IIWIHIIIWIHIWWIll!WWll!IIHIWIJHIIIHHII _cm\>co 7‘4ng This is to certify that the thesis entitled 'I‘HEEFFECI'SOFDAYIMANDTMERA'IIRECN'IHE HIBERNATINGRHYTEMOF'IHEMEAWJUMPMIVDUSE (@PUSHUDSG‘IIUS) presented by Alan E. Muchlinski has been accepted towards fulfillment of the requirements for Ph.D. degree in Zoology film; Major professor Date 5 July 1979 0-7639 OVERDUE FINES ARE 25¢ PER DAY > PER ITEM Return to book drop to remove this checkout from your record. l‘l llfl‘ll‘l‘lltll'lr'lllll“! I." 'HIEEFFECI'SOFDAYIE‘CD-IAND'I'D’IPERAIUREONTHE WWOFMWMMWSE (ZAPUSHUDSONIUS) BY .Alan.E.IMuchlinSki A.DISSEKDNEBJN submitted to Nfichigan.State'University in.partial fulfillment of the requirements for the degree of DOCTOR.OF PHILOSOPHY Department of Zoology 1979 ABSTRACT 'IHEEFFECI‘SOFDAYLMAND'IE'IPERAIURECNTHE PHEERNATII‘ERIWOFTHEWJMDBMJUSE (ZAPUSHUDSG‘IIUS) By Alan E. Muchlinski Results fran four treatmmt ooubinations involving different levels of daylength (ID 15.33:8.67, U) 12:12) and temperature (5 and 20°C) damnstrated that significantly more animls prepared for and entered into hibernation when held under short daylengths than under long daylengths (P < 0.001). Within a photoperiod level, temperature had no effect upon the nunber of animals initially preparing for and entering into hibernation (P > 0.20) . Rhythms in hibemating- nonhibernating activity were demcmstrated only under short daylengths but the period 1mgths were not the same at the two temperatures (332.0 :24.3 SD days at 5°C vs 157.2 260.9 SD days at 20°C, P < 0.005). In addition, cycles in body weight under ID 12:12, 20°C were shorter than the cycles in hiberrmtion—norflmibernation activity ( 57.2 160.9 SD days vs 102.8 :71.2 SD days, P < 0.05). In a fifth treatment canbination, animls were exposed to three-mnth clock- shifted , gradually decreasing daylmgths (sinnlated June 21 daylength on Septenber 21) . Seven of the eleven animals prepared for and six of these entered hibernation at sinulated dates corresponding to the time span animals in nature undergo these changes. Two animals prepared for and entered hibernation two Alan E. Muchlinski weeks aheadofthefirst occm‘renceseeninnatureandtmanimals underwent these changes very early. /'I‘hese results indicated that meadow junping mice are using the decreasing daylengths of late surmer and fall as a stimilus for preparation and initiation of hibernation./ This differs from t1~e stinnlus used by Zapus princeps. Arousal of field animals is correlated with soil tarperature . The first males emerged fran hibernation when the soil temperature at 100 cm equals approximately 7°C while the first females aroused two weeks after the males when the soil temperature at the same depth approxi- mated 9°C. The following scenario is then proposed for Zagis hudsonius. The long daylengths of spring and smmer stinnlate reproduction which con- tinues through mid-August. Beginning in late August, the decreasing daylengths begin to stinnlate animals to prepare for and enter into hibernation and this continues through mid to late October. Hibernation then continues until mid-April to early May depending upon the level of the soil temperature around the hibernaculun. When the soil tempera- ture increases to a certain level, the animals energe from their hibernaculun and beccme active for the sumer. 'I‘nerefore, it appears that two factors are important in the yearly cycle of Zapus hudsonius . Photoperiod signals the end of the active interval and, possibly, soil tanperature signals the end of the hibernation interval. I wish to thank the members of my guidance carmittee, Richard W. Hill (Chairman), Rollin H. Baker, Donald L. Beaver, and Lester F. Wolterink, for their assistance during my stay at Michigan State University. Controlled enviroment roans were nede available to me byJackKingdeACSandIthankthanfortheir support. Thefield portion of this research project was conducted on land made available by the Institute of Water Research at Michigan State University . Much of the credit for the success of this project must go to my wife, Michele, who helped me through many hours of laboratory and field work. ii TABIEOFCONI‘ENI‘S Page Ackmwledgeients ......................... ii List of Tables .......................... v List of Figures ......................... vi INTRODUCTION ........................... 1 MATERIALS AND MEI‘EKDS ...................... 7 Laboratory Experiments ..................... 7 Field Procedure ........................ 15 RESULTS ............................. 18 laboratory Results ....................... 18 ID 15.33:8.67, 20°C ..................... 22 ID 15.33:8.67, 5°C ...................... 22 ID 12:12, 20°C ........................ 25 ID 12:12, 5°C ........................ 32 Clock- Shifted Photoperiod Ehcperiment ............. 43 Field Results: Timing of Events ................ 46 Field Results: Survival .................... 53 DISCUSSION ............................ 58 Photoperiod .......................... 59 Presence of an Ehdogenous Circannnual Rhythn ......... 59 Other Types of Endogenous Clocks ............... 61 Endogaious Rhythns Not Present ................ 64 Syntl'esis .......................... 65 iii iv Tarperature .......................... Ehrelution ........................... BIBLIWRAPHY ........................... Page 66 67 74 Table l . Table 2 . Table 3 . Table 4 . Table 5 . Table 6. Table 7 . Table 8 . Table 9 . LISTCFTAELB Animals which survived mre than 137 days under ID 15.33:8.67, 20°C ................. Days spent in the hibernating or active state over the course of the experiment for those animals that hibernated at least once. ID 12:12, 20°C ...... Period lengths for weight cycles and maximum-minim weights for animls that went through at least two weight gains under ID 12:12, 20°C .......... Days spent in the hibernating or active state over the cause of the experiment for those animals that hibernated at least once. ID 12:12, 5°C ....... Period lengths for weight cycles and madam-minim weights for animals that went through at least two weight gains under ID 12:12, 5°C ........... Summary of T.C.'s SW, LW, SC, LC ........... The date of first appearance above ground by males and females in the spring .............. Distribution of first captures in relation to soil temperature (hiring the spring ............ Maxinun weight to mximm weight period lengths for four animals captured after undergoing weight increases in two consecutive years .......... Page 24 26 31 33 40 42 46 49 56 LISI‘ OF FIGURES Page Figure 1. Experimental design and results .......... 8-9 Figure 2. Mean maximm body weight attained within 137 days inthefour laboratoryT.C. 'sandinasarrpleof 56 field animals .................. 20-21 Figure 3. Body weight graphs for four animals frcm ID 15.33: 8.67, 20°C ..................... 23 Figure 4. Body weight graphs for four animls fran ID 12:12, 20°C ........................ 27 Figure 5. Body weight graphs for four animls fran ID 12: 12, 20° ........................ 28 Figure 6. Body weight graphs for four animals from ID 12:12, 20°C ........................ 29 Figure 7. Body weight graphs for four animls fran ID 12:12, 5°C ........................ 35 Figure 8. Body weight graphs for four animals fran ID 12:12, ........................ 36 Figure 9. Body weight graphs for four animals from ID 12:12, 5° ........................ 37 Figure 10. Body weight graphs for four animals from ID 12: ~12, ........................ 38 Figure 11. Canparison of weight cycles for one animal at ID 12:12, 20°C (solid line) and one animal at ID 12:12, 5°C (broken line) ............ 41 Figure 12. Respmse of animals to clock-shifted photo- periods ...................... 44-45 Figure 13. Relationship of soil temperature at 100 cm to arousalofmalesandfanales inl977and1978 . . . 48 Figure 14. Figure 15 . Figure 16. Page Percentage of reproductive males arnd females and percent of pregrnant fareles during 1977 trapping season ....................... 50-51 Distribution over time of preparation for hiber- nation in field arnimals .............. 52 Minimrn nurber of animals alive on demgrqnhic grid ........................ 54-55 INI'RDDUCI‘ION Hibernating mammals prepare for end enter into hibernation during a certain time span each year. This preparation for hibernation can take many forms. Sane species cache food in their burrows arnd feed on these stores intermittently during the winter while others undergo drastic weight gains, thereby storing energy in the form of body fat. Successful hibernation can only take place after a sufficient amount of energy has been stored by either or both means. Increased energy reserves are not the only factor involved in the preparation for hibernation. Recent evidence has indicated that many biochemical changes are associated with the preparation for hibernation (Behrisch, 1978) . These changes involve the production of new isozymes capable of maintaining body functions at the low body temperatures achieved during hibernation. Whatever the changes taking place before hibernation, it would be beneficial for a hibernating mammal to begin to prepare for hibernation well before adverse environmental connditions occur. Hypotheses con- cerning the timing of preparation and initiation of hibernation have taken several forms. lbck (1955) suggested that tl'e decreasing day- lengths of late summer and autumn served as the stimulus for the onset of hibernation in the arctic grournd squirrel , Spermphilus urndulatus . He reasoned that an environmental factor must be predictable arnd dependable if it is to be used as a stimulus for preparation and initiation of hibernation, and daylength seemed to be the most depend- able envirormental factor. A test of Hook's hypothesis on the golden—mantled gronmnd squirrel, Spermophilus lateralis, led to a different cornclusion. Pengelley and Fisher (1957) reported that when individuals of this species were main- tained under constant cornditions of photoperiod and temperature for up to four years, hibernating activity alternated with rnonhibernating activity in a pattern samewhat similar to that occurring in nature. These results, later to be expanded by ecperiments utilizing many dif- ferent photoperiod and temperature cornditiorns (Pengelley, 1965; Pengelley and Asmundson, 1969, 1970; Pengelley and Fisher, 1963; Pengelley g Q- , 1975; Scott and Fisher, 1970), led to a hypothesis invoking endogenous rhythnicity. Tm phenomena were noted in regards to this endogenous cycle. First, the period of the rhythm was not exactlyayear inlength. Inmrnst cases theperiodwas less thanone year. This led to the coining of the term "endogenous circannual r hm" which indicates an intrinsic rhythm with a period of approxi- mately one year. Secorndly, the period of the rhythm proved to be temperature independent or nearly so; the length of tie period at 12°C was not statistically different from that at 3°C. This aspect of temperature independence was cornsidered important because it was analogous to the terperatnn'e independence of circadian rhythm and indicated the possibility of an endogenous cellular basis for the rhythm. I-bwever, Brown (1976) has taken this aspect of temperature independence to mean that the rhythm is not endogenous. He bases this on tl'e fact that all know cellular chemical reactions are not terperature independent. I recognize this dichotomy but since I am not atterpting to determine the basis of any possible rhythm, I will use Pengelley and Fisher' 3 description of the rhytl'm for comparative pur- poses. At the present time, no entraining factor (6) (Zeitgeber) has been determined through experimentation to entrain a circamnual rhythm to a period of 365 days. It had been demonstrated by different researchers (see Mrosovsky, 1978 , for a camprehensive reviev) that endogenous (c ircannual rhytl'm) and exogenous (photoperiod ad temperature) factors vary in their importance for different species of mammalian hibernators. It was my purpose in conducting these experiments to determnine the relative roles that endogenous ad exogenous factors play in the preparation for ad entrance into hibernation in the meadow jmping mouse, Zapus hudsonius. Mich of the early research corducted on hibernating mammals dealt with rodents of the squirrel family, Sciuridae, ad this group is where the concept of endogenous circanrmial rhythmicity was developed . Recent research (Cranford, 1978; Mrosovsky, 1977) has deronstrated that some species in different rodent families do not show endogenous circamnual rhythms. This study on the meadow jumping mouse represents the secord look at a member of the Family Zapodidae. Answers to tie following questions were sought: (1) Is the hibernation-mnhibernation cycle and/ or weiglnt cycle in hpus hudsonius governed by a circamnual rhythm of the type described by Pengelley ad Fisher (1957); (2) If a circamnual rhythm as previously described is mt present, are other types of rhythmicity demonstrated ad nmnder what set of controlled conditions might they be expressed; ad (3) What role do photoperiod ad/or terperature play in controlling the preparation for the entrance into hibernation? In particular, I sought to determine if the occurrence of preparation and initiation of hiber- nation was more prevalent under certain corditiorns than others. This wcnuld give insight into the role of photoperiod ad/or temperature in the preparation and initiation of hibernation. Depeding upon which factor (if any) was mere important, I sought to determine if that factor could be used as a stimilus capable of providing information concerning time of the year. To facilitate the urderstading of events occurring during hiber- nation, several terms must be defined. The hibernation interval is that time fram the first evidence of hibernation in the fall to the last evidence of the state in the spring. Ibever, the animal does not retain in hibernation continuously over this time span as arousals and re—entries occur at intervals whose length appears to be characteristic for a species at a specified envirormental temperature (Ta) (Pengelley ad Fisher, 1961; Twente and 'Iwente, 1965). The time period between arousals will be defined as a hibernation bout. Arousals occurring during the hibernation interval will be classified as intermittent arousals whereas the arousal that ends a hibernation interval will be alludedtoasthetermninalarousal. Thetimefranterminalarousalin the spring to first entrance into hibernation in the fall is termed the active interval. The hibernation interval plus the active interval camprise the hibernation-mnhibernation period. Field data can also give some insight into the hibernating life history strategy. In particula', live trapping studies can give insight into the dates of emergence fran ad entrance into hibernation, the hibernation-renhibernation period length, the timing or repro- duction, ad the mortality rate over the hibernation interval. Two irdepth natural history monographs have been written on the meadow jurping mouse (Quimby, 1951; Mnitaker, 1963) as have many shorter notes (Babcock, 1914; Blair, 1940; Dilger, 1948; Hamilton, 1935; Manville, 1956; Steldon, 1934; Whitaker and Mumford, 1971), but these have not dealt in depth with the aforementioned topics for individual animals. For example, information is lacking concerrning the hibernation-mnhibenation period length of individual animals in the wild. Comparison of data on this point frcm the field with those obtairned in the laboratory is important in any discussion concerrning a circamnual rhythm. If the period lengths from the laboratory and field animals are both circamnual, this would mean that the cycles of the field animals are not being entrained to a period of eactly 365 days. An analysis of the dates of entrance into hibernation is important in determining possible modulation of dates of entrance by 1a The number of litters produced per year and their timing is also important for a hibenating species. While rnornhibernators may extend their breeding season from early spring to late fall, ad in some cases through the winter , hibernators must complete their breeding within the brief time span of late spring to late sumer. Juveniles are reported to be the last animals to enter into hiber- nation (Quimby, 1951), being preceded sequentially by the male and female adults. Data fram live trapping will be able to discern if there is differential mortality between these groups. It is of inter- est to krnow what proportion of these various groups survive to consti- tute the following year's population. In particular, it is of interest to note the life spans of meadow junping mice and the relation of life span to the presence or absence of an endogenous circamnual rhythm. MATERIAISANDMEITDDS Laboratory Engperiments Over a three-year period, 92 meadow jumping mice (2% hudsonius) which had either been live-trapped near East Lansing, Michigan, or born to captured females or mated pairs from tie same area, were assigned to five treatment combinations (T.C. '3) involving manipulations of photOperiod ad temperature. Of these five T.C. 's, the first fonn' camposed a completely radam statistical design with a 2 x 2 factorial arrangement of treatments (Figure l) . The factors used in this arrangement were length of photoperiod and level of tempera- ture, each being set at moo levels. The photoperiod levels were ID 15.33:8.67 and ID 12:12, while tle temperature levels were 20° ad 5°C. The ID 15.33:8.67 photoperiod represents that photoperiod observed at the latitude of East Lansing, Michigan, on the longest day of the year, Jane 21. By assigning only animals which had been cap- tured before or within a 14-day period after June 21 to the two treat- ments involving this long photoperiod, it was lmnowrn that the irdi- viduals had not been subjected to any appreciable decrease in daylength during their active phase that particular sumer (7 minutes maximum). The shorter photoperiod was chosen because it had been used in most previous circamnual rhythm experiments (Pengelley, 1965; Pengelley and Asmurdson, 1969, 1970; Pengelley and Fisher, 1963; Pengelley gt _a_._1_. , Figure 1. Experimental design and results. Number in center of each box represents the mmber of animals surviving to 137 days intoeachTHC Pequalsthemrnberofanimalspreparing for hibenation, while H equals the number of animals hibernating out of those animals surviving for at least 137 days. 3 equals ID 12:12, L equals ID 15.33:8.67, C equals 5°C, ad W equals 20°C. TEMPERRTURE 01 F) 20C PHOTOPERIOD LD 12:12 LD 15.33:8.67 T.C. SC 18 P:18 H216 T .C. LC 12 T.C. SN 13 P 12 H: 9 10 1975; Scott ad Fisher, 1970). In this very, direct canparisons could be made between Zapus hudsonius ad other hibernating species. The two tempeature levels were chosen to give a substantial difference in Ta for comparison between treatments . As stated previously, terpeature independence is one property of the circamnual rhytrm described by Pengelley ad Fislner (1957). It was known that hibernation was pos- sible in this animal at both 20° and 5°C (Muchlinski, unpublished data). Therefore, the 15°C tenpeature difference allowed a test of the tempeature indepedence hypothesis . Twenty individuals were assigned to each of these four T.C. '3. Because of the inability to capture 80 animals in one surmer, all four T.C. '3 could not be begurn at the same time. Therefore, the following procedure was used. The first 20 animals (14 males, 6 felales) cantured before or shortly afte June 21, 1976, were assigned to T.C. 1W (ID 15.33:8.67, 20°C). The dates of capture for these animals ranged from May 1 to July 2, 1976. Mortality was high in this T.C. so four additional ani- mals (3 males, 1 female) were put urde this regime on July 5, 1977. The surviving mnembes of the 1976 group wee maintained under these constant conditions for 749 days, while tlne four individuals started in 1977 wee run to a maximum of 370 days. By late August 1976, it was possible to capture a substantial num- be of additional animals, and these wee used to establish T.C.‘s SC (ID 12:12, 5°C) ad SW (ID 12:12, 20°C). Both of these T.C.‘s wee begun on August 24, 1976, ad were canposed of animals captured between July 3 and August 23, 1976. Assignment to these two T.C.‘s was made on basis of date of capture, sex, and age of the individual. mey second 11 individual in orde of capture or birth was assigned to T.C. SC with sauna mine adjustments made to ensure that the numbe of males, fenales, adults, ad juveniles was equal between T.C.‘s (T.C. SC: 9 males, 11 ferales, 5 juveniles included in these mnbes; T.C. SW: 10 males, 10 ferales, 4 juveniles included in these numbes). Those individuals camposing T.C. SW were put unde the prescribed conditions on August 24 while those composing T.C. SC were subjected to ID 12:12 starting on August 24 but were exposed to stepwise drops in Ta . The animals wee loweed fran room tenpeature (approximately 20°C) to 10°C for one week ad then to 5°C. In these two T.C.‘s, sur- ' vivors were maintained urde cornstant conditions for a maximum of 686 days. Treatment combination LC (ID 15.33:8.67, 5°C) was begun on June 21, 1977. Animals used in this T.C. were eithe captured between May 20 ad July 5 or born to mated lab pairs before June 21. As in T.C. SC, stepwise drops in Ta from room terpeature to 5°C wee implemented. This T.C. was accidentally temninated 137 days afte initiation because of the introduction of a toxic substance into the environmental chambe. The majority of the animals wee killed, ad those that sur- vived wee moved to anothe room theeby eding the treatment . In the case of all animals used in the T.C.‘s, an attmt was made to subject animals to similar corditiorns before they were assigned to their respective T.C. As individuals were trapped, they were brought into an animal colony room. Thee the animals wee subjected to room tempeatures (20°-25°C) and simulated natural daylengths using an astronamical timer (Paragon Astro Dial). All animals wee given food (Wayne lab Blox) ad wate ad M. Inn addition, sunflowe seeds 12 were given to all animals on the day of capture. laboratory-born ani- mals came from pairs housed in this same room. All animals were housed individually in plastic cages measuring 27.9 x 17.8 x 12.7 cm or 26.7 x 20.3 x 15.9 cm and wee given a nest box measuring 9 x 9 x 8.2 cm lined with cotton for females ad wood shaving for males. Wood shavings coveed the floor of all cages. Afte assignment to its respective T.C. , each animal was main- tained in a walk-in environmental chambe unde the constant conditions of that particular T.C. Lmntil expiration of tie individual or temi- ration of the expeiment. The Ta in each T.C. was maintained within 1.5°C at the set level, while the light intensity measured at the top of each cage was maintained at approximately 85 111x. All nnonhiberrating animals wee weighed to the nearest 0.5 gram at one week intevals for the first year of the T.C. ad evey second week thereafte. Reproductive condition was also noted at this time with males being scored as testes scrotal, barely scrotal, or nonscrotal while fenales were scored as having easily visible nipples or non- visible nipples. Hibernating animals wee weighed only during or shortly afte an intentittent arousal. This minimized disturbances which may have alteed the length of time spent in hibenation. Hibernation activity was monitored by the use of tracking plates placed in front of the rest box opening. The plates wee coveed with a mixture of alcohol and baby powde, ad evaporation of the alcohol left a smooth surface of talc which would be disturbed by the animal if it left or enteed the nest box. The plates wee checked once pe day, ad the animal was scored as active for the preceding evening if the plate had been distebed or hibernating if the plate had not been l3 disturbed. The first time a plate was found intact, the animal was checked to cornfirm the hibernating state. A continuous record of hibenatirng or rnonhibenating activity was maintained ove the course of each T.C. Seveal powe interuptions and emrironnmental roan breakdowns did occur ove the course of the T.C.‘s. A power outage occurred for approximately two hours in T.C.‘s SL, SW, ad LW on Octobe 3, 1976. Also, a tempeature breakdown occurred in T.C.‘s SW ad 1W for five days starting on January 21, 1977, allowing Ta to rise to 26°C. No tempeature breakdowns occurred in the 5°C T.C.‘s. Therefore, animals in T.C. '3 SW ad 1W were neve exposed to Ta's below 20°C ad animals in T.C.‘s SL ad LC wee neve exposed to Ta's above 5°C. Treatment combination 5 was established in August 1978 to test the influence of a clock- shifted, naturally decreasing plotopeiod on the preparation for and entrance into hibenation. The 12 animals can- prising this T.C. wee captured between July 13 ad August 18, 1978, from tlne East Iansing, Michigan, area. Eight of the animals, thatwere captured before July 20, were placed on ID 15.33:8.67, 20°C on this date. Prior to this date and since the date of capture, the animals were housed in a colony roan maintained at 20°—27°C ad TD 14:10. Four animals wee captured on August 18 ad were immediately placed urde ID 15.33:8.67, 20°C. Each animal was housed individually in a cage measuring 29.8 x 24.1 x 20.3 an. Wood shavings were placed to a depth of 12 cmwhich allowed the animal to burrowunde the shavings for cove. This set-up was used because it had been determined through obsevations that large cages with excess bedding mateial reduced mortality. 14 All animals wee maintained urde tl'ese conditions until Septenbe 21, 1978, at which time they were three months out of phase with the natural environment. An astrornomical time (Paragon Astro Dial, set for latitude 42°N) was set to simulate the June 21 photopeiod on Septembe 21 and theeafte the simulated photOpeiod continued to lag three mornths behind the natural environment. He tempeature was main- tained at 20° rl.5°C at all times, ad tl'e light intensity was approxi- mately 85 1ux. As in T.C.‘s 1W, IE, SW, ad SC, animals wee given food (Wayne Iab Blox) ad wate ad M. Beginrning on August 18, animals in T.C. 5 wee weighed, and repro- ductive condition was ascetained, once pe week. As an animal was obseved to urdego a weight irncrease indicating preparation for hibe- nation, tracking plates wee set into the cage to check for hibernating activity. These plates were then checked daily to ascetain the first day of hibernation. As in the previous T.C.‘s, the animal was checked . at the first sign of a nordisturbed tracking plate to visually obseve the hibernating state. Since the pnupose of this T.C. was only to detemine if decreasing pl'notOpeiods were capable of stimulating pre- paration for ad entrance into hibernation, no continuous record of hibernating activity was maintained afte the initial entrance into hibernation. On Octobe 17, the ovehead neon lights wee innadvetently turned on, and this was not discoveed urntil Octobe 21. The simulated dates for this occurrence were July 17 to July 21. One powe intennption occurred for 25 minutes at approximately 4:00 a.m. on Noverbe 29. The lights were not readjusted. 15 Statistical canparisons among T.C.‘s SC, 18, SW, ad 1W wee based mainly on nonparametric aalyses although parametric tests were used in several instances . Nonparametric procedures wee favored because of the small sample sizes present in most T.C.‘s, the high incidence of nonnequality of variance, and the non-normal distributionn of data. In tl'ose cases where it was detemined that variances wee equal ad normality was present, parametric tests wee used. Normality was tested for by the chi-square goodnness of fit for continuous distri- bution test (Steel and Torrie, 1960) , ad the F-test was used to dete- mine equality of variance (Snedecor ad Cochran, 1967) . Field Procedure In June 1976, a 1.38 hectare live-trapping grid was established at the Institute of Wate Research at Michigan State Univesity, Inghan County, Michigan, for the purpose of gathering demographic data related to the hibenating life history strategy of Zanus hudsonius. The main trapping grid was canposed of 144 Longworth live traps set in 18 rows (25 feet apart) with eight traps (50 feet apart) pe row. In the spring of 1977, two rows (100 feet apart) with eight traps pe row wee setas auxillary lines to thewest of themaingridwhile threerows (100 feet apart) with eight traps pe row wee set to the east of the main grid, giving an additional 1.63 hectares of trapping area. Themaingridwas opeatedontwo connsecutivemornings adthe inteveninng afternoon each week. Trapping was begun in mid-April before emergence of the mice from their hibernacula ad continued through the fall until two weeks went by with no captures. The latte circumstance signnified that all animals residinng on the grid had taken 16 up residence in their hibernacula for hibernation. The auxillary trap lines were rnmn for four weeks in the spring afte energence of the ani- mals ad for four weeks in the fall before the majority of the animals had taken up residence in their burrows. This was an attempt to increase the numbe of animals marked for estimatinng ovewinteinng smr- vival. All traps were baited with wnole oats, cotton was placed in the rear portion of the trap for nesting mateial, ad cove boards wee placed on top of the traps. Individual meadow jumping mice wee marked by toe clippinng and weighed, ad note was taken of their reprochnctive condition. As in the laboratory, males wee categorized as testes scrotal, barely scrotal , or abdonninal, ad ferales wee categorized as nipples easily visible or nonvisible. In addition, female reproductive condition was noted by the condition of the vaginal orifice (peforate or nnonnperforate) ad pregnant females were discennable by obviously bulging abdomens. Lac- tating females could be noted by tne size of the nipples. Individuals newly captured before July 1 each year were classified as adults at the time of capture. Afte July I, new individuals weigh- ing less than 15 g at the time of first capture wee classified as juveniles. Animals caught for the first time afte July 1 ad weighing more than 15 g could not be placed in either category with 100 percent cetainty until the following year when they could be classified as adults. The trapping grid was opeated in the snmner ad fall of 1976, the sprinng, sunme, ad fall of 1977 ad 1978, ad the spring of 1979. Data on the following paranetes wee gathered: (1) minimum mube alive, (2) reproduction, (3) dates of emergence frann ad entrance into 17 hibernation, ad (4) survivorship ove the hibernation inteval. A denographic analysis campute package maintained by Dr . Walt Conley, Department of Fishey and Wildlife Sciences , New mam State Univesity, was used to analyze portions of the data. RESULTS laboratory Results Aninals were obseved to prepare for (inncrease in weight to 3 25 g) and ente into hibenation in all four T.C.‘s camposinng the factorial arrangement of treatments. Howeve, the numbe of indi- viduals udegoing these changes ad their timing varied between T.C. '3. Because of the early temination of T.C. LC, statistical com- parisons concerning the effect of plotOpeiod and/or temperature on the first preparation ad initiationn of hibenation can only be made up to 137 days into each T.C. I-bweve, this alters the outcame of the analysis little, for in T.C.‘s SC, SW, ad IN canbined onnly a sinngle animal enteed hibenation afte 137 days (animal 47 in T.C. IW, which was first obseved in hibenation on day 171). In T.C. LC which was teminated early, the only animal to ente hibenation did so on day 70, while the same animal and a second reached peak weights on days 57 ad 43, respectively. It is unlikely that any additional animals would have eithe prepared for or enteed into hibernation in T.C. IC. The nunbe of individuals preparing for hibenation ad the nunbe that actually enteed hibenation in each T.C. within the first 137 days are sham in Figure 1. A 2 x 4 chi-square test demonstrated that the distribution obseved of tlose preparing for hibenation diffeed fran that expected by chance on2 = 44.36, df = 3, P < 0.005). Six 2 x 2 chi-square tests wee then ccmputed using all possible 18 l9 cambinationns of two T.C. '3. Only tnose canbinations within a proto— peiod level wee not significantly diffeent frann one anothe (SC vs 1C, P < 0.001; SC vs SW, P = 0.42; SC vs LW, P < 0.001; 1C vs SW, P < 0.001; LC vs LW, P = 0.24; Sst LW, P < 0.001; Fisher Exact Probability chi-square Test, Steel and Torrie, 1960) . This indicates that more animals were preparing for hibernation under sl'ort daylengths than ude lonng daylengths. To corroborate this connclusion, a Kruskal- Wallis Main Effects ad Inteaction Test (Bradley, 1968) was conducted on the nmean maximum weight achieved in each T.C. within 137 days (Fignre 2). This demonstrated that only the plnotopeiod main effect was significant, indicating tlat the maximum weight attained diffeed between photoperiod levels (P < 0.001) but was not diffeent within photopeiod levels (0.05 < P < 0.10). Therefore, animals ude ID 12:12 needed the same peak weight regardless of tne tennpeature, as did tnose animals ude ID 15.33:8.67. Howeve, the mean maximnm weight attained ude ID 12:12 was greate than that ude ID 15.33: 8.67. A 2 x 4 chi-square test conducted on tte nunbe of animals hibe- natinng in the four T.C.‘s was also significant (X2 = 33.32, df = 3, P < 0.005). Again, all cambinnationns of two T.C.‘s wee carpared, denonstrating that all between-plotopeiod comparisons , wee signifi- cant (SC vs III, P < 0.001; SC vs SW, P = 0.48; SC vs LW, P < 0.001; ICvs SW, P=0.003; ICvsIW, P=0.40; SstLW, P< 0.001). At ID 12:12, there wee no diffeences between tempeatures in the animals' initial response. Ttose animals innitially preparinng for hibenation at 20°C did so in 57.0 :26.3 SD days while tlose at 5°C did so in 46.5 :15.8 SD days (0.12 < P < 0.66, Wilcoxon Rank Sun Test, Figure 2. 20 Mean maxinum body weight attained within 137 days in the four laboratory T.C.‘s and in a samnple of 56 field animals. F equals field group; S, L, C, andWare the same as in Fignre 1. Sample sizes for the laboratory groups are SC = 18, SW = 13, LC = 12, IN = 12. Vetical bars repre- sent 11 stadard deviation. 21 $59.9 £925 >oom $5.5on LW F SC SW LC Experiment 22 Snedocor ad Cochran, 1967). Tnose animals first hibernating at 20°C did so in 62.5 ill.4 SD days while tlose at 5°C did so in 55.0 118.0 SD days (0.40 < P < 0.50, t = 0.78, df = 23). Therefore, with respect to the initial preparation for and entrance into hibenation, the animals responded the same regardless of the Ta. The important factor was ptotopeiod. ID 15.33:8.67, 20°C As previously stated, animal 47 began to hibenate on day 171. Subsequently this animal went through a hibenation inteval followed by an active inteval (Figure 3). Altlnough two hibenation invevals are shown on the figure, the second was not coinncident with a weight inncrease ad the animal was neve found to be torpid ove the nocturnal peiod. The only time the animal was found torpid was during daylight. 'Ilneefore, this should probably be consideed more like daily torpor than hibenation. All othe individuals remained at low summer weights (example, Fignre 3) and wee judged to be in reproductive condition (testes scrotal or medium nipples). It also appeared that molting proceeded abnormally in flat animals wee often foud with bare patches of skin that wee onnly slowly filled in with hair. A listing of the animals, their dates of capture, and the length of time spent unde these parti- cular conditionns can be found in Table 1. ID 15.33:8.67, 5°C Onnly two of 12 animals prepared for hibenation, and only one of 12 enteed hibenation within 137 days. The two animals preparing for hibernation reacted peak weight on August 23 and Octobe l4, .nemuw sumo mo umfioo umoa Hanan. 5 85m 3 H55: .355. ufimwunmu when x33 9F wN ON .329. an O" e O mN ‘ ON d ‘ Ow @— ON mN Om .wm ov O“ u— ON mN on em Oc (0) “4013" 1009 (0) 1H013H A009 :55: mN ON w~ On . magma: gag .Ooom .Séummda a 8pm mg HEM Mom @5de ”Ema; boom ‘ ‘ ‘ Ian/<34}. 5.29.. wN ON mu Ou .uN (0) 114013" 1009 to L Ov §.DN .m— u; N non nHonan none 00 .mmwumqa 24 TABLEl Annimals which survived more than 137 days unde ID 15.33:8.67, 20°C Annimal No. Sex Date of Capture Length of Survival (Days) 41 M 5 / 2/ 76 260 5 M 6/ 28/ 77 335 10 F 6/ 28/ 77 342 16 M 6/ 28/ 77 355 7 M 6/ 28/ 77 370 62 M 6/ 24/ 76 562 47 F 5/ 23/ 76 565 56 F 6/ 21/ 76 565 60 M 6/ 24/ 76 631 54 M 6/ 20/ 76 749 63 F 7/ 2/ 76 749 64 F 7/ 2/ 76 749 25 respectively, while the individual enteing hibernation did so on Septenbe 6. This timinng indicates that an edogeous rhythm may have been expressed in these individuals because these changes occurred at a time which corresponded to the documented occurrence of these changes in nature (see Field Results section). Howeve, it is also possible that these two animals were stimulated to prepare for ad, in one case, ente into hibenation directly by the low tempeatures. All othe individuals maintained low stable weights ad remained reproductively active. The death of the majority of the animals from a toxic sub-' stance precludes furthe analysis. ID 12:12, 20°C In this T.C., 12 of 13 animals undewent weight gainns to reach a peak weight greate than or equal to 25 grams, ad nine of these ani- mals hibenated. A numbe of animals exhibited more than one hibe- nation inteval (Table 2). Six of the nine animals udewent at least two hibenation intevals, two went through at least three intevals, ad one animal died while in its fourth hibernation inteval. The mean hibenating—nonhibemating peiod length for the animals that hibe- nated at least twice was 157.2 $60.9 SD days, establishing that rhythmicities do occur ude these conditionns with a peiod of approxi- mately five months . Canplete records of weight ad hibenation acti- vity for 12 of the 13 animals in this T.C. are foud in Figures 4, 5, ad 6. For tl'ose animals that did udego two or more hibenation inte- vals, the second inteval was snorte than tne first (P < 0.005, paired t = 5.34, df = 5, a posteiori). In animals 100 and 105, the third .09 9.3 mo gm um weigh on on 952 983 mama—Ham ewes: .393 and» 5 mead Hagan undo mougmu x 26 - - - - - x 8m mm 2 a: - - - - - x NE a: m 8H - x n: on 3 8 we 8 1: a: x mm x 5 H S a a on a 2: - - - x ram 3 a: a a 8 - - - x m3 8 a we HQ 8 - - - x SN 3 a: Q a Na .. - - x 8e N E we a i - - - - - x 8e on a Q wfiumchwfl: o>wuo< wfiuflgflfl m>..flo< game—none.“ o>wuo< wfiumgflm xmm .oz H852. Doom .NHNNH a .095 9..“on um “Cougar pg 3% $93 now mahogany esp mo emu—.60 use .86 33m gunmen Ho gum—goes m5 5 ”2.0% mama mg 27 damn» some mo H9500 ”.on name: 5 53m we Hen—5: dug dame/amuse down—«Edna: ”.8888 88 883 9a .98 .28 S 88 8% .58 8m 888 twat: .68 .e mafia :ezo: . zhzoz 8 8 m: 8 m o 8 8 2 2 m o n . . n . 2 . n . . . S .8 n :_ :18 an. m 8 m .8 m .8 a". 8 u". .8 m .8 m .8 m .8 m and t 00 a t 8 8 rezoz zezoz 8 8 B 2 m c 8 8 2 S m o . . . . n S . . . . . 2 .2 .2 m m 8 m 8 m .8 w. 48 u”. I .8 m .8 m .8 m .8 m as t 3 n .Ov .Ov 28 .ndmnw :86 mo uncuoo umma Home: a.“ 83m m.“ Hone—E Hanna .mamfiaufi gum—twee: ”.8888 m8 x83 2:. 8.8 .282 S .88 main use 8m 8&8 823 .68 .n 888 :czoz xczoz 8 8 2 2 m o . 8 8 2 2 m o q q 4 d ¢ Om - 1 d 1 c On .2 .2 u w .3 m .8 m .8 a". .8 u". 48 m .8 m | 4w” m I___ AM” m 2:. n 3.. n 8 8 :ezo: :hzo: 8 8 2 2 m o 8 8 2 S m o . . . . . S . . . . . 2 2 .2 m m .8 m .8 m 48 m. r8 m. . 0 I - a m - I .s m. .8 m .8 m an I vs ( 29 .nemuw some mo HmcHoo ”flea Hon—94 5 55m 3 Hun—HE .355. 0888." 28 x83 89 .98 .812 S .88 888 use you 388 any team a mafia :FZOt mN ON my Ou O . . . 8.. zczoz mN ON mu On O . n n 8.. O— .3 ON .eN on em Ot O— m— ON MN on en Ov (0) “ENE" A009 (0) lHOIBH A009 zhzoz mN ON mu O— .o. O q ‘ n. ‘ J :53: mN ON mu O" e O m0— Ou mg .ON eN .Om en .Ov O— mu ON uN Om um Ot . egos“ 26.“quon (0) 19013?! A009 (0) 154013" A009 30 hibenation inteval vas storte than tlne second, but small nunbes did not allow a statistical test. If the animals are arranged according to tie length of tire they survived ude the expeimental conditions afte tte beginnning of the first hibenation inteval, survival is correlated with the length of the second active inteval (r = +0.9882, P < 0.05). This indicates that tlose animals which renaired active and did not hibenate for a third time survived the lonngest. The othe nine simple correlations possible with these tabulated data wee not significant. The fact that hibenation did not occur afte evey weight increase signalling preparation for hibenation signifies that hibe- nation depeds upon more than just a sufficient enegy reseve. Because hibenation did not occur evey time an animal undewent a weight inncrease (Figures 4, 5, ad 6), the hibenating—nonhibenating peiod length may not be a true measure of the existing rhythmicity. This is connsistent with a hypothesis of multifactor causation for the appearance of hibernation. The peiod length of weight cycles with amplitndes greate than or equaltoS grams was 102.8 1.71.2 SD days, which is significantly less than the peiod length for hibenation- nonnhibernation cycles (P < 0.05, df = 34, t = 2.05). The peiod lengths of weight cycles for individual animals are shown in Table 3. As can be seen, all of the peiods are less than six months except for one cycle by animal 109 which was approximately equal to six months ad one cycle by animal 84 whichwas just ove one year in length. Sinnce only one cycle out of 27 approached one year in length, it is probable that it was brought about by the shipping of one or two shorte cycles . 31 .U..H USU MO gm Um mGHflgfi 3 OH 5 083 mg GWUFHH - - - - 0.8 0.2 0.8 0.3 8.08 0.2 0.8 8 .8 .02 z 00H - - - - - .. 0.00 0.8 0.00 0.8 0.8 .02 .80 m 50H - - - - 0.8 0.2 0.8 0.: 0.8 0.2 0.8 80 .83 .02 Ha 00H 0.88 0.82 0.80 0.08 0.08 0.28 m.e~ 0.88 0.00 0.8H 0.58 00H .08 .H0 .88 .00 m 00H .. .. 8.8 0.2 0.8 0.2 0.8 0.2 0.8 0.5 0.8 08 8.: .80 .88 a 00 - - - - - - - - 0.8 0.2 0.8 9: H2 08 .. - - - 0.8 0.8 0.8 0.2 0.8 0.8 0.8 8.0 .2... .02 a .8 - - - - 0.08 0.00 8.8 0.2 0.8 0.2 8.8 a .08 .8: .0 8 - - - - 0.8 8.2 0.88 0.2 0.08 0.2 0.8 .8 .R .02 a .8 aeeeeaaaeae is... in... OoON .NHHNH a .895 mfimw usages» cap “5me um swooped. 803 own”. 3885... you 3:38» 88.055.58.05: 98.. 8393 “Eugene Mom mauwcma doeuwm mg 32 The fact that the first ad last might cycles for animal 84 wee storte (105 and 42 days) supports this theory. In males that prepared for hibenation, the testes regressed snortly before or during the weight increase even if the weight increase did not end in hibenationn. At the temination of a hibe- nation inteval, males had barely scrotal testes. In fenales , the nipples went frem easily visible (medium) to not visible (small) shortly before or durinng the weight inncrease, and the nipples became visible shortly afte temninal arousal. Annimals were neve found with bare patches of skin, which indicated that molt was proceeding in a regular fashion. ID 12:12, 5°C In this T.C. , all 18 animals prepared for hibenation. Only 17 of the animals reached a peak weight of 25 grams or more, but animal 104 hibernated afte achieving a peak might of 23.5 grams. Sinnce an indi- vidual that hibenates snould be prepared to hibernate, this animal was classified as preparing for hibenation. Sixteen animals hibernated and of tlnese, l4 survived tlnrough the hibenation inteval (Table 4). Only five of these animals went into a second hibenation inteval. Annimal 70 died snortly afte the teminal arousal frcm the first hibe- nation inteval, but tl'e other animals survived lonng enough to possibly ente into a second hibenatien inteval . The five animals that did udego a secand inteval had relatively snort active intevals (69-121 days), while tl'ose animals not enteing hibenation for a second time survived fromaminimumof 160 to amaximnmof 457 days afte their temninal arousal. 33 TABIE4 Days spent in the hibenating or active state ove the course of the expeiment for ttose animals tlat hibernated at least onnce. ID 12:12, 5°C Annimal No . Sex Hibenating Active Hibenating Active 68 F 258 80 218 x - 70 M 168 14 x - - 73 M 180 160 x - - 75 F 211 282 x - - 79 M 81 x - - - 83 F 247 257 x - - 87 F 218 413 x - - 89 F 240 88 75 271 x 91 M11 20 292 x - - 95 M 186 221 x - - 97 F 56 x - - - 104 17.11 241 102 234 x - 106 F 289 69 273 7 x 108 1 172 121 169 157 x 110 M 214 271 x - - 111 M11 25 453 x - - X deotes tlat animal dies in that phase. 1These animals were known to be juveniles at start of the T.C. 34 For ttose animals tnat hibenated a second time, the two hibe- nating intevals were almost equal in duration. The only exception was animal 89 which expeienced a much done second hibenation inteval. Three of the five repeat hibenators survived tl'e second hibenation inteval. Qne animal died slortly afte emegence frem the second inteval, but the othes remained active for a consideable length of time. In the latte two, the length of the seconnd active inteval (157 ad 271 days) was longe than the maximum initial active interval for any animal that enteed into hibernation twice . 'Iheefore , it appears likely that these two animals would not have returned to hibenation for a third time. For tl'ose animals that denonstrated two hibenationn intevals, the mean peiod of the first cycle was 332.0 824.3 SD days. The mean is significantly greate than the peiod length demnstrated ude ID 12:12, 20°C (P < 0.005, Wilcoxon Rank Sum Test) ad indicates that the rhythm obseved unde these conditions can be classified as circamal. Howeve, the majority of the animals (8 of 13) did not denonnstrate any rhythm of repeated hibenation even tlough they survived lonng enough to do so. Complete records of might ad hibenation for 16 of the 18 animals in this T.C. are to be found in Figures 7, 8, 9, ad 10. The diffeence in the lengths of tie rhythms ude LD 12:12, 20 ad 5°Ccanbeassiged to thediffeence intl'e lengthofthehibe- nation inteval. The length of the inteval at 5°C was significantly greate, 187.0 :79.7 SD days, than the inteval at 20°C, 50.0 243.6 SD days (P < 0.05, Wilcoxon Rank SunTest). Even if only the first hibe- nation inteval is used to campare the T.C. '5 (because the first inte- val at 20°C was lonnge than the second), the inteval length was still 35 1.53: wN ON 2 O“ e O ‘ I A. ‘ no :53: eN ON mu On e O p0 O— m— ON mN On an Ot O— mu ON wN OM en .Ov (0) .lHOIBH A009 (0) 114013" A009 5.29.. mN ON 2 O" e O d I ‘ Oh I. Ou mu ON eN On .un Ov O“ m— ON wN OM en Ov gamma :8 mo 8:80 on“: 8%: 5 85w 8.“ 895: H.955 .8828 gage: 8888.... 82 x83 88. .088 .882 B .88 2858 850 80 £088 0.808 .88 (0) .lHOIBH A009 (‘0) “4013" A009 .5 9.588 36 imam :08 mo 858 ”0me 8&5 5 85» m.“ 843: .355 8.25805 03880:”: ”.888.“ when 883 men. .Oom .NHNNH a Baum 8% 86m 8m magnum 8.383 boom .m 08w?“ :ezo: :e200 8 8 2 02 m 0 8 8 2 2 m 0 a . J n - 2 n . n 4 . 02 .2 .2 8 8 ow m ow m .8 .m. .8 ...". .8 m .8 m .8 m .8 m as n E. n .00 0. zezoz zezoz 8 8 2 2 m 0 8 8 2 2 8 0 . . n - . 02 . . - . . S .2 .2 8 8 .8 m .8 m .8 .m. .8 ...". .8 m .8 m .8 m .8 m 8: ... a: ... Ov Ov 37 :35: ON ON m— O“ ‘ ‘ ‘ ...—20... ON ON 3 Ou NO O" m— ON ON On an .Oc Ou w~ ON ON Om mm Ov .838 :03 no 8258 umma .895 5 53m ma 8.9.85 Hung 888888 83 883 9:. .98 .812 S 888 8358 .58 .88 £828 8888 .68 .m 858; (O) .lHOIBH A008 (0) “4013" A009 :35: .ON ON 3 Ow O O d 1 d 1 ...—20... ON ON 2 Ou ..o. O 1‘ I d 1‘ OO— O" .3 ON ON on mm .Ov Ou w— ON .ON OM OM Ov . 38535 gag (0) “4013" 1009 (0) 1W1!" £009 .namuw :03 mo .338 ”.de 88am: 5 55m 88 .845: 35:4 .mgufi gag 888888 83 883 9:. 8.8 .288 S 8.8 38?. .58 88 28.8 £8883 .68 .2 258: 8828: 8828: 8 8 2 2 8 o 8 8 2 2 8 a . - . . . 2 T . . . . 2 . 2 . 2 m M 8 m 8 m 4 8 ...". 4 8 m. . 8 m . 8 m . 8 m . 8 m 3: ... 2: c 3 9 8828: 8828: 8 8 2 2 8 o 8 8 2 2 8 o q 8 4 . 4 on . J 4 J 4 o— . 2 . 2 m m 8 m 8 m . 8 ...". . 8 m. . 8 m . 8 m l 4 mm m l I 4 mm m OO—L 008 Ov O? 39 greater at 5°C (P < 0.05, Wilcoxon Rank Sun Test). The lengths of the active interval were the same at the two tenperatures (P > 0.50, t = 0.46, df = 12). The weight cycles under LD 12:12, 5° (Table 5) and 20°C were also not temperature independmt (Figure 11), being longer at 5 than at 20°C (277.6 i804 SD days vs 102.8 171.2 SD days, P < 0.001, t = 8.08, df = 39). As in T.C. SW, hibernation did not follow every weight increase. Reproductive condition changes occurred as in T.C. SW. A sumary of T.C.‘s SC, LC, SW, and LW is in Table 6. Because reproductive cycles are under hornnnal control , it would be of interest to know more about the levels of circulating hormones in the blood stream. In the experiments cormcted in this study, all the males that underwent weight gains also underwent gonadal regression. This was true whether hibernation occurred or not. It would be of interest to lmow whether or not those animals that did hibernate had different levels of circulating hormones than those :mimals that did not hibernate. It is possible that subtle differences in hormne levels could make the difference between hibernation and activity. A large amount of variability in the lengths of hibernation and weight cycles was denonstrated in the two short daylength T.C.‘s. Although it may be tempting to extrapolate these laboratory results to a discussion of variability in natural populations (i.e. , sane animals may die in nature because they end hibernation in mid-winter) , caution should be exerted because the laboratory animls were mintained under ccnstant conditions which never occur outside of the laboratory. It is possible that sane of the variability is present because the animals are affected in an adverse way by the constant conditions. For 40 .od. «nu mo p.838 um 83.8822“. on. 8 E52 6.8.8.3 8.35.88 0895..” - - - - 8.8 8.2 0.08 28 22 2: .. - - - 8.8 8.2 0.08 88 z 0: .. - 0.2 0.2 0.8 0.2 0.8 808 .200 202 82 .. - - - 8.88 0.2 8.28 08 .8 82 - - - .. 0.8 0.2 8.8 38 20.8 02 - - - - 8.2 8.2 8.88 888 z 80 - - - - 8.8 0.2 8.8 2 22 a 0.8 0.2 0 08 0.: 8.8. 0.2 0.8 H: .82 .20 8 08 - .. - - 8.8 8.2 8.8 $8 .8 a8 - - - - 0.8 8.2 0.8 :8 .8 88 - - - - 0.8 8.2 8.8 88 8 2 - - - - 0.08 0.2 0.8 $8 8 88 A80 A80 A80 A80 A80 280 80 £868 028.... 68 .02 35.2. 8888 33 8888 33 e888 33 8888 Dom .NHHNH a Hons: 288w unmade» 95 ammo.“ um. smegma 883 285 385888 How 2339» 58.82.883.89: new 8398 ”imam; now 23833 3.8.8.8 mg 41 s'c 73 Month 19 40-1 A3 2225 8.30 'IO-i 20°C (solid line) and one animal at 1D 12:12, 5°C (broken line) . Canparison of weight cycles for one animal at 1D 12:12, Figure 11 . 42 gm QC 5 98 Q8 08 8.88 08 8.88 08 8.888 08 8.888 08 8.888 08 8.088 88.888 88 88.88 08.88 8.808 8.808 0.08 8.888 38 08 8.88 08 8.88 08 8.088 08 8.088 08 8.888 08 8.888 88.8.88 88 88.2 88.88 8.888 0.88 0.888 0.888 08 888.88 88880 880 880 8888.8 88883 88888 88.888 88.58088. 88.88888 8588888888 8888883 8888883 888888 888888 88888 888888 858.88 88808 8868 82.88 88888888 8823888 8882888882 828.8888: 58.88882 8888883 98888. 88895288 .888283888 8.88: 8.88.ngng 8.882 g E on .8 .5 .3m m..o..H mo E25 9% 43 example, an animal in nathre would be exposed to increasing daylengths upon arousal in the spring, mereas laboratory animals were exposed to constant daylengths upon terminal arousal. Also, animals in nature would be in total darkness (hiring periodic arousals whereas laboratory animals may be exposed to light during a periodic arousal. Clock—Shifted PhotOperiod qu>erimemt Of the 12 animals which started this experiment, one died on December 3 while still at a low body weight, 11 survived to prepare for hibernation, and 10 of these hibernated (Figure 12) . Two of the ani- mals underwent weight gains at simulated times of late Jme and early July while two other animals reached peak weight during simulated early to mid-August. The remaining seven animals prepared for hibernation chming a simulated time span which corresponded with the time for pre— paration for hibernation of animals in the field (late August through mid-October) . Concerning the animals that prepared for hibernation outside of the simulated natural time span, it is possible that the two animals which mderwent weight gains during August were still being stimulated by the decreasing daylengths. These individuals prepared for hiber- nation about two or three simulated weeks ahead of the first occurrence observed in nature (August 23). With the possibility of these animals being adults (they were captured after July 1 and weighed more than 15 g, hence they could not be classified as juveniles with 100 percent cer- tainty) and their having access to Ed; £13135; food, this early entrance should not be taken as being abnormal. However, difficulty does arise in interpreting the preparation for and entrance into hibernation of Figure 12. Response of animals to clock-shifted photoperiods . Relationship of weight gain (up to maximun weight attained) to similated and actual dates. Animals increasing in body weight to 1 25 g (dashed line) have prepared for hiber- nation. X indicates that the animal did not hibernate. Dashed line at simulated late August represent first occurrence of hibernation in nature while dashed line at actual mid-October represents approximately the last date an animal would enter into hibernation if an endogenous rhythm was present. Tic marks on X-axis indicate first day of month. M = male, F = fenale, J = juvenile. 45 0800 .0238 van >02 - L I In F ‘1! [I \'4‘ i I..- ‘\' ION 8“. .\//\ - . .............. .8... -.....u..-::. ......m...::.- ...: ............88 x E u n E 8 s . 8 m s E .08 . E . _ . u u .88 . n n . .8 ..uO arm 9? 2E. a.m.... 28 8222.5 (6) H5!0M ‘POB 45 the two individuals during actual late Septetber and early October. These events are what one would predict if an endogenous circannual rhythm or some other type of endogenous clock mechanism (e.g., l'eur- glass timer) were present in Zapus hudsonius. The results fran this ameriment were compared to T.C. LW which was used as a control. I-bwever, this was not an ideal control group because it was not conducted simultaneously with the shifted daylength T.C. To strengthen the conclusions from the shifted daylength T.C. , I would suggest that it be conducted again with a control group held under long daylength and warm temperatmtes . If enough animals are available, a three-month, clock-shifted daylength experiment in the opposite direction (simulated Septarber 21 daylength on June 21) may also provide some insight. Field Results: Timing of Events Animals aroused fran hibernation at different times in the spring during the three years of the field study (Table 7). The first males TABLE7 The date of first appearance above ground by males and females in the spring MES FD’IAIES Year Date Year Date 1977 April 20 1977 May 5 1978 May 3 1978 May 17 1979 April 26 1979 May 9 47 aroused approximately two weeks earlier than the first females, and the arousal of both sexes was correlated with soil temperatlm'e , with males appearing abovegroundwheithe soil texperature at 100 cmbelowthe surface reached approximately 7°C and the females appearing when the soil terperature at the same depth reached apprcmimately 9°C (Figure 13 and Table 8). In 1977, the animals aroused two weeks earlier than in 1978 and one week earlier than in 1979. When males emerge fram hibenation in the spring, the testes are barely scrotal (that is, they are not fully scrotal but are starting to descend). By the time the fetales emerge, males have scrotal testes. Same families have a perforate vagina at the time of energence; after two weeks of activity, 100 percent of the females have a perforate vagina (Figure 14). The vast majority of the males born during the sumer develop scrotal testes and most juvenile females do develop a perforate vagina, but no ferales were found to became pregnant in their juvenile year. It is not known if juvenile males can impregnate adult females. Figure 14 shows two peaks in the percent of females pregnant, corresponding to the production of two litters per year. The first litter starts to enter the trappable population in early July while a second burst of juveniles enters the trappable population in early to mid-August. There were no major differences between years in the timing of entrance into hibernation (Figure 15) . Adult males undergo weight gains in late August and are out of the trappable pOpulation by early September. Adult females and first-litter juveniles enter hibernation by mid to late Septerber, while second-litter juveniles are the last to enter into hibernation, doing so in early to mid-October. In all 63853 n m .8888 u 2 8888.888 8888 88.88.68.888. 96am 85.888588 88.8.8.3 888.888.5888 msouhd .wan 88:88 83.8 5 moan—8mm Em no.3: mo awesome ou Eu o3 um 8888.88.38.53 8808.8 mo 88888383338 .3 mubwam >5. dzmc roam: mm ON m8 o— m on mm ow w— 08 m on mm ON 2 8 8 8 8 8 1 8 8 8 8 8 8 8 8 on D I .. 8 1 mm 3 \II .I 0 c 88.88 . n G 8- 1 I. E 3 D .l l u ( IO N 8 08 a C 2.3 .. U 4 l. 0 n 0 j a 1 3 L T- U n J 88 ... E 4 I R l 0 U 0 T l R 3 mu. 4 H P 1 on m J \J T «U 4 3 L J .0 I L .8 .. .. mm I 1 1 m8: 49 TABLES Distribution of first captures in relation to soil temperatnm‘e during the spring Soil Temperature at Week Year Number of Animals 100 cm (°C) 1 1977 5 male, 0 female 7.0 l 1978 3 male, 0 female 7.0 2 1977 5 male, 0 female 8.5 2 1978 8 male, 0 female 8.0 3 1977 1 male, 2 fenale 9.0 3 1978 3 male, 1 fexale 9.0 4 1977 1 male, 3 fenale 10.0 4 1978 2 male, 1 female 10.5 5 1977 0 male, 2 female 11.5 5* 1978 -- 6 1977 0 male, 4 fenale 13.0 6* 1978 -- i arousal temperature for males 1977 = 8.04 :1.01°C N=12 i arousal teIperature for males 1978 = 8.31 -1.06°C N=l6 + No significant difference between years for males P< 0.50; t 3 0.68; df = 26 i arousal tenperature for females 1977 = 11.18 i1.65°C N=ll i arousal tenperature for females 1978 = 9.75 11.06°C N= 2 Significant difference with 1977 between males arnd females P<0.001; t = 5.61; df = 21 *Due to trap predation by raccoons , no animals were captured duringweeks 5and6of 1978. Figmre 14. 50 Percentage of reproductive males and females and percent of pregnant females during 1977 trapping season. This graph represents a percentage of the total population (adults plus juveniles) because of the uncertainty in classifying juveiile animals. 51 TESTES SCROTRL 100- 75- q 0 5 hzuuxmm 25. JUNE JULY RUG SEPT OCT HHY HONTH 52 a - w m 7 9 9 9 1 I 1 1. ‘I m 1 F 1 3 F. "F. F 2 1.2 2 "F F. .HF. 21. 3 1.5 "F "F. "F 23 23 42 —HF "F. "F. 31 61. 21 1 prp-ppb.p- P..bp.--. .p....p.p 0 5 0 5 0 5 1 1. 1. m4c>¢upz~ gum—3 uzo z~ ammo: mo 0 mm o... Fromm: z~ ozuwcuzuz: zouhczzwmur com 0225.5... @4532: ...o mum-.52 10-1 10-15 11-1 9-15 9-1 8-15 DRTE M-males, F = females. field animals . Figure 15. Distribution over time of preparation for hibernation in 53 three years during which fall trapping was conducted, all animals had entered hibernation by October 20. The fact that the annimals prepared for and entered into hiber- nation at the samne time of year in all three years, but did not arouse at the same timne every spring, irndicates that the active interval is shorter in some years than in others. Specifically, the 1978 active interval was two weeks shorter than the 1977 active interval. The late arousal of animals in 1978 influenced the timing of the reproductive events occurring during the summer. In 1978, the movement of the second litter into the trappable population was approximately two to three weeks later than in 1977 (Figure 16). This mneans that the 1978 second- litter animals had two to three weeks less timne to prepare for hiber- nation than the 1977 second-litter animals. It was possible to obtain data on the hibernating-rnonhibenating period length from four animals which were captured after undergoing weight increases in two consecutive years (Table 9). One arnimal (#20) was known to be a juvenile at its first hibenation. The other animals were not of know age. The important fact about these data is ttat the period length for all four animals is less than 365 days which indi- cates that these four arnimals underwent weight increases earlier their second year than their first. These animals did not demonstrate a rhythm under natural conditions which was exactly a year in length. Field Results: Survival If all animals, regardless of weight, first captured after July 1 are classified as juveniles (an overestimate, case 1), 16.67. of the 1977 juveniles and 10.17. of the 1978 juveniles survived to the 54 Figure 16. Minimum number of animals alive on demgraphic grid. 76 = 1976, 77 = 1977, 78 = 1978, 1 = entrance of second litter into trappable population in 1977, 2 = entrance of second litter into trappable population in 1978. 55 76 78 77 I 1 I I O O O O V m N P SAI'IV UHBWHN WfllNlNIIN MONTH 56 TABLE9 Maidmun weight to maximum weight period lengths for four animals captured after undergoing weight increases in two consecutive years Animal No . Maximum Weight to Maximum Weight Period Length # 13 September 9 (25.0 g) to 349 days maximum August 24 (28.5 g) # 14 September 23 (25.0 g) to 347 days maximum September 4 (27.0 g) # 20 September 30 (27.0 g) to 340 days minimumn Septenber 4 (25.5 g) #130 September 24 (22. 5 g) to 349 days maximum Septenber 8 (32.5 g) following spring. 0n the other hand, if only animals first captured at 15 g or less are regarded as juveniles (an underestimate, case 2), 12.5% of the 1977 juveniles and 15.67.. of the 1978 juveniles survived to the following spring. The true percentage of juvenile survival most likely falls samewhere between these extremes (12.5 to 16.670 for 1977 and 10.1 to 15.6"/" for 1978). Broken dom into first-litter and second-litter survival rates (the first litter comprises tlnse animals captured between July 1 and August 1, the second litter comprises tlnose animals first caught between August 2 and the end of trapping) , case 1 first-litter survival equals 14.27.. for 1977 and 0.07. for 1978 as compared to case 2 survival of 22.2% for 1977 and 0.07. for 1978. The snm'vivorship for 1978 first- litter animals is probably low because of trap predation by raccoons . Second-litter case 1 survival equals 18.77. for 1977 and 10.9% for 1978 as compared to case 2 survival of 0.07. for 1977 and 17.27. for 1978. 57 The low survival for 1977 second-litter animals (case 2) may be due to a small sanple size (8). In this case, the case 1 survival rate may be closer to reality. The case 1 survival rate for second-litter animals indicates that the shorter active interval for 1978 may have had the effect of decreasing juvenile survival over the winter; however, the case 2 estimates show the Opposite result. Pbre data need to be col- lected on this aspect of differential mortality before any definite con- clusion can be denonstrated. If a juvenile born in a given year survives the winter to become a breeding adult, it will have a 28.5% chance of surviving through a second hibernation interval and a 8.57.. chance of surviving through a third hibernation interval. These statistics are based on 35 animals known to be adults in the spring of 1977. Ten of these animals sur- vived to 1978 and three survived to 1979. DISCUSSION The results indicate that an exogenous factor , photoperiod, is very important in controlling the timing of preparation for and entrance into hibernation in Zapus hudsonius . In fact , clock- shifted naturally decreasing daylengths can initiate a body weight gain and hibernation response simnilar to that seen in nature. Ambient tenpera- ture, on the other hand, lnad no effect on preparation and entrance into hibernation, but soil temperature is correlated with arousal from hibernation in the spring. The following scenario is then proposed for Zapus lnudsonius. The long daylengths of spring and summer stimulate reproduction which con— tinues through mid-August. Beginning in late August, the decreasing daylengths begin to stimulate animals to prepare for and enter into hibernation and this continues through mid to late October. Hiber- nation then continues until mid-April to early May depending upon the level of the soil temperature around the hibernacula. When the soil temperature increases to a certain level, the animals emerge from their hibernacula and became active for the summer . Therefore , it appears that two factors are important in the yearly cycle of Zapus hudsonius . Photoperiod signals the end of the active interval and, possibly, soil temperature signnals the end of the hibernationn interval. I will first takeuptheissueofphotoperiodandthenccxmebacktodiscuss therole of temperature later in the paper . 58 59 Photoperiod It is possible that photoperiod has its effect upon this species in one of two ways. Either daylength acts as a Zeitgeber and provides information concerning timne of year that entrains an endogenous self- sustaining oscillator or daylength stimulates preparation for and entrance into hibernation in the absence of any endogenous rhythm. In the following sections, I will explore both of these possibilities by examining hypotheses generated by different researches (Mrosovsky, 1977, 1978; Pengelley and Fisher, 1957; Sansum and King, 1976). Presence of an Endogenous Circannual Rhythm The results indicate that no rhythmicities were present nmnder T.C. LW and mast likely that no rhythmicities would lnave been demnstrated under T.C. LC. In addition, temperature independence was not a prop- erty of the rhythms that did anpear in the two short light cycle T.C.‘s. Althougln the period of the rhythms under T.C. SC could be termed circannual, that under T.C. SW could not. These factors reject the hypothesis that an endogenous circamnual rhythm of the type described by Pengelley and Fisher (1957) for Spemphilus lateralis is present in Zapus hudsonius. Honaever, the possibility still exists that Zapus hudsonius possesses an endogenous circamnual rhythm which is expressed only under certain environmental conditions. This concept of rhythm nonexpression has a firm basis in circadian rhythm research. For ecample, electroshock of a rat will totally eliminate the locamtor rhythm for a certain time interval but when the animal regains its loccmntor ability, the rhythm reappears in phase as 60 though no perturbation had occurred (Richter, 1965). This indicates that same measuring process was present even though the observable expression of the rhytkm was eliminated. The presence of circannual rhythm under certain conditions but not others occurs in many species. For example, deer demnstrate circamnual cycles in antler growth under ID 6:18 or 18:6 but not under ID 12:12 (Goss, 1976), ad starlings show circamnual cycles in testi- cular growth under ID 12:12 but not under ID 11:13 or 13:11 (Schwab, 1971) . In addition, other experiments deIonstrating circamnual cycles in birds for body weight, mnolt, gonad size, and locanrntor activity usually have used the plnotoperiod ID 12:12 (Gwimner, 1971; lofts, 1971) . Pocket mice (Perggnattmns longimembris) show circamnnnal cycles in above-ground activity nmnder ID 12:12, 16°C but not LD 12:12, 31°C, ID 9:15, 18°C, ID 15:9, 18°C, or ID 9:15, 8°C (French, 1977). As yet it has not been shown that the results presented by the authors mnentioned above are analogous to the rhythm noneanession con— cept seen in circadian rhythm. However, rhythm nonexpression does remain a viable hypothesis and one that cannot be refuted by the data that have been presented. In fact , the preparation for ad entrance into hibernation of one individual in T.C. LW ad the early preparation ad entrance of two individuals in the decreasing daylength experimnt led same support to this hypothesis. It is possible that some indi- viduals have rnarrower limits of enviriormental conditions within which the rhythm can be expressed than others. For ecample, if a population of animals is considered to represent a normal (bell-shaped) curve with respect to their capacity for exhibiting a rhythm mde diffeent photo- periods, one would expect to see fewer and fewer animals demonstrating 61 rhythms the further the photoperiod deviated fran some optimal photo- period level. Some evidence suggests that ID 12:12 may be the optimal phntoperiod (French, 1977; Gwinmner, 1971; lofts, 1964; Schaab, 1971). At some level of daylength, all of the animals would ecceed their limits of rhythm expression and a particular experiment would record the absence of any rhythmicity. The three animals that prepared for and entered hibernation in the long daylength and decreasing daylength T.C. would have wider limits of rhythm etpression than the majority of animals, but ID 15.33:8.67 would not have exceeded their limnits of expression. Other Types of Endogenous Clocks InZapus hudsonius, therhythms inbodyweightadhibernation exhibited under T.C. SW are not circannual, but the possibility still exists that they are endogenous. In this case, the body weight ad hibernation rhythms would not be terperature independent , since they were shorter at 20°C than at 5°C. Mrosovsky (1977) feels that one mechanism for the production of a temperature depedent rhythm such as that seen in Zapus lnudsonius would be an internal oscillator that is very sensitive to body temperature (Tb). If the period of the oscillation has been lengthened by a lower- ing of Tb, thenatural period of the rhythm must be less than or equal to the shortest cycle at the highest Tb achieved during hibenation. An alternate hypothesis put forward by Mrosovsky (1977) is that temper- ature only affects the expression of the rhythm. When animals are at higher Tb's, cycles of shorter length are expressed ad when the ani— mals are at lower Tb's, longer cycles are expressed. 62 A proposed example of an endogenous, temperature depedent rhythm is discussed by Mrosovsky (1977) for the dormouse (Elk 313) . In Mrosovsky's study, dormice held under ID 12:12, 20°C had a mean body weight cycle length of 53 days while animals held under ID 12:12, 5°C lnad a mean body weight cycle length of 162 days. Mrosovsky (personal camunication) also states that dormice show cycles of similar lengths under ID 6:18, ID 18:6, and natural plnotoperiod. Although Zapus hndsonius does not respond identically to G_l§ g_]_._i£ in detail under all eqnerimnental conditions, the response of g. hudsonius and Q. g1_is_ are similar under ID 12:12, 20 and 5°C. This derpnstrates that g. lnudsonius fits the general concept of Mrosovsky (1977) but diffes in detail fram Q. g_l_i_s_. Conceptually, it may be possible that temperature dependent rhythms are present in g. hudsonius nmnder LD 15.33:8.67, 20 ad 5°C, but not expressed. If this is the case, both Mrosovsky's concept and the rhytlm nonexpression concept could be involved. If an exogenous stimulus (such as photoperiod) is capable of set- ting in motion a series of physiological events inside the animal which run for one year, the animal could prepare for ad enter into hiber- nation in successive years at the correct time provided the stimulus was present. This corresponds to an hour-glass timer (Biirnning, 1973), which is capable of providing the animal with a clock, but it is not exactly an endogenous rhythm in that no rhythm could be demnstrated witlnout the enogenous stimulus. This could be the method by which g. hudsonius measures time over the hibenation and active interval. In the particular case of g. lnudsonius, the decreasing daylengths of late 63 summer ad fall could provide the exogenous stimulus that would ini- tiate the hour-glass timer. The hibernation interval and active inter- valcouldbeconmnledtooneamtheradtheirlengthsgovemedby physiological processes. For enamle, the active inteval length could be determined by the physiological processes of sperm/ egg production ad breeding while the hibernation interval length could be determined by hormonal influences (see synthesis section) . Mrosovsky (1978) has postulated that the yearly cycle of hiber- nation in the thirteen-lined gronmd squirrel (Spermnphilus tridecemlineatus) may be controlled by an hour-glass timer which is reset by warm temperatures. Mrosovsky bases his postulate on the fact that a high percentage of the animals maintained in a cold room did not show persistent cycles but became stuck in the reproductive phase after the first hibernation interval. Mrosovsky theorized that warm tenpera— tures may be needed to reset the hour-glass timer so that another hibernation interval could follow. The fact that circannual cycles may tend to be shown by more animals in a warm environment (Mrosovsky, 1978) leds same support to the hour-glass hypothesis. Several meadow jumping mnice responded to L0 12:12, 5°C in a manner simnilar to Spemophilus tridecemlineatus under similar conditions (Figure 8, animals 75 and 95; Figure 9, animals 87 and 110). These four animals became arrested in the reproductive state ad did not enter into a second hibernation interval. However, a number of animals did not become arrested in the reproductive state and were able either toenter intoasecondhiberrnation intervalorgo tlnrougha second weight gain and loss. Therefore, although same meadow jumping mice behave like S. tridecemlineatus under ID 12:12, 5°C, the data are not 64 strong enough to draw a firm cornclusion about the possibility of an hour-glass timer being present in _Z_. hudsonius. Endogenous Rhythms Not Present It is possible that the rhythms demnstrated in _Z_. hidsonius under constant conditionns are not endogenous but are in fact forced upon the animals by the particular photopeiod chosen for the experiment (TD 12:12). This particular plnotopeiod is viewed by Sansum and King (1976) as being short enough to break the plnotorefractory state in an animal but still lonng enough to provide photostimulation ornce the ani- mal becames photosensitive. The alternate states of photosensitivity and photorefractoriness would then be expressed creating a rhythm. The major assunption behind this hypothesis is that thee is no endogenous oscillator present under corditionns when a rhythm is expressed, and underlying nonexpressed rhythms are not present under othe experimental conditionns . The rhythm produced is totally a result of the expeimental conditions peceived by the animal. In _Z_. hudsonius, the denonstration of rhythm urde ID 12:12 ad not ID 15.33:8.67 does fit this theory. One important fact about the cycles present in g. hudsonius unde ID 12:12 that is diffeent fram other species is that although the hibernation intervals are nnot the same length at the two expeimental tenpeatures, the active intevals are. This might be expected if the active inteval must rnmn for a certain time once it has been initiated. Thiswouldhappenifittakesasetamnmntoftimetoallowforgonnadal growthadspem/eggprochictiontobeginsotheanimaliscapableof 65 breeding. The hibenation inteval would still be depedent upon body tempeature, leading to cycles of diffeent lengths. The equality of the active inteval lengths under 5 ad 20°C for g. hudsonius is diffeent fren a species such as §_. latealis which denanstrates a circannual rhythm as described by Pengelley and Fishe (1957). In this species, the active interval is much longe at the highe tempeatures to canpensate for the starter hibenation inteval (Pengelley and Asmurdson, 1969) . Synthesis 1 do rat believe that it is possible, with the data that I have presented, to distinguish clearly between ttne hypottnesis of rhythm ran- expression, othe types of endogeaus clocks, or na edogenous rhythm being present. However, I have shown that individual meadow jumping mice do respond to clock- shifted decreasing daylengths by preparing for and enteing hibenation. Therefore, daylength must be consideed as a sufficient exogenous cue capable of stimulating the preparation and initiation of hibenation in _Z_. hudsonius at the apprOpriate time of the year. The action of daylength on the animal may be mediated through the pineal gland. In the pineal glad, stimuli fram the eyes conveyed via the optic nerve and superior cervical ganglion cause a decrease in the production and release of melatonin which lessens the inhibition on cetain physiological events (Wurtman ad Axelrod, 1965) . Since the production of melatonin occurs during the dark phase of the 24-h cycle (light inhibits synthesis ad release), starter days in the late summer ad fall would mean a greate amount of melatonnin production. It may 66 be that ttnere is a threstald for melatonnin accumlation in the body, and wtnen that ttnrestald is surpassed, certain ctnanges such as gonadal regression occur. Reite (1969, 1973) has contended that one of the mast important functions of the pineal glad may be to central seasonal reproductive rhythm in ptatosensitive animals. tbwever, less is krawn about the effect of the pineal upon hibenation. Spafford ad Pengelley (1971) reported that the inhibition of synthesis of seotonnirn (the precursor or melatonin) disrupted hibenation, ad Faber and Riedesel (1976) denanstrated that daily subcutaneous inj ections of melatonin resulted in an increase in the incidence ad duration of hibernation. threver, the work by ttnese two sets of authors was con- ducted on the golden-mantled gronmd squirrel , Spermaphilus latealis , a species which does not exhibit a ptatopeiodic response. Inj ectiorns of melatonin into a ptatopeiodic species such as g. tnudsonnius when indi- viduals are held under lorng daylengths would be instructive. Tempeature Although soil tenpeature is correlated with spring arousal of the field populationn, it was noted in the laboratory experiments that ani- mals did arouse wittaut any change in ambient tempeature . Thee are two possible explanations for this. First, there may be a lower thres- tald for enegy reseves below which the animals camat retain in hibenation. When an individual reaches a certain enegy reseve, arousal might take place. Secondly, the pineal glad may play a role in arousal. The gonads of hibenating mammals undergo spontaneous recrudescence afte about 25 to 30 weeks in camplete darkness (Hoffman at al_., 1965; Popovic, 1960; Reite, 1972). Since gonadal involution 67 is brought about via the pineal gland (Cusick and Cole, 1959; Hoffman and Reite, 1965; Hoffinan gt a_1_. , 1965), it is possible that the gonnads became refractory to pineal influence or that the pineal glad becomes ednausted and is no longe capable of maintaining gonadal involution afte 25 to 30 weeks of canplete darkness (Reite, 1972). Since gonadal regression seems to be a preequisite for hibenation, recrudescence of the gonads late in the hibernation interval may cause edocrine changes that can temninate hibenation. Soil tempeatnnre may just tnave a fine tnmning effect on the date of arousal, allowing slight madulatiorn (one to two weeks) according to above-ground environmental conditions . EVolution Hibernation has been obseved in species of manuals camprising many diffeent ordes and famnilies. Attempts to study the evolution of this pheumenontavebeenhampeedbecausehibernation seems tobea well conntrolled and camplicated physiological phenomenon yet it is distributed among mammals that are often consideed to be primitive (Hudson, 1973). Additional problems arise in attempting to obtain a talistic view of the factors affecting hibenation because researcters have studied the problem of timing of preparation ad entrance into hibernation unde a wide variety of experimental conditions. Some researches attacked the problem by attenpting to determine the pres- ence or absence of an edogeaus rhythm wtnile othes bypassed the rhythm conncept and ornly sought to find exogeaus factors capable of stimulating hibernation. 68 Howeve, it appears that at least one unifying concept does came through this vast liteature on hibenationn. Same animals have been found to possess an edogeaus rhythm which is vey resistant to influ- ence by exogeaus factors while other species, wtnether they have been fonmd to possess an edogeaus rhythm or not, are more easily influ- enced by exogenous stimuli. Much of the early work done on hibernators concerning factors cap- able of stimulating hibenation was confined to the rodent Family Sciuridae (the squirrel family). It has been demonstrated that edog- eaus circamnual rhythms which are resistant to influence by erogeaus factors are present in Spemqphilus latealis (Pengelley ad Fishe , 1957), Tamius striatus (Scott ad Fisher, 1972), ad Marmata mannax (Davis, 1967). It has also been denanstrated that rhythms close to circamnual or circamnual but diffeing in detail from _S_. latealis (e. g. , a lower pecentage of animals denanstrate the rhythms) are pres- ent in Spermaphilus rictnardsonnii and Spemaphilus columbianis (Scott and Fishe, 1970) , ad Spermaphilus variegatus, Spemophilus beechevi, Spemaphilus mahavensis and Spemaphilus teeticaudus (Pengelley ad Kelly, 1966) . Spemaphilus beldingi stnows a circannual cycle in repro- ductive condition at ID 12:12, 16°C, but hibernation does rat occur at that tempeature (Helle and Paulson, 1970) . Theefore, it appears that edogenous circamnual cycles are widely distributed within the hibernators in the Family Sciuridae although ttnere may be some exceptions. For exanple, it appears that tenperature may be an important emogeaus factor capable of initiating an taur- glass timring mectnanism in Spemaphilus tridecemlineatus (Mrosovsky, 1978) . Also, Spermaphilus nmndulatus may respond to ptatopeiod 69 (Dresche, 1967), but more work needs to be done with both of these species before a circannual rhythm can be ruled out. Endogenous circannual rhythm have also been reported for hibe- nators not in the squirrel family. The European hamnste, Cricetus cricetus (Family Cricetidae) , has denonstrated circannual cycles under ID 12:12, 20°C ad possibly at ID 12:12, 7°C. Seveal animals denon— strated circamnual cycles under connstant light at 7°C (Canguilhen gt _a_l_., 1973). The tedgetag, Erinacens europaeus (Orde Insectivora, Family Erinicidae) , is cited as stawing a circannual rhythm in body weight ad hibernation (Kristoffesson ad Suomalainen, 1964) , but the lighting conditions wee rat constant as light also enteed the animal roam ttnrough a window (Mrosovsky, 1978). This sheds some doubt on the presence of an endogenous rhythm. In addition, the garden dormause, Eliomnys gercinus (Dean, 1973), ad the pocket mouse, Peognathus lorngimnembris (French, 1977) , are reported to show circannual cycles in torpor. Howeve, ttne data on the garden dormouse make a firm statement about circamnual rhythmicity tenuous because diffeent populations respond diffeently to similar experimental conditions, and pocket mnice show circannual cycles only urde a limited range of photOpeiod ad tenpeature conditions, which also sheds same doubt on the obseved rhythm being endogeaus . It seems, then, that firmly denanstrated edogenous circamnual cycles in body weight and hibernation which are resistant to exogeaus influence are, with same possible exceptions (such as Cricetus cricetus) , probably limited to the Family Sciuridae (Orde Rodentia) . Outside of this family, it appears that erogeaus factors are usually more important than edogeaus factors. For example, Cranford (1978) 70 denonstrated that an edogeaus cycle is rat present in Zapus princeps (Family Zapodidae), and I have stawn in this pape that it is possible that an endogeaus cycle is not present in Z_apus hudsonius (a diffeent hypothesis can account for the appearance of cycles). Also, ptato- peiod plays a vey important role as an exogenous factor in _Z_. tnudsonius. In addition, golden hametes (Mesocricetus auratus, Family Cricetidae) are ptatopeiodic and do rat demonnstrate circamnual cycles (Reite, 1973), ad Peonmyscus leuc0pus (Family Cricetidae) , a species capable of daily torpor, has been stam to be ptatopeiodic although the possibility of an endogenous rhythm has nat been looked at (Lynch _e_t_ g. , 1978). Theefore, it appears that the majority of the species examined which are rat sciurid rodents rely on erogenous cues and have not evolved an edogeaus rhythm that is impevious to a wide variety of erogenous factors . It also appees that the peceptionn of cueing factors has evolved to meet the requirenents of particular species. For example, alttaugh Zapus hudsonius ad Zapus princeg are vey closely related, they use diffeent extenal cues to stimulate the preparation ad initiation of hibernation. In this study, g. hLdsonius responded to photopeiod wtnile in the study by Cranford (1978), _Z_. pr._rinc§ps was found rat to rely on ptatopeiod but rather on seed availability. This diffeence is probably related to the habitats these species occupy. In Michigan, _Z_. tmidsonius inhabits prairie meadows, and the animals are active above gronmd for approximately five months of ttne year. Q. Emceps, on the othe had, inhabits mountain meadows with starte summers ad is active above ground for anproximately 3.5 to four months (Cranford , 1978). In the habitat of _Z_. We, seeds do not became available in 71 ttne diet until about 50 days afte arousal (Cranford, 1977). In late years of arousal and plant growth, if the animals were cueing on day— length and always enteed hibenation at the same time, they miglnt nnot have eaugh time to gain the necessary weight to maintain themselves ove the winte. Oueing in on seed availability allows the animals to gain weight wtnen the food supply is most favorable. Meadow jumping mice, on the othe had, do not expeience such a limited active inte- val and the availability of ttne seed resource seem to be spread ove more of the active interval. The intense burst of seed availability does not seem to be present for _Z_. hndsonius , ad theefore, daylength is a more reliable characte. The question then arises, why may some animals rat exhibit edog- eaus circanmmal rhythms when they occupy the same habitat as some species that do use these rhythms? For example, g. tmdsonius relies tneavily on exogenous cues wtnile the woodchuck, Marmata monax, which lives in the sane area, has a well denonstrated edogeaus circannual rhythm ad is vey insensitive to exogenous stimuli (Davis, 1976; Mrosovsky, 1978) . I believe that an important factor in ttne evolution of an edogenous circannual rhyttm is the life span of the animals involved. In this study, I deranstrated that a vey low percentage of the meadow jumping mice survived lonng enough to ente into hibenation in ttnree consecutive years. On the othe had, it has been deIonstrated by two groups of researches (Armitage and Dowrntawer, 1974; Michener ad Michene, 1977) that in some species of sciurid rodents W flaviventris and Spemophilus richardsoniQ , a greate proportionn of the population may survive through three or more hibenation intevals. For example, Armitage and Downhowe (1974) calculated 1x values of 72 0.215 for 2-3 year old, 0.161 for 3-4 year old, and 0.121 for 4-5 year old female yellow-bellied marmots, Marmota flaviventris. In most of the hibernating species that have been stndied (exam- ples: Carl, 1971; Michener and Michene, 1977; Quimby, 1951), the juveniles born in any given year ente into tnibenation afte the adults. The following year, these same animals then ente into hibe nation on an earlie date than the previous year as I have shown for Zapus tnudsonius. Only in preparing for and enteing into hibernation in the third year would wild animals cane very close to stawing an inteval of 365 days between one hibenation interval and the next. This would then continue for subsequent years. Bdogeaus circamnual rhythm may not be advantageous for a species in which vey few indi- viduals survive to enter into ttnree hibenation intevals because very few animals survive to the point wtnere they would be undegoing pre- paration for ad entrance into hibenationn at the same time in two con- secutive years. On the other hand, some of the sciurid rodents men- tioned above would survive long eaugln to stnow a patten closer to a year in length. In this case, possessing an endogenous circannual rhyttnm may be more advantageous than relying more heavily on cetain exogenous cues. Howeve, the particular advantage (if any) is not knawn. In geneal, edogeaus rhythms are ttauglnt to be adaptive if for same reason an event must take place at a time that is not marked by an obvious environmental refeence point (Menaker, 1974). This does not follow for circamnual cycles such as hibenation because photo- peiod is a clear ad accessible environmental cue. A second reason wtny this dictatamy in edogeaus and exogenous rhythms is present may be the polyphyletic origin of hibernation. 73 (I-bweve, it stauld be noted that hibenation has been connsideed to be a monaphyletic, residual trait in primitive mammals, e.g., Cade, 1964.) If an edogeaus rhythm was present in the basal group of a lineage, one might expect to obseve this rhythm in the more recently deived species. As an example, the genea Tinmi_'_u_s_ and Eutamius are proposed as the basal group for the Family Sciuridae (Black, 1963) . Theefore, edogenous circamnual rhythm in the Family Sciuridae seem logical because they are reported to be present in the gems firtamius (Belle ad Paulson, 1970). On ttne othe had, endogenous circamnual rhythms may rat have been present in the basal groups of some othe families which contain hibernators, ad theefore, exogeaus factors have a more imnportant role. BIBLICXIRAPHY Armitage, K. B., and J. F. Downtawer. 1974. Denography of yellow- bellied marmot populations. Ecology, 55:1233-1245. Babcock, H. L. 1914. Notes on the meadow junping mouse, especially regarding hibenation. Ame. Natur. , 48:485-490. Betnrisch, H. W. 1978. Metabolic econamy at the biochemical level: The hibenate. In: Strategies in Cold: Natural Torpidity ad Themagenesis, ed.L...CHWangandJWHndson. Academic Press, pp. 461-498. Black, C. C. 1963. A review of the North American tetiary sciuridae. Bull. Mus. Carp. 2001., Harvard Univ., 130:109-248. Blair, W. F. 1940. Heme ranges and populations of the jumping mouse. Amer. Midland Nat., 23:244-250. Bradley, J. V. 1968. Distribution-Free Statistical Tests . Prentice- Hall, 388 pp. Brown, F. A. J. 1973. Biological rhyttms. In: Camparative Animal th'nyszoéogy, Vol. 1, ed. C. L. Prosse. W. B. Saundes Co., pp. 9- 5 . Bunning, E. 1973. The Physiological Plock. Circadian Rhythms and Biological Chronometry. English Univ. Press, 258 pp. Cade, T. J. 1964. The evolution of torpidity in rodents. Anmn. Acad. Sci. Fenrn. A. TV. Biol., 71:77-112. Canguilhen, B., J. P. Schiebe, andA. Koch. 1973. Rhythme circannuel pondéral de hamte d' europe (Cricetus cricetus). Influences respectives de la ptatopériode et de la temperature extene sur son déroulenent. Archives des Sciences Physiologogigues, 27: 67- 90. Carl, C. A. 1971. Population control in arctic ground squirrels. Ecology, 52:395-413. Cranford, J. A. 1977. The ecology of the western jumping mouse, 1% $5” Unpublished Ph. D. dissertation, Urniv. Utah, Salt 74 75 Cranford, J. A. 1978. Hibernation in the westen jnmping mouse (Za ms). J. Manual” 59: 496-509. Cnnsick, F. J., ad H. Cole. 1959. An improved method of breeding golden hamtes. Tex. Rep. Biol. Med., 17:201-204. Dean, S. 1973. Peiodicity of heteottnemy in the garden dormouse, Eliomys quecinus (I..). Nettnelads J. of Zoology, 23:237-265. Davis, D. E. 1967. The annual rhythm of fat deposition in woodchucks (Marmota manax). Physiol. Zool., 40:391-402. 1976. Hibernation ad circannual rhyttma of food consump- Zion' in marmats and ground squirrels. Quart. Rev. Biol., 51: 77-514. Dilge, W. C. 1948. Hiberation site of the meadow junping mouse. J. Mammal. , 16:187-200. Dresche, J. W. 1967. Environmental influences on initiation ad maintenance of hibenation in the arctic ground squirrel, Citellus undulatus. Ecology, 48:962-966. French, A. R. 1977. Circannual rhythmicity ad entrainment of surface activity in the hibenator, Peograttnus longimenbris. J. Mammal. , 58:37-43. Goss, R. J. 1969. Ptatopeiod control of antle cycles in deer. II. Alteationns in amplitude. J. Exp. Zool., 171:223-234. Gwinne, E. 1971. A camperative stndy of circannual rhyttme in warbles. In: Bioc‘rnronnametry, ed. M. t’hnnake. National Academy of Sciences, Washington, D. C., pp. 405-427. Hamilton, W. J., Jr. 1935. Habits of jnmping mice. Amner. Midlad t , 16:187-200. Helle, H. C., and T. L. Poulson. 1970. Circannian rhythm--II. Fndogeaus and exogenous factors controlling reproduction ad hibenation in chipnmnks (Eutamias) ad ground squirrels (Spermaphilus). C(Imp. Moat-Physiol., 33A. 357-383. Hock, R. J. 1955. Ptatopeiod as a stimulus for onset of hibernation. Fed. Proc.,14:73-74. Hoffiman, R. A., and R. J. Reite. 1965. Pineal glad: Influence on gonads of male hamstes. Science, 148:1609-1611. Hoffman, R. A., R. J. Haste, ad C. Townes. 1965. Effect of light and tenpeature on the endocrine system of the golden hamste (tfisocricetus auratus, Watetnouse) . Carp. Biochem. Physiol. , 1 : - . 76 Hudson, J. W. 1973. Torpidity inmamnals. In: Canparative Physiology of Themoregulation, Vol. III, ed. G. C. Whittow. Academic Press, pp. 97-165. Kristoffesson, R, ad P. Suomalainen. 1964. Studies on the physiology of the hibenating hedgetag. II. Changes in body weight of hibernating and nan-hibernating animals. ..Acad Sci. Fem. Sec. A. IV., 76: 1-11. lofts, B. 1964. Evidence of an autonncmous reproductive rhythm in an equatorial bird (Quelea quelea) . Nature, 201:523-524. Lynch, G. R., S. E. White, R. Grudel, adM. S. Beger. 1978. Effects of ptatopeiod, melatonin administration and thyroid block on spontaneous daily torpor and tenpeature regulation in 1:121; ringtcfggooted mouse, Peamyscus leicopus. J. Camp. Physiol., Manville, R. H. 1956. Hibenation of a meadow jumping mouse. J. Mammal., 37:122. Menake, M. 1974. Circannual rhythms in circadian pespective. In: Circarnnual Clocks: Annual Biological Rhythm , ed. E. T. Pengelley. Academic Press, pp. 507-520. Michene, G. R., and D. R. Michene. 1977. Population structure and dispesal in Richardson's ground squirrels. Ecology, 58:359-368. Mrosovsky, N. 1977. Hibenationn ad body weight cycles in dormice: A new type of endogeaus cycle. Science, 196:902-903. . 1978. Circanmal cycles in hibernators. In: Strategies in Cold: Natural Torpidity ad Thermagenesis, eds. L. C. H. Wang ad J. W. Hudson. Academic Press, pp. 21- 65. Palme, D. L., and M. L. Riedesel. 1976. Responses of whole-animal and isolated hearts of ground squirrels, Citellus latealis, to melatonin. Comp. Biochem. Physiol., 530: - . Pengelley, E. T. 1965. The relation of external conditions to the onset and temination of hibenation ad estivation. In: Mammalian Hibernation. III, eds. K. C. Fisher, A. R. Dawe, C. P. Lyman, E. Sctnobaum, and F. E. South. Olive and Boyd, pp. 1-29. Pengelley, E. T. and S. J. Aerurdson. 1969. Free-rnmmning peiods of edogeaus circannian rhythms in the golden-mantled ground squir- re1,Cite11us latealis. Camp. Biochem. Physiol., 30A: 177- 183. . 1970. The effect of light on the free-running circannual 1?th of the golden-mantled ground squirrel, Citellus latealis. Comp. Biochem. Physiol., 32A: 155- 160. 77 Pengelley, E. T., and K. C. Fishe. 1957. Onset ad cessation of hibenation uder constant tenpeature ad light in tte golden- ningntzzled ground squirrel, Citellus latealis. Nature, 180:1371- 7 . . 1961. Rtnythmical arousal fram hibenationn in the golden- mantled groud squirrel, Citellus latealis tescorum. Can. J. Zool. , 39:105-120. . 1963. The effect of terpeature and ptatopeiod on ttne yearly hibenating betavior of captive golden-mantled groud squirrels, Citellus latealis tescorum. Can. J. Zool., 41: 1103- 1120. Pengelley, E. T., and K. H. Kelly. 1966. A "circamnian" rhythm in hibenating species of the genus Citellus with observations on their physiological evolutiorn. Camp. Bioctnem'. Ptnysiol. , 19A: 603-607. Pengelley, E. T., S. J. Asmmdson, B. Barnes, ad R. C. Aloia. 1975. Relationship of light intensity and ptatopeiod to circannual rhythmicity in the hibernatirng groud squirrel , Citellus latealis. Comp. Biochem. Physiol., 53A:273-277. Popovic, V. 1960. Endocrines inhibenation. In: Mammalian Hibe- nation. Bull. Mus. Comp. Zool., Harvard, 124:105-130. Quimby, D. 1951. The life history ad ecology of the meadow junping mouse, Zapus tmdsonius. Ecol., Monagr., 21:61-95. Reite, R. J. 1969. Pineal function in long term blinded male and female hamstes. Gen. Comp. Edocrira1., 12:460-468. . 1972. Evidence for refractoriness of the pituitary-gonad axis to the pineal gland in golden hamstes ad its possible implicationns in annual reproductive rhyttms. Anat. Rec., 173: 365-372. . 1973. Camparative physiology: The pineal gland. Arm. Rev. Ffiysiol. , 35:305-328. Richte, C. P. 1965. Biological Clocks in Medicine ad Psychiatry. Ctarles C. Ttamas Co., 109 pp. Sansum, E. L., ad J. R. King. 1976. Iong-tem effects of constant ptatopeiods of testicular cycles of mite-crowned sparrows (Zonatrichia leucophrLs gambelii) . Physiol. Zool. , 49:407-416. Schwab, R. G. 1971. Circannian testicular peiodicity in the european starling in the absence of ptatopeiodic ctange. In: Biochrona- metry, ed. M. Make, National Academy of Sciences, Washington, D. C., pp. 428-447. 78 Scott, G. W., and K. C. Fisher. 1972. Hibernation of eastern chip- munks (Tamias striatus) maintained under controlled conditions . Can J Zoo ., 50:95-105. Slneldon, C. 1934. Studies on the life histories of Za and Napeozapus in Nova Scotia. J. Mammal., 15:290-300. Snedecor, G. W., andW. G. Cochran. 1967. Statistical Methods. Iowa State Univ. Press, 593 pp. Spafford, D. C., ad E. T. Pengelley. 1971. The influence of neuro- humr serotonin on hibernation in the golden-mantled gronmd squirrel, Citellus lateralis. Carp. Biochem. Physiol. , 38A:239- 249. Steel, R. G. D., ad J. H. Torrie. 1960. Principles and Procedures of Statistics. McGraw-Hill, 481 pp. 'I‘mmte, J. W., and J. A. Twente. 1965. Effects of pore temperature upon duration of hibmnation of Citellus lateralis. J . Appl . Physiol. , 20:411-416. Whitaker, J. 0., Jr. 1963. A study of the meadow jumping mouse, Za hudsonius (Zimmerman), in Central New York. Ecol. angr. , : S-ZSZ. Whitaker, J. 0., Jr., and R. E. Mumford. 1971. Jquing mice (Zapodidae) in Indiana. Proc. IndianaAcad. Sci., 80:201-209. Wurtman, R. J., andJ. Azelrod. 1965. The pineal glad. Sci. Amer., 213:50-60.