71-2038 BROCKE, Rainer Hans, 1933THE WINTER ECOLOGY AND BIOENERGETICS OF THE OPOSSUM, DIDELPHIS MARSUPIALIS, AS DISTRIBUTION­ AL FACTORS IN MICHIGAN. Michigan State University, Ph.D., 1970 Ecology U n iversity M icrofilm s, A XER Q \C om pany, A nn A rbo r, M ichigan (§) Copyright by RAINER HANS BROCKE 1971 iii THE WINTER ECOLOGY AND BIOENERGETICS OF THE OPOSSUM, DIDELPHIS MARSUPIALIS, AS DISTRIBUTIONAL FACTORS IN MICHIGAN By Rainer Hans Brocke A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Fisheries and Wildlife 1970 ABSTRACT THE WINTER ECOLOGY AND BIOENERGETICS OF THE OPOSSUM, DIDELPHIS MARSUPIALIS, AS DISTRIBUTIONAL FACTORS IN MICHIGAN By Rainer Hans Brocke The opossum’s recent northward population expansion in North America suggests that it possesses effective adapta­ tions to temperate winter cold. Interrelationships of the opossum's anatomy, physiology, behavior and environment were studied in southern Michigan during winter. Field observations, electrically monitored activity of penned opossums and data from 154 road-killed opossums collected throughout southern Michigan indicate that opossums leave their dens to forage in winter on those days when the air temperature exceeds the freezing point. Almost all foraging activity ceases at air temperatures below -5°C. Snow appears to affect opossums primarily by covering the food supply and hindering travel. in winter dug by other mammals. Opossums use ground dens Between December 1, 1967, and March 29, 1968, the lowest daily mean temperature for six artificial ground den cavities in outdoor pens holding opossums was 0.6°C (of 1440 measurements, 1 m below ground surface), while the lowest recorded temperature for air was Rainer -Hans Brocke -22.8°C. The mean winter air temperature fluctuation for the 24 hour period was 9.9°C. For ground den cavities, it was 0.3°C. Thermal characteristics of eight natural dens paralleled those of artificial dens. With the approach of spring and warm weather, penned opossums gained a thermal advantage as they occupied artificial tree dens more fre­ quently than ground dens. Analysis of stomach contents of 20 road-killed specimens showed animal remains in 85% of all stomachs. The resting body temperature (TB ) of five acclimatized opossums declined with declining air tempera­ ture (T^), according to + 0.048TA . the regression: Tg(C°) = 33.65 Oxygen consumption of four opossums (mean weight 3.48kg) is given by the equation: 0.51 - 0.014 Ta (C°).; RMR (ml 0 2 /g/hr) = The B M R .calculated from the equation is 0.15 ml 02/g/hr or .72 kcal/kg/hr, among the lowest of rates for mammals. Winter energy requirements of opossums confined in outdoor pens with artificial tree and ground dens was determined by food balance calorimetry. The mean metabolizable energy (ME) derived from ad libitum daily ingestion of whole rabbit is 131 kcal/kg body wt./day (four opossums, 43 trials). The mean ME expenditure of seven opossums in artificial ground dens (for minimum activity and sleep at TA g just above freezing) is 1.96 kcal/kg/hr. The mean ME expenditure of eight opossums (mean weight 2.25 kg) at higher activity levels is 2.92 kcal/kg/hr. The mean ME equivalent of body weight loss for seven opossums is Rainer .Hans Brocke 4.3 8 kcal/g weight lost. Approximately one-third of the opossum1s winter energy expenditure is provided by catab# olism of body substance, and two-thirds by food. The body fat fraction of road-killed opossums was calculated from specific gravity determinations of eviscerated carcasses. Decline in body fat of 54 opossums related to days of the winter period (December 1 = Day 1; March 29. = Day 120), is described by the regression: Y = 27.13 - 0.157 X; where Y represents the lipid fraction of body weight in percent, and X represents days of the winter period. Likewise, winter decline in body weight of 138 road-killed opossums is described by the regression: Y = 3012.4 - 7.67 X; where Y represents body weight (g) and X represents days of the winter period. One gram of body weight supplies 5.0 3 kcal of ME according to fat and weight decline data. Maximum weight losses of penned animals over the winter period ranged from 40 to 45%. Three mathematical models of winter survival were calculated for a theoretical winter period of 120 days. The number of enforced winter sleep days (days with sub-freezing temperatures) determines how many poten­ tial foraging days remain in the 120 day winter period. The mandatory energy intake (energy deficit) per foraging day increases northward as the potential number of foraging days decreases. Assuming complete foraging success (315 kcal per foraging day) opossums can theoretically survive 90 days of enforced winter sleep if 30 foraging days with complete Rainer-Hans Brocke foraging success are available. Presently, the northern distributional limit of the opossum in Michigan coincides with a winter severity level of 70 enforced days of sleep, or the approximate southern limit of the pine-hemlock ecotone region. are discussed. Limitations of deep snow on distribution The hypothesis is presented that winter energy limitations will generally restrict opossum popula­ tions in northeastern North America to points south of this ecotonal boundary. Perhaps the opossum's physiological- behavioral response to winter is unique among temperate mammals because potential generalized ecological niches to accommodate its low-key mode of life are restricted. "There are wood rate here ae big ae a French eat, whiah have white fur— . The female hae two ekine under her belly which givee the effect of a aloak oloeed at the top and bottom and open in the middle• They have ae many ae eight young, which they carry inside when they walk. Same eavagee brought me a eouple onae during the winter* X hoped to eond them to France, but I wae surprised eome days afterward to find their tails mieeing. The aold had froaen them ana they had broken off like glass. Sometime later their ears also dropped off, eo that I wae obliged to kill them. Some eavagee to whom X told this informed me that the mothers always kept them in their holes imtil they were ae big ae themeelvee, and that they never want out when it wae very cold." Memoir of Pierre Llette (a French army officer) concerning the Illinois country* 1702* from "The Western Country In the l?th Century"* "by M. Qnalfe (ed). quote courtesy of the Citadel Press. • n • I To my families the Brockes and Biebesheimers and to Dr. E. Paul Reineke Professor of Physiology Michigan State University ACKNOWLEDGMENTS For their guidance and generous assistance in this research project, I express my sincere thanks and deep appreciation to my major professor, Leslie W. Gysel, and to members of my doctorate committee, Professors E. Paul Reineke, Duane E. Ullrey, Manfred D. Engelmann and Rollin H. Baker. I am grateful to personnel of the Michigan Department of Natural Resources who collected specimens for this study, particularly to Ralph I. Blouch and C. T. Black. This research was supported by the Michigan Agricultural Experiment Station, in part through a Graduate Research Assistantship for the winter of 19 68. I appreciate the financial support accorded by the faculty position I held for five years in the Department of Natural Science. Finally, I am deeply grateful to my wife, Mary, for her patience and encouragement. v TABLE OF CONTENTS LIST OF TABLES............................................... viii LIST OF F I G U R E S ................................................ * CHAPTER I INTRODUCTION................. ........... . . . . 1 CHAPTER II METHODS AND MATERIALS .......................... 5 Study Area....................................... 5 Field Observations........................... . . 7 Experimental Enclosures and Artificial Dens............................... 7 Calorimetry by Food B a l a n c e ................... 12 Oxygen Consumption and Temperature R e g u l a t i o n .................................. 15 Collection of Road-killed S p e c i m e n s ........... 21 Measurement of Body F a t ........................22 CHAPTER III RESULTS AND DISCUSSION: A C T I V I T Y ............. 24 CHAPTER IV DEN MICROCLIMATE AND ECOLOGY.................... 57 CHAPTER V FOOD AND FORAGING BEHAVIOR...................... 75 CHAPTER VI TEMPERATURE REGULATION.......................... 83 CHAPTER VII ENERGY REQUIREMENTS BY OXYGEN CONSUMPTION............................. . . .90 CHAPTER VIII ENERGY REQUIREMENTS BY FOOD BALANCE CALORIMETRY. . Metabolizable Energy of Experimental F o o d ....................................... Maximum Energy Intake ....................... Energy Requirements for Sleeping and Activity in W i n t e r .................... Energy Equivalent of Body Weight Loss . . . . Energy Requirements Concurrent with a Body Weight Loss of 40%.................. vi 10 3 103 107 109 120 120 CHAPTER IX DECLINE IN BODY WEIGHT AND FAT IN WILD O P O S S U M S . ............................. 124 CHAPTER X ECOLOGICAL SYNTHESIS: MODELS OF WINTER ENERGETICS ANDSURVIVAL ............ 143 Assumptions for Model 1 ...................... 147 Sample Calculation - Species Winter Energy Expenditure at Battle Creek, M i c h i g a n ................................... 15° Results and Inferences........................ 151 CHAPTER XI DISTRIBUTION.................................. 16 0 CHAPTER XII C O N C LU SI ON S....................................179 BIBLIOGRAPHY APPENDICES .............................................. APPENDIX A .......................... 181 192 APPENDIX B ............. ...................... 200 LIST OF TABLES Table 1. Winter temperatures of natural ground dens at East Lansing, Michigan (Winter of 1967-1968).......... ...61 Table 2. Stomach analysis of 20 opossums collected between Jan. 2 3 and Mar. 20, 19 68..... 77 Table 3. Resting oxygen consumption and metabolism calculated from the equation RMR (ml 0 2 /gm/hr) = 0.51 - 0.014 T^. Table 4. Basal metabolic rates of some winter-acclimatized mammals 92 Table 5. Energy values (kcal) determined by bomb calorimeter for feces and urine of opossums on a maintenance diet in autumn. ............................. .104 Table 6. Food intake and excretory output of opossums on a maintenance diet in autumn..........................106 Table 7. Energy intake, excretory energy loss and metaboliz­ able energy of opossums on a maintenance diet in autumn. . 10 6 Table 8. Ad libitum daily ingestion of whole rabbit by opossums in winter.............. 108 9. Metabolizable energy (M.E.) requirements of opossums for sleeping and minimum activity in winter........ 117 Table 10. Metabolizable energy (M.E.) requirements of opossums for periods at differing activity levels........... 118 Table 11. The metabolizable energy (M.E.) equivalent of weight loss while opossums were in the den in winter.^ Table 12. Metabolizable energy (M.E.) expended and percentages of M.E. derived from food and loss of body substance, while opossums were losing weight at the approximate daily rate of .3 3% of the initial winter weight, equivalent to a total weight loss of 40% of the initial winter weight over the 120 day winter period. Table 122 # • ♦ VXXX Table 13. Body weight statistics of opossums collected January through March, 1968, in southern Michigan... 134 Table 14. Regression of body weight (g) on days of the winter period (Dec. 1 = Day 1; Mar. 29 = Day 120) of opossums collected January through March, 1968, in southern Michigan.................................... 135 Table 15. The lipid fraction of body weight of opossums collected January through March, 1968, in southern Michigan............................................. 139 Table 16. Regression of the lipid fraction (%) of body weight on days of the winter period (Dec. 1 = Day 1; Mar. 29 = Day 120) of opossums collected from January through March, 1968, in southern Michigan.•• 140 Table 17. Total energy expenditure for the winter period (120 days) and energy deficit per foraging day for various combinations of foraging days and days spent in sleep, for Models 1, 2 and 3 ............... 152 Table 18. Effect of lack of foraging success on the energy deficit (kcal) per foraging day for various combinations of foraging days, and days spent in winter sleep (120 day winter period)................ 153 Table 19. Location description of ground dens................. 192 Table 20. Opossum activity as a function of temperature in activity units per/animal per/hour.................. 193 Table 21. Opossum activity as a function of time of night in activity units per animal per/hour............... 19 4. Table 22. An example of data used to calculate the maintenance energy requirement for an opossum (No. 1)........... 195 ix LIST OF FIGURES Figure 1. Diagram of experimental pens, showing overhead, side and cross-sectional views.............. Figure 2. Diagram of respirometer used in oxygen consumption determinations........... *................. 19 Figure 3. The daily collection of road-killed opossums in southern Michigan compared to the maximum tempera­ ture at Lansing, Michigan, for the period January through March, 1968............................... 26 Figure 4. The mean daily collection of road-killed opossums collected in southern Michigan fromJanuary through March, 1968, plotted according to 5°Cmaximum temperature classes at Lansing, Michigan......... 29 Figure 5. Activity intensity (activity units/animal/hour) of opossums confined in experimental pens plotted according to 5°C temperature classes, February 18 to May 3, 1968..... ..31 Figure 6. Diagrams of probing circuits and a short foraging trip............ 35 Figure 7a. A probing circuit made in an aspen stand on the night of March 12-13, 1968........................ 37 Figure 7b. A probing circuit made by the same opossum on the following night of March 13-14............... 3 7 Figure 8a. The mean daily activity distribution related to mean air temperature for 8 opossums for the 15 day period February 18 to March 4 (120 opossum nights)............................................ 41 Figure 8b. The mean daily activity distribution related to mean air temperature for 7 opossums for the 15 day period March 4 to March 19 (10 5 opossum nights). ... 43 The mean daily activity distribution related to mean air temperature for 6.9 opossums (daily mean) for the 15 day period March 19 to April 3 (103 opossum nights)....................................45 Figure 8c. x Figure 8d. The mean daily activity distribution related to mean air temperature for 7.8 opossums (daily mean) for the 15 day period April 3 to April 18 (117 opossum nights)............................... 47 Figure 8e. The mean daily activity distribution related to mean air temperature for 7.6 opossums (daily mean) for the 15 day period April 18 to May 3 (114 opossum nights)............................... 49 Figure 9a. and 9b. An opossum weighing 1.83 kg, walking in fresh snow about 2 8 cm de ep ...................... ....... 54 Figure 10. Three-day temperature means for air and ground dens at East Lansing, Michigan for the winter of 1967-1968.......................................... 59 Figure 11. Five-day soil temperature means for various depths at East Lansing, Michigan for the winter of 19671968................................................63 Figure 12. The winter occupancy of tree dens and ground dens during daylight hours by eight experimentally penned opossums.................................... 67 Figure 13. The winter occupancy of seven natural ground dens in Area I by opossums and other mammal species... .72 Figure 14. Opossum body temperatures (C°) are plotted against ambient (environmental) temperatures for the period January through March, 1966................ 86 Figure 15. Resting oxygen consumption versus ambient tempera­ tures for fully acclimatized animals, determined from January through July, 1966................... 94 Figure 16. Sequence of resting oxygen consumption measurements for two opossums, showing metabolic fluctuations associated with shivering at low T^'s............. 99 Figure 17. The decline in body weight of a typical experi­ mental opossum (No. 5,£) during the period from January 1 to March 29, 1968...................... Ill Figure 18. The decline in body weight of an experimental opossum (No. 2,$ ) during the period from January 1 to March 29 , 19 6 8................................. 113 Figure 19. The decline in body weight of an experimental opossum which failed to survive (No. 7, 18 Figure 2 . Diagram of respirometer used in oxygen consumption determinations. Between determinations, the syringe was disconnected from the. oxygen input port, providing an air outlet. During determinations the air supply was clamped off. OXYGEN INPUT L=L nu ^ *i RADIATOR < THERMOMETER in u m w w il iii i i i »ii r 15 [^-W IN D O W rr,7 ■ SUPPORTING SCREEN / MANOMETER CO^ ABSORBER HJ) ABSORBER SEAL /Si of each measurement period were not identical. Air was circulated through the chamber between oxygen consumption measurements. Mean barometric pressure recorded at the site was used to correct for STP conditions. Carbon dioxide and water vapor were absorbed by sodium hydrate asbestos (Ascarite) and indicating anhydrous Ca£j^ (Drierite, residual H20 in gas .005 mg/1) respectively Generally, the total absorption surface of 400 cm2 for each absorbant was found to be quite adequate. In other, similar investigations, small circulating fans have been used in closed circuit respirometers to maintain air in constant contact with absorbing chemicals. However, a fan was deemed undesirable here from the standpoint of duplicating the thermal environment in a ground den. Initial oxygen consumption data obtained without a fan followed expected patterns and it soon appeared that inhalations and exhala­ tions of experimental opossums provided sufficient air circulation in the relatively small chamber space. Oxygen consumption values obtained here agree with energy determinations by food balance calorimetry. The abrevia- tions "BMR" and "RMR" used here stand respectively for "Basal Metabolic Rate" as it is usually defined, and "Resting Metabolic Rate". The latter category applies to sleeping and quietly resting animals. Collection of Road-killed Specimens The opossum may well be the most frequently killed mammal on southern Michigan highways. Winter road-killed specimens collected before the end of March are a good source of study material. Road-killed opossums collected from January through March, 1968, provided the following information: 1. Winter weight, sex and age characteristics, and decline in body weight through the winter season. 2. Characteristics of the fat depot and decline in body fat. 3. Air temperatures at which animals were active. H. An estimate of population distribution and range. 5. A qualitative analysis of stomach contents for a few specimens. The Michigan Department of Natural Resources issued a statewide field order to its personnel to collect roadkilled opossums. Each collected animal was sealed into a heavy duty polyethylene (turkey size) bag and pertinent information recorded on the attached tag. Specimens awaiting pick-up were placed in freezers or outdoors in barrels away from the sun. Generally, opossums processed as late as one month after collection were still in good condition for the purposes of this study. All animals were sexed, and aged as juveniles or adults. Females were aged according to pouch discoloration 22 (Petrides, 1949). by size alone. was used. Many males and females could be aged In a few cases, dentition (McCrady, 1939) At the beginning of winter, one useful criterion is the condition of ears and tail. These appendages are blunted by freezing in animals that have lived through a previous Michigan winter. Measurement of Body Fat Theoretically, an animal body can be divided into fat and fat-free components. On this basis, if the densities of fat and fat-free components are known, the unknown amount of fat in an animal body can be calculated if the specific gravity of the carcass is determined by water immersion. 4 By this technique, body fat has been estimated for men (Behnke et al 1942, Fidanza et al. 1953), guinea pigs (Rathbun and Pace 1945, Morales et al. 1945), albino rats (Da Costa and Clayton 1950), cattle (Kraybill et al. 1951) and swine (Brown et al. 1951, Kraybill et al, 1953). More recently, the specific gravities of guinea pigs, swine and humans have also been calculated by helium displacement (Liuzzo et al. 1958, Gnaedinger et al. 1963, Hix et al. 1967). Body fat was determined for 54 road-killed opossums. Hair was removed with animal clippers. The viscera were dissected out, leaving genital organs, kidneys and fat posterior to the kidneys along the back. By weighing each carcass to .1 gm in air and in water, specific gravity was calculated from the relationship: 23 Specific Gravity = Weight in Air Weight in-Air -HiJeight in Water The. eviscerated carcass was suspended from a triple beam balance pan by means of a small wire hook through the animal's nose. Detergent was added to the water to facilitate removal of air bubbles adhering to the carcass. CHAPTER III RESULTS AND DISCUSSION: ACTIVITY The opossum tends to avoid freezing temperatures in winter and forages almost exclusively during periods of thaw. These generalizations are confirmed by my field observations, data from road-killed specimens and recorded activity of penned animals. In Figure 3, the daily collection of road-kills in southern Michigan from January through March is compared with the daily maximum temperature at Lansing for the same period. The daily ntaximum temperature, which usually occurs in the after­ noon, has been chosen here for correlation with road-kill frequency because opossums are not entirely nocturnal in winter, but may venture forth in the late afternoon when temperatures are permitting. This is particularly true when the plane of nutrition is low, as will be shown in later chapters. Temperatures at Lansing were selected as being fairly representative for southern Michigan. It is evident that collection success was greatest when the maximum temperature exceeded the freezing level. Conversely, on days when the maximum temperature declined below freezing, collection success was minimal (Figure 3). These road-kill data, graphed according to five degree maximum temperature * 2H Figure 3. The daily collection of road-killed opossums in southern Michigan compared to the maximum temperature at Lansing, Michigan, for the period January through March, 1968. 15 10 IS 20 JANUARY 25 30 4 0 14 10 24 25 FEBRUARY 5 10 15 20 25 20 MARCH 27 classes (Figure 4) shows a striking correlation between high collection success and above-freezing temperatures. Grouping all specimens collected (154 animals) according to two major temperature classes, i.e. above and below 0°C, the mean number of animals collected per day at temperatures below 0°C was .56 (22 animals, 39 days); the mean was 2.64 per day for temperatures exceeding 0°C (132 animals, 50 days). These means differ at the 1% level of significance (t = 5.15, d.f. = 87; (Snedecor 1956). It appears that the movement of opossums drops off markedly below 0°C, judging from these data and assuming that automobile traffic remains constant. The activity of opossums shows similar relationships with temperature (Figure 5; Appendix A, Table 20). Activity was low or non-existant at sub-zero temperatures. It should be noted that each night as the temperature dropped, the activity of animals would usually span two % temperature classes. However on some nights, total activity was included in one temperature class, or as many as four classes when the temperature declined as much as 15°C during the night. For the whole period of 75 nights (5 59 opossum nights or 7.45 opossums per night), the mean activity span was 8.10 hours/animal/night, and the range of mean nightly activity spans was 2.2 to 13.7 hours/animal/night. The mean activity intensity was 5.07 activity units/animal/hour and the range of mean nightly activity intensity was 1.79 to 11.62 activity units/animal/hour. 28 Figure The mean daily collection of road-killed opossums collected in southern Michigan from January through March, 19 68, plotted according to 5°C maximum temperature classes at Lansing, Michigan. Lower limits of temperature classes are shown in the figure, the classes extending for 4.9°C, e.g. 0° to 4.9°, 5° to 5.9°C, etc. 01 o 01 o N« 01 I* 6 «•€ O MEAN DAILY COLLECTION — OPOSSUMS • “ T * DAILY n ^ MAXIMUM O' TEMP. N3 to 01* CLASSES Figure 5. Activity intensity (activity units/animal/hour) of opossums confined in experimental pens plotted according to 5°C temperature classes February 18 to May 3, 1968. Data are for 559 opossum nights, or 7.45 opossums per night for 75 nights. Temperatures were measured at the pen site. Lower limits of temperature classes are shown in the figure, the classes extending for *+.9°C e.g. 0° to 4.9°, 5° to 5.9°C, etc. Probing activity is explained in the text. ACTIVITY U N IT S A N I M A L ^ / l l O U N I s1 K oa 8J + + + + f 32 At sub-freezing temperatures, the opossum displays an activity pattern which I shall term "probing activity." Apparently, opossums monitor temperatures and other conditions of the macroenvironment by leaving the nest chamber and coming up to the ground den entrance. Probing activity may be extended into a small travel circuit in the vicinity of the den. Typically, a probing animal will stand at the den entrance, sniffing the air. of time at the den entrance varies. The length For example, on the night of February 20-21, 1968, the air temperature was -8°C at 6:00 p.m. and declined to -20.5°C by 1:00 a.m. Six of the eight experimental animals came to the entrance from one to 14 times between 6:00 p.m. and 1:00 a.m. The. other two animals did not come to the den entrance and may have slept without interruption. The length of time each animal stood in the entranceway varied from 2 minutes to 15 minutes (21 observations) and the mean probing time was 5.8 minutes per animal. On the following night, the temperature was -10.5°C at 5:00 p.m. and -15°C at midnight. Probing activity was of shorter duration on this night, varying from 1 to 8 minutes in length, and averaging 3.2 minutes (21 observations six animals). Most probing activity on the night of February 21-22 began at 7:00 p.m. and ended around 10:00 p.m. In terms of activity units (units/animal/hour), I have observed that most activity up to about 2.5 units/animal/hour represents probing activity. In the field, short trails made in the 33 snow at air temperatures below 0°C are often a manifesta­ tion of probing activity. Such probing trails are quite distinctive, being limited track circuits that leave the den entrance but soon return (Figure 6). Sometimes probing ^circuits of several successive nights may be recognized at natural dens (Figure 6 and 7). In this connection, Wiseman and Hendrickson (19 50) studying Iowa opossums, state, "...on two occasions, when night temperatures were 20°F (-6.7°C), tracks from ground dens made a circle of a few feet and returned." They also state that opossum activity is negligible below 20°F. There is some individual variation in the inclination of opossums to expose themselves to freezing temperatures. One large penned male remained conspicuously above ground at temperatures as low as -5°C, sleeping in the tree den in preference to the ground den. Apparently the ground den was quite suitable to other opossums rotated through this pen. It is perhaps significant that this experi­ mental animal was the only one which ultimately died of starvation. Contrary to much popular opinion and frequent state­ ments in the literature, sub-freezing temperatures per se are apparently not detrimental to the health of opossums, when they are well-fed. In this study, caged animals held in an unheated garage for oxygen consumption and body temperature measurements remained in good health exposed to winter temperatures as low as -15°C. However, these Figure 6 Diagrams of probing circuits and a short foraging trip. The foraging trip appeared to be entirely unsuccessful. 35 DCN ENTRANCE TWO PROBING C IRCUITS MADE ON THE NIOHT OF FEBRUARY 3 - 4 . TO TAL DISTANCE 14* t M AIR TEMP* - 4 * C OROUND DEN TEMP. // i*c » PROBINO C IR C U IT FEBRUARY 3 * 4 ' DISTANCE I M r — 0 10 M # FORAOINO TRIP FEBRUARY 2 * 3 DISTANCE 122 M A IR TEMP. - 2 * C « AIR TCMR —4# C :> O C N TEMR 4* 4*C *# / • i i A % i i i \ v « 4 / i i i % \ / SMl OPOSSUM W ALKED ON LOjBS / 36 Figure 7a. A probing circuit made in an aspen stand on the night of March 12-13, 1968. The ground was covered by 10 cm of new snow and the minimum temperature during the night was -12°C. The total distance traveled by the opossum from its den located in the left background was 1.2 m. Figure 7b. A probing circuit made by the same opossum on the following night of March 13-14. This circuit, at the left of the photograph, leads from the den located in the left back­ ground. The probing circuit at the right is the one from the previous night, pictured in Figure 7a. The snow-free area in the fore­ ground was caused by melting of snow disturbed by the photographer. About 7 cm of snow remained on the ground from the previous day, and the minimum temperature during the night of March 13-14 was -14°C. 38 animals showed signs of discomfort at temperatures below -10°C as they shifted positions quite frequently- Their ears were frozen back to the head and their tails were shortened considerably by frost-bite. At winter’s end these animals appeared in excellent condition, apart from their trunkated appendages. The opossum appears to be more temperature sensitive than other predators of similar size associated with it. The striped skunk, raccoon and red fox occurred with the opossum on both study areas, and on occasion, occupied the same dens with i t . Hence the relative temperature sensitivity of these species was quite apparent. At least some nocturnal activity was noted for the red fox for all winter nights for which data were recorded, regardless of temperature. Abies (19 69) found that lower temperatures were associated with increased red fox activity in Wisconsin. Skunks and raccoons, were not active on the coldest nights, particularly with deep snow on the ground. However, these species were frequently active at lower temperatures than the opossum. For example, on Area I, on the night of February 3-4, 1968 the air temperature at 9;00 p.m. was about -4°C and snow covered the ground to an approximate depth of 4 cm. opossums made probing circuits. At two dens, However, skunks and raccoons made active forays that night. On the night of February 9-10, the air temperature was about -7°C at 9:00 p.m. Skunks and raccoons were active, but opossums did not leave 39 their dens. On the night of February 19-20, the air temperature was about -8°C at 9:00 p.m. Again, opossums did not leave their dens while skunks and raccoons were active. Sharp and Sharp (1956) report that the lower temperature limit for winter activity of Nebraska raccoons, discouraging 90% of the population, is 24°F (-4.5°C). In this study I found track evidence on several occasions of raccoon activity at air temperatures below -4.5°C when there was little snow cover on the ground. The mean distribution of daily activity for 15 day periods from February 18 to May 3 is graphically illus­ trated in Figures 8a to 8e (Figures based on data in Table 21, Appendix A). It appears that the opossum is highly nocturnal when not restricted by low temperatures (Figures 8d and 8e). Generally, the first animals leave their dens around sunset; activity rises to a peak around midnight or slightly later, then declines and ceases around sunrise (Figures 8d and 8e). well with field observations. These data agree During the warm months of the year from late spring onward, wild opossums are to be seen at night with rare exceptions. During late winter and early spring the mean activity distribution appears to be strongly modified by low temperatures (Figure 8a). Mean activity commences at a low level at sunset but declines abruptly again around 8:00 p.m. as the mean nocturnal temperature dips below the minimum tolerance 40 Figure 8a. The mean daily activity distribution related to mean air temperature for 8 opossums for the 15 day period February 18 to March 4 (120 opossum nights). The mean times of sunrise and sunset are designated by symbols. ACTIVITY UNITS /A N IM A L /H O U J I o — *» c# ^ Figure 8b. The mean daily activity distribution related to mean air temperature for 7 opossums for the 15 day period March 4 to March 19 (105 opossum nights). The mean times for sunrise and sunset are designated by symbols. X MAX. TEMP.« 7-7*0 r AT t : 4 0 P.M. TEMP. X MIN. TEMP.« - S I*C AT « :0 Z A.M MEAN ACTIVITY MAX 4 - MAX IX 105 OPOSSUM NIOHTS ■P CO 10 II IS ! t MIDNIGHT 3 4 5 A.M. f Figure 8c. The mean daily activity distribution related to mean air temperature for 6.9 opossums (daily mean) for the 15 day period March 19 to April 3 (103 opossum nights). The mean times for sunrise and sunset are designated by symbols. % •TV XHOINOIN H'4 ACTIVITY UNITS / A N IM A L / HOUR Figure 8d. The mean daily activity distribution related to mean air temperature for 7.8 opossums (daily mean) for the 15 day period April 3 to April 18 (117 opossum nights). The mean times of sunrise and sunset are designated by s y m b o l s . 1 ACTIVITY UNITS / — m c* * m * 1 1 1 • ■ ^ ■ NOUN j- • L *Ma4 ■ ANIMAL / / XHSINQIN / N*V / Z.+1 / r/ Figure 8e. The mean daily activity distribution related to mean air temperature for 7.6 opossums (daily.mean) for the 15 day period April 18 to May 3 (114 opossum nights). The mean times of sunrise and sunset are designated by symbols. 1 X MAX. TEMP. * 15 AT 4:05 P.M. •10* C TEMP 15 X MIN .TEMP. * t*7# C AT 5 :4 4 A.M. 10 5 MEAN ACTIVITY APE. 18- MAYS 114 OPOSSUM NIOHTS & ho II IE I MIDNIGHT A.M. to 50 level of about -5°C. One of the major ecological problems encountered by this highly nocturnal species in the north is illustrated by the truncated activity distribution in Figure 8a. In late winter and early spring, most opossums are near the starvation point. During this period, often the only available foraging hours occur in the early afternoon when daily temperature maxima exceed the freezing level. Hence, in February and early March, opossums may be occasionally seen foraging in broad daylight, an extremely unusual circumstance at any other time in the year. Fitch and Sandidge (19 53) describing the winter activity of Kansas opossums, write, "These occasional daytime forays seem to occur almost always in animals driven by hunger on winter days when the temperature has suddenly risen after periods of severely cold weather that have imposed activity and fasting." I have observed a number of such foraging individuals in late winter, usually in an emaciated condition. One animal made periodic visits to an apple orchard to feed on rotten apples in March (Area I). Its route from den to orchard and back was identical in four visits. Its trips coincided so predictably with daily maximum tempera­ tures that I anticipated its coming one warm afternoon and observed it feeding. To a lesser extent, the "tempera­ ture squeeze" on daily activity is also apparent in Figures 8b and 8c. The high level of activity in early April (Figure 8d) is a reflection of increased sexual activity which reached a peak during this period. Although all animals were confined separately and without visual contact, males and females responded to each other. Animals would make clicking sounds by opening the mouth suddenly and slightly causing the lips to smack (see also McManus 1967). They would move around rapidly in the pens and occasionally utter a harsh cry. Judging from field observations and activity data of penned animals (Figures 8a to 8e), two prominent features of the opossums daily activity pattern are as follows: (1) Activity occurs almost exclusively between sunset and sunrise except in late winter when it is apparently modified by a combination of extreme hunger and low environmental temperatures; and (2) nocturnal activity has a unimodal distribution with a peak near midnight. These activity characteristics appear to differ from those of other nocturnal species occurring with the opossum in one, or both respects. The raccoon is generally nocturnal in Michigan, yet tends to be somewhat active in daytime, especially in spring (Berner and Gysel 1967). Abies (1969) found crepuscular peaks of activity for Wisconsin red foxes, which appear to coincide with maximum prey activity. Abies reports that daytime fox activity is greater in winter than in summer.. The cottontail rabbit is largely nocturnal, with an apparent,annual cycle of 52 activity between 5:00 p.m. and 7:00 a.m., according to roadside counts (Lord 1963). In spring and summer, rabbit sightings along roads reach a peak before sunset (Illinois, Lord 1963) and after sunrise (Lord 1963, and Kline 1965, Iowa). Apparently voles and mice also show bimodal activity peaks at dawn and dusk (Hatfield 19 35, 1940, Hamilton 194 3, Calhoun 1945, Brown 1956). Winter snow appears to affect opossums primarily by covering the food supply and hindering travel. The opossum is poorly adapted to travel on soft snow, as it literally wallows through the snow with its low-slung body (Figure 9). Pruit (19 60) classifies the opossum as a chionophobe (snow hating) after Formozov's (1946) classification system. Reluctance to emerge from the den after a heavy snowfall was illustrated by penned animals in mid-January, 1968. On the night of January 14-15, the air temperature remained at the freezing point most of the night. Two animals made probing circuits in their pens and two animals did not emerge from their dens but fed on snow at their den entrances. In the remaining four pens, den entrances were either completely blocked by snow or were marked by small openings in the snow. In the field (Areas I and II), sub-freezing tempera­ tures prevailed for the first half of January and the ground was covered by about 40 cm of snow. I found no evidence that opossums had emerged from their dens during this time. On January 17, air temperatures rose above Figures 9a and b. An opossum weighing 1.83 kg, walking in fresh snow about 2 8 cm deep. Travel by this animal was extremely laborious; its tracks sunk in to a depth of 12 to 14 cm and its belly was dragging constantly. This opossum was placed in the snow for the photograph. At no time did I find indications that wild opossums venture forth in fresh, soft snow of this depth. 54 1 55 the freezing point marking the beginning of a six-day thaw. Tracks showed that some opossums had been active on the night of January 17-18, but apparently not in open areas where the snow was still soft. I followed two foraging trails in a pine plantation (Area I) where dripping melt-water from the foliage had frozen to a crust on the snow surface. These trails were entirely confined within the limits of the crusted area under the pines. These animals veered away whenever they encountered the soft snow at the periphery of the planta­ tion. A possible indication of discomfort in the snow is the opossum1s frequent habit of walking on fallen trees and logs where these are available (Figure 6), rather than on snow covered ground. Perhaps discomfort is due to the rapid heat loss from the bare, uninsulated toes in contact with snow. The opossum's locomotion appears to be greatly restricted by deep, soft snow. However, within the opossuirfs range in southern Michigan, restriction of loco­ motion per se by deep snow is probably a minor limiting / factor. There are, I believe, two reasons for this: Firstly mean snow depth is considerably less in southern Michigan than in many areas farther north, as will be discussed in Chapter X I . Usually, frequent thaws rapidly reduce deep snow cover following each snow storm in this region. Secondly, an ice crust forms rapidly op fresh snow during 56 periods of thaw when opossums are active. Only a thin crust is necessary to support an average sized opossum of about 2.5 kg. This species has a fairly slow and even mode of locomotion and each foot places minimum stress on the substrate as the body weight is gradually shifted onto it. CHAPTER IV DEN MICROCLIMATE AND ECOLOGY In winter, opossums generally use ground dens accord­ ing to my field observations in southern Michigan. Stuewer (1943) similarly found that opossums used ground dens exclusively in winter in south-western Michigan. Preference for ground dens over other den sites in the northern environment has been reported by Reynolds (1945) in Missouri, Fitch and Sandidge (1953) in Kansas, Wiseman and Hendrickson (19 50) in Iowa, and Hamilton (19 58) in New York. Ground dens provide an extremely stable and amelio­ rated microclimate in winter. Winter temperatures of artificial ground dens in the pens compared to external air temperatures are given in Figure 10. Since ground den nest cavities were occupied by opossums most of the time, den temperatures are 1.0° to 1.5°C higher than they would have been if dens had been empty. A mean temperature increment of 1.51°C due to body heat was calculated from 109 comparative temperature measurements of dens with and without animals. Between December 1, 1967 and March 29, 1968, the highest recorded air tempera­ ture was 21.1°C while the highest recorded artificial 57 Figure 10. Three day temperature means for air and ground dens at East Lansing, Michigan for the winter of 1967-1968. Each plotted point for ground den temperatures represents the mean of 36 measurements, the maximum and minumum measurements daily for six dens over three day periods. The mean winter temperature fluctuation for ground dens was 0.3°C. On this graph, the daily temperature fluctuation would be represented by slightly more than the thickness of the den temperature line. DEM TEMPERATURE in (D AIR TEMP. MAXIMUM‘S I 14 ■ I 20 ■ I 26 M > I1 I 1 I M » I 1 I ■ I 1 I T T 1 I 1 I I 7 13 19 25 31 6 12 18 24 I 7 JAM. PER. MAR. MIDPOINTS. THREE - DAY CLASSES 60 den temperature was 6.4°C (n = 1440). The highest recorded daily mean for artificial dens was 6.3°C. The lowest recorded temperature for air was -22.8°C while the lowest recorded artificial den temperature was -1.1°C. o The lowest daily mean for artificial dens was 0.6 C. The mean winter air temperature fluctuation for the 24 hour period was 9.9°C, and for ground den cativities it was 0.3°C. The winter temperatures of natural ground dens Cn = 53) are given in Table I. A description of these dens is given in Table 19, Appendix A. In general, these values parallel those for artificial dens. The dominant influence on ground den temperatures appears to be the heat budget of the soil. Soil tempera­ tures for December through March at various depths are given in Figure 11. Soil temperature means are based on U. S. Weather Bureau data for East Lansing. It is apparent that the winter soil temperatures do not decline below the freezing point at the mean depth of ground den nest cavities (100 cm). However, soil temperatures appear to decline to their lowest level in March more slowly than ground den temperatures (Figure 10 and 11). Quite probably this difference is due to the acceleration of convective heat loss by burrow air circulation. Measurements presented here agree with those of most previous studies of den temperatures (Gerstell 1939, Linduska 1947, Hayward 1965, and Davis 1966) which have been of shorter duration and limited in scope. Winter Table 1. Winter temperatures of natural ground dens at East Lansing, Michigan (Winter of 1967-1968). The range of air temperatures, 1 m above ground, was 21.1°C to -21.1°C for January through March, 1968. Number of Measurements Mean Insertion Depth of Probe, CM 1 10 213 2 7 221 3 2 4 Mean Den Temperatures C° Jan. Feb. Mar. Den Temp. Range (JanMar) C° -1.6 -1.6 2.2 -5.5 to 5.0 - .9 3.6 -1.1 to 3.6 180 - 1.7 - 0 to 3.3 2 200 _ .3 - - .6 to 1.1 5 8 158 - 1.3 1.7 0 to 3.3 6 13 212 3.4 CO • CM Den No. 3.6 -1.1 to 7.8 7 8 241 - 1.8 3.3 .6 to 4.4 8 3 200 - 1.4 - Mean (53 total) 203 .9 1.0 2.9 1.1 to 1.7 Figure 11. Five-day soil temperature means for various depths of East Lansing, Michigan, for the winter of 1967-1968. The soil is level Miami fine sandy loam under fescue grass. The graph is based on United States Weather Bureau data, U. S. Dept, of Commerce. 50*6 CM 05 CO DEPTH^ 2 0 9 *2 CM 101*0 CM , APPROXIMATE MEAN DEPTH OP 0 T " 'l """"'"I MAR. It ft 27 64 temperatures of one ground burrow as reported by Berner and Gysel (1967) are about 5°C to 7°C lower than values given here. The difference may be due to vertical insertion depths of the temperature probe less than those of the present study (Berner, personal communication). The great stability of the winter microclimate within ground dens is similar to that beneath the snow surface, although snow temperatures are somewhat lower (Formozov 1946, Pruitt 1956, Vose and Dunlap 1966). Stephenson (1969) measured winter temperatures in a Canadian beaver lodge. He found that they remained near the freezing point throughout the winter even though air temperatures declined to a mean minimum of -21°C. The winter temperatures of other natural shelters such as tree cavities, hollow logs and brush piles tend to be lower and far less stable than ground den tempera­ tures. A few temperature measurements of such shelters are as follows: On the morning of February 11, the air temperature was -9°C while an 8 m.p.h. wind was blowing. In the base of a large black oak tree (Area II), the cavity temperature was .5°C higher than the external air temperature. The thermometer probe was inserted upward into the tree trunk for a distance of 120 cm. On the same morning, temperatures were measured within a hollow elm log, 45 cm in diameter, and at ground level near the center of three large brush piles. In all cases, external and internal temperatures were found ,to be identical. Other investigators report similar observations. Linduska (19**7) found that temperatures at ground level near the center of brush piles in Area II averaged l.**°C higher than external air temperatures for January, February and March. Gerstell (19 39) in Pennsylvania reported that January temperatures within a hollow log were the same as those of the external environment; temperatures were about 2°C lower within a stump than outside. According to the data of Berner and Gysel (1967) for a raccoon tree cavity near the present study areas, the mean temperature within the tree cavity for January, February and March was -6.7°C, or 1.3°C cooler than the mean external temperature for the same period (these values are recalculated from their data). In mid-January, temperatures within this cavity declined to a minimum of -21°C (Berner 1965). With the approach of spring and warm weather, opossums tend to use above-ground den sites more frequently than ground dens as the above-ground sites gain heat more rapidly. In the experimental pens, most opossums utilized tree dens or ground dens in spring in such a way that they appeared to gain a thermal advantage. In Figure 12, it is evident that the occupancy of artificial tree dens increased markedly as the mean air temperature rose above the temperature for ground dens, around March 16. By April 10, a majority ' F animals (six out of eight) were sleeping in tree dens during daylight, hours (Figure 12). Figure 12. The winter occupancy of tree dens and ground dens during daylight hours by eight experi­ mentally penned opossums. Occupancy is compared to activity intensity, activity span, air temperature and ground den temperature for 196 8. All plotted points and bars represent five-day means. Percent utilization of tree dens is complementary to that of ground dens. For example if 75% of the animals (6 animals) occupy three dens, 25% (2 animals) would occupy ground dens. Animals generally remained in their respective dens for the duration of the daylight period. m MEAN ACTIVITY INTCNtlTY. UNITS/ANIMAL/HOUR CD MEAN ACTIVITY SPAN , HOURS /ANIMAL/NIOHT It 10 iLMiMMfiMflxi s- OROUND OEN TEMPERATURE _ I V --------- r ^ __ MAX., MEAN* ' J rMIN.AIR TEMPERATURE • UTILIZATION, TREE PENS to o . ______CL . . c u b i * "i SO tS I S II FEBRUARY I.. IS tl SB MARCH SI "T — S I ""I ■"■"T"' 11 | 10 IS to » I APRIL MAY 68 Utilization of ground dens increased again toward the end of April when the mean air temperature declined below that of ground dens. During the summer months I have observed that opossums will use a variety of other den sites besides ground dens. Opossums do occasionally occupy bulky leaf nests in winter in sheltered locations other than ground dens. I saw three such dens during the winter of 1967-1968, two in garages and one in a tree cavity. Temperature measure­ ments within one of these nests in a tree cavity are of particular interest because they appear to confirm that a primary requirement for the winter den environment are temperatures which do not significantly decline below the freezing level, rather them a specific requirement for a particular den location. Opossums used nine ground dens in Areas I and II; however one animal in Area II occupied a cavity within a large spreading base of a black oak (Figure m , Appendix B) for a period between January 21 and February 29. This large cavity was closely packed with oak leaves. On February 11, when the external air temperature was -7°C, the cavity temperature (within the nest) at a probe insertion depth of 190 cm was -.9°C; the opossum was in the den. Judging from two other temperature measurements of the same den in late February while the opossum was in the den, the temperature remained close to the freezing point in this winter. tree cavity for the duration of The microclimate within this heavily insulated 69 tree cavity is very comparable to that of ground dens and very unlike that of most other tree cavities. The two other leaf nests were in garages* one of which I investigated. This nest was a bulky affair* jammed behind implements and partly built into the rear wall of an unused garage. Two female opossums occupied the nest, sleeping about 30 cm apart. The nest was dismantled and contained about four bushels of maple and elm leaves. The garage owner said that he had seen the animals first in August, and that he had found a trail of leaves leading to the nest through the garage soon afterward. It is quite likely that opossums built these nests in the very interesting manner of carrying leaves in the coils of their tails (Lincecum 1872, Pray 1921, Smith 1 9 m , and Nestell's letter in the appendix, 1969). Although I have not observed this behavior pattern directly, I found a trail of grass and debris leading into three ground dens within the experimental pens. The opossum's occasional winter use and more extensive summer use of den sites other than ground dens in Michigan suggests that its winter occupancy of ground dens is a matter of thermal advantage (selection is probably guided by comfort) rather than innate preference. Some evidence that denning behavior may vary with latitude is provided by Yeager (19 39) who found that only 20% of winter dens in Mississippi were ground dens while 47% of the dens were in trees and logs. 70 Contrary to occasional statements in the literature, the opossum does not dig its own ground dens. It is obliged to use the ground dens excavated by other mammals, apparently because its grasping type of feet are unsuited for effective digging. Animals within experimental pens managed to escape by digging on only two occasions, even though the bases of the masonite pen walls were covered with dirt to a depth of merely 10 cm. One animal attempted to escape from a holding cage placed on the ground. It broke through the wire mesh floor of the cage but did not manage to dig under the cage. its cage in the garage. Another animal escaped from Its attempts to dig through the soft dirt under the garage door were quite ineffective although the opportunity to dig extended over the entire night period. At best,-the opossums digging attempts result in shallow scrapes. Generally, only a single opossum occupies a ground den at one time. However, individual opossums may spend the winter in the same den complex in company with other mammal species. The winter occupancy of seven ground dens in Area I, determined primarily by observation of tracks, is diagrammed in Figure 13 (temperatures for most of these dens are given in Table I; descriptions are given in the appendix). Den No. 5 with a single entrance held an opossum and a striped skunk continuously for at least two months. Den No. 7 (Figure 42, Appendix B) showed most signs of use by different species. During the course of the 71 Figure 13. The winter occupancy of seven natural ground dens in Area I by opossums and other mammal species. Each bar represents occupancy by a single animal; parallel, contiguous bars indicate simultaneous tenancy of the same den. Bars with square ends represent complete occupancy periods as shown. A slanted end to a bar indicates uncertainty as to the exact beginning or end of the occupancy period. Triangles represent burrow investigation or cursory use. IM 11! WINTER OCCUPANCY OF GROUND OCNS RACCOON FOX (F ) SKUNK(S) WOODCHUCK CW) OPOSSUM (O ) ifi F F //sss .-yy.'./• ,^ / s nmti DEN FFFF AAA4 F O ▲ ▲ JAN. I nr 15 F //////* i FES. I 15 mmrm NAR.I FFF JAL r 15 ” 1 ...... AM.I I 15 73 winter it was occupied by five mammal species. There were four entrances to this den, only one of which was consistently used. On February 28, an opossum, striped skunk, raccoon and woodchuck occupied this den simultaneously, presumably in different chambers of the den complex. On February 29, the woodchuck emerged from hibernation and as I approached the den obscured from its vision, it uttered a series of gutteral barks. until it saw me. It continued barking in a bellicose manner Subsequently, the woodchuck was the sole tenant of this den complex, the opossum moving to a very inferior, untenanted den (Den No. 10) about 150 m away (Figures 7 and 13) where it remained until the middle of March. The woodchuck remained in Den No. 7 until early April when a family of red foxes raised in Den No. 2, took over Den No. 7. It is of interest that a red fox stayed overnight on March 2 3-24 in Den No. 7 (Figure 13) simultan­ eously with the woodchuck during a heavy snowstorm which left 27 cm of snow on the ground. In the Literature, Lay (1942) reported simultaneous den occupancy by an opossum and an armadillo in Texas. Sandidge (19 53) in Kansas found evidence of one den housing both an opossum and a woodchuck. However, Reynolds (1945) in Missouri wrote,, "At no time were opossums found occupying dens with other species of mammals." These observations on the winter den tenancy of the opossum are not extensive. However, they do suggest that the opossum's winter den ecology neatly dovetails with that of other mammal species. During the winter, when energy conservation through the prevention of heat loss is vital to the opossum's survival, it occupies ground den complexes quite harmoniously with other species. In spring, the opossum's thermal dependency on the ground den ceases coincidentally with increased breeding competition for desirable dens by other mammals. However, the opossum's marsupium obviates the necessity of competition for dens. entering into breeding Hence, that portion of the opossum's ecological niche concerned with den use appears to overlap to a minimal degree with niches of other mammals, a circum­ stance predicated on the opossum's adaptable denning behavior. CHAPTER V FOOD AND FORAGING BEHAVIOR The opossum appears to be a carnivore by preference and omnivore by necessity. The rapacity with which captive animals will attack living prey brought near them leaves little doubt about the opossum^s dietary predilectious. Judging from the contents of twenty stomachs collected during January, February and March (Table 2) the winter diet of opossums consists largely of animals. Animal remains, primarily mammalian, occurred in 8S% of the stomachs and plant remains were found in 35% of the specimens. Most plant material was fibrous debris and grass, apparently swallowed incidentally. The importance of animal food in the winter diet has been noted by other workers. Taube (1942) examined 52 stomachs of Michigan opossums collected in September, November and December. He found that 81% of the total volume consisted of animal material, and 16.5% was vegetable. Eight specimens collected in March, April and May (Taube 1942) showed animals comprised 91% of the volume, and vegetable matter 9%. Dearborn's (19 32) Michigan specimens showed an incidence of 70% animal material and 30% plant 76 material. In New York (Hamilton 1958), the mean frequency of occurrence (n = 65) of various food items for the months of December through March are as follows, (means calculated from Hamilton's data): Mammals 60%, amphibians 20.2%, earth worms 32.5%, insects 2 2.9%, fruits 19.7% and green vegetation 48.5%. Apparently, certain plant foods are utilized when they are available. It has been previously noted that one opossum was observed to make frequent forays to an apple orchard in March, where it would feed on rotten apples. The one specimen (Table 2) containing corn was killed adjacent to a large corn field close to Area I. Wiseman and Hendrickson (19 50) reported that the bulk of the opossum's winter food in Iowa was corn, with insects next in importance. The high incidence, of mammals (70% occurrence, Table 2) in the winter diet is remarkable in view of the fact that the opossum can be outrun by almost all Michigan mammals. Apparently the opossum overcomes its disadvantage of low speed by surprising its victims at close quarters, or by eating carrion. Within lunging distance of its jaws, about 20 to 40 cm, the opossum visually recognizes mammalian food and acts swiftly. Yet, at a distance of about 1 m, it has difficulty recognizing a live mouse, judging from the behavior of captive specimens. I believe the ground burrow is of prime importance in providing a concentration point and occasional trap for the opossum's victims. Rabbits and mice are undoubtedly captured in the close confines of 77 Table 2. Stomach analysis of 20 opossums collected between January 2 3 and March 20, 1968. Frequency of Occurrence Striped Skunk Muskrat Cottontail Rabbit Whitetail Deer Opossum Shrew Field Mouse Unidentified Mammals Garter Snake Leopard Frog Insects Earthworms Plant Fibers 3 2 7 78 burrows where their escape routes are cut off. Probably many, if not most, earthworms, insects, snakes and amphibians (Table 2) in the winter diet, are obtained below the frost line in ground burrows. Apparently some mammal food is carrion. (Table 2) fed on whitetail deer carcasses. Two animals The muskrat remains in one stomach consisted of skin and fur, with no flesh or bone, indicating carrion origin. I doubt that the striped skunk remains in one specimen are the result of predation. Opossums probably locate some carrion in ground dens. Allen and Shapton (1942) reported finding dead striped skunks in three ground dens and a rabbit in another of the 36 dens they excavated and examined. Opossums will readily eat their own kind. I have observed young opossums, three months old, feeding on a dead litter mate. Only one clear case of cannibalism was noted in the 20 specimens examined, although some opossum hair occurred in almost all stomachs. A case of cannibalism was observed in the experimental food balance study. A female opossum (2.08 kg) was introduced into a pen contain­ ing a dead opossum in the den cavity. This animal fed on the dead opossum from February 19 to March 16, apparently sleeping next to the carcass. March 5. The carcass began smelling on During this period, the total calculated energy cost of survival for the 2.08 kg opossum was 4167 kcal (metabolic rate based on food balance data). Of this amount, 1212 kcal were provided by dog food, leaving 2955 kcal to be 79 provided by cannibalism. Using the combustion value of 1.37 kcal/gm wet weight for whole field mice (Golley I960), the dead 2.0 kg opossum must have provided 2740 kcal of energy. These calculations indicate that the carcass was probably entirely consumed. From a nutritional standpoint, the unusually good health of this individual seemed to confirm that cannibalism provides the best ratio of amino acids in the diet. Cannibalism may be a significant factor in the winter survival of opossum populations. late in the year. Many opossums are born Most of these specimens are rat-sized (less than 2 kg) at the beginning of winter and have a small chance of survival. Most opossums of this size collected in January, February or March, were at the starva­ tion point, with exhausted fat depots (see Chapter IX). Such animals are quite likely to die in ground dens where they provide a potential energy source for other opossums. It is conceivable that the protein pool created by these late-born individuals contributes to the winter survival of some breeding stock. An interesting sidelight of opossum nutrition in winter is the practice of coprophagy by this species. ing opossum trails in winter snow, I have of coprophagy several times. While follow­ observed evidence Three of the experimental animals were observed to coprophagize. Refection by one of these opossums (No. 4, Figure 30, Appendix B) was quite regular in the first seven days in the metabolism cage. 80 Initially, a segment of the wire mesh cage bottom was covered by a small sleeping platform. Feces collecting on this platform were eaten by this individual. the platform was removed, preventing coprophagy. Subsequently, There was no change in the weight level maintained by this individual after coprophagizing had ceased, indicating that the energy derived from feces must have been minimal. A number of workers have reported on significant nutritional benefits derived from coprophagy by rabbits and rats. Recycling of the gut contents harnesses bacterial activity in the lower intestinal tract and apparently has positive effects on the availability of the following nutrients: Vitamin K (Nightingale et al. 1947, Johanssen and Sarles 1949, Barnes and Fiala 1959, Mameesh and Johnson 1960, and Wostmann et al. 1963); Riboflavin and Niacin (Kulwich et al. 1953); Pantothenic acid (Fridericia et al. 1927, Kulwich et al. 19 53 and Daft et al. 1963); and Vitamin B^2 (Kulwich et al. 1953, and Barnes et al. 1957). What specific role coprophagy plays in the opossum's winter nutrition is an open question. It may be that refection supplements certain nutrients which decline to a critical level as much of the energy for winter survival is derived from body tissue catabolism. The varied winter diet of the opossum reflects its opportunistic foraging habits. Winter trails appear to to wander aimlessly through the countryside. seems to be a pattern. Yet, there For example, on January 21, I followed 81 the trail of one individual continuously for 2.55 km in 15 cm of snow. The maximum radial distance of this trail from the den of exit and reentry was O.1! km. The trail led through several ecological types; percentages of the total length of the foraging trip are as fellows: 65% lowland hardwoods including elm, ash and swamp white oak, 10% lowland swale with marsh grasses, cattails and buttonbush, 5% upland grassland, 5% along upland fence rows and 15% upland woods consisting largely of black oak. The opossum proceeded from one bare spot of ground to another in its search for food. This is typical. I have found that opossums seldom attempt to locate food beneath snow unless the snow blanket is less than about 5 cm in depth. These snow-free areas are usually located at the base of trees, logs and debris ("snow shadow", Pruitt 1960). At one point the opossum vigorously attacked the rotten underside of a log.; beetle remains testified to its search for insects. On several occasions, it walked the full length of logs and wind-fallen tree limbs rather than in the snow. there was evidence of coprophagy. Twice The opossum crossed an open clearing and deliberately walked through a large brush pile and two rose thickets. For about 75 m, the animal walked on the ice in the middle of a small stream before crossing to the other side. directly to the den. Finally, the trail returned Thus, opossum trails are less haphazard than they may appear to be. In the process of investigating nooks which might hold food, the trail seems 82 to wander. There is one strong impression I have obtained in following opossum trails. Opossums appear to be quite unsuccessful at capturing any of the homiotherms above ground in winter. Unlike foraging conditions in summer, young and helpless warm blooded prey are not available in winter. In view of the high incidence of warm blooded animals in the winter diet, it seems that many of these . prey items must be obtained as carrion or captured alive in dens. In summary, on a typical winter foraging trip, an opossum may find miscellaneous leavings such as scats, bones and fur of other carnivores, odd insects, berries and other plant items. On rare occasions it may come upon a bonanza in the form of a carcass. Success on these above-ground foraging trips appears to be highly variable, often extremely poor. Yet, even minute quantities of certain food items may provide essential nutrients such . as trace minerals and vitamins which may reach a marginal level as much of the energy for winter survival is provided by tissue catabolism (see Chapters VIII and IX). CHAPTER VI TEMPERATURE REGULATION Body temperatures (Tg) of resting opossums plotted against ambient temperatures (Figure 1*0. show a linear relationship The regression equation Tg (C°) = 33.65 + 0.048 TA was calculated from 103 observations on five animals by the method of least squares. The "t" test for the regression coefficient is significant (p ^ 0.05, Snedecor 19 56). The resting Tg curve (Figure 1*0 shows a lability of 2.16°C over a TA range of47°C. On the basis of the regression equation, the Tg is 33.6°C at a TA of 0°C and Tg equals TA at an ambient temperature of 35.4°C. Greater variation in body temperature was displayed by individual animals. A large male (mean weight 4.79 kg) exhibited a Tg fo 32.0°C at TA -12.0°C, and a Tg of 36.8°C at Ta 31.8°C. This is a Tg range of 4.8°C. weight 2.73 kg) had a T o A female (mean of 29.8°C at T. -10.0°C (the lowest A value recorded in this study) and a Tg of 35.5°C at TA 31.8°C, o representing a range of 5.7 C for resting Tg. It should be noted that animals showed no signs of torpor at all TA *s and responded normally to stimuli at the lowest resting Tg values recorded here. 83 84 Body temperatures recorded during moderate activity (Figure 14) are one to three degrees higher than for resting animals. All activity values were obtained at night when animals were normally active. The difference between Tg's of active and resting animals is most pronounced at low ambient temperatures; this difference is only about 1°C or less at TA 35°C (Figure 14). The data indicate that on those winter days when opossums are active for a few hours, their Tfi 's would show an apparent diel cycle, being higher at night when animals are active. Comparison with Morrison's (1946) data for Didelphis marsupialis etensis is interesting. This tropical subspecies on Barro Colorado Island showed no diurnal Tfi cycle. Body temperatures varied at random over a 2.5°C range, while the TA 's ranged from 24° to 29°C. Morrison's mean value for day and night Tfi (50 measurements, two animals) is 35.5°C. Apparently the northern and tropical forms of Didelphis marsupialis differ little in the charac­ teristics of their labile thermoregulation. The pronounced diel Tg range of the northern form during the cold months is probably an obligatory response to the greater difference between Tg and prevailing TA 's. Scattered body temperature values are given by several investigators. These studies have been of short duration and under laboratory conditions. Johnson (19 31) obtained five rectal temperatures from one animal ranging from 33.6°C to 35.5°C, with a TA range of 25° to 29°C. This animal had a Tg varying from 33.5 to 34.7°C in a refrigerator Figure 14. Opossum body temperatures (C ) are plotted against ambient (environmental) temperatures for the period January through March, 1966. Hashed lines include one standard deviation on either side of the regression line. 86 i 87 (T^ 4° to 6°C). Wislocki (1933) obtained nine rectal temperatures ranging from 32° to 34.5°C at 14° to 26°C. Britton and Atkinson (1938) state, "The opossum, Didelphia virginiana...showed practically no reduction in body temperature after exposure to approximately 0°C for several days." Response of the opossum's body temperatures to controlled changes in ambient temperature have been studied by McManus (1969). His animals were kept outdoors at late summer temperatures (July, August and September). Rectal temperatures of these opossums remained essentially constant at approximately 35.5°C as the ambient chamber temperature was experimentally lowered to 3°C at the rate of 1°C every 6.8 minutes. McManus also recorded rectal temperatures at prevailing environmental temperatures of opossums kept outdoors during June, July, August, November and December. These measurements (interpreted from graph) show a slight drop in Tg of about 35.8°C at T^ 35.0°C, to about 34.8°C at T^ 5.0°C. The latter Tg measurements agree with those of this study (Figure 14) if the present activity and resting Tg values are lumped (McManus did not separate Tg measurements according to activity level). The Tg measurements of McManus (1969) obtained at experimentally declining T^'s compared to Tg measurements of winter-acclimatized animals in the present study, illustrate the difference between acclimation, a short-term response and acclimatization, a long-term response. 88 Hart (1957) has discussed the importance of distinguishing between these thermoregulatory responses. The opossum apparently regulates quite precisely when it is subjected to short-term temperature changes t as the data of McManus show. However, the present data indicate that the opossum's thermoregulatory responses to seasonal changes in the environment are distinctly different from its short-term response. This is of considerable ecological importance. Significant energy savings can be atrributed to the opossum's lower resting Tg in the normal winter environment, as I will show in Chapter VII. It is important that conditions are specified under which temperatures are recorded. These conditions include depth at which the Tg measurement was taken, ambient temperature, level of activity, length etc. of acclimation, Scott (19 38) quotes rectal temperatures ranging from 32.0° to 37.1°C in a laboratory with a TA of 22° to 26° C. It is possible that Scott's values were recorded under varying conditions, e.g., the present study showed that depth of thermometer insertion is important. At a TA ranging from -12° to 9°C, the mean of 20 Tg readings taken simultaneously at a depth of 3 cm is 31.39°C, an average of 1.89°C lower than deep body readings. Again, it was found that Tg of fed animals was about 1°C higher than for the fasting condition. Low body temperatures and labile thermoregulation (i.e., imprecise thermoregulation) have been reported for 89 other marsupials (Enders and Davis 1936, Morrison 1946, Bartholomew and Hudson 19 62) and other mammals including monotremes (Martin 1903), pangolins (Eisentraut 1960), edentates (Wislocki and Enders 1935, Irving et al. 1942, Johansen 1961) and fossorial rodents (McNab 1966). Bartholomew (1956) measured the rectal temperature of the Australian marsupial Setonyx brachyurus, a small wallaby. He found that generally the Tg remained most values between 37°C and 38°C. quite stable with The Tg of two out of three animals declined to a low of about 36.5°C when the T^ was experimentally dropped from 20°C to3°C and held at the lower value for four hours. Most mammals with labile thermoregulation are tropical or subtropical forms. In the temperate environment, the opossum may exhibit the lowest normal body temperatures compared with other non-aestivating, non-hibernating mammals associated with it. Morrison and Ryser (1952) give a mean Tg of 38°C for 17 species (excluding monotremes, marsupials, edentates and bats) weighing from one to 10 kg. The mean Tg of 56 species weighing from 10 gm to 1000 kg is given as 37.8°C. These values are considerably higher than the opossum's body temperatures. CHAPTER VII ENERGY REQUIREMENTS BY OXYGEN CONSUMPTION Resting oxygen consumption curves for four opossums are plotted in Figure 15. The equation for the line (fitted by eye) representing all data is: = 0.51 - 0.014 TA (C°). are given in Table 3. RMR (ml (^/g/hr) Values calculated from the equation The general RMR line is projected to 35.4°C, the point of zero thermostatic heat requirement (where Tfi equals T^). The mean of 17 oxygen consumption values for slight to moderate activity (standing and slow movement) is 0.93 ml Og/g/hr for three animals (s = 0.3). The latter measurements for moderate activity were made at a mean TA if 5QC and a range in TA of -4° to 11°C. The type of apparatus used here precluded metabolic measurements at higher activity levels. The extremely low general BMR level is striking (Figure 15). A male (mean weight for the study was 4.79 kg) had a mean BMR of 0.15 ml Og/g/hr (n =22), at TA fs between 23.2°C and 2 8.6°C. A female (mean weight for the study was 2.73 kg) had a mean BMR of 0.16 ml O^/g/hr (n = 19) at TA 28.9°C. Basal metabolic rates of comparable, non-hibernating mammals are given in Table 4. It is apparent that the observed BMR values for the opossum are 90 91 Table 3. Resting oxygen consumption and metabolism calcu­ lated from the equation RMR (ml 0 2 /g/hr) = 0.51 - 0.014 Ta . Ta ml 02/g/hr kcal/kg/24 hrs. -10 .65 74.88 -5 .58 66. 81 0 .51 58.75 5 .44 50.69 10 .37 42.62 15 .29 33.41 20 .22 25.34 25 .15 17.28 Basal 30 .15 17.28 Basal . 01 0 .73 00 d- -15 Table Basal metabolic rates of some winter-acclimatized mammals. (1945) equation: BMR (kcal/kg/24 hrs) = 70.5 wt. (kg)“*27 Animal and Reference Weight kg Critical Observed1 Temp. BMR Tc °C kcal/kg/24 hrs Calculated from Brody's Observed Calculated BMR BMR ml 02/g/hr ml 02/g/hr BMR ratio Observe^ Calculated ml 02 /g/hr Opossum, present study 4.79 21 16.81 .15 .40 .37 Opossum, present study 2.73 29 17.82 .16 .47 .34 Red Fox, Irving et al. 1955 5.01 -13 56.40 .50 .40 1.25 Arctic Fox, Scholander et al. 1950 5.50 -40 64.18 .58 .39 1.49 Arctic Fox, Scholander et al. 1950 4.60 -40 58.26 .52 .40 1.30 Arctic Fox, Scholander et al. 1950 4.00 -40 58.50 .52 .42 1.24 Porcupine, Irving et al. 1955 3.20 -12 50.76 .45 .45 1.00 Porcupine, Irving et al. 1955 6.70 -12 48.50 .43 .37 1.16 Porcupine, Irving et al. 1955 7.66 -12 38.35 .34 .35 .97 1 Liter oxygen is equivalent to 4.8 kcal 93 Figure 15. Resting oxygen consumption versus ambient temperatures for fully acclimatized opossums, determined from January through July, 1966. 1 Values to the left in the graph were collected n in winter and values to the right in summer. Solid symbols represent oxygen consumption measurements. Open symbols represent mean valued for the individual animals for two degree Ta | classes. The four dashed lines have been fitted! by eye to mean values for individual animals. The solid line is based on all data. «• w ♦ less than 40% of the value predicted by Brody's (1945) equation relating BMR to body weight in mammals, where: -.27 BMR (kcal/kg/24 hrs) = 70.5 wt(kg).. The extremely low BMR of the opossum compares with 0.25 ml Og/g/hr of the armadillo (mean weight 3.7 kg. Johansen 1961) and 0.13 ml 02/g/hr of the three-toed sloth (mean weight 5 kg, Irving et al. 1942). For any mammal with a relatively stable Tg, the RMR below the critical temperature may be described by the following equation (Kleiber 1961, Gordon et al. 1968): RMR = (Tb - Ta ) Where C, often called "conductance", is a constant equiva­ lent to the rate of heat loss and is given by the slope of the line for RMR measurements at various TA 's. The value of C is a function of the temperature gradient maintained between Tg and T^. However, for a labile homiotherm like the opossum, RMR is displaced downward from original pro­ portionality as the Tg declines with declining TA . Hence the value of C as obtained from the RMR line, which is 0.014 ml 02/g/hr/°C for the opossum (Figure 15), differs from C values calculated on the basis of actual temperature differentials maintained between Tg and TA , at specified T a 's . For example, at T^ 2 5°C (the lower critical tempera- - ture, where Tg is 34.8°C) the value of C is 0.016 ml 02/g/hr/ while at 0° (Tg 33.6°C), the value of C is .015 ml 02/g/hr/°C The opossum realizes some energy economy due to sliding Tg. If the opossum were to maintain a theoretical Tg of 35.4 in the manner of a rigid homiotherm, the energy cost would be greater by 8% of the normal energy cost above the basal level at TA 0°C (on the basis of data for all four animals, and C = 0.015 ml 02/g/hr/°C). In the case of the large male CH.79 kg), the equivalent energy savings is 12%. McNab (1966) reported a strong downward displacement of RMR (experimental TA 's) for the tropical mole-rat Heterocephalus glaber. shows remarkably poikilothermy. He found that this tropical mammal poor thermoregulation, verging on It is well known that homiotherms realize various levels of energy economy by a reduction of metabolic rate and associated Tg. Generally, the latter is a controlled response to specific stimuli. Such energy economy has been reported for torpid homiotherms (Pearson 1953, Lasiewski 1963, Hudson 1965, Tucker 1966 and others) and hibernating mammals (Kayser 1965 and others). Steen (1958) reported readjustment of Tg downward by small northern birds exposed to cold at night when they appeared unable to maintain their body temperature. Unlike these homiotherms, the opossum apparently achieves energy economy as an obligatory by-product of imprecise thermoregulation. The critical temperature (T^) of the opossum is approximately 25°C, a relatively high value for a temperate mammal. Yet, the opossum's rate of heat loss or energy expenditure at TA 0°C is comparable to that of a placental mammal with a BMR of 0.5 ml Og/g/hr and a T q of 0°C (see Table U and Figure 15). In this connection, metabolic slopes of mammals with different BMR's have been 97 graphically compared (Scholander 1950, 1955, Gordon et al. 1968 and others) by converting BMR to a common basal level (Basal 100 etc.). When the mammals being compared have widely different BMR's such comparisons will considerably distort relationships between metabolic rates. The energy cost at low T^'s of any mammal with a relatively high and low BMR (like the opossum) will appear inflated. A noteworthy aspect of opossum metabolism is the normal occurrence of shivering thermogenesis. Fluctuations in RMR below approximately 10°C are correlated with shivering (Figure 16). The regularity of shivering thermogenesis is astonishing. I have observed sleeping opossums showing alternate periods of shivering and quiet for hours at low T^'s. The comparative importance of shivering and non­ shivering thermogenesis in mammals is presently not clear (Hemingway 1963). Apparently, with acclimation, placental mammals rely less on shivering and more on a non-shivering metabolic response to cold. Hemingway reviews evidence that shivering is an emergency mechanism and functions only in severe cold. In contrast to most northern mammals, shivering is employed frequently by winter acclimatized opossums, warranting consideration as a normal form of thermogenesis. Johansen (1961) reports the common occurrence of shivering among armadillos at low T ^ ’s. Shivering and attendant increase in temperature were occasionally observed at a T^ of 30°C. (Johansen's armadillos were not acclima­ tized to experimental T^'s). West (1965) concludes that Figure 16. Sequence of resting oxygen consumption measure­ ments for two opossums, showing metabolic fluctuations associated with shivering at low T^'s. Dash marks represent shivering. Ta 4*C •ft 4 •2 •I 1*0 J <0 2 * 7 3 KG to RMR ML • TIME , M IN. 100 shivering and muscular activity of extra heat production in are the principle means birds. of non-shivering thermogenesis. He found no evidence While the evidence is meager, there is a hint that shivering thermogenesis may be the more primitive response. In view of the decline in Tg with declining T^, one would infer that a minimum must exist below which a further decline in Tg places an intolerable on the organism. physiological stress The RMR shows increasingly wide fluctuations (Figure 16) as the T^ declines below 0°C. Winter sleeping opossums observed at Ta -12°C showed spells of deep breathing, much shivering and much shifting of positions. Evidence has been presented above that opossums avoid sub­ freezing temperatures by utilizing the micro-environment of ground dens when the general air temperature declines below the freezing point. Hence, it appears that the "Achilles heel" of the opossum's unusual thermoregulatory-metabolic response is temperature sensitivity at a mean Tg below approximately 33°C (at T^ -5°C, the mean Tg is 33.41°C). Irving (1966) has reported a considerable internal tempera­ ture gradient from core to extremities in a number of species at low ?a '8> At winter's end, many wild opossums display recently amputated ears and tails from freezing. This suggests difficulty in the maintenance of minimum temperatures in poorly insulated extremities. McManus (1969) measured rectal and tail temperatures of opossums subjected to ambient temperatures experimentally lowered 101 to 3°C. At 3°C, the Tg's in the tail, 12.7 cm (5 in.) and 20.3 cm (8 in.) from its base, were about 12°C while the rectal temperature was 34.5°C (values interpreted from graph). Thus, McManus* measurements indicate that the temperature gradient between Tg at the end of the tail and was about 9°C at 3°C. From the standpoint of energy economy, the stable microclimate of the ground den with above-freezing T^'s is of pivotal importance. This is well illustrated by the oxygen consumption curve in Fugure 15. It has been pre­ viously noted that the resting energy expenditure of the opossum (Figure 15) does not exceed the basal energy expenditure (about .5 ml 0 2 /g/hr) of comparable cold-adapted mammals at T^'s of 0°C or above. Representative BMR.’s (Table 4) of 0.55 ml 0 2 /g/hr for the arctic fox (Scholander et al. 1950), .50 ml 02/g/hr for the red fox and .43 ml Og/g/hr for the porcupine (Irving et al. 1955) are comparable to the opossum's RMR at 0°C. The ecological importance of the opossum's use of ground dens can hardly be over-emphasized. For example, if the opossum were to sleep in a tree cavity at -15°C (a situation it would not tolerate by choice), the oxygen consumption rate would increase to 0.7 ml 02/g/hr, a 40% metabolic increase over the RMR level in a ground burrow at T^ 0°C. For a 2.5 kg opossum, this metabolic increment per day is equivalent to 58 kcal, or in terms of prey, the energy provided by one additional field mouse each day* 102 The data presented here suggest that the opossum's unusual thermoregulatory-metabolic complex, coupled with its denning behavior, serve as an effective alternative to adaptive fur insulation. As Scholander (1955) has pointed ou t, insulation has been the main avenue for climatic adaptation in mammals. If indeed this alternative physio­ logical response is as successful as it appears to be for the opossum, why is it such a rare phenomenon among temperate mammals? Perhaps the answer is that this adaptive response (particularly low metabolic rate) is necessarily associated with a low key mode of life for which the availability of accommodating ecological niches is limited. The almost universal occurrence of higher metabolic rates (excluding the special case of hibernation) and a Tg of about 38°C (Morrison and Ryser, 1952) for most mammals, suggest that these physiological characteristics confer to mammals a competitive advantage and a greater ability to extract energy from potential ecological niches. CHAPTER VIII ENERGY REQUIREMENTS BY FOOD BALANCE CALORIMETRY Metabolizable Energy of Experimental Food The designation "metabolizable energy" (M.E.) used here refers to the gross energy (combustion value for food) minus urinary and fecal energy. Energy values (kcal) of feces and urine of five opossums are given in Table 5. It will be noted that the dry and wet energy content of fecal material is quite similar for the five experimental animals. There is greater variation in the urinary energy between individuals, both in the wet and dry conditions. The relatively low energy value of wet urine for animal No. 5 is a consequence of the copious amount of urine produced by this individual. The mean urinary energy value of 1.467 kcal/g dry weight may seem low compared to the mean value of 3.322 kcal/g dry weight reported for bobcat urine (Golley et al. 1965). However, the fraction of non-combustible material in the urine was high. For example, for animal No. 2, the mean weight of non-combustible residue in the urine com­ prised 54.12% of the total weight (n = 5, range = 54.16 54.36%). The combustion value of the ash-free dry material averaged 3.248 kcal/g which is higher than the calculated 10 3 Energy values (kcal) determined by bomb calorimeter for feces and urine of opossums on a maintenance diet in autumn. Opossum No. and Sex 10 Mean f c ? f c CO 9 9 Urinary Energy Content dry wet kcal/g kcal/g 3.311 1.120 1.207 .025 3.45 3.307 1.359 1.477 .040 2,44 3.439 1.124 1.212 .014 3.29 3.538 1.369 1.607 .046 2.01 3.585 1.341 1.830 .047 2.84 3.436 1.263 1.467 .034 • 5 9 Fecal Energy Content dry wet kcal/g kcal/g o 2 CO 1 Mean Weight 401 Table 5. 105 combustion values for urea (2.52 kcal/g) and uric acid (2.74 kcal/g) containing the major energy fraction in mammalian urine. The energy value of the kibbled dog food used in the winter food balance energy determinations is 4.010 kcal/g. The energy value of the dog food batch used to calculate energy values in Table 7 is 4.220 kcal/g. These values are slightly lower than 4,365 kcal/g quoted by the Ralston Purina Company (personal communication, Corbin, 1968). The food intake and excretory output of opossums on a maintenance diet is given in Table 6 . These values and the values from Table 5 were used to calculate energy intake, excretory energy loss and metabolizable energy, which are given in Table 7. The high variability of fecal output of animal No. 2 (Table 6 ) is due to the fact that this animal defecated every second day and output is expressed on a daily basis. When animals are compared, gross and metabolizable energy requirements and excretory energy loss per unit body weight, increase with decreasing body weight (Table 7) as might be expected with the increase in ratio of surface area to body weight as the weight declines (proportional to W*^®, Kleiber 1961). However, it is interesting to note that per animal, the gross and meta­ bolizable energy requirements and excretory energy loss are similar for all five animals (Table 7). The latter effect is predictable because the decrease in metabolic rate due to the decrease in total body size (or weight) tends to be Table 6. Food intake and excretory output of opossuas on a Maintenance diet in autumn. or minus one standard deviation. Mean Food Food Intake Feces Output H-an Intake Per Unit Per Animal Opossum Days of Weight Per Animal Body Wt. (Wet) Mo. and Sex Experiment (kg) (g/day) (g/kg/day) (g/day) l 9 2 2. 5 9 10 V 9 o' Means (n = 5) IS IS 15 21 21 3.03 3.>tS 2.4*1 3.29 2.01 65.0 6S.0 6S.0 65.7 60.0 21.4 18.8 26.6 20.0 29.8 28.9 28.1 34.9 3S.S 30.5 17.4 2.84 64.1 22.5 31.6 ± 19.1 t 28.6 t 12.0 1 9.7 t 14.0 Feces Output Per Unit Body Wt. (g/kg/day) 9.S 8.1 14.3 10.8 15.2 11.1 t i i i i 6.3 8.3 4.9 2.9 7.0 Mean values are given plus Urine Output Urine Output Per Animal Per Unit Body Wt. (ml/kg/day) (ml/kg/day) 64.4 100.5 218.0 85.1 73.5 t ± t t t 23.9 17.7 61.0 24.8 32.7 108.3 21.2 29.1 89.3 25.9 36.6 7.9 ± 5.1 ± 25.0 ± 7.5 t 16.3 t 38.1 H o CD Table 7. Energy intake* excretory energy loss and metabolizable energy of opossums on a maintenance diet in autumn. 1 2 5 £ & 9 10 % » Mean % (kcal/kg/day) Fecal Energy Per Unit Body Wt. (kcal/kg/ day) Urinary Energy Per Animal (kcal/day) Urinary Energy Per Unit Body Wt. (kcal/kg/ day) M.E. Per Animal Per Day (kcal/ day) 260.65 260.65 260.65 277.25 253.20 86.02 75.55 106.82 84.27 125.97 32.37 38.19 39.23 48.60 40.90 10.68 11.07 16.08 14.77 20.35 1.63 4.02 3.05 3.91 3.45 .54 1.16 1.25 1.19 1.72 226.65 218.44 218.37 224.74 208.85 74.80 63.32 89.49 68.31 103.90 86.95 83.80 83.78 81.06 82.48 262.48 95.73 39.86 14.59 3.21 1.17 219.41 79.96 83.61 3.03 3.45 2.44 3.29 2.01 2.84 Opossum Mean No. and Weight Sex (kg) Gtobs Energy Intake Per Unit Body Fecal Energy Per Animal (kcal/day) Gross Energy Intake Per Animal (kcal/day) M.E. Per Unit Body Wt. Per Day (kcal/kg/ day) M.E. as t Energy Intake 107 compensated by an increase in metabolic rate due to the greater surface area to weight ratio in smaller animals. The efficiency of dog food in terms of metabolizable energy is given in Table 7. These values do not deviate more than 3.34% from the mean of 83.61% for all five animals. For the bobcat, Golley et al. (1965) report that the fraction of metabolizable energy as percent of energy intake ranged from 79% to 82% for three animals which did not gain weight. These values are quite similar to those determined here, even though the diet of bobcats was deer meat. Other M.E. fractions for bobcats which gained weight on a rabbit and chicken diet (Golley et al. 1965) ranged from 61% to 84%. Maximum Energy Intake An estimate of maximum energy intake per day was required for calculating the winter energetics models in Chapter X. Forty-three feeding trials were conducted. Dead cottontail rabbits opossums in winter. were fed ad libitum to caged Usually, a 1.5 kg rabbit was consumed by an opossum in three or four nights. completely consumed in all cases. Rabbits were almost Often, the only remnants were the distal portion of a leg or piece of fur. Opossums attacked carcasses with avidity. Their strong- jaws opened carcasses at almost any point, often at the ventral posterior end. Body juices were licked out; forepaws were used for grasping and holding. Sometimes the head was P p v4 ' 4 108 eaten first. female. On one occasion, smelt was offered to a 3*0 kg She ate 531 g of smelt in one night (value not included here). Table 8 . Ad libitum daily ingestion of whole rabbit by opossums in winter. The mean weight of rabbit ingested is given plus or minus one standard deviation. Body wt., of opossum (kg) 2.10 o* 2.50 t 2 .0<* 2.08 2.05 2.38 2.15 2 .0>* 2.19 2.13 Mar. 4 - 16 Mar. 17 - 23 Mar. 2<* - 30 13 7 7 Feb. 19 - 28 Feb. 29 - Mar. 2 Mar. 9 - 1 3 10 Feb. 16 - 21* Mar. 17 - 25 Mar. 2i* - Apr. 2 9 9 10 Feb. 16 - 28 13 2.23 2.40 3.75 2.79 3.08 3.43 2.75 3.09 2.61 2.89 4.05 3.18 2.52 3.08 2.82 2.95 1.84 Feb. 2i» - Mar. 5 Mar. 8 - 1 2 11 5 Means (n = 19) 2.2S Mar. 21* - 30 3 5 7 6.9 8.3 10.3 3.2 3.3 2.5 6.5 5.7 4.8 2.6 6.4 10.9 10.8 3.5 7.3 13.5 5.9 2.7 2.88 6.7 2.28 2.58 4.66 11.2 3.6 7.0 10.1 3.5 2.92 8.2 4.1 3.7 3.7 119 An equation for opossum B.M.R. can be calculated from the oxygen consumption data. The mean weight of the four animals uued in oxygen consumption measurements was 3.48 kg. The basal energy expenditure at the latter weight is 60.2 kcal/24 hrs (see data given under oxygen consumption). On this basis, the equation for opossum B.M.R. becomes: Opossum B.M.R. (kcal/24 hrs) = 23.6 W(kg ) * 75 The factor of 23.6 is probably among the lowest for mammals! Agreement between energy expenditure estimates of sleeping opossums by food balance calorimetry and by oxygen consumption is good. tion; R.M.R. Using the equation for oxygen consump­ (ml/g/hr) .= 0.51 - 0.014 T^ (C°), and assuming 4.8 kcal/1 of oxygen consumed at a mean body weight of 2.38 kg and a mean winter temperature for ground dens of 2 .2 °C, the energy expenditure estimate by oxygen consumption is 2.3 kcal/kg/hr. The equivalent estimate by food balance is 1.96 kcal/kg/hr for a mean body weight of 2.38 kg, as given in Table 9. Energy requirements for higher activity levels are given in Table 10. It is evident that the M. E. values vary widely between and within measurements for various individuals. There is some correlation with mean activity intensity measured by switch closures at ground den entrances. However, the correlation is poor. and tree 120 Energy Equivalent of Body Weight Loss The energy equivalents of body weight loss are given in Table 11. Energy equivalents were calculated from minimum M.E. rates measured for individual animals (Table 9) applied to their known weight losses during periods when these animals stayed in their dens continuously. The mean M.E. value for seven animals is 4.38 kcal/g weight lost; the range is 3.03 to 5.35 kcal/g weight loss (Table 11). If the mean value of 4.4 kcal/g weight loss approximates the true value, then a considerable amount of moisture must comprise the weight loss in opossums. Brown et al. (1951) report values for high density belly fat of swine as follows: ether extract 63.36%, protein 8.59%, moisture 27.76%, and ash .37%. On this basis, assuming a ratio of: Protein _ 1 Fat = 7 * for composition of weight loss, and the combustion value for lipid is 9.5 kcal/g, and the combustion value for protein is 5.6 kcal/g, (Brody 1945) then to produce an energy value of 4.4 kcal/g of weight lost, approximately 42% of the . weight lost is lipid, 6 % is protein and 52% is moisture and ash. Energy Requirements Concurrent with a Body Weight Loss of 40% The M.E. expended while opossums were losing weight at the daily rate of .3 3% of the initial winter weight, or a total weight loss of 40% during the 120 day winter period, is given in Table 12. Measurement periods were established Table 11. Opossum Number and Sex The metabolizable energy (M.E.) equivalent of weight loss while opossums were in the dens in winter. Most of the time was spent in sleep. The minimum M.E. expenditure rate for each animal was used in calculations (see Table 5). Mean Weight for Period kg Weight Decline for Period gs/day M.E. rate used in Calculations kcal/kg/hr M.E. kcal/g weight loss 17 2.68 28.3 2.36 5.35 Measurement Period Days Jan. 2 - 1 8 1 ? 2 9 Jan. 2 - 1 8 17 3.60 23.4 1.44 5.31 3 $ Jan. 2 - 1 3 12 3.17 37.5 2.03 4.11 4 $ Jan. 5 - 1 7 13 2.89 39.2 1.95 3.44 5 (f Jan. 2 - 1 4 13 3.14 34.6 1.94 4.22 6 Feb. 5 - 1 5 11 2.21 21.3 2.09 5.20 7 Feb. 8 - 1 5 8 3.26 49.5 1.92 3.03 13 2.99 33.4 1.96 4.38 Mean (n = 7) Table 12. Metabolizable energy (M.E.) expended and percentages of M.E. derived from food and loss of body substance, while opossums were losing weight at the approximate daily rate of .33% of the initial winter weight, equivalent to a total weight loss of 40% of the initial winter weight over the 120 day winter period. Mean Opossum Body Number Weight and Sex kg Measurement Period Days in Period Metabolizable Energy Expended Total from Loss kcal of Body From kcal/x leg f Period Substance Food kcal/day Body wt/day % % 2.09 Jan. 18 - Mar. 28 70 10,329 35.7 64.3 147 70.6 9 3.36 Jan. 4 63 8,167 57.2 42.8 130 38.6 3 9 2.54 Jan. 13 - Mar. 26 73 8,321 40.5 59.5 114 44.9 4 9 2.36 Jan. 18 - Mar. 3 46 6,700 28.7 71.3 146 61.9 5 Cf 2.66 Jan. 17 - Mar. 9 52 7,143 33.7 66.3 137 51.5 6 cf 2.31 Jan. 22 - Feb. 28 38 4,634 39.3 60.7 122 52.8 7 ■ M e o «0 N II % <>* % il II * *1 :| 3 < 8IH M nl Ml *1 aI™ I & * el % 1\ lOO* I %** * ***I'**® V»* I Figure 23. Regression of the lipid fraction (%) of body weight (Y) on days (X) of the winter period (December 1 = Day 1, March 29 = Day 120) of 54 opossums collected from January through March, 1968, in southern Michigan. See text for determination of "heavy" and "light" animals. FRACTION OF BODY WEIGHT , PERCENT 26 LIPID 1.0 2.4- 4* 6 .0 9 9 * 2 6 *2 6 HEAVY ANIMALS « 27*16 - 0*167 X A L L A N IM A LS ( N *6 4 ) Y « 2 4 * 6 6 - 0*1 4 6 X L IG H T A N IM A LS C N *32 ) T I 6 DEC. 26 7 22 JAN. M ID P O IN TS i 16 T I I 6 21 7 FES. MAR. DAY CLASSES I' I 22 26 134 Table 13. Body weight statistics of opossums collected January through March, 1968, in southern Michigan. Description Males Adults Juveniles Adults and Juveniles Females Adults Juveniles Adults and Juveniles Males and Females Adults Juveniles Adults and Juveniles Sample Size n 40 66 106 12 36 48 52 102 154 Mean Body Weight S.D. 3556.5 2230.7 2731.0 :456.7 547.8 824.6 2780 770 770 2350.0 1628.6 1808.9 429.7 456.8 550.6 2040 710 710 3278.1 2018.2 2443.6 620.9 590.8 859.7 2040 4620 3290 710 710 - 4620 Range(g) — — — — — - — — 4620 3290 4620 3240 3140 3240 Table If. Regression of body weight (g) on days of the winter period (Dec. 1 = Day 1; Mar. 29 = Day 120) of opossuas collected January through March, 1968 in southern Michigan. Description Saaple Sise n %0 66 106 Feaales Adults Juveniles Adults and Juveniles Males and Fenales Adults Juveniles Adults and Juveniles ? = %325.8 ? = 256%.5 Y = 3185.8 - 9.82*X f.50 X 6.28 X %316 2560 3179 3727 229% 2809 3137 202% 2%32 27.3 20.9 23.5 12 36 ’ 48 Y = 359%.3 - 13.1%*X Y = 2383.5 - 9.50 X Y = 2630.2 - 9.57 X 3581 237% 2621 2806 1813 2056 2017 12%3 1%82 %3.7 %7,6 %3.% %6 92 138 $ = %359.6 - 13.89*X Y = 2%61.5 - 5.93*X Y * 3012.% - 7.67*X %3%6 2%56 3005 3526 2106 2552 2693 1750 2092 38.0 28.7 30.% 1 Y = body weight (g) X = days of winter period * t test of the regression coefficient significant at 95% level of confidence. 135 Males Adults Juveniles Adults and Juveniles Regression Equation1 Values Calculated from the Regression Equation iody wt.tg) Body wt.(g) Swt. Body wt.(g) Mar. 29 or Dec. 1 or loss / Day 1 Day 120 120 days Day 60 136 were processed, of which 6 8 .8 % (106) were males (Table 13) and 31.2% (>48) were females (a ratio of 2.21 : 1.0), and 66.2% (102) were juveniles and 3 3.8% (52) were adults (ratio of 1.96 : 1.0). Juveniles are defined here as animals less than one year old. The higher sex ratio in favor of males may represent an actual preponderance of males, or a greater cruising radius for males, or both. The latter appears most likely. Sandidge (19 53) reported that 53.9% of 560 opossums in Kansas (of which 426 speci­ mens were pelts), were males; 58% of Hamilton's (1958) New York specimens were males. Fitch and Sandidge (1953) report a 1 : 1 sex ratio for 117 trapped specimens in Kansas. The greater fraction of juveniles in the sample, approximately in the ratio of 2 : 1 , is probably fairly representative of the population as a whole. Possibly the actual fraction of juveniles is higher because young individuals I have observed in the field have a smaller cruising radius than adults. Fitch and Sandidge (1953) found that the juvenile fraction ranged from 60.8% to 75% in three years of study in Kansas. The data in Table 14 and Figure 20 indicate that the percent weight loss for adults (27.3%) exceeds that for juveniles (20.9%). I believe that this indicated relation­ ship is erroneous - actually, the smallest juveniles apparently die before winter's end and hence the lightest cohorts are not represented in late winter samples. Thus, the weight loss of juveniles tends to be minimized. In this connection, the five smallest individuals were collected between January 19 and February 16. No specimens of this small size appeared in the collection after February 16. The body fat determinations of three of these small animals are as follows: January 21, 1074 g weight, 8.5% body fat January 31, 774 g weight, .5% body fat February 6 , 710 g weight, 11.8% body fat These fat percentages are below the level required for winter survival, judging from the comparative value of 17.7% (Table 16) for mid-winter for all opossums analyzed. Also, the high significance (95% level) for the regression of juvenile weight is misleading because in individual male and female regressions, the considerable deviations of class means for January 22 and February 6 are complementary (Figures 21 and 22), resulting in significance for the pooled juvenile regression. It appears that in general, the percent weight loss of females (Table 14, Figures 21 and 22) calculated from the regression equation is greater than that of males. This is to be expected since the calculated mid-winter mean weight of females is 753 g (Table 14) less than males, indicating highe metabolic requirements due to smaller size.. It is of considerable interest that the percent weight loss for females appears to exceed the 40% level (Table 14). Judging from observations on penned animals,' this is at, or near the limit of weight loss that can be sustained. Maximum recorded weight loss of penned animals was as follows: No. 1, 44. 7%; No. 3, 42.5%; No. 4, 41.5%; No. 5, 39.5% (this animal almost died); No. 7, 42% (this animal died of starvation). Following are data reported by other investigators on the winter weight loss of opossums and other mammals. Hamilton (195 8 ) in New York recorded the weights of 38 different individuals in the months of December through March. Weights in March were 31% lighter for males and 38% lighter for females. Fitch and Sandidge (195 3) in Kansas recorded the weights of the same individuals, trapped in the wild, October through March. An approximated mean weight loss (five individuals, interpreted from graph) for this period was 39%. One adult male, recaptured several times, had weights as follows: November 28, 1950, 4.54 kg; December 23, 5.00 kg; March 6 , 3.08 kg; June 18, 2.62 kg. This animal lost 48% of its December weight in six months. In comparison, Michigan raccoons show a winter weight loss of 25% for females and 30% for males (computed from the data of Stuewer, 1943). Hibernating ground squirrels lose about 30% of their weight (Kayser 19 62). Bailey and Davis (196 5) report that the average woodchuck loses about 39% of its mean prehibernation weight. These data suggest that the opossum may lose a large fraction of its weight in winter, apparently larger than many other mammals. r The lipid fraction of body weight of opossums collected January through March, 1968, in southern Michigan. Description Males S3ults Juveniles Adults and Juveniles Sample Size n Mean Body Weight (g) 11 3548 23 34 2111 Range of Body Weights (g) Lipid Fraction of Body Weight (%) 17.6 13.7 14.9 5.3 - 0.0 0.0 - 2576 2980 - 4170 770 - 3090 770 - 4170 2040 - 3240 710 - 2040 710 - 3240 20.6 12.8 6.2 — 31.0 22.0 15.6 3.2 3.2 — 31.0 2040 - 4170 710 - 3090 710 - 4170 18.8 13.4 15.2 Females Adults Juveniles Adults and Juveniles 20 2551 1492 1863 Males and Females Adults Juveniles Adults and Juveniles 18 36 54 3160 1888 2312 7 13 Range of the Lipid Fraction of Body Weights (%) - 5.3 — 0.0 0.0 - 25.4 28.2 28.2 31.0 28.2 31.0 gcY Table 15. Regression of the lipid fraction (%) of body weight on days of the winter period (Dec. 1 = Day 1; Mar. 29 = Day 120) of opossums collected fron January through March* 1968* in southern Michigan. Descripton Sample Sixe n 2 Regression Equation Values Calculated from % Lipid 1 Lipid Begin Winter Middle Winter Day 60 Day 1 the Regression Equation \ Lipid t Decline in Initial End Winter Lipid Fraction for Day 120 120 Day Winter Period Heavy opossums1 22 ? = 26.25 - .097 X 26.1 20.4 14.6 44.1 Light opossums1 32 ? = 24.66 - .149 X 24.5 15.7 6.8 72.3 All opossums 54 ? = 27.13 - 1.57 X 27.0 17.7 8.3 69.3 To divide "heavy" from "light" animals, the regression equation Y = 3012.4 - 7.67 X (Y = wt. in g, X = days of winter period) for all 154 animals for which weight was determined* was used. "Heavy animals” were above the line, "light" ones were below it. * Y s Lipid fraction of body freight in t. X * days of the winter period. Ohl Table 16. 141 Statistics and regressions for winter decline in body fat are given in Tables 15 and 16 and Figure 23. The total sample size analyzed for body lipid content was 54 indi­ viduals, a number too small to be subdivided into age and sex groups for calculating regressions. Rather, lipid regressions were calculated for two sub-groups with weights falling above the weight regression for all animals (Table 14 Y = 3012.4 - 7.67 X*) - the "heavy opossums", and those falling below the regression - the "light opossums". results are given in Table 16 and Figure 23. These The regressions for.percent fat are significant at the 1 0 % level. The mean lipid fraction of 18.8% (Table 15) for adult animals is higher than the corresponding value of 13.4% for juveniles. This is to be expected as the mean body weight of juveniles is 1.27 kg less than adults and hence have a higher metabolic rate per unit body weight. Likewise, the percent decline of body fat in light animals (Table 16) is greater than heavy animals by 28.2%. The fat lost per unit body weight lost can be calculated from the lipid regression equation for all animals analyzed (Table 16), and their body weight regression (Y = 2853.4 6.9X* where Y = weight and X = days). fat lost/g body weight lost. This valueis 0.73 g Thus, using a combustion value of 9.5 kcal/g fat, the energy provided by weight loss would be 6.9 kcal/g weight lost. Forheavy animals, the equivalent value is 0.53 g fat/g body weight lost(using the regression for weight of Y = 39 86.1 - 10.IX*, and the -im corresponding equation for lipid loss in Table 16. At 9.S kcal/g fat, 1.0 g body weight supplies 5.03 kcal, which does not differ greatly from 4.4 kcal/g weight loss calculated by food balance for experimental animals. CHAPTER X ECOLOGICAL SYNTHESIS: MODELS OF WINTER ENERGETICS AND SURVIVAL Several important features regarding the opossum's winter energy economy emerge from the data and observations presented in previous chapters. 1. These are: Labile thermoregulation and low body tempera­ tures contribute to a low resting metabolic rate, among the lowest for non-hibernating mammals. Consequently, energy expenditure of sleeping opossums is low. 2. Denning behavior and the thermal characteris­ tics of the ground burrow are such that the opossuiris resting metabolic rate does not exceed the basal metabolic level of wellinsulated species. 3. A great proportion of time is spent in winter sleep, suggesting considerable energy savings. 4. Depot fat appears to play a major role as an energy reserve. However, the fat depot supplies only enough energy for a maximum survival length approximately one-half of the winter period (unlike a true hibernator such as the woodchuck). 143 5. Generally, foraging is restricted to nights with temperatures around 0°C or above. Foraging is also restricted by deep snow. It is a fact that the opossum survives the temperate winter and is eminently successful alongside its specialized and well insulated competitors. The data and observations presented here suggest that the opossum compensates for its lack of predatory agility and inferior insulation by means of behavioral thermoregulation and low energy require­ ments. However, the opossum appears to have an ecological Achilles Heel in the form of two major limiting factors; Cl) its physiological inability to forage at sub-freezing temperatures, and (2 ) the limitations imposed on foraging by deep snow. Both an excess of winter days with sub­ freezing temperatures and deep snow would tend to limit energy intake. Both factors are a function of locality and latitude and in Michigan they tend to increase with increasing latitude. The objective of the mathematical models which follow is to quantify the limitations imposed by these factors on the opossum's winter energy budget and hence on its survival and distribution. Let us assume that a hypothetical winter period of X days consists of various combinations of days spent in sleep (days when the maximum temperature does not exceed 0 °C) and days on which animals forage (days when the maximum temperature exceeds 0°C). An energy parameter most critical to the opossum's winter energy balance* is the minimum 145 mandatory energy intake on those days available for foraging. This energy deficit per foraging day would tend to increase as fewer days for foraging are available in the winter period. The energy deficit per foraging day may be calcu­ lated for various winter combinations of sleeping and forag­ ing days using a "representative opossum" of mean weight' and characteristics. We may formulate as follows: Key to symbols: Dp = Foraging days in the winter period Dg = Sleeping days in the winter period = Foraging days with complete lack of foraging success (ie., no energy intake) Eg = Rate of energy expenditure per sleeping day Ep = Rate of energy expenditure per foraging day E^ = Energy provided per unit body weight lost Et = Total winter energy expenditure e TF= Total energy to be provided by foraging Ed = Energy deficit/foraging .day (All energy values are kcal of metabolizable energy) W = Maximum weight loss K = Constant 146 Assuming that Dg + Dp = K and total energy expenditure (E^) for any given combination of foraging and sleeping days is (1) Et = Ds Eg + Dp E f and the total energy provided by maximum weight loss is W EW and the total energy which has to be provided by foraging p) is (2) ETF = ES ES ^ EF eF ” ^ EW eTF = 16,710 — 3630 kcal E^.p = 1 1 , 0 80 kcal :151 Energy deficit per foraging day (Ej>) is (3) Ej^ - Dg Eg + Dp Ep — W Ey DF ED E d = 158,3 kcal/day Energy deficit per foraging day (ED > with a complete lack of foraging success for 15 days (D^; ie., no energy intake for 15 days of foraging) is (4) Ed = Dg Eg + Dp Ep - W Ew df - »L En = 11,080 D 70-15 Ep = 201.4 kcal/day Results and Inferences Energy relationships were calculated for Models 1, 2 and 3 on the basis of assumptions and equations given above and in the Appendix. These values are presented in Tables 17 and 18, and Figures 24 and 25. Mean species energy relationships (Model 1) are visually represented in Figure 24. It will be noted that the energy deficit increases ever more sharply in relation to a decreasing number of foraging days, or an increasing number of days spent in sleep. Assuming that an opossum is capable of garnering the maximum of 315 kcal on every foraging day Table 17. Total energy expenditure for the winter period (120 days) and energy deficit per foraging day for various coaibinations of foraging days and days spent in sleep; for Models 1, 2 and 3. Available foraging days in the winter period (120 days)* and days spent in winter sleep* in parentheses. Model 2 Adults S =3.30 kg 100 (20) 90 (30) 80 (W0) 70 (50) 60 (60) SO (70) W0 (80) 30 (90) 20 (100) 18*960 18*510 18*060 17,610 17,160 16,710 16*260 15,810 15,360 1W.910 1W.W60 111 117 12W 133 1WW 158 177 2QW 2W3 309 WW1 21*290 20,7*40 20,190 19,6W0 19,090 18,SW0 17,990 17,WW0 16,890 16,3W0 123 130 138 1W9 162 180 205 2W2 17,960 17*WW0 16*920 16,WOO 15*880 15,360 1W*8W0 Total 21.8W0 (kcal/ 120 days) befieit 117 (kcal/ foraging day) Total (kcal/ 18*W80 120 davs) Deficit = 2.05 kg (kcal/ 11W foraging day) Model 3 Juveniles S 110 (10) 152 Total Model 1 (kcal/ All Aniaals Deficit S = 2.WO kg Iccal/ foraging day) 120 (0) 120 126 135 1W5 158 176 201 305 1*4,320 13*800 238 300 W29 13*280 W2W Table 18. Effect of lack of foraging success on the energy deficit (kcal) per foraging day for various combinations of foraging days, and days spent in winter sleep (120 day winter period). Calculations are for the whole population (ie. adults and juveniles, Model 1) Number of Days with Complete Lack of Foraging Success 30 (90) 0 (Complete success) 111 117 124 133 144 158 177 204 243 309 5 116 123 131 141 154 170 193 226 278 371 10 121 129 138 150 165 185 213 254 324 464 15 127 135 146 160 177 201 236 291 389 619 20 133 143 155 171 192 222 266 339 486 928 25 140 151 166 184 210 246 304 407 649 1860 30 148 161 177 200 231 277 354 509 973 9280 153 120 (0 ) Available foraging days in the winter period (120 days) and days spent in winter sleep in parentheses____ 110 100 80 70 90 50 40 60 (40) (50) (30) (60) (70) (80) (1 0 ) (2 0 ) 15& | Figure 24. Model 1. Relationships of winter energy deficit! per foraging day (ie. , minimum mandatory energy intake per foraging day) to the number of ! available enforced sleeping days and foraging j days in the 120 day winter period. Calculations! are based on a theoretical opossum representing | the species as a whole, with mean species :| characteristics. Solid curves indicate the effect of a complete lack of foraging success | for given numbers of d a ys , on the energy j deficit per foraging day. The maximum energy ! deficit per foraging day which can be sustained i (315 kcal) is represented by the horizontal ;> dashed line. Projected intersections of the energy deficit curves (solid lines) with the i maximum energy deficit line (horizontal dashed line) indicate the maximum number of enforced | winter sleep days which can be survived (see text). | J H : % r P SO ES SO IS 10 8 DAYS DAYS DAYS DAYS DA DA KNCMV DEFICIT COMPUTE LACK OF FORAOINO SUCCESS FOR 155 < KCAL) / FOAAOIHO OAV FORAOINO DAYS / ISO DAY NO * FOAAOINO SUCCESS t l ♦ —LIMIT ISO DAY 156 1 ■I ■ ■i I Figure 25. Model 2 (Adults) and Model 3 (Juveniles). | Relationships of winter energy deficit per 1 foraging day (ie., minimum mandatory energy j intake per foraging day) to the number of I available enforced sleeping days and foraging days in the winter period (heavy solid curves). Calculations for each model are based on a theoretical opossum with mean characteristics of either adults (Model 2) or juveniles (Model 3) The maximum energy deficit per foraging day which can be sustained by adults and juveniles (U30 and 270 kcal respectively) is represented by the horizontal dashed lines. Projected intersections of the two energy deficit curves % (solid.lines) with maximum energy deficit lines (horizontal dashed lines) indicate the maximum number of enforced winter sleep days which can be survived (see text). The broken, slanting line represents the total energy expenditure for the 120 day winter period for Model 1 in relation to the number of sleeping and foraging days. 100 FORAOINO DAYS / SO SO 120 DAY WINTER PERIOD 70 SO 90 40 PERIOD 10*000 WINTER +M “AX- lt(000 17,000 H CJi 'J ENEROY EXPENDED (KCAL) / 1*0 DAY ADULTS 10,000 15,000 14*000 13,000 100 WINTER SLEEP DAYS / 120 DAY WINTER PERIOD J j4 U M IT -» l0 0 ADULTS 16*6 available, it could theoretically survive a winter with 91 enforced days of sleep and only 29 available foraging days. It is assumed, of course, that all body fat reserves are depleted. A lack of foraging success decreases the theoreti­ cal maximum number of enforced days spent in winter sleep (a function of winter severity) which can be survived (Figure 24). Each additional unit of five days of complete lack of foraging success will decrease by six the number of enforced winter sleep days which can be survived, or increase by six the number of necessary foraging days (Figure 24). No more than 55 enforced winter sleep days can be tolerated assuming a complete lack of foraging success for 30 days or some partial equivalent of the latter. Energy deficit curves for Model 2 (adults) and Model 3 (juveniles) and total energy expenditure for Model 1 (mean species representative) are given in Figure 25. As might be expected, the total energy expenditure declines with increasing number of days spent in winter sleep. It is surprising that the energy deficit per foraging night is practically identical for juveniles and adults, (Table 17, Figure 25). There are two apparent reasons for this. Firstly, the greater individual mass of adults and conse­ quently their greater energy expense, tends to be compensated by the higher metabolic rate of juveniles. It will be noted that the total energy expenditure of juveniles and adults (Table 17) differs only by a little over 3000 kcal for any combination of winter conditions. As I have observed previously (Chapter VIII), experimental opossums on a maintenance diet similarly had identical or nearly identical food requirements, regardless of body size. Secondly, the difference of approximately 3000 kcal of total energy expenditure between adults and juveniles is minimized farther by the additional 2 950 kcal of energy in the fat depot of adults (7750 kcal) compared to juveniles (*+800 kcal). Could it be that this equivalence of energy deficit is true for other wild species? The near identical energy deficit per foraging day for juveniles and adults does not mean that their chances for winter survival are similar. The absolute theoretical limit of winter severity which can be survived is 86 days of enforced sleep for juveniles versus 100 days for adults (Figure 25). This theoretical difference is simply due to the greater daily intake capacity of adults compared to juveniles. One may also infer that the predatory success of adults is greater compared to that of juveniles. A large opossum (I have had them as large as 7.2 kg) can overpower most potential prey animals whereas a small opossum might conceivably fail to kill than itself. a rabbit as large or larger In the final analysis,long-term population survival depends on juvenile survival. The age ratio in the road-kill collection was 2 : 1 in favor of juveniles. It stands to reason that the winter energy relations of juveniles considered here are most relevant to population survival and species distribution. CHAPTER XI DISTRIBUTION Elton (1927) wrote, "...Animals are in practice, limited in their distribution by their habits and reactions, the latter being so adjusted that they choose places to live which are suitable to their particular physiological requirements and their breeding habits." In previous chapters, I have attempted to show that energy limitations on the opossum's winter survival are a function of its physiology and ecology. In this section, data will be presented supporting the hypothesis that the northern distributional limit of the opossum is determined by the prevailing number of sub-freezing winter days and snow depth limiting winter energy intake, and hence limiting survival. Theoretically and conservatively, energy limitations illustrated in Models 1, 2 and 3 (Chapter X) suggest that opossums cannot survive an excessive number of sub-freezing days (with enforced winter sleep), such that the energy deficit per foraging day exceeds the daily intake capacity of the individual. One may infer that the actual maximum limit of "winter severity" which can be survived is some lesser number of sub-freezing days and greater number of foraging days than the theoretical maximum. 160 This appear* 161 to be the case, judging from distributional data. Distribu­ tion and relative density of opossums in Michigan are related to the number of sub-freezing winter days (enforced winter sleep days) in Figures 26 and 27. The opossum is rare to absent in areas with 70 or more winter days in which o the maximum temperature does not exceed 0 C. (Note, the isotherms have been plotted on the basis of the annual number of days with daily maximum temperatures of 0°C or below. About 90% of these days occur during December through March, the period used as a basis for calculations. Therefore, the approximate observed limits of tolerance as interpreted from the map in Figure 26 and 27 would be slightly lower than indicated). Conservatively, the observed limit of tolerance appears to be approximately 70 o winter days with a maximum daily temperature of 0 C or below, equivalent to 70 winter sleep days and SO foraging days. The calculated total energy expenditure for 70 sleeping days and 50 foraging days for the "average opossum" (Model 1, Table 17 and Figure 24, Chapter X) is 15,810 kcal and the energy deficit per foraging day is 204 kcal, equivalent to approximately 18 days of complete foraging failure. It is interesting to note that the greatest density of opossums in southern Michigan (Taube 1942) appears to follow and occur south of the "50 day" isotherm (Figure 27). Figure 26. Locations of road-killed opossums collected in Michigan from January through March, 1968, related to the number of days in winter with mean daily maximum temperatures equal to or 1.less than 0°C (numbers and heavy solid lines). Collecting effort was distributed throughout the state (collecting was done by the Michigan Department of Natural Resources). Climatalogical data are from the Weather Bureau, U. S Department of Commerce. 16 3 •2 100 •s 02 70 00 Nft WINTER DAYS * MAX. TEMP. * 0 * C 00 40. 00 SO 40 4S 40 Opossum distribution and density in Michigan related to the number of days in winter with mean daily maximum temperatures equal to or less than 0 C (numbers and heavy solid lines). The relative abundance index represents opossums trapped per square mile from 1937 to 1941 (Taube, 1942). The heavy broken line represents the approximate northern distribution boundary, based on questionnaires. The letters represent cities, as follows: A Allegan, B Benton Harbor, C Cadillac, D Detroit, E Escanaba, G Gaylord, H Houghton Lake, J Jackson, L Lansing, M Marquette. Climatalogical data are from the Weather Bureau, U. S. Department of Commerce. RARE TO AMENT LOCALLY COMMON TO RARE ARUNOANT TO COMMON RELATIVE AtUNDANCE INDEX SB 166 Comparison of distribution with winter snow in the state (Figure 28) shows that the observed limit of distribu­ tion also correlates approximately with snow depth. It is difficult to isolate the two climatic factors considered here. They complement each other in their negative effects and their combined negative effects increase northward. However, the apparent importance of snow depth is indicated by a population density comparison at the "60 day" isotherm _ V near Lansing (Figures 27 and 28), and at the "60 day" isotherm farther north (Figures 26, 27 and 28). In the area of Lan-ing with relatively less snowfall, opossums are common. The population density drops off sharply at the more northerly locations where the mean -Iriumber of winter days with accumulated snow depth on the ground of 28 cm or more exceeds 30 days (Figures 27 and 28). Shelford (1911) noted, "The geographic range of any species is limited by the fluctuation of a single factor (or factors) beyond the limit tolerated by that species." In this vein, the main thesis proposed here is that "winter jeverity" is relevant to population survival only insofar as winter energy intake is limited by two factors: (1 ) the number of available foraging days is determined by the number of days in the winter period with maximum temperatures above 0°C, and (2) foraging success of the species is negatively affected by deep snow. If indeed these factors are important, it would follow that population declines are associated with Figure 2 8 . Mean number of winter days with accumulated snow depth on the ground of 28 cm or more. The heavy broken line represents the approximate northern distributional limit of the opossum. Snow map adapted from A. H. Eichmeier, 1964, Michigan Weather Service and Weather Bureau, U. S. Department of Commerce. 168 severe winters. A well-documented population decline occurred in Allegan County, southwestern Michigan, following the severe winter of 1939 - 1940 (Stuewer, 1943). In 1939, 127 opossums were trapped in the wildlife refuge. In 1940, 24 opossums were trapped, with only five recaptures from the previous year. Allegan County has 41 days (mean) in December through March in which the maximum temperature does not exceed 0°C. There were 12 fewer potential foraging days in the winter of 1939 - 1940 in Allegan County (inter­ preted from U. S. Weather Bureau records). Thus, 67 foraging days were theoretically available which is quite adequate under normal conditions of snow (Figure 27). Assuming a maximum energy intake of 315 kcal/foraging night (Model 1, Figure 24, Chapter X), more than 30 days with a complete lack of foraging success, or some partial equivalent of i t , could have been sustained. Apparently, excessive snow depth adversely affected foraging to an equivalent of more than 30 days of complete foraging failure. Stuewer (1943) reports "two or more feet" of snow on the ground in January and March. Allegan County had a mean unmelted snowfall of 30 cm for the four winter months of 19 39 - 1940 (U. S. Weather Bureau). The average snowfall for Michigan for 1940 was the greatest on record, amounting to 210 cm compared to the normal 143.2 cm. Metzger (1955) ascribes the decline of a local opossum population in Ohio to a severe winter. Peterson and Downing (1956) suspect that population fluctua­ tions in southern Ontario have been caused by severe winters. * From my own observations, a significant population decline occurred at Kensington Park in south-eastern Michigan (a natural area) following the winter of 1961 - 1962. Apparently, both a lack of available foraging days and deep snow caused the decline. This winter had the lowest mean temperature since 1936, and the greatest snowfall (mean of 16 cm) in January, February and March since 1943. There were 67 days when the maximum temperature did not exceed freezing, five more days than the mean value. Taube (1942) reports declines in opossum trapping success in Michigan following "hard winters." Of particular interest is an apparent population decline of major proportions which occurred around 1850 (Taube 1942). Early sighting records compiled by Taube show that this species occurred in the southern-most portion of Michigan from 1822 to 1850. He found no records of opossums occurring in the state between 1850 and 1897. An exceptionally severe winter was reported to have occurred around the middle of the century (cannot be verified by Weather Bureau). From data compiled by a Wisconsin conserva­ tion department official (Taube 1942), the opossum was scarce in Wisconsin during approximately the same period. A synchronous population decline of such wide magnitude implicates climatic factors. 171 The question arises whether there are any other important factors limiting northern distribution in Michigan. I believe this is quite unlikely. The opossum occupies an extremely diversified ecological niche, occurring in most terrestrial habitats within its range. Today, the opossum is found from the tropics north through the temperate region of eastern North America and has been locally successful in the West. Climatic factors are most likely to affect the distribution of a species with such non­ specialized ecological requirements. This inference is supported by comparisons between ecologically similar areas in Michigan which differ greatly in degree of opossum abundance. For example, the oak forest association occurring on the sandy soils of Allegan County in southwestern Michigan (see city of Allegan, Figure 27) harbors a moderate opossum population. In the similar oak forests of Roscommon County (see Houghton Lake, Figure 27) and other similar areas to the north, the species is rare to absent. Opossums are common in the mixed hardwood forest, shrubland and alder thickets on the podzol soils of Gratiot County in the southern part of the state. They are absent from practically identical ecological associations occurring frequently in northern Michigan. Opossums are common in the beech-maple region of south-central Michigan (see Lansing, Figure 27). They are rare to absent from the beech-maple forests of Wexford County (see Cadillac ., Figure 27). Another tempting iH (if improper) comparisonj the raccoon occurs throughout the state (common in the Lower Peninsula, rare in the Upper Peninsula according to^Stuewer 19H3 and Burt 195*0 compared to the absence of the opossum from roughly onehalf of the state. Both animals are omnivorous and both spend much of the winter period in sleep. The most sensitive indicators of climate are ecological. It is therefore useful to compare opossum distribution with regional vegetational and faunal associations. In Michigan, the present northern distributional limit of the opossum (Figure 29) coincides approximately with the southern edge of the pine-hemlock ecotonal region (or lake forest, which also includes yellow birch, sugar maple and beech as dominants) between the boreal coniferous forest region and the deciduous forest region (Nichols 1935, Shelford and Olsen 19 35, Shelford 1963). This region has climatic, edaphic, vegetational and faunal affinities with the boreal coniferous forest. As a tentative hypothesis, I propose that the southern edge of the pine-hemlock ecotone approxi­ mates the northern potential distributional limit for the opossum in eastern North America. North of this boundary, winter energy availability is limited by the climatic factors which have been considered in this study and survival is therefore impossible. As of this writing, distributional data appear to support this hypothesis (Figure 29)i in a general way, the distribution of the opossum has not extended north of this line. It is conceivable that local populations Figure 29. The present approximate northern distributional limit of the opossum in the northeastern United States and adjacent Canada in relation to the southern edge of the pine-hemlock (lake forest) ecotone region (Nichols 19 35, Shelford 1963). The northern distribution of the opossum is based on the following sources: Wisconsin, Jackson (1961), Long and Copes (1968); Michigan, present study; Ontario, Peterson and Downing (1956), Peterson (1966); New York, Hamilton (1958); Vermont, New Hampshire and Massachusetts. Waters and Rivard (19 62). APPROXIMATE SOUTHERN LIMIT OF PINE* HEMLOCK ECOTONE OR LAKE FOREST j I Q OPOSSUM DISTRIBUTION 175 may become established to the north of the theoretical boundary wherever man has created an exceptional source of available energy, such .as a garbage dump. However, it is doubtful that such marginal populations will long survive the vagaries of climate. The validity of the latter hypothesis remains to be tested. For the Lower Peninsula of Michigan, there is good evidence that the opossum has marched as far north as it cam go, as its range has not changed substantially for the last 25 years (Taube 1942 and present study). In Wisconsin, the opossum may extend its range northward (Figure 29). There have been recent records for northern Wisconsin (Long and Copes, 1968) and along the southern political boundary of Michigan's Upper Peninsula (Ozoga and Gaertner 1963, Rafferty, personal communication 1967). Perhaps the opossum will extend its range northward in the New England states. The ecotonal boundary has less predictive value there as the ecology of the region is far more complex than illustrated in Figure 29. The northward spread of the opossum appears to be a phenomenon of historic times (Guilday 1958). Opossum remains are associated with archeological sites to the south of Michigan along the Ohio River and its tributary and Miami River, the southern shore of Lake Erie and south of the Hudson River valley along the Atlantic coast (Guilday 1958). Apparently it occurred at the edge of the prairie, south of Chicago (Liette, 1702). Early sightings of the 176 opossum in Michigan (summarized by Taube 1942) followed the white man's pioneering activities. Corresponding north­ ward population spread.occurred throughout the East and Mid-west (Seton 1909, Goodwin 1935, Peterson and Downing 1956, Hamilton 19 33, 1958, and others). One may infer that agricultural practices and associated early serai stages provided favorable conditions for popu­ lation expansion. In Michigan, the white man vastly increased (artificially) the prairie and forest edge habitats from the few, small, natural prairie remnants present in the original climax forest of the southwestern portion of the state (Veatch 1927). Pioneering caused a general 4 increase in forest edge species (Seton 1909), such as the fox squirrel (Allen 1943) while the climax forest association declined. In the case of the opossum, this species is not a member of the forest-edge faunal association by virtue of any particular specialization or requirements. Man's pioneering activities may have caused population expansion by benefitting the species winter energy economy. Farm crops provide concentrated, easily available energy which may be important under survival conditions. Also, various foods are more available at ground level (where the opossum forages) in early serai stages than they are in the climax forest. There is another possibility suggested by the findings of this study. Perhaps man has been the agent for the opossum's success by inducing an increase in the abundance 177 and variety of burrowing mammals characteristic of the prairie and forest edge. It has been noted previously that the opossum is obliged to use the burrows of other mammals. Survival in the temperate winter depends on the ground microclimate, judging from the data presented. The increase in burrows may have provided minimal requirements for popula­ tion expansion. Today, the principle terrestrial burrowing mammals ih Michigan occurring throughout the state are the woodchuck, red fox, striped skunk and badger. Seton (1909) presents evidence that the red fox was not present in the primeval temperate deciduous forest. The striped skunk is an animal of the forest edge (Seton 1909 , Burt 1954) and apparently increased following settlement (Seton 1909). The badger frequents the prairie and forest edge (Seton 1909, Burt 1954) and probably also increased. Today, the badger is uncommon in Michigan, although it is distributed throughout the state. legacy in dens. It does, however, leave a significant Seton (1909) expresses the opinion that the woodchuck increased coincident with lumbering and farm­ ing. It is interesting to note that the opossum did occur along the edge of the prairie, south of Chicago before settlement began (Liette 1702). Other (typical) species present in the primeval forest climax community either did not produce ground dens (gray fox) or could hardly have produced a significant number of ground dens (gray wolf). One may infer that ground dens increased after settlement permitting population expansion of the opossum. If indeed 178 this was the case, the opossum has invaded much of the temperate environment on the coattails of the burrowing mammals. CHAPTER XII CONCLUSIONS By most measures, the didelphids are a successful group. They have survived with little morphological change for 70 million years or more (Simpson 1935) while many specialized marsupials and placentals have faded into extinction. With the re-establishment of the Pleistocene land bridge, the opossum was one of the few South American species which successfully invaded the domain of North American placentals. The opossum's success has been attributed to its structural generalization and non­ specialized behavior. Its physiological-behavioral response to winter outlined here, appears to be but another facet of its generalized nature. Winter survival for mammals is, in large measure, a problem of energy economy. Placental mammals have evolved mechanisms to cope with the temperate winter such as effective body insulation, huddling behavior, use of insulated shelters, food storage, hibernation and other means. From an energetics standpoint, the opossum's labile thermoregulation and low metabolic rate, coupled with its use of ground dens in winter is a notable and unusual response among mammals affecting winter energy 179 180 economy. If the opossums physiological-behavioral response is as successful as it appears to be, why is it so poorly represented among northern mammals? answer has ecological roots. I believe that the The general trend in mammalian evolution has been one of specialization. Through specialization, utilization of energy in the eco­ system has been more efficient. It appears that the price for specialization has been a higher level of metabolism. Yet it seems that unexploited portions of many other niches do exist, particularly in the disturbed communities asso­ ciated with man, which provide enough energy to sustain a species specializing in generalization. the adaptable opossum. That species is In short, perhaps the opossum's physiological-behavioral response to winter is a unique one because potential generalized niches to accommodate a low-key mode of life are restricted. BIBLIOGRAPHY BIBLIOGRAPHY ABLES, E. D. 19.69. Activity studies of red foxes in southern Wisconsin. J. Wildl. Mgt. 33(1): 145-153. ALLEN, D. L. and W. W. Shapton. 1942. An ecological study of winter dens with special reference to the eastern skunk.Ecology 23: 59-68. ALLEN, D. L. 1943. Michigan fox squirrel management. Game Div., Dept, of Conservation, Lansing, Mich, 404p. BAILEY, E. D. and D. E. Davis. 19 65. The utilization of body fat during hibernation in woodchucks. Canad. J. Zool. 43: 701-707. BARNES, R. H . , G. Fiala, B. McGehee and A. Brown. 1957. Prevention of coprophagy in the rat. J. Nutr. 63: 489-498. BARNES, R. H. and G. Fiala. 1959. Effects of the prevention of coprophagy in the rat. VI Vitamin K. J. Nutr. 6 8 : 603-614. BARTHOLOMEW, G. A. 1956. Temperature regulation in the macropod marsupial Setonix brachyurus. Physiol. Zool. 29(1): 26-40. BARTHOLOMEW, G. A. and J. W. Hudson. 19 62. Hibernation, estivation, temperature regulation, evaporative water loss and heart rate of the Pygmy Possum, Cercaertus nanus. Physiol. Zool. 35: 94-107. BEHNKE, A. R . , B. G. Feen, and W. C. Welham. 1942. The specific gravity of healthy men. J. Am. Med. Assoc. 118(7): 495-501. BERNER, A. and L. W. Gysel. 19 67. Raccoon use of large tree cavities and ground burrows. J. Wildl. Mgt. 31(4): 706-714. BRITTON, S. W. and W. E. Atkinson. 19 38. in the sloth. J. Mammal. 19: 94-99. 181 Poikilothermism 182 BRODY, S. 19*15. Bioenergetics and growth. Originally Reinhold Publ. C o . , reprinted 1964 by Hafner Publ. Co., N. Y. 1023p• BROWN, A. E. 1909. The tuberculin test in monkeys with notes on the temperature of mammals. Proc. Lond. Zool. Soc. p. 81. BROWN, C. J., J. C. Hillier and J. A.Whatley. 1951. Specific gravity as a measure of the fat content of the pork carcass. J. Anim^ Sci. 10: 97-10 3. BROWN, L. E. 1956. Field experiments on the activity of the small mammals Apodemus, Clethrionomys, and Microtus. Lond. Zool. Soc. Proc. , T2o(*+Ti £49-564. BURT, W. H. 1954. The mammals of Michigan. Press, Ann Arbor, Mich., 288p. Univ. of Mich. CALHOUN, J. B. 1945. Diel activity ry thins of the rodents Microtus ochrogaster and Sigmodon hispidus hispidus. EcoTogy “2 6 'C?>: 751-77 3 . --------- --------- ----CORBIN, J. E. 1968. Personal communication regarding the energy content of Purina Dog Chow. Ralston Purina Co., St. Louis, Mo. DAFT, F. S., E. G. McDaniel, L. G. Herman, M. K. Romine and J. R. Hegner. 1963. The role of coprophagy in the utilization of B vitamins synthesized by intestinal, bacteria. Fed. Proc. 22: 129-133. DEARBORN, N. 19 32. Foods of some predatory fur-bearing animals in Michigan. Bull. No. 1, Univ. of Mich. Sch. of Forestry and Conservation, Ann Arbor, Univ. of Mich. Press. DA COSTA, E. and R. Clayton. 1950. Studies of dietary restriction and rehabilitation I I. Interrelationships among the fat, water content and specific gravity of the total carcass of the albino rat. J. Nutr. 41: 597-606. DAVIS, D. E. 19 6 6 . The role of environmental factors in hibernation of woodchucks (Marmota monax). Ecology 48(4): 683-689. EICHMEIER, A. H. 1964. Snow depths in Michigan. Michigan Weather Service and Weather Bureau, U . S . Dept. of Commerce, 95p. 183 EISENTRAUT, M. I960. Heat regulation in primitive mammals and in tropical species. Bull. M u s . Comp. Zool. , Harvard, 124: 31-43. ELTON, C. 1927. Animal Ecology.- Sidgwick and Jackson Ltd., London. 209p. ENDERS, R. K. and D. E. Davis. 1936. Body temperatures of some Central American mammals. J. Mammal. 17: 165-166. ENGELMANN, M. D. 1961. The role of soil arthropods in the energetics of an old-field community. Ecol. Monogr. 31: 221-228. FIDANZA, F . , A. Keys and J. T. Anderson, 1953. The density of body fat in man and other mammals. J. Appl. Physiol. 6 : 252-256. FITCH, H. S. and L. L. Sandidge. 1953. Ecology of the^ opossum on a natural area in northeast Kansas. Univ. Kans. Publ., Mus. Nat. Hist. 7C2): 305-338. FORMOZOV, A. N. 1946. Materials for fauna and flora of the U.S.S.R. New Series Zoology 5 (XX): 1 - 152. English Edition, Occasional Paper No. 1, Boreal Inst., Univ. of Alberta, Edmonton, Alberta, Can. FRANCQ, E. N. 1967. Feigned death in the opossum Didelphis marsupia?is . Unpublished Ph.D. Thesis, Penn State Univ., 57p.. FRIDERICIA, L. S., P. Freudenthal, S. Gudjonnsson, G. Johansen and N. Schoubye. 19 27. Refection, a transmissible change in the intestinal content enabling rats to grow and thrive without Vitamin B in the food. J. Hyg. 27: 70-102. GEIGER, R. 1966. The climate near the ground. Press, Cambridge, Mass., 611p. Harvard Univ. GERSTELL, R. 19 39. The value of groundhog holes as winter retreats for rabbits. Penn. Game News, Sept.: 6-9. GNAEDINGER, R. H . , E. P. Reineke, A. M. Pearson, W . D . Van Huss, J. A. Wessel, H. J. Montoye. 1963. Determination of body density by air displacement, helium dilation, and underwater weighing. Ann. N. Y. Acad. Sci. 110: 96-108. GOLLEY, F. B. 1960. Energy dynamics of a food chain of an old-field community. Ecol. Monogr. 30: 187-206. 184 GOLLEY, F. B. and J. D. Gentry. 1964. Bioenergetics of the southern harvest ant, Pogonomyrmex badius. Ecology 45: 217-225. GOLLEY, F. B. , G. A. Petrides, E. L. Rauber, J. H. Jenkins. 1965. Food intake and assimilation by bobcats under laboratory conditions. J. Wildl. Mgt. 29(3): 442-447. GOODWIN, G. G. 1935. The mammals of Connecticut. Geol. and Nat. Hist. Survey Bull. 53: 1-221. State GORDON, M. S., G. A. Bartholomew, A. D. Grinnell, C. B. Jorgensen and F. N. White. 196 8 . Animal function: Principles and adaptations. MacMillan Co., N. Y . , 560p. GUILDAY, J. E. 1958. The prehistoric distribution of the opossum. J. Mammal. 39: 39-43. HAMILTON, W. J., Jr. 1933. The northward spread of the opossum in New York. J. Mammal. 14: 151-152. HAMILTON, W. J . ,Jr. 1943. The mammals of eastern United States; an account of recent land mammals occurring east of the Mississippi. Comstock Publ. Co., Ithaca, N. Y . , 432p. HAMILTON, W. J., Jr. 1958. Life history and economic relations of the opossum (Didelphis marsupialis . . virginiana) in New York State. Cornell Univ. Agr. Exp. Sta. Mem. 354: 1-48. HART, J. S. 1957. Climatic and temperature induced changes in the energetics of homiotherms. Rev. Can. Biol. 16: 133-174. HATFIELD, D. M. 1935. A natural history study of Microtus californicus. J. Mammal. 16(4): 261-271. HATFIELD, D. M. 1940. Activity and food consumption in Microtus and Peromvscus. J . Mammal. 21(1): 29-36. HAYWARD, J. S. 1965. Microclimate temperature and its adaptive significance in six geographic races of Peromvscus. Can. J. Zool. 43: 341-350. HEMINGWAY, A. 1963. Shivering. Physiol. Rev. 43: 397-422. HIGGINBOTHAM, A. C. and W. E. Koon. 19 55. Temperature regulation in the Virginia opossum. Am. J. Physiol. 181: 69-71. HIX, V. M . , A. M. Pearson, E. P. Reineke, T. A. Gillett and L. J. Giacolletto. 1967. Determination of specific gravity of live hogs by suppressed zero techniques. J. Anim. Sci. 26(1)-: 50-57. HODGEMAN, C. D . , R. C . W e a s t and S. M. Selby. 1964. Handbook of chemistry and physics. 45th Edition, Chemical Rubber Co., Cleveland, Ohio. HUDSON, J. W. 1965. Temperature regulation and torpidity in the pygmy mouse, Baiomys Taylori. Physiol. Zool. 38(3): 243-254. IRVING, L . , P. F. Scholander and S. W. Grinnell. 1942. Experimental studies of the respiration of sloths. J. Cell. Comp. Physiol. 20: 189-210. IRVING, L., H. Krogh and M. Monson. 1955. Insulation and metabolism of arctic mammals in winter and summer. Physiol. Zool. 28: 173-185. JACKSON, H. H. T. 1961. Wisconsin Press. Mammals of Wisconsin. Univ. of JOHANSEN, K. 1961. Temperature regulation in the ninebanded armadillo. Physiol. Zool. 34: 126-144. JOHANSSON, K. R. and W. B. Sarles. 1949. Some considerations of the biological importance of intestinal micro-organisms Bact. Rev. 13: 25-45. JOHNSON, G. E. 19 31. Hibernation in mammals. Biol. 6(4): 439-461. Quart. Rev. KAYSER, C. 1962. Effet du sejour a basse temperature associe & 1 ’obscurity sur l'apparition prScoce de 1 'hibernation chez Le hamster (Cricetus cricetug). Compt. Rend. Soc. Biol. 156: 498-5015 to -10.1 -10 to -5.1 -5 to -0.1 0 to 4.9 5.0 to 9.9 10 to 14.9 15 to 20 Fifteen day Periods 0(7) Mar. 4 - Mar. 19 (105 opossum nights) 0(4) .15(13) Mar. 19 - Apr. 3 (103 opossum nights) - - Apr. 3 - Apr. 18 (117 opossum nights) - Apr. 18 - May 3 (114 opossum nights) - .68(49) 1.54(65) 2.23( 29) ' •v 1.46(22) 3.28( 37) 4.16( 45) 4.79( 17) 3.08( 2) 0 ( 3) 2.41( 48) 4.20( 38) 8.47( 31) 4.12( 25) - - - 5.87( 23) 6.38( 40) 6.S9C 37) 7.97( 35) 4.43(20) - - 2.93( 13) 4.2S( 38) 5.79( 50) S.07( 46) S.14( 3) Total Period .57(62) 1.47(90) 3.16(150) 4.74(161) 6.04(135) 5.75(108) 4.52(23) Standard Deviation S .47 1.05 1.66 1.70 1.77 3.46 1.52 Standard Error S * .06 .11 .13 .13 .15 .33 .32 Feb. 18 - May 3 (75 nights, 559 opossum nights) 0<11> 193 Feb. 18 - Mar. 4 (120 opossum nights) Opossum activity as a function of time of night, in activity units/animal/hour. All activity values are hourly means. 12-1 1-2 2-3 Feb. 18 - Mar. 4 Mar. 4 - Mar. 19 .09 Mar. 19 - Apr. 4 3-4 Hours of the Night ptht ----6-7 7-8 4-5 5-6 .06 .09 .21 1.53 10-11 2.82 2.56 1.87 1.48 .41 .41 .48 .68 1.94 3.48 3.84 3.52 4.27 .05 .01 .08 .22 .42 1.16 3.07 3.59 4.01 4.19 .02 .12 .30 3.23 5.38 5.79 7.09 .02 1.29 3.15 4.78 5.53 .97 2.76 3.70 3.99 4.51 Apr. 18 - May 3 .02 9-10 .23 Apr. 4 - Apr. 18 Mean - All Periods 8-9 .05 .08 .10 .15 .27 11-12 12-1 1-2 Hours of the Night A.M. 4-5 5-6 2-3 3-4 Feb. 18 - Mar. 4 1.38 1.25 .92 .34 .29 .26 .13 .01 Mar. 4 - Mar. 19 3.75 3.05 2.58 2.32 1.68 1.33 .92 .38 .07 Mar. 19 - Apr. 4 4.45 4.27 4.78 4.48 4.15 3.17 2.42 1.28 .23 .13 Apr. 4 - Apr. 18 7.40 8.78 8.55 8.02 6.85 5.60 3.51 .61 .07 .04 Apr. 18 - May 3 5.95 6.10 5.99 5.62 4.93 4.27 1.83 .15 Mean - All Periods 4.58 4.71 4.58 4.16 3.58 2.93 1.76 .47 .07 .03 Midniffht 6-7 7-8 8-9 . 46T Table 21. 19 5 Table 22. An example of data used to calculate the maintenance energy requirement for an opossum (No. 1). The period used to calculate maintenance energy for minimum activity was Feb. 24 to Mar. 3 when body weight remained essentially constant. Other periods used representing higher activity levels were Mar. 3 to 26, Mar. 9 to 17 and Mar. 9 to 26. Date 24 25 26 27 28 29 Mar. Food Eaten Cg> Body Weight (g) 115 1830 6 0 0 0 0 0 10 0 0 100 0 0 0 7 90 8 0 110 0 0 0 0 1 2 3 4 5 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 — 140 0 0 110 0 0 120 0 0 0 ■ — — — — 1615 — — 1814 — Activity Span (hours) Total Activity Units Mean Temp. (C°> - 8.6 - 6.6 -3.3 - 2.8 -5.5 -9.2 -4.7 -6.3 -2.7 0.5 -1.9 -3.9 3 25 0 0 2 0 0 6 4 9 4 17 4 9 9 1 10 2 3 6 — 4 8 — — — — — 1758 — — — — 1701 — — 1758 — — 1729 — — — 140 1672 0 120 1814 — 5 13 18 7 4 7 7 12 2.2 35 48 17 12 8 40 7 14 7 13 63 49 17 8.9 5.0 -1.9 -1 . 1 -9.2 -7.9 -0*5 4.4 4.4 7.0 7.7 66 11. 6 8 23 4.4 4 7 20 0.0 15 16 48 72 -4.4 -4.2 0.3 8.3 20 10. 0 8 15 11 8 8 15 20 21 196 FORMULA FOR THE CALCULATION OF PERCENT BODY FAT FROM DENSITY DETERMINATIONS Specific gravity is giyen by the general formula: Specific Gravity = _____ Weight of the Body Weight of an equal Volume of Water Specific Gravity = Weight of Body in Air W t . o t Body in Air - Wt. of Body in Water or On this basis, Kraybill et al. (1951, 19 52) developed a formula for calculating the percent fat in an animal’s body using density estimates for the fat and lean components of the carcass and determining the specific gravity of the whole animal. The derivation of this formula is given below. Symbols: Wg - Weight of body in air Wp = Weight of fat in air D^ = Density of lean component Dp = Density of fat component Sq = Specific gravity of the eviscerated carcass wF Wp W cdl Bp - d f) _ wB (PL - sB> Wp Dp Dj^ - WB “ dL SG = 5q Dl Dp Sq Dp Dl = yG ^ L " '^F7 SG Dp ' s^T^’-^iT « 197 Assume that Dp = .912, based on the work of Morales et al. (19if5), for guinea pigs. This is similar to fat density values reported by other workers. D^ = 1.11. Assume that The latter value is based on the four lowest densities of whole (eviscerated) opossums as follows: 1.102 (animal died of starvation), 1.107, 1.108 and 1.102, with a mean value of 1.105. (The percent fat calculated for a body density value of 1.10 5 using D^ = 1.11 is 1.9%, a reasonable value for a starved animal). Using Dp = .912 and DL * 1.11, the general formula for opossums is derived as follows: WF df dl Sq Sq