The four-night meens of the per cent time measure end both the passage end the feeder tunnel activity rate measures in the field experiment were examined by nonparametric Kruskal-Wallis end Wilcoxin's rank-sun tests. Parametric procedures could not be used because the raw data could not be trensformed to meet the required assnnptions. I-hvever , the nonparametric procedures used in this study allowed the same comparisons as those possible in a 2 x 4 factorial analysis of variance. The field experiment had one missing replication due to 58 equipment malfunction. The missing datnm was replaced with the group mean rather then replacing all replication scores with the group meen as suggested by Bradley (1968) . This procedure was followed because the large number of ties would have magnified the existing differences. Four—night meens and nightly scores were used for comparisons of the per cent time and passage tunnel activit;i rate measures in the laboratory experinent. The raw data for each replication were trens- formed to the arcsin end W scales , respectively. Hmever, only the four-night meen scores were used for canparisons of the feeder tun- nel activity rates . The raw data for each of these replicationn meens were trensformned to the W scale. Each dependent variable above was enalyzedbyalenfactorial enalysis ofvariance. Canparisons of the changes in juvenile weights in the field end laboratory experiments were analyzedby a 2 x2 x4 (3.3., locationx juvenile sex it adult class) factorial enalysis of variance. Trens- formetion was not required for the weight chenge data. chnpgeneity of variances was confirmed using the F-mex test. Dif- ferences among sane simple main effect means were exemined by _a_ posteriori Student-Nevmn—Keuls tests . The parametric methods were described by Sokal end Rohlf (1968) end Winner (1971). The nonparametric procedures followed Bradley (1968) end Sokal end Rohlf (1968) . A pr__ior_i_ significance levels were set at alpha - .05 for all statistical tests. Notailedtestswereusedthrougl'eut. Theenalyses ofvarianceand dnisquaretestsweredoremntteWeng600-14calenlatorintlemchigan State University Museun. RESULTS Theresults oftheexperimentshavebeenorgenizedinto three sub- sections under tlne headings of juvenile settlement, juvenile weight change, end descriptive measures. The first subsection contains the enalyses of the settlenent measures end juvenile activity in the passage tunnel . The effects of the expe'imental treatments on the weight changes of juveniles are evaluated in the second subsection. The wound- ing, behavioral observation, end reproduction data are presented in tlne final subsection. Juvenile Settlement Nesting Area. The propensity of juveniles to rest within field or laboratoryenclosures occupiedbyenadultwas eneninedtodetermine the effects of the social factors on juvenile settlenent. Prior to testing for differences in individual main effects by pooling data for tie remaining main effects, it was necessary to determine if differ- encesedstedepngthel6treamentgroupsbyusinga16x2contin- gency table. The percentages for this comparison are given in Table 3. The four-day total proportim of juveniles that nested within the resi- dent area was not jointly independent of location of the experiment, juvenile sex, end adult class (XZadj - 28.3820, df = 3, .005 > P). Therefore, the data could not be pooled to test for my single main effect. The four-day total nesting area frequencies obtained in the 59 a‘—.~ unv a 3 HM HEEHQMWOH aim-EU NH“ I I is. F VI High“; - I FNI [Ind Al, n 1 Flu \Aflfi as W HIV I. I “In! R. H EU mu: ~ ~ . - IN “‘0 b NV ENE NV .6 Q m. H PH 1' ‘1 O I . U 'l .H. ‘1‘.“ F . I 6O .xon umcc w on ecuoHUmcH bummed—gem: was 3.295.». m 553 new «co uom q u zep .3.“qu ennuhfiom m .83 acumen meager—an m mammouemu Eudoo none I w 8 8 8 8 2 S 8 8H Ewen.“ o 8 8 8 8 8 8 8H e ea 8 8 8 8 8 o 8H 8H m e8 8 8 8 8 8H o 8H 03 N e8 8 8 8 8 8 8 8H 2: H ea fines. ea 3mg wise 2e: 9:3. ezw «Tam wise 3e: Hausa @53qu £88382 33 3580 mended 353:2 3.2 «Have «:88 dea «fig pan—fig 33.30an 38 SH 8 8 8 8 8 8 8H tease 8H 8 8 8 8 8 8 8H e as con 8 8 8H 8 8 8H 8H m .3 8H *8 8 8 8 8 8 R: N e8 8H 8 8 8H 8 8 8H e03 a e8 GEE ea dame wine me: 992 2% 326m mime an: 3.88 gauged 35882 “32 39.8 wfiuflufl 338de “32 wise 3e85, mesa «Hausa. cg Bu: £98 ”.5339 m5 5 e33: menu mega «nu mo mowwufiwocwe 888 9.6% H38 $3M new .033 can no g .m 033. l I vac...— Rm fermile 61 field and laboratory experiments were analyzed separately to determine the effects of the social factors on juvenile settlenent. In the field experiment, the proportion of juveniles that nested within the resident area was jointly inndependent of juvenile end adult social conditions (8 x 2 contingency table, Table 4). These data could be pooled to test for differences between male end fenale juveniles end ampng adult classes. A 2-way contingency enalysis showed that there was nno signnificent difference in the proportionns of juvenile males end juve- nile fenales that nested within the resident area (Table 5) . Differ- ences anong adult classes for data pooled fun the juvenile sexes were also nonsignificent (Table 6). Inn conclusion, under field enclosure conditions, the use of nesting areas by juveniles was nnot dependent uponn the sec of juveniles or the sex of adults/reproductive stage of adult fenales. In the laboratory experiment, the proportion of juveniles that nested within the resident area was not jointly inndependent of juvenile end adult social factors (8 x 2 continngency table, .005 > P, Table 7). It was necessary to test for differences between juvenile mles end fenales separately for each adult class (2-way contingency tables) end to test for differences among adult classes separately for juvenile neles end fenales (4 x 2 contingency tables). . Therewereno significannt differences intheproportions ofjuve- rnile mles end juvenile fenales that nnested within the resident area for adult males (Table 8); for nonlactatinng fenales (P 2 . 065 , Table 9); for lactating fenales (Table 10); end for conntrols with nno adults (Table 11). However, there was a signnificannt difference among adults for juvenile males (.01 > P > .005, Table 12). Inspection of Table 12 62 Table 4. Chi square enalysis of the four-day total nesting area frequencies of juveniles for the eight treatment canbinations used in the field experiment. Juvenile Pele Juvenile Feels Area: Resident Dispersal Resident Dispersal Adult Male 20 0 l8 2 Nonlactating Feels l8 2 13 7 lactating Feels 16 4 13* 6* Control (No Adul.t) 16 4 20 0 xzadj - 4.5542, df - 3, .5 > P > .1; N - 19* Table 5. Chi square enalysis of the fouroday total nesting area frequencies of male end feels juveniles during the field experiment. The data were pooled for the adult classes. Resident Area Dispersal Area Juvenile Male 70 10 Juvenile Feels 64* 15* Xzadj - 0.3156, df= 1, .9 > P > .5; Na 79* ] ab? it 63 Table 6. Chi square enalysis of the four—day total nesting area frequencies of juveniles during the field experiment. The data were pooled for the juvenile sexes to examine differences sung the adult classes. '— Area: Resident Dispersal Adult Male 38 2 Nonnlactating Feels 31 9 lactating Feels 29* 10* Control (No Adult) 36 4 Xzadj - 2.1853, df - 3, .9 > P > .5; N = 39* Table 7. Chi square enalysis of the four-day total nnesting area frequencies of juveniles for the sight treatment ccubinations used in the laboratory experiment. 1 1: l 1.— Juvenils Male Juvenile Feels Area: Resident Dispersal Resident Dispersal Adult Male 20 0 l4 6 Nonlactating Feels 19 l 8 12 lactating Fenale 2 18 5 15 Control (No Adult) 15 5 8 12 Xzadj - 15.4865, df - 3, .005 > p Table 8 . Chi square analysis of tlne four-day total nesting erea frequencies of juvenile mles end fenales for adult mles during the laboratory experiment. Area: Resident Dispersal Juvenile Male 20 0 Juvenile Feels l4 6 xzadj =- 1.7645, df = 1, .5 > P > .1 Table 9 . Chi square enalysis of tlne four-day total nesting area frequencies of juvenile males end feelss for nonlactating fenales during the laboratory experiment. Area: Resident Dispersal Juvenile Nels l9 1 Juvenile Feels 8 12 Xzadj =- 3.4473, df =- 1, p e .065 65 Table 10. Chi square analysis of the four-day total nesting area frequencies of juvenilenmales and females for lactating females during the laboratory experiment. Area: Resident Dispersal JUVenile Male 2 18 JUvenile Female 5 15 xzadj - 0.3896, df - 1, .9 > p > ,5 Table 11. Chi square.ana1ysis of the four-day total nesting;area frequencies of juvenilewmales and.fEma1es for the control*with no adult during,the laboratory'experinenm; Area: Resident Dispersal Juvenile Male 15 S JUvenile'Female 8 12 xzadj - 1.2532, df - 1, .5 > P > .1 Table 12. Chi square annalysis of the four-day total nesting area frequencies of juvenile mles during the laboratory experiment. Area: Resident Dispersal Adult Male 20 0 Nonlactating Fenale l9 1 Lactating Fenale 2 18 Control (No Adult) 15 5 XZadj - 12.2619, df = 3, .01 > P > .005 Table 13. Chi square analysis of the four-day total nesting area frequencies of juvenile feales during the laboratory ennperiuent. l _ Area: Resident Dispersal Adult Male 14 6 Nonlactating Femle 8 12 Iactatirng Fenle 5 15 Control (No Adult) 8 12 xzadj - 2.1714, df - 3, .9 > P > .5 67 reveals that juvenile males nested in the resident area less frequently when lactating fenles, canpared to adult males, nonlactatirg fennales, and the conupl conditian with no adult, were encountered. 'lhe propor- tion of juvenile fenles that nested within the resident area was similar for all adult classes (Table 13) . These results show that under laboratory conditions, the use of nestingareasbyjuvenileswasnotdependentnpontheirsen. I-bweve', juvenile mles tended to nest in resident areas occupied by nonlactating fenles unre frequently than did juvenile fenales at a borderline signi- ficannce level of alpha 2 .065. The use of nesting areas by juveniles was s tly dependent upon the sex of achnlts/reprochnctive stage of adult fenles for juvenile males, but not juvenile fenales. lactat— ing fenles, cepared to the other adult classes, repulsed juvenile mles. Figure 6 gives a sunnary of the four-day total frequencies of juvenile nesting areas. Nesting Arranger-ant. The propensity of juveniles to nest with adults was exedned to determirne the effects of the social treatments on a second aspect of juvenile settlenent under two environmental condi- tions. Daysonwhichjuvenilesnestedinthedispersalareaswerenot included in this analysis since it was of interest to examine whether differences in nnesting arrangenent due to social factors existed when juveniles nnested in areas occupied by adults. Inclusion of juveniles irrespective of their nnesting area would have distorted the findings. Also, the control juveniles were omitted fren this analysis. Aswas thecaseforthenestingareaueasure, thefour-day total proportion of juveniles that nested with adults was nnot jointly inde- pendent of location of the experiment, juvenile sen, and adult class Figure 6. Four-day total nesting area frequencies of juveniles (N = 20 juvenile days). The designations for adults (0" = male, 3 - nonlactatinng fenale, IAC 3 - lactating feele, and GN-cmntrolwithnoadult) willbewedforother figures in this paper. 3. LAB EXPERIMENT 0— JUVENULE MALE A = JUVENILE FEMALE cncn n FYDF‘RIMFNT A .= JUVENILE MALE 69 NO. IN DISPERSAL AREA m '— 2 lil E a: Ill O. X I3 ILI < 2 S .1 III “- d I.“ d ICC-I...III-IIOIUOIIIIOIIOCOCCU z .III'II-II'ICCIIIII'III...IIII.IIOIIIICIIIIICOIIIOOIIIIII-.IIIII I '1 I.I.I > D '5 4 II 0 <1 .A A I J I I l l l 1 L l L A. FIELD EXPERIMENT Co A A I 1 — vaav maelsau NI 'ON Figure 6. n5 0 no 0 F ADULT 7O ()(Zfld:I - 19.0299, df - 2, .005 > P). The percentages for this compari- son are given in Table 14. The effects of the social factors on nesting arrengenent were enamined by conducting separate enalyses of the four- day total frequencies obtairned in the field end laboratory enperinents. In the field experiment, the proportion of juveniles that nested with adults was not jointly independent of juvenile end adult social conditions (6 x 2 contingency table, .01 > P > .005, see Table 14 for the rev data). It was necessary to test for differences between juvenile mles end juvenile fenles separately for each adult class (2-way con- tingency tables) end to test for differences annng adult classes sepa- rately for juvenile mles end juvenile fenales (3 x 2 contingency tables). There were nno significant differences in the prOportionns of juve- nile males end juvenile femles that nested with adult males (Table 15) , but juvenile mles nested with nnonlactating fenales unre frequently then did juvenile fenles (.025 > P > .01, Table 16). The proportions of the juvenile sexes that nnested with lactating feeles were similar (Table 17); onne juvenile mle nested for three days with a lactating fenle that had abendoned her litter. Differences annng achnlt classes were mnsignificant for juvenile males (Table 18) end juvenile fenales (Table 19) . But juvenile fenles tended to nest are frequently with adult mles then with either nonlactating fenles or lactating fenales at a borderline significence of alpha 2 .07 (Table 19) . These results show that, under field enclosure conditions, the nnesting arrengeent of juveniles was significently dependent upon the sex of the juveniles, but only when nonlactating fenales are encoun- tered. Juvenile mles nnested with rnonnlactating fenales unre frequently .md Havel. Henri! ELVE‘C Law-«k In: :ruaafi 329 MC I...u..wn~u~fi.vU.HnXm Havana: 13;de HEQUJ \AfldlnuHSmw «Ufa. Kdewnv Quen— LO ENC-Hid (NH. mvNQmflH 71 6838.." mum «mum ”cavemen an... 5 Bung moafifign £0.33 so 936 3:0 8qu .86 ”.83me m5 5 when 3an no .133 .6633 u p .336 m5 £6688 03m 9.83%me ugfimmumfifieogmfiwu umfi agnmouwflufllw 6 o 6 8H 96 8 6 o 86 .e 6 2 38 agape 6 o 6 8H 6 8a a o 6 9: A98 e s8 6 o 6 8H 6 8H 6 o 6 8 6 8 m be 6 o 6 8H 6 ma 6 o 6 8H 6 8 N be 6 o 6 8H 6 fl 6 o 6 8 6 8 H s8 mam—Um mam—5m can! magnum mam—5..“ 6.32 36303 3680382 3:3 wfiufiufl wfiufiuqesz #52 mam—8m edge mam: «dug 86 o Gd o 86 S 86 2 ad 8 p86 8 H38 beéé 6 o 6 o 6 on 6 mu 6 8H 6 8 .6 ea 6 o 6 o 6 8 6 ma 6 8H 6 8 m a8 6 o 6 o 6 on 6 no. 6 on 6 8 N 65 6 o 6 o 6 8 6 o 6 8 w6 8 H .3 mam game was wine mine «as: 36803 wane—862 ”:52 wfiuflofl 9330382 #53 mam—Um «Ugh mam: «Hg pg 35 69.—”3e fig wound: udau weigh «5 mo $635039 :38 e595 H38 hmvnfifim use hide «5 mo 55 .3. MEN.“ 72 Table 15. Chi square enalysis of the four-day total nesting arrengeent frequencies of sale end fenale juveniles for adult unales during the field enqneriment. Nested With Did Not Nest Adult With Adult Juvenile Male 13 7 Juvenile Fenale 11 7 fiwjanme,e=1,9>P>,5 Note: Only days on which juveniles nested in the resident area are included. Table 16. Chi square enalysis of the four-day total nesting arrengeent frequencies of sale end fenale juveniles for nnonlactating fenales during tlne field enperinent Nested With Did Not Nest Adult With Adult Juvenile Male 15 3 Juvenile Feele 0 13 xzadj - 5.2474, df - 1, .025 > P > .01 Note: Only days on which juveniles nested in the resident area are irncluded. 73 Table 17. Chi square enalysis of the four-day total nesting arrengenent frequencies of mle end fenale juveniles for lactating fenales during the field experiment. Nested With Did Not Nest Adult With Adult Juvenile Male 3 13 Juvenile Fenale O 13 X2361 - 0.6796, df 1, .5 > P > .1 Note: Only days on which juveniles nested in the resident area are included. Table 18. Chi square enalysis of the four-day total nnesting arrengenent frequencies of juvenile unales for adults during the field ecperinent . Nested With * Did Not Nest Adult With Adult Adult Male 13 7 Nonlactating Fenale 15 3 lactating Fenale 3 13 xzadj - 3.7996, df :- 2, .5 > P > ,1 Note: Only days on which juveniles nested in the resident area are included. 74 Table 19. Chi square enalysis of the four-day total nesting arranngeent frequencies of juvenile feales for adults (lining the field enperiment. Achlt With Adult Adult Male 11 7 Nonlactating Fenale O 13 Lactating Fenale O 13 xzadj =- 5.2963, df - 2, P = .07 Note: iny days on which juveniles nnested in the resident area are included. Table 20. Chi square enalysis of the four-day total nesting arregeent frequencies of male and feele juveniles for adult mles during the laboratory eqneriment. ‘66—:— I: I Nested With Did Not Nest Adult With Adult Juvenile Male 14 6 Juvenile Fenale 12 2 X2363. =- 0.2826, df - 1, .9 > p > ,5 Note: Only days on which juveniles nnested in the resident area are inncluded. O‘- k— W. .- ‘9 fl- 75 then did juvenile fenales. The four-day total proportions of juvenile males end females that nested with adult males or lactating fenales were similar; juveniles tended to nest with adult sales, but tended not to nest with lactating feales. There was some evidence to indicate thatthenestingarrengeentis dependentnqnontheachlt class. The four-day total proportionns of juvenile males that nested with adult males or nonlactating fenales were such higher, though nnot significent, then the four-day total proportien of juvenile mles that nnested with lactating fenales . Juvenile fenales never nnested with nonlactating fenales or lactating fenales, but frequently nested with adult mles. This difference in four-day total proportions closely approacled signi- ficence (P = .07) . In the laboratory ecperiment, the proportion of juveniles that nested with adults was again not jointly independent of juvenile end adult social conditions (6 x 2 contingency table, .05 > P > .025, see Table 14 for the raw data). No significant differences were found between the proportions of juvenile mles end juvenile feales that nested with the adults when cepared separately for adult males (Table 20) end nonlactating feales (Table 21) . A chi square test for lactat- ingfealeswasjudgednot tobeneeningful; theenpected frequencies were ectrenely small sinnce juveniles selden nested in areas occupied by adults. On nneither of the two days that juvenile mles nested within enclosures containing lactating fenales nor on nnone of the five days that juvenile fenales nnested within enclosures conntaining lactating fenales did the juveniles nest with the adults. Differences emg adult males , nonlactating fenales , end lactating females when cepared 76 Table 21. Chi square analysis of the four-day total nesting arrengeent frequencies of male end feels juveniles for nonlactating fenales during the laboratory ecperiment Nested With Did Not Nest Adult With Adult Juvenile Male 16 3 Juvenile Fenale 8 0 xzfi-OfifildfaL.9>P>5 Note: Only days on which juveniles nnested in the resident area are included. Table 22. Chi square enalysis of the four-day total nesting arrengeent of juvenile mles for adults during the laboratory enperiment. Nested With Did Not Nest Adult With Adult Adult Male 14 6 Nonnlactating Fenale 16 3 lactating Fenale 0 2 xz-Len,e=2,5>P>.1 Note: Only days on which juveniles nested in tl'e resident area are included. 77 separately for juvenile mles (Table 22) end juvenile fenales (Table 23) were nonsignificant. These findings are in conflict with the results of the chi square test for overall 3-way independence. Inspection of Table 14 indicates that juveniles nested with adult sales and nonlactating fenles more frequently then with lactating fenales. waever, a 3 x 2 contingency table using pooled data showed that signnificence was onnly borderline at alpha 2 .06 (Table 24) . This probability statenent must be viewed with cautionbecausethetestms suggestedupminspectionofthedata. The failure to detect significent differences in the pooled data my be due tothesnnllmnberofjuveniles thatreeinedintheresidentareafor acne treatment groups, whichmayhavebeencanpoundedby theoverly con- servative adjusted total chi square value. The four-day total nesting e-rengenent frequencies are given in Figure 7. These results show that, in the laboratory enperinent, the nesting arrengeent of juveniles end adults was not dependent upm tlne sen of juveniles. Both juvenile males end feales tended to nest with adult uales or nonlactating fenles, but juveniles nnever nested with lactating fenales. This difference in the adult classes was very close to being significent. TimeintheResidentArea. Theenuntoftine, enpressedasthe percentoftheactiveperiod, juveniles spentintlneresidentareasof the field end laboratory enclosures was enamined to determinne the effects of the social factors on a third aspect of juvenile settleent. Significantly nore juveniles reained in the resident areas during the activepeniodinthefieldeqnerimentcanparedtothelaboratory 78 Table 23. Chi square enalysis of the four-day total nesting arrangenent frequencies of juvenile fenales for adults during the laboratory enperinent. Nested With Did Not Nest Adult With Adult Adult Male 12 2 Nonlactating Female 8 O lactating Feale 0 5 xzadj - 4.5184, df =- 2, .5 > P > .1 Note: (Mly days on which juveniles nnested in the resident area are uded. Table 24. Chi square enalysis of the four-day total nesting arrengenent frequencies of pooled juvenile males end feales for adults during the laboratory enperinnent. Adult Male 26 8 Nonlactating Fenale 24 3 lactating Fenle 0 7 2 - - .. x adj 5.7170, df 2, P - .06 Note: Only days on which juveniles nested in tlne resident area are included. Figure 7. 79 Four—day total nesting arregenent frequencies . Only days on which juveniles nested in the resident areas are sham. . =JUVENILE MALE 0 =JUVEN|LE MALE A =JUVEN|LE FEMALE 80 NO. NOT NESTING WITH ADULT ”a '0 O 31 I I I I r T I W— I I I I T I I 1 I I I B. LAB EXPERIMENT DIG....II'IIIII...IIII...-I.III-CIIIIIOIIII.IIII...III-IIIIOIOIIO......OO.IIIIIIIIIII-IIOIIIOOIIIII.‘ fl‘// ’ V /’,// " / I: " ' f xv" 6‘ ///’ / / -. / ,, I/x / ,. <( )4- WI .44; / /,1/ / /. ....I A. FIELD EXPERIMENT q- 4JJIJIILELILILLLJLI f T a é '0 _ .L'II'IGV HllM |.-')NI.I.SSN 'ON 0.- Figure 7. ADULT Jam 518m] byrm 81 enperinent (.005 > P, Table 25). This probability statment nust be viewedwithcautionbecause the testwas snggestednqnmenaflnationof the data. 'The signnificent difference above can be attributed to the environmental treatments for two reascns. Firstly, all but one of the juveniles in the field enperiment used the apparatus (passage end feeder tunnels) dwingthepretestperiodpriortotheplacenentofthe restraining grids (Table 26). Secondly, all juveniles were required to go throngh the holes in a section of 1.3-cm-mesh hardware cloth prior to the eqnerinents end usually for at least one week following the enperinents in order to secure food end water. Therefore, it was assured that all juveniles were capable of squeezing through the holes. The data obtained frcm the field experinent were enalyzed with nonparametric statistics (Table 27); the fair-night group ueens are given in Table 28 . No significent differences were present in either theuaineffects ofjuvenilesenoradult class, endthejuvenile sennn adult class interaction was also nonsignificant (Table 27) . Therefore, unnder field enclosure connditions, the four-night meen per cent tine juveniles spent intheresidentareadnn'ingtheactiveperiodwasnot significantly influenced by the sen of juveniles or the sen of adults/ reproductive stage of adult feales. I-bwever, there was a sliglnt ten- dency for juvenile fenales to spend less time within enclosures occupied by nnonlactating or lactating fenale coupared to enclosures containning adultmalesorthecantrol connditionofnoadult. Nosuchtendencywas apparent for juvenile males. The back trensforned four-night end nightly group meens of the time juveniles spent in the resident area during the active period in the laboratory enperiment are given in Table 28 . These treasures were Tabl E /§§" 82 Table 25. Chi square analysis of the proportions of juveniles thatreminedintheresidentareaenringallfour nights of the enperinnental test period. Remined for 1007. Did Not Renain for of the Tine 1007. of the Time Field 26 14 Tab 4 36 x2 - 25.8133, df =- 1, .005 > P Table 26. Sumnery of the passage tunnel end/or feeder tennelusebyjuvenilespriortotheplacenent of restraining grids during the pretest period. Used Did Not Use N Apparatus Apparatus Field 40 39 1 lab 40 40 0 Table 27 . Nonparenetric enalyses of the four-night ueen per cent tinnejuvenileswereintheresidentareadxingthe field eqneriment. Connparison Test Wn or HID Value Prob. Juvenile mzfi wn - 25 P > .1 Adult $31,268,: H/D 1.6232 .9 > P > .5 mggmt fifilgm H/D - 0.2666 .975 > P > .9 83 Add Nmav Engaging .moumuagmoufiufigugmoxooaufiemumfisgnmogfllw .guggouflfielg. 8.3-8.3 8.898 8.898 8.8-96 68.96 . 88-98 8.8-9.6 88-96 98 98 98 98 98 H88 98 98 e .88 688-98 8.898 8.898 8.8-98 8.8-3.8 68.98 8.8-96 8.83.6 98 9H 98 H8 98 898 98 98 m 88 A8.Hm-.-V Ae.~8-o.88 au.ae-o.ov A8.mm-o.mm8 A8.88-m.8mv Ao.ea-o.ov A8.88-m.omv Am.8m-o.HmV 98 98. . 98 98 98 3.8 «.8 93 .e. .88 68-96 8.82-96 8.8-98 69.-96 88-9va 8.8-96 8.82-96 8.8-96 9.: 9E 98 98 98 98 98 98 H 88 8.3-96 8.8-6.8 8.8-98 8.896 6896 8.8-8.8 8.8-9.6 68-96 a 93 98 9mm 98 98 93 98 898 snap-Ham pea—583 boom-Hons 6 6 6 n6 6 6 6 86 m 98 9: 9: 9H8 98 98 98 98 938-868 685888 388 838 68 36.8 838.8 8H8: 882 68 wine «Hana 8H5 8H5? H9588 @3303 86836382 H6888 @5363 858883 mam—uh 6% 3d: «H.658 |I| Annaggflgflfignofigmouufifiugflgnmfiuuggg .83“an 84 evaluated by analysis of variance after the individual scores for each replication were transformd to the arcsin scale. The tables and figm'es slowing the differences anong means give the transformed values. The four-night man min effect of juvenile sex was not signi- ficant, but the mean main effect of adult class had a significant F- value (.05 > P > .025, Table 29). Inspection of Table 30 reveals that juveniles spent less time in the presence of lactating fenales coupared to the remaining adult classes. Statistical exandnatim of the adult class main effect was not conducted because the juvenile sex x adult class interaction was also significant (.025 > P > .01, Table 29) . Analysis of the siuple main effects indicated that the adult class main effect was significant for juvenile males (.005 > P > .001, Table 29), but not for juvenile farales. A Student—Neman—Kenls _a_ posteriori test showed that juvenile males spent a smaller percentage of the active period in the residait area with lactating females conpared to adult mles, mtflactating females, and the control condition with no adult (.01< P, Figure 8). Also, juvenile males spent mre tine with non- lactating famles than did juvenile famles (P 2: .005, Table 29). These mans are graphed in Figure 9. In conclusion for the laboratory experimt, the aununt of time juvmiles spent in the resident area during the active period was influ- enced by both the sex of juveniles and the sex of adilts/reprodmtive stage of adult fenales , but the significant statistical interaction indicated that the effects were not consistent for either the juvenile or adult social factors. In re3pect to the juvenile sexes, mles spent significantly more time with nonlactating famles than did the juvenile females, but males and females spent similar annunts of tine with adult 85 Table 29. Analysis of variance of the four-night mean per cent time juveniles spent in the resident area during the laboratory aqaeriment. Source of Variation df NBS F-value Prob . Juvmile 1 735.3920 1.9228 .25 > P > .1 Adult 3 1,218.0712 3.1849 .05 > P > .025 Juvenile x Adult 3 1,499.7718 3.9214 .025 > P > .01 Error 32 382.4571 Total 39 (SS = 21,127.5486) Simple Main Effects of Juvenile For: Adult Male 1 108.9000 0.2849 .75 > P > .5 Nonlactating Fenale 1 3.513.7503 9.1873 P = .005 lactating Female 1 1,228.5506 3.2123 .1 > P > .05 No Adult 1 383.1610 1.0018 .5 > P > .25 Sinple Main Effects of Adult For: Jimmile Male 3 2,335.7062 6.1071 .005 > P > .001 Juvenile Female 3 382.2766 1.0000 .5 > P > .25 Error 32 382 . 4571 Cu ml = 86 Table 30. Mean per cent time juveniles spent in the resident area for each of the adult classes during the laboratory experinent (N = 10). *Four-_-night lst *2nd *3rd *4th Adult x Night Night Night Night Male 55.4 51.7 56.2 59.4 51.5 Nonlactating Female 49.5 54.0 48.5 47.7 47.2 Lactating Female 30.5 48.3 24.0 19.8 19.2 Gontr wi NoAdnonlt th 49.3 55.4 57.3 45.3 43.4 * = significant difference among mans. 87 Figure 8. The four-night group means for the per cent tine juvenile males spent in the resident area (lining the laboratory exper iment. The mean marked by the asterisk (*) is signi- ficantly different from the other means at the .01 level (N = 5 . 88 at- LAC Q CON ADULT 0+ m m . . m e m 7 a w m a 2 3_. 50.me m>.._.0< “5* zD_.. .00—mm."— m>_._.0< "—0 R Z P). These means are given in Table 30, and it can be seen that juveniles spent less tine with lactating fenales, ccupared with the renaining adult classes, (hiring the last three nights. Statis- tical examination of the nightly adult class main effects was not con- ducted because each of the nightly juvenile sex x adult class inter- actions were also significant (.05 > P). Analysis of the sinple main effects showed that differences due to the adult classes were significant for juvenile mles, but not juvenile fenales, on the last three nights. An 3 posteriori Student-Nem- Keuls test showed that on these nights juvenile mles spent less time in the resident area with lactating fenales canpared to the time spent with Table 31. Analysis of variance of the per cent tine juveniles spent 92 intheresidentareaonthefirstnightdrringthe laboratoryexperiment. Source of Variation df PBS F-value Prob . Juvenile 1 88.2981 0.1966 .75 > P > .50 Adult 3 98.6867 0.2197 P > .75 Juvenile xAdult 3 1,547.0618 3.4445 .05 > P > .025 Error 32 449.1407 Total 39 (SS = 19,398.0445) seen: 3.... Adult Male 1 1.6000 0.0036 P > .75 Ncnlactating Fenale 1 3,727.2164 8.2985 .01 > P > .005 Lactating Fenale 1 725.0523 1.6143 .25 > P > .10 No Adult 1 275.6250 0.6137 .5 > P > .25 Simple Main Effects of Adult For: Juvenile Male 3 992.3585 2.2095 .25 > P > .10 Juvenile Fenale 3 653 . 5388 1.4551 P = . 25 Error 32 449.1407 HI- Table 32 . 93 intheresidentareaonthesecondnightduringthe laboratory experiment . Analysis of variance of the per cent time juveniles spent Source of Variation df MSS F—value Prob . Juvenile 1 557.9343 1.3059 .5 > P > .25 Adult 3 2,401.5095 5.6210 .005 > P > .001 Juvenile x Adult 3 1,819.5864 4.2590 .025 > P > .01 Error 32 427.2366 Total 39 (SS = 26,892.7915) Simple Main Effects of Juvenile For: Adult Male 1 106.2760 0.2488 .75 > P > .5 Nonlactating Fenale 1 3,950.1563 9.2458 .005 > P > .001 lactating Fenale 1 1,718.7210 4.0229 1 > P = .05* No Adult 1 241.5723 0.5654 5 > P > .25 Simple Main Effects of Adult For: Juvenile Male 3 3,505.2879 8.2046 .001 > P Juvenile Fenale 3 715.7218 1.6752 .25 > P > .10 Error 32 427.2366 * Indicates borderline significance where the calculated value was within 0.20 units of the critical value. Table 33 . Analysis of variance of the per cent time juveniles spent intheresidentareaonthethirdnightduringthe laboratory experiment. Source of Variation df NBS F-value Prob . Juvenile 1 614.0290 1.0776 5 > P > .25 Adult 3 2,783.5504 4.8850 .01 > P > .005 Juvenile x Adult 3 1,975.7248 3.4673 .05 > P > .025 Error 32 569.8143 Total 39 (SS = 33,125.9108) Siuple Main Effects of Juvenile For: Adult Male 1 103.4266 0.1815 .75 > P > .50 Nonlactating Fenale 1 3,770.5872 6.6172 .025 > P > .01 lactating Fenale 1 2,099.8908 3.6852 1 > P > .05 No Adult 1 567.0090 0.9951 .5 > P > .25 Sinple Main Effects of Adult For: Juvenile Male 3 4,035.6675 7.0824 .001 > P Juvenile Fenale 3 723.6400 1.2700 .5 > P > .25 Error 32 569.8143 95 Table 34. Analysis of variance of the per cent time juveniles spent in the resident area on the fourth night during the laboratory experiment. Source of Variation df PBS F-value Prob . Juvenile 1 834.2082 1.7865 .25 > P > .10 Adult 3 2,086.6736 4.4686 .025 > P > .01 Juvenile x Adult 3 1,994.5878 4.2714 .025 > P > .01 Error 32 466.9589 Total 39 (SS = 28,020. 6784) Sinple Main Effects of Juvenile For: Adult Male 1 6.8724 0.0147 P > .75 Nonlactating Fenale 1 3,275.3760 7.0143 .025 > P > .01 lactating Fenale 1 1,855.5888 3.9738 .1 > P > .05 No Adult 1 1,680.1344 3.5980 .1 > P > .05 Sinple Main Effects of Adult For: Juvenile Male 3 3,575.7644 7.6576 .001 > P Juvenile Fenale 3 505.4971 1.0825 .5 > P > .25 Error 32 466 . 9589 ,lll’lllllnlllll ni-nl flllfl i ll- 96 adult nnales, nonlactating fenales, and the control condition with no adult (.01 > P). These differences are graphed in Figure 10, and the results for the first night are given for ccuparison. Juvenile mles, coupared with juvenile fenales, spent nore time with rnonlactating fenales on all four nights (.025 > P, Figure 11). On the second night, juvenile fenales spent more time in tlne resident area with lactating fenales than did juvenile males; this difference closely approached significance (Table 32). These data, and the canparisons for the retaining nights, are given in Figure 12. It is apparent that, subsequent to the first night, the significant differences due to the main effect of the adult classes were consistent with the analysis of the overall four-night means. The difference between the juvenile sexes for nonlactating fules was consistent on all four nights. Finally, the effects of the social factors in general were consistent over nights since only one couparison, juvenile sex for lactating fenales, showed a difference (but not significant) which was not found in the analysis of the four-night mean values. In sumary, the analyses of the three settlenent measures obtained in the field and laboratory experiments indicate that both the sex of juveniles and the sex of adults/reproductive stage of adult fenales may affect the outcome of juvenile-adult social interactions. Therefore, both juvenile and adult social factors any influence settlenent. Inspection of Table 35 reveals two major points concerning the effects of social factors. Firstly, the outcanes of juvenile-adult social interactions were not consistent in the field and laboratory experi- ments. This is indicated by the failure to detect similar patterns of significant differences anong social factors in tlne analyses of the 97 Figure 10. Nightly mans of the per cent tim juvenile males spent in the resident area during the laboratory ecperinent. The asterisk (*) indicates a significantly different man at the .01 level (N a 5). _— E __. H -— fl __ NIGHT 2 NIGHT 4 98 NIGHT 1 NIGHT 3 ..=. _DOEwn— m>_._.0< "_O X z P) in the tlycannpariscns ofthemanper cent timjuvenile esandfemles spentintheresidentareasoccupiedby nonlactating fenales during the laboratory experinent (N = 5 for each night). N 9' 3183 .5 ‘1' $35? a. 9‘ I?" «‘3 TIME [MEAN % OF ACTIVE PERIOD] JUVENILES WERE IN RESIDENT AREA 1'1" Figure 11 , 100 NON LACTATING FEMALE MALE NIGHT 1. = NIGHT 2 NIGHT 3% NIGHT 4 FEMALE FEMALE JUVENILEM 101 Figure 12. Nightlycanparisansof themanpercent tim juvenile mlesandfemles spentintheresidentareasoccupiedby lactating females during the laboratory experiment (N - 5). $8 5 i? 13’? 3851583 3 TIME [MEAN % OF ACTIVE PERIOD] JUVENILES WERE IN RESIDENT AREA 32‘ E! H N MALE 102 LACTATING FEMALE NIGHT 1 n U.PIC-I'll.IUIIUIIIDICICUIICIIIIDID-CDIDIIIIIIIIIIIDODIDOIIIIIOI NIGHT 3 .‘C.’.'..IIIID.IIDOIIIIDOIIIIII-DOD...-IIIIIUIOIIIIIIIIIIIIIIIII-IUC FEMALE MALE JUVENILE NIGHT 2 NIGHT 4 Y FEMALE 103 Table 35. Snmnary of the analyses of the three settlenent variables. Main Effect Interaction Settlement Dependent Variables Juvenile Adult Sex Class Field Experiment Nesting location" - - - Nesting arrangenent* o o + Per cent tim in _ _ - Resident area (a) laboratory quneriuent Nesting location* 0 o + Nesting arrangeIent* o o 1- Per cent titre in _ + + Resident area (b) ) nonmaranetric test. (b) - analysis of variance. + 8 significant social factor main or interaction effect. - = nonsignificant. o = test was not mde because statistical interaction was present. rEpj 104 field and laboratory data. Secondly, the significant statistical inter- actions between juvenile sexes and adult classes show that, in these cases, juvenile sexes differed in their response towards the various adult classes. A summry of the significant statistical interactions is given in Table 36. The sec of juveniles significantly affected the settlenent measures only when nonlactating fenales were encountered. Juvenile fenales, canpared with juvenile males, nested with nnonlactating fenales less frequently in the field experiment and spent less time within the enclo- sures occupied by nonlactating fenales in tlne laboratory experiment. Also, in respect to the juvenile sexes, juvenile males tended to nest within enclosures occupied by nnonlactating fenales in the laboratory (P = .065). The sex of adults/reproductive stage of adult fenales sig- nificantly affected the settlenent treasures only for juvenile mles during the laboratory experiment. These juveniles nested less fre- quently ad spent less time within enclosures containing lactating fenales ccupared to those containing adult males, nonlactating fenales, or the control condition with no adult. There was sane evidence to indicate that juvenile fenales nested more frequently with adult males than with nonlactating fenales or lactatinng fenales in the field experi- ment (P = .07). Also, both juvenile sexes tended to nest with lactating females less frequently than with adult males or nonlactating fenales in the laboratory (P = .06). On the basis of these results, it can be conncluded that adult fenales may limit the settlenent of juveniles in P. m. I_n_a_i_r_dl. However, the limitationis depedentuponthe sexofjuveniles aduponthe reproductive stage of adult fenales. Compared with juvenile males, 105 :8 AmA no. 98:3 .7353 madam moi-.833 .- e dams and mm: ammo u o §:mfiuwwm§n u dune-Swan Hagan @8398 3wa u n 58me 59a wage none .33me n doom-8mm? ugwgm gm 033 n m 385m dam made «8 a o o «3.9% 828.11 mg a» - 398m mag be + - - mNH + mama mad-05H. "How mmflo “.3 c3388 8888 mamfimm 955303 38 + - m3 + 8 a flame mfiufiqflfia man: ”.3 “How 83 3882:- 88383 bougomfl Bean 88883 8.2 “838m 5 may B88855 8382 23983 .9882 name 8m :82 flag-“Boa gang u§\3§> ”.6988 . egg ugfluumm ooh—fl 9.3 How moan-mac udfim dam 880m 3an 5953 mdoHuumhaufi Queumwumum agave-Ewan 6:... mo $9393 9? mo baa—5m .om 3an 106 juvenile females may be limited by nonlactating fenales. Compared with adult mles and rnonlactating fenales, lactating females may limit the settlenent of juvenile males . Also, nonlactating and lactating fenales , canpared with adult males, may limit the settlenent of juvenile females, but the evidence is not conclusive. The present study provided no evi- dence to indicate that adult males limit the settlenent of juvenile mles or juvenile females to a greater degree than either nonlactating or lactating fenales. Finally, the physical envirmment appeared to be an inportant factor in determining the effects of social factors on the juvenile settlenent treasures . Activity Between the Resident and Mspgsal Areas. The juvenile activity in the passage tunnel, enanessed as the amber of passage tun- nel trips per hour of the active period, was examined in order to deter- mine if social conditions influenced the rate at which juveniles entered and left the resident area. As was pointed out previously, a smaller prOportion of juveniles used the passage tunnels in the field experi- ment coupared to the laboratory experinent (.005 > P, Table 25). The data obtained from the field encperinent were analyzed with non- parametric statistics (Table 37); the four-night group mans are given in Table 38. No significant differences were found for the main effects of juvenile sex or adult class, and the juvenile sex x adult class interaction was mnsignificant as well (Table 37) . The results above indicate that the social conditions did not affect the rate of juvenile activity between areas in the field-enclosure experiment. This finding is consistent with the failure to detect differences anmng treatment groups for the annunt of time juveniles spent in the resident area. 107 Table 37 . Nenparanetric analysis of the four-night mean juvenile activity rate in the passage tunnel during the field ecpe'i‘nent. Canparisan Test Wn or H/D Value Prob. Juveni] Wilcoxin's e rank-sun test wnn = 25 P > .1 Kruskal- A‘mt Wallis test H/D - 1.6224 .9 > p > ,5 Juvenile 1: Adult Kruskal— Interaction Wallis test an) ' 0-5812 .9 > P > .5 The back transformd four-night and nightly group means of the juvenile activity rate obtained in the laboratory exper-imnt are given in Table 38. These treasureswere evaluated by analysis of variance after the individual raw scores for each replication were transformed to the/W172 scale. The tables and figures showing the differences among means give the transformed values. The four-night mean min effect of achlt class had a significant F—value (.05 > P > .025), but the juvenile sexuain effect and the juvenile sex it adult class interaction were nonsignificannt (Table 39). The adult treatment group means are given in Table 40. Juvenile acti- vity was significantly higher when juveniles encountered adult males canpared to lactating fenales (Student-Nem—Kenls test, .05 > P, Table 41) . None of the other camarisons was significantly different. Figure 13 shows the mgnitudinal relationships mg the four-night adult treatment means. The nightly activity rates were evaluated with separate analyses of variance to determirne if the effects of social factors were 108 Add .88 mg 388mg £03 I a Amer—.5 «managfiuacguflflmoqgnwogfllu .35 a? 3 «a a I 2x. 89573.8 Acme-8.8 2.3-9.8 85.3320 A8.¢-m.~.8 8868.8 Sod-3.8 8364-888 8.x «fin 3.8 3.5 86 8.3 No...” 86 .V E 85678.8 883-888 8938.8 anew-No.8 883888 8908.8 88513.8 883-3 .8 36 856 min 86 8..» $2” mad 5N6 m b8 888-320 8~.mN-8..8 God-$.88 88.8-38.8 88348.8 , 8868.8 85.3.3.8 88.3.8.8 8.8 55.5 .36 3:3 no.8 5N3 5N.m 8.5 N 58 88385.8 “3.3-8.8 8~.3-8.8 85688.8 85673.8 Gad-8.8 856-398 8.8-8.8 No.8 86 «5.0 8.3 .36 5N3 5n.n 3.3 H 5 383-898 Qumran-8.8 88313.8 8.18-8de Sada-8.8 cane-8.8 838-888 888-338 m 8.5 3.m m5..~ 8.3 59.» 8.3 36 pend flan-.50..” ugh—mg 5933 8 3 3 *8 8 8 8 a8 « nmd 3.0 8.0 3.0 3.3 mad $8 3.0 unwed-noon cg 33m 332 9e mass... «asap «3. 332 as wise mine 03.. adds. 39.8 5323 33383 H880 95393 wfifiumusz 3.8mm mg «a: 3.8.2. .An-eeqagflfighhghanfibghaamohfiamn .833. 109 Table 39. Analysis of variance of the four-night man juvenile activity rate in the passage tun-e1 during the laboratory experiment. I Source of Variation df PBS F-value ' Prob . Juvenile 1 3.5367 3.1722 .1 > P > .05 Adult 3 3.3633 3.0167 .05 > P > .025 Juvenile 2: Adult Interaction 3 0.2171 0.1948 P > .75 Error 32 1. 1149 '1me 39 (SS- 35.6771) consistent over nights (first night, Table 42, second night, Table 43; third night, Table 44; and famth night, Table 45). There was a simi- ficant adult class main effect on the first night only (.01 > P > .005, Table 42) . The juvenile sex main effect and juvenile sex 1: adult class interaction were never significant. An a posteriori Student-Nam- Keuls testontheachnlt treatna1tnaansforthefirstnightslnwedthat juveniles unved more frequently between the resident and dispersal areas whm adult males were encountered carpeted to not-lactating fenales, the control condition (.05 > P), and lactating fenales (.01 > P, Table 46). The latter three group means were not significantly different. Figure 14 shows the mgnitudinal relationships among the achlt treammt means for the first test night. The results show that the significant differences sung the four- night mean activity rates can be attributed to the significant differ- encesmnngtheachllt classesonthefirst testnight. 'Iherewere significant differences amt-g treamaIt groups for the ammt of time 110 dame mo 38w m wag umfifl 805.8%me “€8meme umsu @38ng .4 N: H 8H 35 H 3N mic H H: Rd H m: de H SH flag-Maw 3.0 H $4 #5 H 34 9:0 H o: 3.0 H 84 2.0 H a: 3%”:qu 8.0 H 8H 8.0 H a; 3.0 H 8H Hg H Rd 3.0 H SH wfifiwmmfim 8.0 H o: 3.0 H :3 mg H 35 3.5 H 85 3.0 H m: 32 £sz Hafiz Hafiz “Ema « H32 fiH _ Em EN Hm? Hfiéésfi .SH n E H.363 houmuofia m5 wfiufi 8330 58m mfi .Ho 68 How GEE mwmmma H5 5 32 .0338 «HHBEH 69m H HV :8: .3 ~33 111 Table 41. Student-Naman-Keuls test for differences among the four-night adult class means of juvenile activty rate in the passage tunnel during the laboratory experiment. Adult Four-Night Mean 1 l S.E. 4: Male 3.19 i- 0.40 Control 2.57 i 0.32 * Nonlactating Female 2.11 t 0.25 lactating Female 1.87 i 0.35 *Any two means (ilS.E.) not joined by the sauna vertical line are significantly different at .05 > P (N = 10). 112 Figure 13. Four-night adult group means for the juvmile activity rate in the passage tunnel during the laboratory 1- mt. Thevertical lines indicate i 1 S.E. (N= 10 . 113 T 3.6j r k 3 D T I“? IF_! Ira, Jt inwvuaau... aanmflzmw OEmn. m>_._.0< .m:\wn=m._. .02 z._._>_._.0< mn=zm>_.=. 1.6“ CON LAC ADULT Q FigurelB. 114 Table 42. Analysis of variance of the juvenile activity rate in the passage tunnel on.the first night during the laboratory experimt . _-: Source of Variation df PBS F-value Prob. Juvenile 1 2.4572 1.6052 .25 > P > .1 Adult 3 7.1087 4.6439 .01 > P > .005 Juvenile x Adult Interaction 3 0.8657 0.5655 .75 > P > .5 Error 32 1.5308 Total 39 (SS = 75.3647) Table 43. Analysis of variance of the juvenile activity rate in thepassagetmmlmthesecondnightdmfigthe laboratory experiment . _l Source of Variation df PBS F-value Prob. Juvenile 1 6.2110 3.8585 .1 > P > .05 Adult 3 2.7589 1.7139 .25 > P > .1 Juvenile x Adult Interaction 3 0.8108 0.5037 .75 > P > .5 Error 32 1.6097 Total 39 (SS - 68.4305) 115 Table 44. .Analysis of variance of the juvenile activity rate in the passage tunnel on the third night during the laboratory experiment. Source of variation df MES Favalue Prdb. Juvenile 1 3.3126 2.2781 .25 > P > .1 Adult 3 1.7124 1.1776 .5 > P > .25 fit; “8:51;; “Mt 3 0.4986 0.3429 P > .75 Error > 32 1.4541 Tbtal 39 (SS = 56.4776) Table 45. Analysis of variance of the juvenile activity rate in the passage tunnel on the fourth night during the laboratory experiment. Source of variation df ‘MSS Favalue PrOb. Juvenile 1 2.7794 2.8971 P = .1 Adult 3 2.3943 2.4957 .1 > P > .05 Juvenile ngdult Interaction 3 0.3231 0.3367 P > .75 Error 32 0.9594 Tbtal 39 (SS = 41.6317) 116 Table 46. Studaxt-Neunan-Kmls test for differences anong the first night adult class mans of juvenile activity rate in the passage tmnel (hiring the laboratory experimt. Adult man (:1 S.E.) mle 3.87 1: 0.50 + Caitrol 2.55 .t 0.31 * + Nonlactating Fenale 2.33 1 0.24 lactating Fenale 1.93 t 0.44 Note: Any two meme (:1 S.E.) not joined by the sane vertical line are significantly differ- entat .05>Pkor .01>P+. 117 Figure 14. First night adult group means of the juvenile activity rateintl'nepassagetunelmringthelaboratoryecperi- ment. 'meverticallmesindicateils.E. (NalO. 118 NIGHT 1 CON H n... w H __ _. "fir r T _ J. _ “Hawwmuumnmummwwm _ z‘ 09me w>_._.0< .m:\wn=m._. .0 252. m._.._._>_._.0< m..=zm>3... Figure 14 . ADULT 119 juveniles spent within the resident area during the last three test nights (see Tables 32, 33, and 34). The inconsistency of these results indicates thateventl'mnghthemmtoftinejuveniles spent inthe resident area was affected by social conditions, the covenant rate of juveniles in and out of the resident area was not affected on the last three nights. Therefore, when social conditions led to a relatively exclusive or mnemlusive use of a confined space, based on a tine unea- sure, thepropensityofjuveniles toenterandleave that spacewasnot affected. JuvenileWCtmge Amltnaleg.g.austerushavebeenreportedtoretardt1'egrowth rates of juveniles under free-ranging and laboratory conditions (Healey, 1967). Inanattenpttoreezanineandenpamimonthisreportedinflu- ence,theweightdnangesofjmaenilesinthepresentecperineitwere examined. No distinctionwasmadebetween juveniles that always renainedintl'eresidentareaendthosethatspentsanetiminthedis- persal area, since dispersing juveniles in free-rengingpopulationsmy raminnearresidentadiltsorleavetheresidents'l'mermges. A snmneryofthejuvenileweightdmgegrmpneensisgiveninTableM. A2x2x4factorialana1ysisofvariancewasperforuedmthe migltchmgedataoadninedfranthefieldendlaboratoryeqaeriment (Table 48). Neitherthejuvenilesecnorachnlt classmineffectswere significant, hmjuvenileshadhighergrowthratesinthefield experi- mntthaninthelaboratoryewerimnt(.01>?>.005,seeFignn:e15). Nene of the interaction F-values involving the locatian min effect ms significant, however, the juvenile sex 1: adult class interaction was 120 Table 47. Group means (i l S.E.) of the weight change of juveniles. Negative values indicate a mean loss of weight within a group (N = 5). Field Experiment Laboratory Pbcperimnt Juvenile Juvenile Juvenile Juvenile Male Fenale Male Fenale Adult Male 2.00 i 0.66 0.02 .+. 0.22 0.04 .+. 0.22 —0.76 i 0.57 mating 0.20 i 0.58 0.66 i 0.41 0.30 1 0.41 0.62 i 0.34 mm 0.98 n 0.54 0.26 i 0.38 0.74 i 0.72 -0.12 i 0.52 Cmtml 0.14 i 0.19 0.66 1 0.65 -0.50 a: 0.32 -O.48 1 0.29 (No Adult) significant (P 2 .025). Table 49 shows the interaction group mans. 'D'nese results suggest that the physical environment of the laboratory experiment (gg. , confined and sterile space, disturbance by the experi- menter, etc.) may have led to a reduction in the growth rates of juve- niles. But the effects of the experimtal social factors appeared to be consistent between the field and laboratory experinents. Analysis of the simle main effects of the juvenile sex x adult Class interaction stowed that the sex of juveniles was important only when adult males are encountered. Juvenile males had higher growth rates when interacting with adult males than did juvenile fenales (.005 > P > .001, Table 48). This difference is illustrated in Figure 16. Also, there was a significant F-value in the couparison of adult classes for juvenile males (.05 > P > .025, Table 48). Juvenile mles had Table 48 . Analysis 121 of variance of juvenile weight change. Source of Variation df PBS F-value Prob . Location 1 8.0645 7.3476 .01 > P > .005 Juvenile 1 2.8880 2.6313 .25 > P > .10 Adult 3 1.1192 1.0197 .5 > P > .25 location x Juvenile 2 0.0500 0.0456 P > .75 location x Adult 3 1.9218 1.7510 .25 > P > .10 Juvenile x Adult 3 3.6727 3.3462 P 2' .025 m? x Ju‘mfle 3 0.6840 0.6232 .75 > p > .5 Error 64 1.0976 Total 79 (SS = 103. 44) Sinple Main Effects of Juvenile For: Adult Male ' 1 9.6605 8.8015 .005 > p > .001 Nonlactating Fenale 1 0.7605 0.6929 .5 > P > .25 Lactating Fenale 1 3.1205 2.8430 .1 > P > .05 No Adult 1 0.3645 0.3321 .75 > P > .5 Simle Main Effects of Adult For: Juvenile Male 3 3.0809 2.8070 .05 > P > .025 Juvenile Fenale 3 1.7109 1.5588 .25 > P > .10 Error 64 1.0976 g 122 Figure 15. Graph of the significant difference between the man weight changes for juveniles used inn the field experiment ad the laboratory experiment. The vertical lines indicate i 1 S.E. (N = 40). 123 m. :2 .H. ”—71 a a 4 a .. 000000 0.. 35.5 5.2m; m...zm._>:_. EXPERIMENT Fignme 15 . 124 Table 49. $111me of the juvenile sex x adult class interaction group mans (i 1 S.E.) of the weight change of juveniles (N = 10). Negative values irndicate a man loss of weight within a group. Adult Juvenile Halea Juvenile Fenale Male 1.02 2‘. 0.46 * -0.37 i 0.32 Nonlactating Fenale 0.47 .+. 0.31 0.64 t 0.25 lactating Fenale 0.86 1 0.43 0.07 i 0.31 Control With No Adult -0.18 .+. 0.20 0.09 i 0.38 * = significant difference between juvenile sexes. a = significant difference amng adult classes. Table 50. Stndent-Neunan-Kenls test among adult class mans (1 1 S.E.) for the weight change of juvenile males (N = 10). Negative values indicate a man loss of weight with a group. Juvenile Tested With 7': Adult Male 1.02 1 0.46 9: lactating Fenale 0.86 i 0.43 Nonlactating Fenale 0.47 i 0.31 Control With No Adult -0.18 .t 0.20 *mansnot connectedby the sanevertical line are significantly different at .05 > P. 125 Figure 16. Graph of the significant difference between the man might changes for juvenile mles and juvenile fenales tested with adult mles. The vertical lines indicate it 1 S.E. (N =- 10). 126 ADULT MALE 151' 1 .. 1 2.. to III oIIOCIOOOQOO il'i’lbl’ll QIOICCIdodI-n‘l oloolo I. ’I’I’bF’IDI’DDDD O IIOOIQOOO .bIIIIPUP IIIDFI Pill JUVENILE Figure 16. 127 higher growth rates when adult males were encountered canpared to the control condition of nno adult (StLdent-Nemmn-Kenls test, .05 > P, Table 50) . Figure 17 slows the differences in nagnitude of these mans. In conclusion, the growth rate of juveniles was significantly influenced by their sex only when adult males were encountered. In this case, juvenile fenales gained less might than did juvenile males. There were nno significant differences in the effect of tine adult social factors for either mle or fenale juveniles. However, juvenile male growth rates my be inhibited by the absence of conspecifics since con- trol juvenile mles gained significantly less might than the juvenile tales that interacted with adult males. Descriptive Pbasures m. Eight (407.) of the 20 juveniles tested with lactating fenales were wounded; onne juvenile fenale died the day after the experi— mnt was canpleted, presnmably fran wounds received from a lactating female. No juveniles were wcnunded by adult nnales, ad onnly onne juve- nile, a female, was wounded by nonlactating fenales (Table 51). The fact that conntrols had no wounding indicates that juveniles were not lacerated while squeezing through the restraining grids. These find- ings support the conclusion that adult mle and nnonnlactating fenale E. 13. ibai_n_:_<_i_i_ are nnot agressive towards strange juveniles of either sex. I'bwever, lactating fenales are aggressive towards both male ad fenale juveniles. Direct Observations . The mice were observed in order to obtain measures of the frequencies of fighting, chasing, and avoidance beha- viors. The mnber of lS-second intervals during which both the 128 Figure 17. Graph of the man weight changes for juvenile males. The vertical lines indicate 1. l S.E. (N = 10). 129 1' nF—n 4L 1. 7 _ l an 1 m .24 1 .20. m02<=0 PIG—m3 w._3_. 0. 1 8 O 0 0 an O #1 2 0 ADULT Figure 17. 130 o 333. as o 3.080 dam—Um 3333 $8.8an flame anufiqflcfi mad: he ma 093 ozv HOHUHS 380m $3803 36%. 380m 308382 0:2 O O O O O H on O O O N O O O O O on O O O N O O O O O H BE my; 3503 38.53 pg gag «€238.53 gag .oz «HEQE 85302:. .02 mam—Um 0% mam: 0% G n E gen—33 0.382%. mo gm 96.4w organs :3 3an 131 juveniles ad the adults were observed in the resident area was low. Consequently, the frequencies of behaviors were too small to mrit statistical analysis. Only one fight was observed; a nonnlactating fenale initiated a locked fignt with a juvenile male on the first night of the test period. In general, there tended to be sore chases on the first night compared to the fourth night (Table 52) . Lactating fenales were observed to chase both juvenile mles ad juvenile fenales, but nonlactating fenales never chased juvenile fenales. Juvenile mles and juvenile fenales chased nonlactating fenales ad adult mles , respectively, but never chased lactating fenales. Juveniles tended to avoid adults mre than adults avoided juvenile (Table 53). Mutual avoidance was nnot prevalent between juvenile mles and nonlactating fenales. Conclusions, beyond the observation that both aggressive ad avoidance behaviors were noted, cannnot be firmly mde. Reproduction. Juvenile mles ad fenales were involved in preg- nancies (Table 54) . This indicates that sane of the juveniles were Sexually mtnnre at a maxim of 44 days of age. Therefore, sexual behavior my be involved in determining the settlenent pattern of dis- Persing juveniles. Since the pregnancies occurred during the four-day Period, pair-bonding in 11.13. gird; seem to be a rapid process. 132 Table 52. The total number of chases per treatment group. Each cell represents 300 15-secord intervals . First Night Fourth Night Adult Juvenile Adult Juvenile Chased Chased Chased Chased Juvenile Adult Juvenile Adult Juvenile Male Adult Male (17) 5 0 (ll) 1 0 ‘F'mana‘lacletatmg (69) 6 6 (28) 0 0 15:3sz (55)* 10 0 (0) 0 0 givenile Fenale Adult Male (22)* 0 0 (50) 2 2 muting (6)* 0 0 (16) 0 0 £25,213ng (20) 26 0 (55) 0 0 Note: The nunbers in parantheses are nunbers of lS-second intervals which mice were observed together in resident area . * = 292 lS-second intervals. 133 0.3335" 98000an «mm a a .000 ”.5303 5 H050?» Ban-030 0H03 00d: fig wag 303005 @8000an we 080a...» 0.8 00%....3 5 mg 05. "0002 N m o 33 o o o 88 weave 0,8863 o n H 80 N o o is wise 0.58er o a m 80 o m o 3.38 202 :8... Hanan 3985. o o 0 § 0 e o 13 388 050863 o o o as on N m as waste wfiufloedaz o m 0 3d 0 m o as on: $.13 632 3862. 698ng 03.8.4. 68882. 698ng HEB. £863. 39: 8362 88ng 83:: §HB< e363. 34663. 092 68088 632 65sz 88.8 3sz 8th ill 03.005" @80073 com 35.00% :00 some 3an cam—500.0 03 33223 05020 H5548 Eu. 00%va mo .3qu 38 m5 .mm. 0.309 1.34 Table 54. Treatment group sunnary of the umber of juveniles involved in pregnancies (N = 5). Emerimat Adult thenile Male Juvmile Famle Male - 2 Field Fenale 1 - Male - -* Laboratory Fenale 1 - *N = 4 becaus 1 juvenile was inadvertently sacrificed prior e to the end of the three-week post-test period. DISCUSSIG‘I The general hypothesis of this study was that the outcams of juvaiile-adult interactions in _11. g. ba_ir__di would depend upon both juve- nile and adult social factors . Interactions between a juvenile male or famle and an adult mle, a nonlactating famle, or a lactating female were studied in field or laboratory enclosures over four-day periods. Juveniles with no adults were used as controls against envimmtal factors extrinsic to the acper‘ixnents. The specific mrking hypotheses were thatthesexofjuvmilesandthe sexofadults and/ortherepro- ductive stage of adult fenales would affect both (1) the settlenent of juveniles and (2) the weight change of juveniles. Also, it was expected that the effects of the biological classes would be similar mast field- enclosure cmditions and laboratory conditions. The first working hypothesis concerning juvenile settlenent was par- tially confirmed in respect to both juvenile and adult social factors; the effects of social factors varied depending upon the social carbina- tions of interacting animals . Contrary to that expected, the effects of the biological classes on juvenile settlenent were not consistent in the field and laboratory experimmts. The second working hypothesis con- cerning juvenile weight change was partially confirmed in respect to juvenile sex, but it was rejected for the adult social factors. As expected, the effects of the biological classes on juvenile weight change were consistent in the field and laboratory experinents. 135 136 Juvenile settlement and weight change are discussed separately in refer- ence to: (l) the effects of social factors, (2) the juvenile-adult social relationships, and (3) the consequences for free—ranging g. I_n_. _ba__i_rc_l_i_ populations. Juvenile Settlement The importance of juvenile and adult social factors in affecting juvenile settlement in g. m. b_a_i_r_d_i was evaluated experimentally by determining the propensity of juvefiles: (1) to nest, during the day- time, in enclosures occupied by adults (nesting location), (2) to share nests with adults, if juveniles nested in enclosures occupied by adults (nesting arrangement), and (3) to rennin, string the nighttime active period, in enclosures occupied by adults (time variable). Each of these dependent variables served as a measure of juvenile settlement, and significant differences among social combinations were judged to indi- cate relative effects of adults in limiting juvmfle settlemmt. Effects of Social Factors. Nonlactating fenales limited the settlenent of juvenile fenales coupared with juvenile males based upon the sigmificant differences in the nesting arranganent variable during the field enqaerimentandm theuean timevariable chringthe laboratory experimmt. Also, in the laboratory experimt, juvenile fenales tended to nest in resident areas occupied by nonlactating famles less fre- qtmtly than did juvenile mles. Differmces in juvenile male and female settlement were not significant when adult males or lactating femles were encountered. 'Ihat social interactions, and not a funda- mental difference between juvenile sexes, led to differences in juvenile settlenent was indicated by the failure to detect significant 137 differences between the male and female juveniles that were placed in the control conditions with no adult. Differences in juvenile settlenent due to the effect of the sex of adults and/or the reproductive stage of adult fenales were detected for the man time variable in the laboratory experiment. Lactating fenales, canpared to adult males, nonlactating fenales, and the control condition with no adult, limited the settlenent of juveniles, but this difference was significant for juvenile males only. The analysis of the nesting arrangenents for the laboratory experiment indicated that the settlenent of both male and fenale juveniles tended to be limited by lactating fenales. A similar finding was obtained in the analysis of the nesting arrangenents in the field experiment, except that the settlenent of juvenile fenales appeared to be also linnited by nonlactating fenales. The latter two nesting arrangenent differences closely approached sig- nificance. thneriments were repeated under two environmental conditions in order to determine if the social variables had the same relative influ- ence in two conditims of space quality. Contrary to that eoected, the experimental physical environment was an inportant factor in deter- mining the outccues of juvenile-adult interactions in respect to juve- nile settlenent . Nesting arrangements were significantly different anong social canbinations in the field experiment, and both the nesting location and the mean variables were significantly different amng social combinations in the laboratory experiment. The confirned and sterile space of the laboratory enclosures prob- ably led to an increase in the mnber of contacts between the interact- ing prairie deer mice, which, in turn, increased the frequency of the 138 behaviors that resulted in better resolution of the rnesting location variable and the mean time variable over that in the field experiment. However, since feder juveniles nested in the resident area during the laboratory experiment, sample size was decreased to a point where the nesting arrangenent measure did not significantly differentiate between the effects of social factors, though borderline significance was obtained. 0n the other hand, the nesting arrangement measure could be used to detect significant and borderline to significant differences amng social combinations in the field experiment. In general, it can be concluded that the laboratory experiment obtained more conclusive evidence for the effects of social factors on juvenile settlenent . The point should be made that the validity of the laboratory results was not reduced by the proposed high frequency of behavioral interactions because juveniles could escape the presence of adults by moving to the dispersal areas. The claim that the frequency of behavioral interactions was depen- dent upon the enclosure size is in agreement with the findings of Calvin (1973) . This investigator found that the frequency of agonnistic postures of Microtus males was inversely related to enclosure size . The grassland habitat of the field enclosures in the present experiment may have also served to decrease the frequency of contacts and behav- ioral interactions and/or increase avoidance by providing cover and refuge sites. The effects of changes in weather conditions are not known. The four-night means of the time measure were reliable in deter- mining consistent differences among social combinations in the labora- tory eiqneriment . Contradictory significant differences were not 139 detected in the analysis of the individual nightly values . The signifi- cant differences between juvenile males ad fenales that interacted with nnenlactating fenales were present on all four: nights, ad significant differences anong adult classes for juvenile males were found on the last three nights. These results suggest that the respective social relationships were established at least by the second night and ramined stable during the course of the experiment. Social Relationships. The results of this study on _I_’_. Q. Lair_d_i indicate that nonlactating and lactating fenales, but nnot achnlt males, may limit the settlement of dispersing juveniles depending upon their sen. These findings are contradictory to the canclusions of Sadleir (1965) and Healey (1967), for g. g. ansterus, and Flowerdew (1974), for A. sylvaticus, but are consistent with Phtzgar's (1971) findings for g. _IEEQEE: waever, the effects of different femle reproductive stageswerennot detenninedinthelatter studyandtlneresults ofthe present investigation are more specific. 'nne different conclusions may reflect upon differences in the social behavior of the different taxa and/or upon the failure of previous investigators to consider the biological determinants of social behavior. Peragyscus mniculatus ansterus and A. sylvaticus males are aggressive towards juveniles, but nonlactating fenales typically are not (Healey, 1967, and Flowerdew, 1974). I-bwever, in 3. g. _:ba_i;_<_1_i_, lactating fenales frequently wounded juveniles, but adult males and nonnlactating fenales were typically nonaggressive. Direct observations provided no support of the relative aggressiveness of lactating fenales , basedtqnonwonnnding, becansetheprairiedeermdceinthepresent study infrequently interacted, and canclusive patterns in the relative 140 levels of agonnistic behaviors could not be determined. Hill (1970) reported low frequencies of behavioral interaction between adult 2. M and _I:. m. 3.13.4.1. and suggested that high levels of social interaction mnay not be necessary to produce significant effects on the spatial relationships of Peronyscus . The relative limitation of juvenile male settlenent by lactating fenales can be enqalained by the presence of maternal aggression dir- ected towards strange juveniles . There was some evidence to indicate that lactating fenales, compared with adult males, also limited the settlenent of juvenile females , though the results of the statistical analyses were not conclusive. It may be suggested that immature female 3. 3. E1511; do not recognize differences among adult social classes as mrkedly as do young males and therefore do not form different social relationships as readily. On the other hand, lactating females may be less aggressive towards juvenile fenales than they are towards juvenile males. Although, it is doubtful that differential aggressiveness exists because Savidge (1974) found that, in g. g. b_a_i_r_c_l_i_, there is no differ- ence in the dispersal rates of male ad fenale juveniles in response to aggression by their mother. This implies that the adult fenales were similarly aggressive towards nale and fenale offspring. Also, the pos- sibility exists that the treatment combination sanple sizes of the pre- sent experiment were not sufficiently large to detect true differences , if they existed. A functional explanation for the failure of adult males to limit the settlement of juvenile males in the present study is not available. It may be suggested that E. n_nn_. ba__i_1_'_d_i males are more social than pre- viously reported (I-kmard, 1949) , and juveniles do not avoid adults upon 141 encounter. Myton (1974) implied that 3. 1w males are more social than fenales, and Metzgar (1971) found that resident female g. M appeared to limit thne number of transient fenales that settled in a local population, but males did not limnit the settlenent of juvenile males. The sane relationship may exist in g. n3. ba___i_r___di even though Hill (1970) observed nno difference in the levels of aggression during adult fenale-fennle and male-male encounters . In the present study, nonlactating fenales limited the settlenent of juvenile fenales canpared with juvenile males , but nanlactating females selden wournded juveniles. It appears that the juvenile fenales avoided the nonlactating fenales in the absence of overt aggressionn. Terman (1962) and Hill (1970) indicated that Percmnyscus commonly avoided conspecifics upon encounter, and suggested that animals may be organ- ized in space without aggressive interaction. It should be mentioned that agonnistic interaction may involve behavioral patterns that do not include direct fighting (Scott, 1972) . These would not be detected by anexaninationofmndingaswasdoneinthepresent study. The differences in settlenent between the male ad female juveniles that interacted with mnnJactating fenales can also be explained in terms of the establishment of sexual relationships between the young mnales and the nonlactating fenales . Sexual relationships were present between juvenile fenales and adult mnales as well. Canpared with lactating fenales, the failure of adult mnales and nonnlactating females to limnit the settlenent of juvenile females and males, respectively, might be expected because home ranges of opposite sexes in PenLnyscus usually overlap more thnan those of individuals of the same sex (see Stickel, 1968) and g. n_n_. ba__i._r__di form mon'noganous pair-bonds aboard, 1949). 142 Successful pair-bonds may be formed over a four-day period because both male ad female juveniles were involved in pregnancies. That most (about 807.) of the juveniles were nnot involved in pregnancies during the experiments indicates that individual differences in sexual matnn'ity existed annong juveniles because all adults were in breeding condition. Of course, another possible explanation is that pair-bonding and impreg- nation typically takes lonnger than four days. Hill (1970) found that nine out of ten _l_’_. 13. gird; adult pairs bred within a seven-day period. Finally, the relative differences in thne laboratory enqneriment nesting area and per cent time measures were not consistent with the lack of differences in the rate at which juveniles entered and left the resident areas . It appears that the relatively exclusive or nonexclu- sive use of space by juvenile and adult g. g. iba_i_r_di_, as determined by social interactions, is nnot permanent. That is, the degree of mutual use of the resident areas by juveniles and adults was nnot related to the frequency at whnich juveniles entered and left the resident areas. This finding is in accordance with the reported influx of neighboring g. M into depopulated areas (Stickel, 1946) . Peromyscus my be envisioned as connstantly probing the periphery of their home ranges, and using new space or establishing new social relationships if the opportunity arises. Hill (1970) reported that Percuyscus use space as it becanes available to then. Yet uost individuals in free-ranging populations are arranged spatially in stable hens ranges, though sane individuals may switch hone ranges (see Stickel, 1968) . The social behaviors determining the spatial organizations of Perenyscus popula- tionns are not apparent, but it is known that Peregyscus is nnot terri- torial (Hill, 1970) . 143 Different rodent taxonanic groups have been reported to vary in respect to male and fennale aggressiveness (Steiner, 1972, and Michener, 1973 for Spernnophilus, and Eisenberg, 1963 for Perchyscus) . It is not suggested hnerein that aggressive males are unimportant in determining the juvenile settlenent for some groups because much evidence is avail- able to docunent the contrary (3.3. , Carl, 1971). The point to be made is that generalizations fran such groups (3.3. , E. m. austerus, Healey, 1967) to other taxa (_e_.g. , _11. n3. _b_8;__irg_i) are premature until future comparative studies are conducted. Further, the results of this study show that settlenent may be influenced by botln juvenile ad adult social factors , and social relationships thnat do not involve aggression may be fanned during the settlenent process. In smmary, the social organizations of small rodents may be characterized by various social relationships thnat differ anong taxa. That these differences may exist between Peramyscus subspecies is not surprising because significant differences have been reported in habitat selectionn (Harris, 1952) and several other behavioral responses including activity (Foster, 1959) for g. maniculatus subspecies . Correlates of the potentially different social organizations such as nodes of environmental exploitation, habi- tats, ad life histories present intriguing topics of future compara- tive research. These investigations should add insight into the func- tional significance of the various social systems. The role of lactat- ing fenales should not be overlooked in such studies because maternal aggressionn may be an important general phenomenon in Perenyscus ad othner rodents. Consequences of Juvenile-Adult Interaction. The results of the present study allow sane inferences to be made concerning the 144 consequences of juvenile-adult interactions on juvenile settlenent in 1:. g. m, if the social relationships in free-ranging populations are similar to those found using enclosures. The population biology information has been taken from hbward's (1949) report. It has been assured hnerein that the sex ratio of dispersing juveniles is constant over time, although there may be a larger number of males than fenales born into a population as was found by Terman ad Sassanan (1967) for a laboratory colony of _1:. m. M. iny intraspecific social inter- actions have been considered, but interaction with other rodent species (Grant , 1972) ad availability of physical environment requirements such as nest sites (Rose and Gains, 1976) may influence settlenent in rodents. Dispersal and subsequent settlenent occurs during the breeding season (warmer months) in g. g. m. During periods of peak repro- duction (spring and late summer-fall) when a high percentage of the resident females are nursing young ad exhibiting maternal aggression, juvenile males should be less likely to settle than at other times when reproduction within the pOpulation is lower (mid-summer) . Since young fenales appear to be similarly limited by nonlactating ad lactatinng females, the ratio of juvenile males-to-fenales that successfully settle should vary directly with the ratio of ronlactating—to—lactating fenales in the population. There is same indicationn that the settlenent of juvenile fenales is limited by adult females, but there is no indi- cation that adult males limit the settlement of juvenile males . To summarize, the settlenent of young males is suggested to be related to the reproductive activity of adult fenales withnin the populationn, 145 whereas the settlenent of inmature fenales may be related to the popu- lation density of resident fenales. It is noteworthy to nention sane relationships between thne propor- tion of g. maniculatus fenales in local papulations and population den— sities as determined by Terman and Sassanmn (1967) from analyses of NACSM data. These data were obtained fran local populations in succes- sive seasons (spring-fall-spring) or the sane season through successive years (spring-spring, fall-fall). In the successive seasons ad spring- spring couparisons , the proportion of fenales increased or decreased when population densities were increasing or decreasing, respectively, significantly more frequently thnann when papulation densities were decreasing or increasing, respectively. A sinnilar pattern existed for the fall-fall carparisons, but it was not significant. This relation- ship between the proportion of fenales in the papulation ad the popu- lation density changes may indicate that the ratio of juvenile males- to-fenales that settled into the populations varied directly with the ratio of mnlactating-to-lactating fenales. Though it was not men- tioned if changes in the population densities were related to chnanges in reproductive activity within the populations. The same couparisons for g. M populations were nonsignificant. Three points should be mentioned with the proposed settlenent uodel in mird. Firstly, since reproductive activity in E. m. bairdi and other Peranyscus species (3.3. , Rintanoa, gt 3],, 1976) decreases during mid-smmer for unlqrmn reasons , the increased settlenent of males mild probably not be due to sexual relationships being famed with rnonlactating fenales. 'I‘rne causative factor would nore likely be a decrease in the percentage of resident fenales that are lactating. 146 Of course, if young mles settle near nonlactating fenales, then the chances are great that thney would be involved in reproduction during the fall breeding peak. The second point is that the assunption that juveniles disperse only as far as is required to reach a suitable settlenent site is inplicit in this verbal nodel. It follows that dispersal distance for mles would vary in response to changes in the ratio of nonlactating- to-lactating fenales . However, several field stndies on PM have obtained evidence for slightly greater dis- persal distances in males (see Stickel, 1968, ad Howard, 1949 for g. in. gene). though these studies did not attenpt to relate dispersal distance with the papulation reproductive activity. Thne reasons for the differing dispersal distances are not known, but these may be related to differences in activity or trapability (Termann, 1968) . The third ad final point is that whnile g. m. ba__irgi hnas been reported to form nonoganous pair-bonds (Howard, 1949) , the present research eval- uated social interaction between two animals only to obtain Specific informtion. The behavior of adult mles toward juveniles nay deped upon the presence of an adult fenale. Still, the mechanisms associated with pair-bonding are rot known. The fact that Birdsall ad Nash (1973) reported that scnne g. maniculatus litters (ll of 107 ecanined) con- tained at least three different alleles of parental origin indicates that the uorogauous pairs are not consistently mintained. The proposed settlenent nodel can be explained on the basis of individual selection. If an individual remined in the presence of hostile social conditions ad/or where the chances of successful repro- duction were low, the fitness of that individual would be correspond- ineg low. Therefore, juveniles should continue to disperse in order 147 to increase the probability of reproductive success as long as the potential gain of moving is greater than tho potential cost of renain- ing in an area. Mir-ray (1967) gave a similar and more detailed argu- ment, but did not consider the role of social relationships othner than dominance-subordinance. Healey (1967) discussed the individual selec- tive advantages of aggressiveness in g. E- austerus adult males in respect to their role in affectirng the settlenent ad snn'vival of imnigrating juvenile conspecifics. These proposed advantages are con- sistent with Murray's suggestions. Thne evolutionary significance of the proposed relationships in g. 13. 29.-Iii. between the papulation reproductive activity and the settlenent of juvenile mles ad the population density of resident fenales and the settlenent of juvenile fenales is not clear. Hmever, these proposed correlates may be due to tho selection for different social relationships that consequently effect populationn regulation, though the present study was rot designed to evaluate mechnanisms of papulation regulation ad future research is required . Tromnbersofyoungg. TB. MWedareineccessoftho nunbers that settle within a population. local populations should be limited at an absolute density that can be supported by tho area with the surplus juveniles dispersing into neighnboring populations or being taken by predators. It can be speculated that the ultimate determinant of population density in g. g. _bgigdi is tho umber of resident, breed- ing fenales within a populationn. If thne number of adult fenales is limited at sonno maxinum level, the absolute production of young will also be limited. however, the production of young per female will be madndzedatalevel thatcanbesupportedbytheresources intlne 148 natal bane ranges. The survival of offspring will be enhanced if lac- tating fenales repulse strange juveniles that nould be conpetitors for the available resonnrces in the natal hono ranges. Nonlactating fenales nnight be expected to repulse juvenile fenales because a single home range may not provide adequate resources for two lactating fenales and thoir litters in the event that adults ad juveniles become pregnant. (In tho other hand, ronlactating fennales should not repulse juvenile males because of the potential for future breeding. Whether males conn- pete for available hero ranges and notes , as is tho case in other rodents (9g. , Carl, 1971), emits future study. It should be noted that nother-offspring interactions nnay be influenced by different selection pressures than those between strange animals because the nother ad offspring are related by one-half. Off- spring might be repulsed in order to reduce the chances of parent-off- spring natings , especially if tho parental mles rennains near the natal site. hbwever, Williams (1976) suggested that, uder certain condi- tions, inbreeding may be adaptive. Also, the inclusive fitness of the nother may be increased if shne permits her offspring to establish breed— inghonorangesnearthenatal sitebecause thorisk incurredto tho offspring while traveling across unfamiliar terrain would be reduced. Alternatively, the inclusive fitness of the nother mighnt be reduced if offspring breed nearby because the natal area may not contain suffi- cient resources to support several lactating fenales ad thoir litters . Savidge (1974) reported that adult 2. g. bfldi fenales differ in aggressiveness and in the consequent repulsion of offspring. however, whnether differences exist in the behavior of adult fennales towards 149 offspring and strange juveniles has not been determined ad should be tho tapic of finture research. Dispersal may affect the densities, age and social structures, and sex ratios of local populations (Lidicker, 1975). Advances in the understanding of the population biology of small normals can be node by studying the thnree phases of dispersal independently. Eventually, the attainment of a conplete uderstanding of dispersal will require that investigations be conducted on the abiotic ad biotic factors that influence all thnree phnases (leaving tho natal area, crossing unfamiliar space, ad settlenent) in field populations. Then this understanding can be integrated into the study of nortality ad natality at the popu- lation level. At present, studies such asthose, as voll as an evalua- tionn of tho proposed settlenent nodel for _l_’_. g. M, are not feasible because adequate sarpling technologies are not available. The results of this study suggest two liros of research that could be conducted using enclosures to provide adequate central over experi- mental aninnals. Firstly, the preference of juveniles for adults should be examined. This study would be especially holpful in giving greater resolution in the juvenile male-achilt nnale ad the juvenile mle-non- lactating fenale relationships . Secondly , ad of nore innodiate inter- est, the effect of adult pairs on juvenile settlenent should be eval- uated. If adult fenales, ad rot adult males, limit tho settlenent of dispersing juveniles, then the presence of an adult nole along with an adult fenale should not influence tho response of juveniles toward adults and vice versa. This study would connclusively determine (1) whothner adult fenales limit tho settlenent of juveniles ad (2) whnethor juvenile—adult male social relationships are dependent upon tho 150 presence of an adult fenale. Taken together, tho results of these two tests muld challenge the existence of the social relationships neces— sary for tho prOposed settlenent nodel. Juvenile Weight Change Tho inportance of juvenile ad adnlt social factors in affecting the growth of juvenile F. m. begin; was evaluated by examining the dif- ference between initial and final weights of juveniles. Both the sex of juveniles and the sex of adults ad/or the reproductive stage of adilt fenales were hypothesized to affect juvenile growth. It was expected that the effects of tho biological classes would be similar uder field-enclosure conditions and laboratory connditions. Effects of Social Factors. A significant difference in juvenile growth rates do to tho sex of juveniles was detected for those juve- niles that interacted with adult males. In this case juvenile fenales had lower growth rates thnan juvenile males. Differences between tho male and fenale juveniles that interacted with nonnlactating fenales or lactating fenales or encountered the control condition were ronsigni- ficant. There were no significant differences in tho growth rates of male or fenole juveniles anong tho adnlt classes, but central juvenile males gained less weight than did those males that interacted with adnlt nnales. The weight gain of conntrol juvenile fenales was similar to that of tho fenales that interacted with tho respective adnlt classes. Juveniles gairod significantly nore weight drring tho field experiment canpared to the laboratory experiment. however, thore were no statistical interactions between tho effects of location ad either tho effects of juvenile or adult social factors. Therefore, it can be 151 concluded that tho effects of tho social factors on juvenile growth were similar in both experiments. Social Relaticnnships. Tho results of tho present study on g. g. M do not indicate that adnlt males inhibit tho growth of juveniles canpared to nonnlactating fenales or lactating fenales, although social interactionn with adult males caused a difference in tho weight gain of juvenile males ad fenales . This conclusion is contradictory to that of Healey (1967) for _I_’_. n3. austerus; several possible reascnns exist for tho difference in results . Firstly, Healey allowed adilts to interact with juveniles over a l4-day period ad social interactionn in tho pres- ent study was limited to four days. Clearly, the longer tino period nay have been inportant in increasing tho relative differences in weight gain anong treatnnent connbinations. But a longer period probably was not necessary to prodice significant differences becanse Healey noted a dramatic weight loss within four days in juveniles released onto field plots containing aggressive males , canpared to juveniles released onnto field plots conntaining docile males . The weight loss over tho renaining ten days was relatively slight . Secondly, tho experinontal design used by Healey called for juveniles either to be confined with adults in canparatively snoll laboratory mazes or to be free to enri- grate (disappear) fran field plots ad be unavailable for tho analysis of weight changes. All juveniles in tho present study were allowed to avoid adults by enterinng tho dispersal area, however, these animals were included in the analysis. It is not loom how this design feature influenced tho results of tho present study. however, since juvenile settlenent was not limited by adult nales , canpared to nonlactating fenales or lactating fenales in both tho field ad laboratory 152 enqneriments , it appears that juveniles renainod in tho presence of adilt mnales. This latter point leds support to the third explanation of tho contradictory conclusionns. No attenpt was made to modify aggressionn among males in tho present study. This difference in designs probably is not important in explaining the difference in results because Healey reported that _I_’_. g. austerus males , irrespective of thoir aggressiveness, tended to inhibit juvenile growth to a greater degree than did adult fenales . Tho differences between results may reflect basic differences in tho social behavior of E. g. austerus ad 2. 13. gird; with tho fornor subspecies being more aggressive. Tho difference between the weight gain of juvenile males and females that interacted with adult males apparently was not related to aggressive behavior because, as previously mentioned, adnlt moles never womded juveniles and tho direct observations of encounters indi- cated that adnlt males behnaved sinnilarly towards juvenile males and fenales . Perhaps the sexual relationship between adult males ad juve- nile fenales caused increased activity or decreased food intake and led to tho relative retardationn of growth. The fact that juvenile males tended to have lower growth rates , compared with juvenile fenales, when nonnlactating fenales were encountered is consistent with tho notionn of short-term effects of sexual behnavior on tho growth rates of juveniles. That tho activity rates in the passage tunnel were not different between tho juvenile males and fenales that encountered adilt males or nonn- lactating fenales indicates that food intake , ad not activity, was affected by social interaction. Still, tho possibility renains that activity, other than the one measured, could have chnanged. 153 Consquences of Juvenile-Adult Interactions . The basic thosis of Healey's study nos that aggressive adult males decreased juvenile snu- vival ad growth and effected population regulation in E. g. austerus. Tho inhibition of juvenile maturation resulting from interaction with adults is of great potential importance in determnining tho status of Perenyscus populations, if a decrease in growth rate is related to an increase in tho age at which juveniles become sexually mature. 0thor field investigations are lacking, and data from laboratory studies do not provide conclusive support of either tho existence of inhnibitionn or the specific social correlates of inhibition that can be extrapolated to free-ranging E. g. gird; populations (Terman and Gardner, 1970, ad Thales and Terman, 1975) . Several inferences can be drawn in reference to the effects of social interactionn on juvenile growthn rates in E. g. ba__i_n_'_d_i as a result of this study. Firstly, since the abiotic environment of tho experi- ment proved to be a significant factor in affecting tho juvenile weight changes, growthmoasures may be partly spurious to social fac- tors. Additionnal support of this conclusion is the finding that there were no statistical interactionns between the location of tho experi- ments and the social factors , which indicates thnat those two treaments operated independently of onne aothor . Secondly , failure to detect differences anong juveniles that interacted with tho respective adult social classes may have been due to tho juveniles avoidinng the adults within the resident area or moving to tho dispersal area. In either case, it is likely that juveniles in free-ranging populations escape potentially adverse social conditions that could cause an inhibition of growth. This conclusion is supported by the finding that interaction 154 with lactating fenales, the most aggressive adult class, did not cause an inhnibition of juvenile growth conpared to interaction with tho typically nonnaggressive adult males or nonlactating fenales. Thore- fore, the third inference is that a relative retardation of growth do to short-term social interaction need not be correlated with aggressive behavior. The difference between juvenile males and fenales that interacted with adult males suggests that sexnual relationships may be important in affecting juvenile growth. Fourthly, the differences between juvenile nnales that encountered adult males or the control con- dition indicates that social interaction between male _F_. 13. ba__i._r_di pro- motes growth conpared to situations where interaction is not possible. Perhaps social facilitation influences tho food intake or juveniles. This point reflects uponn aother aspect of the apparent basic differ- ences intho social behaviorofg. g. austerusadfi. g. M, ad agrees well with the finding that adult males do not limit the settle- ment of juvenile males . Future research should be conducted in an attenpt to docnnent tho occurrence of maturation inhibition uder experinnental conditions where juveniles can avoid the presence of adults. A study on an aggressive Peromyscus taxon, such as 1:. mg. austerus, would provide conclusive evi- dence of whother juveniles avoid hostile social conditions and thereby avoid possible retardation of growth. Another study should examine the role of sexual relationships in affecting juvenile growth. Special attention in both studies should be given to a determination of the immediate nechnanisnns responsible for differences in juvenile growth rates. Activity ad food intake may be the most useful dependent vari- ables in short-term experiments. 155 Summary 1. Juvenile-adult social interactions in I_’. IE. bairdi were studied using field (3.8 m2) or laboratory (1.2 m2) enclosures. The working hypotheses were that the sex of juveniles and the sex of adults and/or ' the reproductive stage of adult fenales would affect: (1) the settle- ment of juveniles and (2) the growth rate of juveniles. It was expected that the effects of the biological classes would be similar under field- enclosure conditions and laboratory conditions . 2. Following a three—day pretest period during which all prairie deer mice could becone familiar with the experimental environments , juveniles could becoro familiar with the test apparatus, ad adults could establish residency, a juvenile male or female was placed in an enclosure containing: (1) an adult male, (2) a nonlactatinng adult female, (3) a lactating female, or (4) no adult as a control. Five replications of each of the eight social conbinnations were used in the field and laboratory experiments . Tho juveniles were free to interact with tho adults over four-day periods. During the experiments, juve- niles could nove through a two-way tunnel to an adjacent uoccupied enclosure, but adults could not, and all mnice had access to nest boxes. 3 . Juvenile settlement was evaluated experimentally by determin- ing the propensity of juveniles: (l) to nest, during the daytine, in enclosures occupied by adults, (2) to nest with adults, if juveniles nested in enclosures occupied by adults, and (3) to retain, during the nighttino active period, in enclosures occnupied by adults. Significant differences anonng social combinations were judged to indicate a rela- tive limitation of settlenent. The growth of juveniles was examined by analyzing the difference between their initial ad finnal weights . 156 Juvenile activity between enclosures, juvenile wounding, and reproduc- tive activity were measured, ad direct observations of interactions in the laboratory were conducted. 4. Tho settlemennt of juvenile fenales coupared with the settlement of juvenile males was limited by nonlactating fenales. Differences in settlenent due to the sex of juveniles were not significant when juve— niles interacted with adult males or lactating fenales or encountered tho control condition of no adult. Tho effects of adult social factors wee significant for juvenile males; lactating fenales , coupared with adult males, nonlactating fenales, or the control condition, limited the settlenent of immature males. however, diffeences in settlenent do to tho adult social factors were only bordelino to significance for juvenile fenales; both nonnlactating and lactating fenales, coupared with adult males , teded to limit the settlenent of immature fenales . Similar outcoues of social inteactions in free-ranging populations would result in nonnlactatinng and lactatirng females, but not adult males, limitinng the settlenent of dispersing juveniles depeding upon their sex. Tho effects of the social factors were not consistent in tho field ad laboratory expeiments possibly due to diffeences in the frequencies of behavioral inteaction ad/ or avoidance . 5. It was proposed that contradictions in the results of previous studies on Peomnyscus (9g. , Healey, 1967) ad the present researchn are do to diffeences in social behavior of the different taxa and/or the failure of past workers to conside the biological determinants of social behavior. lactating g. n_n. bairdi freqnuently wouded juveniles, but adult males and nonlactating fenales were typically nonaggressive towards juveniles . Also , sexual relationships were formed between 157 adults ad juveniles because sore juveniles were involved in preg- nancies doing the experinents. 6. The relative differences in the laboratory experiment nesting area and per cent tine measnoes were not consistent with the lack of diffeences in tho rate at which juveniles entered and left the resi- dent areas. That is, the degree of mutual use of the resident areas by juveniles ad adults, as detemined by social inteactions, was not related to the frequency at which juveniles entered and left tho resi— dent areas. Thnis is in agreement with previous reports on tho utili- zation of available space in Peronyscus . 7. Juvenile fenales, conoared with juvenile males, had signifi- cantly lower growth rates when adult males were encounteed. Diffe- ences in growth rates between male and feunale juveniles that inteacted with nonnlactatinng females or lactating females , and diffeences between control juveniles, were nonsignificant. Thnere were no significant diffeences in the growth rates of male or female juveniles do to tho adult social factors, but control juvenile males gained significantly less weight than did those males that inteacted with adult males . 8. These results may also reflect differences in the social behavior of diffeent Peromyscus taxa. However, behaviors othner thnan aggression, such as sexual behavior, may affect juvenile growth by alterinng juvenile activity and/or food intake. The effects of tho social factors on juvenile growth rates were siumilar ude the two con— ditions of space quality. But juveniles in tho field expeiment gained significantly more weight than the juveniles in the laboratory experi- ment. Thnerefore, diffeences iun juvenile growth rates may be partly spurious to social factors, at least on a short-term basis. APPENDIXA 158 APPENDIXA ANALYSIS OF THE DISAPPEARANZE OF JUVENIIES 130le THE FIELD MERIT/TENT No adults were lost fromn tho field enclosnoes, but seveal juve- niles escaped or were taken by predators. A 2-way contingency table showed that the proportions of juvenile males ad fenales that dis— appeared froun these enclosnoes wee close to being significantly diffe- ent (P = .06, Table 74). The data from the field eoeiment were eval- uated under the assumption.of equal disappearance of juveniletmales and females. hbweve, only data collected from juveniles that completed tho eoeriment were analyzed. Table 74. Chi square analysis of the disappearance of juveniles doing tho field expeiment. Juvenile hole Juvenile Fenale Disappeared 9 2 Did not disappear 20 20 x2 = 3.5608, df = 1, P =2 .06 APPENDIXB 159 APPENDIXB JUVENIIEACI‘IVITYWI'I‘I-ENARFAS The activity of juveniles within tho resident ad dispersal areas, expressed as the mine of feede tunnol trips pe hono juveniles had access to feeders doing the active period, was examined to determine if tho expeimental social coditions influenced the movenent of juve- niles within areas. The proportion of juveniles that tripped feede tunnels in the field experiment was significantly smaller than that in tho laboratory experiment (.005 > P, Table 55 for the resident area ad Table 56 for tho dispersal area). Tho foo-night mean activity rates within the resident ad dis- pesal areas are given in Table 57; those data were analyzed with non- parametric statistics (Tables 58 ad 59 , respectively). No significant diffeences were food for the main effects of juvenile sex or adult class. The juvenile sex x adult class inteaction was nonsignificant as well. Tho results for the field expeinent indicate that social conditions did not affect tho juvenile activity rate in eithe the resi- dent or dispesal areas. The back transformed within area activity rate means obtained in tho laboratory expeimont are given in Table 60. These data were examined by analyses of variance after tho raw scores for each replica- tion were transformed to the W scale. The tables ad fignures showing the diffeences anong moans give the transformed values. Neithor the juvenile sex nor adult class main effects were sig- nificant for the resident area activity, but thnere was a significant 160 Table 55. Chi square analysis of the proportions of juveniles that tripped tho feede tunnels in the resident area doing tho field and laboratory expeinonts . Did Not Tripped Tumols Trip Tunnels Field Expeiment 9 31 laboratory Experiment 29 ll x?- = 20.0501, df = 1. .005 > p Table 56 . Chi square analysis of the proportions of juveniles that tripped tho feeder tumols in tho dispesal area during the field ad laboratory experiments . Did Not Tripped Tunnels Trip Tunnels Field Expeiment 9 31 laboratory Expeiment 32 8 x2 = 26.4665, df = 1, .005 > P 161 dag—Ha Hoooom moo condemn pg 835%“. mo won—HE u m dance—flame agenda cu 96 o u za 2v 8V Amv is 8V A6 A8 A3 M 8.0 85 «No 8.5 86 3.0 m: 35 “dashed 82 assumes Ad 8V Ad 40 8V A8 no one a 3.0 8.0 one Ems 8.0 and 3o 85 seashore. 82 ”.83QO 333 6% madam madam on: 33 so wide wade 3m: 53% H3280 ween—woos wfiuouuoaacz Hohcoo wfiunuone wfiuouuflgz manna dengue was finesse .3 n 5 egos 33m of made eased o>..fiom moo mo g and wound. Hose—.5 Hoooom mo .85 one mo mg 39% womanhoom 9n. Km manna 162 Table 58. Nonoaranetric analysis of the foo-night mean juvenile activity rate in the resident area doing the field experiment . Comparison Test wnn or HID Value Prob . . Wilcoxin's ' _ Juvenile rank-sun test Wn - 26 P > .1 Kruskal- _ Adult Wallis test H/D - 0.8261 .9 > P > .5 nan-1.2;.“ edge... um = «9912 .s > P > .1 Table 59. Nonparametric analysis of tho foo-night moan juvenile activity rate in tho dispesal area doing the field experiment. Conoarison Test Wn or H/D Value Prob. . Wilcoxin's Juvenile rank-sum test Wn = 27 P > .1 Kruskal- _ Adult Wallis test H/D - 1.7387 .9 > P > .5 “mm x Adm K‘USkal' H/D = 1.8602 .9 > P > .5 Inteaction Wallis test 163 Add one 58 gonad some n o Gadnwndv $m.~n3.8 39360.8 ANo.nnoo.Hv 8573.8 QHdnoodv Gnu—”78.8 86-3.8 m 84 e: 3.0 8.0 a: £0 o: 84 gingham 82 Homemade 55-8.8 cannons 93500.8 80.738 $.~-00.8 9.3-00.8 $000.8 80600.8 m N: 03 a: 00.0 :4 9.0 3.0 «a; seashore 8.2 ahead 33 05 380m madam 3a: 333 65 dengue Sufism was: $33 Hoodoo moods: wfiodooaSz H8080 winded wand; magma mag ode: mag .Amléufigggflfigoi gwuuofiu wognwomofi Hod—Ego magnum—flea gaugnyom gmnfiuuxuooflfl. .8 038. 164 juvenile sex x adult class interaction (.05 > P > .025, Table 61). Annalysis of the simple main effects revealed that the sex of juveniles affected their activity within the resident area only when nonlactating females were encountered (.01 > P > .005, Table 61). Figure 18 shows that, in the presence of nonlactating females, juvenile fenales had a higher activity rate than did juvenile males. None of tho othor juve— nile sex couparisons wee significant. But at a bordeline significance level, juvenile fenales, coupared with juvenile males, had a higher activity rate when lactating fenales wee encountered (Table 61) . The back transformed means for this conparison are shnownn in Table 60. Thnere were no significant diffeences among the adult classes for eithne juve- nile males or fenales. Finally, thee were no significant F-values for the juvenile activity rates within the dispersal area (Table 62) . In conclusion, juvenile activity within the resident area was influenced by social factors only in tho couoarison of juvenile sexes for nonnlactating fenales doing the laboratory experiment . Juvenile activity within the dispesal area (in the absence of adults) was affected by neither the juvenile nor adult social factors doing the field ad the laboratory expeiuments . As noted earlie (see Methods), activity noasnoed in a feeder tun- nol is subject to confoundingvariables admnightnotprovidea true estimate of tho effects of social factors on locouotor activity. how- eve, the propensity of juvenile rodents to enter ad leave feeder tun- nels may be related to tho innclination to ente live-traps because the same confounding variables are present in both situations. That is, bait prefeence of individuals and thoir response to the physical 165 Table 61. Analysis of variance for tho foe-night mean juvenile activity rate in the resident area during the laboratory eqnerinrent. Source of Variation (if PBS F-value Prob . Juvenile 1 1.0365 3.6036 .1 > P > .05 Adnlt 3 0.0234 0.0813 .9 > P > .5 Juvenile x Adult '3 0.8901 3.0945 .05 > P > .025 Error 32 0.2876 Total 39 (SS = 12.9812) Simple Main Effects of Juvenile For: Adult Male 1 0.2958 1.0284 .5 > P > .25 Nonnlactating Female 1 2.2563 7.8445 .01 > P > .005 lactating Fenale 1 1.1560 4.0191 .1 > P > .05* No Adult 1 0.0000 0.0000 P > .75 Simple Main Effects of Adult For: Juvenile Male 3 0.5429 1.8874 .25 > P > .1 Juvenile Fenale 3 0.3699 1.2861 .5 > P > .25 Error 32 0.2876 *Indicates bordeline significance whee tho calculated value was within .20 units of tho eitical value. 166 Figure 18. Graph Showing the significant difference (.01 > P > .005) in the four-night mnean juvenile male and fenale activity rate in the resident area occupied by nonlactating fenales in the laboratory expeimnent. The vetical lines indicate i 1 S.E. (N = 5). 167 . fl E n I.— A M E F G W T M C n A w + Vfippr-PP snow-mewuawfiomm 22111.1.100 _._._>_._.0< m.=zm>..=. FEMALE MALE JUVENILE Figure 18. 168 Table 62. Annalysis of variance for tho four-night mean juvenile activity rate in the dispesal area during the laboratory experiment . c-f- Sonrce of Variation df IVES F-value Prob. Juvenile 1 0.4016 0.7631 .5 > P > .25 Adult 3 0.8023 1.5244 .25 > P > .1 Juvenile x Adnlt Inteaction 3 0.0680 0.1292 P > .75 Error 32 0.5263 Total 39 (SS = 19.8534) aspects of the feede or trap may influence novenent into those struc- tures. The findings of tho present study are in conflict with sugges- tions that adnlt small rodents decrease tho trapability of juveniles by somohnow restricting juvenile movenonnt (Watts, 1970) . Those data will be presented in a future pape. mm C 169 APPENDIXC ANALYSESOF'IHEDAEYNESI'INGAREAW Table 63. Chi square analyses of the daily nesting area frequencies. These comparisons ShOW’the lack.of joint independence of location of tho expeimont, juvenile sex, and.adult class. df xz-value Prob. Day 1 3 29.1976 .005 > P Day 2 3 33.4616 .005 > P Day 3 3 34.2857 .005 > P Day 4 3 29.8039 .005 > P Table 64. Chi square analyses of tho daily nesting area frequencies of juveniles for the eight treatment coubinations used in the field expeiment. df XZ-value Prob. Day 1 3 4.4444 .5 > P > .1 Day 2 3 4.0306 .5 > p > .1 Day 3 3 8.6275 .025 > p > .01 Day 4 3 5.0000 .5 > p > .1 170 Table 65. Chi square analyses of the daily nosting area ies of mole and female juveniles dnring tho field expeimont. The data were pooled for tho adult classes. df xz-valno Prob. Day 1 1 0.0000 .975 = P Day2 1 0.2423' .9>P>.5 'Day3 1 0.7843 .5>P>.1 Day 4 1 0.1732 .9 > P > .5 Table 66. Chi square analyses of the daily nesting area frequencies of juveniles dnringy tho field t. The data wee pooled for the juvenile sexes to examino diffeences among the adnlt classes. df xz-value Prob. Day 1 3 1.1310 .9 > P > .5 Day 2 3 1.9558 .9 > P > .5 Day 3 3 3.9200 .5 > P > .1 Day 4 3 2.5000 .5 > P > .1 Table 67. Chi square analyses of the daily nosting area frequencies of juveniles for tho eight treatment conbinations used in tho laboratory expeimont. df xz-value Prob. Day 1 3 14.4000 .005 > P Day 2 3 22.9333 .005 > P Day 3 3 17.3737 .005 > P Day 4 3 15.9399 .005 > P 171 Table 68. Chi square analyses of the daily nesting area frequencies of juvenile males and females for adult males during the laboratory experiment. df xz-value Prob. Day 1 1 1.1110 .5 > P > .1 Day 2 1 1.1110 .5 > P > .1 My 3 1 2.5000 .5 > P > .1 Day 4 1 2.5000 .5 > P > .1 Table 69. Chi square analyses of tho daily nesting area frequencies of juvenile males and fenales for nonlactating fenales during tho laboratory enqneriment. df xz-valno Prob. Day 1 1 4.2860 .05 > P > .025 Day 2 1 4.2860 .05 > P > .025 Day 3 1 4.2860 .05 > P > .025 Day 4 1 1.6666 .5 > P > .1 Table 70. Chi square analyses of tho daily nesting area frequencies of juvenile mnales and fenales for lactating fenales during the laboratory experiment. df xz-valno Prob. Day 1 1 0.4760 P z .5 My 2 1 1.1111 .5 > P > .1 My 3 1 1.1111 .5 > P > .1 My 4 1 0.0000 .975 = P Table 71 . 172 Chi. square analyses of the daily nestinng area frequencies of juvenile males and females for the controls*with no adult during the laboratory expeiment . df xz-value Prob. 1 1.6666 .5 > P > .1 1 2.5000 .5 > P > .1 1 0.0000 P = .975 1 4.2850 .05 > P > .025 Chi square analyses of the daily nesting area frequencies of juvenile males dnring tho laboratory experiment. df - xz-value Prob. 3 11.4670 .01 > P > .005 3 19.9990 .005 > P 3 14.7250 .005 > P 3 7.6920 .1 > P > .05 Chi square analyses of the daily nesting area frequencies of juvenile fenales dnring the laboratory eqneiment. df xz-valno Prob. My 1 3 2.4000 .5 > P > .1 Day 2 3 4.0000 .5 > P > .1 My 3 3 2.2222 .9 > P > .5 My 4 3 4.7619 .5 > P > .1 IlS'I‘OFREFEREl‘CES LISTOFREFEREI‘CE‘S Ambrose, H. W. , III. 1972. Effect of habitat familiarity and toe- clippinng on rate of owl predation in Microtus mlvanicus. J. Mum. 53:909-912. Baenninge, L. P. 1973. Aggression between weanling Peogzscus and Microtus. Anim. Behnav. 21:335-337. Barash, D. P. 1973. Tho social biology of the Olympic nonmt. Anim. Behnav. anogr. 6:171-245. . 1974. Tro evolution of meant societies: A general thoory. —_—S—cIence 185 : 415-420. 1977. Sociobiology and behavior. Elsevie, New York. Betts, B. J. 1976. Behaviour in a papulation of Colunbian ground squirrels , Speuophilus colunbiannus colnnnbianus . Anim. Behnav. 24: 652-680 . Birdsall, D. A. and D. Nash. 1973. Occurrence of successful multiple insemination of females in natural populations of dee mice (Pemscus maniculatus) . Evolution 27 :106-110 . Blair, W. F. 1942. Size ofhonerangeandnotesontholife history ofthomodlmddee—mnseandeastenchipnmkinnorthen Michigan. J. Mamm. 23:27-36. . 1953. Pepulation dynamics of rodents and othner small manuals. Adv. inGen. 5:1-41. Bovet, J. 1972. On tho social behnavior in a stable group of log- tailed field mice (Apodems sylvaticus) . II . Its relations with distribution of daily activity. Behaviour 41:55-67. Bradley, J. V. 1968. Distribution-free statistical tests. Prentice- Hall, Innc., New Jesey. Bradt, G. W. 1938. A study of beave colonies inhflchigan. J. Mamm. 19:139-162. Broadbooks, H. E. 1970. Bone ranges and teritorial behnavior of tho Yellow-pinochipmunk, Eutamias amoenus. J. Mann. 51:310-326. 173 174 Bronson, F. H. 1964. Agonistic behnavior inwoodchucks. Anim. Behav. 12:470-478. ad C. Desjardins. 1971. Steoid hormones ad aggressive m inmnmmals, pp. 43-63. _Inn_B. E. Elefthneiou and J. P. Scott (eds.) , Thne physiology of aggression ad defeat. Plenun Press, New York. ad H. M. Marsden. 1964. Male-induced synchrony of estrus in We. Gen. and Conn). findocrinol. 4:634-637. Bron, L. E. 1969. Field expeimnents on tho movements of sylvaticus L. using trapping ad tracldng technniqtos. ia Calhoun, J. B. 1963. The ecology and sociology of tho Norway rat. Bethosda, 1%.; U. S. Department of Health, Edncation, adWelfare. Carl, E. A. 1971. Population control in Arctic grood squirrels. Ecology 52:395-413. Chitty, D. 1967. The natural selection of self- tory behnavior in animal populations. Proc. Ecol. Soc. Annst. 2: 1-78. Christian, J. J. 1970. Social subordination, population density, ad mammalian evolution. Science 168:84-90. 1971. Fighting, maturity, ad population density in Microtus penngylvanicus. J. Mann. 52:556-567. , J. A. Lloyd, andD. E. Mvis. 1965. Tho role of edocriros in tho self-regulation of mammalian populations. Rec. Prog. I-br. Res. 21:501-578. Clark, F. H. 1938. Age of semnal maturity in mice of tho genus Peomyscus. J. Manm. 19:230-234. Conley, W. 1976. Competition between Microtus: A behnavioral hypo- thosis. Ecology 57:224-237. Coove, W. J. 1974. Some reasons for not using the Yates continuity correction on 2 x 2 contingency tables. J. Amer. Stat. Assoc. 69:374-382. Crow, J. F. and M. Kimura. 1970. An introduction to population genetics thoory. Rape and Row, New York. Croweoft, P. ad F. P. Rowe. 1963. Social organization and teri- torial behnavior in thewildmouse (ling-moans L.). J. Zool. 140:517-531. 175 Dice, L. R. 1932. Variation in a geographic race of tho dee-mouse, ngglégscus maniculatus bairdi. Univ. Mich. Occ. Papes Mus. Zool. . , 6 pp . 1936. Age variation in Peronyscus maniculatus Eacilis . W. 17:55-57. ad W. E. Hoard. 1951. Distance of dispesal by prairie —_de—emice from birthplaces to breeding sites. Contri. Lab. Vet. Biol. , Univ. Mich. No 50:1-15. Drabek, C. M. 1970. Ethnoecology of tho round-tailed gourd squirrel, Spermophilus teeticaudus . Diss. Abstr. 313:3764-3765. 1973 . Homo range ad daily activity of tho round-tailed ground squirrel , Spemophilus teeticandns roglectus , Amer . Midl . Nat. 39:287-293. Eisenbeg, J. F. 1962. Studies on tho behnavior of W maniculatus my; ad 2. californnicus parasit cus. 19:177-217. 1963. The intraspecific social behavior of sono eicetine ts of the gens Peomyscus. Ame. Midl. Nat. 69:240-246. . 1968. Behavior patterns, pp. 451-495. In J. A. King (ed.), Biology of Peromyscus (Rodentia). Spec. Publffio. 2, Anne. Soc. Mamm. Errington, P. L. 1939. Uosevations on young muskrats in Iowa. J. Mamm. 20:465-478. Ewe, R. F. 1973. The carnivores. Cornoll Ikniv. Press, Ithnaca, New York. Feguson, G. A. 1966. Statistical analysis in psychology and educa- tion. McGraw-Hill, New York, 2nd edition. Flowedew, J. R. 1974. Field and laboratory expeimnents on tho social behaviorandepulationdynamnics ofthowoodmouse (m sylvaticus). J. Anim. Ecol. 43:499-511.. Fordhamn, R. A. 1971. Field. populations of deemice with supplenontal food. Ecology 52:138-146. Foste, D. A. 1959. Diffeences in behavior ad tenpeament between two races of the dee mouse. J. Mamm. 40:496-513. Fur-re, R. K. 1973. hhning of Peomyscus maniculatus in tho Chamelled Scablands of east-central Washington. J. Mann. 54:466-482. 176 Gardne, R. H. and C. R. Terman. 1970. The relationship between age of gonping and weight of selected organs of prairie deemnice. Res. Pap. Ecol. 12:1-18. Gill, J. L. 1976. Statistics 422, Course notes. Mich. State Univ. Godfrey, G. K. 1953. A technnnique for finding Microtus nests. J. hhmm. 34:503-505. Grant, P. R. 1970. Enqneimontal studies of competitive inteaction in a two-species system. II. Tho behnavior of Microtus, Peroyscus ad Clethrionomnys species. Anim. Behnav. 18: 411- 426. . 1972.1ntespecific competition among rodents. Annnn. Rev. F261. Syst. 3: 79-106. Hamilton, W. D. 1964. Tho genetical thoory of social behavior, I, II. J. Thooret. Biol. 7:1-52. Harris, Van T. 1952. An expeimnental study of habitat selection by prairie and forest races of tho deemouse, P scus maniculatus. Contri. Lab. Vert. Biol., Univ. Mich. No. 56:166. Healey, M. C. 1967. Aggression and self-regulation of population size in deemnice. Ecology 48:377-391. Hill, J. L. 1970. Space utilization of Peogy'scus: Social and spatial factors. Unpublished Phn.D. disseation, Mich. State Univ. , East Lansing, 84 pp. hbward, W. E. 1949. Dispesal, amount of inbreeding, and longevity in a local population of prairie deermice on the George Reseve, zouihen Michigan. Contri. Lab. Vet. Biol., Univ. Mich. No. 3: -50. . 1960. Inmate ad environmental dispesal of individnal vertebrates. Ame. Midl. Nat. 63:152-161. Joule, J. and G. N. Caneon. 1975. Species renoval studies. I. Dispesal strategies of sympatric Signodonn hispidns ad Reithrodontomnys fulvescens populations. J. Mamm. 56:378-396. King, J. A. 1955. Social behavior, social organization, ad papulation dynamics in a black-tailed prairie dog ton in tho Black Hills of South Mkota. Contri. Lab. Vet. Biol., Univ. Mich. 67:1-123. . 1958. Maternal behavior and behavioral development in two Especies of Peonyscus maniculatus. J. Mamm. 39:177-190. . 1968a. Psycholo ,pp. 496-542. In J. A. King (ed.,) Bio logy of Peronyscus tia) . Spec. _Publ. No. ,Ame. 177 , (ed.). 1968b. Biology of Perogyscus (Rodentia). Spec. Publ. No. 2, Amer. Soc. Mamn. . 1973. Tho ecology of aggressive behavior. Annrn. Rev. Ecol. Syst. 4:117-138. Krebs, C. J . 1970. Microtus population biology: Behnavioral chages associated with the population cycle in M. ochrmaste ad M. pemnsylvanicus. Ecology 51(1) :34-52. , M. S. Gaines, B. L. Kelle, J. H. Myes, and R. H. Tanarin. 1973. Population cycles in small rodents. Science 179:35-41. Layne, J. N. 1954. The biology of the red squirrel, Tamriasciurus hnudsonicus M (Bangs), in central New York. Ecol. Moog. . 1968. Ontogeny, pp. 148-253. In J. A. King (ed.), Biology of Peronnyscus (Rodentia). Spec. PublT-No. 2, Amer. Soc. Mamm. Levine, S., G. C. Haltmoye, G. G. Karas, and V. H. Denenbeg. 1967. Physiological ad behavioral effects of infantile stimulation. Physiol. Behav. 2:55-59. Lidicker, W. Z. 1962. Ehmigation as a possible mechanismn permitting the regulation of population density below carrying capacity. Amer. Nat. 96:29-33. 1973. Regulation of nunbes in an island population of tho 4 ifornia vole, a problem in comnnmity dynamics. Ecol. Moog. 3:271-302. . 1975. The role of dispesal in tho derogaphy of small mammals, pp. 102-128. In F. B. Golley, K. Petrusewicz, ad L. Ryszkowski (eds.) , Small—mammals: Their prodnctive and population dynamics. Cambridge Univ. Press , Cambridge. Mettle, L. E. and T. G. Gregg. 1969. Population genetics and evolu- tion. Prentice-Hall, Inc., New Jesey. Metzgar, L. H. 1967. An expeimontal conparison of seeech owl pre- dation on resident ad transient white-footed mnice (Peonyscus M). J. Mamm. 48:387-391. . 1971. Behavioral population regulation in tho woodmouse, Pemcus leicopus. Amer. Midl. Nat. 86:434-448. Michneor, G. R. 1973. Intraspecific aggression ad social organization in ground squirrels. J. Mamm. 54:1001-1003. and D. R. Michene. 1973. Spatial distribution of yearlings E a Richardson's good squirrel papulation. Ecology 54:1138- 1142. 178 Ivbye, K. E. 1971. A prelimninnary physiological model of aggessive behavior, pp. 223-263. In B. E. Eleftheiou ad J. P. Scott (eds.) , 'Ihne physiology of—aggession and defeat. Plenum Press, New York. Mnrray, B. G. 1967. Dispesal in vetebrates. Ecology 48(6):975-978. Myes, J. H. and C. J. Krebs. 1971. Geotic, behavioral, ad repro- dnctive attributes of dispesing field voles, Microtus pennsylvanicus and Microtus ochrogaste. Ecol-W Moog. 41: 53- 78. Myton, B. 1974. Utilization of space by Peonyscus leucopus and othe smnall manuals. Ecology 55:277-290. Nicholson, A. J. 1941. The hnomes ad social habits of the wood mouse W2 3%3mveboracensis) in southern Michigan. Amer. Orr, H. D. 1959. Activity of white-footed mice in relation to environ- mnent. J. Mamm. 40:213-221. Payne, A. P. and H. H. Swanson. 1970. Agonistic behavior between pairs of hamstes of tho sane and site sex in a neutral observation area. Behavior 36:259- 69. Poole, T. B. and H. D. R. Morgan. 1973. Diffeences in aggessive behnaviorbetweenmalemdce (Mnsnnnsculus L.) incolonies of dif- feent sizes. Anim. Behnav. 21: 788- 795. . 1975. Aggessive behnavior of male mice (Mus musculus) towards familiar and unfamiliar opponents. Annim. Behav. 23: 470- 479. Rasmussen, D. I. 1964. Blood goup polymorphism and innbreeding in natnral populations of tho dee mouse Peomyscus maniculatus. Evolution 18:219-229. Richards, S. M. 1974. The concept of domninance and methods of assess- ment. Anim. Behav. 22:914-930. Rinntamaa, D. L., P. A. Maznr, and S. H. Vessey. 1976. Reprodnction dring two annual cycles in a population of Peonyscus lencgpus noveboracensis. J. Mamm. 57:593-595. Rogstad, 0. J. 1965. A life history study of thirteen-lined gomnd squirrels in Southern Wisconsin. J. Mamm. 46:76-87. Rose, R. K. ad M. S. Gaines. 1976. Levels of aggression in fluctuat- ing papulations of the prairie vole, Microtus ochrogaste, in easten Kansas. J. Mamm. 57:43-57. Russell, P. A. 1971. "Infantile stimulation" in rodents: A conside- ation of possible mochanismns. Psychol. Bull. 75:192-202. 11’ Ill him. ”. 1|?! in I. .lnfl .. ll 179 Sadleir, R. M. F. S. 1965. The relationship between agonistic behnavior and population changes in the deemouse, Peonyscus maniculatus (Wagne), J. Anim. Ecol. 34:331-352. . 1970. The establishment of a dominance rank orde in male I%—W maniculatus ad its stability with time. Anim. Behnav. Savidge, I. R. 1970. Social factors contributing to tho departnre of Pemscus mnaniculatus bairdi from their natal site. Unpublished .D. isseation, Mich. State Univ., East Lansing. . 1974. Social factors in dispesal of deer mnice (P maniculatus) from their natal site. Amne. Midl. .flzggg- 261185. Scott, J. P. 1972. Animal behnavior. Univ. Chicago Press, ChicagO, 2nd edition. . 1973. Comparative social psychology,pp 124-160. In D. A. Dewsbnry and D. A.Rethlingshafe (eds. ), Conparative psychology: A modern survey. .McGraw-Hill, New York. Scudder, C. L., A. G. Karczmar, and L. Lockett. 1967. Behavioral developmental stndies of for genea ad seveal strainns of mice. Anim. Behav. 15:353-363. Smith, C. C. 1968. The adaptive natnre of social organization in the genus of tree squirrels Tamiascinrus. Ecol. hbnogr. 38:31-63. Sokal, R. R. ad F. J. Rohlf. 1969. Biometry the prinnciples and prac- tice of statistics in biological research. W. H. Freeman ad Co. , San Francisco. Steiner, A. L. 1972. hbrtality resulting from intraspecific fighting in sono gomnd squirrel populations. J. Mamm. 53:601-603. Stickel, L. F. 1946. The sorce of animals moving into a depopulated area. J. Mamm. 27:301-307. . 1968. Home rages and travels, pp. 372-411. _In_ J. A. King (Ed. ), Biology of Pmsm (Rodentia). Spec. Publ. No. 2, Amer. .Mamm. Terman, C. R. 1962. Spatial and honing consequences of tho introdic- tion of aliens into semi-natnral papulations of prairie deermice. Ecology 43:216-223. . 1963. The influence of diffeential early expeience upon spatial distribution within populations of prairie deemice. Anim. Behnav. 11:246-262. 180 . 1968. Population dynamics, pp. 412-450. In J. A. King zen), Biology of Peonyscus (Rodentia). Spec. Pfibl. No. 2, Amer. Soc. Mamm. ad J. F. Sassanan. 1967. Sex ratio in dee mouse popula- tIB-ns. J. Mamm. 48:589-597. Thiessen, D. D. 1973. Foothnolds for survival. Amne. Sci. 61:346-351. Thorns D. and C. R. Teman. 1975. The effects of diffeential pre- natal and postrnatal social environments on sexual maturation of yong prairie deemnice (Pegyscus maniculatus bairdii). Amer. Behnav. 23:241-248. T‘rives, R. L. 1974. Parent-offspring conflict. Amer. 2001. 14:249- 264. Van Valen, L. 1971. Gronp selection ad the evolution of dispesal. Evolution 25:591-598. Watts, C. H. S. 1969. Regulation of woodmouse ( modean 3 1vaticus) nunbes in WythannWoods, Berkshire. J. Anim.fl.m:§85-3M. . 1970. A field expeiment on intraspecific inteactions in HE red-backed vole, Clethrionomys gappei. J. Mamm. 51:341-347. Wecke, S. C. 1963. The role of early expeience in habitat selection by the prairie dee mouse, Perogyscus mnaniculatus bairdi. Ecol. honogr. 33(4) :307-325. Williams, C. C. 1966. Adaptation ad natLral selection. Prinnceton Univ. Press, Prirnceton, New Jesey. 1975. Sex and evolution. Princeton Univ. Press, Princeton, Jesey. Wilson, S. 1973. The deve10pment of social behavior in the vole (Microtus agestis). J. Zool. 52:45-62. Winne, B. J. 1971. Statistical principles in experimental design. McGrav-Hill, New York, 2nd edition. will”!11111111111111”: £32114”: 3129’