v v v .' _ _ - ' __ ' —______“ . 2. ‘..‘c.‘ oxvcsn cousuwvnon AND ACTIVITY PATTERNS IN THE MEADOW vow yscnmggg ggwsfiygmcg (one). find: for Ma 0% 6I M. S. MICHIGAN STATE UNIVERSITY Richard G. Winger? I958 mflljll l 21E um gmlmmflml L/ [OXYGEN CONSUMPTION AND ACTIVITY PATTERNS IN THE MEADOW VOLE, MTGROTUS PENNSYLVANICUS (0RD), by Richard G. Wiegert AN ABSTRACT ’Submitted to the“ College of Science and Arts Michigan State University of Agriculture and Applied Science in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Zoology 1958 Approved by 77/6 (15/ flayed/£44 Richard George Wiegert ABSTRACT The relationship between oxygen consumption and weight in the meadow vole, Microtus _p_. pennsylvanicus, was investigated. The oxygen consumption was measured in an apparatus utilizing the prin- ciple of indirect calorimury. Measurements were made on both fasting and non-fasting animals. The hourly oxygen consumption of the non-fasting voles was measured hourly over a 24-hour period. Variations in hourly readings provided a means of estimating the activity cycles of the voles. A direct measurement of locomotor activity cycles was made in cages provided with recording activity wheels. ‘ Measurement of oxygen consumption at various ambient tem- peratures showed that the temperature of thermoneutrality for the meadow vole lies within the range 250-290 Centrigrade. Voles showed a lower rate of oxygen consumption when huddled together than when measured individually. This was presumed to be due to lessened heat loss and/ or reduced activity. The oxygen consumption of fasting animals of different weights was related according to the exponential equation: cc. Oz/vole/hour = 9. 2(body wt. in gms.)° 64. The oxygen consumption of the non-fasting animals was estima- ted by the equation: cc. 02/vole/hour = 20. 6(body wt. in gms.)' 52 No significant difference between the slopes of the two regression lines was determined. Oxygen consumption in the non-fasting voles was higher, no doubt, due to the increased activity and the specific Richard George Wiegert dynamic action of feeding. A two to four-hour cycle in the oxygen consumption of the voles was demonstrated. This short cycle was probably due to activity con- nected with feeding, since other investigators have found that it per- sists in total darkness. A 24-hour cycle in oxygen consumption was found in newly-cap- tured voles. This cycle was not present in voles measured one week after capture. The author postulated that this 24-hour cycle was due to variation in activity induced by environmental conditions present in the field since it did not persist under laboratory conditions. Tem- perature and light both probably play an important role in the condi- tioning of such a cycle. Voles whose activity was measured in cages provided with re- cording exercise wheels showed no correlation of activity with times of day. As a group they did not show any correlation of activity with light or dark. A highly significant difference was present between voles with regard to the activity exhibited in the wheels. The greatest distance a vole traveled in the wheels during any one 24-hour period was 15. 3 kilometers, while one of the animals did not use the wheel at any time during one of the 24-hour periods. The equation relating oxygen consumption of the non-fasting animals to body weight was shown to be useful in estimating the main- tenance energy consumption of voles under 'natural' conditions. The data on activity patterns showed that measurement of diel variation in oxygen consumption provided a more reliable means of estimating activity patterns than did direct measurement of locomotor activity by means of exercise wheels. 'OXY GEN CONSUMPTION AND ACTIVITY PATTERNS IN THE MEADOW VOLE, MICROTUS PENNSYLVANICUS (0RD). by Richard G. Wiegert 'A THESIS Submitted to the College of Science and Arts Michigan State University of Agriculture and Applied Science in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Zoology 1958 r-w - / -, Approved by /’,.’./’m:,.// 47.4 V’fl/{Ly C- »‘~. ACKNOWLEDGEMENTS The author wishes to thank Dr. Don W. Hayne. Fisheries Institute, University of Michigan Museum Annex, who suggested the problem and served as chairman of the Graduate Committee during the initial phase of the work. Special thanks are due Dr. Marvin Hensley who assumed the duties of chairman after the de- parture of Dr. Hayne. Thanks are gratefully extended to Dr. E. P. Reineke, of the Department of Physiology, for help with many technical aspects of the problem and to Dr. Phillip Clark for his untiring efforts in di- recting the statistical treatment of the data. The author is grateful for the time devoted by Dr. T. Wayne Porter to a critical reading of the manuscript. Acknowledgement is made of the help received from Dr. Frank Golley, now of the Department of Zoology, University of North Carolina, who assisted with measurement of the range of themo— neutrality, and Mr. Robert Burns, who took the photographs of the apparatus. Grateful acknowledgement is also extended to Mrs. Bernadette Henderson, departmental secretary, for her generous aid with the many administrative details associated with such a program. TABLE OF CONTENTS Page INTRODUCTION ..................................... 1 Energy consumption ............................... 1 Activity patternS. . . . ..... . ..... . .................. 4 METHODSANDMATERIALS............... .......... 7 Apparatus for long term oxygen consumption determination.OOOOOOOOOOOOOOOOOOOOOOIO0.0.0.... 10 Apparatus for short term oxygen consumption determinations O 0 O O O I O O O O O O I O O O O O O O O O O I O I I O O O 0 O 1 7 Exercise wheel apparatus. . . . . . . ....... . . . . . . . . . . . . 20 AnimalsOOOOOOOOOOOOIOOOOOOOOOO ...... O. 00000 .00... 21 RESULTS 22 Thermoneutrality and grouping experiment 22 Fastingmetabolic rate 26 Metabolic rate with food and water available......... 32 Daily activity based on variation in oxygen consumptiODOOOOOOOIOOOOOOOOOOOOOOOOOO0.0.0.... 37 Activity as measured in exercise wheels . ........... 48 DISCUSSION .......... 56 Thermoneutrality and grouping ....... . ...... . . . . . . . 56 Fasting energy consumption. . . ................ . . . . . 57 Maintenance energy consumption . . . . . . . . . . . . . . . . . . . 60 Activity patterns based on oxygen consumption ....... 62 SUMMARY.......... ............. ......... . ..... 65 LITERATURE CITED ................................ 68 LIST OF TABLES TABLES l. 10. ll. Calculation of regression line relating oxygen consumption to body weight. Data from 50 wild voles, 67 determinations. Alldeterminations made under fasting conditions...................... Calculation of regression line relating oxygen consumption to body weight. Data from 8 voles, 14 determinations. Animals laboratory bred and fasted before measurement ............ . ...... . . . . . Comparison of the regression line equations from TableslandZ.......... ........ . .......... . ..... Calculation of regression line relating oxygen consumption to body weight. Data from 15 wild voles, 26 determinations. All determinations 24 hours in duration. Food and water available . . . . . . . . Comparison of the slopes of the regression line from Tables 1 and 4 ........ Night versus day rates of oxygen consumption in the meadow vole. All figures represent 7o of diel total per 12-hour period . . . ......... . . . ..... . ..... Analysis of variance in the 24-hour oxygen con- sumption record. Eight voles measured the day afterCaptureo.0.00....0.00....OOOOOOOOOOOOOOOOOOO Analysis of variance in the 24-hour oxygen con- sumption record. Twelve voles measured after from three to eight days in captivity. . . . . . . . . . . . . . . . Three-factor analysis of variance in the 24-hour oxygen consumption record. Two voles measured during the first, second, and third week following capture. ...... . ........ Three-way analysis of variance of activity patterns of the meadow vole as day vs. night variations. Seven voles (wild) measured in activity wheels for sevensuccessivedays”Hun”. ............ Distances traveled by the meadow vole in activity wheels. All figures are expressed as kilometers per day (24-hour period) and represent the mean value for each vole during the seven day measure- ment period ...... PAGE 29 29 31 36 36 39 46 46 49 50 53 TABLES PAGE 12. Analysis of variance in the 24-hour activity wheel record. Seven voles measured after one weekin captivity.OOOIOOOIOOOOOOOOOIOOOOOO00.0.0... 55 13. Analysis of variance in the 24-hour activity wheel record. Four voles measured after 55 several months in captivity ........................ _/ LIST OF FIGURES FIGURE 1. 10. 11. 12. 13. 14. 15. Diagram of apparatus for the determination of hourly oxygen consumption over a 24-hour period.... ....... . ........... .......... Diagram of the recording exercise wheel apparatus . . Apparatus for determination of hourly oxygen consumption over a 24-hour period . . . . . . . . . . . . ..... Contents of the animal chamber of Figure 3 . . . . . . . . . Apparatus for SiInultaneous oxygen consumption determinations on several animals. Front view Showing manometers ....... o o o o o o o' 00000000 o oooooo Apparatus for simultaneous oxygen consumption determinations on several animals. Rear view showing animal chambers . ....... . ................ Details of a single unit of the apparatus shown inFigures5and6................................ Oxygen consumption of meadow voles at various ambient temperatureSOOOOOOOOOOOI0..00.00.000.000. Effect of grouping on oxygen consumption . . . . . . . . . . . Graph of fasting 02 consumption in relation to bOdywej-ght.OOOOOOOOOOOOOOO0....00.000.000.000... Non-fasting 02 consumption in relation to body weight OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOIOOOOOOO Graph of average maximum, average minimum, and mean hourly values for 02 consumption over a 24-hour period, plotted by weight class . . . . . . ..... Frequency and duration of the short activity cycle. . . . Activity patterns of meadow voles 1 day after capture. Based on diel variation in 02 consumption. . Activity patterns of meadow voles 3 to 8 days after capture. Based on diel variation in 02 consumption PAGE 12 12 18 18 19 24 25 30 33 38 41 44 45 INTRODUCTION In recent years, within ecological circles, there has been in- creasing concern with the question of energy dynamics of populations. Some means of estimating the energy consumption of a population has become desirable as well as ability to assign values to different types of energy utilization such as energy used for maintenance, growth, and activity. ' This study was inaugurated in an attempt to answer some of the above problems with regard to energy consumption in the meadow vole, Microtus pennsylvanicus. The method of measurement of the energy consumption permitted the investigation of some other aSpects of the life of this species, specifically those of activity, critical and lethal temperatures, and the effect of huddling on the energy consumption. Some of these were only briefly considered but results should prove of interest, both for their inherent value and as a stimulus to further investigation in these areas. Terminology and definitions of technical terms are in accordance with Brody (1945). Energy Consumption In attempting to determine the energy consumption of various weight classes in a given population the investigator is immediately faced with the problem of a decrease in the basal metabolic rate as the weight of an animal increases. Benedict (1915), Brody (1945), Kleiber (1947), and Sherwood (1936) have all shown this phenomenon in homeotherms. Zeuther (1953) showed that the relationship holds for species ranging in size from bacteria to large mammals. Brody (1945) found that metabolism, either between species or within species, showed a relationship to weight which can be defined by the generalized equation: Y = a Xb where: *4 II amount of oxygen consumed a constant for the species or group in question {D II X = weight of the animal D" n slope of the regression line (regression co-efficient) The value (b) receives the most attention since it expresses the relationship existing between the oxygen consumption and the weight of the animal. A rather extensive literature has developed concern- ing the value of (b) for comparisons of different species of homeotherms. Brody (1945), and Kleiber (1947), have rather conclusively shown that (b) is significantly different from zero. Brody found that a graph of the oxygen consumption of mammals ranging in size from mice to elephants, when plotted against body weight, permitted the calcula- tion of a regression line whose exponent (b) was equal to o. 73. Kleiber suggested that this figure should more appropriately be 0. 75. Within a species Galvao (1947) found a (b) value of 0. 90 for dogs. Chambers and Summerson (1950) presented a general review of energy metabolism. The reasons for the general decrease in rate of metabolism with increase in weight are not fully known. McCashland (1951) proposed that the cells of an animal's body decrease in physiological activity as the animal ages. Kleiber (1947) discussed several proposals and concluded that there was a general decrease in rate of oxygen consump- tion of the tissues as weight increased. He found that tissues measured in vitro showed a lower rate of oxygen consumption when they were taken from heavier animals. In contrast to this, Bertalanffy and Estwick (1953) found no difference in oxygen consumption between tissues taken from animals of different weights. There is, at the present time, some disagreement as to whether the oxygen consump- tion of the same tissues in vitro adequately reflects the oxygen con- sumption of the same tissues in vivo. Since the reason for the de- crease in rate of metabolism with increase in weight was not within the scope of this study further discussion is unnecessary. Further information may be found in a paper by Martin and Fuhrman (1955). Few data are available concerning the metabolic rate of Microtus. Hatfield (1939) measured the carbon dioxide production of individual voles for short periods of time. Hatfield's data were recalculated as oxygen consumption by Pearson (1947) on the basis of an assumed R. Q. of 0. 8. This gave a metabolic rate of 5. 2 cc. of oxygen per gram of body weight per hour. Pearson postulated that this high value indicated that Hatfield's voles were active during the period of measurement. However, there are many errors in- herent in the estimation of oxygen consumption from observed car- bon dioxide production. Pearson (1947) and Morrison (1948 a) re- ported data on the oxygen consumption of Microtus penngvlvanicus measured over a 24 hour period, with the voles provided with food and water. Their studies, however, were of a preliminary, compara- tive nature involving a series of five runs on only two voles. To determine the relationship between basal energy metabolism and weight the animals must be in a fasting condition, inactive, and the measurements must be made at the temperature of thermo- neutrality. Conversely, if the energy cost is to be estimated for conditions occuring naturally in the field, where food and water are available and normal activity occurs, then some method of measure- ment must be devised in the laboratory whereby these conditions are provided when the oxygen consumption of the animal is measured. Hitherto there has been little concern with the measurement of oxygen consumption in other than the fasting condition. This is generally due to the fact that measurements, made where food and water are avail- able and activity permitted, are assumed to show more variation than measurements made under fasting conditions. The results presented here seem to show that the latter statement may not be correct in the case of small wild mammals. The present study shows that there is a great need for more investigations of oxygen consumption under con- ditions approximating those found in the field. Oxygen consumption determinations under 'basal' conditions are of little use in the estima- tion of energy consumption by mammals in their natural environment. Activity Patterns The activity pattern exhibited by Microtus is a subject which has received much attention by previous investigators and has resulted in some disagreement as to what time of day this rodent is most active. Hamilton (193 7a, 1937b) presented trapping data which showed the vole to be active at all times of the day with peaks just after dawn and shortly before dusk. Seton (1909) described M; pgmsylvanicus as being chiefly diurnal in the far north, while Wood (1910) stated that Microtus is not strictly nocturnal. Hatt (1930) found this vole to be active at all times of the day. Emlen et. a1. (1957) measured the activity of the meadow vole by the use of dropping boards and showed activity peaks between the hours of 7-10 a. m. and 7-10 p. m. Several laboratory studies of the locomotor activity of the genus Microtus have been made. Davis (1933) measured a single female LII: agrestis and found peaks before dawn and at dusk, plus a short 2-4 hour cycle which he attributed to feeding activity. Hatfield (1940) found the same 2-4 hour short cycle in Microtus and showed that light had no modify- ing effect. Mann (1954) extensively investigated locomotor activity as exhibited by _N_I_. pennsylvanicus in electrically recording cages and showed a significant degree of nocturnality in his animals but no well defined bimodality in the 24 hour cycle. Morrison (1948a) de- termined the activity patterns, both the 24 hour cycle and the shorter 'feeding' cycle of the meadow vole, by measuring diel variation in oxygen consumption. He found the familiar 2-4 hour short cycle and also found the voles to be nocturnal with an activity peak at midnight. His method had the further advantage of giving a more quantitative record of acti- vity than measurement of locomotor activity alone. The rationale behind the use of variation in oxygen consumption as a measure of activity is well supported by the work of Hart (1952a), who showed that the greater part of daily variation in oxygen consumption in the albino laboratory mouse was due to activity. Fuhrman et. al. (1946) also showed a correlation of activity with changes in oxygen consumption in the white mouse. Hart (1950) showed that a given amount of activity produced the same increment of oxygen consumption over the basal rate regardless of the environmental temperature. The validity of measurements of activity made in the laboratory versus those based on field observation have been questioned. Elton et. a1. (1931) showed a correlation between the activity of the wood mouse, Apodemus sylvaticus, as measured in the laboratory and in the field. Johnson (1926) found no difference between the activity of the deer mouse, Peromyscus s . , in the laboratory and that exhibited in the field. Johnson (1939) found that continuous light had little effect on the activity patterns of Peromyscus leucopus, while Davis (1933) found that being kept in continuous darkness for 24 days had little effect on the activity cycle of Microtus agrestis. In contrast to the above, Calhoun (1945), in a study of activity patterns in the prairie vole, Microtus ochrogaster measured in activity wheels, found that the daily cycle could be completely reversed over a period of 10 days. Furthermore, Hatfield (1935) found that the daily cycle in M. ' californicus was modified by environmental temperature changes. In view of the above noted disagreements the author decided that a secondary purpose of this study should be the systematic in- vestigation of the effect of continued captivity on the activity pattern of the meadow vole with conclusions to be based on data derived from hourly variation in oxygen consumption. As a corollary, the activity patterns of several voles should be measured using values for loco- motor activity obtained through the use of recording exercise wheels. The results permitted a comparison to be made between the two methods of activity measurement. METHODS AND MATERIALS Introduction The investigator attempting to determine the energy consumption of a small mammal is faced with a variety of methods. Metabolism cages are often used where the food and water intake of the animal is measured and the feces and urine collected. Determinations of the energy content of these materials then enables one to calculate the energy consumption of the animal. Since it does not give any indication of activity patterns, nor will it permit the investigation of changes in metabolic rate with changes in weight, this method was not deemed adequate for this study. Some form of animal calorimetry is thus seen to be a more sa- tisfactory choice. Brody (1945) gave a good discussion of the methods of direct versus indirect calorimetry. He showed that, in general, the method of indirect calorimetry gives the most accurate results with the least expenditure of time. The cost of the apparatus ne- cessary for the direct method is also often prohibitive. In indirect calorimetry the oxygen consumption is measured and the energy con- sumption can be estimated by multiplying by the factor 4. 825. This factor represents the average caloric equivalent of oxygen in Kilo- calories per liter (Brody, 1945). The production of carbon dioxide is often measured at the same time and the R. Q. (respiratory quo- tient or ratio of moles carbon dioxide produced to the moles oxygen consumed) is calculated. This R. Q. value is corrected for protein oxidation by measuring the urinary nitrogen and is then used to calculate the relative amounts of carbohydrate and / or fat being oxidized by the animal. Theoretically this permits a more accu- 80 rate estimation of the caloric equivalent of oxygen. In practice, how- ever, the method is often superfluous. At times it may even be mis- leading since carbon dioxide production is often changed by such tem- porary physiological conditions as hyperventilation or incomplete oxidation of foodstuffs. Richardson (1929) specifically questions the use of carbon dioxide production when the animal is permitted to exercise or form fat deposits. In addition, the range of caloric equivalents of carbon dioxide is larger (5. 0 - 6. 7 Cal / liter) than that of oxygen (4. 7 - 5. 0 Cal / liter). Thus, it is apparent that small errors in the measurement of carbon dioxide production will intro- duce larger errors into the energy estimation than will comparable errors in the measurement of oxygen consumption. Furthermore, the entire range of values of the caloric equivalent of oxygen is often within the limits of experimental error (Brody, 1945). For these reasons the measurement of carbon dioxide production was omitted during this study. Many methods for determining the oxygen consumption of small mammals, using the principles of indirect calorimetry, have been described in the literature and a complete enumeration of them is impossible within the scope of this paper. Some recent techniques are those of Davis (1932), Dewar and Newton (1948), Grad (1952), Hart (1950), Holtkamp et. a1. (1955), Lilienthal (1949), Maclagan and Sheahan (1950), Morrison (1947), Morrison (1951), Smith (1955), Topliff and Cullum (1955), Watts and Gourley (1953). A modification of the method described by Maclagan and Sheahan (1950) was found to be most satisfactory for the purposes of this study. The apparatus is of the closed circuit type where the animal rebreathes air from which the carbon dioxide has been removed by soda lime placed :3 r Lazl Dr 2 7 [o [0 I’d Figure 1. Diagram of apparatus for the determination of hourly oxygen consumption over a 24-hour period. El ”i “3 ha Figure 2. Diagram of the recording exercise wheel apparatus. 10. in the bottom of the chamber. The apparatus can be easily modified to permit the hourly measurement of oxygen consumption over a 24 hour period. The method of Maclagan and Sheahan can be used with smaller animal chambers to simultaneously measure the short term fasting oxygen consumption of several animals. Economy and sim- plicity of design and construction were also factors of considerable importance in the selection of the apparatus. Apparatus for lggtirm oxygen consumption determination The modified apparatus for long term metabolic determinations is shown diagrammatically in figure 1. Two units were used with the animal chamber of each unit consisting of a 250 mm. Pyrex desicca- tor (A). This was provided with a cover into which a two-hole rubber topper was fitted. A thermometer (B) was inserted through one of the holes and the other permits a glass tube to lead to the mercury manometer (C). In the construction of these manometers the scale was made from wooden meter sticks. The U-shaped tube was made from glass tubing with a 5-mm. inside diameter. A plastic float was placed on top of the column of mercury in the side of the tube open to the atmosphere. This in turn contained a hollow metal wire, connected to a plastic scriber which contacted the drum of the kymo- graph (D). The kymograph was constructed by using a standard elec- tric alarm clock, with the hands removed, plus a drum made of a frozen juice can which was connected to the minute hand axis so that the drum would make a complete revolution hourly. Beyond the branch to the closed end of the manometer, the glass tubing continues to the two way stopcock (E). This permitted evacuation of the desicca- tor by the filter pump (H) and the subsequent refilling of the apparatus with oxygen from the cylinder (G). The gas bag (F) was used as a 11. reservoir for the oxygen from the cylinder. The principle of the apparatus depends on the fact that in a con- stant volume system, under stable conditions of temperature and barometric pressure, the volume of gas in the system will decrease proportionately with the internal pressure. When an animal is placed in the chamber it consumes oxygen. The carbon dioxide produced is ab- sorbed by soda lime in the bottom of the chamber causing the internal pressure to drop. The oxygen consumed can then be calculated from the relationship: (1) oxygen consumption in m1. /hr. = (V-Va) x P x 273 7615' 273% T Where: V = net volume of unit in ml. Va: volume of animal (assuming 1 gm. = 1 m1.) P = pressure difference registered over time interval T t = temperature inside animal chamber in degrees C. T = time interval of determination in hours The formula assumes the constancy of both barometric pressure and chamber temperature during the period of measurement. It is obvious that over a 24-hour interval one, or both, of these factors are going to lchange. Before inserting (P) in the above equation corrections for changes in barometric pressure which occured during the period of measurement are easily made. To correct for changes in tem- perature we use the relationship: (2) Q = P2 " (P2 Jul) 72- where: ——r .‘ ‘l .‘ . . w ' 'l .q H J um 91* " m fl ‘. u «my ‘7 ill)“ '13 - ? m, A Figure 3. Apparatus for determination of hourly oxygen consumption over a 2M-hour period. -' . , I. s - - - h \— v' i ' _ r :--,£-M "Mm-‘1'" “"?n.- - ' .--l.- ‘ Figure 11. Contents of the animal chamber of figure 3. of. 13. Q = factor by which P must be corrected before insertion into equation (1). P2: barometric pressure - P. t1 = original temperature reading in degrees Kelvin. t2 = temperature reading at end of run in degrees Kelvin. The temperature during a measurement period was never lower than 20° Centigrade nor above 30°. Therefore the fraction t1 was Tz’ taken as a constant having the value 0. 996 for a one degree change in temperature. Utilizing.this factor a table of values was calculated which showed the correction for any change in temperature combined with any value of P2. For simplicity the values of P2 were taken at intervals of 50 mm. of mercury since the change in the correction factor (Q) for smaller intervals was negligible. It must be remembered that if the change in barometric pressure is positive, then its value must be subtracted from (P) before substitution in equation (1). If the temperature change is positive, then the correction factor (Q) will be positive and vice versa. Having established the theoretical basis the discussion may now proceed to the description of the apparatus in use. In figure 4 a close- up of the animal chamber shows its contents (nest box, containers for holding water and good, a large Petri dish filled with soda lime, and the wire screening for supporting the animal). The volume of the sys- tem was determined by first weighing the empty desiccator. It was then filled with distilled water of known temperature and re-weighed. From the difference in the two weights and the calculated density of the water the volume of the desiccator was found. From this value 14. the volume of the contents (food dishes, wire, etc.) was subtracted. The volumes of these items were found using their densities as given by Hodgeman (1952-53). The volume of the connecting tubing and other fittings was calculated and added to that of the desiccator. The animal was placed in the chamber (A) of figure 1. The lid of the desiccator was sealed with vaseline. With the stopcock (E) turned to the off position, the gas bag (F) was filled from the oxygen cylinder (G). Then the stopcock (E) was turned so as to permit the filter pump (H) to draw a vacuum in the desiccator (A). This should approximate 200 mm. of mercury as measured on the manometer (C). The stopcock (E) was then turned so that the vacuum was re- placed with pure oxygen from the gas bag (F). Since Krogh (1916) showed that an increase of the partial pressure of oxygen of this magnitude had no effect on the metabolic rate in humans it was assumed that the same would hold true for Microtus. The system was now charged with oxygen and the stopcock (E) again turned to the off position. The method of charging the apparatus described above is a modifi- cation of the original Maclagan-Sheahan apparatus devised by Dr. E. P. Reineke of the Department of Physiology, Michigan State Uni- versity. The modified version permits a quantitative measurement of the amount of oxygen injected into the system in contrast to the method of charging the desiccators described in the original paper. This quantitative method also permits calculation of the point beyond which the animal would be subjected to partial pressures of oxygen lower than those prevailing in the atmosphere. The author has cal- culated that after replacing a 200 mm. vacuum with pure oxygen, the animal may be permitted to draw a vacuum of approximately 160 mm. 15. of mercury before the partial pressure of oxygen in the system will be reduced to that of atmospheric air. This value will vary from day to day depending on the barometric pressure and the percent of oxygen in the atmOSphere. For the purposes of this study such variation is negligible. In a few cases where large voles were being measured it was realized that the 160 mm. limit would not suffice for the usual 12-hour period of measurement and in these cases a vacuum of 300 mm. of mercury was replaced with pure oxygen preparatory to start- ing a period. Having turned the stOpcock (E) to the off position, the apparatus is left for one hour to establish temperature equilibrium. This period also gave the animal time to become accustomed to the chamber. After one hour the temperature of the air inside the animal chamber and the pressure difference (P) between the air in- side the chamber and the atmosphere was recorded. A reading was taken of the level of mercury in each arm of the manometer tube and the difference between them recorded as the value of (P). This pro- cedure prevented errors which would have been caused by irregu- larities in the tubing if readings were taken from zero on one side of the manometer only and then doubled. Having recorded the tem- perature and the pressure difference, the scriber was brought into contact with the smoked paper on the drum of the kymograph. The scriber was held against the drum of a fine, flexible guide wire (Figure3). The attachment of the kymograph to the apparatus is a modification devised to permit hour by hour measurement of the oxy- gen consumption of the experimental animal. The point where the period of measurement started was recorded on the smoked drum. As the level' of mercury in the right arm of the manometer dropped, reflecting the consumption of oxygen by the animal, the scriber 16. described a continuous spiral around and down the drum. The vertical distance between any two marks was then equal to the value of (P/ 2) for one hour. The kymograph drum was removed at the end of a measure- ment period. The distance between the lines was then measured, re- corded, and the drum resmoked before being replaced. The tempera- ture, value of (P) on the manometer, and the barometric pressure were also recorded at the end of a period. For adult Microtus the apparatus had to be recharged after 12 hours. Since the process of removing the kymograph, letting oxygen into the system, and replacement of the drum resulted in a loss of some time, the hourly measurements of the last half of a 24-hour run were about five minutes out of phase. However, measurements were started at various times of the day and night, thus this difference was assumed to have no measurable effect on the resulting calculation of activity patterns. The young animals, due to their lower total oxygen consumption, could often be left for the full 24 hours without recharging the apparatus. In practice a single kymograph could serve for two manometers by merely varying the width of the tip of the scriber to permit recog- nition of the individual marks. The relationship of the manometers to the kymograph is shown in figure 3. The hourly oxygen consump- tion values cannot be corrected for changes in barometric pressure and/or temperature. However, these factors generally changed only a small amount over the course of one hour and are negligible when compared to the experimental error of 5 to 10 per cent encountered in making these hourly readings. 17. The experiments were conducted in laboratories where the tem- perature remained relatively constant, variations greater than three degrees Centigrade were never encountered in a single 24-hour period. Thus temperature change was never a large factor and such correc- tions as were made were minor. Apparatus for short term oxygen consumption determinations A second aspect of this study, that of measurement of the fasting metabolic rate of Microtus, presented problems differing somewhat from the above in that it was desired to have simultaneous short term measurements of the oxygen consumption of several animals. The apparatus devised for this purpose is shown in Figures 5, 6, and 7. The principle is the same as that of the previously described appara- tus except the kymograph is omitted and quart Mason jars were used as the animal chambers instead of desiccators. The single small de- siccator shown in Figure 6 was used to hold several animals in the grouping experiments. The smaller jars permitted an oxygen consump— tion measurement of good accuracy (experimental error of less than 3%) in a shorter period of time. One-half hour was usually chosen for the length of this measurement period with six animals being mea- sured simultaneously. The animals were placed in a small wire cage as depicted in Figure 7. This served to restrict movement and ren- dered the metabolic rate more truly representative of the theoretical 'basal' rate. Two-piece metal jar lids were used on the chambers with a copper tube soldered into the center of each. Stopcock grease was applied to the rim of the jar each time it was sealed to prevent leakage. Short thermometers were placed inside the chambers to record the temperature during the course of the run. After charging the apparatus Figure 5. Apparatus for simultaneous oxygen consumption determinations on several animals. Front view showing nanometers. ' Figure 6. Arnaratus for simultaneous oxygen consumption determinations on several animals. Rear view showing animal chambers. :aciovnvo‘JUA.‘ zu- ‘gmcvovl0|8"' .. . Figure 7. Details 0‘ a 9‘? “"I e ' 1- " L 1. .- H . [111113 Of ,he a_’-paIEJtUS shown 20. with oxygen the voles were left undisturbed for 20 minutes to allow temperature equilibration to take place before a period of measure- ment was begun. During this time the stopcocks to the gas bag were left open so that the animals could draw in oxygen to replace that used in respiration. The starting time of a period was recorded at the mo— ment of closing the stopcocks. Exercise wheel apparatus Measurement of the activity patterns of a group of Microtus using recording activity wheels was conducted in a laboratory where the room temperature was maintained at approximately 23 degrees Centigrade, the humidity controlled, and the only light source was artificial. The lights were turned on at 7:30 a. m. and remained burn- ing until 9:30 at night. The apparatus consisted of a series of 13 acti- vity wheels. The essential parts are diagrammed in figure 2. Each consisted of a plywood wheel (A) with a wire screen bottom (B), open on one side, and mounted on the front hub of a bicycle (C). From the Open side of the wheel the vole could pass through a small hole in the plywood frame into a small wire cage (D), which was large enough to contain a porcelain food dish (E) and the spout of a drinking bottle (F). The animals were fed a diet of rolled oats and carrots during the course of the experiment. The wheel was connected to a counting de- vice (G), by means of an eccentric (H) and lever (I). The voles were placed in the cages four days prior to starting the recording period in order to become accustomed to the cages. It was noticed that many immediately began to run in the wheel while others were still not using the wheels at the end of the four day period. Readings were taken each morning one-half hour after the lights were turned on and 21. each evening one-half hour before the lights were turned off, except for one 24-hour period during which readings were made every two hours. Animals The animals used in the study were live trapped from areas in the vicinity of the laboratory on the campus of Michigan State Uni- versity. All trapping was done in the evening. The traps were set and baited with rolled oats about 4:00 p. m. and the voles removed be- tween 9:00 and 10:00 p.m. The Microtus population in this area was very high during the period of the study and no difficulty was encoun- tered in obtaining suitable numbers of animals. All voles which were to be fasted for 12 hours were placed in cages with water but no food and left overnight. The next morning at 9:00 a.m. they were placed in the apparatus. At the conclusion of a period of measurement they were fed and at noon released at the trap in which they had been cap- tured. The procedure for handling those voles which were to be placed in the apparatus with food and water was the same except that they were given food overnight and were not released until after a 24-hour oxygen consumption determination had been made. 22. RESULTS Thermoneutrality anchroupmgchperiment The term 'fasting metabolic rate' as used in this discussion is de- fined as the rate of oxygen consumption of the animal measured under conditions of thermoneutrality, in the postabsorbitive condition, and with activity restricted to a minimum. This corresponds to the commonly used term 'basal metabolism'. Since the activity of wild mammals can seldom be reduced to zero perhaps 'fasting metabolism' is a more accurate term. It implies that some extraneous activity is present during the period of measurement. Extensive investigations have been carried out on the range of thermoneutrality in the albino laboratory mouse. Morrison (1948b) reports that the work of several authors has given a range of 28 to 34 degrees Centigrade for this species. To extrapolate such values to different Species is difficult however, and since no known data are available on the critical temperature in the meadow vole, the oxygen consumption of three fasting voles was measured at various ambient temperatures. The apparatus used was that shown in Figures 5, 6, and 7. The three Mason jars containing the voles were submerged in a small water bath. Measurements were made at eight different temperatures. The actual environmental temperature was determined by means of small mercury thermometers placed inside the animal chambers. The length of a period of measurement was 20 minutes. One—half hour was allowed for temperature equilibration after raising the tem- perature of the water bath and before starting the next period of mea- surement. The temperatures inside the animal chambers were one to 23. two degrees higher than that of the water bath. The graph in Figure 8 shows the results of this experiment. Be- cause of differences in total oxygen consumption between voles the graph is drawn showing the percent of the minimum oxygen consump- tion plotted as a function of the environmental temperature. The re- sults indicate that the lowest oxygen consumption occured between 25 and 25 degrees Centigrade. All subsequent metabolic determinations were made within this temperature range. As the environmental temperature was increased the voles exhi- bited signs of heat distress such as rapid breathing and flattening out on the screen of the animal cages. When the temperature of the cham— bers was increased from 34. 5 to 39 degrees Centigrade one animal died. The other two animals lived for only 40 minutes at an ambient temperature of 39 degrees. For this reason the oxygen consumption at 39 degrees could be measured only for 10 minutes. It would appear that under the conditions of the experiment the lethal temperature for the meadow vole lies in the range of 34 to 39 degrees Centigrade. This agrees with the value of 36 degrees Centigrade given by Dice (1922) for the prairie vole, Microtus ochrogaster. Fasting metabolic studies are often made by measuring groups of animals in the same chamber. Since in this study fasting rates were investigated with a view to determining the relation of weight to oxy- gen consumption, the animals were measured individually. However, it was deemed advisable to measure several individuals, both singly and in groups, to determine what differences, if any, existed between the two conditions with regard to oxygen consumption. meadoaaccsou unodpsa udo«aw> ve moflob aoceos Mo soapgfi9maoo.aownko .w canmdh .o o .amQEdno dosage ceased uhapmamasoa 3 mm on an mm on ma mu R... «N om 2 2 Id u u d d d q q q 1 OHmEmm .mm .t Illl. 3% .S a. .I-| 39s .8 a ulll O O N O \O 0 no 3 mnmtutu exoqe % as uotzdmnsuoo ZO COM Oxygen consumption ~ cc./gram/hour 14.0 L. O 0— Single ._Groured ‘0 3.5 L- O’ 3.0 ‘> O o o o o o O (D I) 2.5 __ . O O O 2.0 - O 1 l 1 l 1 1 l J 1 1 2 3 a 5 6 7 8 9 # mice -——- 3 3 3 b b 3 5 5 5 Term. °o.-23 2h 26 27 27 27 29 29 29 Figure 9. - Effect of grouping on oxygen consumption. 26. Nine determinations were made in which several voles were mea- sured individually and then the same animals were measured as a group. The size of the groups varied from three to five voles. No attempt was made to assess the effect of group size on oxygen consumption. All determinations were conducted in temperature range of 25 to 29 de- grees Centigrade. All results were expressed as the rate of oxygen consumption or cc. / gm. / hour, and are shown in Figure 9 along with the temperature at which each group was measured. The voles, with one exception, had a lower rate of oxygen consumption when in a group. This difference was subjected to statistical analysis using the non- parametric Wilcoxson test described by Siegel (1956). Since the hypo- thesis that grouping would actually lower the rate of oxygen consump- tion was advanced before the start of the test, the one-tailed value for (p) was applicable. The difference in rate of oxygen consumption due to grouping was found to be highly significant (p = less than 0. 005). No increase in the tendency to huddle was noted at the lower ambient temperatures. In the single case where the oxygen consumption rates were about equal in the grouped and individually measured voles, it was observed that the animals did not huddle together when grouped but remained separate and even showed some degree of activity. Fa stingmetabolic rate A total of 50 newly-trapped voles were used in the investigation of the relation of fasting oxygen consumption to body weight. Repeat ob- servations were recorded on 12 animals and were averaged in order to obtain a mean value for the individual? Because of the scarcity of voles in the 10 to 15 gram weight class, eight laboratory bred voles were used to determine the oxygen consumption of juveniles. The wild 27. trapped voles were used as they were trapped, no attempt being made to select for size. This accounts for the preponderance of animals in the 20 to 30 gram weight class. The voles were removed from the traps at night, marked, fasted for 12 hours, and then placed in the apparatus. The period of measurement of oxygen consumption was 30 minutes. The temperature during the periods of measurement was in the range 25 to 29 degrees Centigrade, with most determinations made at 26 or 27 degrees. In most determinations the animals were relatively quiet. The data from periods during which they showed high levels of activity were excluded. All data are shown in figure 10 as a double logarithm plot of oxygen consumption against weight. Data for wild trapped voles are shown as dots. while values from juvenile, laboratory bred voles are represented by open circles. Equation lines were fitted by the method of least squares (Brody, 1945). The basis for this method is that the exponential equation (Y = aXb) reduces to the linear equation (log Y = log a + b(logX). By taking logarithms of the data the treatment is essentially the same as that for a straight line regression. The per- tinent calculations are shown in Tables 1 and 2. Details of the calcu- lation of (b), (Syx), and (r) were adequately treated by Brody (1945). The five percent confidence limits of (b) were calculated by the method of Snedecor (1956). The significance of the correlation coefficient (r) was obtained from Table 7. 6. 1 in Snedecor (1956). Although the use of (r) was fully justified in the case of the wild trapped voles, it's vali- dity in the case of the juvenile, laboratory animals was questionable since there was some selection for size. The value of the regression coefficient (b) for each group was shown to be significantly different from zero. The correlation coefficient (r) for the wild voles was 28. significant at the one percent level while that for the laboratory voles was significant at the five percent level. The scattering of the data in figure 10 reflects the difficulty of accurately measuring the fasting metabolism of small animals. This scattering is probably due to increased activity owing to their being deprived of food. Although fasting metabolic measurements made on laboratory white mice and rats uSually show less scattering than those made with food and water available, it is probably only due to the better tolerance of prolonged fasting exhibited by the latter animals. A comparison was made between the two regression lines of figure 10 using the method of covariance (Snedecor, 1956). This me- thod offered a way of comparing the two values for (a) and the two values obtained for (b) in Figure 10. The data and computations of this analysis are given in Table 3. No significant difference could be determined between the slopes of the two regression lines. A signi- ficant difference does exist between the two values for (a). This in- dicates that the relationship between oxygen consumption and body weight follows the same exponential relationship in the two groups but that the juvenile, laboratory bred voles show a generally higher level of metabolism. The reasons for this are not apparent. The juvenile animals were somewhat more active while in the metabolism chambers. Perhaps the diet of these captive voles exerted some influence which resulted in a higher metabolic rate. More probably the higher rate of metabolism was due to the above noted increased activity since diet should have had little effect, all the voles having been measured soon after weaning. Table l 0 Calculation cf regression line relating OP consumption to body weight. Data from 50 wild volesf67 detemlinations. All determinations made under fasting conditions. Sum log,X. = 70.7355 n.= 50 Sum log Y’ = 93.6817 Sum log XI: 100.5682 Sum.log'Y = 175.9319 .61; log 3* = 0.961395 + 0.61211 X Y = 9.2 X b = 0.6“.111 fidflcial limits Cf (b) x "9 0.1358} and 0.5265 II \II + stardaré deviation - standard deviation — ’1 I 1414 4 = 0.5052 . = less than 0.01 ’d Table 2 0 Calculation of regression line relating 0 consumption to body weight. Data from S voles, l” de§orninztions. Animals laboratory bred and fastrd before measurement. Sum lcgrx == 9.7507 n = 8 sum loggY = 1h.7038 “.um log—:1: = 11.,198 sum iog“: = 27.0%; .79 r I‘ - . .- w 10?: ‘1. = O.“ 9:9 -{' C07C’lY K v =2 { .76 a: - O O -/ + stfnbard deviation = 11.0 3 r = 0.7060 - standard deviition : 9.9 C p = less than 0.05 150 190 130 120 110 H O Q ‘3 0 consumption — cc./vole/hour 03 Q °2 fl A0 A 60 50 Figure 10. - Graph of fasting 02 body weight. l 30 hO Body weight in grams consumption in relation to Table 3. Comparison of the regression line equations from tables 1 and 2. 2 2 Sum y2--‘2 ‘Tean df Sum.x Sum x.y Sum y b f (Sum xy) /Sum x2 Square Lab. voles 7 .0u61 .0565 .0t09 .792 6 .0120 .0020 Hild voles us .u920 .5199 .M067 .602 no .2015 .00h2 Eithin 5M .2135 .0062 Peg. Coeff. 1 .0009 .0009 Comion 56 .5uu1 .556h .hu76 .655 55 .21u2 .0039 Adj. Moons 1 .0619 .0619 Total 57 .6661 .u096 .h66o 56 .2721 For (b) the value (F) = .COO9[.0062 = less than 1.0. The difference between the slones is non—significant. For (a) of the constant of the equation, the value (F) == .0619/.0039 = 15.9 df.= 1,55. The difference is significant at the One “extent level. 32. Metabolic rate with food and water available The relation between oxygen consumption and body weight was in- vestigated in voles which were provided with food and water and were free to move about and exhibit 'normal' activity patterns. The appara- tus used was that shown in Figures 1, 3, and 4. Although the animal chamber offers some freedom of movement, it is still rather restricted when compared to field conditions. The animals were placed in the chamber along with a weighed quantity of rolled oats, carrots and lettuce. Water was provided in small dishes. Periods of measurement were started during different times of the day to minimize any effect which starting time might have had on activity patterns and /or rate of oxygen consumption. All observations were made at the temperature range of thermoneutrality. Figure 11 shows a double logarithm plot of the data representing the oxygen consumption as the mean hourly value for each animals plotted against body weight in grams. Because of the uneven distri- bution of the size classes in the field and the relatively small number of mice used, weight selectivity of the animals was deemed desirable in order to insure adequate representation. The 15 solid dots repre- sent data from 26 determinations. Thus, some of the points are mean values of from two to four measurements on the same individual. The weight change in the animal between replications ranged from two to six grams, with a mean value of 2. 8 grams. The increase in accuracy of estimation of the individual metabolic rate as a result of these re- plications outweighed the error introduced by the relatively small changes in weight. The two open circles of Figure 11 represent data from determinations on juvenile, laboratory-bred voles. Each of these open circles represents the mean hourly oxygen consumption of two 24-hour periods of measurement. In general these results appear 170L 160 150 lbO 150 H N O 110_ consumption — co./vole/hour H O 0 On C. 80 K5 1 l 20 30 no Body weight in grams 1 Figure 11. - Non—fasting 02 consumption in relation to body weight. 34. to be higher than would be expected on the basis of the extrapolated regression line for the wild-trapped voles. These data are shown to further delineate graphically the oxygen consumption weight relation- ship for the lighter weight animals. The small amount of data avail- able did not permit quantitative interpretations such as were made be- tween wild and laboratory bred voles in the section concerned with fasting oxygen consumption. The data for the oxygen consumption mea- sured with food and water available shown much less scattering than the fasting metabolic data. This can be readily seen from a considera- tion of the standard deviations, which are only half as great in the case of the non-fasting voles. The equation representing the regression line and the standard deviations were calculated using the least square method (Brody, 1945). Morrison (1948) reported the mean hourly rate of oxygen consumption of two meadow voles. The regression equation of Figure 11 fits his data fairly well. Some deviation is to be expec- ted due to individual variation between voles. All calculations of the regression coefficient (b), and the standard deviation (S), and the calculation of the confidence limits of (b) are shown in table 4. Because of the fact that values of weight were se— lected, the correlation coefficient (r) was not used. The calculations of Table 4 show that the value of (b) for the population from which the sample was taken is within the range 0. 39 to 0. 65 unless the sample was of the aberrant type occuring one in 20 times. Apparently the value of (b) for meadow voles measured in a non-fasting, active state is significantly different from the value 0. 75 proposed by Brody (1945), or the value 0. 75 proposed by Kleiber (1947). The situations are not comparable, however, since measurements were made on growing animals of the same species while Brody and Kleiber used data from 35. adult, fasting animals of different species. The regression line equation of Figure 11 was compared to that representing the fasting metabolic rate of wild Microtus (Fig. 10), in order to test whether the provision of food and water to the voles along with permitting 'normal' activity had a significant effect on the value of (b). The method of covariance analysis was used (Snedecor, 1956). The calculations and results of this comparison are shown in Table 5. No significant difference was shown to exist between the values of (b) in the two regression lines. The evident difference between the levels (values of a) of the two lines is so great that no statistical comparison seemed necessary. This higher rate of oxygen consumption in the non- fasting animals was undoubtedly due to the specific dynamic action of the food consumed plus the higher activity level. A measure of the diel variationin oxygen consumption is given by the maximum and minimum hourly rates which were recorded during the 24-hour period of measurement. The mean values per weight class for maximum, minimum, and mean hourly oxygen consumption per vole are graphically illustrated in Figure 12. When calculating the means all individual determinations were given equal value, regardless of whether they were repeats on the same animal. The number of 24- hour periods of measurement represented by the average is shown in parentheses below the weight class designation. The mean hourly oxygen consumption for each animal was calculated and this value was used to calculate the mean value for the weight class. Superimposed on the curve of the mean oxygen consumption per weight class is the curve. of the equation (Y = 20. 6 X' 52) which represents the relation of oxygen consumption to weight in non-fasting voles. The expected value is the 10 to 15 gram weight class is actually an extrapolated one Table h, ”falculzztion of regression. line rolrtir (, ornsmtption to bofy height. Data frcn 15 wild voles, 26 det'rufu'tions. all determinations an hours in duration. Food and nster available. Sum.long = 21.3526 n = 15 Sun loggY'= 30.7899 C”um log‘x = 30.7060 dum.log Y': 65.2963 - .52 log y-= 1.51u0 + 0.5126 x r": 20.6 x ' b = 0.5126 5 t fiducirl limits of (b) 0.3g57 and 0.6519 standard deviation - standard deviation + II Na. Etc) gig; Table 5 o Conparison of the slepes of the regression lines from tables 1 and no 2 2 Reg. 1.18811 f Sun x Sum x.y Sum y Coeff. f Sum dev. Square Fasting L'9 .5980 .5199 .h067 .6h20 no .2015 .oon2 Non-fasting in .5105 .1611 .0990 .5122 15 .0157 -0009 Within Fog. Coeff. 61 .2170 .0056 1 .0025 .0025 Common 65 .8055 .9810 .5057 .59P9 62 .2195 F'= .0025/.0056 = less than 1.0 The difference between the slopes is not significant. 37. since the seven voles in this class were all juvenile laboratory-bred animals and the data were not used in the calculation of the regression equation. Daily activitLbased on variation in oxygen consumption The introduction illustrates how diel variation in the oxygen con- sumption of small mammals reflects changes in their activity. Thus, it is possible to use the data on hourly oxygen consumption which were obtained during this study to quantitatively investigate the activity patterns of the meadow vole. Data from 29 uninterrupted 24-hour runs on 18 animals were available and are used to: (1) determine whether the voles were nocturnal or diurnal, (2) determine the duration of the short 'feeding' cycle is present, and (3) determine the presence or absence of any well defined 24-hour cycle of activity. Following the suggestion of Welsh (1938) the term 24-hour cycle, rather than the more ambi- guous term 'daily rhythm', is used to designate the correlation of activity with times of day. To measure the night / day ratio of activity the oxygen consumption of each vole was calculated for the period of darkness and for the period of daylight. Since all measurements were conducted with the apparatus placed in a window ledge, the period of darkness and of daylight could be found by reference to a table showing the hours of sunrise and sunset. Naturally the intensity of light was somewhat reduced from that which would prevail in the field. Since the study was conducted over a period of approximately three months the relative lengths of the periods of night and day changed as the season progressed. Thus, all data are expressed as a percentage rate of oxygen consumption. For example, a value of 53 for oxygen consumption during the night means that if the \4‘. r— 8 \ ave. Ina-x. \ mean N) j rrfrcveion line 0‘. (“70. min. 0 o o g Q U‘ c neumntion er cc./gu./h0ur 0’) - . , o °o., .'.'. 3 h . .'o.. 'O .- ° . o .0 . . ... -, .' o- - ° 0 2 L 1 1m 1 J__ l l___ t 01365- 10-155 15.20 20—25 25-30 30-35 3 0 149-59 “n are. -- 7 9 3 5 2 8 3 figure 12. - Graph of average maximum, average .zinimm, and mean hourly Values for 0? consunmtion arm a zit-hour period, nlO‘Etvd by weight class. Table 6. Night versus day rates of 0 consumption in the meadow vole. All figures represent % of iel total per 12-hour period. Wild Captive Lahorztory bred ni *ht :31 n1 aht 1&3: ni {ht 931 n9.6 5o.h 51.7 u8.5 55.6 nu.u 51.6 M8.u 52.1 n7.9 5u.7 M5.5 lfales 55.8 n6.2 51.5 M8.7 50.u M9.6 50.6 L9.u 51.7 M8.5 55.9 M6.1 55.8 nu.2 51.1 ia.9 51.0 U9.0 55.6 M6.u Fearles 55.8 M6.2 55.6 n6.h T*sing the fiilcoxson non-parenetric test: T: 1.5 N: 18 p = less than 0.01 40. night/day periods were equal in length, the animal would have con- sumed 53 percent of the total for the 24-hour period during the night. Table 6 shows the values for 18 voles on which uninterrupted 24-hour determinations were made. Wild-trapped males, wild-trapped females, and laboratory bred juvenile voles are shown separately. In cases where more than one measurement was made on a single individual the mean is used to prevent bias of the statistical test. All voles ex- cept one showed a higher rate of oxygen consumption during darkness. Most of thegvoles were measured within two days after capture but a few were used after being in captivity for periods varying from two to four days in length. The significance of the difference between night and daylight rates of oxygen consumption was tested using the non- parametric Wilcoxson test (Siegel, 1956). The higher oxygen consump- tion during the night revealed by Table 6 was found to be highly signifi- cant. The magnitude of the difference, however, was very slight. The night / daylight ratio of oxygen consumption for all animals was 1. 09. The ratio for the laboratory voles alone was 1. 25. The same ratio calculated from data on wild-trapped voles alone was 1. 06. This latter vale is the same as that reported by Pearson (1947). I The short cycle, or 'feeding' cycle as it is called by some authors, was investigated using the data on hourly oxygen consumption. Since the hourly oxygen consumption of each vole was determined by calcu- lation from the distance between lines on the smoked drum of the kymo- graph, the following method was used to determine the length of a short cycle. The 24-hour record of each animal was examined and the start of a cycle was taken as the point where there was a one millimeter in- crease in the distance between the lines on the smoked paper, con- 34 26 22 H 03 % frequency 9" p 10 +‘F—" .r.__. L__J, L Z'B'h's'e'vfa' Length of cycle in hours Figure 13. - Frequency and duration of the short activity cycle. 42. versely, a one-millimeter decrease then signified the end of a cycle. By this means each 24-hour record was divided into several cycles having durations of from two to eight hours. These are plotted in Figure 13 as frequency in percent against duration in hours. In cases where repeat runs were available for a single animal the mean number of cycles of each duration for each such animal was used to avoid bias- ing the results. The three-hour cycle was the most frequent, in occur- ance followed by the two-hour cycle. These results agreed well with the short cycle lengths reported for the genus Microtus by Morrison (19483), Pearson (1947), Davis (1933), and Mann (1954). The investi- gations of Morrison and Pearson were based on variation in oxygen consumption. Mann and Davis used records of locomotor activity. The mean value for cycle duration, that is, the number of cycles di- vided by the total number of hours during which measurements were made, is 3. 6 hours. As noted earlier, the literature shows much disagreement regard- ing the presence, or absence, of a well defined 24-hour cycle in m- Lu_s_, A careful study of the known habits, physiology, and habitat of the species Microtus pennsylvanicus, seems to indicate that the probable origin of any 24-hour cycle is to be found in the environmental condi- tions prevailing in the field and that no single factor can be determined to be primarily responsible. To investigate the activity patterns under laboratory conditions the hourly oxygen consumption data from several determinations on both wild trapped animals and those which had been kept in captivity for definite periods were utilized. The animels used in these determinations were subjected to relatively constant environ- mental conditions except for the factor of day length, which approximated 43. field conditions. The oxygen consumption of eight voles was measured for 24 hours the day following capture. Six voles were subjected to the same treatment after being kept in captivity for periods of time varying from three to eight days in length. Since Morrison (1948a) found the presence of a 24-hour cycle was often more easily shown by using two-hour periods (thus averaging out the short cycle), the author decided to adopt the same method. Thus, all following conclusions regarding the presence, or absence, of a 24-hour cycle are based on oxygen consumption per vole per two-hour period. Tables 7 and 8 present the results of a two way analysisof variance of this data. A description of the method of analysis is published in Snedecor (1956). The data of Table 7 were from voles measured the day following cap- ture. The effect of time of day on oxygen consumption (reflecting changing levels of activity) was determined to be highly significant (p = less than 0. 01). There was also a significant difference between voles with regard to their total oxygen consumption for the 24-hour period. This latter fact was undoubtedly due to weight variation be- tween voles. The data of Table 8 were obtained from voles measured after ' being held in captivity for periods varying in length from three to eight days. A significant effect of times of day on oxygen consumption could not be shown. A highly significant difference was still noted between animals with regard to total oxygen consumption for the 24-hour period. These results indicated that the 24-hour cycle exhibited by voles in the field was apparently maintained for only a short time after they were brought into the laboratory, even though light conditions were not radically changed. It is, of course, possible that the intensity of light was a more important factor than its mere presence of absence. This m .momob m .cewprodmeco 0 er cthpefinee dofim so ocean .oadunco acuHm .nm aofiop ioneoc we makeupec nufiefiwo¢ I . 3H 0.2.1. «rm bow M0 9.39. none 04. m m a N mags”: OH w 0 .4 N flood ‘ “.m 1%. . c4. . né V‘ 0 CC! O O\ .5 noriumnsuoo 60 {atp {8101 ;o .moHob m .aofiuoadmcoo No a“ :ofipewewp Hofie co venom .oasaodo amour whee m on m moan» aoedea we acuappep hawhuaud I .ma oududm hum Mo mafia coon as m e a N oaaeefie OH m m a N a8: +14 . wt . . q . . . . 747 . .mm. _filuund.u. -nmw 0 Idip'tvaoa J C‘) - C uotiumnsuco I Q. 0 r" \l ’7 U 'fiat) 153 7 0 Analysis of variance in the 2n-hour 02 consumption record. Eight voles measured the day after capture. Source of variation Sum of squares df. Kean square vo1es 257,501 7 33.929 3': hu.6 p = 410.01 Time of day 57,202 11 5,582 3': &.u Error 53,617 77 761 Total 550,520 95 38,072 Table 8. Analysis of variance in the 2M-hour O2 consumption record. Imelve voles measured hfter from three to eight days in captivity. m Source of variation Sum of Squares df. Vean square vo1es 156,hu6 11 27,289 F = 52.1 p =(OoOl Time of day 12,069 5 1,09 r = 1.29 p =>Oo® error 96,795 55 850 Total 195,255 71 29,256 47. possibility however, is believed to be rather remote. The data discussed above are shown in Figures 14 and 15. The oxygen consumption (expressed as the percent of the diel total) is graph ‘ against the time of day. Figure 14 shows a tendency for a bimodal pattern with peaks between 6 p. m. and 8 p. m. and again be- tween 4 a. m. and 6 a. m. The large drop in oxygen consumption be- tween 8 a.m. and 10 a.m. was characteristic of all voles in this group. The data graph in Figure 15 are from voles which were kept in captivity before measurements were made. The decrease in oxy- gen consumption between 8 a.m. and 10 a.m. is not present and the bimodal pattern is not nearly so well defined, a tendency being noted for a third peak to occur between 10 p.m. and 2 a.m. However, even though the two graphs show dissimilarities there still would be no way of indicating whether either showed a significant correlation of times of day with oxygen consumption without subjecting the data to statisti- cal analysis. To further investigate the effect of captivity on the activity patterns of the meadow vole, two voles were selected and measured after being kept in captivity for stated intervals of time. The oxygen consumption of each animal was determined during the first, second, and fourth week of captivity. The results were subjected to a three factor analy- sis, mixed model, two factors fixed and one random (Snedecor, 1956). The various factors and their tests of significance are tabulated in table 9. By numbering the sources of variation the method of calcu- lating the (F) values is shown. None of the effects were significant with the exception of the interaction of voles with weeks of captivity. This is interpreted to mean that length of captivity had an effect on the 48. total oxygen consumption, but did not act in the same way on each vole. An important conclusion however, is that in these two animals the times of day did not significantly effect oxygen consumption, and furthermore, this lack of effect was not caused by a change in activity pattern due to further captivity nor to any demonstrable difference in the activity patterns of the two voles. Activity as measured in exercise wheels All measurements of activity patterns in the meadow vole which have been discussed thus far were based on diel variation in oxygen con- sumption. This involves the assumption that these changes are primarily due to variation in levels of activity. In order to provide data on loco- motor activity, a group of voles was placed in cages provided with exer- cise wheels which were attached to a recording device. Data were obtained for the number of revolutions of the wheel during the periods of darkness and light on seven days in succession, using seven wild trapped voles and four voles which had been kept in captivity for over four months previous to the experiment. During the fifth day of the experirnent, readings were also taken at two-hour intervals to demonstrate the presence, or absence, of a 24-hour cycle of activity. The data on oxygen consumption indicated that correlation of times of day with levels of activity ceased during the first week of captivity. In view of this discovery, it is unfortunate that the two-hour readings of voles in the exercise wheels could not have been taken immediately after capture. The completely erratic behavior of some animals pre- cluded the recording of any meaningful data until after a four-day equili- bration period since they refused to use the wheel during the first and se- cond day in the cages. Even after the four day period there was significant ~s~.:a Ha omw.wm Hoooe ooaqa mm Hmfi.:m onooa-moao>-oeae any Ho.ouvu o mm.~ u hav\xm. u.m oma.m m wmm.oa oeooa-ooaoo “my o.H.v u A~V\xmv u a mam mm mam.am oxoozuoaae “my mo.o.n u o mH.H u h~v\xav u.m mam.a Ha mao.:H ooaoouoEee “av e.H.v u Amv\xmv u.m Nmm m mam.a oxooa any 04 v u :33 u a mm: H 8: moao: av no.0». u a 94 u 3(73 u m 2mm; 2 $0.0m owe yo and. ,3 oneowm nook. .w.c mopoodm 90 Sam nowpmfinwp Ho moon50m .9888 moaaodoueooa. anwm.o=...oomaom .oomw'mewowmmua- $889.. $6» 92.. .caoooa aofipoeomooo No Hdogufim onp a“ moqwflpwp no mammaono nonowmnooaoe . m 035. mm mmm.Hmm.m Hopoe H :o».mm mooHoooHHoom Hms.:H pm mnm.mmm gonna xmv Ho.ouvu o hem.wv ~6.H u ”ml\nev n a amm.am m mwm NaH .o\eoe-ooooom-moHop lav Ho.0uvu a hem.mv mo.e u xwv\flmv n a mew.mw m mmm.mmm eoHooooomumoHop ”my mo.0u.n o hem.mv mm.m u xwv\xmv u a amm.Hm m mmm.HHm ooHoouoeiHo\aoc ”my mo.o.nu o Am.Hv mm.m u ”NV\A:V u a mem.wm H mew.mo ooooomuoemHo\soo as» Ho.ouvu o new.wv ma.OH u xmv\xmv u a www.mmH m mmo.wmm ooHoe any o.Hv u 6HmmooSm nopom now mHooos mpHpHpow aH uoaowwee AcHHsv moHop nopom .HOHpoHaob pann .m> how we 0H0» Scones on» Ho mnhoppop mprHpow Ho ooowwnap no mHthmom hosnoonnE .CH oHQoB 51. difference between animals, with regard to their use of the wheels, far greater than any difference in activity between voles as measured by variation in oxygen consumption. Bailey (1924) also noted differences between voles with regard to use of exercise wheels. Since the length of the period of darkness in the animal room was 11 hours and that of light 13 hours, all data are expressed as the mean number of revolutions of the wheel per hour, making possible a com- parison of the two periods as to rate of activity. Owing to a certain amount of disturbance during the day by personnel working in the ex- perimental room, a large screen was placed around the cages during the middle of the third day and left there during the remainder of the period of measurement. The data from the darkness versus light readings were analyzed, using only data from the first two days and from the last two days of the period of measurement. By this means the effect of adding the screen could be evaluated. The computations of this analysis are shown in Table 10. The information provided there— in includes only data from the seven wild voles in order to provide re- sults comparable to those obtained from the measurements of oxygen consumption. A three-way analysis of variance was used (Snedecor, 1956). As before, the method of calculating the (F) values is illustrated in each case by numbering the sources of variation. Table 10 shows that the effect of light conditions on activity did not reach the five percent level of significance. Furthermore, screening the voles had no significant effect on the 24-hour cycle of activity. There was a highly significant difference between voles with regard to total daily activity (pi = less than 0. 01). The interaction between light conditions and screening the voles was not significant, thus the inability to show a significant 24-hour activity cycle was not due to changes in activity pattern caused by plac- ing the screen around the cages. The interaction between light conditions 52. and voles was significant (p = less than 0.05, greater than 0. 01). In the absence of any significant main effect of light conditions, this in- dicates that some voles were more active at night while others were more active during the day. The interaction between voles and the effect of placing the screen was highly significant (p - less than 0. 01). This, coupled with the fact that there was no significant main effect of screening, indicates that placement of the screen around the cages re- sulted in changes in the total daily activity of the voles but did not act in the same manner on each vole. That is, the presence of the screen apparently lowered total daily activity in some voles and increased it in others. The three-way interaction was not significant. Individual variation in rate of activity between voles was strikingly demonstrated by some calculations of the distance traveled during a 24-hour period. The mean distance traveled per day by each vole, both wild and captive, was calculated using the circumference of the activity wheel times revolutions per 24-hour period. The results of these cal- culations are tabulated in Table 11. These are believed to be reason- ably accurate since the voles did not exhibit the tendency exhibited by albino white mice, of hanging on and riding the wheel. The voles were occasionally observed to leave the wheel while it was still revolving, but the error in all instances was very small. Some idea of the great variations in activity between voles can be obtained from a considera- tion of the maximum and minimum values for distances traveled in a 24-hour period. These values are shown at the bottom of Table 11. The data for activity during two-hour periods, which was obtained during the fifth day of the expreiment, were analyzed in the same manner as the comparable data obtained from oxygen consumption values. A Table 11. fistances tereled by the meadow vole in activity wheels. All figures are exrres:ed as kilometers per dty(2M-hour period) and represent the mean value for each vole during the seven day measurement period. s==== ================================== Wild Captive 5.2 0.6 1.0 Ifales 1.9 7.7 14.5 9.6 Fenales 8.9 0.1 0.3 1.0 Vaximum during one 2u-hour period = 15.3 kilometers 'Ninimum during one 2u-hour period was zero. 54. two-way analysis of variance was made on the data obtained from the seven wild voles and on the data from the four animals which had been kept in captivity. The data were expressed as the mean number of re- volutions per two-hour period per vole. The computations are shown in Tables 12 and 13. In each case the results were the same, indicat- ing no significant effect of times of day on the level of activity. In each case there was a significant difference (p = less than 0. 01) between voles with regard to total activity per 24-hour period. Since no statistically significant 24-hour cycle could be determined, the graphs of the data are not reproduced herein. However, the com- posite graph for the wild voles and that for the captive animals showed peaks between 10 p. m. and 12 p. m. and again between 6 a.m. and 8 a. m. although graphs of the individual voles showed great differences from one another. This again serves to emphasize the fact that the method of interpretation of cyclic phenomena, based on composite graphs, may be open to serious question unless accompanied by a statistical analysis demonstrating significance. Table 12. Analysis of variance in the EM-hour activity wheel record. Seven voles measured after one week in captivity. — Source of variation Sum of SqUPTPS d.f. Mean square voles 29,297,622 6 9,879,609 r = 18.5 (6,66) Time of day 9,765,182 11 955,198 F = 1.69 (11,66) Error 17,919,526 .66 265,857 Total 51,927,550 85 Table 130 Analysis of variance in the 29-hour activity wheel record. Four voles measured after several months in captivity. E W Source of verirtion Sum of squ~res d.f. "ean Square voles 10,059,662 5 5,555,221 pa: 15.5 (5,55) Time of day 5,698,926 11 518,059 3': 2.06 (11,55) P = >0-05 Prror 3.309.973 33 Total 29,068,061 97 56 . DISCUSSION Thermoneutrality and grouping In the absence of any published data regarding the temperature range of thermoneutrality in the meadow vole, an attempt was made to evaluate this range for use in further metabolic determinations. This attempt was in a large part successful, although further investigation would be of value in narrowing the range from that of 25 to 29 degrees Centigrade, which was used in this study. The results of the oxygen consumption of grouped versus individual voles (Figure 9) demonstrated that the combining of animals in a group tended to give a more accurate estimate of the fasting oxygen consump- tion. Pearson (194 7) reported that grouping tended to lower the oxy- gen consumption of the house mouse, Mus musgulus. Most of his work was done at ambient temperature below that of thermoneutrality. Some fragmentary results, however, showed that even at the temperature of thermoneutrality, Pearson's mice exhibited a lower oxygen consump— tion when measured in groups. Sealander (1952) found that huddling significantly raised the survival time of various species of the genus Peromyscus when the mice were exposed to temperatures below the lethal limit. It would appear from these results that one of the main effects of huddling is to reduce heat loss and, by this means, to re- duce oxygen comsuption when the animals are measured at an ambient temperature below that of thermoneutrality. The lowering of oxygen consumption by huddling, even when animals are measured at the temperature of thermoneutrality, apparently is due to reduction in activity since a reduction in heat loss at this temperature would actually result in a rise in oxygen consumption. 57. Grouping of animals is precluded in any study in which the oxygen consumption is to be correlated with weight unless groups can be se- lected within which there is very little weight variation. Unfortunately it was impossible to so select the voles during this study, since the results of the fasting metabolic determinations were based on measure- ments of random trapped individuals. Still, the grouping experiment was of value in showing that much of the scattering of data to be ob- served in Figure 10 is no doubt due to uncontrolled activity, consist- ing mostly of small movements of the head and limbs. Any gross move- ments such as running and / or jumping were precluded by the small wire cages in which the voles were confined during the measurement period. Fasting energy consumption The values for fasting oxygen consumption in both wild and labora- tory bred voles are shown in Figure 10. The calculations of Tables 1 and 2 indicate that the fasting oxygen consumption of a vole of a given weight can be estimated by the formula: (1) Y = 9.2 x-64 with standard deviations of plus 16 percent, and minus 13. 8 percent. The five percent fiducial limits of (b) for the population lie between the values 0. 46 and 0. 83. This range of values indicates that (b) is not significantly different from the values reported by Brody (1945) and Kleiber (1947) for adult animals of different species. Since the neuro- endocrine system involved in the control of metabolism is not fully de— veloped until relatively late in life, and both body composition and body conformation are changing rapidly throughout the growth period, varia- tions would be expected in the metabolic curves of growing animals 58. (Brody, 1945). No breaks were noticed in the 'metabolic curve' in this study however, the reason probably being that no animals younger than weaning age were used. The results of this phase of the study indicate that, when growing animals of a single species are to be considered, the equation relating metabolism to body size will be more accurate if the calculations are based on data from that species, rather than adapting a curve which is designed to fit all species of mammals. Brody (1945) expresses the view that the value of (b) should be accurate to only one decimal place, and that therefore, 0. 73 should be replaced by 0. 70. The value for (b) which was determined for fasting animals in this study, 0. 64, is thus not so different from Brody's figure as was at first apparent. The results of determinations of the fasting oxygen consumption in eight juvenile voles bred in the laboratory were tabulated. Since the slopes of the two regressions were not significantly different, the effect of being raised in captivity apparently changes only the level of oxygen consumption and does not alter the basic relationship between weight and metabolic rate. It may even be that this higher level is not an effect due to captivity but that it is only an indication of a normal break in the metabolic curve of growing animals. The limited data available in the lighter weight classes do not permit any objective con- clusions to be drawn, however. The factor 4. 825 cal/cc. oxygen consumed is generally used in es- timating the energy value of the oxygen consumption of hemeotherms. This is calculated on the basis of an assumed R. Q. value of 0. 8. (Brody, 1945). This is an approximation but, as has been shown herein, the caloric equivalents of oxygen, even at the extremes of R.Q. values, are very close to one another. Thus, the error introduced by using the 59. average value in energy estimations is very small and is even less than the error which would be introduced by trying to determine the R.Q. of the experimental animal. Using this caloric equivalent the energy con- sumption of the fasting meadow vole of a given weight can be estimated according to the relationship: (2) Gram calories per day per vole = 9. 2 (weight in grams)‘ 64 x 4. 8 x 24 The above relationship actually defines the basal energy consumption measured at the temperature of thermoneutrality, plus the increment due to uncontrolled activity during the measurement period. Since this error due to activity is probably very small, the relationship given above should suffice for fairly accurate estimates of the basal energy cost of the meadow vole. Certainly, it will provide a degree of accu- racy which has hitherto been lacking in all such energy estimates. It should be recognized that lower environmental temperature in the field will increase this energy cost. The correction factor which would be needed for different temperatures was not assessed during this study. Much has been done by Hart (1950, 1952a, 1952b, 1953, and 1956) on the relationship between energy metabolism and ambient temperature in the white footed mouse, Peromyscus leucopus, and the albino labora- tory mouse. Morrison and Ryser (1951) reported measurements of oxy- gen consumption on a meadow vole weighing 51 grams at various ambient temperatures ranging from 0 to 30 degrees Centigrade. Their results showed that every degree decrease in temperature resulted in an in- crease of O. 139 cc/gm. hr. in the rate of oxygen consumption. The relation was seen to be linear within the limits of variability, showing that the insulation of the vole was not changing. 60. Maintenance energlconsumption A second phase of this study was undertaken to provide an estimate of energy consumption of the animal which would be referrable to 'natural conditions'. Figure 11 presented a double logarithm plot of the non-fasting oxygen consumption as a function of body weight. The regression line was represented by equation: (3) Y = 20.6 x52 with standard deviations of plus 8 percent and minus 7. 4 percent. The calculations of Table 5 indicated no significant difference be- tween the slopes of the regression line equation (1) for fasting voles and equation (3) for non-fasting voles. In Table 4 the five percent confidence limits of the exponent (b) of equation (2) were calculated. The values lie between 0. 39 and 0. 64. The upper confidence limit approaches the rounded off value for (b), O. 70, proposed by Brody (1945). Undoubtedly there is some underlying relationship between oxygen con- sumption and body weight (metabolic body 'size' of Brody, 1945). In the case of the non-fasting animal this relationship is partially obscured by the activity exhibited by the animal. In any case, it would seem that the best estimate of a relationship between body size and oxygen consump- tion is determined on the basis of data from the species to which the re- lationship is to be applied. The data for Figure 11 enables the energy estimate to be obtained from the equation: (4) gram calories/day/vole = 20. 6(body wt. in gma.)' 52x 4. 825 x 24 = 2386 (body weight in grams)- 5 In contrast to the previous estimate of energy consumption derived from fasting metabolic data, the equation above estimates the total 61. maintenance energy cost to the animal. The voles generally exhibited a slight increase in weight during a 24-hour period of measurement but this productive energy use would not be indicated by the oxygen con- sumption. Equation (4) is based on the assumption that the level of activity in the metabolism chamber is the same as the vole would exhibit in the field. This is probably an underestimation since voles in the exercise wheels showed themselves capable of traveling distances as great as 15 kilometers per day. The author wishes to stress that es- timation of energy consumption also depends on the degree of accuracy of the caloric equivalent of oxygen (4. 825 Cal/ liter). In the absence of information on the R.Q. of the animal the degree of accuracy of this factor can never be ascertained. Thus, any energy values derived from equation (4) are to be regarded only as reasonable estimates, but even so, estimetes which have a higher degree of accuracy than any which have hitherto been available to workers in this area. The daily amplitude of the oxygen consumption was calculated by weight class and plotted in Figure 12. The amplitude becomes pro- gressively greater as the weight of the animal decreases. This may be analyzed by noting that the maximum rate of oxygen consumption tends to follow the same logarithmic curve as does the average hourly rate. The minimum rate shows a more linear relationship to weight. The anomalous result in the 30 to 35 gram weight class is probably due to the small number of animals that were averaged. The energy cost of a stated increment of work, or activity, should be proportional to body weight to the power 1. 0 rather than to fractional power of body weight (Brody, 1945). Since this is apparently not true in this case, where the maximum oxygen consumption rises with a fractional power of weight, it may be assumed that the younger, lighter animals can attain a higher 62. level of activity than the adults and that they also are capable of periods of complete rest. This phenomenon can be seen in many mannals, including man. Activity patterns based on oxygen consumption The data from Table 6 indicate that wild trapped voles, measured in the laboratory, have a significantly higher rate of oxygen consumption during the night, although the magnitude of this difference is very slight. The small absolute difference in the amounts of oxygen consumed dur- ing the night, over that consumed during the day for the wild trapped voles, may actually represent a greater. night / day difference in amount of activity than is apparent at first glance. This is because a continuous basal level of oxygen consumption is maintained even in the complete absence of activity, thus making a rather large absolute increase in activity apparent only as a small percentage increase in total oxygen consumption. Measurements of the night/ day ratio of activity were also made by using activity wheels. Interpretations based on the statistical analysis of these data have already been made. Failure to show statistical correlation between light conditions (night versus day) and activity was no doubt due to the extreme variance encountered between individuals in their use of the wheels. The screen placed around the apparatus during the third day of the experiment had no effect on the night / day ration of activity but did affect the total activity of the voles. The ex- periment in the wheels was of value in illustrating the extreme range of variability in the distance an individual vole will travel in a single 24 hour period. Table 11 shows that over the seven day period of the experiment the average distance traveled per 24-hour period varied 63. from O. 1 kilometers per day to 8. 9 kilometers per day. There is no apparent difference between captive and wild voles with regard to the distance traveled in the wheels. The distance traveled during any one 24——hour period varied from zero to 15. 3 kilometers. A short cycle of activity was found in the meadow vole. Figure 13 shows that the most frequent length of the cycle was three hours, the next most frequent period was two hours long. This is an agreement with the short cycle of two to four hours in length which has been re- ported by several authors. In general this short cycle is attributed to feeding activity. The author sees no cause for disputing this hypo- thesis and the short cycle is therefore considered as a feeding pheno- menon, although no attempt was made during this study to prove such an origin. Tables 7 and 8 show that voles whose oxygen consumption was mea- sured the day after capture possessed a well defined 24-hour cycle, while voles retained in captivity for periods of from three to eight days lacked such a cycle. All the voles which were measured in the activity wheels had been in captivity longer than eight days and none of them showed a statistically significant 24-hour cycle of activity. Table 9 indicated that after the initial loss of the 24-hour cycle, diel varia- tion in the oxygen consumption did not change significantly over the next four weeks. From the above observations some conclusions can be drawn con- cerning the activity patterns of Microtus pennsylvanicus. Continued captivity may have little effect on the night versus daylight activity pattern since even animals which had been kept in captivity for several weeks before measurement exhibited a significantly higher rate of oxy- gen consumption during the night. However, captivity does have a 64. detrimental effect on the 24-hour cycle of activity, that is, the cycle wherein the activity is correlated with definite times during the 24— hour period. In this study the newly-trapped animals had a well de- fined 24-hour cycle. Being least active in the forenoon, they gradually became more active in the afternoon, with two partially defined acti- vity peaks, one between 6 p. m. and 8 p. m. and another between 4 a.m. and 6 a.m. The short cycles which were assumed to be due to feeding activity were superimposed on the 24-hour cycle. In the labo- ratory, where all conditions except light were controlled, the 24-hour cycle soon ceased. The short feeding cycle was retained, as was the significantly higher rate of oxygen consumption during the night period. The 24-hour cycle must be determined by factors other than light alone, since it was not maintained in the laboratory, where light con- ditions approximated those found under field conditions. Calhoun (1945-1946) states that most animals which are subjected to fluc- tuations of light in their normal habitat possess a 24-hour cycle and that this cycle in inhibited by continuous light. However, in an earlier paper Calhoun (1945) mentions temperature as having an important effect on cycles. Since the lethal temperature for the meadow vole was shown by this study to be within the range of environmental tem- peratures occuring in southern Michigan during the summer, it would follow that the voles are regularly exposed in the summer and fall to temperatures which exceed the lethal limit during the day. Certainly, this would modify to a great extent, any activity cycle induced by light alone. The effects of the many limiting factors in the environment of the meadow vole will remain only speculative until the time arises when more factual material is available. 65. SUMMARY 1. Preliminary investigations of the temperature range of thermoneu- trality in Microtus pennsylvanicus indicated that this range lies some- where between 25 and 29 degrees Centigrade. 2. Measurements of the oxygen consumption of animals grouped in the metabolism chamber as compared with the same animals measured alone, showed that the effect of grouping, even within the temperature range of thermoneutrality, tends to lower the oxygen consumption, pro- bably due to lessened activity. 3. Measurement of the oxygen consumption of animals of different weights indicated that the rate of oxygen consumption has a definite relation to weight. In the case of fasting, relatively inactive animals this relationship was expressed by the equation: (1) Oxygen consumption as cc/vole/hr. = 9. 2(wt. in gms. )' 64 For non-fasting voles exhibiting 'normal' activity, the oxygen consump- tion was related to body weight by the following equation: (2) Oxygen consumption as cc/vole/hr. = 20. 6(wt. in gms. )' 52 No difference was determined to exist between the slopes (b values) of equations (1) and (2). 4. Equation (1) provided a means of estimating the minimum energy cost to the animal under conditions of thermoneutrality, while equa- tion (2) permitted the estimation of the total energy utilization, exclu- sive of productive processes, of a given weight vole under the same c onditions . 66. 5. By means of maximum and minimum rates of oxygen consumption it was determined that the younger and lighter animals exhibited more frequent and more intense bursts of activity as well as indicating a tendency for a more relaxed state of rest. 6. The meadow voles used in this study exhibited a short 'feeding' cycle of oxygen consumption whose mean duration was 3. 6 hours. The mode was three hours. 7. The voles used in the study exhibited a slightly higher rate of oxy- gen consumption during the night. 8. Newly captured voles possessed a well defined 24-hour cycle which showed peaks after dark and before dawn. Voles which had been kept in captivity for periods of from three to eight days did not possess this cycle. 9. Measurement of activity patterns using recording exercise wheels demonstrated that there is a significant difference in the use of the wheels by different voles. No statistically significant 24-hour cycle was present, nor was there any correlation of activity with light con- ditions. The distance traveled in the wheels varied from 0 to 15. 3 Kilometers per 24-hour period. 10. The measurement of diel variation in oxygen consumption gives a better quantitative estimate of activity patterns than measurement of locomotor activity in exercise wheels. 11. The 24-hour activity cycle of the meadow vole is conditioned by factors other than, or in addition to, light. 67. 12. Further research on the limiting effect of environmental factors such as light, humidity, temperature, etc. , on Microtus Rennsylvanicus is greatly to be desired. 68. LITERATURE CITED Bailey, V. 1924. Breeding, feeding, and other life habits of meadow mice (Microtus). Jour. Agric. Res., 27:523-536. Benedict, F. G. 1915. Basal metabolism. Jour. Biol. Chem., 20:263-299. Bertalanffy, L. and R. R. Estwick. 1953. Tissue respiration of musculature in relation to body size. Amer. Jour. Physiol. , 173(1):58-60. Brody, S. 1945. Bioenergetics and growth. Reinhold: New York. 1023 pp. Calhoun, J. B. 1945. Diel activity rhythms of the rodents Microtus ochro aster and Sigmodon hispidus hispidus. Ecology, 26(3):§51-273. Calhoun, J. B. 1945-46. Twenty-four hour periodicities in the animal kingdom. II-the vertebrates. Jour. Tenn. Acad. Sci., 20(2):228-232, 20(3):291-308, 20(4):373;‘378, 21(2):208-216, 21(3):281-282. Chambers, W. H. and W. H. Summerson. 1950. Energy metabolism. Ann. Rev. Physiol., 12:289-310. Davis, J. E. and H. B. van Dyke. 1932. The measurement of oxygen consumption of small animals. Jour. Biol. Chem. , 95:73-78. Davis, D. H. s. 1933. Rhythmic activity in the short-tailed vole (Microtus). Jour. Animal Ecol., 2:232-238. Dewar, A. D. and W. H. Newton. 1948. The determination of total metabolism in the mouse. Brit. Jour. Nutrition, 2(2):123-141. Dice, L. R. 1922. Some factors affecting the distribution of the prairie vole, forest deer mouse, and prairie deer mouse. Ecology, 3:29-47. Elton, C., E. B. Ford, and J. R. Baker. 1931. Health and para- sites of a wild mouse population. Proc. Zool. Soc. London, pp. 657-721. Emlen, J. P., R. L. Hine, William A. Fuller, and P. Alfonso. 1957. Dropping boards for population studies of small mammals. Jour. Wildl. Mgt., 21(3):300-314. 69. Fuhrman, G. J., E. D. McLin, and M. L. Turner. 1946. The effect of time of day on the metabolic rate of albino mice. A mano- metric method. Amer. Jour. Physiol., 147(2):284-288. Galvao, P. E. 1947. Heat production in relation to body weight and body surface. Inapplicability of the surface law on dogs of the tropical zone. Amer. Jour. Physiol., 148(2):478-489. Grad, B. 1952. A simple method for the measurement of oxygen consumption and heart rate of rats. Endocrinology, 50(2):94-99. Hamilton, W. J. 193 73. Growth and life Span of the field mouse. Amer. Nat., 71:500-507. Hamilton, W. J. 1937b. Activity and home range of the field mouse, Microtus pennsylvanicus (0rd). Ecology, 18:255-263. Hart, J. S. 1950. Interrelations of daily metabolic cycle, activity, and environmental temperatures of mice. Canad. J our. Res. , D, 28:293-307. Hart, J. S. 1952a. Use of daily metabolic periodicities as a measure of the energy expended by voluntary activity of mice. Canad. Jour. 2001., 30(1):83-90. Hart, J. S. 1952b. Effects of temperature and work on metabolism, body temperature and insulation: Results with mice. Canad. Jour. Zool. , 30 (1):90-98. Hart, J. S. 1953. Energy metabolism of the white—footed mouse, Peromyscus leucopus noveboracensis, after acclimation at vafious environmental temperafilres. Canad. Jour. Zool. , 31:99-105. Hart, J. S. 1956. Seasonal changes in insulation of the fur. Canad. Jour. 2001., 34:53-57. Hatfield, D. M. 1935. A natural history study of Microtus californicus. Jour. Mamm., 16:261-271. Hatfield, D. M. 1939. Rate of metabolism in Microtus and Peromyscus. Murrelet, 20:54-56. Hatfield, D. M. 1940. Activity and food consumption in Microtus and Peromyscus. Jour. Mamm., 21:29-36. Hatt, R. 1930. The biology of the voles of New York. Roosevelt Wildl. Bull., 5(4):513-623. Hodgeman, C. D. (ed.). 1952-53. Handbook of chemistry and physics. Chemical Rubber Pub. Co. , Cleveland. 34th Ed. 2950 pp. 70. Holtkamp, D. E., S. Ochs, C. C. Pfeiffer, and A. E. Henring. 1955. Determination of oxygen consumption of groups of rats. Endocrinology, 56(1):93-104. JOhnson, M. S. 1926. The activity and distribution of certain wild mice in relation to biotic communities. Jour. Mamm. 7:245-277. Johnson, M. S. 1939. Effect of continuous light on the periodic, spontaneous activity of white-footed mice (Peromyscus). Jour. Exp. Zool., 82:315-328. Kleiber, M. 1947. Body size and metabolic rate. Physiol. Rev. , 27(4):511-541. Krogh, A. 1916. Respiratory exchange of animals and man. Longmans, Green: London. 173 pp. Lilienthal, J. L. Jr., K. L. Zierler, and B. P. Folk. 1949. A simple volumeter for measuring the oxygen consumption of small animals. Bull. Johns Hopkins Hosp. , 84(3):238-244. Maclagan, N. F. and M. M. Sheahan. 1950. The measurement of oxygen consumption in small animals by a closed circuit method. Jour. Endocrinology, 6(4):456-462. Mann, P. M. 1954. Diel activity rhythms of several species of small mammals. M.Sc. Thesis, U. of West. Ontario, London, Ontario, Canada. Martin, A. W. and F. A. Fuhrman. 1955. The relationship between summated tissue respiration and metabolic rate in the mouse and dog. Physiol. Zool., 28(1):18-34. McCashland, B. W. 1951. A study of metabolic changes in young rats. Growth, l5(l):1-9. Morrison, P. R. 1947. An automatic apparatus for the determi- nation of oxygen consumption. Jour. Biol. Chem., 169(3):66 7-6 79. Morrison, P. R. 1948a. Oxygen consumption in several small wild mammals. Jour. Cell. and Comp. Physiol., 31(1):69-96. Morrison, P. R. 1948b. Oxygen consumption in several mammals under basal conditions. Jour. Cell. and Comp. Physiol., 31(3):281-291. Morrison, P. R. 1951. An automatic manometric respirator. Rev..Sci. Instruments, 22(4):264-267. 71. Morrison, P. R. and F. Ryser. 1951. Temperature and metabolism in some Wisconsin mammals; Mus, Peromyscus, Clethrionomys, Zapus, Eutamias, Microtus, Mustella, Gfaucom s, Scalopus, attus, Citellus, Bla'rina. Federation Proc., ): - . Pearson, O. P. 1947. The rate of metabolism of some small mammals. Ecology, 28:127-145. Pearson, O. P. 1948. Metabolism of small mammals with remarks on the lower limit of mammalism size. Science, 108:44. Richardson, H. B. 1929. The Respiratory Quotient. Physiol. Rev., 9:61-125. Sealander, J. A. 1952. The relationship of meat protection and huddling to survival of Peromyscus at low temperatures. Ecology, 33:63-71. Seton, E. T. 1909. Life histories of northern animals. Charles Scribners Sons: New York. Vol. I. 673 pp. Sherwood, T. C. 1936. The relation of season, sex, and weight to the B. M. R. of the albino rat. Jour. Nutrition, 12(3):223-236. Siegel, S. 1956. Non-parametric statistics for the behavioral sciences. McGraw-Hill Co.: Hew York. 312 pp. Smith, E. 1955. A new apparatus for long term measurement of oxygen consumption in small mammals. Proc. Soc. Exp. B101. and Med., 89(4):499-501. Snedecor, G. W. 1956. Statistical methods. Iowa State College Press: Ames. 534 pp. Topliff, E. D. and E. L. Cullum. 1955. A device for filling a spirometer automatically. Jour. App. Physiol., 8(2):243-244. Watts, D. F. and D. R. H. Gourley. 1953. Simple apparatus for determining basal metabolism of small animals in student laboratory. Proc. Soc. Exp. Biol. and Med., 84(3):585-586. Welsh, “J. H. 1938. Diurnal rhythms. Quart. Rev. Biol., 13:123-139. Wood, F. E. 1910. Mammals of Champaign county. Bull. 111. State Lab. of Nat. Hist., 8:501-613. Zeuthan, E. 1953. Oxygen uptake as related to body size in organisms. Quart. Rev. Biol., 28(1):1-12. ELUH 1113i M”! Diroulltiua gq~r -36; m 1119025.. MW" JHN in 1081 F ”#119996”. Mfr ii